The Mirage of The H Economy: Previous Page

Previous Page
THE MIRAGE OF THE H2 ECONOMY 477
pockets and resulting perturbations. J. Chem. Phys. 117: Following is a hypothetical example of such a reaction:
3753–3761.
8. Bowron, D.T., Soper, A.K., and Finney, J.L. (2001). Temper- MeO + Heat −−−→ Me + 1/2O2
ature dependence of the structure of a 0.06 mole frac-
tion tertiary butanol–water solution. J. Chem. Phys. 114: Me + H2 O −−−→ H2 + MeO + Heat
6203–6219.
9. Hummer, G. et al. (1996). An information theory model where Me is any metal. No such reaction was ever
of hydrophobic interactions. Proc. Natl. Acad. Sci. 93: discovered, and all cycles proposed were quite complex.
8951–8955. One does not need to be an experienced cost estimator to
10. Botti, A. et al. (2004). The microscopic structure of concen- realize that such a cycle would be very expensive and have
trated NaOH solutions. J. Chem. Phys. 120: 10154–10162. a low thermal efficiency.
11. Bowron, D.T. and Finney, J.L. (2003). Structure of a
The H2 was supposed to be used to replace natural
salt–amphiphile–water solution and the mechanism of salt- gas in existing pipelines. Several obvious questions were
ing out. J. Chem. Phys. 118: 8357–8372. not asked:
12. Sacco, A., De Cillis, F.M., and Holz, M.J. (1998). Chem. Soc.
1. A pipeline for methane cannot be used to transport
Faraday Trans. 94: 2089.
H2 . To avoid leakage, one needs totally different
valves and connectors. Furthermore, as the volume
of H2 is triple that of methane, one needs larger
THE MIRAGE OF THE H2 ECONOMY
pipes and different compressors. Of course, the same
REUEL SHINNAR factors would apply to the distribution of the gases
The City College of the CUNY to customers. None of the valves or controls would
New York be tight enough, and all burners would have to
be changed.
2. As nuclear reactors are available, why not use
In a recent paper (1), I analyzed the new ‘‘hydrogen electricity instead of methane for most applications?
economy’’ and explained why, for most uses, it makes Electricity is far more efficient than H2 or methane
no sense to convert fuels, electricity, or any other form of (by a factor of two) for most uses. Let me illustrate.
energy to hydrogen, only to use the hydrogen as a fuel for Today, in the United States and the world, a large
fuel cells. As H2 is not a resource but an inefficient energy amount of electricity is generated from natural gas.
carrier, it would appear that the H2 economy is just a This electricity, on a BTU basis, costs by a factor
mirage. If alternative energy sources are ever developed, of about 2.2 more than the natural gas feed. The
an economy based on electricity would be cheaper (by a fact that we do it obviously means that electricity
factor of three) and much easier to adopt than H2 . Oddly, is more valuable than natural gas. At the same
price, it would have driven out natural gas. H2 as
I had covered the same topic twenty years ago in a paper
a fuel has an even lower value than natural gas. If
that analyzed the hydrogen economy of the 1970s (2).
we now convert electricity to H2 , we again lose at
Then, the investment in H2 research was about 10 billion
least one-third of the energy. Thus, H2 made from
wasted dollars before it was phased out. In this column, I
electricity has a value about four times lower than
will try to explain why the scientific establishment engages
the electricity. This simple calculation illustrates
in such obviously futile behavior and why this mirage the stupidity of the idea to convert nuclear or
attracts so many followers. solar energy to hydrogen. At that time, no one was
In the mid 1970s, I became aware of a strange foolhardy enough to suggest using H2 generated by
phenomenon. Many leading scientific journals were a nuclear reactor as fuel for a fuel cell. Furthermore,
suddenly giving enormous coverage to hydrogen, the an electricity distribution system is already in place
fuel of the future; a new term, hydrogen economy, and can grow with demand.
was born. International meetings attracted thousands of 3. Why not make H2 by electrolysis, a well-known
participants and the United States, Germany, and Japan process? Nobody asked why—in contradiction to our
funded the necessary development with vast resources. cumulative technical experience—a set of chemical
The central idea, put forward by the International reactions should generate a large increase in T
Atomic Energy Research Center in Vienna, used a high- more cheaply or efficiently than electrolysis. The
temperature nuclear reactor as the heat source. In this main item in the cost of electricity is the nuclear
reactor, pressurized helium was used as the coolant. To reactor, and if a thermochemical cycle is less
generate electricity, the hot helium was first expanded efficient, it is inherently more expensive.
through a turbine and then used for steam generation 4. The cost of switching to hydrogen is astronomical.
before being fed back to the reactor. To produce the equivalent of one million BTU of
The ‘‘thermochemical cycle,’’ the name applied to this hydrogen per day by electrolysis would require a
new process, does not include a turbine, and the heat from 25-kilowatt nuclear reactor at an investment cost
cooling the hot helium is used to drive a set of closed-cycle of $150,000, which means that a standard nuclear
chemical reactions; the net result is that water is split into reactor of one gigawatt costing $6 billion will be
hydrogen and oxygen. able to supply 40,000 million BTU/day of H2 the
478 THE MIRAGE OF THE H2 ECONOMY
equivalent of 6700 barrels of oils. Building the numbers. This story illustrates a key aspect of the
capacity for supplying a new pipeline that could hydrogen economy. Many excellent scientists lack the
carry 12% of the current U.S. capacity of natural gas skills needed to deal with economics and costs. Later, I
would have to be built from the start at a cost of published another paper demonstrating, from elementary
close to $1 trillion. For thermochemical cycles, the design principles, that any thermochemical process must,
cost would be significantly higher. All of this must be by necessity, cost much more than electrolysis (2).
invested before a pipeline can be started. The same It is true that many scientists could have erred because
energy needs could be supplied by electrical energy they were not knowledgeable about process economics.
at only one-quarter the cost because electrical energy However, I also came to realize that other forces were
has a higher efficiency for almost all purposes. driving the program. Although they are inherently safer,
The investment for electricity can be spread over high-temperature nuclear reactors of the type required by
many years because a grid is already in existence. thermochemical cycles were deemed less competitive and
Furthermore, just the energy from the nuclear more costly than boiling water reactors, which created an
reactions would cost $50 per million BTU H2 . opportunity for their developers to seek more government
5. The most dangerous of all fuels, pressurized H2 support. A large nuclear research community was hungry
ignites with an invisible flame. Its explosive limits for grants and contracts.
are very wide, and its minimum ignition energy is Interestingly and unfortunately, the episode of the
ten times smaller than that of any other gaseous 1970s has been ignored or forgotten. However, before
fuel. This problem occurs not only when H2 is being we place large national resources behind an effort to
distributed or used, but even more so when it is again deal with such major problems as renewable energy,
generated in a thermochemical cycle. When H2 is global warming, and energy independence, we have to
produced by electrolysis, one can locate the plant understand what went wrong the first time a so-called
far away from the nuclear reactor generating the ‘‘hydrogen economy’’ was attempted.
electricity, which, however, is not true for helium, The paper I just published is about the reincarnation
which must be pumped in such large quantities that of the hydrogen economy. This time, all the proposed
it cannot be sent over long distances at reasonable processes are feasible, but the program itself seems to
cost. Society has accepted the risks of explosion make even less sense. As before, the main emphasis is
in conventional fuels but tries to limit exposure. on the replacement of conventional fuels with a national
Thus, even a small propane storage tank may not distribution system of H2 at a tremendous loss in efficiency
be transported through a tunnel, but what level of in the production and transmission steps. As before, the
risk from a potential explosion at a nuclear power same intractable safety and cost problems still persist.
plant complex is acceptable? None of these critical Similar driving forces remain at work. Large research
questions were asked then—nor are they being budgets are projected to find a use for a technology in
asked today. which DOE has invested huge sums of money to little or
no avail. This time, it is fuel cells powered by hydrogen,
These questions have been on my mind for a long a technology that is good for space applications, where
time. In the ‘‘age of technology,’’ how could the scientific H2 is used as a rocket fuel, and therefore available in
community have been misled by such obvious nonsense? small quantities, and other specialty applications, but
And why had so many engineering firms and large useless for applications that couple H2 generation with
companies come up with totally unrealistic cost estimates, the generation of electricity. Such systems have half the
at least by a magnitude too low, and thermal efficiency efficiency of combined cycle turbines, cost over five times
estimates that are unrealistically high? more (1,4), and create at least double the greenhouse
The intellectual climate of the time can be illustrated by emission, probably much more.
a small story. Science carried a six-page article, featured Today, there even is talk about switching cars from
on the cover, about a new cycle developed by General gasoline to H2 , which makes the transition problem even
Electric that promised to produce H2 at $2.50/million more difficult as the new supplying system cannot start
BTU H2 (at that time it would have cost $7.00/million locally, and this would also require the use of pressurized
BTU to produce H2 from oil). For fun, I flow-sheeted the H2 at 6000 psi, a pressure very seldom used in industry.
process and published the mass and heat balance in a Nobody mentions that in industry an H2 tank with
letter to Science (3) suggesting that if GE can reduce costs the capacity proposed for a family car requires storage
to $2.50, they should forget about hydrogen and change in a special room with a blowout wall. Any technician
the economy. By comparison, us poor chemical engineers approaching such a pressurized tank to check the valves
could not achieve such results at a cost of $250/million swings a wooden broom to prevent being fried by an
BTU H2 . Three months later, I was invited to GE’s central invisible flame which may come from a leaking valve,
laboratory where the Director of Chemical Research as above 1500 psi, an H2 leak is self-igniting (1). For a
offered me a consulting job. The head of the lab was a gas station, a minimum-sized storage tank would have the
chemist who had read about the need for a thermochemical explosive power of two thousand tons of TNT or two of the
cycle, and he actually developed the first one that worked. largest bombs used by the Air Force. A terrorist could just
His engineers had given him this cost estimate. When my open a valve and detonate a small bomb after 20 seconds.
letter was published, he asked Corporate Engineering The most obvious fallacy of the new H2 economy,
to make their own estimate and they confirmed my however, is that it solves all of our problems (5,6). For
THE MIRAGE OF THE H2 ECONOMY 479
those of us who accept the first and second laws of break on the investment. When Luz became large, the tax
thermodynamics, it is obvious that the use of H2 would breaks were taken away. If Luz had been allowed to build
increase the use of other energy resources. Arguments a few gigawatts, the price would have come down and
are made that, once we have a hydrogen distribution other companies would have come in; we might have had
system, we could capture the CO2 produced and sequester affordable solar energy today.
it. However, we have no idea how to sequester such vast We have today the resources and the technology to
amounts of CO2 safely. A successful policy can only be do something useful about global warming if we really
formulated after all the elements are solved and a good had the resolve. For example, for $50 billion a year
deal is known about costs. At present, the United States over twenty years, we could replace 40% of coal power
releases 100 million tons of sequestered CO2 back into the plants with solar power plants using proven technology;
atmosphere each year. we just have to engineer them better and provide them
In a system in which the basic science is known, with sufficient storage capacity. We are the only western
costs are reduced by innovative ideas that develop during nation who has enough sun and enough area for solar
the implementation stage of a project rather than by power. Alternatively, we could eliminate about 80% of
more research. The cost of liquid natural gas plants and our gasoline used for cars. We could easily afford to do
combined cycle power plants decreased in the last ten both, but we have to realize that no research can make
years by a factor of two even though the basic technology solar, nuclear, or other alternative energy competitive
did not change because many were built to compete in the with cheap oil or gas.
market. So, research, to be effective, needs to be coupled Our society has solved far more difficult large-scale
with a clear plan for implementation. technical problems successfully, e.g., space travel and the
In order to promote the use of fuel cells or other most advanced defense weapons in the world. Applying
new technologies, which are by one to two orders of our capabilities to global warming or to resource depletion
magnitude more expensive than available technologies, to preserve a livable world to our grandchildren would
the research community has invented the myth of require not only conviction but a strong political will to
‘‘learning curves.’’ In the last 40 years, computers became, do so.
through miniaturization, cheaper by several magnitudes, To give a recent example, thirty years ago, we faced the
but this option does not apply to energy systems. In the last problem of increasing pollution from our coal power plants.
20 years, the cost of solar cells was reduced by less than a We had the technology (scrubbers) to reduce the pollution
factor of two despite a tremendous research budget. Better substantially. It would have cost about $20–30 billion, but
engineering and mass production can reduce the cost of politics and the influence of senators from coal-producing
most new technologies by a factor of two to three. However, states prevented the adoption of such a solution. Instead,
this would require the building of large plants based on we spent $20 billion on ‘‘clean coal’’ research programs
competitive bidding, not a new research breakthrough. that failed to produce results that could help solve this
Large research budgets over the past 40 years have not problem. However, the ‘‘clean coal’’ effort did provide large
made the cost of fuel cells more competitive. research budgets along with the illusion that we were
The alternative sources that really reduce CO2 accomplishing something. Thirty years later, the same
emissions include solar and nuclear energy and biomass. plants are producing more electricity and consuming more
The use of H2 just makes every one of these options several coal—still poisoning the fish we eat with mercury, causing
times more expensive. Solar and nuclear energy produce smog that blocks us from seeing the sky, destroying our
electricity, and, for most purposes, electricity is far more forests, and, most importantly, impairing the health of all
efficient to use than H2 . To provide electricity to the motor of us more than any other form of pollution.
in a H2 car requires three to four times the solar or nuclear The private car is another example of the obstacles
capacity required for providing the electricity directly via that a serious national policy faces. In the 1970s, in
a battery. Although we don’t have good electric cars yet, response to the Arab oil embargo, oil price increases,
Toyota has just come out with a hybrid car with a plug-in and an alleged likelihood that the oil resource base
battery large enough for a 40-mile drive. Such a car could would become depleted within the next several decades,
reduce total gasoline consumption by 80%, but electric cars Presidents Nixon, Ford, and Carter proposed new energy
would only reduce greenhouse emissions if the electricity policy initiatives that included the goal of improving auto
were generated by solar energy. As for biomass, which fuel economy. These initiatives resulted in the Corporate
is a valuable but limited future resource, conversion by Average Fuel Economy standards (CAFÉ), which required
fermentation to alcohol followed by conversion to gasoline that the fleet fuel economy of new light-duty vehicles
or diesel is far cheaper and thermally more efficient than (5500 lbs or less) reach 27.5 mpg by a certain, date a
conversion to hydrogen. 30% reduction in gasoline demand. This standard limited
In the early 1990s, an Israeli company, Luz (7), built the ability of the companies to produce large cars and
solar electric power plants using a high-temperature heat stationwagons. The same provision set lower standards in
transfer fluid, which included sufficient storage to supply mileage and emissions for light-duty trucks weighing less
Los Angeles with 300 megawatts of electricity on a reliable than 5500 lbs.
basis. The Luz plant had another major advantage; the Nine years ago, the car industry realized that one could
design allowed for conventional fuel to be used as a backup produce large cars (SUVs) that fit all the definitions of
because, even in the desert, 10% of the days may be a light truck, such that one should be able to fold all
rainy. To be profitable, all Luz needed was a 50% tax seats to the floor to provide a major space occupying
480 HYDROGEN ION
the car and several others. It was a great success. As of the hydrogen atom. The latter has been extracted
light trucks convertible to luxury passenger cars do not mostly from Sienko and Plane (1). This article should be
count as passenger cars for the CAFE laws and are not studied in conjunction with the pH and Hydronium Ion
considered in the fleet average, this allowed car companies articles.
again to produce gas guzzlers. Eight years later, 30% of An ion in general terms can be any grouping of one or
the passenger cars sold in the United States are SUVs, and more atoms that carries a charge (1). The characteristics
all the efficiency gains of 30 years have been wiped out. of an ion include:
As we face the potential of global warming and
resource depletion that might endanger the future of 1. the sign (type) of charge: positive (cation) or
our grandchildren, we escape our responsibilities by negative (anion);
pretending that all problems can be solved better and at 2. the number of charges or valence (z): monovalent
lower cost through the mirage of more research. Thus, the divalent, etc;
H2 mirage allows car companies to continue to produce
3. the ionic radius (r): the distance from the center of
gas-guzzling, highly polluting SUVs while touting the
the nucleus to the outer shell of the atomic particle;
glorious future, an ‘‘H2 ’’ car that will miraculously be
energy efficient and pollution-free. And, as in the 70s, 4. the ionic potential: z/r, the ratio of the charge of the
thousands in the scientific and political communities are ion to the ionic radius;
promoting this mirage to the public. 5. the electronegativity: the power of the ion to
Thus, mirages like the ‘‘hydrogen economy’’ play an attract electrons;
important psychological function in our society, and many 6. the molar volume: the volume of one mole
in the technical community play along. The role engineers (cm3 /mol); and
and scientists should be playing today is to educate the 7. the ionic conductance: the ability to carry electri-
public and the politicians about the real options. We should cal current.
be reminded of an old cartoon in the Pogo series that bears
the caption ‘‘We have met the enemy and he is us.’’ When an ion is surrounded by many other ions/molecules
such as in a solution, its other properties such as
BIBLIOGRAPHY ionic strength and ion activity coefficient come to light.
1. Shinnar, R. (2003). The hydrogen economy, fuel cells, and Ottonello (2) is a major reference where many properties
electric cars. Technology in Society 25(4): 453–576. of various ions are tabulated.
2. Shinnar, R., Shapira, D., and Zakai, S. (1981). A thermochem- Hydrogen, the lightest of all gases, is colorless, odorless,
ical and hybrid cycles for hydrogen production. A differential and tasteless. Together with helium, it occupies the first
economic comparison with electrolytes. IEC Process Design & period of the periodic chart of elements and is the most
Development 20: 581. abundant element in the solar system. After helium, it has
3. Shinnar, R. (1975). Thermochemical hydrogen generation heat the second lowest boiling temperature of all elements (2).
requirements and costs. Science 188: 1036. Because of its low boiling point, liquid hydrogen is used
4. Shinnar, R. (2003). Fuel-cell faux pas. Chemical & Engineering as a cryogenic fluid (to produce low temperatures), and
News 81(14): 4. because of its high heat of combustion (120 kJ/g), it
5. Rifkin, J. (2003). The hydrogen economy. Environmental is a valuable rocket fuel. Molecular hydrogen, H2 , can
Magazine. January. combine directly with most elements. Mixtures of H2 and
6. Bush, G.W. (2003). State of the Union Address. January 28. O2 can be explosive. Much of the industrial consumption
7. Shinnar, R. (1993). The rise and fall of luz. ChemTech 23: of hydrogen is in hydrogenation reactions, the addition of
50–53. H2 to other molecules. Water and power are required to
produce hydrogen. In its compounds, hydrogen is found in
READING LIST three oxidation states, +1, −1, and 0. Hydrogen is such
Energy Research of the Department of Energy. (2001). Was It an important element that it has its own association. The
Worth It? National Research Council. National Hydrogen Association (WWW. Hydrogenus.org),
Effectiveness and Impact of Corporate Average Fuel Economy for example, was formed in 1989 (in the U.S.) to act as a
(CAFE) Standards. (2002). National Research Council. focal point in hydrogen related matters. Anyone interested
in water related matters should know the important
characteristics of hydrogen because of its place in the
HYDROGEN ION water molecule.
GHOLAM A. KAZEMI
Shahrood University of DEFINITION OF A HYDROGEN ION
Technology
Shahrood, Iran A hydrogen ion, H+ , is a monovalent cation that occurs
when a hydrogen atom loses its only electron:
INTRODUCTION H −−−→ H+ + e−
This article includes an introduction to the ion and its Therefore, the hydrogen ion has only one proton and
properties followed by an overview of the characteristics no electron. This is why ‘‘proton’’ is another name for
HYDROGEN ION 481
hydrogen ion. Some references state that there is no such Model 420A are examples. Measurement of pH cannot
ion as a hydrogen ion and what occurs, is actually a be more accurate than 0.02 pH units because the standard
hydronium ion, H3 O+ , rather than a hydrogen ion (3). solutions or buffers which have been calibrated by the U.S.
The hydride ion, H− , occurs when a hydride such as NaH Bureau of Standards (NBS) have an absolute accuracy in
dissociates: the range of ±0.01 to 0.02 pH units (7). In field conditions,
NaH −−−→ Na+ + H− where disturbing factors are ample, the accuracy is
usually ±0.05 pH units. Colorimetry is one older method
In this reaction, the hydride ion radius is 1.46 Å. Molecular of measuring H+ concentration.
hydrogen is formed from the reaction of a hydride ion and
a hydrogen ion (4): Some Examples
Pure neutral water at 25 ◦ C has a pH of 7, a H+
H+ + H− −−−→ H2
concentration of 10−7 mg/l. A 1-molal solution of HCl
will have an H+ concentration of 1 and a 0.01-molal
H3 + is an ion that is artificially generated by the
solution will have an H+ concentration of 10−2 mg/l. The
interaction of H+ and H2 . For instance, generation of
concentration of hydrogen ion in a 0.01-molal solution
H3 + is the most critical specification of the ion source in a
of a weak such as H2 CO3 is only 10−4.2 mg/l, and it is
mass spectrometer (5).
10−10.8 mg/l in an aqueous solution of pH 3.2.
Many authors prefer to use the term ‘‘hydrogen ion
CHARACTERISTICS OF THE HYDROGEN ION activity’’ rather than hydrogen ion concentration. The
reason that the ‘‘pH’’ concept has been developed is the
Some major characteristics of the hydrogen ion are very low concentration of hydrogen ion and the difficulty in
presented in Table 1. comprehending such low, sometime ambiguous, numbers.
In chemistry and geochemistry, it is assumed that In solutions, the concentration of H+ never reaches zero;
the standard heat capacity, C0p ; standard partial molal there is no solution that is 100% basic or 100% acidic.
Gibbs free energy from the elements, G0f ; standard partial
molal enthalpy of formation from the elements, Hf0 ; PRODUCTION OF HYDROGEN ION IN WATER
p
standard partial molal entropy, S0 ; and standard partial
0
molal volume, Vp , of the hydrogen ion equal zero at all Hydrogen ion is produced in natural water from the three
temperatures and pressures (2,7). following main sources as well as from humic and fluvic
acids, volcanic gases, acid rain, and short-chain organic
MEASUREMENT OF HYDROGEN ION CONCENTRATION acids present in some oil field brines (8).
More than 100 years ago, Arrhenius suggested that acid 1. Hydrolysis
properties are attributable to the presence of an easily
detachable hydrogen ion, H+ (1). Now, everyone knows H2 O −−−→ H+ + OH−
that the concentration of hydrogen ion in any solution is
a measure of the solution’s acidity or pH (pH = − log H+ ). As stated above, however, it might be more appropriate to
Thus, the measurement of H+ concentration is better write the previous equation in the form:
described in the pH article of the encyclopedia and is
only briefly dealt with here. 2H2 O −−−→ H3 O+ + OH−
The H+ ion is the easiest ion concentration to measure.
This is done using a hydrogen electrode, which is formed 2. Dissociation
by bubbling H2 over a platinum electrode surface (7). A
H2 CO3 −−−→ H+ + HCO3 −
variety of these electrodes are commercially available;
Metrohm pH meter Model 744 and Orion pH meter
3. Oxidation
Table 1. Notable Characteristics of the Hydrogen Ion, H+a

2FeS2 + 7.5O2 + 7H2 O −−−→ 2Fe(OH)3
Type Cation + 8H+ + 4SO4 2−

Valence 1
Radiusb —
IMPORTANCE OF THE HYDROGEN ION
Ionic potential (IP)c —
Electronegativity 2.2
Molar volume 0 (This is based on convention)
Some of the important functions and behaviors of the
Ionic conductance 349.8 cm2 × (equiv )−1 hydrogen ion are described as following. Considering the
Charge to mass ratio (q/m) 9.6 × 107 C/kg huge and universal importance of H+ , these should be
regarded only as examples.
a
References 4, 6, 7.
b
It is difficult to define because the electron of the hydrogen atom has gone,
and only the proton remains. So, there is no outer shell left to measure its
1. The radius of the hydrogen ion is very close to the
distance to the proton. size of the neutron. This similarity is the principle
c
It cannot be calculated because there is no specified ionic radius. used to design and develop compensated neutron
482 THE HYDRONIUM ION
logs (CNL) and neutron probes. The former is used 11. Wilson, E.M. (1984). Engineering Hydrology, 3rd Edn.
as an important tool for calculating the porosity of Macmillan, London.
petroleum reservoirs and aquifers (9,10), and the 12. Deutsch, W.J. (1997). Groundwater Geochemistry: Funda-
latter is of extensive use in agricultural practices for mentals and Applications to Contamination. Lewis, Boca
measuring soil moisture (11). Raton, FL.
13. Tebbutt, T.H.Y. (1995). Principles of Water Quality Control.
2. The concentration of hydrogen ion, pH, is a
Reprint of 4th Edn. Butterworth Heinemann, Oxford.
master variable of the groundwater system because
the hydrogen ion participates in many important
chemical reactions such as (12):
THE HYDRONIUM ION
a: mineral dissolution—precipitation CARL W. DAVID
CaCO3 + H+ = Ca2+ + HCO3 − University of Connecticut
Storrs, Connecticut
b: aqueous complexation
Fe3+ + H2 O = Fe (OH)2+ + H+
Naked protons cannot exist in aqueous solution. Even
c: adsorption –desorption when one writes
≡ FeOH + Cu+ = ≡ FeOCu + H+ HCl(g) → H+ (aq) + Cl− (aq) (1)
3. The hydrogen ion plays a major role in denitrification one knows that the H+ (aq) is a representation of a
through which various forms of nitrogen in the soil significantly more complicated moiety (or, to be more
revert to N2 (6): precise, set of moieties). The simplest representation of
H+ (aq) is the hydronium ion (sometimes also referred to
4NO3 − + 4H+ −−−→ 2N2 + 5O2 + 2H2 O as the oxonium ion or the hydroxonium ion),
H3 O+ (aq), but this is just the first in a series the next
It also plays a major role in redox reactions and few members of which are H5 O2 + , H7 O3 + , H9 O4 + , and
acid–base reactions. in summary (H2 O)n H3 O+ , the hydrated hydronium ions
4. In sanitary engineering and water treatment, pH (Fig. 1).
is an important factor in chemical coagulation, The hydronium ion is isoelectronic with ammonia (NH3 )
disinfection, water softening, corrosion control, and and therefore, ought to be pyramidal with the oxygen at
biological processes (13). the apex of the pyramid and the three protons disposed
5. The most remarkable function of hydrogen ion is its below, with symmetry C3v . Like ammonia, it should invert,
control of the acidity or alkalinity of solutions. with the O[or N] passing through the plane of the three
protons from one side to the other. In the planar transition
state, when all four nuclei are coplanar, the symmetry
BIBLIOGRAPHY should be D3h . Also, like ammonia, there should be a
barrier to this umbrella inversion; appropriate tunneling
1. Sienko, M.J. and Plane, R.A. (1979). Chemistry: Principles has been observed spectroscopically (ν2 ). It is thought
and Applications. McGraw-Hill, New York. that the O–H distance is about 0.96 Å, and the H–O–H
2. Ottonello, G. (1997). Principles of Geochemistry. Columbia angle about 115◦ . The barrier to inversion, a saddle point
University Press, New York. on the energy hypersurface, is thought to be between 1
3. Fetter, C.W. (1988). Applied Hydrogeology. 2nd Edn. CBS and 5 kcal/mol (Fig. 2).
Publishers and Distributors, Delhi. The simplest occurrence of the hydronium ion can be
4. Lagowski, J.J. (1973). Modern Inorganic Chemistry. Marcel seen in the gas phase water dimer, (H2 O)2 . The proton
Dekker, New York. transfer reaction inside the dimer would be
5. Attendorn, H.G. and Bowen, R.N.C. (1997). Radioactive and
Stable Isotope Geology. Chapman & Hall, London.
H2 O–H2 O(g) → H3 O+ (g) : OH− (g) (2)
6. Evangelou, V.P. (1998). Environmental Soil and Water
Chemistry: Principles and Applications. John Wiley & Sons,
forming an ion pair (Fig. 3).
New York.
7. Langmuir, D. (1997). Aqueous Environmental Geochemistry.
Prentice-Hall, Upper Saddle River, NJ.
8. Hounslow, A.W. (1995). Water Quality Data: Analysis and +
Interpretation. Lewis, New York. H
9. Hearst, J.R., Nelson, P.H., and Paillet, F.L. (Eds.). (2001).
Well Logging for Physical Properties: A Handbook for O
Geophysicists, Geologists, and Engineers, 2nd Edn. John
Wiley & Sons, New York. H H
10. Todd, D.K. (1980). Groundwater Hydrology, 2nd Edn. John
Wiley & Sons, New York. Figure 1. The hydronium cation.
THE HYDRONIUM ION 483
H
H +
O H
H
O H − +
HO O
O H
H
H H
H H
Energy
O
H O
55.35 cm −1 H H
Figure 4. With three molecules of water, one can form a hydrated
+ hydronium cation and a hydroxide anion.
H + H
O H H O
Clearly, in liquid water, the hydroxide and hydronium
H H ions are solvated but not statically enough to allow
definition of a moiety.
Inversion coordinate
In aqueous media, for the analog of the proton transfer
Figure 2. Potential energy surface for the hydronium ion’s reaction, one could write
inversion coordinate.
2H2 O(l) → H3 O+ (aq) + OH− (aq) (3)
the prototype of all acid–base chemistry in an aque-

H
ous medium.
O H In aqueous acidic media, the hydronium ion is rarely
H
O H − written. Instead, the older notation is employed. For
O +
O instance, for acetic acid, CH3 COOH, one would write
H H
H H CH3 COOH(aq) ⇔ H+ (aq) + CH3 COO− (aq) (4)
Figure 3. Dimer water forming an ion pair through pro-
ton transfer. partial dissociation into a proton and a acetate anion,
rather than the more accurate
H2 O(l) + CH3 COOH(aq) ⇔ H3 O+ (aq) + CH3 COO− (aq)

In the dimer, one molecule of water is a proton donor to (5)
the other (which is a proton acceptor). The former water The conductance properties of water are attributable in
is a Bronsted acid, the latter a Bronsted base. On the line part to the hydronium cation and its peculiar properties.
of the hydrogen bond, HO–Hdonor ::::OH2 , the movement of The Grotthuss mechanism for proton transfer in ice
the donor proton from one oxygen to another is equivalent employs hydronium ions hydrogen bound to adjacent water
to the proton transfer (above). The O–O distance in the molecules, in which, like the dimer, an excess proton
dimer is about 3 Å, and the OH distances are about 1 Å, migrates (or tunnels) from one oxygen to an adjacent one,
so the shift of 1 Å in a proton’s position (between two moving only a tiny distance, whereupon the charge has
oxygens) creates an OH− :::H3 O+ ion pair. shifted enormously (one O–O distance).
Hydrated versions of this ionization process illustrate A similar mechanism for OH− (aq) charge migration
that the actual process in liquid water is not cleanly most likely also obtains. Here a water donor adjacent
definable. Figure 4 shows the results of three water to the hydroxide anion shifts its donating proton to
molecules involved in ionization, whereas Fig. 5 shows the hydroxide, thereby itself becoming a hydroxide anion
the results with four water molecules. (Fig. 6).
O H
H
O H −
O +
H H O H
H H H
O H
O O
H O
H H Figure 5. With four waters, hydrated hydroxide
H H
and hydrated hydronium ions are possible.
484 INFILTRATION AND SOIL MOISTURE PROCESSES
H
Ground
O Exfiltration
H O surface
H H
Infiltration
H
O
H
O
H
H
H O Redistribution
Plant
H uptake
Interflow
Water Capillary rise

table
Recharge
H −
H Ground water
H
O
O H
H O Figure 1. Unsaturated zone definition and active processes (1).

H
O
+
H
H CONTROLS ON INFILTRATION AND SOIL MOISTURE
H O
To characterize soil moisture and infiltration, the

H physical controls on these processes must be considered.
The primary soil controls will be considered in this
Figure 6. The Grotthuss mechanism for positive and negative chapter; however, other factors such as soil chemistry,
charge creation and migration in aqueous solution.
thickness, soil layering or horizons, and preferential flow
paths, as well as vegetation cover, tillage, roughness,
topography, temperature, and rainfall intensity also exert
INFILTRATION AND SOIL MOISTURE important controls (4).
PROCESSES A soil’s particle size distribution has a large impact
on its hydraulic properties. Soil particles less than 2 mm
PAUL R. HOUSER
in diameter are divided into three texture groups (sand,
(from The Handbook of Weather,
silt, and clay) that help to classify broad soil types and
Climate, and Water:
soil water responses (Fig. 2) (5). The type of clay and the
Atmospheric Chemistry,
Hydrology, and Societal
coarse material over 2 mm in diameter can also have a
Impacts, Wiley 2003) significant impact on soil water properties. An overview of
methods for determining particle size properties is given
by Gee and Bauder (6).
Infiltration is the process of water entry from surface Bulk density, ρb (M/L3 ) is the ratio of the weight of
sources such as rainfall, snowmelt, or irrigation into dry solids to the bulk volume of the soil, and porosity, φ
the soil. The infiltration process is a component in (M3 /M3 ), is the total volume occupied by pores per unit
the overall unsaturated redistribution process (Fig. 1) volume of soil:
(1) that results in soil moisture availability for use Va + Vw ρb
by vegetation transpiration, exfiltration (or evaporation) φ= =1− (1)
Vs ρm
processes, chemical transport, and groundwater recharge.
Soil moisture, in turn, controls the partitioning of where Vs (L3 ) is the total volume of soil, Va (L3 ) is the
subsequent precipitation into infiltration and runoff, and volume of air, Vw (L3 ) is the volume of water, and ρm
the partitioning of available energy between sensible and (ML−3 ) is the particle density (normally about 2.65 g/cm3 ).
latent heat flux. The volumetric water content, or soil moisture, θ (L3
Because of the importance of soil moisture on multiple L−3 ) is the ratio of water volume to soil volume:
processes, its definition can be elusive (2); however, it is Vw Ww ρb
most often described as moisture in the unsaturated sur- θ= = (2)
Vs Wd ρw
face layers (first 1 to 2 m) of soil that can interact with
the atmosphere through evapotranspiration and precipi- where Ww (M) is the weight of water, Wd (M) is the
tation (3). weight of dry soil, and ρw (M/L3 ) is the density of water.
INFILTRATION AND SOIL MOISTURE PROCESSES 485
100 0.60
10
90
20 Porosity, f
80
0.50
30
lay
Pe
70
tc
Clay
rc
igh
40
en
60
we
tb
0.40
yw
50
by
Proportion by volume
50
eig
nt
Silty
Field
rce
Sandy clay
60
ht
clay
capacity, qfc
Pe
40
sill
Silty Clay
Clay loam loam 70 0.30
30
Sandy clay loam
80
20
Loam Silt loam
Sandy Loam 90 0.20 Wilting
10 Loamy Silt point, qpwp
Sand Sand 0
10
100
90
80
70
60
50
40
30
20
10
0.10
Percent by weight sand Clay
Figure 2. Soil textural triangle describing the relationship Loam Heavy
between texture and particle size distribution (5). clay loam
0 Fine Clay loam
Sand sandy loam
Light clay loam
sandy loam Silt loam
“Absorbed” Water
Fine sand
Figure 4. Water holding properties of various soils (8).
“Capillary”
water
Suction
Clayey soil
Particles
Figure 3. Capillarity and adsorption combine to produce suction Sandy soil

(7).
Water content
Soil moisture can vary in both time and space, with a Figure 5. Effect of texture on water retention characteristics (9).
theoretical range from 0 to ϕ, but for natural soils the
range is significantly reduced due to isolated pore space
and tightly held or ‘‘adsorbed’’ water (Fig. 3) (7). If a soil is matric potential, ψ (L), increases as the radii of curvature
saturated, then allowed to drain until the remaining water of the meniscus or water content decreases (Fig. 5) (9).
held by surface tension is in equilibrium with gravitational Matric potential is expressed in reference to atmospheric
forces, it is at field capacity, θf . Vegetation can remove pressure, so for saturated soil ψ = 0 and for unsaturated
water from the soil until the permanent wilting point, θw , soil ψ < 0.
is reached. Therefore, the available water content for plant The hydraulic conductivity, K (L/T), is a measure of the
use, θa = θf − θw . Typical ranges of porosity, field capacity, ability of the soil to transmit water that varies nonlinearly
and wilting point for different soils are given in Fig. 4 (8). over a large range depending on both soil properties and
In unsaturated soils, water is held in the soil against water content (Fig. 6) (10). Many laboratory and field
gravity by surface tension (Fig. 3). This tension, suction, or hydraulic conductivity measurement methods exist for
10−3
10−4
Sand
Hydraulic conductivity, K, ms−1
10−5
10−6 Loam
10−7
10−8
10−9 0 0.1 0.2 0.3 0.4 0.5 Figure 7. Changes in water retention characteristics between
sorption and desorption (16).
Water content, q, m3 m−3
Figure 6. Effect of texture and soil moisture on hydraulic
conductivity (10). Combining Darcy’s law with the law of conservation of
mass results in a description of unsaturated flow called
Richards equation (19):
use with various soils; see Bouwer and Jackson (11) or
∂θ ∂ K ∂θ ∂K
Green et al. (12) for details. = − (4)
∂t ∂z C ∂z ∂z
Soil water content can significantly impact infiltra-
tion by (1) increasing the hydraulic conductivity, which where C = −∂θ/∂ψ is the water content change in a unit
increases infiltration, and (2) reducing the surface tension soil volume per unit matric potential, ψ change. The
that draws moisture into the soil, which reduces infiltra- Richards equation is the basis for most simulations of
tion. The net effect of these impacts depends on the water infiltration and redistribution of water in unsaturated
content itself, the water input rate, and duration and the soil. Using some approximations, analytical solutions of
distribution of hydraulic conductivity. the Richards equation are available (20,21) that show good
The water retention characteristic describes a soil’s agreement with observations (22). The Richards equation
ability to store and release water and is defined by is based on saturated flow theory, and does not account
the relationship between soil moisture and the matric for all of the processes active in natural systems, so it may
potential (Fig. 5). This is a power function relationship not always perform well (23).
that has been described by Brooks and Corey (13)
and Van Genuchten, (14) among others. The water
INFILTRATION ESTIMATION
tension characteristic is usually measured in air pressure
chambers where the water content of a soil sample can be Some basic principles that govern the movement of water
monitored over a wide pressure range (15). into the soil can be used to predict infiltration. The
The water retention relationship may actually change infiltration capacity, f (L), is the maximum rate that a
between drying and wetting due to the entrapment of air soil in a given condition can absorb water and generally
in soil pores (Fig. 7) (16). For practical applications, this decreases as soil moisture increases. If the rainfall rate
effect, called hysteresis, is usually neglected (17). is less than the infiltration capacity, then infiltration
proceeds at the capacity rate. However, if the rainfall rate
exceeds the infiltration capacity, then infiltration proceeds
PRINCIPLES OF SOIL WATER MOVEMENT at the capacity rate, and the excess rainfall ponds on the
surface or runs off. As the time from the onset of rainfall
Through experiments on saturated water flow through increases, infiltration rates decrease due to soil moisture
sand beds, Darcy (18) found that the rate of flow, Q increases, raindrop impact, and the clogging of soil pores,
(L3 /T), through a cross-sectional area A (L2 ), is directly until a steady-state infiltration rate is reached (Fig. 8) (24).
proportional to head loss (e.g., water elevation difference), Existing infiltration models use empirical, approximate, or
H (L), and inversely to the flow path length, l (L): physical approaches to predict infiltration.(25)
H Empirical. Empirical infiltration models generally uti-

Q = KA (3) lize a mathematical function whose shape as a function of
l
5 where S is a parameter called sorptivity, t is time

from ponding, and A is a constant that depends on soil
Rate of rainfall, infiltration, and runoff (cm/hr)
properties. In this model, the infiltration rate approaches a

Rainfall
4 constant equal to the hydraulic conductivity at the surface
Runoff water content, and the wetting front advances without
changing its shape and approaches a constant velocity.
3 Physical. Recent advances in numerical methods and
computing has facilitated the practical application of the
Richards equation to realistic flow problems. Such pack-
2 ages can simulate water infiltration and redistribution
using the Richards equation and including precipitation,
runoff, drainage, evaporation, and transpiration processes
1 (30).
Infiltration
INFILTRATION MEASUREMENT
0 Infiltration rates can be measured at a point using

20 40 60 80 100 120 140
a variety of methods described here, each appropriate
Minutes from onset of rainfall
for certain conditions. However, because of the large
Figure 8. Idealized relationship between rainfall, infiltration, temporal and spatial variability of infiltration processes,
and runoff rates (24). catchment average infiltration rates may be desired, which
can be obtained through the water balance analysis of
rainfall–runoff observations (31).
time, t, matches observations and then attempts a physical
explanation of the process. Ring Infiltrometer. This simple method is most appro-
Kostiakov (26) proposed the simple infiltration rate, priate for flood irrigation or pond seepage infiltration. A
f (L/T) model: cylindrical metal ring is sealed at the surface and flooded.
f = αγ tα−1 (5) Intake measurements are recorded until steady-state con-
ditions are reached (32). If the effects of lateral flow
where α and γ are constants that have no particular are significant, then a double-ring infiltrometer can be
meaning and must be evaluated by fitting the model to used. Due to ponding conditions within the ring, observed
experimental data. infiltration rates are often higher than under natural
Horton’s (27) infiltration model has been widely used conditions (33).
in hydrologic simulation. It relates infiltration capacity to
initial infiltration rate, and f0 , the constant infiltration Sprinkler Infiltrometer. This method is appropriate for
rate at large times, fc : quantifying infiltration from rainfall. Artificial rainfall
simulators are used to deliver a specified rainfall rate
f = fc + (f0 − fc )e−βt (6) to a well-defined plot. Runoff from the plot is measured,
allowing computation of the infiltration rate (34,35).
where β is a soil parameter describing the rate of decrease
of infiltration. Tension Infiltrometer. The tension or disk infiltrometer
employs a soil contact plate and a water column that
Approximate. Analysis approximations to the Richards is used to control the matric potential of the infiltrating
equation are possible if several simplifying assumptions water. By varying the tension, the effect of different size
are made. Most approximate infiltration models treat the macropores can be determined (36,37).
soil as a semi-infinite medium, with the soil saturating
above a wetting front. Furrow Infiltrometer. This method is useful if informa-
Green and Ampt (28) assumed in a soil with constant tion on infiltration of flowing water in irrigation furrows
hydraulic properties, the matric potential at the moving is desired. Either the water added to a small section of
wetting front is constant, leading to a discontinuous blocked off furrow to maintain a constant depth or the
change in soil moisture at the wetting front: inflow—outflow of a furrow segment can be monitored
to determine the infiltration characteristics of the system
(ϕ − θi )Sf (38).
f =K 1+ (7)
F
SOIL MOISTURE MEASUREMENT
where Sf (L) is the effective suction at the wetting front, θi
is the initial water content, and F (L) is the accumulated Soil water content can be determined directly using
infiltration. gravimetric techniques or indirectly by inferring it from a
Phillip (29) proposed that the first two terms in a series property of the soil (39,40).
of powers of t1/2 could be used to approximate infiltration:
Gravimetric. The oven-drying soil moisture measure-
f = 12 St1/2 + A (8) ment technique is the standard for calibration of all
other methods but is time consuming and destructive. be used to characterize catchment (or even larger) scale
The method involves obtaining a wet soil sample weight, processes. It is clear from the nonlinear character of soil
drying the sample at 105 ◦ C for 24 h, then obtaining the water processes that catchment average infiltration cannot
dry sample weight [see Eq. (2)]. be computed based on catchment average soil properties. It
is also clear that the physical meaning of a soil property,
Neutron Thermalization. High-energy neutrons are say porosity, is relative to the volume over which it is
emitted by a radioactive source into the soil and are averaged [47]. However, there is a need to understand
preferentially slowed by hydrogen atoms. The number and reduce this complexity for the purposes of prediction
of slow neutrons returning to the detector are a measure and management. Several approaches, including dividing
of soil moisture. the catchment into hydrologically similar subareas [48],
Gamma Attenuation. The attenuation in soil of gamma various statistical approaches [49], and scaling and
rays emitted from caesium-137 is directly related to similarity theory [50,51] have made headway toward an
soil density. If the soil’s bulk density is assumed to be understanding of infiltration and soil moisture spatial
constant, then changes in attenuation reflect changes in variability, but are not being widely used in practical
soil moisture (41). applications.
One of the most important recent findings in this regard
Time-Domain Reflectometry (TDR). TDR measures the is the scale invariance of soil water movement(i.e., the
soil’s dielectric constant, which is directly related to soil Richards equation) remain uniform across spatial scales
moisture, by measuring the transmit time of a voltage [52]. This new understanding of the underlying symmetry
pulse applied to a soil probe. of the Richards equation may help to facilitate a workable
scale invariant analytical soil water dynamical model.
Tensiometric Techniques. This method measures the Finally, there is acontinuing need for the observation of
capillary or moisture potential through a liquid-filled soil properties, soil moisture, and infiltration processes at
porous cup connected to a vacuum gage. Conversion to multiple scales to facilitate understanding and prediction
soil moisture requires knowledge of the water retention of these complex and socially significant processes. It is
characteristic. likely that remote sensing of soil moisture and other land
Resistance. The electrical resistance or conductivity of surface factors will be instrumental in this respect.
a porous block (nylon, fiberglass, or gypsum) imbedded
in the soil depends primarily on the water content of BIBLIOGRAPHY
the block. However, because of salinity and temperature
sensitivity, measurements of these sensors are of limited 1. Dingman, S.L. (1994). Physical Hydrology. Prentice-Hall,
accuracy (42). Englewood Cliffs, NJ, p. 211.
2. Lawford, R.G. (1992). An Overview of soil moisture and its
Heat Dissipation. Changes in the thermal conductivity role in the climate system. In: F.J. Eley, R. Granger and
of a porous block imbedded in the soil depend primarily L. Martin (Eds.). Soil Moisture; Modeling and Monitoring
on the water content of the block. The dissipation of a for Regional Planning, National Hydrology research Centre
heat pulse applied to the block can be monitored using Symposium No.9 Poceedings, pp. 1–12.
thermistors, then the soil water content can be determined 3. Schmugge, T., Jackson, T.J., and McKim, H.L. (1979). Survey
from calibration information. of in-situ and remote sensing methods for soil moisture
determination. In: Satellite Hydrology. American Water
Remote Sensing. Soil moisture can be remotely sensed Resources Association.
with just about any frequency where there is little 4. Rawls, W.J., Ahuja, L.R., Brakensiek, D.L., and Shirmoham-
atmospheric absorption (43). But, it is generally accepted madi, A. (1993). Infiltration and soil water movement. In:
that long wavelength, passive microwave sensors have Handbook of Hydrology. D.R. Maidment (Ed.). McGraw-Hill,
the best chance of obtaining soil moisture measurements New York, pp. 5.1–5.51.
that contain little error introduced by vegetation and 5. Hillel, D. (1982). Introduction to Soil Physics. Academic, New
roughness and offer great potential to remotely sense soil York, p. 29.
moisture content with depth due to differential microwave 6. Gee, G.W., and Bauder, J.W. (1986). Particle size analysis. In:
absorption with varying dielectric constant (44). Methods of Soil Analysis, Part I—Physical and Mineralogical
Methods, 2nd Edn. A. Klute (Ed.). American Society of
SPATIAL AND TEMPORAL VARIABILITY Agronomy Monograph 9, pp. 383–411.
7. Hillel, D. (1980). Fundamentals of Soil Physics. Academic,
Natural soils exhibit considerable spatial heterogeneity New York, p. 143.
in both the horizontal and vertical directions, and at 8. Dunne, T. and Leopold, L. (1978). Water in Environmental
all distance scales from the pore to the continent to Planning. W.H. Freeman, New York, p. 175.
a degree that it is difficult to capture this variability 9. Hillel, D. (1980). Fundamentals of Soil Physics. Academic,
in routine measurements [45,46]. This large variations New York, p. 150.
in soil properties, infiltration, and soil moisture over 10. Marshall, T.J., and Holmes, J.W. (1988). Soil Physics, 2nd
relatively small areas makes it difficult to transfer Edn. Cambridge University Press, p. 87.
the understanding of processes developed at a point to 11. Bouwer, H. and Jackson, R.D. (1974). Determining soil prop-
catchment scales. Many hydrological models assume that erties. In: Drainage for Agriculture. J. van Schilfgaard (Ed.).
a single spatially representative average soil property can American Society of Agronomy, Madison, WI, pp. 611–672.
12. Green, R.E., Ahuja, L.R., and Chong, S.K. (1986). Hydraulic 32. Bouwer, H. (1986). Intake rate: Cylinder infiltrometer. In:
conductivity, diffusivity, and sorptivity of unsaturated Methods of Soil Analysis, Part I—Physical and Mineralogical
soils—field methods. In: Methods of Soil Analysis, Part Methods, 2nd Edn. A. Klute (Ed.). American Society of
I—Physical and Mineralogical Methods, 2nd Edn. A. Klute Agronomy Monograph 9, pp. 825–844.
(Ed.). American Society of Agronomy Monograph 9, 33. Lukens, R.P. (Ed.). (1981). Annual Book of ASTM Standards,
pp. 771–798. Part 19: Soil and Rock, Building Stones, pp. 509–514.
13. Brooks, R.H. and Corey, A.T. (1964). Hydraulic Properties of 34. Agassi, M., Shainberg, I., and Morin, J. (1988). Effects on
Porous Media, Hydrology Paper 3, Colorado State University, seal properties of changes on drop energy and water salinity
Fort Collins, CO. during a continuous rainstorm. Aust. J. Soil Res. 26: 1–10.
14. Van Genuchten, M.Th. (1980). A closed-form equation for 35. Peterson, A. and Bubenzer, G. (1986). Intake rate sprinkler
predicting the hydraulic conductivity of unsaturated soils. infiltrometer. In: Methods of Soil Analysis, Part I—Physical
Soil Sci. Soc. Am. J. 44: 892–898.
and Mineralogical Methods, 2nd Edn. A. Klute (Ed.).
15. Cassel, D.K. and Klute, A. (1986). Water potential: Tensiom- American Society of Agronomy Monograph 9, pp. 845–870.
etry. In: Methods of Soil Analysis, Part I—Physical and
36. Ankeny, M.D., Kaspar, T.C., and Horton, R. (1988). Design
Mineralogical Methods, 2nd Edn. A. Klute (Ed.). American
for an automated tension infiltrometer. Soil Sci. Soc. Am. J.
Society of Agronomy Monograph 9, pp. 563–596.
52: 893–896.
16. Hillel, D. (1980). Fundamentals of Soil Physics. Academic,
37. Perrouix, K.M. and White, I. (1988). Designs of disc
New York, p. 153.
permeameters. Soil Sci. Soc. Am. J. 52: 1205–1215.
17. Rawls, W.J., Ahuja, L.R., Brakensiek, D.L., and
38. Kincaid, D.C. (1986). Intake rate: Border and furrow. In:
Shirmohammadi, A. (1993). Infiltration and Soil Water
Methods of Soil Analysis, Part I—Physical and mineralogical
Movement. In: Handbook of Hydrology. D.R. Maidment (Ed.).
Methods, 2nd Edn. A. Klute (Ed.). American Society of
McGraw-Hill, New York, pp. 5.5–5.6.
Agronomy Monograph 9, pp. 871–887.
18. Darcy, H. (1856). Les fontaines publiques de la ville de Dijon.
39. Gardner, W.H. (1986). Water content. In: Methods of Soil
Dalmont, Paris.
Analysis, Part I—Physical and Mineralogical Methods,
19. Richards, L.A. (1931). Capillary conduction of liquids in 2nd Edn. A. Klute (Ed.). American Society of Agronomy
porous mediums. Physics 1: 318–333. Monograph 9, pp. 493–544.
20. Kuhnel, V., Dooge, V.C.I., Sander, G.C., and O’Kane, J.P.J. 40. Schmugge, T.J., Jackson, T.J., and McKim, H.L. (1980).
(1990). Duration of atmosphere-controlled and soil-controlled Survey of methods for soil moisture determination. Water
phases of infiltration for constant rainfall at a soil surface.
Resour. Res. 16(6): 961–979.
Ann. Geophys. 8: 11–20.
41. Jury, W.A., Gardner, W.R., and Gardner, W.H. (1991). Soil
21. Sposito, G. Recent advances associated with soil water in the
Physics, 5th Edn. Wiley, pp. 45–47.
unsaturated zone. In: Reviews of Geophysics, Supplement,
U.S.National Report to International Union of Geodesy and 42. Hillel, D. (1982). Introduction to Soil Physics. Academic, New
Geophysics 1991–1994, pp. 1059–1065. York, 1982, pp. 57–89.
22. Whisler, F.D. and Bouwer, H. (1970). A comparison of 43. Schmugge, T.J. (1985). Remote sensing of soil moisture. In:
methods for calculating vertical drainage and infiltration in Hydrological Forecasting. John Wiley & Sons, New York.
soils. J. Hydrol. 10: 1–19. 44. Ulaby, F.T., Bradley, G.A., and Dobson, M.C. (1979). Potential
23. Nielsen, D.R., Biggar, J.W., and Davidson, J.M. (1962). Exper- application of satellite radar to monitor soil moisture. In:
imental consideration of diffusion analysis in unsaturated Satellite Hydrology, American Water Resources Association,
flow problems. Soil Soc. Am. Proc. 26: 107–111. pp. 363–370.
24. Dunne, T. and Leopold, L. (1978). Water in Environmental 45. Jury, W.A., Gardner, W.R., and Gardner, W.H. (1991). Soil
Planning, W.H. Freeman, New York, p. 169. Physics, 5th Edn. John Wiley & Sons, New York, pp. 268–
293.
25. Rawls, W.J., Ahuja, L.R., Brakensiek, D.L., and
Shirmohammadi, A. (1993). Infiltration and soil water 46. Dingman, S.L. (1994). Physical Hydrology. Prentice-Hall,
movement. In: Handbook of Hydrology. D.R. Maidment (Ed.). Englewood Cliffs, NJ, pp. 247–250.
McGraw-Hill, New York, pp. 5.21–5.23. 47. Bear, J. (1972). Dynamics of Fluids in Porous Media. Elsevier,
26. Kostiakov, A.N. (1932). On the dynamics of the coefficient of New York.
water-percolation in soils and on the necessity for studying 48. Springer, E.P. and Cundy, T.W. (1987). Field-scale evaluation
it from a dynamic point of view for purposes of amelioration. of infiltration parameters from soil texture for hydrologic
Trans. Sixth Comm. Intern. Soil. Sci. Soc. Russian, part A, analysis. Water Resour. Res. 23: 325–334.
17–21. 49. Berndtsson, R., and Larson, M. (1987). Spatial variability
27. Horton, R.E. (1940). An approach toward a physical inter- of infiltration in a semi-arid environment. J. Hydrol. 90:
pretation of infiltration-capacity. Soil Sci. Soc. Am. J. 5: 117–133.
399–417. 50. Sharma, M.L., Gander, G.A., and Hunt, G.C. (1980). Spatial
28. Green, W.H., and Ampt, G.A. (1911). Studies on soil physics: variability of infiltration in a watershed. J. Hydrol. 45:
1. Flow of air and water through soils. J. Agric. Sci. 4: 1–24. 101–122.
29. Phillip, J.R. (1957). The theory of infiltration: 1. The 51. Wood, E.F., Sivaplan, M., and Beven, K. (1990). Similarity
infiltration equation and its solution. Soil Sci. 83: 345–357. and scale in catchment storm response. Rev. Geophys. 28:
30. Ross, O.J. (1990). Efficient numerical methods for infiltration 1–18.
using Richards equation. Water Resour. Res. 26: 279–290. 52. Sposito, G. (1995). Recent advances associated with soil
31. Soil Conservation Service, Hydrology. (1972). In: SCS water in the unsaturated zone. In: Reviews of Geophysics,
National Engineering Handbook. U.S. Department of Agri- Supplement, U.S. National Report to International Union of
culture, Washington, DC, Sec. 4. Geodesy and Geophysics 1991–1994, pp. 1059–1065.
490 ION EXCHANGE AND INORGANIC ADSORPTION
ION EXCHANGE AND INORGANIC Table 1. Distinction Between Physical Adsorption and
Chemisorption (1)
ADSORPTION
Physical Adsorption Chemical Adsorption
K. KADIRVELU
Low heat of adsorption (<2 High heat of adsorption (> or 3
JYOTSNA GOEL
or 3 times latent heat of times latent heat of
Centre for Fire, Explosives, and evaporation.) evaporation.)
Environmental Safety Nonspecific. Highly specific.
Timarpur, India Monolayer or multilayer. Monolayer only.
No dissociation of adsorbed May involve dissociation. Possible
species. Only significant at over a wide range of
The presence of inorganic ions in water is of major relatively low temperature.
concern for water quality because they can interfere with temperatures.
legitimate use of water for various applications. Inorganic Rapid, nonactivated, Activated, may be slow and
ions, particularly heavy metals, are seriously questionable reversible. irreversible.
because of their toxicity and threat to human life and the No electron transfers, Electron transfer leading to bond
although polarization of formation between sorbate and
environment.
sorbate may occur. surface.
Ion exchange and adsorption share so many common
features in regard to removal of inorganic ions from water
that they can be grouped together as unified processes
for inorganic ion treatment. These processes involve the sulfonic acid groups; less commonly, groups are carboxylic,
increase in concentration of a particular component at the phosphonic, and phosphinic, and so some resins involve
surface or interface between two phases. quaternary ammonium groups (strongly basic) and amino
The objective of this article is to enumerate application groups (weakly basic).
of adsorption and ion exchange processes for inorganic Most ion exchangers in large-scale use are based on
ion removal from water. This article presents a review synthetic resins—either preformed and then chemically
of potential adsorbents and ion exchangers for inorganic reacted as styrene or formed from active monomers.
ion removal. It introduces important topics regarding Natural zeolites were the first ion exchangers, and natural
inorganic ion removal by ion exchange and adsorption and synthetic zeolites are in use today.
processes. Factors effecting metal removal are also Ion exchange may be thought of as a reversible reaction
presented and discussed in sections of the article. Research and of chemically equivalent quantities. A common
results from batch studies are presented to illustrate basic example of ion exchange is the familiar water-softening
concepts. Accordingly, emphasis is placed on providing reaction
a review of adsorbents and ion exchangers available for Ca++ + 2NaR ⇔ CaR2 + 2Na+ (1)
inorganic ion removal.
where R represents a stationary univalent anionic site in
ADSORPTION AND ION EXCHANGE: IN GENERAL the electrolyte network of the exchanger phase.
Adsorption involves, in general, the accumulation (or FACTORS AFFECTING INORGANIC IONS REMOVAL BY
depletion) of solute molecules at an interface (including ADSORPTION
gas-liquid interfaces, as in foam fractionation, and liquid-
liquid interfaces, as in detergency). Adsorbent surfaces Factors affecting inorganic ion removal by adsorbent
are often physically and/or chemically heterogeneous, and include (a) nature of the adsorbent (b) solution pH,
bonding energies may very widely from one site to another. (c) presence of foreign ions, and (d) temperature.
In discussing the fundamentals of adsorption, it is useful
to distinguish between physical adsorption, involving only
Nature of the Adsorbent
relatively weak intermolecular forces, and chemisorption,
which involves, essentially, the formation of a chemical Surface Area and Pore Structure. Surface area with
bond between the sorbate molecule and the surface of active functional groups is one of the principle charac-
the adsorbent. The general features, which distinguish teristics affecting the adsorption capacity of an adsorbent.
physical adsorption and chemisorption, are presented in The adsorption capacity of a solid adsorbent is generally
Table 1. proportional to the specific surface area of the mass adsor-
Ion exchange is an effective means of removing bent; that is, the adsorption of certain solutes increases
heavy metals from wastewaters. It is a reversible with an increasing surface area. Table 2 shows the physi-
chemical reaction, where the removal of heavy metals cal properties of a few adsorbents that have been already
is accomplished by the exchange of ions on the resin for employed for inorganic removal.
those in wastewater. In ion exchange, ions of positive However, the specific surface area alone is frequently
charges (cations) and negative charges (anions) from inadequate to explain the adsorption capacity of porous
the fluid (from aqueous solution) replace dissimilar solids, such as activated carbons, for different solutes. In
ions of the same charge in solids. The ion exchanger view of this, pore size distribution in activated carbon
contains permanently found functional groups of opposite indicates that microspores contribute a major portion of
charge-type. Cation exchangers generally contain bound the specific surface area.
ION EXCHANGE AND INORGANIC ADSORPTION 491
Table 2. Properties of Adsorbents Employed for Inorganic Ion Removal (2–6)

Calgon Coir-Pith AC from AC Carbon
Property CAL Carbon Arudo Donax Cloth Aerogel
Surface area (m2 /g) 1100–1300 575 1142 1689 700

Porosity (%) 65.0 93.11 76.0 66.5 36.2
Micropore Volume (g/ml) 0.12 0.0913 0.47 0.964 0.10
Bulk density (g/ml) 0.56 0.12 0.5 0.13 0.644
Moisture content (%) 2.0 5.88 9.3 2.0 18.0
Ash content (%) 7.0 7.87 21.56 13.0 3.5
Depending on their preparation, active carbons also carbon aerogel equilibrated with Pb (II). SEM of activated
contain larger pores, known as macro and mesopores in carbon surfaces before and after the adsorbent was
the classification proposed by Dubinin (7) and now adopted equilibrated with the metal ion solution clearly shows
by the International Union of Pure and Applied Chemistry the presence of Pb (II) ions.
(IUPAC) (8). The definition of the different types of pores As a result of these investigations, various researchers
is based on their width w, which represents the distance have postulated the existence of functional groups,
between the walls of a slit-shaped pore or the radius such as carboxyic, phenolic, lactonic, aldehydes, ketones,
of a cylindrical pore. One distinguishes micropores, for quinones, hydroquinone, anhydrides, and ethereal struc-
which w does not exceed 2 nm; macropores, for which w tures. Also, analysis of physicochemical properties of
exceeds about 50 nm; and mesopores of intermediate size adsorbent ions provides a deep insight into adsorption
(2.0–50 nm). mechanism involved during the adsorption process.
A linear relationship has also been demonstrated
between the specific surface area and sorbate molecular
Solution pH
size (9).
Extensive research investigations have been reported for
Chemistry of Adsorbent Surface. The nature and the adsorption of inorganic ions at different solution pH
presence of specific functional groups on the surface values (9–11). The dependence of adsorbent’s adsorption
of the adsorbent impart certain characteristics that capacity for inorganic ion removal from aqueous solution
affect the adsorption process. It is now known that pH has been attributed to changes in surface chemical
surface functional groups are formed during the carbon characteristics of adsorbent and species distribution
(adsorbent) activation process. of the inorganic ions varying with solution pH. In
Attempts have been made to identify and estimate the general, the cationic species removal will increase with
surface oxygen chemical structures (functional groups) increasing solution pH (Fig. 2), whereas in the case
using several physical and chemical techniques, which of inorganic ions present as anionic species, removal
include neutralization of acid and base (known as Boehm’s generally increases with decreasing pH, which is mainly
method); desorption of the oxide layer; potentiometer, because of electrostatic repulsion, which can be understood
thermometric, and radiometric titrations; direct analysis by removal of Cr (VI) by various adsorbents at lower
of oxide layers by specific chemical reaction; Infrared pH range.
(IR) spectroscopy; and X-ray photoelectron spectroscopy Specifically, pH affects the status of the outer hydration
(XPS). The surface morphology and metal ion distribution sheaths of the metal ion, metal speciation of inorganic ion
adsorbent samples can be visualized via scanning electron in aqueous medium, complexation and solubility, and the
microscopy (SEM). Figure 1 shows SEM micrograph for electrochemical behavior of the carbon surface.
(a) (b)
Figure 1. (a) SEM of carbon aerogel; (b) SEM of carbon aerogel loaded with lead.
98
96
% Removal of Pb (II)
94
92
90
91.8 xxx xxx

88 xxx xxx
10 0.1
8
6 0.08
Figure 2. Plot for percent removal of Pb (II) versus 4 Adsorbent dosage (g)
pH 2 0.06
pH and adsorbent dosage using carbon aerogel as
0 0.04
adsorbent.
Figure 3 depicts possible inorganic ion removal mecha- adsorption was hypothesized to be the primary removal
nisms for carbon aerogel in relationship to pH, at which mechanism. Also, it has been reported that metal removal
solution precipitation begins for Pb (II), Hg (II), and Ni by activated carbon was found to be inversely proportional
(II). Figure 3 shows clearly that at any pH, the removal to the solubility of the metal (13).
of metal ions is greater by adsorption in the presence of Namasivayam and Kadirvelu (14) reported an ion
carbon aerogel than by precipitation in the absence of exchange phenomenon for Hg (II) removal by the formation
carbon. A rapid increase in percentage removal of metal of the following surface complexes when Hg (II) is present
ions is observed after pH 5.4 and 7.5 for Pb (II) and Ni in the solution with the functional groups present on GAC
(II), respectively, because of precipitation in the absence (ACG-50).
of carbon aerogel.
The adsorption of metal ions depends on both the nature 2(CX OH+ ) + Hg2+ −−−→ (CX O)2 Hg2+ + 2H+ (2)
of the adsorbent surface and the species distribution of the
metal ions in the aqueous solution, which is, in turn, According to the surface complexation theory, Hg
governed by solution pH. Removal noticed in a pH range (II) removal increases as pH function increases can
of 3–5 for Pb (II) and Hg (II), because of their partial be explained on the basis of a decrease in adsorption
hydrolysis resulting in the formation of hydrolyzed metal competition between proton and metal ion species for
ion species [PbOH]+ , [HgOH]+ Pb (OH)2 , and Hg (OH)2 . the adsorbent surface sites by the decrease in positive
In the case of Ni (II), the maximum adsorption range surface charge, which results in a lesser columbic repulsion
was shifted to 4–7, which may be because of formation of forward of the Hg (II) ion.
[NiOH]+ and Ni (OH)2 at these pH (12). Low solubilities
of hydrolysis metal ion species may be another reason Presence of Foreign Ions
for maximum adsorption. At pH values below pHprec ,
The presence of complexing ligands and competing
adsorbates can alter metal removal from that observed
in the mono-component system. The degree of competition
120 is dependent on the type and concentration of competing
ions, number of surface sites, and affinity of the surface for
100
adsorbate and the ionic property of adsorbate. Presence
80 of other cocations can influence the adsorption capacity of
% Removal
Pb(II) Adsorption adsorbent with respect to one adsorbate. This phenomenon

60 Hg(II) Adsorption is illustrated in Fig. 4, where the adsorption equilibrium
Ni(II) Adsorption curve for Hg (II) removal by GAC is considerably effected
40 Pb(II) Precipitation by the presence of other metal ions like Cu (II) and
Hg(II) Precipitation Pb (II) (15,16). Similar results have been reported for
20 Ni(II) Precipitation the effect of EDTA on lead removal using GAC column
0 studies (17).
0 1 2 3 4 5 6 7 8 9 10 11 12
Initial pH
Temperature Effect
Figure 3. Effect of solution pH on the removal of lead, mercury, As the process of adsorption is spontaneous, it is
and nickel using carbon aerogel. accompanied by a decrease in the system’s free energy. A
20 Hg(II) Table 3. Thermodynamic Parameters for the Mercury

Hg(II) + Cu(II) Adsorption on Carbon Aerogel
Hg(II) + Pb(II) Temperature G0 H 0 S0
15
Hg(II) + Pb(II) + Cu(II) (K) % Removal (J mol−1 ) (Jmol−1 ) (Jmol−1 K −1 )
qe (mg g−1)
293 60 −987.714 +25708.82 +91.11444

10
303 68 −1898.86 +18990.88 +68.94302
313 73 −2588.29 +32921.47 +113.4497
5 323 80 −3722.79 +46023.63 +154.0137
333 87 −5262.92 +28133.9 +100.2908
343 90 −6265.83 — —
0
0 10 20 30 40 50 60 70
Ce (mg l−1) Raji et al. (20) have also presented alike thermodynamics
Figure 4. Equilibrium curve for Hg (II) ion in binary and tertiary for the uptake of copper and mercury, respectively.
system with Cu (II) and Pb (II) ions using GAC as adsorbent.
ADSORPTION MECHANISM—ION EXCHANGE PROCESS
decrease in entropy always exists because of loss of degrees With some of the adsorbents, like coir pith carbon,
of freedom of the solute in passing from the dissolved state adsorption mechanisms have been explained based on
to the adsorbate state. It follows from the relationship the ion exchange model (12). A pure carbon surface is
considered to be nonpolar, but in actual practice, some
G = H − TS (3) carbon–oxygen complexes Cx O, COx and Cx O2 are usually
present (21–23). The surface oxygen complexes hydrolyze
that the adsorption process must always be exothermic; water molecules as shown below:
that is H must have negative value regardless of the
nature of interaction forces. An increase in temperature, Cx O + 2H2 O −−−→ Cx OH2 2+ + 2OH− (8)
therefore, will result in a reduction of the equilibrium
adsorption capacity, whereas lower temperature will favor COx + xH2 O −−−→ C(OH)x + xOH− (9)
an increased capacity. 2 2+ −
Cx O + H2 O −−−→ Cx O + 2OH (10)
As reported by J. Goel et al. (6), according to the
adsorption isotherm, the amount of Hg (II) by carbon As the carbon is prepared with H2 SO4 , (NH4 )2 S2 O8 , and
aerogel increases with increasing temperature. Values of NaHCO3 , groups such as Cx ONa+ , Cx ONa2 2+ , Cx SO3 Na,
the thermodynamic parameters, such as G0 , H 0 , and and Cx ONH4 are also present. Na+ in the above groups
S0 , describing mercury uptake by carbon aerogel were also exchange with H+ in the medium as follows:
calculated using the following thermodynamic equations:
Cx ONa+ + H+ −−−→ Cx OH+ + Na+ (11)
G = −RT. ln K
0
(4)
Cx ONa+ + 2 H+ −−−→ Cx OH2 +2 + Na+ (12)
H 0 = [R × T1 × T2 /(T2 − T1 )] × Ln (K2 /K1 ) (5) 2+ + +2 +
Cx ONa2 + H −−−→ Cx OH2 + 2 Na (13)
S0 = (H0 − G0 )/T (6) + +
CxSO3 Na + H −−−→ Cx SO3 H + Na (14)
where R is the gas constant and K, K1 , and K2 are
Excess of Na+ was introduced into carbon where it was
the equilibrium constant at temperature T, T1 , and T2 ,
washed with NaHCO3 to neutralize any free sulfuric acid
respectively. The equilibrium constants was calculated
(H2 SO4 ) during the preparation of carbon. When metal
from:
ion is present in solution, its adsorption will free some
K = Ceq,s /Ceq,1 (7)
H+ and the pH increase will be lower than in the blank.
At the same time, Na+ will also be released according to
where Ceq,s and Ceq,1 were the equilibrium concentration
reactions as given below:
of Hg (II) in solution and on the adsorbent, respectively.
Values found for G0 , H 0 , and S0 , as presented in 2Cx OH+ + M2+ −−−→ (Cx O)2 M2+ + 2H+ (15)
Table 3, are indicative of the spontaneous nature of the
2+ 2+ 2 +
uptake process. CX OH2 +M −−−→ CX OM + 2H (16)
The positive H 0 value confirms the endothermic + 2+ 2+ +
2Cx ONa + M −−−→ (Cx O)2 M + 2Na (17)
nature of the sorption process. As diffusion is an
2+ 2+ 2 +
endothermic process, it would be expected that increased CX ONa2 +M −−−→ CX OM + 2Na (18)
solution temperature would result in increased uptake of + 2+ + +
2Cx SO3 H + M −−−→ (Cx SO3 )2 M + 2H (19)
Hg (II) ions from aqueous solution.
2+ +
Anoop Krishnan et al. (18) has reported that the 2Cx SO3 Na + M −−−→ (Cx SO3 )M + 2Na (20)
endothermic nature of the adsorption process, where
uptake of mercury on the bagasse piths activated carbon, is which indicates that apart from exchange of H+ ions on the
favored by increase in temperature. Panayotova (19) and adsorbent with M2+ ions, significant amount of Na+ ions,
which were present in the adsorbent, were also exchanged where R is the styrenedivinylbenzene copolymer. After
with M2+ ions. Hg (II) ions sorption, the ion exchanger Imac TMR
can be regenerated by means of the concentrated
hydrochloric acids.
ION EXCHANGERS FOR INORGANIC ION REMOVAL Metal ions and other cationic impurities are removed
on the cation exchanger, whereas chromates are removed
By ion exchange, undesirable ions are replaced by on the anion exchangers:
others, which do not contribute to contamination of the
environment. The method is technologically simple and nRSO3 H + M+ −−−→ nRMn+ + nH+ (23)
enables efficient removal of even traces of impurities
2ROM + CrO4 2− −−−→ R2 CrO4 + 2OH− (24)
from solutions. Examples of selective removal of heavy
metal ions by ion exchange are presented in Table 4. They R2 CrO4 + CrO4 2− + H+ −−−→ R2 Cr2 O7 + OH− (25)
include removal of Pb (II), Hg (II), Cd (II) Ni (II), V
(IV, V), Cr (III, VI), Cu (II), and Zn (II) from water and The NaOH solutions are applied for anion exchanger
industrial wastewaters by various means of modern type regeneration. Further developments of conventional ion
of ion exchangers. exchangers, like polystyrenesulphonate cation exchangers
By ion exchange, either all ions can be removed and strongly basic anion exchangers, were aimed at
from a solution or substances are separated. Therefore, large exchange capacity and better chemical, thermal,
selective removal of ionic contamination and complete and mechanical stability. Kinetic capability of ion
deionization can be distinguished. The choice between exchangers could be improved by means of macroporous
selective removal and complete deionization depends ion exchangers developed in the 1970s. The range of
mainly on the composition of solution and on the extent of application of the ion exchange method was remarkably
decontamination required. extended, first by invention and application of new organic
Hg (II) ions sorption on the ion exchanger Imac TMR and inorganic ion exchangers.
(Styrenedivinylbenzene Copolymer) can be described by
means of the following reactions:
ADSORBENTS FOR INORGANIC ION REMOVAL
2+ +
2R–SH + Hg −−−→ R–S–Hg–S–R + 2H (21)
Large numbers of adsorbents have been developed so
+ +
RSH + HgCl −−−→ R–S–HgCl + H (22) far for the removal of inorganic ions, which primarily
Table 4. Selective Removal of Heavy Metal Ions by Ion Exchangers

Metal Ion Ion Exchangers Composition and Functional Groups
Pb (II) Leawtit TP 207 (24) Chelating ion exchanger with

functional iminodiacetate group
Ethanotricarboxylate, Gel type Copolymer Of Vinyl benzene
1,1-ddicarboxylate-2- Chloride, Styrene and Divinyl
etanophosphonate (25) Benzene
Polyacrylamidoglycol Resin (26) Amide, OH and carboxylic groups
Clinoptilolite (27) Aluminosilicate structure
Hg (II) Imac TMR Styrenedivinylbenzene Copolymer
Varian BTAM (Nike), Varian BTKM, Containing nitrogen and sulphur
Varian BSM, Ionac SR-3 (28–31) functional group (22),
Thiocarbamate group (23),
Mercapathile group (24), Thiourea
functional group (25)
Dowex A-1 (32) Iminodiacetate group
Cd (II) Dowex 50 W-X4 Polystyrenesulphonic cation exchanger
Diphonix (33) Sulphonic groups and diphosphonic
groups
KB-2 and KB-2-12P (34) Polyacrylate carboxylic cation
exchangers
Phillipsite and Chabazite (35) Natural zeolite from volcanic rocks of
Central Italy
Ni (II) and V Wofatit CA-20 (36) Macro porous carboxylic cation
(IV, V) exchanger
Chelex −100 (37) Iminodiacaetae ion exchanger with
Na/H/Ni in exchange system
Sodium titnate, Na4 Ti9 O20 (38) Layered material with exchangeable
Na ions in the titanium oxide layer.
Cr (III, VI) Lewatit MP-500A (39) Basic anion exchanger
Amberlite IR-120 (40) Polystyrenesulphone cation exchanger
Table 5. Summary Table of Maximum Reported Adsorption Capacities (mg/g)

Hg Pb
Sorbent Source Cd (II) Cr (III) Cr (VI) (II) (II)
Bark (41–44) 32 19.45 400 182

CPEI cotton (45) 1000
Chitin Cr, Hg, Pb (46) 558 92 100 796
Chitosan (47, 48) 16.50 57 1123 58
Clay (49, 50) 28 116
Dead biomass (51) 43.0
Leaf mould (46–52) 150 1587
Lignin Cd, Cr, Hg and Pb (46) 87 17 632 1587
Modified wool (53) 46.65
Moss (54, 55) 5.058 76 43.9 16.2 135
Peat (54, 55)
Sawdust (56) 16.05 230
Seaweed (57, 58) 215 344
Xanthate (59, 60) 33.27 19.67 1.149 18
Coir pith carbon (3, 61) 94.0 154.0 264.0
Fly ash (62) 198 .073 444.0
Red mud (63) 67.0 7.1 165.0
Peanut hull carbon (64) 89.0 220.0 210.0
Furnace Sludge (65) 7.4 68.0
include activated carbon prepared from a variety of 6. Goel, J., Kadirvelu, K., Rajagopal, C., and Garg, V. K. (2004).
materials of biological origin, industrial wastes, dead Carbon. Elsevier Press, Amsterdam, In press.
biomass, organic wastes, etc. Besides, several low-cost 7. Dubinin, M. M. (1983). Carbon 19: 321.
adsorbents such as chitin, chitosin, clay, fly ash, peat, 8. IUPAC. (1972). Manual of symbols and terminology,
moss, zeolites, xanthates, seaweed/algae/alginate, lignin, Appendix 2, Pt.1, colloids and surface Chemistry. Pure and
and bark/tannin-rich material have also been explored Appl. Chem. 3: 578.
for heavy metal containing wastewaters (37–56). Table 5 9. Calgon Corporation. Basic Concept of Adsorption on Activated
presents some of the highest adsorption capacity reported Carbon. Calgon, Pittsburgh, PA.
for various low-cost adsorbents. 10. Kadirvelu, K., Thamaraiselvi, K., and Namasivayam, C.
Although a significant number of low-cost adsorbent (2001). Sep. and Purifi. Technol. 24: 497–505.
materials from sources, the commercial activated carbon 11. Ricou, P., Lecuyer, I., and Le Cloirec, P. (1998). Environ.
(CAC) is still being used intensively to date. Based Technol. 19: 1005–1062.
on its size and shape, activated carbon is classified as 12. Kadirvelu, K. (1998). Characterization of coir pith carbon
follows: Powder (PAC), granular (GAC), Fibre (ACF), and its utilization in the treatment of metal bearing
and cloth (ACC). As a result of the different sources wastewater. Ph.D. Thesis, Bharthiar University, Coimbatore,
of raw materials, the extent of chemical activation, India, pp. 59–60.
and the physicochemical characteristics, each type of 13. Sigworth, E.A. and Smith, S.B. (1972). J. AWWA 64:
activated carbon has its specific application with respect 386–391.
to physicochemical properties, as well as depending on 14. Namasivayam, C. and Kadirvelu, K. (1999). Carbon 37:
the inherent advantage and disadvantages in wastewater 243–256.
treatment. 15. Goel, J., Kadirvelu, K., and Rajagopal, C. (2004). Ads. Sci.
Tech. 22: 257–274.
Acknowledgment 16. Goel, J., Kadirvelu, K., and Rajagopal, C. (2004). Environ.
Authors are grateful to Sh. A. K. Kapoor, Director, Dr. Technol.
Chitra Rajagopal, Scientist ‘‘F’’, CFEES, DRDO, for granting 17. Reed, B.E. and Arunachalam, S. (1995). ASCE J. Env. Eng.
permission to write this article and their constant encouragement 121(9): 653–663.
and support.
18. Panayotova, M.I. (2001). Waste Mgmt. 21: 671–676.
19. Anoop Krishnan, K. and Anirudhan, T.S. (2002). J. Haz.
BIBLIOGRAPHY Mater. B92: 161–183.
20. Raji, C. and Anirudhan, T.S. (1996). Removal of Hg (II) from
1. Ruthven, D.M. (1984). John Wiley & Sons, New York. aqueous solution by sorption on polymerized sawdust. Ind. J.
2. Netzer, A. and Hughes, D.E. (1984). Water Res. 18: 927. Chem. Tech. 3: 49–54.
3. Kadirvelu, K. and Namasivayam, C. (2000). Environ. Tech- 21. Huang, C.P. and Ostovic, F.B. (1978). J. Env. Eng. Div. ASCI.
nol. 21: 1091–1097. 104: 863–878.
4. Basso, M. C., Cerella, E.G., and Cukierman, A.L. (2002). Ind. 22. Sharma, D.C. and Forster, C.F. (1994). Bioresource Technol.
Eng. Chem. Res. 41: 180–189. 47: 257–264.
5. Kadirvelu, K., Faur-Brasquet, C., and Le Cloirec, P. (2000). 23. Yenki, M. K.N. and Natarajan, G. S. (1991). Separation Sci.
Langmuir 16: 8404–8409. Technol. 26: 661–674.
496 IRON
24. Bayer, A.G. (2000). Leverkusen. 53. Low, K. S., Lee, C. K., Cheong, F.Y., and Lim, G. S. (1993). J.
25. Trochimczuk, W.A. (1999). Europ. Polym. J. 35: 1457–1464. Environ. Sci. Health A 28(8): 1689–1704.
26. Rivas, B.L., Maturana, H.A., Villegas, S., and Pereira, E. 54. Tong, P., Baba, Y., Adachi, Y., and Kawaza, K. (1991). Chem
(1998). Polym. Bull. 40: 721–728. Lett. 9: 1529–1532.
27. Petruzzelli, D., Pagano, M., and Tiravanti, G. (1995). Ind. 55. Kertman, S. V., Kertman, G. M., and Chibrikova, Z.H. (1993).
Eng. Chem. Res. 34: 2612–2617. J. Appl. Chem. USSR 66(2): 465–466.
28. Bolto, B.A., Pawlowski, L.E., and Spoon, F.N. Ltd. (1987). 56. Dikshit, V.P. (1989). Nat. Acad. Sci. Lett. 12(12): 419–421.
London. 57. Holan, Z.R., Volesky, B.N., and Prasetyo, I. (1993). Biotech.
29. Warshawsky, A. and Erlich-Rogozinski, S. (1977). Microm. J. Bioeng. 41(8): 819–825.
22: 362–368. 58. Leush, A., Holan, Z.R., and Volesky, B. (1995). J. Chem. Tech.
30. Dorfne, K. (1991). Walter de Gruyter, Berlin. Biotechnol. 62: 279–288.
31. Inoue, K. et al. (1993). Bunseki Kagaku. 42: 725–731. 59. Bricka, R.M. and Hill, D.O. (1989). Metal immobilization
by solidification of hydroxide and xanthate sludges. In:
32. Rengan, K. (1997). J. Radioanal. Nucl. Chem. 219: 211–215.
Environmental Aspects of Stabilisation and Solidifications
33. Hubicki, Z., Jakowicz, A., and Lodyga, A. Application of of Hazardous and Radioactive Wastes. P.L. Cote and
the ions from waters and sewages. In: Adsorption and T.M. Gilliam (Eds.). ASTM, Philadelphia, PA, pp. 257–272.
Environmental Protection, Studies in Surface Science and
60. Flynn, C. M., Jr., Carnahan, T.G., and Lindastrom, R.E.
Catalysis. A. Dabrowski (Ed.). Vol. 120. Elsevier, New York.
(1980). Adsorption of heavy metal ions by xanthated sawdust.
34. Kauspedience, D., Snukiskis, J., and Gefeniene, A. (1998). J. Report of Investigation #8427. United States. Bureau of
Radioanal. Nucl. Chem. 229: 129–132. Mines.
35. Colella, C., De Gennaro, M., Langella, A., and Pansini, M. 61. Kadirvelu, K. and Namasivayam, C. (2003). Advances in
(1993). Cadmium removal from wastewater using chabazide Environmental Research.
and phillipsite. In: Natural Zeolites 93. D.W. Ming and 62. Apak, R., Tutem, E., Hugul, M., and Hizal, J. (1998). Water
F.A. Mupton (Eds.). Brockport, NY, p. 377. Res. 32: 430–440.
36. Halle, K., Fischwasser, K., and Fenz, B. (1982). Technol. 63. Suemitsu, R., Venishi, R., Akashi, I., and Nakana, M. (1986).
Umweltschutz. 25: 1201032. J. Appl. Poly. Sci. 31: 75–83.
37. Leinonen, H. and Lehto, J. (2000). Rect. Funct. Polym. 43: 64. Periasamy, K. (1993). Adsorptive removal of heavy metals
1–6. from synthetic indisustrial wastewater using peanut hull
38. Lehto, J. et al. (1996). J. Radio Nucl. Chem. 208: 435–443. carbon. Ph.D. Thesis, Bharthiar University, Coimbatore,
39. Untea, I., Tudorache, E., and Florea, C. (2000). Revista de India.
Chimie. 51: 123–126. 65. Lopez-Delgado, C.P. and Girgis, B.S. (1998). Water Res. 32:
40. Drela, I. (1996). Rudy Met. Niezelaz. 41: 321–323. 1169–1177.
41. Randell, J. M., Bermann, R.L., Garrett, V., and Waiss, A.C.
(1974). J. Foorest Prod. J. 24(9): 80–84.
42. Teles de Vasconcelos, L.A. and Gonzalez Beca, C. G. (1994). IRON
Eur. Water Poll. Control 4(1): 41–51.
WILLIAM R. ROY
43. Masri, M. S. and Friedman, M. (1974). J. Appl. Polymer Sci.
Illinois State Geological Survey
18: 2367–2377.
Champaign, Illinois
44. Randall, J. M., Hautala, E., and Waiss, A.C., Jr. (1974).
Removal and recycling of heavy metals ions from mining
and industrial waste water streams with agricultural
byproducts. Proceedings of the Fourth Mineral Waste
CHEMISTRY OF IRON
Utilization Symposium, Chicago, IL, pp. 329–334.
45. Roberts, E.J. and Rowland, S.P. (1973). Env. Sc. Technol. The average concentration of iron in the earth’s crust
7(6): 552–555. is about 5%, and it occurs in a wide variety of oxide,
46. Masri, M. S., Reuter, F.W., and Friedman, M. (1974). J. Appl. carbonate, silicate, and sulfide compounds either as
Poly. Sc. 18: 675–681. metallic ores or in rock-forming minerals. Iron also occurs
47. Prades, E.G. et al. (1994). J. Chem. Tech. Biotechnol. 59: naturally in water from the dissolution of iron-containing
289–295. minerals. Iron in water can occur in a dissolved form,
48. Cadena, F., Rizvi, R., and Peters, R.W. (1990). Feasibility a colloidal form that is combined with organic matter,
studies for the removal of heavy metals from solution and as suspended, iron-containing particles. Dissolved
using tailored bentonite. In hazardous and industrial wastes. iron occurs in water in two oxidation states: ferrous
Proceedings of the 22nd Mid-Atlantic Industrial Waste iron, Fe (II), and ferric iron, Fe (III). In solution, iron
Conference. Drexel University, Philadelphia, PA, pp. 77–94. can occur as a cation (Fe2+ ) and in combination with
49. Matis, K.A. and Zoublious, A.I. (1994). Biotech. Bioeng. 44: other ions to form complexes and neutral ion pairs. Iron
354–360. does not generally occur as the trivalent cation (Fe3+ )
50. Niu, H., Xu, X.S., and Wang, J. H. (1993). Biotech. Bioeng. 42: in significant concentrations in water. Trivalent iron
785–787. generally exists in combination with other constituents.
51. Sharma, D.C. and Forster, C.F. (1994). Biores. Technol. 49: Depending on the occurrence of anions, such as sulfate,
31–40. carbonate, and phosphate, and the pH and reduction-
52. Srivastava, S. K., Singh, A.K., and Sharma, A. (1994). Envi- oxidation potential (Eh ), the major iron species found
ron. Technol. 15: 353–361. in natural waters include a sulfate complex (FeSO4 0 ),
IRON 497
16 recommended dietary allowance for iron in healthy

Fe(OH)3
14 adults is 10 mg per day for men and 15 mg per day
for premenopausal women (2). Dissolved iron in potable
12
Fe2+
water is more of a nuisance than a potential heath
Iron, mg/L
10 issue. It can impart an unpleasant metallic taste to

8 drinking water and alter the taste of food cooked in
6 the water. The taste threshold for ferrous iron is about
Fe(OH)2+ 0.1 mg/L. Concentrations exceeding 0.3 mg/L of iron can
4
stain plumbing and clothing (3). The secondary maximum
2 contaminant level for public water supplies is 0.3 mg/L,
0 and 1.0 mg/L for freshwater aquatic life (4). However,
−300 −200 −100 0 100 200 300 400 500 600 700 bioassays by Roy et al. (5) suggested that concentrations
Eh, mV less than 1.0 mg/L may be toxic to freshwater aquatic life.
Figure 1. The distribution of 10 mg/L of iron in water at 15 ◦ C
and pH 7.0 as a function of oxidation-reduction potential (Eh).
LEVELS OF IRON IN WATER
a carbonate form [Fe(HCO3 )2 ], phosphate complexes The amount of dissolved iron in water depends on pH
(FeHPO4 0 , FeH2 PO4 + ), and various hydroxide species and Eh, but concentrations of less than 0.01 to 10 mg/L
such as FeOH+ , Fe(OH)2 + , Fe(OH)3 0 , Fe(OH)4 − , and are typical for natural waters (6). Well-aerated surface
FeOH2+ . The relative proportion of ferrous iron and ferric water usually contains less than 0.05 mg iron/L. Iron
iron depends on the Eh of the water; ferrous iron is the concentrations in aquifers in Illinois have ranged from
dominant form in oxygen-poor water (reducing conditions). <0.05–21 mg/L (7). Iron in acid-mine drainage from coal
In oxygen-rich water (oxidizing conditions), ferric forms mines has ranged from 10–5,625 mg/L (5,6).
of iron dominate. For example, in a solution containing
10 mg/L of iron [Fig. 1], about 98% of the iron would occur
SOLUBILITY OF IRON IN WATER
as Fe2+ in a relatively reducing system. At an Eh greater
(i.e., more oxidizing) than about 200 mV, ferrous iron is
The amount of dissolved iron in water can be limited
converted to the ferric forms Fe(OH)3 0 and Fe(OH)2 + . At
by the solubility of iron-containing solid phases. Schwab
an Eh of 700 mV, about 96% of the iron occurs in ferric
and Lindsay (8) concluded that when pH + pE (pE is
forms. When a groundwater sample is collected, it may
related to the redox potential Eh by pE = (EhF)/(2.3RT)
initially appear clear. However, if the sample is allowed to
react with air and oxidize, the ferrous iron in the initially where R = gas constant, T = absolute temperature, and
clear, reduced sample will be converted to ferric forms: F = Faraday constant) is less than 8.5, the concentration
of ferrous iron in groundwater is limited by the solubility
4Fe2+ + O2 + 8OH− + 2H2 O
4Fe(OH)3 (1) of siderite, FeCO3 :
As a consequence, the sample may develop a reddish- Fe2+ + CO3 2−

FeCO3 (2)
brown color, and iron-containing particles may settle at
the bottom of the container. The rate at which ferrous iron When pE + pH is greater than 8.5, the solubility of ferrosic
is oxidized depends on temperature, pH, and the amount hydroxide, Fe3 (OH)8 , limits the amount of ferrous iron:
of dissolved oxygen and iron.
+
Microorganisms can influence the oxidation state and 2Fe3+ + Fe2+ + 8H2 O
Fe3 (OH)8 + 8H (3)
distribution of iron in water. Iron can be used as an
energy source via oxidation by microbial activity or When the concentration of iron is greater than the
reduced because iron can act as an electron acceptor. For solubility product of either siderite or ferrosic hydroxide,
example, Crenothrix and Leptothrix are diverse groups the iron precipitates, from solution. Ferrosic hydroxide
of iron-precipitating bacteria that are widely distributed is thermodynamically unstable, and other minerals may
in freshwater and soil. After ferrous iron is oxidized to form, causing ferrosic hydroxide to dissolve. For example,
the sparingly soluble ferric form, it may precipitate as amorphous ferric hydroxide, [Fe(OH)3,a ] and goethite (α-
ferric hydroxide. The bacterial cell sheaths can become FeOOH) are less soluble than ferrosic hydroxide and given
coated with the precipitate, and the resulting gelatinous time, may control iron solubility. Moreover, amorphous
mass can clog water pipes and drains. There are bacteria ferric hydroxide can dehydrate (9) to form the more stable
and fungi that can reduce ferric iron. However, much mineral hematite (Fe2 O3 ):
less is known about iron-reducing microorganisms (1).
The Geobacter are a group of bacteria that can reduce Fe2 O3 + 3H2 O
2 Fe(OH)3,a (4)
ferrous iron.
The prevalent form of iron in surface water in the
HEALTH AND REGULATORY ISSUES iron–carbon dioxide–water system can be illustrated by
constructing a pH–pE stability diagram [Fig. 2]. This
Iron is not toxic in human consumption. Iron is an type of diagram illustrates which forms of iron are
essential element for the formation of hemoglobin. The most thermodynamically stable at various values of pH
498 IRON
+20
Fe3+ Up
per
Fe(OH)2+ lim
it o
fw
ate
r st
abi
lity
+10
amorphous
Fe2+ Fe(OH)3
pE
Fe(OH) −4
Low
er l FeCO3
imi
t of
wa
ter st
−10 abi
lity
Fe(OH)2
Figure 2. pE–pH diagram for the iron–

carbon dioxide–water system at 25 ◦ C 0 3 6 9 12 15
and 1 atm (11). pH
and pE rather than at specific concentrations. Ferrous (12). In the AAS method, a water sample is aspirated into
iron is the stable form in equilibrium with relatively an air-acetylene flame and atomized. A monochromatic
acidic and reducing waters, whereas ferric iron, as both beam of light at a wavelength corresponding to iron is
Fe3+ and FeOH2+ , is in equilibrium with acidic and passed through the flame. The amount of light energy
oxidizing waters. In relatively less oxidizing and alkaline sorbed is proportional to the amount of iron present.
waters, the most thermodynamically stable forms may When analyzed at a wavelength of 248.3 nm, the optimal
be amorphous iron hydroxide, siderite (FeCO3 ), or solid concentration range for iron is about 0.3 to 10 mg/L.
ferrous hydroxide. Dissolved iron can be determined by argon-supported
The concentration of iron in water may also be influ- inductively coupled plasma (ICP) spectroscopy. In ICP,
enced by sorption–desorption interactions with mineral the liquid sample is aspirated into a high-temperature
surfaces. Iron oxides and hydroxides have been studied (6000–8000 K) flame, and atomic emission is measured
as sorbents for other metals, but ionic iron is sub- to determine the concentration of the iron present. The
ject to the same electrostatic interactions that attract concentration of iron is proportional to the intensity of
cations to clays and organic matter. Ferric iron can be the emissions. Iron has many such emission lines in its
sorbed by silica, clay minerals, and iron and manganese
spectrum, but the wavelength of 259.94 nm is commonly
oxides (10). The sorption of ferrous iron has not been
used for concentrations ranging from about 0.007 to
studied in detail.
100 mg/L.
Dissolved iron also can be determined by the phenan-
METHODS FOR DETERMINING IRON throline method. In this method, iron in solution is reduced
to the ferrous form, and then allowed to react with 1, 10-
The amount of iron dissolved in water can be determined phenanthroline to form a reddish-orange product. The
quantitatively by atomic absorption spectroscopy (AAS), absorbance of the complex is measured in the range of
atomic emission spectroscopy, and by colorimetric methods 460–520 nm. Both AAS and ICP methods determine total
ISOTOPES 499
iron, but the amount of ferric and ferrous iron can be ISOTOPES
determined separately by the phenanthroline method (13).
ALDO CONTI
METHODS FOR REMOVING IRON FROM WATER Frascati (RM), Italy
The presence of large concentrations of dissolved iron is

a common water-quality problem, especially for ground- Isotopes are atoms with the same atomic number, Z, but
water (1). Methods for removing iron are summarized in different atomic weight, A. Different isotopes, therefore,
Lehr et al. (3) and Vigneswaran and Visvanathan (2). The have the same chemical but different physical properties,
oxidation of soluble, ferrous iron into relatively insoluble such as melting and boiling point. The name isotope means
ferric iron, followed by filtration of the suspended mass ‘‘at the same place’’ and comes from the periodic table,
is common to many of the methods. Depending on its where different isotopes of the same element are at the
concentration in the untreated water, the oxidation of same place. The scientific way of labeling isotopes reflects
ferric iron can be promoted by aeration, catalytic oxida- this. They have the symbol of the corresponding element,
tion with manganese oxides, and chlorination. Household but followed by a hyphen and the number of nucleons
water softeners are designed to remove calcium and mag- (neutrons and protons) of the isotope. So, C-12 is the
nesium from hard water using an ion exchange resin, but normal carbon (six protons and six neutrons) and C-14 is
they can also remove dissolved iron. Lime softening of a heavy unstable isotope (six protons and eight neutrons).
municipal water supplies can also promote the precipita- Isotopes of the same element share the same chemical
tion of ferric iron. The presence of iron-fixing bacteria may properties because they depend only on the number of
necessitate the use of relatively large amounts of bleach protons, Z. But different isotopes can have very different
(‘‘shot chlorination’’) to reduce the amount of iron in the weights, which is particularly evident in hydrogen. Normal
final product. hydrogen has only one proton and no neutrons. Deuterium,
or H-2, has one neutron and its weight is nearly twice that
BIBLIOGRAPHY of hydrogen. Then, tritium, H-3, has two neutrons and
is three times as heavy, which is the reason why, when
1. Chapelle, F.H. (2001). Ground-Water Microbiology and Geo-
deuterium combines with oxygen, the resulting molecule is
chemistry. John Wiley & Sons, New York.
called heavy water. Due to the extra weight of deuterium,
2. Vigneswaran, S. and Visvanathan, C. (1995). Water Treat-
heavy water boils at 101.42 ◦ C (214.56 ◦ F) and freezes at
ment Processes. Simple Options. CRC Press, Boca Raton,
FL.
3.81 ◦ C (38.86 ◦ F), at standard pressure. The difference in
weight affects the speed of chemical reactions involving
3. Lehr, J.H., Gass, T.E., Pettyjohn, W.A., and DeMarre, J.
(1980). Domestic Water Treatment. McGraw-Hill, New York, water. For this reason, heavy water is not really safe to
pp. 172–178. drink. Experiments performed with mice have shown that
4. U.S. EPA. (1999). National Recommended Water Quality large amounts of heavy water in the body can inhibit
Criteria–Correction. EPA 822-Z-99-001. mitosis, which affects rapidly dividing tissues, such as the
5. Roy, W.R., Skelly, T.M., Krapac, I.G., Griffin, R.A., Dickerson, stomach lining.
D.R., Schuller, R.M., Suloway, J.J., and Shimp, N.F. (1984). Isotopes do not differ only in physical properties. The
Geochemical Properties and Toxicological Effects on Aquatic extra neutrons make some of the nuclei highly unstable,
Life of Coal Wastes. Environmental Geology 107, Illinois State which is the reason that some isotopes are radioactive.
Geological Survey, Champaign, IL, p. 124. A typical example is uranium. Its isotopes, such as U-
6. Patterson, J.W. (1975). Wastewater Treatment Technology. 235, decay radioactively following a long chain, which
Ann Arbor Science, Ann Arbor, MI. leads to lead. Some of this decay involves a neutron
7. Warner, K.L. (2000). Analysis of Nutrients, Selected Inor- becoming a proton by emitting an electron. In this case, the
ganic Constituents, and Trace Elements in Water from Illinois nucleus becomes a new element, with the atomic number
Community-Supply Wells, 1984–91. Water-Resources Inves- increased by one. The energy emitted during the decay
tigations Report 99-4152, U.S. Geological Survey.
can be harvested in nuclear plants or used destructively
8. Schwab, A.P. and Lindsay, W.L. (1983). Effect of redox and in atomic bombs.
pH on the solubility and availability of iron. Soil Sci. Soc. Am.
The instability of their nuclei makes isotopes useful for
J. 47:201–205.
many applications. A famous one is carbon dating. Carbon
9. Garrels, R.M. and Christ, C.L. (1965). Solutions, Minerals,
comes mainly in two isotopes, the normal C-12 and the
and Equilibria. Freeman Cooper, San Francisco, CA.
heavy, unstable, C-14. When an organism grows, it uses
10. Anderson, M.A. and Rubin, A.J. (1981). Adsorption of Inor-
them in the same proportion in which they are found in the
ganics at Solid–Liquid Interfaces. Ann Arbor Science, Ann
Arbor, MI. environment. After the death of the organism, the ratio
11. Stumm, W. and Morgan, J.J. (1981). Aquatic Chemistry. John
changes continuously, as the heavy C-14 decays. C-14 has
Wiley & Sons, New York. a half life of 5730 years, which means that after this period
12. American Public Health Association. (1992). Standard half of it has gone. By this method, it is possible to date
Methods for the Examination of Water and Wastewater, 18th any biological remains or artifact up to about 50,000 years
Edn. Washington, DC, Section 3500. Iron, 3-65–3-68. old. For older objects, the amount of C-14 becomes too
13. Stucki, J.W. (1981). The quantitative assay of minerals small to be measured reliably. Isotopes are often used as
for Fe2+ and Fe3+ using 1, 10-phenanthroline: II. A labels. By replacing a stable atom in a molecule with an
photochemical method. Soil Sci. Soc. Am. J. 45: 638–641. unstable radioactive isotope, it becomes possible to follow
500 ISOTOPE FRACTIONATION
the chemical reactions the molecule undergoes. Our ability researchers use ‘‘enrichment factor (Ef)’’ in an equation
to detect radioactive isotopes makes them very useful in slightly different from the fractionation factor to deduce
the fields of medicine and biology, where they are used source/s of solutes in atmospheric precipitation (3,4). If
as tracers. the Ef is greater than unity, the ion in question (in the
sample) is derived partly or wholly from a source other
than oceanic origin. If it is less than unity, the ion could
ISOTOPE FRACTIONATION have been derived wholly from oceanic sources:
GHOLAM A. KAZEMI (Ion/Cl)sample

Shahrood University of Ef = ||
(Ion/Cl)seawater
Technology
Shahrood, Iran
In this example, the enrichment factor equals 0.91.
(0.5)
INTRODUCTION Ef = = 0.91
(0.55)
The following paragraph extracted from the major
textbook by Bowen (1) describes fractionation as it is dealt DEFINITION
with in general chemistry. The subsequent paragraph
defines fractionation in general hydrochemistry. Then, Isotope fractionation is defined by Domenico and
isotope fractionation is discussed. Shwartz (5) as a change in an isotopic ratio (for example,
In general chemistry, fractionation is the separation of 18
O/16 O) because of a chemical reaction (for example, water
a mixture, usually of chemically related components, into changes to water vapor). It is also defined by Coplen (6) as
fractions that possess different properties. A mixture of the fractionation or partitioning of isotopes by physical or
liquids having different boiling points, may be separated chemical processes which is proportional to the differences
by collecting fractions that boil at different temperatures in their masses. Coplen also classifies isotopic fractiona-
in a fractionating column. This is a long, vertical column tion processes into two categories, chemical and physical
containing plates or bubble caps attached to a still. As isotopic fractionation processes: physical processes such
a consequence of internal reflux, a gradual separation as ultrafiltration or gaseous diffusion of ions or molecules.
occurs between high and low boiling point fractions of the Chemical processes are those that lead to redistribution
liquid mixture. of the isotopes of an element among phases or chemical
In hydrochemistry, the fractionation factor is a ratio species and are divided into two broad categories equilib-
which shows the depletion or addition of ions in rainwater rium isotopic reactions and kinetic isotopic reactions (6).
compared to seawater (2). In other words, the fractionation Gilreath (7) in a simple yet valuable text describes the
factor is the ratio of two ions in rainwater to the ratio of fundamentals of isotope fractionation in detail and refers
the same ions in seawater. For example, for Cl− relative to it as ‘‘isotope separation.’’ Anyone intending to deal
to Na+ , the fractionation factor FNa is (2): with isotope fractionation is advised to consult this text
first. The point of primary importance is that fraction-
(Cl/Na)rainwater
FNa = ation of isotopes is important in the geologic environ-
(Cl/Na)seawater ment only for elements of low atomic number. Bowen (1)
describes three main processes which cause fraction-
In this equation, the units of concentration can be mg/l,
ation: (1) isotope exchange reactions; (2) unidirectional
mmol/l, meq/l, etc.
reactions; and (3) physical processes such as evapora-
Example: What is the fractionation factor of Na
tion, condensation, crystallization, melting, adsorption,
for a rainwater sample whose chloride and sodium
desorption, and diffusion. Isotopes may also be fraction-
concentrations are 1 and 0.5 mg/l, respectively?
ated in the laboratory by a variety of means (URL Ref.
http://www.c14dating.com/frac.html). One method that
Cl/Narainwater sample = 1/0.5 = 2
may result in considerable fractionation during isotopic
Cl/Naseawater = 1.81 analysis is the recently introduced technique of secondary-
ion mass spectrometry (SIMS) (8).
(2)
FNa = = 1.11 The following are the factors that control the rate of
(1.81)
isotope fractionation:
A fractionation factor greater than unity indicates
that the rainwater sample is depleted (in comparison to 1. Fractionation is minimized when reactions are
seawater) in the ion in question. If it is less than unity, it fast (9).
shows the enrichment (in comparison to seawater) of the 2. The higher the mass difference between various
sample in the ion. Appelo and Postma (2) also propose the isotopes of the element, the more effective the
fractionation factor as a quantitative approach to find the fractionation process. Deuterium, for example, has
activity of a single species that forms part of a given the largest mass difference from common hydrogen
total concentration, for example, for HCO3 , mHCO3 − = compared to isotopes of other elements. Hence, it is
α × CO2 , where α is the fractionation factor. Some subject to greater fractionation.
ISOTOPE FRACTIONATION 501
3. Isotope fractionation decreases as temperature and Ottonello (13); the last is a thorough mathematical
increases—at finite temperatures, all isotopic treatment of the topic. Here, in the Encyclopedia of Water,
species are well mixed (6). the discussion of isotope fractionation is limited to the iso-
4. Isotopic exchange between minerals and water in topes of hydrogen and oxygen, specifically, the atoms that
deep, basinal flow systems or in geothermal systems form the water molecule, as well as other isotopes that
induced by high temperature helps the fractionation are of frequent use in water related/hydrologic studies.
process to proceed (5). The latter group includes isotopes of the elements C, N,
5. In evaporation from a pond, fractionation is more Cl, S, Sr, and less used U, Th, and Si. These are, often,
efficient if the vapor produced is constantly removed, referred to as environmental isotopes (6). Some workers
for example, by wind (10). such as Haldorsen et al. (11) classify the environmental
isotopes into two groups: hydrologic isotopes (O, D, T)
Fractionation Factor and biochemical isotopes (C, N, S). Although this classi-
fication covers only a limited number of environmental
The fractionation factor is the ratio of isotopes (for isotopes, it is sound for fractionation. For the first group,
example, 18 O/16 O) in a liquid phase to the same ratio the main fractionating processes are physical or chemi-
in the vapor phase; cal reactions, biological/bacteriological/microbial processes
are dominant in the fractionation of the second group.
(18 O/16 O)liquid Before discussing all environmental isotopes one by one,
α= (11)
(18 O/16 O)vapor it should be pointed out that in some cases, considerable
isotope fractionation has led to the usefulness of those
or more clearly/generally defined by Haldorsen et al. (11) isotopes in hydrologic studies, for example, H and O
‘‘as the increase or decrease of a given isotope in a isotopes (enrichment and depletion processes are the
substance that undergoes a phase change or simply in fundamentals for studying lake–groundwater interaction,
terms of delta notation: recharge temperature, and many other applications). In
contrast, in some other cases, lack of isotope fractionation
1000 + δ 18 Oliquid has been helpful. Strontium isotopes, for example, are
α=
1000 + δ 18 Ovapor used to study the origin of solutes in the streams because
they are not subjected to substantial fractionation.
Example: In the condensation of water vapor to liquid,
typical values of δ 18 O for the liquid and vapor are, −5% and H and O
−14%, respectively. Therefore, the fractionation factor is,
Hydrogen and oxygen atoms have three isotopes each
1000 + (−5) 995 (see also Heavy Water and Deuterium articles): 1 H, 2 H,
α= = = 1.0092 3
1000 + (−14) 986 H, and 16 O, 17 O, and 18 O. Water molecules containing
the heavy isotopes are less likely to evaporate and more
Each fractionating process results in a specific frac- likely to condense than those containing lighter isotopes.
tionation factor. For example, the oxygen fractionation Hence, when a body of water evaporates, the vapor
factor in freezing differs from that from condensation. The phase is depleted and the residue, the liquid phase, is
former and the latter are 1.0111 and 1.003, respectively. enriched in heavy isotopes, 2 H, 3 H, 17 O, and 18 O. During
This means that freezing is more effective than conden- melting/freezing of water, heavy isotopes are slightly
sation for isotope fractionation; freezing leads to stronger more inclined to freeze and less inclined to melt than
fractionation than condensation. lighter isotopes, making the melt water depleted in 18 O
Fractionation factors approach unity as temperature and deuterium compared to the snowpack. Ultrafiltration
increases; when α = 1, the isotope ratios become equal (passage of water through shale micropores) enriches
in both phases being considered, and hence no further the residue (not passed through water) in O and H
isotopic fractionation can occur (1). Higher fractionation isotopes (6), but transpiration is not a fractionating
factors mean more effective fractionation processes. process. The principles of hydrogen isotope fractionation
are the scientific foundation for producing heavy water,
Isotopic Fractionation Curve which is used in the nuclear industry.
The fractionation factor for deuterium is greater
A curve defines the equilibrium isotope fractionation
than that for oxygen-18 because of a larger difference
between coexisting phases as a function of temperature (1).
between deuterium and hydrogen masses compared to the
It usually takes the form of an exponentially decreasing
difference between 16 O and 18 O masses. The fractionation
curve; there is a negative correlation between isotope
factors of these two elements at various temperatures are
fractionation and temperature. The shape of this curve is
calculated via the following equations (1):
different for different isotopes.
For oxygen,
Fractionation of Various Environmental Isotopes
1000 ln (α 18
O) = 2.644 − 3.026 × 103 /T + 1.534 × 106 /T 2
As can be inferred from the information provided so
far, isotope fractionation is an extensive topic in chem- For hydrogen,
istry/geochemistry and has been dealt with in detail in
many references such as Bowen (1), Gat (12), Gilreath (7), ln (α D) = 0.0771 + 13.436 × 103 /T 2
502 ISOTOPE FRACTIONATION
(note that the temperature is in K). S

Sulfur in near-surface environments frequently changes
Example: between the oxidized and the reduced forms, so it acquires
The fractionation factor for 18 O at 293.15 K (20 ◦ C) is a wide variety of 34 S/32 S ratios. Isotope fractionation
occurs in many processes of the sulfur cycle. The principal
ln (α 18
O) = (2.644 − 3.026 × 103 /293.15 + 1.534 reaction leading to the fractionation of sulfur isotopes
× 106 /293.152 )/1000 = 1.01000 in nature is the reduction of sulfate ion by anaerobic
bacteria (13). Bacterial reactions are important because
they are kinetic, fast, favor the lighter isotope, and result
The fractionation factor for 18 O during calcite precipitation
in depleting 34 S in natural biogenic H2 S and hence enrich
from water is 1.0349, which means that calcite is 35%
the remaining sulfate (11). Brucher et al. (14) studied
more enriched in 18 O relative to the water from which it
isotope fractionation during bacterial sulfate reduction
precipitates. However, not all precipitated calcite is richer
and found that in addition to the variations in the rates
in 18 O compared to ‘‘parent water.’’ For instance, if calcite
of bacterial sulfate reduction, genetic and physiological
precipitation is due to freezing, the 18 O content of calcite
differences between various genera of sulfate-reduction
is lower (11).
bacteria strongly influence isotope fractionation.
In the precipitation of sulfide minerals from solution,
C the fractionation and enrichment of 34 S depend on the
Oxidation–reduction processes in which the heavier relative bond strengths and the temperature of solu-
isotope is concentrated in the more oxidized forms are tion (13). Sulfide minerals in newly deposited sediment
important in carbon fractionation. The most influential can be enriched in 32 S by about 50% compared with
factor in carbon isotope fractionation is selective uptake associated marine sulfate (1). The equilibrium isotope frac-
of lighter CO2 by plants. During photosynthesis, plants tionation factor between SO4 2− and H2 S is 1.074 and is
discriminate against the heavier isotopes of carbon; the 1.020 between water and dissolved sulfate at 88 ◦ C (6).
isotope 13 C in organic matter (soil) is depleted by 1.8% During dissolution of evaporites, the largest reservoir of
compared to its natural ratios in the atmosphere. It is sulfate, no sulfur fractionation takes place.
believed that all organisms discriminate against 14 C about
twice as much as against 13 C and the ratio between Sr
the stable 12 C and 13 C atoms can be used to correct Because of its relatively high mass, no detectable isotope
for the initial depletion of 14 C. The amount of isotope fractionation accompanies precipitation of Sr within
fractionation depends on the photosynthetic pathway used minerals (6).
by the plant. Plants are divided into three categories
in this regard (11). At high temperatures, equilibrium
fractionation of isotopes may occur between CH4 and CO2 , ARTIFICIAL ISOTOPE FRACTIONATION FOR
and this is sensitive to changes in pH, oxygen fugacity, and ‘‘SEPARATION OF ISOTOPES’’
ionic strength as well as to changes in temperature (13).
As an example, the fractionation factor for 13 C in CaCO3 The most interesting and applied part of the isotope
in equilibrium with CO2 is 1.01074. fractionation topic is artificial separation (‘‘fractionation’’)
of isotopes, a subject that is scientifically simple, but, in
Cl engineering terms very complicated. Historically, the first
successful attempt to separate isotopes was that by Aston
The author has not come across a remark in the (the inventor of the mass spectrograph and the Nobel prize
literature about the occurrence of fractionation in chlorine winner in 1922) who separated neon into two fractions of
isotopes. This may have happened because chloride is different atomic weights by repeated diffusion. The topic
a highly mobile, soluble, and geochemically conservative gained importance with the discovery, in 1939, that 235 U
(is not removed from solution by mineral interaction or is a fissionable material and in succeeding years that
secondary mineral formation) element and has a relatively uranium-235 is the key material for producing nuclear
high mass. weapons and power (7).
N
VARIOUS TECHNIQUES FOR SEPARATING ISOTOPES
Fractionation of nitrogen isotopes occurs mostly through ARTIFICIALLY
biological processes and some isotope exchange reactions.
Kinetic isotope fractionation often dominates nitrogen As stated before, separating isotopes is an intensive
reactions and because of this, during denitrification, complicated engineering task, and only a few countries
δ 15 N values of residual nitrate increase exponentially have the capacity. Four principles that constitute the
as concentrations decrease (6). Bowen (1) reports that foundation of isotope separation should be considered
conversion of ammonia to nitrate (nitrification) by when devising a separation process (7):
nitrifying organisms produces isotope fractionation up to
20% and even greater values (up to 30%) are associated 1. At a given temperature, gaseous molecules, regard-
with denitrification. less of mass, have the same average kinetic energy.
MARIOTTE BOTTLE—USE IN HYDROLOGY 503
2. The gravitational or centrifugal forces acting on 7. Gilreath, E.S. (1985). Fundamental Concepts of Inorganic
a particle of matter are directly proportional to Chemistry. 19th printing, International Student Edition.
its mass. McGraw-Hill, Singapore.
3. The electric or magnetic forces affecting ionized 8. MacRae, N.D. (1995). Secondary-ion mass spectrometry and
atoms or molecules follow Coulomb’s laws for geology. Can. Mineralogist 33: 219–236.
these forces. 9. Langmuir, D. (1997). Aqueous Environmental Geochemistry.
Prentice-Hall, Upper Saddle River, NJ.
4. Among the lighter elements, isotopic species may
10. Mazor, E. (1991). Applied Chemical and Isotopic Groundwa-
vary slightly in their chemical reactivities.
ter Hydrology. Open University Press, Milton Keynes.
Considering these four principles, a variety of techniques 11. Haldorsen, S., Riise, G., Swanson, B., and Sletten, R.S. (1997)
Environmental isotopes as tracers in catchments. In: Geo-
have been developed for isotopic enrichment (7). Each of
chemical Processes, Weathering and Groundwater Recharge in
these techniques has advantages and disadvantages and
Catchments. O.M. Saether and P.D. Caritat (Eds.). Balkema,
is suitable for specific types of conditions and isotope(s). Rotterdam, pp. 185–210.
12. Gat, J.R. (1981). Isotopic fractionation. In: Stable Isotope
1. The gaseous-diffusion method is used for separating
Hydrology, Deuterium and Oxygen-18 in the Water Cycle.
and enriching uranium isotopes. The production
J.R. Gat and R. Gonfiantini (Eds.). IAEA, technical report
plant which uses this technique is huge; the huge series, No. 210, pp. 21–34.
diffusion plant at Oak Ridge, USA, in operation
13. Krauskopf, K.B. (1979). Introduction to Geochemistry, 2nd
from 1945, represents one of the greatest technical Edn. McGraw-Hill, New York.
achievements in the history of humankind.
14. Bruchert, V., Knoblauch, C., and Jorgensen, B.B. (1999). Sul-
2. The electromagnetic method is suitable only where fur isotope fractionation during bacterial sulfate reduction.
cheap electricity is available. In: Geochemistry of Earth’s Surface. H. Armannsson (Ed.).
3. The thermal diffusion method is good for laboratory- Balkema, Rotterdam, pp. 303–305.
scale production.
4. The centrifuge method was not developed into a
large-scale production plant. MARIOTTE BOTTLE—USE IN HYDROLOGY
5. The electrolytic method is used only for separating
JOSEPH HOLDEN
hydrogen and deuterium.
University of Leeds
6. The chemical exchange method is used mostly for Leeds, United Kingdom
separating C and N.
7. The distillation method is only for enrich- A Mariotte bottle is a device that allows constant flow
ing deuterium. of a fluid from a container, even when the fluid level in
the container changes. It is named after Edme Mariotte
A sideline and an interesting coincidence is that while (1620–1684), a French physicist, who, independently from
this article was being written, the Board of Governors Boyle, discovered the law that states that the volume of
of International Atomic Energy Agency (IAEA) postponed a gas varies inversely with its pressure. Mariotte bottles
its decision on Iran’s (the home country of the author) have a wide range of applications in engineering and
uranium enrichment and nuclear program activities (CNN science (such as fuel supply) and are used in a variety
Online edition, Nov. 22, 2003). The uranium enrichment of hydrologic applications ranging from river discharge
and nuclear activities program of Iran has been in the gauging to infiltration measurement.
news headlines for quite a long time. Figure 1 shows the general design of a Mariotte bottle.
It consists of a bottle that has an escape pipe or tap to allow
BIBLIOGRAPHY fluid to exit and a vent tube to allow air to enter the bottle.
The bottle is sealed so that the only location where air can
1. Bowen, R. (1988). Isotopes in the Earth Sciences. Elsevier, enter is at that bottom of the vent tube, b. The fluid (such
London.
as water) exits through a tap (or any sort of escape tube) at
2. Appelo, C.A.J. and Postma, D. (1999). Geochemistry, Ground- c, at a rate that is controlled by the size of the escape tube
water and Pollution. Balkema, Rotterdam.
and the head of water. The head of water, h will always be
3. Kazemi, G.A. and Mehdizadeh, H. (2001). Solute content equal to the distance between the outlet tap and the bottom
and chemical composition of atmospheric precipitation of
of the vent tube. Therefore, regardless of the water level
Shahrood Region, Proc. of 5th Iranian Geol. Convention,
Tehran, September 2001, pp. 401–405 [in Persian].
inside the bottle, there will be a constant head of water.
If the vent tube is kept at a constant level, then as
4. Panettiere, P., Cortecci, G., Dinelli, E, Bencini, A., and
Guidi, M. (2000). Chemistry and sulfur isotopic composi- a result of the reduction of the fluid in the bottle, the
tion of precipitation at Bologna, Italy. Appl. Geochem. 15: air above it expands and its pressure drops. When the
1455–1467. pressure of the air has dropped so that it plus the pressure
5. Domenico, A.P. and Shwartz, F.W. (1991). Physical and of the water column from c to b (the vent tube opening) is
Chemical Hydrogeology. John Wiley & Sons, New York. less than the atmospheric pressure of the air outside the
6. Coplen, T.B. (1993). Uses of environmental isotopes New Mariotte bottle (and therefore also inside the vent tube),
York. In: Regional Ground-Water Quality. W.M. Alley (Ed.). then air enters the Mariotte bottle at b and rises through
ITP, pp. 255–293. the fluid. Therefore, the pressure at c will always be equal
504 MARS EXPLORATION ROVER MISSION
Vent tube
Air
Air
Water
c
Fluid
d
e
Water
h Figure 2. Mariotte bottle used to keep a constant water level

above the ground surface.
a then it can be used to determine how much water has been

supplied during any given period. Frequently, Mariotte
Outlet pipe
bottles are used in instruments designed to measure soil
infiltration rates such as rain simulators or other types of
Figure 1. A Mariotte bottle. infiltrometers where a constant rate of water supply to the
soil surface might be required.
to atmospheric which is also equal to the pressure at b. In constant head infiltrometers (see entry for
The excess pressure, under which the fluid is ejected, is INFILTROMETERS), the Mariotte principle is used to ensure
the pressure of the fluid between the levels b and a. It that there is a constant depth of ponded water above the
remains constant until the fluid drops to the level of b, and soil surface. This is done by placing the vent tube for the
the discharge from the escape pipe will remain constant Mariotte bottle at the same end of the bottle as the water
until then. Thus, Mariotte bottles can be used wherever exit pipe as in Fig. 2. Whenever the water level on the
there is an application for a constant flow of water from a ground drops below the height of the base of the vent tube,
supply reservoir. d, air is allowed to enter the water supply bottle, and water
If the vent tube is raised to a greater height, the head can be released from e until the water level on the ground
will increase, and therefore, the rate of water flowing out has risen back to the base of the vent tube. In this way, a
of the bottle will increase. It is therefore possible to control constant height or head of water can be maintained on a
the rate of water release from the bottle by controlling the soil surface. This principle can also be used for irrigation
head of water, and this is, in turn, controlled by the height to maintain a constant surface water level or a constant
of the base of the vent tube above the outlet at a. If the groundwater level.
vent tube is moved down to the level of the escape opening,
a, the flow stops because then the height of the column
becomes zero. The discharge also ceases if the vent tube is MARS EXPLORATION ROVER MISSION
sealed preventing air from entering the bottle.
GUY WEBSTER
There is a wide range of common hydrologic applica-
Jet Propulsion Laboratory
tions for Mariotte bottles. They are used frequently for
Pasadena, California
adding a constant rate of solution (such as a dye or tracer)
to watercourses. Most commonly this is for discharge gaug- DONALD SAVAGE
ing where, if a constant rate of solution (such as sodium NASA Headquarters
chloride solution) is added to river, then the discharge of Washington
the river can be calculated based on the equation, OPPORTUNITY ROVER FINDS STRONG EVIDENCE
MERIDIANI PLANUM WAS WET
Q = q(Ci − Cd )/(Cd − Cb )
where Q is the river discharge, q is the rate of injection of Scientists have concluded the part of Mars that NASA’s
tracer solution, Ci is the concentration of the tracer, Cb is Opportunity rover is exploring was soaking wet in
the background concentration in the river, and Cd is the the past.
downstream sampled concentration of the tracer.
It is often useful to use a Mariotte bottle to supply a This article is a US Government work and, as such, is in the
constant rate of water and also, if the bottle is graduated, public domain in the United States of America.
MARS EXPLORATION ROVER MISSION 505
This image, taken by Opportunity’s microscopic imager, shows a portion of the rock outcrop at
Meridiani Planum, Mars, dubbed ‘‘Guadalupe.’’ View Opportunity images from this press release.
Evidence the rover found in a rock outcrop led scientists Laboratory, Pasadena, Calif., because of the good fortune
to the conclusion. Clues from the rocks’ composition, to have the spacecraft arrive next to an exposed slice of
such as the presence of sulfates, and the rocks’ physical bedrock on the inner slope of a small crater.
appearance, such as niches where crystals grew, helped The robotic field geologist has spent most of the past
make the case for a watery history. three weeks surveying the whole outcrop, and then turning
‘‘Liquid water once flowed through these rocks. It back for close-up inspection of selected portions. The rover
changed their texture, and it changed their chemistry,’’ found a very high concentration of sulfur in the outcrop
said Dr. Steve Squyres of Cornell University, Ithaca, with its alpha particle X-ray spectrometer, which identifies
N.Y., principal investigator for the science instruments chemical elements in a sample.
on Opportunity and its twin, Spirit. ‘‘We’ve been able to ‘‘The chemical form of this sulfur appears to be
read the tell-tale clues the water left behind, giving us in magnesium, iron or other sulfate salts,’’ said Dr.
confidence in that conclusion.’’
Benton Clark of Lockheed Martin Space Systems, Denver.
Dr. James Garvin, lead scientist for Mars and lunar
‘‘Elements that can form chloride or even bromide salts
exploration at NASA Headquarters, Washington, said,
have also been detected.’’
‘‘NASA launched the Mars Exploration Rover mission
At the same location, the rover’s Mössbauer spectrom-
specifically to check whether at least one part of Mars ever
had a persistently wet environment that could possibly eter, which identifies iron-bearing minerals, detected a
have been hospitable to life. Today we have strong evidence hydrated iron sulfate mineral called jarosite. Germany
for an exciting answer: Yes.’’ provided both the alpha particle X-ray spectrometer and
Opportunity has more work ahead. It will try to deter- the Mössbauer spectrometer. Opportunity’s miniature
mine whether, besides being exposed to water after they thermal emission spectrometer has also provided evidence
formed, the rocks may have originally been laid down by for sulfates.
minerals precipitating out of solution at the bottom of a On Earth, rocks with as much salt as this Mars rock
salty lake or sea. either have formed in water or, after formation, have been
The first views Opportunity sent of its landing highly altered by long exposures to water. Jarosite may
site in Mars’ Meridiani Planum region five weeks point to the rock’s wet history having been in an acidic
ago delighted researchers at NASA’s Jet Propulsion lake or an acidic hot springs environment.
506 REMOVAL OF ORGANIC MICROPOLLUTANTS AND METAL IONS FROM AQUEOUS SOLUTIONS BY ACTIVATED CARBONS
The water evidence from the rocks’ physical appearance demands of the population. However, the increase in
comes in at least three categories, said Dr. John agricultural and industrial growth has also led to a
Grotzinger, sedimentary geologist from the Massachusetts corresponding growth in pollution related to the two
Institute of Technology, Cambridge: indentations called activities. The use of pesticides to boost food production
‘‘vugs,’’ spherules and crossbedding. has caused increased contamination of drinking water,
Pictures from the rover’s panoramic camera and thus putting pressure on governments to invest in the
microscopic imager reveal the target rock, dubbed ‘‘El provision of clean water. The same can also be said
Capitan,’’ is thoroughly pocked with indentations about a about the increase in industrial activities, which has
centimeter (0.4 inch) long and one-fourth or less that wide, seen, among many of the pollutants, the increase of
with apparently random orientations. This distinctive heavy metals in the water bodies. Since the installation of
texture is familiar to geologists as the sites where crystals water pollution control schemes are not profit-generating
of salt minerals form within rocks that sit in briny water. activities, there is a lot of ongoing research into finding
When the crystals later disappear, either by erosion or efficient but inexpensive ways of separating pollutants
by dissolving in less-salty water, the voids left behind are from drinking water.
called vugs, and in this case they conform to the geometry The ban on the use of atrazine for nonagricultural
of possible former evaporite minerals. purposes in the United Kingdom in 1993 has resulted in
Round particles the size of BBs are embedded in the the introduction of alternative herbicides. Among them is
outcrop. From shape alone, these spherules might be benazolin, which can be a potential groundwater pollutant
formed from volcanic eruptions, from lofting of molten owing to its high water solubility when compared with
droplets by a meteor impact, or from accumulation of that of atrazine. A report compiled by the Environment
minerals coming out of solution inside a porous, water- Agency (1) revealed that, for the first time in England
soaked rock. Opportunity’s observations that the spherules and Wales, benazolin has become one of the 10 most
are not concentrated at particular layers in the outcrop frequently occurring herbicides in surface water. The
weigh against a volcanic or impact origin, but do not potential harmful effects of benazolin as a herbicide
completely rule out those origins. necessitates that it must be removed from water bodies
Layers in the rock that lie at an angle to the main if present at concentrations higher than that required
layers, a pattern called crossbedding, can result from by environmental regulations. However, conventional
the action of wind or water. Preliminary views by water treatment processes, such as filtration, disinfection,
Opportunity hint the crossbedding bears hallmarks of coagulation, and sedimentation, have been reported to be
water action, such as the small scale of the crossbedding more effective in removing less water-soluble and easily
and possible concave patterns formed by sinuous crestlines degradable pesticides (2). Ozone technology is a potentially
of underwater ridges. attractive technique in wastewater treatment. However,
The images obtained to date are not adequate for in some cases, the ozone degradation products can be
a definitive answer. So scientists plan to maneuver equally harmful. Lambert et al. (3) have investigated
Opportunity closer to the features for a better look. ‘‘We degradation efficiency of benazolin and other herbicides
have tantalizing clues, and we’re planning to evaluate this using ozone and a combination of ozone and hydrogen
possibility in the near future,’’ Grotzinger said.
peroxide. They realized that for benazolin, doses of
JPL, a division of the California Institute of Technology
1.0 mg O3 /l, 2.0 mg O3 /l, and 3.1 mg O3 /l, degradation
in Pasadena, manages the Mars Exploration Rover project
efficiencies of 55%, 68%, and 88% could be achieved,
for NASA’s Office of Space Science, Washington.
respectively. The combination of ozone and hydrogen
For information about NASA and the Mars mission on
peroxide further increased the degradation efficiency,
the Internet, visit http://www.nasa.gov/.
but they recommended complementary techniques for
Images and additional information about the project
other herbicides. Adsorption on activated carbon is
are also available at http://marsrovers.jpl.nasa.gov/ and
one of the well-established and effective techniques for
http://marsrovers.jpl.nasa.gov/relocate.html?relocate=
the removal of herbicides from water. The adsorptive
http://athena.cornell.edu/.
efficiency of activated carbon is because of its well-
http://marsrovers.jpl.nasa.gov and http://athena.cornell
developed internal pore structure, surface area, and
.edu.
surface reactivity. Typically, activated carbon consists
of interwoven micropores, mesopores, and macropores,
REMOVAL OF ORGANIC MICROPOLLUTANTS and these are responsible for the high surface area of
the carbons. The existence of surface functional groups
AND METAL IONS FROM AQUEOUS
in an activated carbon matrix means that they can be
SOLUTIONS BY ACTIVATED CARBONS manipulated by thermal or chemical treatments to produce
adsorbents that are tailored for particular functions.
BASU SAHA
As wastewater streams consist of a mixture of
Loughborough University
pollutants that can generally be classified as organic and
Loughborough, United Kingdom
inorganic pollutants, tailoring an adsorbent to remove
INTRODUCTION both classes of pollutants would give a significant process
advantage. Among the inorganic pollutants, heavy metals
World population growth has triggered agricultural and form a major part in the classification. Heavy metals
industrial growth for the sustenance of the day-to-day constitute a class of environmental pollutants that are
REMOVAL OF ORGANIC MICROPOLLUTANTS AND METAL IONS FROM AQUEOUS SOLUTIONS BY ACTIVATED CARBONS 507
of particular concern in the treatment of industrial studies have gone into trying to improve the removal
wastewater because of their potential danger to humans efficiency of the existing adsorbents. In improving the
and the environment. Therefore, they must be treated if adsorption capacity of adsorbents, attention has been
present at concentrations higher than those specified by turned on to the number of variables that affect the
the environmental regulations. adsorption capacity of the activated carbons.
Activated carbons have been proven to be an effective As pollution in wastewater is because of a cocktail
adsorbent for the removal of organic compounds from of pollutants, the use of activated carbon could be a
water (4,5). Research has also shown that they have viable way of simultaneously adsorbing pollutants of a
a potential to remove heavy metals, especially in its different nature, which has a further advantage of cutting
modified state. For example, a number of authors have down on the number of wastewater processing units.
oxidized activated carbons to enhance sorption of heavy Hence, the objective of this research is to investigate
metals (6–9). The oxidation processes have mainly been the effect of simultaneous adsorption of a heavy metal
by use of nitric acid, ozone, and air. The three methods (copper) and a selected herbicide (benazolin) on as-
have always produced the same results, that of producing received and surface-modified activated carbons. The
weakly acidic functionality on the surface of the carbon. adsorbents chosen for this work are as-received F400
Saha et al. (10) compared the metal sorption capacity of a supplied by Chemviron (USA) and various surface-
conventional carbon and its oxidized form. The adsorption modified samples of F400 that were prepared in the
of herbicides and trace metal ions onto activated carbon author’s laboratory. Benazolin is highly soluble in
is discussed in this chapter. In particular, we discuss the water (solubility 600 mg/l) when compared with atrazine
effect of modification of activated carbon on the uptake of (solubility 33 mg/l). Properties, structural formula, and
copper and benazolin. The effect of post-treatment of the three-dimensional representation of benazolin are given in
oxidized carbons on the uptake of the target pollutants is Table 1 and Fig. 1, respectively. Copper has been chosen as
also highlighted. The adsorbents were characterized in a a representative of heavy metals. In order to understand
bid to understand their chemical and physical properties, the sorption mechanism of the target pollutants, physical
which gives a better understanding of the nature of the and chemical characterization were performed for all
association of the target pollutants and the adsorbents. the adsorbents. The other objective of the study is to
investigate the mechanism of adsorption of the chosen
BACKGROUND pollutants. If the mechanism is fully understood, it is
possible to produce tailor-made engineered adsorbents for
Herbicides have undoubtedly contributed to the growth various purposes in water treatment.
of agricultural productivity. For example, it has been
reported that a $4 billion U.S. investment in herbicide EXPERIMENTAL
control saved $16 billion U.S. in crops every year (11).
However, the benefit realized from the use of herbicides Adsorbent Materials
does not factor in the negative effects they have on A coal-based commercial granular activated carbon, F400
human health. was used for surface modification with nitric acid. The
As a result of their negative impact on human
health and the environment, stringent legislation has Table 1. Properties and Structural Formula of Benazolin
been imposed on the purity of drinking water. The
European Union (E.U.) agreed to have the drinking water Solubility
Directive Admissible Concentration for single herbicide Chemical in Water
Herbicide Formula Structural Formula (mg/L) pKa
or pesticide at a maximum of 0.1 µg/L. In Great Britain,
the National Rivers Authority (12) showed that eulan and Benazolin C9 H6 ClNO3 S S 600 3.04
permethrin (mainly in effluents from the textile industry) O
had exceeded their Environmental Quality Standard
N
(EQS) limits. Also, diazinon, a sheep dip insecticide,
had also exceeded its EQS limits. One major concern CI CH2CO.OH
was that the most commonly used, herbicides atrazine,
diuron, bentazone, isoproturon, and mecoprop exceeded
the standard 0.1 µg/L in surface waters, which has been
and always will be a major concern simply because most
groundwater sources used for drinking water have no
treatment facilities designed to remove herbicides. Also,
the most common water treatment processes, such as
sedimentation, disinfection, coagulation, and filtration,
can effectively remove only selected herbicides, especially
those that have low solubility in water. On the other
hand, advanced water treatment processes (mainly using
adsorption onto activated carbon) have proved to be the
most efficient and reliable method for the removal of
aqueous dissolved organic herbicides. Hence, a lot of Figure 1. 3-D molecular representation of benazolin.
carbon sample was obtained from Chemviron (USA). Table 2. Procedures for Solid Phase Extraction (SPE)
Before the oxidation process, the carbon was washed Conditioning 500 mg, s-triazine, 6 ml SPE columns were
in distilled water and dried at 383◦ K for 24 hours. The conditioned by passing 10 ml of
dried sample was then subjected to different modes of HPLC-grade methanol followed by 10 ml of
modification using part of a method described by Saha 2% MeOH in water under gravity.
et al. (7) and Chingombe et al. (13). Extraction 500 ml sample solution were passed through
the column at a flow rate of 1 ml/min.
Sample Preparation Elution 6 ml of acetonitrile was allowed to soak in the
columns for 20 minutes before being passed
A known amount of the washed and dried sample was through at approximately 0.5 ml/min and
introduced into the reaction vessel that contained nitric the eluent collected in 7 ml vials. The
acid and distilled water at a ratio of 1:1 (v/v). The reaction columns were then blown dry with
mixture was heated to 363◦ K, and the reaction was nitrogen and the solvent was evaporated to
allowed to continue for 6 hours with continued stirring. dryness using a stream of nitrogen.
The reaction solution was removed and a fresh reaction Reconstitution 500 µl of 40% acetonitrile, 60% 10 mmol
mixture of nitric acid and water was introduced. The KH2 PO4 buffer at pH 3.0 was added to the
reaction was allowed to continue for an additional 3 hours vials to redissolve the analytes. The vials
were shaken vigorously and their contents
under the same reaction conditions. The oxidized carbon
transferred to HPLC auto-sample vials.
was then washed with distilled water until no further
change in pH could be detected.
The resultant water-washed samples (after oxidation)
showed that correlations were linear over the range
were divided into three portions. The first portion was
(0–8 ng/µl).
denoted AC1, and this is a sample that was washed with
Solid phase extraction (SPE) technique was employed
water after the oxidation reaction. AC2 was a sample
for the preconcentration of benazolin before analysis. The
that was washed with 0.1 M sodium hydroxide to remove
complete procedure for SPE is given in Table 2.
humic acids that were formed during the oxidation process.
The washing was continued until no further coloration
from the humic acids could be detected. Reconditioning CHARACTERIZATION OF ADSORBENTS
of the sample to the hydrogen form was performed using
0.1 M hydrochloric acid, and further washing with distilled All samples were characterized to investigate the effect
water was done until the pH of the supernatant solution of the treatment and the chemical as well as the physi-
stabilized to a pH of about 4.0. The conversion was deemed cal characteristics of these materials. The characterization
complete when the inlet and the outlet pH of the liquid included scanning electron micrographs (SEM), Brunauer-
flowing in the column were the same. The third sample, Emmett-Teller (BET) and Langmuir surface area mea-
denoted AC3, was a result of heating the third portion surements, pore size distribution, Fourier Transform Infra
of the oxidized sample at a temperature of 580◦ K under Red spectroscopy (FTIR) analysis, sodium capacity mea-
vacuum of 2 mmHg for 12 hours to remove humic acids. surement, pH titration, zeta potential measurements,
and X-ray photoelectron spectroscopy (XPS) analysis and
elemental analysis. The detailed characterization proce-
ANALYSIS OF METAL ION AND HERBICIDE dure and the results have been reported by Chingombe
et al. (13). The results indicated that a significant change
Copper
occurred in the structure of the precursor by oxidation
The copper ion content was analyzed using a Var- and subsequent treatments. Characterization of the adsor-
ian SpectraAA-200 atomic absorption spectrophotometer bents revealed that modification of the carbon by oxidation
(AAS). The AAS was set on the flame mode using air- resulted in the introduction of weakly acidic functional
acetylene mixture. groups, and the presence of such groups was confirmed
by FTIR, pH titration, zeta potential measurements, and
HPLC Analysis Method sodium capacity results. Sodium capacity was enhanced
As preconcentration of benazolin before analyzing on the by a factor of up to 8 for the oxidized carbons compared
HPLC was required, a trace analysis method had to with as-received samples.
be developed for the quantification of the herbicide. A
Batch Adsorption Experiments
Hewlette Packard 100 series HPLC consisting of a diode
array detector, column thermostat, auto sampler, and a The sorption of copper on the F400 carbon range is
binary pump was used for analysis. Jones Chromatograph shown in Fig. 2. The results clearly show that AC1 and
supplied a Genesis C18 silica column (4 µm, 150 × 3.0 mm) AC2 have the same adsorptive capacity for copper when
combined with a 1 cm Genesis guard column. The detector, compared with AC3 and unoxidized F400. However, among
flow rate, and injection volume were set at 222 nm, the adsorbents employed, as-received F400 showed the
0.88 ml/min, and 100 µl, respectively, and this gave the least adsorptive capacity. The enhanced capacity observed
retention time of benazolin at 2.01 minutes. The column on AC1, AC2, and AC3 could be as a result of the
temperature was also set at 311◦ K. The eluent consisted oxidation process, which generated carboxylic functional
of a solution of 40% acetonitrile and 60% buffer (10 mmol groups as evidenced in the FTIR spectra and the high
KH2 PO4 at pH 3). The calibration curve for benazolin sodium capacity of the above sorbents. These groups
REMOVAL OF ORGANIC MICROPOLLUTANTS AND METAL IONS FROM AQUEOUS SOLUTIONS BY ACTIVATED CARBONS 509
1.4 analysis (13). It is generally accepted that sorption of an

organic molecule takes place mainly in the micropore
Copper uptake, mmol/g
1.2
region of the adsorbents, because in this region, the contact
1
F400 points between the adsorbent and the molecules is higher
0.8 AC3 than in mesopores and macropores. Also, it is reported that
0.6 AC1 an overlap of potential forces exists in the micropores,
AC2 which results in increased adsorption forces. Therefore,
0.4
naturally, the capacity of AC1 and AC3 is expected to
0.2
be low. Another factor in the difference between the
0 adsorption capability of the adsorbents is the electrostatic
0 0.05 0.1 0.15 0.2 interaction between the adsorbent and the adsorbate.
Equilibrium concentration, mmol/l The zeta potential results show that the isoelectric
Figure 2. Equilibrium sorption results of copper on F400 and point (IEP) of the oxidized carbons are significantly lower
the oxidized carbons. (1.5) than that of the conventional carbon (6.0) because
of the introduction of weakly carboxylic acid groups with
would be dissociated in aqueous solution, and they are dissociation between pH 2 and 6. As the experiments are
responsible for copper sorption. The oxidized and heat- performed at pH 5, it would imply that the surface of the
treated sample AC3 showed reduced capacity compared oxidized carbons is predominantly negative because of the
with AC1 and AC2, which could be because of the dissociation of the carboxylic groups. Also, the pKa value
decomposition of the carboxylic acid groups during the of benazolin (3.04) is lower than the pH of the solution
heat treatment. Toledo-Bautista et al. (14) made similar experiment, and this condition would render mutual
observations when they adsorbed Cr(III) on a range of repulsion between the adsorbent and the adsorbate.
conventional and modified carbons. Adsorption of benazolin is more favorable on F400 because
The formation of metal surface complexes on the of its high IEP. Streat and Horner (4), when comparing
oxidized carbon involving cooperative action seems quite adsorption of some herbicides (including benazolin) on
likely. Approximate calculation of oxygenated functional F400 and hypercross-linked polymers, realized that F400
group density per unit area for F400(ox) yields a value outperformed the polymeric sorbents. They postulated
of 0.02 functional groups per square angstrom. Given
that the oxidized carbons evaluated in the present study
possess a large proportion of pores in the region of 10–20 25
Equilibrium concentration
on adsorbent (µmol/g)
angstrom and the diameter of the hydrated metal ions is

20
approximately 8 angstrom (15), it is reasonable to assume
a cooperative binding mechanism (see Fig. 3). 15 F400
AC1
Figure 4 shows the adsorption of benazolin on all the AC2
adsorbents under study. The results clearly show that 10 AC3
benazolin adsorption on F400 is slightly higher than AC2 5
and reasonably higher than AC1 and AC3. However, AC1
and AC3 seem to have similar sorptive performance, 0
whereas AC2 showed a higher performance than the 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14
other oxidized carbons, which can be expected because Equilibrium solution concentration (µmol/l)
there was a significant loss in micropores for samples Figure 4. Adsorption of benazolin on F400 and the oxi-
AC1 and AC3 as obtained by the pore size distribution dized carbons.
O
C
O O
O C C
O O
C OH
OH C H C
O H
O O O O
C
Cu2+ (H2O)6
O C O Cu O C
C O O O
H
O C H
O
C Figure 3. Postulated complexation re-
O action between copper (II) and oxidized
carbon surface.
that adsorption of benazolin on F400 was through the 25
dispersion interaction of the π electrons in the respective
20
aromatic systems by a donor-acceptor mechanism. They
further noted that the chloro (–Cl) group on benazolin is 15
an electron-withdrawing group, therefore implying that a
reduction of electron density occurs in the π ring system 10
Benaz + 10 ppm Cu
of the herbicide. Hence, benazolin would act as an electron 5 Benaz + 5 ppm Cu
acceptor in the donor-acceptor mechanism involving F400. Benaz
This observation is consistent with investigations carried 0
out by Radovic et al. (16) when they considered the 0 0.05 0.1 0.15 0.2
adsorption of aromatics possessing electron-withdrawing Equilibrium solution concentration (µmol/l)
and electron-donating functional groups on conventional Figure 6. Simultaneous adsorption of copper and benazolin
and chemically modified activated carbons. Chemical on AC1.
modification involved samples that were subjected to
nitration and others that were oxidized. Heat-treated
1.2
Copper uptake, mmol/g

carbons were also considered. They used nitrobenzene and
aniline as the target adsorbates. Nitrobenzene contains 1
nitro (–NO2 ) group, which is known to be electron 0.8 AC3
withdrawing, whereas aniline possesses an amino (–NH2 ) AC2
0.6 AC1
group, which is electron donating. As nitration and
0.4 F400
oxidation reduces the π -electron density in the graphene
layers of the carbon surface, which would in turn favor 0.2
adsorption of aromatics with electron donor functional 0
groups, it was realized that maximum adsorption uptake 0 20 40 60 80
of aniline (with electron donor group) were attained on Initial benazolin concentration, ppb
oxidized carbon surfaces. On the other hand, maximum Figure 7. Effect of the initial concentration of benazolin on
adsorption uptake for nitrobenzene were found on heat- copper uptake.
treated carbon surfaces (which possess higher π -electron
density than the oxidized carbons). Hence, from the
uptake is distinctly affected in the presence of copper ions.
view point of Radovic et al. (16), it would be reasonable
Also, Fig. 7 shows the effect of the initial concentration of
to postulate that benazolin adsorption is favored by
benazolin on the uptake of copper. The results clearly show
adsorbents with high π -electron density. However, in
that as the initial concentration of benazolin increases, a
the oxidized carbons employed in this work, such a
corresponding decrease occurs in the uptake of copper up
mechanism would be highly unlikely because of the
to the point when the initial concentration of benazolin
introduction of oxygen groups in the form of carboxylic
is at approximately 20 ppb, that is when the uptake
acids. The carboxylic acid functional groups are electron
of copper normalizes. The immediate conclusion that
withdrawing, and this condition reduces the π -electron
could be drawn from the two sets of results is that a
density in the ring systems of the carbon. Therefore,
competition exists for the same adsorption sites by copper
the dispersion interaction between the herbicide and the
and benazolin. As mentioned previously, the adsorption
oxidized carbon is reduced. Hence, unoxidized F400 would
of benazolin is favored through electron donor-acceptor
be expected to have a higher benazolin capacity than the
type of a mechanism, which involves π electron in the
oxidized carbons.
ring systems of the carbon. Biniak et al. (6) showed that
Figures 5 and 6 show the adsorption isotherms of
copper ions had the capacity to interact with the π
benazolin in the presence of copper at 5 ppm and
electron in the ring systems of the carbon. In their
10 ppm for unoxidized F400 and AC1 and AC3 samples,
work, they employed XPS and FTIR to study the state
respectively. For both cases, it is apparent that benazolin
of the copper adsorbed on different carbon surfaces. They
investigated three types of modified activated carbons.
25 One of the modified carbons was obtained by annealing
a conventional carbon in a vacuum at 1000◦ K. This
20 carbon had a low oxygen content of 2.72% compared

with oxidized and ammonia-treated carbons, which had
15
10.10% and 6.20%, respectively. The annealed carbon was
10 virtually devoid of acidic functional groups. XPS and FTIR
results indicated that the dominant mechanisms of copper
5 Benaz adsorption on heat-treated active carbons could be dipole-
Benaz + 10 ppm Cu
Benaz + 5 ppm Cu dipole (π -d) interactions between graphene layers and
0 metal ionic species as well as spontaneous electrochemical
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 reduction of copper ions. Hence, it is expected that
Equilibrium solution concentration (µmol/l) benazolin and copper would compete for the same sites
Figure 5. Simultaneous adsorption of copper and benazolin derived from the π electrons in the graphene layers of
on F400. the modified adsorbents. However, conclusive evidence of
MOLECULAR NETWORK DYNAMICS 511
such competition would only be proved by employing an 5. Streat, M., Malik, D.J., and Saha, B. (2004). In: Ion Exchange
annealed carbon adsorbent. and Solvent Extraction. Vol. 16. A.K. SenGupta and
Y. Marcus (Eds.). Marcel Dekker, New York, pp. 1–84.
CONCLUSIONS 6. Biniak, S., Pakula, M., Szymanski, G.S., and Swiatkowski, A.
(1999). Langmuir 15: 6117–6122.
Oxidation of the conventional carbon and post-treatment 7. Saha, B., Tai, M.H., and Streat, M. (2001). Trans. IChemE.
of the resultant carbons had an influence on their 79(Part B): 211–217.
uptake for copper and benazolin. Oxidation resulted in 8. Jia, Y.F., Xiao, B., and Thomas, K.M. (2002). Langmuir 18:
the introduction of weakly acidic carboxylic functionality, 470–478.
which markedly enhanced the uptake of copper ions. The 9. Saha, B., Tai, M.H., and Streat, M. (2003). Chem. Eng. Res.
isoelectric point (IEP) was also reduced as a result of Des. 81: 1343–1353.
oxidation. Oxidation of the carbon resulted in altered 10. Saha, B., Tai, M.H., and Streat, M. (2001). Trans. IChemE.
pore size distribution. Reduction in microporosity was 79(Part B): 345–351.
observed, and it also explained the low uptake of 11. Department of the Environment. (1996). Pesticides in Water.
benazolin by oxidized sample compared with as-received Department of the Environment Report, The Working Party
F400. However, post-treatment by washing with sodium on the Incidence of Pesticides in Water.
hydroxide seemed to open up the pores for AC2 sample. 12. National Rivers Authority. (1998). Pesticides in the aquatic
Simultaneous adsorption of benazolin and copper showed environment. Water Quality Series 26: 12.
that direct competition existed for the active sites, and 13. Chingombe, P., Saha, B., and Wakeman, R.J. (2004). 16th
possibly the two adsorbates targeted the π electron cloud International Congress of Chemical and Process Engineering,
on the graphene layers of the adsorbent. Copper uptake by Prague, Czech Republic.
the conventional and modified samples decreased in the 14. Toledo-Bautista, I., Utrilla-Rivera, J., Garcia-Ferro, M.A.,
presence of benazolin and vice versa. The decrease in metal and Castillo-Moreno, C. (1994). Carbon 32(1): 93–100.
uptake was dependent on the concentrations of metals in 15. Nightingale, E.R.J. (1959). Phys. Chem. 63: 1381–1386.
multicomponent aqueous solutions. The effect of copper 16. Radovic, L.R. et al. (1997). Carbon. 35(9): 1339–1348.
on benazolin uptake was more pronounced than the effect
of benazolin on copper uptake from water. Further work
is necessary to fully understand the precise mechanisms MOLECULAR NETWORK DYNAMICS
of adsorption of complex aromatic molecules onto carbon,
and this forms the basis of our continued work. Research is D.L. MARRIN
in progress in the author’s laboratory to provide conclusive Hanalei, Hawaii
evidence on the interaction of these adsorbates with the
modified carbon surface.
The environment is challenged by other micropol- Perhaps the twentieth century’s most startling discovery
lutants, namely polychlorinated hydrocarbons, organic about liquid water is that it functions as a vast
oxygen compounds, surfactants, pesticides, aliphatic interconnected network, rather than as a random
intermediates developing from the chemical indus- collection of independent molecules. Individual water
try, and, of course, endocrine disrupting compounds molecules serve as network elements, which constitute
(EDCs). Further study of activated carbon continues to the building blocks for the fundamental water network.
explore for potential solutions to these important prob- Elements in the water network are connected to one
lems. another via magnetic-type linkages known as hydrogen
bonds. In solid water (ice), each water molecule hydrogen
Acknowledgments bonds with all four of its nearest neighbors in forming
I would like to express my sincere gratitude to Professor Michael a perfect tetrahedron, which is a three-sided pyramid
Streat for his support, continuous encouragement, and advice comprising the most basic molecular geometry of water
for conducting the research in the area of adsorption and ion
(see Fig. 1). As ice melts into liquid water, some
exchange and development of tailored sorbent materials for
environmental remediation. I am indebted to Dr. Hadi Tai and
of the hydrogen bonds between water molecules are
Purazen Chingombe for their meticulous experimental work. broken (causing the tetrahedron to distort), whereas the
Funding from European Community, EPSRC, Severn Trent remaining bonds transition to an ultra-dynamic state
Water Ltd., and Commonwealth Scholarship Commission is whereby they are switched as rapidly as a trillion times per
gratefully acknowledged. second. The dynamics of this unique molecular network are
believed to underlie many of water’s anomalous properties.
BIBLIOGRAPHY
1. Environment Agency. (2002). The Annual Report of the NETWORK BEHAVIORS

Environment Agency pesticide monitoring programme. UK.
2. Kouras, A., Zouboulis, A., Samara, C., and Kouimtzis, T. This frantic switching of hydrogen bonds, which consti-
(1998). Environ. Pollut. 103: 193–202. tutes the underlying dynamism of water’s molecular net-
3. Lambert, S.D., Graham, N.J.D., and Croll, B.T. (1996). Ozone work, permits water to flow and to behave as a liquid even
Sci. Engineer. 18: 251–269. though it retains much of the molecular geometry of a solid.
4. Streat, M. and Horner, D.J. (2000). Trans. IChemE. 78(B2): Physical chemists have long modeled liquid water as a
363–382. space-filling network of individual water molecules, within
512 MOLECULAR NETWORK DYNAMICS
Hydrogen atom BIOLOGICAL STRUCTURE AND FUNCTION

H
Molecular biologists originally considered water to be
Covalent bond merely an inert solvent within which biomolecules (e.g.,
proteins and nucleic acids) structured and activated
Oxygen atom themselves. More recently, water has been perceived
as a key structural and functional component of many
H O biomolecules, whose interaction with the surrounding
environment is often mediated through hydration shells,
which are immense water networks surrounding the
H biomolecules (5). Water releases its hydrogen bonds with
biomolecular components so that protein or DNA assembly
may begin. Conformational changes in the biomolecule
Lone pair of electrons then restructure both the internal (bound or integral) and
H
external (hydrating) water that, in turn, facilitates more
changes in the biomolecule. This iterative process appears
Figure 1. This diagram represents the simplest tetrahedral unit to repeat until assembly and folding mechanisms yield a
of water’s H-bonded network (1). Two of the four hydrogen atoms bioactive crystal.
comprising the tetrahedron are covalently bonded to the central In addition to the structural importance of water in
oxygen atom (forming an individual H2 O molecule), whereas biomolecules, the insertion and removal of water also has
the remaining two hydrogen atoms (donated by each of two profound effects on their functioning. One of the most
neighboring water molecules) are hydrogen bonded to the oxygen’s remarkable functions that proteins perform is that of a
lone pairs of electrons. Every water molecule in the network can
biochemical catalyst or enzyme, which permits chemical
act as both a hydrogen donor and a hydrogen acceptor in bonding
reactions to proceed at rates that are millions of times
with its nearest neighbors.
faster than would occur under ambient conditions. Water
appears to play several roles in the functioning of enzymes.
which all potential hydrogen bonds were described as First, water molecules located at the interface of the
either unbroken (ON) or broken (OFF) (2). More recently, protein-substrate complex often provide the free energy
scientists have characterized the shuffling (i.e., breaking that actually binds the enzyme to its substrate. Second,
and forming) of hydrogen bonds comprising liquid water’s water may be excluded from the active site until the
network as a nonlinear process and, consequently, have moment of catalysis; at which time, it is allowed to enter
been unable to decipher the rules that govern the chang- the active site and reconfigure hydrogen bonds (permitting
the reaction to proceed). Finally, the enzymes actually
ing of molecular interconnections. This inability does not
imply that the switching rules are haphazard—only that structure the water network or networks around them,
thus enhancing the rate of catalysis (6).
science’s observational and mathematical techniques are
In addition to its role as a structural component of
currently overwhelmed by water’s dynamism. Even scien-
biomolecule, water hydrates DNA as it would any other
tists who study much simpler computer-generated systems
solute in an aqueous solution. It appears that water is
have found it difficult to identify switching rules solely
most highly ordered adjacent to the DNA molecule and
by analyzing network patterns and behaviors. Instead,
is less ordered with each successive hydration shell. This
researchers arbitrarily set the switching rules for these
highly ordered water is sometimes referred to as vicinal
artificial systems and then observe corresponding network
or biological water (6). As was the case for proteins, each
behaviors.
subunit of the DNA seems to have its own water network
Whereas network hierarchies within liquid water have
that interacts with other networks and, through successive
only recently been investigated, structural hierarchies are
hydration shells, with bulk water. If the geometry of
well described. Water exhibits structural self-organization
the DNA molecule is contorted, the hydration shells and
on at least three distinct levels: molecules, clusters
associated geometric clusters seem to be likewise affected.
of molecules (e.g., hydration shells), and clusters of
Similarly, the DNA molecule may contort in response
clusters (e.g., micelles). Self-organization implies that
to changes initiated by the water that hydrates it. The
water molecules or molecular clusters arrange themselves
reasons for twisting and folding DNA molecules may
into increasingly complex states. It is this self-organization
include stuffing it into chromosomes or unraveling it
that reportedly permits a network of simple interconnected
during transcription and replication.
elements, along with their switching rules, to display
coherent behaviors (3). Although the primary water
network is held together by short-range hydrogen bonds COMPLEX DYNAMICS
between neighboring molecules, it may either behave
as an extensive system with global coherence or as Scientists working with the pentagonal water cluster,
individual smaller regions with local order (4). These which serves as a basic building block for many large
two behaviors have been classically linked to network hydration shells, observed that water’s vibrational states
patterns displaying chaotic or disordered (i.e., bulk liquid are caused by a combination of bending and stretching
water) and more ordered (i.e., water clusters) properties, motions of its bonds. Molecular vibrations are expressed
respectively. in the units of cycles per second, or hertz. Although
MOLECULAR NETWORK DYNAMICS 513
water’s covalent bonds vibrate at about 1014 hertz (one- H

hundredth of a picosecond per cycle), its hydrogen bonds Total H H O H
vibrate within the lower frequency range of 1012 to 1013 shell H H
O O H
hertz (one-tenth to one picosecond per cycle). The time O H H O H O H O
H
frames for one cycle of these bond vibrations are shorter H H H
H O H O H
than those measured for the exchange hydrogen bonds in H H O
H O O H H H
H
H
water’s bulk liquid network (more than one picosecond). H H
H O H O O O
One picosecond equals a time period of one-trillionth of H O H
H
a second (10−12 seconds). In addition to the dynamics Primary
O
H H
O H
shell O H O
associated with vibrating and exchanging bonds, water’s H H
H O H O H
molecular network performs several concurrent activities O O
H O
that have been clocked at both faster and slower rates. H H H H H
H
As an example, ultrafast vibrational energy is trans- O H H O H O
H
H
O O
ferred in water on timescales less than one-tenth of a H H H H
picosecond (7). This ultrafast transfer could play a role H H O H O
H
in liquid water’s transporting vibrational energy among O H H O
Solute H H
certain solutes (e.g., proteins, DNA) and, perhaps, in trans-
ferring information within water’s local network. Recent Figure 2. This diagram illustrates the primary and total
spectroscopy studies have suggested that liquid water’s hydration shells surrounding an ionic, or electrically charged,
network may also function on timescales of less than a solute (1). In this case, a positively charged solute (e.g., a sodium
femtosecond (one-thousandth of a picosecond), which is ion in seawater) attracts the negatively charged oxygen atoms of
at least 100 times faster than the ultrafast vibrational surrounding water molecules. A solute’s effect on the orientation
energy. Moreover, some of the studies indicate that most of water molecules in the liquid network is greatest within the
molecules in the liquid network form only two (rather primary shell and diminishes as a function of distance from the
than three or four) hydrogen bonds with their neigh- hydrated ion. Theoretically, the total hydration shell includes all
water molecules in the network that are affected by the solute.
bors, perhaps creating long chains of molecules within a
disordered cluster network (8). In contrast to these sub-
picosecond dynamics, the reorientation of water molecules
and (3) varying orientations of the water molecules that
in a spatially restricted environment (e.g., the interface of
comprise ion hydration shells (9). These factors are
a micelle) occurs on the much slower timescale of hundreds
believed to profoundly affect the structure and dynamics
to thousands of picoseconds. Although further research is
of the water network comprising the world’s oceans, which
required to substantiate the recent observations, models
contain 97% of the planetary water. Perhaps scientists
describing both the structure and dynamics of water’s
will better understand the oceans when the secrets of
molecular network are rapidly evolving.
seawater’s molecular network dynamics are more fully
revealed.
SEAWATER
BIBLIOGRAPHY
Thus far, the discussion of water’s network has dealt
with either pure water or dilute aqueous solutions in
1. Marrin, D.L. (2002). West. In: Universal Water. Inner Ocean
which solutes comprise only a small fraction (e.g., less
Publ., Maui, HI, pp. 100–114.
than 0.1%) of the total solution. By contrast, the salts
2. Rice, S. (1975). Conjectures on the structure of amorphous
(ions) in seawater account for about 3.5% of the total
solid and liquid water. Topics Curr. Chem. 60: 109–200.
solution. Unlike the structure of pure water’s bulk phases
3. Varela, F., Thompson, E., and Rosch, E. (1991). The Embodied
and clusters, the molecular structure of seawater is
Mind. MIT Press, Cambridge, UK, pp. 87–90.
poorly described as a result of difficulties associated
4. Blumberg, R.L., Stanley, H.E., Geiger, A., and Mausbach, P.
with employing the investigative techniques that routinely
(1984). Connectivity of hydrogen bonds in liquid water. J.
probe the structure of pure water networks. Moreover, the Chem. Phys. 80: 5230–524.
high concentration of ions in seawater is hypothesized to
5. Westhof, E. (1993). Structural water bridges in nucleic acids.
significantly affect the water network. As the percentage In: Water and Biological Macromolecules. E. Westhof (Ed.).
of water molecules comprising hydration shells increases, CRC Press, Boca Raton, FL, Chapter 7.
the network structure of bulk water is compromised (see 6. Clegg, J. (1979). Metabolism and the intracellular environ-
Fig. 2). A salting-out effect is observed at very high ment: the vicinal-water network model. In: Cell-Associated
ion concentrations, when the interaction between water Water. W. Drost-Hansen and J. Clegg (Eds.). Academic Press,
molecules and ions is sufficiently weakened to permit the New York, pp. 363–413.
formation of salt crystals. 7. Nitzan, A. (1999). Ultrafast relaxation in water. Nature 402:
The most pertinent changes in the molecular water 472–475.
network that have been observed or hypothesized in 8. Zubavicus, Y. and Grunze, M. (2004). New insights into the
transitioning from dilute aqueous solutions to seawater structure of water with ultrafast probes. Science 304: 974–976.
include (1) the aforementioned ion–ion interactions, 9. Robinson, G., Wilse, S.B.Z., Singh, S., and Evans, M. (1996).
(2) overlapping hydration shells in which a single water Water in Biology, Chemistry and Physics. World Scientific,
molecule simultaneously hydrates more than one ion, Singapore, pp. 245–259.
514 IN SITU CHEMICAL MONITORING
IN SITU CHEMICAL MONITORING is to continuously renew the reagent at the sensor

surface. This strategy depends on molecular diffusion to
MELISSA A. SINGER PRESSMAN transport the analyte to and from the sensor surface;
JOSEPH H. ALDSTADT, III this type of sensor is called a ‘‘renewable reagent
Genetic Technologies, Inc. sensor’’ (4–6).
Testing Institute Chemical analyzers, on the other hand, depend on
Waukesha, Wisconsin active transport of the analyte into and away from
the monitor. Although analyzers lack the mechanical
INTRODUCTION simplicity of sensors, the addition of valves and pumps
in the design can more importantly provide a means by
For routine monitoring of environmental systems, in situ which to accommodate self-calibration of the monitor.
measurements offer several distinct advantages over The addition of these mechanical capabilities also
laboratory analysis (1). In situ analysis allows much allows for multistep, complex chemistries to be used
higher rates of sample throughput. As a result of the rapid as a way to increase sensitivity and selectivity of the
dynamics of most environmental systems, it is important monitor.
to collect frequent samples to prevent undersampling of Continuous flow analysis (CFA) is a more general
the phenomena under study, which is essential to allow term than so-called ‘‘chemical analyzers,’’ referring to
for a more accurate characterization of the true dynamics any measurement process in which the concentration
of the system and, therefore, a better understanding of of an analyte is determined based on a steady flow of
the biogeochemical processes involved (2). Additionally, sample through the instrument (7). Nonchromatographic
in situ analysis eliminates sample contamination that
CFA-based measurement systems are widely used for
inevitably develops during the transport and storage of
industrial process control and environmental monitoring.
samples to the laboratory. Furthermore, current research
As the instrument is often placed in a hostile environment
efforts in several laboratories have demonstrated that
in these types of applications, the relative simplicity
relatively inexpensive monitors can be fielded that can
of the CFA approach is attractive (8). Initial attempts
self-calibrate, can be remotely controlled, and can provide
to use CFA-based instruments were direct adaptations
reliable measurements in realistic environments (3). In
of benchtop (i.e., laboratory) CFA systems that were
particularly, it is the inability to provide high data quality
ruggedized (e.g., water-proofed) and taken to the sampling
at trace levels in complex matrices that has prevented
in situ measurements from gaining more favor. site (9). This approach was burdensome, expensive, and far
In situ monitors have been classified broadly as either from robust.
‘‘chemical sensors’’ or ‘‘chemical analyzers.’’ Chemical An example of an early effort of this type is
sensors depend on diffusion to bring the analyte to the illustrated by the work of Elliott et al. in using a
transducer surface. This type of sensor is mechanically segmented continuous flow analyzer, often called an
simple, sensitive, and practical from the standpoint ‘‘autoanalyzer,’’ for shipboard chemical analyses (10).
of low power requirements and small footprint, but These researchers monitored nitrate concentrations in the
often suffers from poor selectivity and the inability Everglades using a commercially available autoanalyzer
to self-calibrate. More importantly, sensors are based that they modified for the nitrate derivatization chemistry.
on reversible chemical reactions, i.e., the interaction of In their design, air bubbles were created in-line to
the analyte with the transducer surface, which makes segment the liquid sample stream as a means to prevent
it difficult for these sensors to detect environmentally uncontrolled dispersion
relevant levels; that is, to detect trace levels of a chemical More recently, in situ CFA monitors have been
species, the equilibrium binding (formation) constants developed that are not merely ruggedized bench-top
must be very large. For the reaction of analyte A with instruments. One of the more interesting instruments,
ligand B to form a complex AB, the rate equations are (3): and the model which inspired this research, was described
in work published in 1994 by Jannasch et al. at the
aA + bB → ABrate = −(d/dt) [A] = −(d/dt) [B] Moss Landing Marine Laboratory in Monterey, CA (11).
This group of chemical oceanographers was interested
= k1 [A]a [B]b
in accurately measuring nitrate ion in seawater with
xAB → A + Brate = −(d/dt) [AB] = k−1 [AB]x high sampling frequency. To accomplish this goal, they
designed a CFA monitor ‘‘from the ground up’’ and
The equilibrium constant for the dissociation of the incorporated several novel features. In doing so, they
complex is equal to the ratio of the rate constants for created a new paradigm in the science of chemical
the dissociation of the analyte-ligand complex and the measurements.
formation of the complex, as shown in the following In the system developed by Jannasch et al., a well-
equation (3): characterized, irreversible colorimetric method was used
KD = k−1 /k1 to measure nitrate. The dissolved nitrate was determined
by reduction to nitrite in a neutrally buffered (imidazole)
If the complex dissociates slowly, the time required to solution by an in-line copperized cadmium surface. The
return the signal to a steady baseline (‘‘clear down’’ nitrite then merged with an incoming reagent stream
time) is introduced. One way to avoid this problem containing sulfanilamide (p-aminobenzenesulfonamide or
IN SITU CHEMICAL MONITORING 515
PABSA) in acid to undergo reaction to form a diazo- In 1999, Byrne and coworkers at the University of
nium ion. This species was then reacted with N-(1- South Florida developed a monitor based on long path-
naphthyl)ethylenediamine dihydrochloride (NED) to form length absorbance spectroscopy (LPAS), which could
an azo dye of high molar absorptivity that was detected by be used for the determination of nitrate and several
molecular absorbance spectroscopy. other biologically important chemicals in seawater at
The novel in situ monitor that the Moss Landing much lower levels. These workers used the same well-
group developed used nonmechanical osmotic pumps to characterized chemistry as Jannasch et al., but they
both deliver reagents to the system and to propel the benefited from a 1-m liquid core waveguide as the
continuous flow sample stream through the microscale absorbance flow cell. Another difference with the Byrne
manifold (∼1 mm diameter conduits) (11). For the two group’s approach was in the use of mechanical (peristaltic)
reagent osmotic pumps (i.e., for PABSA and NED), the flow pumps instead of osmotic pumps—resulting in a five-
was driven by the osmotic pressure differential developed fold increase in response time. This monitor was field
across the semipermeable membrane that separated a tested recently in the Gulf of Mexico, and detection
saturated brine solution from a solution with deionized limits for nitrate were observed to be approximately
water (12). The pressure created from the diffusion of 2 nM (13,14).
deionized water across the semipermeable membranes The next key development was reported recently by
resulted in compression of inner bags containing the Worsfold and coworkers at the University of Plymouth.
reagents and their pulse-free delivery to the manifold (at A portable in situ monitor for the determination of trace
∼4 µL per hr). For the sample pump, conventional osmotic phosphate in surface waters was developed (15,16). This
pumps were modified by removing the inner bags and work demonstrated a monitor that was built with a
mounting the rigid permeable membranes in an enclosed forged plate similar in design to the work done by
container. Four of these modified osmotic pumps were Jannasch et al. The phosphate monitor had a throughput
then coupled to draw sample through the manifold at of ∼225 determination per hr and a detection limit of
0.15 µM with 2.0% RSD (n = 9) reported. This monitor
12 µL per hr.
was evaluated by mapping the phosphate concentration in
Microscale flow manifolds can be fabricated by ‘‘cold-
Port Phillip Bay in Australia for a brief (2 hr) period. The
forging’’ PVC: A wire bent in the desired flow microconduit
agreement between determinations obtained using the
configuration was pressed onto the surface of a flat-
in situ monitor and data obtained using manual sampling
milled sheet of PVC. Following removal of the wire,
aboard the ship followed by laboratory analysis was within
∼0.5-mm diameter holes were drilled into the plate for
±10%.
connection of components. The optical system in the Moss
Landing nitrate analyzer used a simple light-emitting
diode (LED) source with photodiode detector, with data INSTRUMENT DEVELOPMENT FOR IN SITU MONITORING
collected by a low-power programmable datalogger. The OF CR(VI)
monitor exhibited a detection limit of 0.11 µM and a
linear response to 20 µM nitrate. Measurements were
For the in situ measurement of relatively low levels
taken every 2 min and the response time was ∼30 min.
of important biological nutrients such as nitrate or
Quality control samples were measured automatically
phosphate, the work of the Jannasch, Byrne, and Worsfold
every 48 hours. Jannasch et al. demonstrated the in situ groups (and others) was pioneering. However, clearly
monitor’s stability by operating it continuously for the next challenge in developing in situ monitoring
month-long periods in large tanks at the Monterey instruments was the ability to measure relevant levels
Bay Aquarium. of trace metal ions in surface waters by the in situ
The Moss Landing group’s research incorporated approach. Hexavalent Cr was chosen as the target
several novel features—mainly the microscale flow design analyte because it is regulated at nanomolar levels (low
using nonmechanical osmotic pumps for reagent delivery parts per million) in surface waters. In addition, the
and sample intake—into an in situ monitor capable of metropolitan-industrial areas of southern Lake Michigan,
frequent and reliable measurements in environmentally specifically the Milwaukee Harbor area, contain industrial
complex samples. In addition, through the periodic sources that generate a concentration of total Cr in
addition of nitrate standards, the monitor automatically surface water that ranges from 0.1 to greater than
self-calibrated by correcting for baseline drift. Another 100 µg/L (17). These reasons, coupled with the well-
advantage to this system was its flexibility; using different characterized spectrometric chemistry (18–21), made Cr
irreversible chemistries, the in situ monitor could be a logical choice as a representative analyte to study for
readily adapted to measure other analytes for optical the development of an in situ monitor that measures trace
detection and quantitation. A drawback of the system metals in the environment.
was that it detected relatively high levels of analyte A benchtop method for determining hexavalent
(i.e., micromolar levels), albeit levels that are relevant to chromium in high particulate-containing surface waters
studying this important nutrient. Concerns over particles by sequential injection (SI) was developed (22). For utility
in the water eventually clogging the microconduit channels in particulate-rich waters, the relative performance of two
are less serious, because small particles (approx. <50 µm) membrane-based methods for sampling was compared.
cannot be drawn into the manifold at the low flow The first membrane approach was based on a commer-
rates used. cial design known as the ‘‘supported capillary membrane
516 IN SITU CHEMICAL MONITORING
sampler (SCMS)’’ that uses tubular membranes; the sec- 4. DeGrandpre, M.D., Baehr, M.M., and Hammar, T.R. (1999).
ond approach was based on a conventional parallel-plate Calibration-free optical chemical sensors. Analytical Chem.
dialyzer (PPD) design that uses planar membranes. The 71: 1152.
membranes were evaluated using the colorimetric method 5. Inman, S.M., Stromvall, E.J., and Lieberman, S.H. (1989).
for the determination of hexavalent chromium by com- Pressurized membrane indicator system for fluorogenic-based
fiber-optic chemical sensors. Analytica. Chimica. Acta. 217:
plexation with 1,5-diphenylcarbazide (DPC). Thin-walled
249.
(∼200 µm) microporous (pore size ∼0.2 µm) polypropy-
6. Degrandpre, M.D. (1993). Measurement of seawater pCO2
lene membranes were equilibrated with DPC during each
using a renewable-reagent fiber optic sensor with colorimetric
sampling period. Formation of the DPC-Cr(VI) complex detection. Analytical Chem. 65: 331.
allowed for efficient membrane transport; without the
7. Ruzicka, J. and Hansen, E.H. (1988). Flow Injection Analysis,
membrane, Cr(VI) transport decreased by ∼90%. Factors 2nd Edn. John Wiley & Sons, New York.
optimized included reagent concentration, sampling time, 8. Johnson, K.S., Coale, K.H., and Jannasch, H.W. (1992).
flow rate, and spectrophotometric conditions. Optimal con- Analytical chemistry in oceanography. Anal. Chem. 64:
ditions for the reagents were 2.00 mM DPC and 0.100 M 1065A.
nitric acid, and sampling times of 600 s and 900 s for 9. Valcarcel, M., Cardenas, S., and Gallego, M. (2002). Contin-
the planar and tubular designs, respectively. The planar uous flow systems for rapid sample screening. Trac-Trends
(PPD) design increased the sensitivity relative to the tubu- Analytic Chem. 21: 251.
lar (SCMS) design by ∼225%. The PPD-SI method was 10. Elliott, C.L., Snyder, G.H., and Cisar, J.L. (1989). A modified
applied to the determination of dissolved Cr(VI) in high Autoanalyzer-II method for the determination of NO3 − in
particulate-containing surface water samples. Figures-of- water using a hollow-Cd reduction coil. Communications in
merit included a detection limit of <20 µg/L, precision of Soil Science and Plant Analysis 20: 1873.
1.1% relative standard deviation at 100 µg/L (n = 4), and 11. Jannasch, H.W., Johnson, K.S., and Sakamoto, C.M. (1994).
selectivity for dissolved Cr(VI) in several surface water Submersible, osmotically pumped analyzers for continuous
samples with high levels of particulate matter and poten- determination of nitrate in-situ. Analytic Chem. 66: 3352.
tial interferences. 12. Pump, A.O. (2000). Technical Note. Alzet, Inc.
The optimized method conditions from the SI studies 13. Kaltenbacher, E.A., Byrne, R.H., and Steimle, E.T. (2001).
was then applied to the design, fabrication, and testing of Design and applications of a chemical sensor compatible
with autonomous ocean-sampling networks. IEEE J. Oceanic
an in situ remote monitor for the measurement of sub-part
Engineer. 26: 667.
per million levels in surface waters (23,24). The monitor
14. Bryne, R.K. (1999). Autonomous in situ analysis of the upper
was based on a continuous flow analysis (CFA) design
ocean. Sea Technol. 71.
in which a narrow conduit (1.0 mm) manifold was hot-
15. Hanrahan, G., Gledhill, M., Fletcher, P.J., and Worsfold, P.J.
forged between Plexiglas plates. A miniature low-power
(2001). High temporal resolution field monitoring of phos-
peristaltic pump was used to draw sample (1.5 mL/hr) phate in the River Frome using flow injection with diode
into the manifold where chromium (VI) in the sample array detection. Analytica. Chimica. Acta. 440: 55.
was derivatized to its diphenylcarbazide complex. The 16. Lyddy-Meaney, A.J., Ellis, P.S., Worsfold, P.J., Butler,
absorbance of the complex was then measured in a E.C.V., and McKelvie, I.D. (2002). A compact flow injection
light-emitting diode/photodiode detection cell. Solenoid- analysis system for surface mapping of phosphate in marine
actuated osmotic pumps were used to introduce quality waters. Talanta. 58: 1043.
control standards at user-determined periodicities to 17. Environmental Defense Fund. (2004). Available: http://
achieve acceptable data quality. Figures-of-merit included www.scorecard.org/.
a detection limit of 0.25 mg/L (7 min cycle time), precision 18. Miller-Ihli, N.J. (1992). Chromium. Technol. Instrum. Anal.
of 2.5% relative standard deviation at 1 mg/L (n = 3), Chem. 12: 373.
and selectivity for dissolved Cr(VI) in several surface 19. Lynch, T.K. and Wilson, J. (1984). Speciation of metals
water samples with moderate levels of particulate matter. in solution by flow injection analysis. Part 1: Sequential
Continuous testing of the monitor over a three-week spectrophotometric and atomic absorption detectors. Analyst.
period using Milwaukee Harbor water fortified with Cr(VI) 109: 839.
yielded results that were in close agreement (±5%) with 20. Losi, M.E., Amrhein, C., and Frankenberger, W.T. (1994).
an ICP-MS reference method. Rev. Environ. Contamin. Toxicol. 136: 91.
21. Kotas, J. and Stasicka, Z. (2000). Chromium occurrence in the
environment and methods of its speciation. Environ. Pollut.
107: 263.
BIBLIOGRAPHY
22. Pressman, M.A. and Aldstadt, J.H. (2003). A comparison
of diffusion samplers for the determination of hexava-
1. Hanrahan, G., Ussher, S., Gledhill, M., Achterberg, E.P., lent chromium by sequential injection spectrophotometry.
and Worsfold, P.J. (2002). High temporal and spatial Microchem. J. 74: 47–57.
resolution environmental monitoring using flow injection with 23. Pressman, M.A. (2003). Development of Novel Analytical
spectroscopic detection. TrAC, Trends Analytic Chem. 21: 233. Techniques and Methods for the Determination of Trace
2. Rawls, R. (1998). Wanted: instruments to view ocean’s Chromium in the Environment, Ph.D. dissertation, University
dynamics. Chem. Engineer. News. 31. of Wisconsin-Milwaukee.
3. Johnson, K.S. and Jannasch, H.W. (1994). Analytical chem- 24. Pressman, M.A. and Aldstadt, J.H. (2005). Development and
istry under the sea surface: monitoring ocean chemistry Testing of an in situ Chemical Monitor for Quantitation of
in situ. Naval Research Rev. 3: 4. Hexavalent Chromium in Natural Waters, in preparation.
NITROGEN 517
NITROGEN to be mineralized in the soil by a chain of microbial

processes (Fig. 1). The release of ammonium and ammonia
KENNETH F. STEELE (total ammonia) from organic matter is referred to as
KRISTOFOR R. BRYE mineralization (3) or ammonification. Total ammonia is
University of Arkansas readily oxidized to nitrate. The conversion of nitrogen gas
Fayetteville, Arkansas to fixed form and from fixed form to nitrogen gas includes
four main conversion reactions that form the essential
parts of the nitrogen cycle (Fig. 1):
INTRODUCTION
1. fixation
Nitrogen (N) is ubiquitous in the environment and in 2. ammonification
the correct form is an essential nutrient for plants and 3. nitrification
animals. About 26.5% of Earth’s nitrogen is in rocks and
essentially all of the remainder (73%) is in the atmosphere 4. denitrification
as an inert monotomic gas (N2 ) that comprises 78% of
the atmosphere (1). Only a small amount of nitrogen is Fixation
present in the hydrosphere and biosphere, but nitrogen For N2 gas to be used by plants, it must first be fixed
is a crucial nutrient there. For nitrogen to be used by with hydrogen to form ammonium ions. Fixation is a
organisms, it must be in a usable form; it must be combined microbial process, and even the few types of plants that
or ‘‘fixed’’ with oxygen or hydrogen. The most common can use N2 directly are ultimately dependent on microbes
fixed form of nitrogen in ground water is the anionic also. For example, legumes can use N2 because of the
oxide—nitrate (NO3 − ), and the second most common form symbiotic relationship between the plant and Rhizobium
is the cationic form–ammonium (NH4 + ). Approximately bacteria that live in nodules in the plant roots. Other
85% of the nitrogen in stream water is in the form of means of fixation include conversion of atmospheric N2
organic N and most of the rest (NO3 − and NH4 + ) is to nitrate by lightning and industrial processes, such as
derived from the decomposition of organic matter (2). the Haber–Bosch process for producing ammonia (NH3
N2 is converted to organic nitrogen predominantly by gas) (4).
microorganisms and certain plants. About 96% of the
nitrogen in soil is bound in organic matter, and nitrate Ammonification
is a product of organic matter degradation (3). Nitrogen
occurs in five oxidation states, −3, 0, +1, +3 and +5, that Ammonification (also referred to as mineralization), the
are primarily the result of bacteria-induced oxidation. The process of converting organic nitrogen (e.g., urea) to
nitrogen species are listed in Table 1. ammonium (Eq. 1), generally occurs during decomposition
of animal and plant matter (4). Dissolution of ammonium
carbonate releases ammonium ions:
NITROGEN CYCLE
H2 NCONH2 +2H2 O −−−→ (NH4 )2 CO3 (1)
Fixation by plants can occur; leguminous plants convert Urea Ammonium carbonate
N2 to organic nitrogen compounds, but most plants cannot
use N2 directly. Most plants require atmospheric nitrogen As indicated by Eqs. 2 and 3, the relationship between
converted to ammonium or nitrate ions (4). Organic NH3 and NH4 + is pH dependent. Under alkaline
nitrogen is not directly available to plant roots and has conditions (pH > 9), formation of NH3 is favored due to
neutralization of NH4 + by hydroxide (OH− ) (Eq. 2). In
acid conditions (Eq. 3), ammonia gas is transformed into
Table 1. Common Forms of Nitrogen in the Environment
and Typical Concentrations of Nitrogen
Typical Nitrogen gas
Oxidation Concentration in N fixation N2 (g)
Species State Typical Source Water as N (e.g., NH3 formed)
Denitrification—
N2 0 Atmosphere 14 mg/L Immobilization NO3−
NO3 − +5 Oxidation of <0.6 mg/L background (conversion to protein)
organic matter/ to 250 mg/L for
atmosphere contaminated Nitrate
Organic N
sources NO3−
NO2 − +3 Intermediate form <0.1 mg/L
in nitrification,
denitrification, Ammonification
Nitrification—
(mineralization)
and reduction of NO2−
NO3 −
Ammonium
NH3 −3 Fixation of N2 <0.01 mg/L
NH4+
NH4 + −3 Ionization of NH3 <0.01 mg/L
N2 O +1 Reduction of NO2 − Negligible
Figure 1. Simplified nitrogen cycle.
518 NITROGEN
ammonium ions. For example, at 15 ◦ C, ammonia gas Analytical Terminology

makes up 0.027% of total ammonia (NH3 + NH4 + ) at a pH
Analytically, organic nitrogen (proteins, peptides, nucleic
of 6.0, but makes up 2.7% of total ammonia at a pH of
acid, urea, and other compounds) and ammonia are
8.0. Although 2.7% is a small amount, it is a hundredfold
determined together and are referred to as ‘‘Kjeldjahl
increase of ammonia (5).
nitrogen,’’ a term that refers to the analytical method
used. By expressing all nitrogen species as N, the
NH3 + H2 O = NH4 + + OH− (2)
concentrations of the various species can be added and
NH3 + H+ = NH4 + (3) subtracted. Organic N (−3 charge) is determined by
analyzing ammonia by other methods and subtracting
Other factors that affect the concentration of NH3 in water this value from the Kjeldhal nitrogen concentration (10).
are temperature and ionic strength. The concentration of Inorganic N is the sum of nitrate, nitrite, and total
NH3 increases as temperature increases and decreases as ammonia (ammonia plus ammonium). Some analytical
ionic strength increases. At 15 ◦ C and pH 8.0, the percent methods determine the sum of nitrate and nitrite, but
NH3 in solution is 2.7%, whereas, at 30 ◦ C and pH 8.0, the results are sometimes incorrectly reported as nitrate
the concentration increases to 7.5% (5). The effect of ionic because nitrite concentrations are considered very low.
strength is minor for fresh to slightly saline waters. Total N is the sum of organic N and inorganic N.
Nitrification AQUATIC NITROGEN

Nitrification is the biological oxidation of ammonium ions
to nitrite followed by conversion of nitrite to nitrate. Despite the large proportion of nitrogen in the atmosphere,
Biological oxidation is carried out by Nitrosomonas and its abundance in water is low-measured as a few mg/L
Nitrobacter bacteria (6,7) as shown in Eqs. 4 and 5. In N, mostly in the inert form of nitrogen. Typically, the
addition to these two bacteria, (8) indicate that several most environmentally important form of nitrogen is
different autotrophic organisms mediate nitrification: nitrate; however, dissolved organic nitrogen makes up the
majority of total dissolved nitrogen in watersheds in New
Nitrosomonas England (11). Note that for primary producers ammonium
2NH4 + + 3O2 −−−−−−−→ 2NO2 − (nitrite) + 4H+ may be more available to organisms than nitrate (12). The
bacteria
various common species of nitrogen in the environment
+ 2H2 O (4) and the estimated concentration of each are presented
in Table 1. Mueller and others (13) have demonstrated
Nitrobacter that background concentrations of nitrate as N are less
2NO2 − + O2 −−−−−−→ 2NO3 − (nitrate) (5)
bacteria that 0.6 mg/L.
Nitrogen species are subject to oxidation/reduction
Because oxidation of nitrite to nitrate is a rapid (redox) and reactions with hydrogen ions (pH). In Fig. 2,
reaction and nitrite is metastable, nitrite is not abundant the NO3 − /NO2 − pe–pH boundary is dotted to indicate that
(Table 1) (9). Note that if both oxygen and nitrate are NO2 − is metastable. Nitrate, at a pe above the NO3 − /NO2 −
present in soil, bacteria preferentially use oxygen because boundary, is reduced at the NO3 − /N2 boundary to form
it yields more energy (1). Therefore, nitrification typically nitrogen. Thus, there is little NO3 − available to form
occurs in oxic conditions, whereas denitrification occurs in NO2 − . Because nitrite is a metastable species, it occurs
anoxic conditions. only in low concentrations in natural waters where
denitrification is occurring (9). Nitrate is present in the
Denitrification N2 field, but under natural conditions, it is present in
concentrations less than those of N2 . Ammonium ions are
Denitrification is the biological reduction of nitrate the dominant nitrogen species at low pe and pH. Ammonia
to nitrogen gas by heterotrophic bacteria, such as is the dominant species above pH 9 (Fig. 2) and thus is not
Pseudomonas denitrificans and Thiobacillus denitrifi- present in significant concentrations in natural waters;
cans (8). These bacteria require organic carbon for energy most natural waters have a pH range of 5–9. Dissolved
(Equations 6 and 7). organic N can also be relatively mobile.
Heterotrophic bacteria
NO3 − + organic carbon −−−−−−−−−−−−→ NO2 − (6) Mobility
Nitrate is very soluble in water and because of its negative
Heterotrophic bacteria
NO2 − + organic carbon −−−−−−−−−−−−→ N2 + CO2 charge, is not adsorbed by humus or clay. Nitrate not
taken up by plants can be transported from the upper soil
zone to lower soil zones and eventually to groundwater.
+ H2 O (7)
Ammonium ions are not nearly as mobile as nitrate ions, in
part, because their positive charge allows their transport
Because oxidation of organic carbon by oxygen yields more to be retarded by adsorption on humus and clay. In some
energy, generally the oxygen concentration must be very situations, ammonium ions can be transported to deep soil
low (anoxic conditions) for nitrate to be used (4,6). horizons (3).
NITROGEN 519
be life threatening for children less than 3 years old.

20 Babies who consume water with nitrate concentrations
as low as 1 mg/L as N risk methemoglobinia or ‘‘blue
baby disease’’ because of the impairment of oxygen
16 transport. There are also indications that nitrate (as
related to nitroso compounds) may be associated with
gastric cancer and perhaps liver and kidney cancer. The
NO3− relative consumption of nitrate from water compared to
12
consumption of dietary nitrate is a factor that concerns
researchers attempting to determine the risk of cancer
NO3−
O2 from water (15,16). The maximum concentration limit
8 NO2−
H2O (MCL) of 10 mg/L N for nitrate in drinking water is based
on minimizing methemoglobinia and is considered safe
concerning cancer.
pe
N2(g) CONTAMINATION
0 H2O
H2
The causes of pollution of water are divided into point
NH4+ sources and nonpoint sources. Nitrogen contamination of
−4 water is generally related to nitrate from sewage and septic
systems (point sources) and from agricultural sources.
The agricultural sources are commercial and animal
−8 manure fertilizers (nonpoint sources) (17,18) and from
confined animal operations (point sources). Agriculture is
generally considered the main source of nitrogen pollution
−12 NH30 of water (19). During rainstorms, streams often exhibit
a dilution of nitrate concentration that indicates that
0 2 4 6 8 10 12 14 most of the nitrate is transported to these streams by
pH groundwater. Excessive nutrients in runoff can result in
Figure 2. Stability fields for common nitrogen species in natural eutrophication of surface waters, for example, the Neuse
waters. The diagram is based on the following conditions: River in North Carolina, U.S.A. (20). Best management
temperature of 25 ◦ C, partial pressures of H2 and O2 are each practices have been developed to minimize the negative
1 atmosphere, partial pressure of N2 is 0.77 atmosphere, and impacts of fertilizers (17).
total aqueous N concentration is 0.001 moles per liter.
ISOTOPES
Toxicity
There are two stable isotopes of nitrogen, 14 N and 15 N.
Ammonia is a pungent, colorless gas that is very soluble Although the whole earth abundance of 14 N is 99.63‰ of
in water and is toxic to fish at concentrations greater than total nitrogen, due to fractionation, the 15 N:14 N ratio is
0.2 mg/L. Dissolved ammonia reacts with water to form 0.003677 (21) in the atmosphere. Differences in the ratio
ammonium ions, which are not nearly as toxic (1/50th) of the two isotopes in the various nitrogen species are a
as the gas (3). The ammonia–ammonium equilibrium is result of the faster chemical reaction of the lighter isotope
controlled to a large extent by temperature and pH; only than the heavier isotope. Isotopic ratios of nitrogen are
highly alkaline waters favor the formation of ammonia. expressed as
The pH range of most natural waters is such that
ammonium ions are dominant compared to ammonia δ 15 N = [(Rsample /Rair ) − 1] × 1000
gas (Fig. 2). The U.S. Environmental Protection Agency
bases its ammonia criteria for water on a complex set of where R is the abundance ratio of 15 N to 14 N. The
equations that include the effect of temperature and pH on international standard for δ 15 N is air, which by definition
ammonia concentration. The equations were developed to has a δ 15 N of 0‰. Values greater than 0‰ indicate
account for acute and chronic toxicity to particular types of enrichment in 15 N, and values less than 0‰ indicate
fish. In addition, the criteria are based on specific periods depletion of 15 N. Sources of nitrate have been determined
of time and the frequency of exceeding these criteria (14). using δ 15 N. The premise is that agricultural activity
Typically nitrate is toxic to warm-blooded animals only and sewage (including septic tank effluent) may have
where large amounts of nitrite are formed by the reduction distinct δ 15 N values. A definitive determination of source
of nitrate. Nitrate can be reduced in the gastrointestinal is sometimes thwarted by the possibility of overlapping
tract to nitrite, especially in infants. Nitrite enters the δ 15 N values between possible sources and changes in
bloodstream and reacts with hemoglobin to produce δ 15 N due to volatilization of NH3 from the soil and
methemoglobin that reduces the number of hemoglobin denitrification of nitrate in water. The origin of the nitrate
cells available for oxygen exchange. These reactions can in water can be found by using δ 18 O to recognize different
520 OSMOSIS-DIFFUSION OF SOLVENT OR CAUSED BY DIFFUSION OF SOLUTES?
sources of nitrate and the influence of denitrification on 19. Spalding, R. and Exner, M. (1993). Occurrence of nitrate in
δ 15 N (21,22). A number of studies have had success using groundwater—A review. J. Environ. Qual. 22: 392–402.
N and often O isotopes to determine the source of nitrate 20. Paerl, H.W., Pinckney, J.L., Fear, J.M., and Peierls, B.L.
contamination (23,24). (1998). Ecosystem responses to internal and watershed
organic matter loading: Consequences for hypoxia in the
eutrophying Neuse River estuary, North Carolina, USA. Mar.
BIBLIOGRAPHY Ecol. Prog. Ser. 166: 17–25.
21. Clark, I.D. and Fritz, P. (1997). Environmental Isotopes
1. Hitchon, B., Perkins, E.H., and Gunter, W.D. (1999). in Hydrogeology. Lewis, CRC Press, Boca Raton, FL,
Introduction to Ground Water Geochemistry. Geoscience, pp. 150–152.
Alberta, Canada, pp. 195–200. 22. Ostrom, N.E., Peggy, H., and Ostrom. (1999). Nitrogen
2. Berner, E.K. and Berner, R.A. (1987). The Global Water Cycle. isotopes. In: Encyclopedia of Geochemistry. C.P. Marshall
Prentice-Hall, Inc., Englewood Cliffs, NJ, p. 397. and R.W. Fairbridge (Eds.). Kluwer Academic, Dordrecht,
The Netherlands, pp. 431–433.
3. Canter, L.W. (1996). Nitrates in Groundwater. CRC Lewis,
Boca Raton, FL. 23. Karr, J.D., Showers, Hinson, W.J., and Thomas, H. (2002).
Nitrate source identification using d 15 N in a ground water
4. Jeurgens-Gschwind, S. (1989). Ground water nitrates in
plume near an intensive swine operation. Ground Water
other developed countries (Europe)–relationships to land
Monitoring and Remediation 68–75.
use patterns. In: Nitrogen Management and Ground Water
Protection. R.F. Follett (Ed.). Elsevier, New York, pp. 75–138. 24. Aravena, R., Evans, M.L., and Cherry, J.A. (1996). Stable
isotopes of oxygen and nitrogen in source identification of
5. U.S. EPA. (1976). Quality Criteria for Water. U.S. Environ-
nitrate from septic systems. Ground Water 31: 180–186.
mental Protection Agency, Washington, DC, pp. 10–13.
6. Manahan, S. (1999). Environmental Chemistry. CRC Press,
Boca Raton, FL.
OSMOSIS-DIFFUSION OF SOLVENT OR
7. Stumm, W. and Morgan, J.J. (1996). Aquatic Chemistry, 3rd
Edn. John Wiley & Sons, New York, pp. 927–934. CAUSED BY DIFFUSION OF SOLUTES?
8. Kendall, C. and Aravena, R. (1999). Nitrate isotopes in
KARL ERIK ZACHARIASSEN
groundwater systems. In: Environmental Tracers in Subsur-
face Hydrology. P.G. Cook and A.L. Herczeg (Eds.). Kluwer Norwegian University of
Academic, Boston, pp. 261–298. Science and Technology
Trondheim, Norway
9. Kehew, A.E. (2001). Applied Chemical Hydrogeology. Prentice-
Hall, Upper Saddle River, NJ.
10. Eaton, A.D., Lenore, C.S., and Greenberg, A.E. (Eds.).
Osmosis is the movement of water through a semiperme-
(1995). Standard Methods for the Examination of Water
and Wastewater. American Public Health Association, able membrane separating two solutions to the solution
Washington, DC, pp. 4-75–4-96. with the higher solute concentration. The membrane is
11. Campbell, J.L. et al. (2000). Dissolved organic nitrogen semipermeable in the sense that it is permeable to water,
budgets for upland, forested ecosystems in New England. but not to the solutes. The osmotic movement of water
Biogeochemistry 49: 123–142. can be blocked by the application on the high concentra-
12. Peterson, B.J. et al. (2001). Control of nitrogen export from tion side of a hydrostatic pressure, which counteracts the
watersheds by headwater streams. Science 292: 86–90. so-called osmotic pressure that moves the water through
13. Mueller, D.K., Hamilton, P.A., Helsel, D.R., Hitt, K.J., and the membrane.
Ruddy, B.C. (1995). Nutrients in ground water and surface The prevailing theory ascribes osmosis to a diffusion
water of the United States—an analysis of data through of water molecules through the membrane down a water
1992. U.S. Geological Survey Water-Resources Investigations concentration gradient across the membrane. The solutes
Report 95-4031, p. 74. are supposed to dilute the water, and the dilution will be
14. U.S. EPA. (1999). 1999 Update of Ambient Water Quality strongest at the side with the highest solute concentration.
Criteria for Ammonia. U.S. Environmental Protection Agency. However, several investigators claim that this explana-
National Service Center for Environmental Publications, tion fails to explain important characteristics of osmotic
Cincinnati, OH.
phenomena. Osmotic movement of water is considerably
15. Bogardi, I. and Kuzelka, R.D. (Eds.). (1991). Nitrate Con- faster than should be expected from a diffusive movement
tamination: Exposure, Consequence and Control. NATO ASI
of water. In addition, the water concentration of solutions
Series. Springer-Verlag, New York.
of different solutes are not correlated with the osmotic
16. Follett, R.F. and Walker, D.J. (1989). Ground water quality
activities of the solutions. Furthermore, the osmotic pres-
concerns about nitrogen. In: Nitrogen Management and
Ground Water Protection. R.F. Follett (Ed.). Elsevier, New
sure is related to the total concentrations of dissolved
York, pp. 1–22. particles, and van t’Hoff pointed out that the number of
17. Steele, K.F. (Ed.). (1995). Animal Waste and the Land–Water
solute particles and the solution volume are related to
Interface. CRC Press, Lewis, Boca Raton, FL. the osmotic pressure in the same way as the number of
18. Steele, K.F. and McCalister, W.K. (1991). Potential nitrate
molecules, pressure, and volume of a gas, i.e., osmotic
pollution of ground water in limestone terrrain by poultry lit- phenomena, follow the universal gas laws. No theory has
ter, Ozark Region, USA. In: Nitrate Contamination: Exposure, been proposed to explain the agreement between osmotic
Consequence and Control. I. Bogardi and R.D. Kuzelka (Eds.). phenomena and gas laws on the basis of diffusion of water
NATO ASI Series, Springer-Verlag, New York, pp. 209–218. down a water concentration gradient.
PARTITIONING AND BIOAVAILABILITY 521
In 1903, George Hulett (1) published a theory that osmotic and molal concentrations, which have m as the
explains the agreement between osmotic phenomena official unit.
and gas laws and that is in agreement with the other To add to the confusion, the official unit for osmolality
observations on osmotic phenomena. Hulett emphasized has led some authors to introduce Osmoles/L solution
that the solute particles diffuse inside solutions in the as a unit for osmolarity. However, the melting point is
same way as gas molecules diffuse within the space in claimed to be −1.86 ◦ C both for a 1 osmolal and for a 1
which the gas is distributed. In the same manner, as gas osmolar solution. Hence, the distinction between units
molecules create a pressure when they collide with the for osmolality and osmolarity is merely semantic and
gas boundary, solute particles create a pressure when of no practical consequence. Osmolality and osmolarity
they collide with the boundary of the solution. Both are different terms for the same thing, and furthermore,
for gases and solutions this pressure will tend to cause they are just quasi-measures for concentration. Solutions
the volume to expand, which is easy for a gas if the of different solutes with the same molar or molal
boundaries can yield to the pressure and allow expansion concentrations all have different osmolalities, and for
of the volume. However, the water in which the solutes of a given type of solute, no exact proportionality exists
a solution are distributed is essentially incompressible, between molal concentrations on one side and colligative
and the expansive pressure created by the diffusing parameters on the other, which is at least in part because
solutes cannot lead to an expansion unless more water the amount of hydration water associated with each solute
can enter the solution and increase the solution volume. particle differs substantially from substance to substance,
The semipermeable membrane of osmotic processes allows and for dissolved salts, the degree of dissociation varies
such an influx of water from a solution where a lower from substance to substance and with their concentration.
solute concentration creates a lower expanding pressure, Although osmolality is not a precise chemical concentra-
whereas the solutes that create the pressure are prevented tion measure, it is a useful tool for comparing total solute
from passing. Water will subsequently move across the concentrations of fluids in physiology, and it is a useful
membrane until the pressures on both sides are equal. measure to control the chemical solute concentrations of
According to the gas laws, an ideal 1 molar solution body fluids and solutions.
(1 mole of particles dissolved in a volume of 1 L), which
is separated from pure water by a semipermeable, will BIBLIOGRAPHY
create a pressure of 22.1 atm, which corresponds to a water
column of impressive 224 m. It will have a melting point 1. Hulett, G. (1903). Beziehung zwischen negativem Druck und
of −1.86 ◦ C, and a solution with this osmotic concentration osmootischen Druck. Z. Phys. Chem. 42: 353–368.
is said to be a 1 osmolar solution. Hammel, H.T. and Scholander, P.F. (1976). Osmosis and
Proportionality exists among the osmolality, the Tensile Solvent. Springer Verlag, Berlin.
osmotic pressure, and the melting point depression. This
proportional relationship with osmotic pressure is also
displayed by vapor pressure and boiling point elevation. PARTITIONING AND BIOAVAILABILITY
Because of the exact proportionality that exists among
these parameters, they are frequently referred to as JIM PHILP
the colligative parameters (from the Latin colligare = FIONA STAINSBY
bind together). SANDRA DUNBAR
A closer inspection of the relationship between Napier University
mole-based concentration measures and the colligative Edinburgh, Scotland, United
properties reveals that it is the molal concentration that is Kingdom
most closely related to the colligative parameters. For this
reason, some investigators prefer the term osmolality.
INTRODUCTION
UNITS FOR OSMOTIC CONCENTRATIONS As soon as a chemical is spilled it becomes a pollutant,

but the fate of that pollutant depends to a large degree
Some confusion exists about the units for osmotic on its chemistry, which determines how it will partition to
concentration. The officially approved unit for osmolality various ‘‘compartments,’’ or phases, of the environment.
is Osmoles/kg water. Some investigators find that this This idea is enshrined in risk-based strategies for
unit is misleading, because it invites us to consider an assessing contaminated land and water. In many cases
Osmole as a defined amount of solute, which when diluted a contaminant spill on land quickly becomes a water
in 1 kg of water, gives an osmolality of 1 Osmole/kg pollution issue, be it surface or groundwater. As many
water. Accordingly, when diluted in 2 kg of water, the countries are heavily reliant on groundwater as a source
same osmole of solutes should give an osmolality of 0.5 of potable water, investigating the fate of pollutants is
Osmole/kg water, but this is never the case. An Osmole an essential prerequisite to any remedial action plan for
is not a defined amount of solute, and the concept of contaminant cleanup.
osmolality has a meaning only for solutions. For this Partitioning is governed by some fundamental prop-
reason, some investigators use the unit Osm, where erties of chemicals, but partitioning should be viewed in
the letter m emphasizes the close relationship between relation to the toxicity of the chemical in question and its
522 PARTITIONING AND BIOAVAILABILITY
degradability in the various environmental compartments. n = number of moles of the chemical

In this article, the important properties of chemicals that R = gas constant
relate to partitioning are examined, and the theory is T = absolute temperature
reinforced by example. The examples chosen are ben-
zene, phenol, and some chlorinated phenols. All of these Another way to think of vapor pressure is that it is
compounds are important industrial chemicals. the pressure that a vapor exerts on the atmosphere:
Bioavailability is a simple concept, but practically it thus it will be clear that it is the driving force for
is a complex issue that remains to be resolved. The volatilization. Vapor pressures of organic compounds
bioavailability concept is also enshrined in the risk are often compared to the vapor pressure of mercury
assessment approach to contaminated land and water. If a (atmospheric pressure is 760 mmHg). Thus the vapor
pollutant is present in soil or water in high concentration, pressure of benzene is 87.2 mmHg at 25 ◦ C. Volatilization
but is not available to the biota, then surely it presents potential can be misleading in the absence of information
minimal risk. Regardless of the apparent simplicity of on the water solubility of the compound, however. A
the concept, there is great urgency to find methods for compound that has a low vapor pressure may still have
quantifying bioavailability. Some approaches to solving a high tendency to ‘‘escape’’ if the water solubility is
this quandary are discussed. low. For example in Table 1, 2,4,6-trichlorophenol has
a much lower vapor pressure than 3-chlorophenol, and
yet its tendency to escape to air is higher. This can be
PROPERTIES THAT DETERMINE PARTITIONING explained when the water solubilities are compared: the
much lower value for 2,4,6-trichlorophenol results in it
The compartments of the environment relevant to having a Henry’s constant about a magnitude order higher
partitioning are air, water, soil, and sediment and, through than 3-chlorophenol, even though its vapor pressure is
bioaccumulation/magnification through food chains, biota. much lower.
Partitioning into the biota is beyond the scope of this
article, and it will deal principally with partitioning to Henry’s Law Constant
nonliving compartments. Figure 1 represents a common
To study partition phenomena, then, Henry’s law constant
scenario in environmental remediation, contamination
is more valuable than vapor pressure. Henry’s law
from old gas stations. constant is the ratio of the equilibrium concentration
of a compound in air and its equilibrium concentration
Vapor Pressure
in water. It can therefore be regarded as the partition
This is the partial pressure of a chemical in a gas phase coefficient between air and water, and is the best indicator
that is in equilibrium with the pure liquid or solid phase of the tendency of a chemical to volatilize from water.
of the chemical. Vapor pressure is greatly influenced by Henry’s law constant can be written as:
temperature, and figures are usually quoted at a constant
P
temperature (usually 20 ◦ C or 25 ◦ C). If the volume of the H= (2)
containing vessel is also known, the vapor pressure can be S
used to calculate the concentration of the chemical in the where H = Henry’s law constant (atm-m3 /mol)
headspace of the vessel from the ideal gas law:
P = vapor pressure (atm)
n P S = water solubility (mol/m3 )
= (1)
V RT Water Solubility
where P = vapor pressure (atm) This is of central importance to understanding partition
V = volume phenomena in the environment. It is the maximum, or
Direct contact Public area

soil/vegetable
intake
River
Vapours Vapours Vapours

Operational
spills Fuel tank
Contaminated soil
Contaminated soil Underground
Figure 1. Partitioning of fuel as a con- leaks
taminant from a gas station that has
Fuel on groundwater surface
a leaking underground storage tank.
Because fuels are multicomponent liquids,
different components partition to different Fuel dissolved in groundwater
compartments of the environment.
Table 1. Important Partition Variables for a Range of Compounds. The Figures were Computed from U.S. EPA Software
(EPI Suite v.3.11 and ECOSAR)
Vapor Level III
Pressure, Water Fugacity Toxicity Fish Biodeg.
mmHg, Henry’s Solubility, Log Log Mass Amount, 96-h LC50 , Probability,
25 ◦ C Constant mg/L Kow Koc % mg/L MITI Non-Linear
Benzene 87.2 5.55 × 10−3 1790 2.13 2.22 Air: 37.6 59.2 .73
Water: 48.4
Soil: 13.8
Sediment: 0.20
Phenol 0.323 3.33 × 10−7 77, 900 1.46 2.43 Air: 1.33 27.7 .71
Water: 47.3
Soil: 51.3
Sediment: 0.098
3-chloro-phenol 0.119 3.45 × 10−7 26, 000 2.50 2.64 Air: 1.35 15 0.39
Water: 39.4
Soil: 59.1
Sediment: 0.21
2,4,6-trichloro-phenol 0.00536 2.6 × 10−6 800 3.69 3.07 Air: 2.55 3.7 0.043
Water: 22.9
Soil: 72.4
Sediment: 2.17
Pentachloro-phenol 1.08 × 10−5 2.45 × 10−8 14 5.12 3.53 Air: 0.0665 0.8 .0031
Water: 5.59
Soil: 70.9
Sediment: 23.4
saturation, concentration of a compound that dissolves in because the halogens increase the molecular volume of
water. It also is significantly affected by temperature. As organic compounds (1).
the water solubility of compounds is defined by a host of Other chemical functional groups affect water solu-
factors, comparisons should always be made at a constant bility in rather specific ways. The presence of polar
temperature, usually 25 ◦ C. In natural water systems, the functional groups tends to increase the water solubility of
concentration of pollutant chemicals is usually rather low, molecules compared to unsubstituted hydrocarbons: alco-
making this especially important. hols, amines, ethers, ketones, and organic acids are good
Polarity refers to the extent to which charge is unevenly examples. Even with these examples, though, a general
distributed within a molecule. Water itself is highly polar. rule of increasing molecular size and decreasing water sol-
Therefore charged or highly polar organic and inorganic ubility still exists. The short-chain fatty acid acetic acid is
molecules, hydrophilic molecules, are readily soluble in much more water soluble than a long-chain fatty acid. The
water. Conversely, nonpolar, hydrophobic molecules find additions to the molecule are not more polar functional
solution in water a less energetically favorable state, and groups but simply more nonpolar C–H2 groups that make
they have low water solubility. In the examples in Table 1, the molecule larger but less polar.
benzene has moderate water solubility, but the addition The situation is further complicated by more specific
of the polar hydroxyl group makes phenol much more chemical reactions. The chlorophenols in Table 1 illustrate
water soluble. The solubility of any solute in any solvent this point. Chlorophenols behave as weak acids when
is mediated by forces of attraction; the main ones are: dissolved in water due to the loss of a proton from
the hydroxyl group to leave a phenoxide ion. The more
• van der Waals forces; chlorines that are added to the ring, the more electrons
• hydrogen bonding; are withdrawn from the electron-rich phenoxide ion. This
• dipole–dipole interactions. lowers the phenoxide negative charge and therefore lowers
its ability to hold a proton. Thus the more chlorines on
The size and shape of a molecule is also critical in the ring, the more easily a proton is released, and the
determining its overall water solubility. Large molecules greater the acidity. The ionized form is usually several
tend to have lower water solubility because they have orders of magnitude more water soluble than the neutral
a higher molar volume. The addition of halogens to un-ionized form. The relative proportion of each species is
an organic molecule might be expected to increase the highly dependent on pH.
water solubility of the new molecule as halogens are
Log Kow , a Measure of Hydrophobicity
electronegative (electron-withdrawing) species and would
thus tend to increase the polarity of the molecule. Polarity and water solubility do not give direct measures
However, in the increasingly chlorinated series of phenols of the hydrophobicity of a compound. A hydrophobic
in Table 1, this is not the case: increasing chlorination compound might be expected to dissolve in a lipophilic
very markedly decreases the solubility of chlorophenols phase, and this is the basis for the octanol–water
partition coefficient, Kow . This is the ratio of a chemical’s THE RELATIONSHIP TO TOXICITY AND
concentration, at equilibrium, when dissolved in octanol BIODEGRADATION
and in water, when both phases are saturated with
each other.
This test was devised by the pharmaceutical industry Hydrophobicity and Toxicity
to study partitioning of drugs as it roughly equates to the
partitioning between water and body fat. It has proven Particularly Kow and log Kow have become increasingly
very useful in predicting how chemicals partition in the used as predictors of the behavior of pollutants in the envi-
environment. Large molecules of low polarity have a high ronment. It has been correlated with toxicity. The most
Kow and are more likely to partition to solids in the common mode of action regarding the toxicity of industrial
environment. Small, polar molecules dissolve in water pollutants is probably narcosis. Nonpolar narcosis results
rather than octanol, have a low Kow , are much less likely from the perturbation of cellular membranes (2) due to
to partition to solids in the environment, and so are more the entry of hydrophobic (lipid-soluble) molecules into the
likely to remain in an aqueous phase: phospholipid bilayer of the membrane. Narcosis has been
shown to be a nonspecific physiological effect which is
concentration of pollutant in independent of chemical structure, and log Kow is highly
water-saturated octanol (mg/L) correlated with nonpolar narcosis toxicity (3).
Kow = (3)
concentration of pollutant Phenols have greater toxicity than predicted by
in octanol-saturated water (mg/L) nonpolar narcosis (4). They act by polar narcosis, which
The range of values for Kow is very wide, and tabulated probably results from a strong hydrogen-bonding group
values are more often expressed as log Kow . In the that makes the compound more polar. Nevertheless,
examples in Table 1, it can be seen that, as the molecule there is still a strong correlation between the toxicity
becomes larger through increasing chlorination, the water of these compounds and log Kow . Pentachlorophenol has
solubility decreases and log Kow increases. greater toxicity as a result of a further mechanism:
its acidity makes it a weak acid respiratory uncoupler
Log Koc , Partitioning to Soil (WARU). Terada (5) noted that WARUs possess an acid-
dissociable moiety (e.g., a hydroxyl group), a bulky
The most basic statement of the composition of soil is hydrophobic moiety, such as an aromatic ring, and
that it contains sand, clay, organic matter, water, and air. multiple electronegative groups, such as halogens. During
Although typically the organic matter content of soil is in
weak acid uncoupling, ATP synthesis is inhibited with
the range of 0.1 to 5%, this is where organic pollutants sorb,
no effect on the respiratory chain (6) because electron
not to the soil as a whole. The soil adsorption coefficient,
transport and phosphorylation reactions are uncoupled
Koc , takes account of this. For a liquid organic pollutant,
from each other.
mass of pollutant sorbed
to soil organic matter (mg/g)
Koc (mL/g) = (4) Hydrophobicity and Biodegradability
mass of pollutant in
the aqueous phase (mg/mL)
It is believed that only molecules of hydrocarbons
The range of values for Koc is very wide, and therefore the that are dissolved in the aqueous phase are available
use of log Koc is preferred. The soil partition coefficient for intracellular metabolism (7). The rate at which a
is greater, the more hydrophobic the pollutant. However, particular organic compound dissolves in water is critical
it has been measured for far fewer chemicals than Kow . to both the toxicity and biodegradability of the compound
It might be expected to be a more realistic measure than as this governs the rate of transfer to the organism (in
Kow , but as data are relatively rare for Koc , a series of biodegradability, microorganisms are important). The rate
correlation equations has been developed that relate Koc to of transfer is determined by the equilibrium and actual
Kow . The increase in log Kow seen in Table 1 as chlorination concentration in the bulk phase and the aqueous phase.
increases is mirrored by increasing log Koc . This is central to the concept of bioavailability.
From the compounds in Table 1, the increasing degree
Fugacity of chlorination decreases the rate of biodegradation.
The term fugacity literally means the ‘‘tendency to flee,’’ Increasing chlorination also leads to a marked drop in
and refers to a molecule’s driving force to escape from water solubility and a corresponding increase in hydropho-
the compartment in which it currently exists. With bicity, which influence biodegradability. The toxicity of
knowledge of all of a chemical’s partition coefficients, it the compound also adversely affects biodegradation. The
is possible to predict its relative concentration in air, interaction between toxicity and biodegradation is not
water, and soil at equilibrium. The gradient of fugacity well understood, but they are clearly linked and compet-
between two compartments (e.g., soil and soil porewater) ing processes: at low concentrations, a compound may
determines the potential for the chemical to move from be biodegradable, but this is arrested at higher concen-
one compartment to the other. Fugacity therefore has trations by toxicity. If the concentration is very low for
units of pressure, but this can be related to the chemical biodegradation, then this may not be enough to support
concentration through modeling, and thus the distribution microbial growth and reproduction, the concept of thresh-
through various compartments predicted (Table 1). old elaborated in detail by Alexander (8).
BIOAVAILABILITY direct or indirect and biological or chemical (14), although,

as pointed out by these authors, only organisms can deter-
Bioavailability is a concept recently borrowed from mine whether a chemical is bioavailable, so direct chemical
nutrition to explain the behavior of chemical pollutants methods are not possible.
in the environment (9). There are several definitions in
circulation. In its simplest form (10), bioavailability is Direct Biological Methods. These are techniques that
defined as ‘‘the amount of contaminant present that can measure the actual amount of a chemical taken up by a
be readily taken up by microorganisms and degraded.’’ But target organism, and this may ultimately be the most
this ignores bioavailability to higher organisms. Another accurate measure of bioavailability, although it is by
definition that takes account of this and alludes to the risk no means routine. For example, the determination of
assessment practice (11) is ‘‘a measure of the potential of chemical levels in earthworms is relatively expensive and
a chemical for entry into ecological or human receptors.’’ time-consuming. The great strength of these direct tech-
The bioavailability quandary mentioned previously has niques, however, is that they integrate all the biotic and
been accurately articulated by the International Union of abiotic modifying factors of chemical bioavailability (14).
Pure and Applied Chemistry (IUPAC).
Indirect Biological Methods. It is possible to observe a
‘‘In the IUPAC Conference of Pesticide Chemistry in London in
response to a chemical in an organism without actually
1998, there was a controversial discussion on the topic whether
agrochemicals in the environment are really hazardous. The measuring the concentration of the chemical. A range
conventional wisdom is that chemicals in soil are available to of responses is possible, such as lethality, enzyme
microorganisms, plant roots, and soil fauna like earthworms induction or inhibition, and reproductive effects. These are
and animals via dermal exposure. Then bioaccumulation regarded as indirect biological methods because the effect
through the food web may induce indirect exposure to may be quantifiable in the organism, but the chemical
higher organisms. National governments are simply reducing concentration remains unknown.
the threshold levels of chemicals in the environments for A promising, novel approach is the use of genetically
larger safety margins in their guidelines. However, the modified microorganisms that can detect and quantify
question raised at the Congress was that chemical residues
specific pollutants. This type of biosensor is based on
in the environment are not always bioavailable, so that
the actual exposure of biota to chemicals is rather different
the highly specific genetic control mechanisms used
from the amount (concentration) present in the environment by microorganisms to ensure that specific proteins are
(authors’ italics). In addition, the persistence and efficacy expressed only when needed; for example, to detoxify
of agrochemicals are affected by their bioavailability in a particular toxic substance. This control is exerted by
soil. There is a lack of comprehensive understanding of inducible promoters, consisting of a specific DNA sequence
the bioavailability of chemicals with different chemical upstream of the genes to be controlled and a DNA-binding
characteristics, such as non-polar organochlorines, polar protein that either activates or prevents transcription
agrochemicals and etc. Thus, there is a strong need to clarify in response to the presence or absence of the target
the scientific basis for bioavailability: definition, estimation compound. Biosensors of this type are easily generated
methods and affecting factors.’’
by fusing such a controllable promoter to a reporter gene
that generates a detectable signal when the promoter is
Measuring Bioavailability
activated (15). The most popular reporter is the bacterial
Measuring or even estimating bioavailability is problem- lux system that produces light (Fig. 2), because light can
atic as it is affected by many individual and interacting be easily quantified and, due to its rarity in biology, there
conditions relating to soil and water chemistry, pollutant is no interference from the background biochemistry of the
chemistry and partitioning, and biological transformation host organism.
and concentration factors (12). Moreover, the ‘‘aging’’ pro- Such biosensors can be both highly specific and
cess brings about a time-dependent decrease in bioavail- responsive to very low concentrations, and the microbial
ability as compounds become sequestered in soil over time cell membrane makes this approach truly a measure of
by ill-understood mechanisms (13). The techniques cur- bioavailability. There are limitations to developing such
rently being researched can be categorized broadly as biosensors for detecting organic pollutants as often the
Figure 2. (a) Schematic of a genetically modified

whole bacterial cell biosensor for detecting bioavail-
able metal. The metal may be expected to be
transported into the bacterial cell via a specific
(a) Out In (b)
transporter protein mediating uptake and efflux. If
the metal is in a nonbioavailable form, for example,
as a precipitate, it will not be transported. The inter-
Bioavail. action of the metal with the specific metal regulatory
M element initiates transcription resulting in synthe-
sis of the luciferase enzyme and the production
Nonbioavail. M o/p lux genes
of light. (b) The naturally bioluminescent bacterium
M Direction of transcription Vibrio fischeri producing light from colonies and
liquid culture.
required specificity is lacking. For heavy metals, specificity 9. Madsen, E.L. (2003). Report on Bioavailability of Chemical
may not be a problem. Nevertheless, careful selection of Wastes with Respect to the Potential for Soil Bioremediation.
the host strain is required for a variety of reasons (16). U.S. Environmental Protection Agency report EPA/600/R-
03/076.
Such biosensors offer the prospect of a much more routine
measure of bioavailability. 10. Maier, R. (2000). Bioavailability and its importance to
bioremediation. In: International Society for Environmental
Biotechnology: Environmental Monitoring and Biodiagnos-
Indirect Chemical Methods. Indirect chemical methods tics. J.J. Valdes (Ed.). Kluwer, Boston, pp. 59–78.
usually involve extracting a fraction of the chemical 11. Loehr, R.C. and Webster, M.T. (1997). Effects of treatment
(metals or organics) from a soil; extractability is defined on contaminant availability, mobility, and toxicity. In:
by the chemical itself, the nature of the extractant(s), and Environmentally Acceptable Endpoints in Soil. D.G. Linz
the experimental conditions applied. The origins of this and D.V. Nakles (Eds.). American Academy of Environmental
approach are in sequential extraction procedures (17) Engineers, Annapolis, MD, pp. 137–386.
that attempt to quantify the speciation of toxicants into 12. Hund-Rinke, K. and Kördel, W. (2003). Underlying issues in
those weakly bound and those strongly bound to the soil bioaccessibility and bioavailability: experimental methods.
matrix. These tests often assume that the weakly bound Ecotoxicol. Environ. Saf. 56: 52–62.
toxicants are those that are more bioavailable, although 13. Chung, N. and Alexander, M. (2002). Effect of soil properties
evidence for this is incomplete. A recent advance in this on bioavailability and extractability of phenanthrene and
atrazine sequestered in soil. Chemosphere 48: 109–115.
technique is sequential accelerated solvent extraction (18)
that reduces sample preparation time and maintains 14. Lanno, R., Wells, J., Conder, J., Bradham, K., and Basta, N.
(2004). The bioavailability of chemicals in soil for earthworms.
relatively constant extraction conditions. A technique
Ecotoxicol. Environ. Saf. 57: 39–47.
under development is the use of selective supercritical
15. Philp, J.C., French, C., Wiles, S., Bell, J.M.L., Whiteley, A.S.,
fluid extraction (19).
and Bailey, M.J. (in Press). Wastewater toxicity assessment
A number of methods are based around the human by whole cell biosensor. In: Handbook of Environmental
gastrointestinal tract (the physiologically based extraction Chemistry. D. Barceló (Ed.). Vol. 5. Water Pollution, Emerg-
test, PBET; 20). To mimic human conditions, they ing Organic Pollutants in Wastewaters, Springer-Verlag,
incorporate gastric juices and enzymes, mix at 37 ◦ C, and Berlin.
use soil residence times similar to those found in children 16. Rensing, C. and Maier, R.M. (2003). Issues underlying use
after ingestion of food. Analysis of all solutions produced in of biosensors to measure metal bioavailability. Ecotoxicol.
the PBET extractions is completed by analytical chemistry Environ. Saf. 56: 140–147.
techniques under matrix-matched conditions. 17. Tessier, A., Campbell, P.G.C., and Bisson, M. (1979). Sequen-
However sophisticated, any such technique is ulti- tial extraction procedure for the speciation of particulate trace
mately limited by the lack of an organism. There will be metals. Anal. Chem. 51: 844–851.
no universal chemical test for bioavailability as the phe- 18. Tao, S., Guo, L.Q., Wang, X.J., Liu, W.X., Ju, T.Z., Daw-
nomenon is organism- and even species-dependent (21). son, R., Cao, J., Xu, F.L., and Li, B.G. (2004). Use of sequen-
tial ASE extraction to evaluate the bioavailability of DDT and
its metabolites to wheat roots in soils with various organic
BIBLIOGRAPHY carbon contents. Sci. Total Environ. 320: 1–9.
19. Björklund, N.T., Bøward, S., Pilorz, K., Mathiasson, L., and
1. Yalkowsky, S.H. and Valvani, S.C. (1979). Solubilities and Hawthorne, S.B. (2000). Introducing selective supercriti-
cal fluid extraction as a new tool for determining sorp-
partitioning. 2. Relationships between aqueous solubilities,
tion/desorption behavior and bioavailability of persistent
partition coefficients and molecular surface areas of rigid
organic pollutants in sediment. J. Biochem. Biophys. Methods
aromatic hydrocarbons. J. Chem. Eng. Data 24: 127–129.
43: 295–311.
2. Cronin, M.T.D. and Schultz, T.W. (1998). Structure-toxicity
20. Ruby, M.V., Davis, A., Schoof, R., Eberle, S., and Sell-
relationships for three mechanisms of action of toxicity to
stone, C.M. (1996). Estimation of lead and arsenic bioavail-
Vibrio fischeri. Ecotoxicol. Environ. Saf. 39: 65–69.
ability using a physiologically based extraction test. Environ.
3. Cronin, M.T.D. and Dearden, J.C. (1995). Review: QSAR in Sci. Technol. 30: 422–430.
toxicology. 1. Prediction of aquatic toxicity. Quant. Struct.
21. Reid, B.J., Jones, K.C., and Semple, K.T. (2000). Bioavailabil-
Act. Relationships 14: 1–7.
ity of persistent organic pollutants in soils and sediments—a
4. Cronin, M.T.D. and Schultz, T.W. (1997). Validation of Vibrio perspective on mechanisms, consequences and assessment.
fischeri acute toxicity data: mechanism of action-based QSARs Environ. Pollut. 108: 103–112.
for non-polar narcotics and polar narcotic phenols. Sci. Total
Environ. 204: 75–88.
READING LIST
5. Terada, H. (1990). Uncouplers of oxidative phosphorylation.
Environ. Health Perspect. 87: 213–218. Hemond, H.F. and Fechner-Levy, E.J. (2000). Chemical Fate and
6. Schultz, T.W. and Cronin, M.T.D. (1997). Quantitative struc- Transport in the Environment. Academic Press, San Diego.
ture-activity relationships for weak acid respiratory uncou- Landis, W.G. and Yu, M-H. (2003). Introduction to Environmental
plers to Vibrio fischeri. Environ. Toxicol. Chem. 16: 357–360. Toxicology, 3rd Edn. Lewis, Boca Raton, FL.
7. Sikkema, J., de Bont, J.A.M., and Poolman, B. (1995). Mecha- Nirmalakhandan, N.N. and Spreece, R.E. (1988). Prediction of
nisms of membrane toxicity of hydrocarbons. Microbiolo. Rev. aqueous solubility of organic compounds based on molecular
59: 201–222. structure. Environ. Sci. Technol. 22: 328–338.
8. Alexander, M. (1999). Threshold. In: Biodegradation and Watts, R.J. (1997). Hazardous Wastes: Sources, Pathways,
Bioremediation. Academic Press, San Diego, pp. 105–116. Receptors. John Wiley & Sons, New York.
PHYSICAL PROPERTIES 527
PHYSICAL PROPERTIES number is used in calculating friction losses, pipe sizing,

and pumping requirements
KEVIN S. MCLEARY In U.S. Customary Units, density is expressed in
slugs/ft3 , and in SI units, it is expressed in kg/m3 .
Like specific weight, density varies with temperature.
The density of pure water varies from 1.940 slugs/ft3
STRUCTURE (999.8 kg/m3 ) at 32 ◦ F (0 ◦ C) to 1.860 slugs/ft3 (958.4 kg/m3 )
at 212 ◦ F (100 ◦ C). Tables 1 and 2 show the variation
Water is a molecule composed of two atoms of hydrogen of density in U.S. Customary Units and SI units,
bonded to an atom of oxygen. It has a volume of 2.97 × respectively.
10−11 µ3 (29.7 Å3 ) (1). As a result of the hybridization
of the oxygen atom, the molecule is bent at an angle
of 104.45◦ . Due to the nonlinear structure and to the SPECIFIC GRAVITY
difference in electronegativity between oxygen (3.44) and
hydrogen (2.20), the molecule has a dipole moment of 1.85 Specific gravity is defined as the ratio of the density
Debye (2). of one fluid at a specified temperature to that of a
Water molecules are held together by intermolecular reference fluid at a reference temperature. For liquids,
attractions. In water, the intermolecular attractions are the reference liquid is usually water, although varying
hydrogen bonds. The negatively charged oxygen atom reference temperatures have been used, for example, 39◦
in one molecule of water is electrically attracted to the and 68 ◦ F (4◦ and 20 ◦ C). Although specification of the
positively charged hydrogen atom of another molecule. reference temperature is required to ensure accuracy, it is
Hydrogen bonding governs the structure of water up to often not given because the density of water varies by only
the boiling point (2). 0.2% over the usual temperature range (5). The specific
gravity of water is usually taken as 1.0.
SPECIFIC WEIGHT (γ )
ABSOLUTE (DYNAMIC) VISCOSITY (µ)
The specific weight of a substance is its weight per unit
of volume. In U.S. Customary Units, it is often expressed The viscosity of a fluid is expressed as a measure of the
as pounds-force per cubic foot (3). In SI units, it is often resistance to flow. Mathematically, it is defined as the
expressed in kN/m3 . Although a value of 62.4 pounds- ratio of the shear stress of a fluid to its velocity gradient or
force per cubic foot (9.81 kN/m3 ) is commonly used for the as the proportionality constant between the shear stress
majority of engineering purposes, the specific weight of and the velocity gradient:
water varies with temperature, from 62.42 pounds-force
per cubic foot (9.805 kN/m3 ) at 32 ◦ F (0 ◦ C) to 59.8 pounds- F
=µ
dv
(1)
force per cubic foot (9.399 kN/m3 ) at 212 ◦ F (100 ◦ C) (4). A dy
Table 1 shows the variability in the specific weight of
water with temperature in U.S. Customary Units. Table 2 Viscosity can be measured by taking two plates of area A,
shows the variability in the specific weight of water with which are separated by a fluid whose thickness is y. The
temperature in SI Units. bottom plate is held stationary. The top plate is moved at
Specific weight is also a function of gravity, and it varies a constant velocity v. The force required to maintain the
slightly with elevation. One cubic foot of water at 32 ◦ F constant velocity is measured and used to calculate the
at sea level has a specific weight of 62.42 pounds-force viscosity µ of the fluid.
per cubic foot (9.805 kN/m3 ); the same volume of water at The chief cause of viscosity in fluids is molecular
10,000 feet above sea level has a specific weight of 62.36 cohesion. Molecular cohesion decreases as temperature
pounds-force per cubic foot (9.796 kN/m3 ). Because the is increased, therefore, viscosity also decreases as
earth is not a perfect sphere, but bulges at the equator, temperature increases. Changes in pressure do not affect
specific weight is also a weak function of latitude. However, viscosity (5). Impurities also exert little influence on
these variations are small enough to warrant the use viscosity (1). Tables 1 and 2 show the effect of temperature
of a constant value of 62.4 pounds-force per cubic foot on viscosity.
(9.81 kN/m3 ) at all elevations and latitudes (4).
KINEMATIC VISCOSITY (ν)
DENSITY (ρ)
There are many times in fluid mechanics and hydraulics
The density of a fluid is defined either as its mass per when the fluid’s absolute viscosity is divided by its
unit of volume or as the ratio of its specific weight to the density. To ease calculations, this term has been defined
acceleration due to gravity. Density ranks among the most as kinematic viscosity because it has kinematic units of
important hydraulic properties of water (1). It is a factor length and time, ft2 /sec or m2 /s.
in the settling velocity of particles and in the filtration Both absolute viscosity and density are a function of
of water through a porous medium. A fluid’s Reynolds temperature, so kinematic viscosity is also a function of
number, the most important dimensionless number in temperature. Tables 1 and 2 show how kinematic viscosity
hydraulics, is a direct function of density. The Reynolds varies with temperature.
528 PHYSICAL PROPERTIES
Table 1. Physical Properties of Water in U.S. Customary Unitsa

Modulus of Dynamic Viscosity, Kinematic Viscosity, Surface Tension,c Vapor Pressure,
Specific Weight, Density, ρ, Elasticityb µ × 105 , ν × 105 , σ, pv ,
Temp., ◦ F γ , lbf /ft3 slug/ft3 E/103 , lbf /in.2 lbf · sec/ft2 ft2 /sec lbf /ft lbf /in.2
32 62.42 1.940 287 3.746 1.931 0.00518 0.09

40 62.43 1.940 296 3.229 1.664 0.00614 0.12
50 62.41 1.940 305 2.735 1.410 0.00509 0.18
60 62.37 1.938 313 2.359 1.217 0.00504 0.26
70 62.30 1.936 319 2.050 1.059 0.00498 0.36
80 62.22 1.934 324 1.799 0.930 0.00492 0.51
90 62.11 1.931 328 1.595 0.826 0.00486 0.70
100 62.00 1.927 331 1.424 0.739 0.00480 0.95
110 61.86 1.923 332 1.284 0.667 0.00473 1.27
120 61.71 1.918 332 1.168 0.609 0.00467 1.69
130 61.55 1.913 331 1.069 0.558 0.00460 2.22
140 61.38 1.908 330 0.981 0.514 0.00454 2.89
150 61.20 1.902 328 0.905 0.476 0.00447 3.72
160 61.00 1.896 326 0.838 0.442 0.00441 4.74
170 60.80 1.890 322 0.780 0.413 0.00434 5.99
180 60.58 1.883 318 0.726 0.385 0.00427 7.51
190 60.36 1.876 313 0.678 0.362 0.00420 9.34
200 60.12 1.868 308 0.637 0.341 0.00413 11.52
212 59.83 1.860 300 0.593 0.319 0.00404 14.70
a
References 3 and 7.
b
The modulus of elasticity is given at atmospheric pressure.
c
Surface tension is given for water in contact with air.
Table 2. Physical Properties of Water in SI Unitsa

Modulus of Dynamic Viscosity,
Specific Weight, Density, ρ, Elasticity,b µ × 103 , Kinematic Viscosity, Surface Tension,c Vapor Pressure,
Temp., ◦ C γ , kN/m3 kg/m3 E/106 , kN/m2 N·s/m2 ν × 106 , m2 /s σ , N/m pv , kN/m2
0 9.805 999.8 1.98 1.781 1.785 0.0765 0.61

5 9.807 1000.0 2.05 1.518 1.519 0.0749 0.87
10 9.804 999.7 2.10 1.307 1.306 0.0742 1.23
15 9.798 999.1 2.15 1.139 1.139 0.0735 1.70
20 9.789 998.2 2.17 1.002 1.003 0.0728 2.34
25 9.777 997.0 2.22 0.890 0.893 0.0720 3.17
30 9.764 995.7 2.25 0.798 0.800 0.0712 4.24
40 9.730 992.2 2.28 0.653 0.658 0.0696 7.38
50 9.689 988.0 2.29 0.547 0.553 0.0679 12.33
60 9.642 983.2 2.28 0.466 0.474 0.0662 19.92
70 9.589 977.8 2.25 0.404 0.413 0.0644 31.16
80 9.530 971.8 2.20 0.354 0.364 0.0626 47.34
90 9.466 965.3 2.14 0.315 0.326 0.0608 70.10
100 9.399 958.4 2.07 0.282 0.294 0.0589 101.33
a
References 3 and 7.
b
The modulus of elasticity is given at atmospheric pressure.
c
Surface tension is given for water in contact with air.
SURFACE TENSION (σ ) has units of force per unit area (ft-lbf /ft2 , or lbf /ft) and
decreases as the temperature of the liquid increases. It is
Surface tension is caused by the attractive forces between also a function of the gas contacting the liquid (5). Table 1
molecules in a liquid. In a bulk liquid, each molecule is gives the variation of surface tension of water in contact
attracted to its neighboring molecules, and the total force with air as a function of temperature in U.S. Customary
on the molecule is zero. However, a molecule at the surface Units, and Table 2 shows the variation in SI units.
experiences an attraction to other molecules deeper in the
liquid, but there are no molecules outside to balance these
attractions. The molecule at the surface experiences a net VAPOR PRESSURE (PV )
attractive force toward the bulk liquid.
Surface tension is the amount of work necessary to Water is a collection of a large number of molecules in a
increase the surface area of the liquid by one unit. It liquid phase that move about at random. The molecules
ENVIRONMENTAL PHOTOCHEMISTRY IN SURFACE WATERS 529
collide with each other and rebound in different directions. used for this quantity of heat on a molar basis. If a mass
Each molecule moves with a certain velocity (energy). basis is used, the term ‘‘specific heat’’ is frequently used.
Molecules on the surface of a liquid moving away from The heat capacity of water is the basis of the unit of heat
the bulk liquid have no other molecule to collide with. in both the English and SI systems. Therefore, the heat
If such a molecule has sufficient energy to escape the capacity (specific heat) of water is 1 Btu/pound-mass/ ◦ F in
intermolecular attractions, it may escape the liquid into the English system and 1 cal/g/ ◦ C in the SI system.
the vapor space above the liquid. It evaporates. Evaporated Heat capacities of liquids are not functions of
water molecules also move about at random. Some of the temperature, except at high reduced temperatures.
water vapor condenses into the liquid water phase. (A reduced temperature is a thermodynamic quantity
If the water were placed in a closed container at relating the temperature of a material to its critical
a constant temperature and the air evacuated from temperature. The critical temperature for water is
the vapor space above the liquid, one could measure 647.1 K.) (6). At reduced temperatures exceeding 0.7–0.8,
the pressure exerted by the water vapor. The pressure liquid heat capacities are large and are strong functions of
would increase until equilibrium is attained—when the temperature.
rate of molecules evaporating and the rate of molecules
condensing are equal. The pressure of the water vapor then
BIBLIOGRAPHY
is constant. This is the vapor pressure of the water for the
given temperature. Table 1 gives the vapor pressure of
1. Fair, G.M., Geyer, J.C., and Okun, D.A. Water and Wastewater
water in psi. Table 2 gives the vapor pressure in kN/m2 . Engineering. John Wiley & Sons, New York.
2. Atkins, P.W. Physical Chemistry, 2nd Edn. W.H. Freeman,
MODULUS OF ELASTICITY (E ) San Francisco.
3. Metcalf & Eddy, Inc. Wastewater Engineering: Treatment,
Although water is usually considered incompressible, it is Disposal, and Reuse, 3rd Edn. McGraw-Hill, New York.
slightly compressible. The modulus of elasticity is defined 4. F.S. Merritt (Ed.). Standard Handbook for Civil Engineers, 3rd
as the increase in pressure that results in a specific Edn. McGraw-Hill, New York.
percentage change in volume: 5. Lindeburg. Civil Engineering Reference Manual.
6. Smith, J.M. and van Ness, H.C. Introduction to Chemical
p Engineering Thermodynamics, 3rd Edn. McGraw-Hill, New
E= (2)
V/V York.
7. Culp/Wesner/Culp. Handbook of Public Water Systems. Van
where p is the increase in pressure, V is the resulting Nostrand-Reinhold, New York.
change in volume, and V is the original volume of water.
The modulus of elasticity has units of pressure and is a
function of temperature. Tables 1 and 2 show the variation ENVIRONMENTAL PHOTOCHEMISTRY IN
of the modulus of elasticity with temperature. SURFACE WATERS
MARK A. YOUNG
SPEED OF SOUND (C )
University of Iowa
Iowa City, Iowa
Sound is the propagation of a pressure wave through a
material and is dependent on the material’s modulus of
elasticity and density. The speed of sound, c, in a liquid is
INTRODUCTION
given by
Egc Surface waters make up a large fraction of the available
c= (3)
ρ terrestrial surface area, accounting for over 70% of the
total area with most of that consisting of ocean waters. A
The speed of sound in water is important in calculating the correspondingly large fraction of the incoming solar flux
pressure increase due to water hammer. When the velocity will impinge on natural waters, penetrating to a depth
in a pipe is decreased suddenly, the pipe experiences an determined by the scattering and absorption properties
increase in pressure. The increase in pressure, p, caused of the water system. While most of the short wavelength
by a decrease in velocity of v, is solar radiation is attenuated in the atmosphere, mainly
by absorbing species such as O2 and O3 , the surface flux
ρcv
p = (4) of energetic photons in the near-UV and visible remains
gc appreciable. While water itself is not photoactive, the
irradiated volume of typical surface waters contains a
HEAT CAPACITY (CP ) plethora of potential chromophores. The chromophoric
material consists of dissolved species and suspended
Heat capacity is defined as the amount of heat needed particulate matter that can have strong electronic
to raise the temperature of a given amount of substance absorptions that significantly overlap with the available
by 1◦ . It is usually expressed in units of heat per unit of solar flux. Absorption of solar photons can initiate a rich
substance per degree. The term ‘‘heat capacity’’ is often variety of photochemical processes with sufficient kinetic
530 ENVIRONMENTAL PHOTOCHEMISTRY IN SURFACE WATERS
rates to manifest a substantial impact on the chemistry

Dissolved and colloidal chromophores
taking place in the aqueous environment. Photochemically x
driven chemistry, in turn, can have enormous implications
for global geochemical cycles, climate, and biogenic DOM, NO3−/NO2−, > Fe(III)-L,
processes, making photochemistry a critical consideration
in the study of environmental processes.
+ hv
H2O, O2, R
BACKGROUND
Natural surface waters can be either saltwater (marine, or

ocean, waters) or freshwater (continental waters). Within
each of these reservoirs, widely varying environments are
found: remote and coastal areas in the oceans, and rivers,
lakes, streams, bogs/marshes on the continents, as well
as interfacial environments, such as estuaries. In addition Reactive intermediates
to differences in salinity and pH, there will be variations
e−(aq), O2−, HO2−, RO2−, 1O2+
in the types and concentrations of chromophoric species
residing in each reservoir. Potential chromophores include OH−, H2O2, > Fe(II), DOM+ −, R−
dissolved organic material (DOM), inorganic NOx − ions,
metal complexes, anthropogenic pollutants, and colloidal
H2O, O2, R
matter consisting of crustal minerals and particulate
organic material (POM). Organic chromophores may also
adsorb on, or complex with, the surface of mineral
colloids and participate in charge transfer processes.
Likewise, metals can become incorporated into organic
material. The distinction between dissolved and colloidal
material is not sharp and often depends on the analytical
procedure used to assay concentrations. The relative
importance of the various chromophores will depend on Final products
the details of the aqueous environment. Chromophores
can absorb solar photons and undergo photochemical Figure 1. Schematic representation of photochemical processes
transformation while yielding a variety of primary and in surface waters. Chromophores consist of DOM, nitrate/nitrite
secondary reactive species, as depicted schematically in ions, dissolved and colloidal metal complexes [represented by
Fig. 1. The photogenerated intermediates can, in turn, Fe(III)], and anthropogenic pollutants (such as substituted
react with other species present in the water at rates that aromatics). Reactive intermediates consist of solvated electrons,
various radical species, both cationic and anionic, peroxides, and
are often diffusion controlled to form final products. In
reduced metal centers.
some instances, the chromophore may be regenerated,
resulting in a photocatalytic cycle that can be highly
efficient (Fig. 1). Photophysical processes, luminescence (fluorescence and
Excitation of a chromophore can initiate a number phosphorescence), internal conversion, intersystem cross-
of photophysical and photochemical processes that are ing—Eq. 2—and collisional quenching—Eq. 3—will re-
outlined in the following equations: turn the chromophore to the electronic ground state and
convert the photon energy to heat. The excited state may
J also ionize, Eq. 4, or react, Eq. 5, converting the photon
S + hν −−−→S∗ excitation; (1) energy into chemical potential. The reaction could be uni-
kL ,kI molecular, such as dissociation, or bimolecular in nature,
S∗ −−−→S (+hν) luminescence, internal
involving other components of the water system. Products
conversion, and intersystem
might include the reactive intermediates in Fig. 1, which
crossing; (2)
can then participate in secondary chemistry. Energy trans-
kQ fer, or sensitization, Eq. 6, could also induce a concurrent
S∗ + Q−−−→S + Q quenching; (3)
change in the electronic spin state, yielding a reactive
∗ + −
S −−−→ S + e photoionization; (4) triplet excited state.
kR
The photochemical sequence is initiated by absorption
−−−→ of a photon, Eq. 1, with a photolytic rate constant, J. From
S∗ P reaction; (5) a simple consideration of Beer’s Law, J is seen to depend
[R]
kET
on the incident solar flux, I, and the (molar, decadic)
S∗ + R−−−→S + R∗ energy transfer; (6) extinction coefficient of the chromophore, ε. The relative
kCT
efficiency, or yield, of a reaction can be specified by the
S∗ + R−−−→S+/− + R−/+ charge transfer. (7) quantum yield, φ, which is the number of product species
formed divided by the number of photons absorbed. The C–C, 348 kJ; O–O, 146 kJ; C–OH, 322 kJ; benzylic C–H,
absorption coefficient and quantum yield are generally 368 kJ; phenolic O–H, 362 kJ; C–Cl, 342 kJ). Over the
wavelength dependent. The flux depends on wavelength wavelength range of photochemical interest, 300–500 nm,
as well as the solar zenith angle and other factors, such as Fig. 2 indicates that the noon-time flux to surface
latitude, longitude, and meteorological conditions, such as waters is about 4 einstein · m−2 · h−1 . Typical penetration
cloud cover. The rate constant for a photoinduced process depths range from 1 to 20 m and so the molar rate
can then be written as an integral over the wavelength of photon absorption is 0.2–4.2 × 10−3 einstein·L−1 ·h−1 .
region of interest: Dissolved organic carbon (DOC) is the major absorbing

species in most natural waters (see Fig. 2) and has
J = 2.303 ε(λ)φ(λ)I(λ)dλ = 2.303 ε(λi )φ(λi )I(λi )λi an approximate average concentration of 5 mg C·L−1 .
i While highly diverse in nature, an average chromophore
(8) molecular weight of 120 g C·mol−1 (10 carbon atoms)
The flux in this case is termed the spectral actinic flux and can be assumed and so the absorber concentration is
is a spherically averaged irradiance that takes into account approximately 0.04 mM. The estimated photon absorption
radiation incident on the chromophore over 4π steradians. rate for a well mixed volume would correspond to
For practical purposes, the integral is often evaluated an average excitation rate of 5–100 h−1 , meaning that
as a discrete sum over small wavelength intervals, as a near-surface chromophore is excited approximately
expressed in the last term of Eq. 8, where ε, φ, and once each minute (2,3). Furthermore, it is estimated
I represent average quantities over the interval λi that the accumulated annual radiation dose is about
centered at λi . 1300 times larger (2). Given the time scales for transport
The solar actinic flux at noon (zenith angle = 0◦ ) for a and chemistry in surface waters, it is clear that
northern latitude (47.4◦ N) is shown in Fig. 2. The flux is photochemically driven reactions will be an important
expressed in terms of the photon flux per unit wavelength component of chemical cycles in the hydrosphere (Fig. 2).
interval with units of einstein · m−2 · s−1 · nm−1 , where an
einstein corresponds to 1 mole of photons (Avogadros’s REACTIVE INTERMEDIATES
number, NA = 6.022 × 1023 mol−1 ). The UV-B radiation
(280–320 nm) is strongly attenuated by the ozone layer, A variety of reactive intermediates have been identified
effectively limiting the useful actinic flux to about 300 nm. from the irradiation of natural waters at relevant solar
Shorter wavelength radiation does reach the surface but wavelengths (see Fig. 1). Selected properties for some of
at 293 nm, the flux has decreased by more than two orders the important species are listed in Table 1, along with
of magnitude relative to 300 nm. UV-A (320–400 nm) is estimated steady-state concentrations in typical surface
not strongly absorbed by ozone and the flux increases up waters (generally corresponding to a noon-time solar
to the maximum in the visible at approximately 575 nm. flux). Under conditions of constant solar illumination,
A photon has a molar energy of pseudo-first-order kinetics result and the steady-state
concentration for an intermediate, [I]SS , can be used in
hc
E = NA J · einstein−1 (9) conjunction with the second-order rate constant, k, for a
λ particular reaction to calculate the reaction half-life as
A 300-nm photon has an energy of 399 kJ · einstein−1 , ln 2
which is on the order of typical bond enthalpies (e.g., τ= (10)
k[I]SS
1.0 10
0.8 8
Einstein × 106 m−2⋅s−1⋅nm−1
(e)
0.6
Absorbance
0.4 (d) 4
(c)
Figure 2. Left ordinate: Absorption spectra of

0.2 (b) 2 surface water samples with varying concentra-
tions and sources of DOC in mg C·L−1 : (a) = 3.0,
(a) (b) = 7.8, (c) = 13.4, (d) = 13.4, (e) = 15.4. Path
length is 1.0 cm. Data are adapted from Ref-
0.0 0
erence 7. Right ordinate: Solar actinic flux for
200 250 300 350 400 450 500
noon-time sun (zenith angle = 0◦ ) for a northern,
Wavelength, nm midlatitude. Data is from Reference 1.
and, thus, estimate the importance of a given reaction reaction with e− (aq) has been shown to degrade organic
pathway. For example, half-lives for reactions with phenol, pollutants (2,10), it is unlikely to play an important role
a representative organic, aromatic compound, have been in photochemical processes in most surface waters.
included in Table 1.
Primary transient species are produced rapidly, on O2 −
the order of <1 ns. These transients can then react
with other components in the aquatic environment to Superoxide, O2 − , can be formed from reaction with e− (aq),
produce secondary products. Besides H2 O, one of the as above, in charge transfer reactions with sufficiently
most important interactions involves dissolved oxygen, reducing species, and via electron trapping at the
which, in well aerated surface waters, will be present in surface of semiconductor colloids. Photolysis of DOM is
concentrations determined by Henry’s Law, a major source of superoxide but photolysis of trace metal
complexes may also contribute to O2 − formation (11).
[O2 ] = KH PO2 (11) Superoxide is a mild reductant/oxidant (Table 1) and
can participate in both one-electron oxidation and
where KH = 1.3 × 10−3 M−1 · atm−1 is the Henry’s Law reduction reactions. The steady-state concentration is
constant (8). The mixing ratio of O2 in the atmosphere uncertain but from limited formation rate measurements,
is 21%, yielding a concentration of [O2 ] = 0.27 mM at the concentration is [O2 − ]SS ≈ 10−9 –10−7 M (9,10). The
298 K. Primary radicals will react with dissolved O2 relatively high concentration suggests that superoxide
at diffusion controlled rates (diffusion controlled rate may be important in photochemical degradation of various
constant, kD ≈ 7 × 109 M−1 · s−1 at 298 K) corresponding species (7). However, O2 − is unstable in water with respect
to a time scale of ≈ 0.4 µs, which will limit the lifetime of to disproportionation to form hydroperoxyl radical and
many photogenerated transients (the ‘‘oxygen wall’’) (9). hydrogen peroxide (3,12)
e− (aq) K
O2 − + H+ −−
−−−−HO2 log K = −4.8 (13)
Electrons generated from photoionization of DOM are
HO2 /O2 − + H+ −−−→ H2 O2 + O2 (14)
rapidly solvated in aqueous environment to produce,
e− (aq). The solvated electron reacts with dissolved oxygen
to form the superoxide, RO2 •
k
Peroxyl radicals can be formed from the photolysis of DOM
e− (aq) + O2 −−−→O2 − (12) and metal complexes. Formation rates for RO2 species are
thought to linearly scale with the DOM concentration in
where k ≈ 2 × 1010 M−1 · s−1 [which is larger than the typical waters (2), but the reactivity of the radical is highly
diffusion limit due to a tunneling mechanism for e− (aq) dependent on the type of humic substance involved, which
transport] (4,5). The time scale for Eq. 12 is on the order is itself not well characterized. These uncertainties make
of 200 ns. The short lifetime coupled with measured it difficult to estimate a steady-state concentration for
typical production rates leads to a very small steady-state peroxyl radicals. HO2 is a mild oxidant/reductant (Table 1)
concentration of [e− (aq)]SS ≈ 10−17 –10−15 M (2,9). While but DOM-derived peroxyl radicals exhibit significant
Table 1. Selected Properties of Some Important Reactive Intermediates in

Aqueous Photochemistrya
Radical
Property −
e (aq) O2 −
HO2 OH 1
O2 ∗
Gf 0 (kJ·mol−1 ) 276b 13b

Hf 0 (kJ·mol−1 ) 277b 80c 138c −7b 94.3d
E0 (V)e −2.9f −0.33g 0.79h 1.9i 0.65j
[I]SS (M) 10−17 –10−15 10−9 –10−7 10−19 –10−17 10−14 –10−13
k (M−1 ·s−1 )k 3.0(7) 5.8(2) 8.5(3) 1.0(10) 0.2(7)
τ (days × 103 ) 0.3–30 0.0001–0.01 0.08–8 0.04–0.4
a
The rate constant data is for reaction with C6 H5 OH (phenol) except for the HO2 radical, which is for
reaction with 1,4-C6 H4 (OH)2 .
b
Data from Reference 4.
c
d
Gas phase value.
e
f
H2 O + e− → e− (aq)
g
O2 + e− → O2 −
h
HO2 + e− → HO2 −
i
OH + e− → OH−
j1
O 2 ∗ + e− → O 2 −
k
Number in parentheses indicates the exponent value. Data from Reference 7.
reactivity with alkylphenols (13) and so may be an CHROMOPHORES

important reaction channel for some pollutants.
NO3 − /NO2 −
1
O2 ∗ Inorganic nitrate and nitrite ions are perhaps the most
well defined and easily understood photoactive component
Singlet oxygen, 1 O2 ∗ , is an electronically excited state
of natural waters. Nitrate concentration in surface
of oxygen with 94.3 kJ of excess energy. Triplet excited
waters is on the order of 1 mg·L−1 although it may be
3
DOM∗ , generated by intersystem crossing from the substantially higher in waters subject to an influx of
initially excited singlet state, can sensitize the formation agricultural drainage due to the heavy use of nitrogen
∗
of 1 O2 through energy transfer, Eq. 6. Most of the singlet fertilizers. The absorption band maxima for nitrate and
oxygen will be rapidly quenched back to the triplet ground nitrite are at 302 nm (ε = 3.1 M−1 ·cm−1 ) and 352 nm (ε =
∗
state by H2 O (τ ≈ 4 µs) and only a fraction of the 1 O2 will 9.6 M−1 ·cm−1 ), respectively (17,18). Ultraviolet excitation
be available for subsequent reaction (2). The production of of nitrite leads to photodissociation (18):
1 ∗
O2 generally scales with the amount of DOC and steady-
state concentrations of 0.3–3.0 × 10−14 M per mg·L−1 of NO2 − + hν −−−→ NO + O
ž −
(16)
DOC have been measured for typical surface waters (14).
Singlet oxygen is a reactive electrophile, particularly
The ž O− species, the conjugate base of hydroxyl radical, is
toward cyclic 1,3 dienes, phenolate anions, and some
rapidly protonated at pH < 12 [pKa (ž OH) = 11.9] (4):
polynuclear aromatics (2,14,15) and can play an important
role in aqueous photochemistry.
O− + H+ −−−→ OH
ž ž
(17)
OH The quantum yield for ž OH production from nitrite

varies from 0.015 to 0.08 over the wavelength range
The hydroxyl radical is a powerful oxidant (Table 1) 298–371 nm (18,19). Photolysis of the nitrate ion yields
that exhibits high reactivity toward a range of organic either nitrite ion or nitrogen dioxide, which can then
species. It is mainly formed from photodissociation of reform nitrate and nitrite:
NO3 − /NO2 − but also via photolysis of metal-hydroxides,
decomposition of H2 O2 , and DOM photochemistry. The NO3 − + hν −−−→ NO2 − + O (18)
hydroxyl radical is rapidly scavenged by DOM and
−
NO3 − + hν −−−→ NO2 + O
ž ž
the steady-state concentration of OH in surface waters (19)
is generally small, [OH]SS ≈ 10−19 –10−17 M (9,10). The − − +
2 NO2 + H2 O −−−→ NO2 + NO3 + 2H
ž
(20)
OH radical readily reacts with many organic species,
via abstraction and addition to double bonds, but the
For wavelengths ≥290 nm, the quantum yield for
low steady-state concentration suggests that hydroxyl
formation of ž OH from NO3 − , via Eq. 19, is in the
will not be the dominant photooxidant for most surface
range of 9.2–17 × 10−3 (17,18). Hydroxyl is a highly
waters. For example, a ‘‘typical’’ OH rate constant (4)
reactive radical and a strong oxidizing agent (see
is 5 × 109 M−1 · s−1 , corresponding to a reaction half-life
above). Various laboratory studies have shown that
of τ ≈ 160 d at the higher OH concentrations. However,
trace organic molecules, such as biologically refractory
for environments with a larger concentrations of NOx −
pollutants, are degraded by irradiation in the presence of
relative to DOC, hydroxyl radical reactions can be nitrate/nitrite (17–19).
important pathways due to near diffusion-controlled OH
rate constants (2).
DOM
Dissolved organic matter (DOM) is a major and ubiquitous
H2 O2
component of natural surface waters, having numerous
Hydrogen peroxide is a relatively stable, nonradical sources such as soils, sediments, and aqueous biological
species produced from photochemical processes in natural activity (20). DOM concentrations are often expressed in
surface waters. It is not a primary product but, terms of the dissolved organic carbon (DOC) fraction,
rather, is generated in secondary reactions such as which comprises about 60% of the total DOM. Continental
the disproportionation of superoxide, Eqs. 13 and 14. waters have typical levels of 1–10 mg C·L−1 DOC,
Hydrogen peroxide absorbs only weakly in the UV although some waters, bogs, marshes, and shallow
(ε = 1 M−1 · cm−1 at 310 nm) (16), and decomposition to streams, for instance, can have concentrations exceeding
yield hydroxyl radicals may be dominated by biological 50 mg C·L−1 . Marine DOC is about 1 mg C·L−1 and
mediation and redox cycles involving trace metals represents an enormous carbon reservoir, approximately
equal in size to the CO2 reservoir in the atmosphere (2).
DOM is nominally defined as the organic content of
H2 O2 −−−→ 2OH (15) water that passes through a filter of some pore size,
0.1–0.5 µm, and so will have a colloidal component.
The slow decomposition of H2 O2 leads to relatively high The DOC in surface waters can be divided into two
steady-state concentrations in surface waters, [H2 O2 ]SS = fractions; a lower molecular weight nonhumic fraction and
10−8 –10−7 M (2,9). higher molecular weight humic component. The nonhumic
fraction is about 10–60% of the DOC and is made α-Fe2 O3 , hematite; γ -FeOOH, lepidocrocite; Fe(OH)3 ,
up of small compounds of mainly biogenic origin, such ferrihydrite) may be present as dissolved species but are
as peptides, carbohydrates, fats, and chlorophyll, which predominantly in colloidal form. Iron has two common
are subject to rapid biological degradation. In addition, oxidation states, Fe(III) and Fe(II), and participates
anthropogenically derived species, such as biocides in redox reactions that are an important component
(often halogenated compounds) and polycyclic aromatic of aqueous chemical cycles. Fe(III) can form strong
hydrocarbons (PAHs, from combustion processes) are complexes with naturally occurring organic ligands, such
present. The higher molecular weight fraction (>300 Da), as polycarboxylate anions (e.g., oxalate, malonate, citrate),
comprising 40–90% of the DOC, is termed humic matter and humic and fulvic acids (11,26). Photoexcitation of a
(HM). Humic matter is a complex material composed Fe(III) complex promotes a ligand to metal charge transfer
of condensed biopolymers and is much more biologically process yielding a free, oxidized ligand and a reduced,
refractory. HM is mildly acidic, has a significant aromatic Fe(II), center. Subsequent reactions generate reactive
content, and is highly functionalized with carboxyl, phenol, intermediates (11,12), as schematically depicted below for
hydroxyl, carbonyl, ether, and ester functional groups the oxalato (C2 O4 2− ) ligand:
present. HM can be further divided based on solubility
properties into humic acids (HAs), which precipitate under >Fe(III)C2 O4 − + hν −−−→ >Fe(II) + C2 O4 − (21)
acidic conditions, and fulvic acids (FAs), which are soluble −
C2 O4 −−−→ CO2 + CO2 −
(22)
at all pH values. FAs are more highly substituted and
− −
have a higher oxygen content than HAs and also have a CO2 + O2 −−−→ CO2 + O2 (23)
lower average molecular weight. >Fe(II) + O2 −−−→ >Fe(III) + O2 − (24)
DOM is a highly colored material with an absorption
spectrum that overlaps significantly with the solar actinic >Fe(II) + H2 O2 −−−→ >Fe(III) + OH (25)
flux (see Fig. 2). The UV absorption of surface waters
has been found to be highly correlated with the DOC where ‘‘>Fe(III/II)’’ represents either an iron colloid
content (10,21,22). Studies have shown that photoexcited surface site or a dissolved species. The superoxide will form
DOM can yield many of the radical species outlined hydrogen peroxide and hydroperoxyl radicals, as discussed
in Fig. 1 and Table 1 through primary and secondary above. In colloidal material, the surface Fe(II) is generally
processes (9). The radical production rate varies with the oxidized back to Fe(III) by the ligand or O2 . However, due
character of the DOM and the concentration, as well as to weakened binding with neighbor atoms, the reduced
the location and type of surface water environment. One species may also be released to the solution (3,26). The
study found that coastal marine waters yielded radical result is photodissolution of the mineral colloid and an
production rates of 0.18–1.8 µM·d−1 in the summer due increase in soluble iron species, facilitating redox reactions
to irradiation of humic material in the surface layer (to and promoting microbial activity.
a depth of 1.5 m) (23). The active spectral region was
determined to be 300–385 nm, suggesting a quantum BIBLIOGRAPHY
yield for radical production of 1–2 × 10−3 . Measurement
of light absorption rates in highly colored, organic-rich 1. Finlayson-Pitts, B.J. and Pitts, J.N., Jr. (1986). Atmospheric
Chemistry: Fundamentals and Experimental Techniques.
waters that exhibit significant absorption at longer, visible
Wiley-Interscience, Hoboken, NJ.
wavelengths suggest the radical production rate will be
2. Hoigne, J., Faust, B.C., Haag, W.R., Scully, J.F.E., and
orders of magnitude higher (9). These radicals dominate
Zepp, R.G. (1989). Aquatic humic substances as sources
the photochemistry of most natural water samples, and sinks of photochemically produced transient reactants.
particularly in marine environments. In: Aquatic Humic Substances. Vol. 219, I.H. Suffet and
The DOM will be altered by photolysis, yielding P. MacCarthy (Eds.). American Chemical Society, Washing-
a cationic radical product, for instance, but it can ton, DC, pp. 363–381.
also be attacked by the photogenerated radicals. DOM 3. Stumm, W. and Morgan, J.J. (1996), 3rd Edn. John Wiley &
decomposition can lead to photobleaching, allowing solar Sons, New York.
radiation to penetrate deeper into the water and impacting 4. Buxton, G.V., Greenstock, C.L., Helman, W.P., and Ross, A.B.
microbial photosynthesis. DOM is generally biologically (1988). Critical review of rate constants for reactions of
refractory, resulting in long carbon reservoir residence hydrated electrons, hydrogen atoms and hydroxyl radicals
times, estimated at 6000 years for deep ocean DOM (24). (OH/O− ) in aqueous solution. J. Phys. Chem. Ref. Data 17:
Photodecomposition of DOM may be an important removal 513–886.
pathway with significant implications for the global carbon 5. Spinks, J.W.T. and Woods, R.J. (1990). An Introduction to
cycle (9,15,24,25). The lower molecular weight degradation Radiation Chemistry, 3rd Edn. John Wiley & Sons, Hoboken,
products are biologically labile and can provide a major NJ.
fraction of the biological carbon uptake in certain surface 6. Wardman, P. (1989). Reduction potentials of one-electron
waters (21,25). couples involving free radicals in aqeous solution. J. Phys.
Chem. Ref. Data 18: 1637–1755.
Fe(III/II) 7. NDRL Radiation Chemistry Data Center. Available:
http://www.rcdc.nd.edu/.
Iron is an abundant crustal metal and is one of 8. Seinfeld, J.H. and Pandis, S.N. (1998). Atmospheric Chem-
the most important trace metals in natural water istry and Physics: From Air Pollution to Climate Change.
systems. The Fe-(hydrous)oxides (e.g., α-FeOOH, goethite; John Wiley & Sons, Hoboken, NJ.
ISOTOPE EXCHANGE IN GAS-WATER REACTIONS 535
9. Mopper, K. and Kieber, D.J. (2000). Marine photochemistry 24. Mopper, K. et al. (1991). Photochemical degradation of
and its impact on carbon cycling. In: The Effects of UV dissolved organic carbon and its impact on the oceanic carbon
Radiation in the Marine Environment. Vol. 10. S.D. Mora, cycle. Nature 353: 60–62.
S. Demers, and M. Vernet (Eds.). Cambridge University 25. Kieber, D.J., McDaniel, J., and Mopper, K. (1989). Photo-
Press, Cambridge, UK, p. 101–129. chemical source of biological substrates in sea water: impli-
10. Cooper, W.J., Zika, R.G., Petasne, R.G., and Fischer, A.M. cations for carbon cycling. Nature 341: 637–639.
(1989). Sunlight-induced photochemistry of humic substances 26. Sulzberger, B., Laubscher, H., and Karametaxas, G. (1994).
in natural waters: major reactive species. In: Aquatic Humic Photoredox reactions at the surface of iron(III) (hydr)oxides.
Substances. Vol. 219. I.H. Suffet and P. MacCarthy (Eds.). In: Aquatic and Surface Photochemistry. R. Helz, R.G. Zepp,
American Chemical Society, Washington, DC, pp. 333–362. and D.G. Crosby (Eds.). Lewis Publishers, Boca Raton, FL,
11. Faust, B.C. (1994). A review of the photochemical redox reac- pp. 53–73.
tions of iron(III) species in atmospheric, oceanic, and surface
waters: influences on geochemical cycles and oxidant for-
mation. In: Aquatic and Surface Photochemistry. G.R. Helz, ISOTOPE EXCHANGE IN GAS-WATER
R.G. Zepp, and D.G. Crosby (Eds.). Lewis Publishers, Boca REACTIONS
Raton, FL, pp. 3–37.
12. Zuo, Y. and Hoigne, J. (1992). Formation of hydrogen peroxide TOM WYDRZYNSKI
and depletion of oxalic acid in atmospheric water by photolysis WARWICK HILLIER
of iron(III)-oxalato compounds. Environ. Sci. Technol. 26: The Australian National
1014–1022. University
13. Faust, B.C. and Hoigne, J. (1987). Sensitized photooxidation Canberra, Australia
of phenols by fulvic acid and in natural waters. Environ. Sci.
Technol. 21: 957–964.
14. Haag, W.R. and Hoigne, J. (1986). Singlet oxygen in surface INTRODUCTION
waters. 3. Photochemical formation and steady-state concen-
tration in various types of waters. Environ. Sci. Technol. 20: Molecular oxygen began to accumulate in Earth’s
341–348. atmosphere about 2.2–2.7 billion years ago, and it is
15. Aguer, J-P., Richard, C., and Andreux, F. (1997). Comparison widely accepted that it originated from the photosynthetic
of the photoinductive properties of commercial, synthetic and oxidation of water (1). The accumulation of O2 in the
soil-extracted humic substances. J. Photochem. Photobiol. A atmosphere marked the most important event in the
103: 163–168.
evolution of life on Earth as the atmosphere changed from
16. Bolton, J.R. and Cater, S.R. (1994). Homogeneous pho- a highly reducing to an oxidizing condition. All preexisting
todegradation of pollutants in contaminated water: an intro- life forms had to either find an anaerobic ecological niche or
duction. In: Aquatic and Surface Photochemistry. G.R. Helz,
adapt to the ever-increasing O2 levels of the atmosphere.
R.G. Zepp, and D.G. Crosby (Eds.). Lewis Publishers, Boca
Today on Earth, the photosynthetic O2 generators are
Raton, FL, pp. 467–490.
widely distributed on land and in the oceans in the form of
17. Zepp, R.G., Hoigne, J., and Bader, H. (1987). Nitrate-induced
plants, algae, and cyanobacteria. The current atmospheric
photooxidation of trace organic chemicals in water. Environ.
pool of O2 contains about 4 × 1019 mols and takes about
Sci. Technol. 21: 443–450.
3.7 million years to turnover. The unique enzyme complex
18. Alif, A. and Boule, P. (1991). Photochemistry and the
that produces the O2 by photosynthetic organisms is called
environment part XIV. Phototransformation of nitrophenols
photosystem II (PSII), which acts as a water/plastoquinone
induced by excitation of nitrite and nitrate ions. J. Photochem.
Photobiol. A 59: 357–367.
oxido-reductase (2). PSII contains a special photochemical
reaction center, which at excitation can generate strong
19. Marbury, S.A. and Crosby, D.G. (1994). The relationship
oxidants that can be coupled to the splitting of two
of hydroxyl reactivity to pesticide persistence. In: Aquatic
and Surface Photochemistry. G.R. Helz, R.G. Zepp, and
water molecules into O2 , four protons, and four electrons.
D.G. Crosby (Eds.). Lewis Publishers, Boca Raton, FL, The protein motif within PSII that catalyzes the water
pp. 149–161. oxidation reaction contains a cluster of four Mn ions, a Ca
20. Beckett, R. (1990). The surface chemistry of humic substances
ion, and a redox-active tyrosine and is often referred to as
in aquatic systems. In: Surface and Colloid Chemistry in the O2 evolving complex (OEC).
Natural Waters and Water Treatment. R. Beckett (Ed.). The oxidation of water by PSII is driven photochemi-
Plenum Press, New York, pp. 3–20. cally by the capture of light quanta (3,4). The chlorophyll
21. Kieber, R.J., Zhou, X., and Mopper, K. (1990). Formation of binding proteins associated with peripheral regions of
carbonyl compounds from UV-induced photodegradation of the PSII complex serve as light-harvesting antenna and
humic substances in natural waters: fate of riverine carbon direct spectrally diverse quanta to the special reaction
in the sea. Limnol. Oceanogr. 35: 1503–1515. center chlorophyll cluster called P680. At absorption of a
22. Miller, W.L. (1994). Recent advances in the photochemistry light quantum, P680 generates an excited state with its
of natural dissolved organic matter. In: Aquatic and Sur- pheophytin partner (P680-Pheo)∗ , which in turn is in equi-
face Photochemistry. G.R. Helz, R.G. Zepp, and D.G. Crosby librium with the exciton in the antenna and the radical
(Eds.). Lewis Publishers, Boca Raton, FL, pp. 111–127. pair P680+ -Pheo− . The time course for this reaction has
23. Zafiriou, O.C. and Dister, B. (1991). Photochemical free been under some debate, but it is in the order of 3–20 ps.
radical production rates: Gulf of Maine and Woods Hole- The primary charge separation is rapidly followed by a
Miami transect. J. Geophys. Res. 96: 4939–4945. charge stabilization event, in which the electron is rapidly
536 ISOTOPE EXCHANGE IN GAS-WATER REACTIONS
transferred to a tightly bound plastoquinone molecule, QA , concerted four-electron event occurs during the S3 to S4
in 300–500 ps, which results in the formation of P680+ - to S0 transition to create the O–O bond (9,10). However,
Pheo QA − . The electron hole in P680+ is subsequently filled definitive experimental evidence to show at which step in
with an electron derived from a redox-active tyrosine, YZ , the S state cycle the substrate water binds and the O–O
located in the D1 protein subunit within some 40–380 ns, bond forms has until recently remained elusive.
which results in the formation of YZ ox P680-Pheo QA − . The binding of the substrate water to the catalytic site
The electron at QA − is then transferred to a second bound can be probed by oxygen isotope exchange measurements
plastoquinone molecule, QB , in 100–200 µs, whereas YZ ox between the solvent water and the photogenerated O2
is reduced by the Mn4 Ca cluster with S-state-dependent using mass spectrometry. Basically, the measurement
kinetics (see below) in the range of 30–1300 µs. involves the transfer of a photosynthetic sample into
One key finding for understanding the water oxidation labeled water of known oxygen isotopic composition (i.e.,
18
reaction is the observation that for dark-adapted, PSII- O-enriched water) and then determining the extent of
containing samples illuminated by a series of brief, isotopic incorporation into the photogenerated O2 .
saturating light flashes, O2 is released with a damped In the first oxygen isotope exchange measure-
periodicity of four. The peak O2 yields occur on the third, ments (11–13), the fastest detectable rate of exchange
seventh, eleventh, and so on flashes and eventually damp was limited to ∼30 s, mainly because of the long stabi-
out after the third or fourth cycle (5, see Figs. 2B and 2C). lization time of the ‘‘open’’ chamber system that intro-
The phenomenological model to explain this observation duced isotopically labeled water to the sample. However,
was developed by Kok et al. (6,7) and entails a cyclic in the more recent measurements from our group, the
reaction sequence in which the OEC traverses through stabilization time was greatly improved, by a factor of
five intermediate states called the S states. Beginning in nearly 5000, through the introduction of a ‘‘closed’’ cham-
S0 and progressing to S4 , each S state is advanced by a ber system (14,15). The results from this experimental
single quantum event at P680. When it reaches S4 , O2 is approach revealed for the first time that one substrate
released within 1–2 ms, S0 is regenerated, and the cycle water molecule is bound to the catalytic site through-
begins anew (see Fig. 1). out the S-state cycle, whereas the second substrate water
The damping in the O2 flash yield oscillations (which molecule is bound in at least the S2 and S3 states, and
is caused by the desynchronization of the forward S that the O–O bond indeed forms during the last step of
state transitions within a given population of O2 evolving the reaction sequence. This article documents the exper-
centers) is explained by a certain probability for a ‘‘miss’’ imental approach and the arguments that come to these
(α) to occur during a flash in which an electron hole fails conclusions.
to accumulate in the OEC and, depending on the flash
duration, by a certain probability for a ‘‘double hit’’ (β) to
RAPID 18 O EXCHANGE MEASUREMENTS DURING
occur in which a double excitation event takes place. The
PHOTOSYNTHETIC O2 EVOLUTION
maximum O2 yield on the third flash is explained by the
condition that the cycle initially starts in the S1 state after
To probe the substrate water bound at the OEC, excess
complete dark-adaptation. The higher S states (S2 and 18
O isotope (added as 18 O-labeled water) is introduced into
S3 ) exhibit lifetimes of tens of seconds to several minutes,
the sample to set up a new isotopic distribution, because
depending on the temperature, and deactivate to S1 if not
natural abundance 18 O in the solvent water is low at
advanced by another incoming quantum, whereas the S0
about 0.2%. The effect of the new isotopic distribution
state slowly relaxes in the dark to the S1 state (8). The S4
on the photogenerated O2 can then be measured at
state being metastable reacts immediately to produce O2
m/e = 32, m/e = 34, and m/e = 36 with an in-line mass
and the S0 state.
spectrometer (e.g., the VG Instruments IsoPrime stable
isotope spectrometer Manchester UK) for the 16 O16 O,
16 18
DURING WHICH STEP IN THE S STATE CYCLE DOES THE O O, and 18 O18 O di-oxygen products, respectively. From
SUBSTRATE WATER BIND? these measurements, the extent of 18 O incorporation
during the O2 evolving process can be determined.
In the original Kok et al. hypothesis, it was implied that Figure 2(a) shows the O2 flash yields at these three
the two substrate water molecules entered the reaction masses for spinach thylakoid membrane samples after
sequence during the last step, just before O2 release. the injection of an aliquot of H2 18 O (e.g., 98.5% enriched
Although this notion has been considered throughout the water from ISOTECH, Miamisberg, Ohio) that gives a
literature, the most recent models for the OEC generally final 18 O isotopic enrichment in the sample of ε ∼ 12%.
predict that the substrate water binds to the catalytic In this case, after complete equilibration, the absolute
site at the beginning of the S state cycle and that a yields (as determined by the extrapolated lines shown in
the figure for the third flash) at the three masses follow
closely the theoretical isotope equilibrium distribution:
hν hν hν hν 32:34:36 = (1 − ε)2 : 2ε(1 − ε) : ε2 = 100%, which indicates
that no major isotopic discrimination occurs during the
S0 S1 S2 S3 S4 O2 evolving process, as shown earlier from steady-state
O2 measurements (16).
Figure 2(b) shows a plot of the normalized O2 yields
Figure 1. OEC Cycle. at m/e = 32 as a function of flash number. The oscillatory
(a)
Mass spectrometer oxygen measurements
m/e = 32
(1940) (b) m/e = 32 Flash pattern
Normalised O2 yield
4
m/e = 34 2
1
(507)
0
2 4 6 8 10 12 14
Flash number
(c) Joliot electrode flash pattern
m/e = 36
Normalised O2 yield
(31) 3
0
2 4 6 8 10 12 14
Light flash Flash number Figure 2.
pattern in the mass spectrometric measurements follows is applied to photogenerate O2 . As the actual instrumental
very closely the flash pattern in Fig. 2(c), which was response time is relatively slow (being about 10 s because
obtained by the conventional amperometric method using of the diffusion of the O2 gas across a semipermeable
a so-called Joliot-type oxygen electrode. Both types of membrane into the inlet line from the sample chamber),
measurements give rise to similar Kok et al. parameters, the flash spacings of a subsequent flash sequence that
which indicates that the two techniques are measuring normalizes the oxygen signals are increased, typically to
the same phenomenon. 20 s. As such, to retard the deactivation reactions of the
To get information on the binding affinities of the higher S states during these flash spacings (see above), the
substrate water in the OEC, the rate of isotope exchange temperature of the sample is reduced (usually to 10 ◦ C),
between 18 O-labeled water and the photogenerated O2 which is accomplished by enclosing the sample chamber in
needs to be determined. To achieve this, our group a cooling block. Details of the setup have been published
designed a ‘‘closed’’ chamber system in which the sample earlier (63,64) and are given in Fig. 3.
volume is considerably reduced (160 µl) and a stirrer
is incorporated to facilitate rapid mixing. By employing S-STATE DEPENDENCE ON THE RATE OF ISOTOPE
a solenoid-actuated Hamilton syringe, 18 O-labeled water EXCHANGE BETWEEN 18 O-LABELED WATER AND THE
can be rapidly injected and equilibrated with the sample PHOTOGENERATED O2
within 6–8 ms. Thus, by measuring the extent of isotope
incorporation as a function of the time after injection of To determine S-state dependence on the rate of isotope
the 18 O-labeled water, the rate of isotope exchange can be exchange, the mass measurements are made as a function
determined. of time between the addition and equilibration of 18 O-
To evaluate the S-state dependence on the rate of labeled water with the sample in a particular S-state
isotope exchange, the ‘‘closed’’ chamber is optically coupled and the first peak yield of the photogenerated O2 . Thus,
to a bank of three xenon flash lamps (e.g., FX-193 lamp, for example, for the S3 state, an S1 -enriched sample
4 µF @ 1 kV capacitor from EG & G, Salem, MA) via suspended in H2 16 O is given two preflashes to advance
a 3-to-1 fiber optic to variously illuminate the sample the OEC into the S3 state. H2 18 O is then rapidly injected
with brief (FWHH ∼8 µs), saturating light flashes. The and followed by a variable delay time (t), which allows
samples are first given a series of activating flashes to for the time-dependent isotopic exchange to occur between
preset the sample into a particular S state (see below). the solvent water and the substrate water. A single
Then, after the rapid addition and equilibration of the flash is then given, and the oxygen signals at m/e = 34
18
O-labeled water, a subsequent turnover flash sequence and m/e = 36 are measured. The oxygen signals are
(spaced 5–10 ms apart, depending on S state, see below) then collected on subsequent flashes (in which complete
PC control
trigger pulses
H2 18O Injection
Mass spectrometer
Xe flash Magnetic
lamp stirrer
Sample chamber
(window, stirrer,
membrane, support)
Water cooling Water cooling
Figure 3. Temperature probe
exchange has occurred because of the 20 s flash spacings) The basic protocol involves the following procedure.
to normalize the data among the different sample aliquots Before loading, the samples are thawed in the dark at room
measured at different t values. temperature and given one saturating preflash followed by
The complete 18 O exchange kinetics were determined a 10-min dark period at room temperature to enrich the
for the different S states using the injection/flash protocols S1 population. The S1 -enriched sample is then loaded
depicted in Fig. 4. into the sample chamber in the dark, equilibrated to
10 ◦ C for 10–12 min, and then given several activating
flashes (spaced 1 s apart). At a given S state, H2 18 O is
rapidly injected into the chamber (t1/2 (mixing) ∼ 4 ms) and
a variable delay time (4 ms < t <200 s) is provided to
18
O2 Products allow the water bound in the OEC to exchange with the
H218O Injection 18
O-enriched solvent water. After the delay time, several
Exchange time rapid turnover flashes (spaced at 5, 10, and 10 ms apart
∆t for S2 , S1 , and S0 states, respectively) are then applied to
S3 S4 S0
advance the sample through to the S3 to S4 to S0 transition,
where the mass ratio of the photogenerated O2 (YM )
∆t
is determined. After this process, several normalization
S2 S4 S0 flashes (spaced 20 s apart) are given and their separate
S1 O2 yields (Yn(1→4) ) are measured. The YM is corrected for
background O2 contributions, double hits, and the injection
∆t response as described earlier (17,18) to give YC .
S1 S4 S0
The resolvable 18 O exchange kinetics for spinach
thylakoid samples as a function of the S-states are shown
in Fig. 5. Here, the corrected, normalized yields of O2
∆t
S0 S4 S0 (YC ) measured at m/e = 34 for the mixed labeled 16 O18 O
product (left side of the figure) and at m/e = 36 for the
Activating preflash(es) Turnover flash(es) Normalisation flashes
double labeled 18 O18 O product (right side of the figure) are
presented. Each point is a plot of Yc at a particular delay
Figure 4. time (t) between the injection of H2 18 O into the sample
and the turnover flash sequences that photogenerated O2 . be 88:12:0. On the other hand, at longer t, when the
The m/e = 36 data exhibit only a single kinetic phase and second substrate water molecule is also exchanging, the
are fit with a simple exponential function: mass distribution will be 77.44:21.12:1.44. Therefore, the
relative contributions of the fast and slow phases will
36
Y = (1 − exp(−36 k t)) (1) be unequal, with the fast phase representing ∼57% (i.e.,
12/21.12) of the total amplitude and the slow phase ∼43%.
In contrast, the m/e = 34 data on the left side in The 0.57:0.43 distribution between the two phases is found
Fig. 5 reveals two distinct kinetic phases. The insets consistently in the m/e = 34 data. Thus, the exchange
show expanded time ordinates for the fast phases. The kinetics in this case is fit exactly by the sum of two
two phases, however, are unequal in amplitude, with the exponentials:
fast phase constituting slightly more than half of the
total signal. The basis for this difference in amplitude 34
Y = 0.43(1 − exp(−34 k1 t)) + 0.57(1 − exp(−34 k2 t) (2)
is well explained by the enrichment condition for two
independent, exchanging sites. As the apparent kinetics
of the two phases differ by at least a factor of 10, the The solid lines in Fig. 5 represent the best fits to the
fast phase of exchange is virtually complete before the data points based on Eqs 1 and 2 for the m/e = 36 and
slow phase begins. Thus, at short t only one substrate m/e = 34 data, respectively. The exchange rates from the
water molecule is exchanging at the catalytic site, which corresponding kinetic fits at 10 ◦ C are given in Table 1.
means at a typical 18 O enrichment of ε = 12%, the mass The entire S-state results for the thylakoids have been
distribution at 32:34:36 for the two oxygen isotopes will presented earlier (17,18).
m/e = 34 m/e = 36
1.0 1.0
0.50 S3
0.5 0.5
Normalised O2 yield
0.25
0 .05 .1
1.0 1.0
1.0
S2
0.5 0.5
0.5
0 .05 .1
5 0 5 1.0 5.0 10.0 0.2 0.5 5 1.0 5.0 10.0

0.2 0.5 0.7 0.7
1.0 1.0
1.0
0.5 S1
0.5 0.5
Normalised O2 yield
0 .05 .1
25 50 75 10
0 0 0
15 30 25 50 75 0 0 0
10 15 30
1.0 1.0
1.0
S0
0.5 0.5 0.5
0 .05 .1 .15 .2 .25
5 0 0 0 5 0 0 0
0.2 0.5 15 30 0.2 0.5 15 30
Time (s) Time (s) Figure 5.
Table 1. Rate Constants for the 18 O Exchange in Spinach the solvent water must be translocated to the binding site
Thylakoid Samples as a Function of the S States at 10 ◦ C across a physical barrier.
S-State 36
k (s−1 ) 34
k1 (s−1 ) 34
k2 (s−1 ) Rates of whole water exchange at a metal site can span
a tremendous range (as much as 18 orders of magnitude),
S3 2.1 ± 0.2 1.9 ± 0.2 36.8 ± 1.9 and examples for various transition metals have been
S2 2.2 ± 0.1 1.9 ± 0.3 120 ± 6 summarized earlier (19–22). The most significant factors
S1 0.022 ± 0.002 0.021 ± 0.002 >175
governing water exchange at a metal site are the charge
S0 18 ± 3 8±2 >175
and ionic radius of the metal center and any electronic
Adapted from Ref. 18. occupancy of d orbitals. As a general rule, water exchange
rates will decrease (by several orders of magnitude) as
the metal center (Mn ) is oxidized and charge is increased;
IMPLICATIONS OF SUBSTRATE WATER BINDING FOR THE i.e., kex (Mn -OH2 ) > kex (Mn+1 -OH2 ). Furthermore, metal-
MECHANISM OF O2 EVOLUTION bound water can undergo deprotonation in response to an
increase in charge (and acidity) at the metal center. Thus,
Several striking features are found in the rate constants it may also be expected for water exchange to decrease
listed in Table 1. In each S state, the slow phase of at deprotonation for a given oxidation state; i.e., kex (Mn -
exchange was resolvable, whereas the fast phase of OH2 ) > kex (Mn -OH). In the case of the deprotonation
exchange could be resolved only in the S2 and S3 states. For of a neighboring ligand (L), the opposite effect may occur
the S0 and S1 states, the fast phase of exchange is equal to because the effective charge of the metal center is reduced,
or faster than the response time of the currently available As such the rate of water exchange will tend to increase;
experimental setup (∼100 − 175 s−1 ). Most importantly, i.e., kex (HL-Mn -OH2 ) < (L-Mn -OH2 ). The protonation state
the rate constants for the slow phase kinetics in the will thus be defined by the pK of the aquo metal complex.
m/e = 34 data are virtually identical to the rate constants Any number of possible intermediates in the water
for the monophasic kinetics in the m/e = 36 data. This last oxidation reaction can exist, and it is important to
result clearly indicates that the extent of 18 O incorporation appreciate that the 18 O exchange measurements could
into the photogenerated O2 is limited throughout the S- reflect not only whole water and hydroxyl exchange but
state cycle by the substrate water molecule undergoing also the exchange of µ-oxo (–O–), peroxo (–O–O–), and
the slow isotope exchange process and that the biphasic oxo (=O) intermediates. Indeed various models in the
behavior in the m/e = 34 data does not originate from PSII literature on O2 evolution invoke one or more of these
heterogeneity. Rather the two kinetic components in the intermediates (9,10). Although a dearth still exists in
m/e = 34 data reflect the binding of the substrate water at studies of 18 O exchange in model compounds, it is expected
two separate sites.
that the isotope exchange in µ-oxo (–O–), perxo (–O–O–),
The most relevant question to ask is at which S state
and oxo (=O) intermediates is much slower (in the order
does the substrate water bind. Based on the kinetic
of hours to days) than whole water exchange because of
resolution of our current mass spectrometric setup, a
tighter bonding and lack of protonation.
measurable 18 O exchange kinetic will conclusively show
The other type of mechanism that could limit the 18 O
the existence of bound water in a particular S state, which
exchange could be diffusion or isotope equilibration across
is justified by the observations that the O2 release on the
a physical barrier. The catalytic site for O2 evolution in
S3 to S4 to S0 transition takes place in ∼2 ms, whereas
PSII is located within the protein domain away from
the rest of the S-state advancement from S0 to S3 takes at
the solvent interface. In this situation, the structural
most another 1 ms (8). The catalytic site thus can make one
complete turnover in ∼3 ms. Therefore, any 18 O exchange barriers provided by the protein matrix may be mediated
that is slower than 3 ms would be indicative of bound through specific water chains or water channels (23),
water. From the rate constants for the slow phase of 18 O something akin to what has been proposed for the
exchange listed in Table 1 for the m/e = 34 data, it is cytochrome c oxidase (24,25). Such a water channel may
clear that one substrate water molecule is bound to the provide an important function in optimizing O2 formation
catalytic site through the S-state cycle. Interestingly, the by controlling the accessibility of the water substrates
results for the fast phase of exchange show that the second and O–O bond formation at the catalytic site, thereby
substrate water molecule is bound in at least the S2 and S3 minimizing side reactions that lead to the formation of
states, but they are inconclusive for the S0 and S1 states reactive oxygen intermediates (23). The inability to resolve
because the fast exchange kinetics in these S states are not the fast phase of exchange in the S0 and S1 , states (Table 1)
resolvable by our current setup (i.e., faster than 6–8 ms). could indicate that the second substrate water molecule
Even though the measurable exchange rates listed in only enters the reaction sequence at the S2 state. However,
Table 1 are slower than the total turnover time of the such a possibility can only be clarified by resolution of the
catalytic site, the magnitudes of these exchange rates fast phase kinetics or by the progressive improvement in
in general are all remarkably fast, where the slowest the resolution of crystal structure of PSII (26–28).
exchange is measured in the S1 state with t1/2 ∼50 s. The S-state dependence of the 18 O exchange kinetics for
Exchange rates in this magnitude range indicate relatively the substrate water are clearly complex, but these results
labile binding sites and can be interpreted for two overall place limits on the types of mechanisms that may be
mechanisms: (1) exchange of water bound to a metal site involved in the oxidation of water during photosynthesis.
or (2) diffusional or isotope equilibration processes, where Although the S-state-dependent changes in the kinetics
RADON IN WATER 541
may in the first instance be interpreted for metal- 22. Richens, D.T. (1997). The Chemistry of Aqua Ions. John Wiley,
centered oxidations and/or substrate or other ligand West Sussex, UK.
protonation/deprotonation events, second-order effects 23. Wydrzynski, T., Hillier, W., and Messinger, J. (1996). Physiol.
mediated by the protein environment, such as H-bonding Plant. 96: 342–350.
networks and structural conformations, will also have 24. Iwata, S., Ostermeier, C., Ludwig, B., and Michel, H. (1995).
to be considered. To sort out the fine details, additional Nature 376: 660–669.
measurements will have to be made under different 25. Tsukihara, T. et al. (1996). Science 272: 1136–1144.
biochemical conditions and in mutants of the PSII. This 26. Zouni, A. et al. (2001). Nature 409: 739–743.
work is currently in progress. 27. Kamiya, N. and Shen, J-R. (2003). Proc. Natl. Acad. Sci. USA.
100: 98–103.
Acknowledgments 28. Ferreira, K.N. et al. (2004). Science 303: 1831–1838.
The authors gratefully acknowledge the various contributions
to this research by Johannes Messinger, Murray Badger, and
Garth Hendry. This research was supported in part by a Human RADON IN WATER
Frontier Science Program grant (RGP 29/2002).
BABS A. MAKINDE-ODUSOLA
Riverside, California
BIBLIOGRAPHY
1. Blankenship, R.E. and Hartman, H. (1998). T.I.B.S. 23:

94–97. Some groundwater contains the naturally occurring
radioactive gas, radon-222 (radon). Radon dissolved in
2. Wydrzynski, T. and Satoh, K. (Eds.). (In press). Photosystem
II: The Water/Plastoquinone Oxido-Reductase in Photosyn- groundwater is released into indoor air when plumbed in
thesis. Springer-Kluwer Academic Publishers, Dordrecht, The water is agitated indoors. Radon released from domestic
Netherlands. water typically contributes less than 2% of total indoor
3. Diner, B.A. and Rappaport, F. (2002). Annu. Rev. Plant. Biol. radon. This article describes the health risks from radon
53: 551–580. in domestic water and the efforts within the United States
4. Barter, L.M.C., Durrant, J.R., and Klug, D.R. (2003). Proc. to regulate radon in drinking water. The U.S. National
Natl. Acad. Sci. USA. 100: 946–951. Research Council (NRC) found that the primary health
5. Joliot, P., Barbieri, G., and Chabaud, R. (1969). Photochem. risks from radon in domestic water are lung cancer
Photobiol. 10: 309–329. and stomach cancer, primarily to ever-smokers. Radon
6. Kok, B., Forbush, B., and McGloin, M. (1970). Photochem. concentration in air could reach levels considered unsafe
Photobiol. 11: 457–475. for underground miners at some water treatment facilities
7. Forbush, B., Kok, B., and McGloin, M. (1971). Photochem. that aerate water as part of the treatment process.
Photobiol. 14: 307–321. Inhalation of radon gas and its progeny may result in lung
8. Joliot, P. and Kok, B. (1978). Bioenergtics of Photosynthesis. cancer (inhalation risk). An estimated 11% of the health
Govindjee (Ed.). Academic Press, New York. risk is from direct ingestion of radon dissolved in drinking
9. Nugent, J.H.A. (Ed.). (2001). Special Issue: Photosynthetic water (ingestion risk). The NRC found that radon is a
Water Oxidation. Biochim. Biophys. Acta 1503. greater source of cancer risk than any other regulated
10. Messinger, J. and Lubitz, W. (Eds.). Special Issue: Photosyn- chemical in drinking water. Some ecological studies
thetic water oxidation. Phys. Chem. Chem. Phys. (In press). contradict using the linear nonthreshold (LNT) model to
11. Radmer, R. and Ollinger, O. (1980). F.E.B.S. Lett. 110: 57–61. assess the health risks from exposure to radon at low levels
typical of indoor air. The U.S. Environmental Protection
12. Radmer, R. and Ollinger, O. (1985). F.E.B.S. Lett. 195:
285–289. Agency (EPA) relied on the LNT model to assess health
risks and propose standards and goal for radon in drinking
13. Bader, K.P., Thibault, P., and Schmid, G.H. (1987). Biochim.
Biophys. Acta 893: 564–571. water. Some water utilities could opt for treatment to
comply with the radon in drinking water standard. Some
14. Messinger, J., Badger, M., and Wydrzynski, T. (1995). Proc.
Natl. Acad. Sci. USA 92: 3209–3213.
other water utilities might avoid treatment by adopting a
multimedia approach to radon mitigation. The multimedia
15. Hillier, W., Messinger, J., and Wydrzynski, T. (1998). Bio-
chemistry 37: 16908–16914.
approach could be more cost-effective and achieve greater
health risk reduction. Noneconomic considerations may
16. Guy, R.D., Fogel, M.L., and Berry, J.A. (1993). Plant. Phys.
101: 37–47.
influence that decision. Health risk reduction and cost
analyses (HRRCA) suggest that most of the benefits would
17. Hillier, W. and Wydrzynski, T. (2000). Biochemistry 39:
4399–4405.
accrue to ever-smokers and smaller water systems could
incur most of the costs.
18. Hillier, W. and Wydrzynski, T. (In press). Phys.Chem. Chem.
Phys.
19. Hillier, W. and Wydrzynski, T. (2001). Biochim. Biophys. Acta INTRODUCTION
1503: 197–209.
20. Hendry, G. and Wydrzynski, T. (2003). Biochemistry 42: Radon, in this article, refers to the specific isotope,
6209–6217. radon-222, the daughter product of radium-226 and a
21. Lincoln, S.F. and Merbach, A.E. (1995). Adv. Inorg. Chem. member of the uranium-238 decay series. The decay
42: 1–88. scheme for uranium-238 is illustrated in Drago (1) and
542 RADON IN WATER
NRC (2). Uranium is found in trace amount in soils the United States. The NRC (2) found that radon is a
everywhere. Radon escapes as a gas from soils into greater source of cancer risk than any other regulated
groundwater, indoor air, and outdoor (ambient) air. The chemical in drinking water. Evaluating, and where neces-
U.S. Geological Survey (3) reported that soil air within sary, mitigating the level of indoor radon is common during
the United States contains between 200 and 2,000 pCi/L real estate transactions in the northeastern United States,
(1 pCi/L = 37 Bq m−3 ) of radon. Radon is a chemically even though compliance with the action level is voluntary.
inert, colorless, and odorless gas whose half-life is about That practice could extend to testing for radon in pri-
3.82 days. Radon emits alpha (α) particles as it decays vately owned wells after the EPA finalizes the maximum
to a series of solid short-lived radionucleotides, such as contaminant level (MCL) for radon in drinking water.
polonium-218 and lead-214. Those physical properties Regulating radon in domestic water had been very contro-
are important in understanding the health risks from versial in the United States and dates back to 1986. In
exposure to radon and the mitigation of such risks. 1986, Congress directed the EPA to establish a manda-
The United Nations Scientific Committee on the Effects tory standard for radon in drinking water as part of the
of Atomic Radiation (4) estimated that about 50% of the amendments to the Safe Drinking Water Act (SDWA).
average annual effective dose from natural sources of The EPA (8) proposed the radon in drinking water rule
ionizing radiation worldwide is from exposure to the alpha- using the risk management approach typically used for
particles produced by radon and its progeny. Potential drinking water contaminants. Amendments to the SDWA
contributors of radon to indoor air include soil gas, natural in 1996 directed the EPA to withdraw the proposed rule
gas, building materials, and to a lesser extent—domestic of 1991 and repropose after an independent risk assess-
water. Radon emanating from soils under homes is the ment by the NRC. The amendments include provisions for
primary source of radon in indoor air, which can range a multimedia approach to achieving an equal or greater
from <1 to 3,000 pCi/L (3). level of health risk reduction. In 1999, the EPA (5) pro-
Radon in drinking water is typically of concern to posed 300 pCi/L and 4,000 pCi/L as the MCL and alternate
sources that rely on groundwater, where there had MCL, respectively, for radon in drinking water. The EPA is
not been sufficient opportunity for significant decay expected to finalize the radon in drinking water rule after
nor release into the atmosphere before reaching indoor 2004. Key inputs and major milestones in developing the
plumbing. A portion of the radon in plumbed water radon rule are illustrated in Fig. 1 of Article MW-66 (9).
used indoors is released into the air, especially during
agitation (e.g., laundering, dishwashing, humidifying, OCCURRENCE OF RADON IN DOMESTIC WATER
toilet flushing, etc.), which increments the level of radon
in indoor air. The NRC (2) estimated that the average Radon levels in water, like those in ambient air and
transfer coefficient is about 0.01% (1 in 10,000) of radon indoor air, vary temporally and spatially. The EPA (5)
in water; that is 10,000 pCi/L of radon in water would analyzed data from the Safe Drinking Water Information
increase the level of radon in indoor air by 1 pCi/L. The System (SWDIS) to estimate the nationwide occurrence of
U.S. EPA (5) estimated that less than 2% of radon in radon in community water systems (CWS). The EPA (10)
indoor air emanates from domestic water. This author found geographic disparity in its occurrence. New England
prefers to emphasize ‘‘domestic water’’ because of the CWS had the highest arithmetic and geometric mean
widespread use of bottled water for drinking and because concentrations of 2933 pCi/L and 1214 pCi/L, respectively.
the NRC (2) found that the major part of the health risk is Some privately owned wells, not monitored nor regulated
from inhalation of radon and its progeny. by the EPA, might contain radon at much higher levels.
Some countries have policies to reduce public exposure The USGS (3) reported that the radon level in groundwater
to radon, especially in indoor environments. Cole (6) ranges from about 100 pCi/L to nearly 3,000,000 pCi/L.
provides the historical development of radon risk policies The EPA (5) estimated the population-weighted national
in the United States. The U.S. Indoor Radon Abatement average concentration of radon in CWS at 213 pCi/L, and
Act of 1988 established the goal of reducing indoor radon a credible range of 213 to 240 pCi/L. In general, smaller
concentrations to natural ambient level. The NRC (2) CWS that rely on groundwater have the highest levels
estimated that the arithmetic national ambient average of radon. Radon levels exceeded 300 pCi/L at more than
level of radon was 0.4 pCi/L. The U.S. EPA established 51% and 16% of the CWS serving fewer than 100 people
a voluntary action level of 4 pCi/L for radon in indoor and 10,000 people, respectively (5). Radon levels exceeded
air. The indoor radon action level in other countries 300 pCi/L and 4000 pCi/L in the domestic water of more
is as high as 27 pCi/L (7). Indoor radon guidelines are than 16 million people and 77,000 people, respectively (5).
22 pCi/L and 11 pCi/L in Canada and the European Union, The EPA (5) and NRC (2) have additional information
respectively (7). The U.S. EPA (5) reported that more than on the occurrence and spatial distribution of radon in
2500 lives were ‘‘saved’’ through the year 2000 since the domestic water, indoor air, and ambient air. SWDIS would
inception of its voluntary indoor radon abatement program contain additional radon occurrence data when the radon
in 1986. rule is finalized and CWS begin compliance monitoring.
Many countries are developing policies to reduce expo- Radon levels at a particular well could fluctuate
sure to radon in domestic water. The World Health Organi- significantly. The EPA (5) proposed a monitoring regime
zation (WHO) is creating a database of national radon reg- that would require CWS using groundwater to monitor
ulations [http://www.who.int/ionizing radiation/env/ for radon initially at each entry point to the distribution
radon/en/]. The rest of this article focuses on efforts within system, quarterly for 1 year. Compliance would be based
RADON IN WATER 543
Figure 1. Estimated fatalities in the U.S. by pathway from radon in drinking water provided by CWS.
on the average of four consecutive quarterly samples. of exposure to radon. The use of the LNT model to assess
Zikovsky and Chah (11) found that concentrations of radon the health effects from exposure to radon at typical indoor
are lognormal in distribution and that the arithmetic concentrations is controversial. The National Council on
mean of concentrations might create a false impression Radiation Protection and Measurements (14) evaluated
about exposure levels. The EPA (5) estimated that the evidence for and against the LNT dose–response
26.8% of sources whose radon concentrations averaged hypothesis and found that the LNT model is more plausi-
between 200 and 300 pCi/L during the first four quarterly ble. The NCRP (14) did not exclude other dose–response
measurements would have long-term exceedances of relationships because of growing evidence that adaptive
300 pCi/L. Field et al. (12) found higher levels of radon in responses, bystander effects, and other variables may
domestic water at the point of use than prior to distribution modify the dose–response relationship. The NRC Com-
because of increments from decay of the radium-226 mittee on the Biological Effects of Ionizing Radiation
deposits within the older portions of water distribution (BEIR) VI did not exclude the possibility of a thresh-
systems. The EPA (5) identified man-made sources of old relationship at the very low levels of radon exposure
radon in groundwater to include releases from wastes typically found in indoor air. The NRC (2) supported
from radium or thorium processing and from phosphate using the LNT relationship to derive quantitative esti-
and uranium mining and milling operations. mates of cancer risks in the absence of credible evidence
to the contrary. The NRC (2) recommended reconsidera-
RISK ASSESSMENT OF RADON IN DOMESTIC WATER tion of the LNT assumption when the ‘‘probability of an
exposed cell becoming transformed’’ and the ‘‘latent period
The health risks from radon in domestic water are pri- before malignant transformation is complete’’ are known.
marily lung cancer and stomach cancer. Radon is a Lorimore et al. (15) observed the potential for interaction
human carcinogen based on epidemiological studies of between irradiated and nonirradiated cells in the produc-
underground miners, cellular mutagenesis studies, and tion of genetic damage and its significant implications for
experimental research in animals (13). A linear nonthresh- risk estimation. Brenner and Sachs (16) suggested that
old (LNT) model is often used to assess the health risks bystander effects from alpha-particle damage could cause
544 RADON IN WATER
downwardly curving dose–response relations and inverse The NRC (2) developed a new model of ingested radon
dose-rate effects. diffusion because its review of published ingestion risks
Parsons (17) presented a biological model of radiation found a wide range in the estimates. The uncertainties
hormesis that invalidated the LNT. Cohen (18) tested were mainly due to the assumed model of radon absorption
the LNT model by analyzing lung cancer mortality rates in the gastrointestinal tract. The NRC (2) noted the lack
and average radon concentration in homes in 1601 U.S. of data on the extent to which radon diffuses through
counties while addressing problems typical of ecological the stomach wall and recommended research to reduce
studies. Cohen (18) found a ‘‘strong tendency for lung can- the uncertainty regarding ingestion risks. Details of the
cer rates to decrease with increasing radon exposure in mathematical models of radon diffusion in the stomach
sharp contrast to the increase’’ expected from the LNT and the physiologically based pharmacokinetic (PBPK)
model. The NRC (2) recommended investigating the sig- model of radon behavior in the body can be found in
nificant negative trend of county lung cancer mortality the NRC (2). The NRC (2) estimated the ingestion risks
with increasing home radon concentration established in using estimates of the amount of water ingested, the
the ecological study by Cohen (18). Puskin (19) found sim- fraction of radon lost from water before ingestion, and
ilar ‘‘strongly negative correlations for cancers strongly calculations of the dose absorbed by tissues exposed. The
linked to cigarette smoking’’ but did not find ‘‘such cor- NRC (2) estimated the confidence limits on the ingestion
relations’’ for cancers not linked to smoking. Puskin (19) risk but noted, ‘‘the actual risk from ingested radon could
explained that the ecological correlation (18) ‘‘provides no be as low as zero depending on the validity of the LNT
substantial evidence for a protective effect of low level dose–response hypothesis.’’ Yamaoka el al. (2004) found
radon exposure.’’ Van Pelt (2003) found that altitude (ele- inhalation of radon gas from radon enriched water induced
vation) explains some of the inverse association between a small amount of active oxygen in the blood, decreased
radon and lung cancer. Van Pelt (2003) postulated that the levels of lipid peroxide and total cholesterol in the
the decrease in lung cancer is ‘‘caused by the carcino- blood of osteoarthritis patients alleviating pain related to
genic effect of higher absolute oxygen concentration in the peroxidation reactions and immune depression.
Table 1 is a summary of the risks estimated by the
inspired air at lower.’’
NRC (2). Table 1 shows that males have higher lifetime
The most direct way to assess the effects of radon
risks than females and that the inhalation risk to ever-
in homes is through case-control studies (20). Krewski
smokers is five times as high as those for never-smokers
et al. (13) summarizes case-control studies of residential
of either sex. Notice that the aggregate health risk to the
radon and lung cancer in North America. Field and
general population is thrice as high as the risk for never-
Duport (21) presented points and counterpoints regarding
smokers due to the inclusion of ever-smokers. Figure 1
lung cancer risks from exposure to residential radon.
shows the relative distribution of cancer fatalities by
The NRC (2) reviewed available epidemiological studies
exposure pathways. Notice in Fig.1 that 89% of the health
of indoor radon and concluded ‘‘it is low and difficult,
risks are from inhalation of the radioactive by-products
if not impossible to detect given the high background
of radon released from water into indoor air. About 11%
lung-cancer mortality in the populations studied.’’
of the health risk is from direct ingestion of tap water
Inhalation of radon gas, especially its progeny, causes (0.6 L/day directly from a tap). An earlier risk assessment
lung cancer primarily among ever-smokers. BEIR VI (20) by the EPA (5) estimated that the proportion of ingestion
evaluated the risks of domestic radon to ever-smokers and and inhalation risks were 53% and 47%, respectively.
never-smokers separately. CDC (22) defines ever-smokers Both the NRC (2) and the EPA (5) found that the
as people who had smoked at least 100 cigarettes in their hazards of exposure to radon during showering are likely to
entire lifetimes. Alavanja (23) found many similarities be small because the buildup of radon progeny is less than
between the pathologic lung-cancer mechanisms observed 4% of the maximum possible value during a typical shower
among those exposed to tobacco smoke and those exposed that lasts 10 minutes. The NRC (2) found it highly unlikely
to radon gas. that radon in drinking water has substantial teratologic or
NRC (2) conducted the most comprehensive review of reproductive effects. The NRC (2) found that information
the health risks of radon in domestic water. That review was insufficient to determine radon risks to ‘‘sensitive’’
included epidemiological studies, BEIR studies, and sub-populations, such as children. However, about 30% of
earlier EPA (8,24) risk assessments. Both BEIR VI (20) fatal stomach cancers are attributed to exposure before
and the NRC (2) reviewed the mechanistic information age 10 (2). The NRC (2) suggested that there might be an
on alpha-particle-induced carcinogenesis. The NRC (2) opportunity interval for lung-cancer risk reduction from
postulated that the biological effects of exposure to radon exposure during childhood. Bilban and Vaupotic (26) found
at the low level of typical indoor air are initiated by higher chromosome aberrations in pupils at elementary
the passage of single alpha-particles that have very high schools that had high indoor radon levels. Fewer than 34%
linear energy transfer (LET). Hei et al. (25) described the of lung cancer patients are diagnosed before age 65 (27).
mutagenesis of a single alpha-particle in mammalian cells. Many researchers have estimated fatalities from radon
Genes involved at low dose are different from those in domestic water. Mills (28) estimated 60, 400, and 5000
involved at higher doses and adaptive responses in cells annual fatalities from radon in surface water, public
exposed to low-level ionizing radiation allow the cells to groundwater, and private wells sources, respectively. The
better withstand a much higher dose later (priming effect) NRC (2) used the risk estimates (Table 1) and information
[http://www.llnl.gov/str/JulAug03/Wyrobek.html]. regarding the occurrence of radon in CWS, daily water
RADON IN WATER 545
Table 1. Lifetime and Unit Risks Posed by Exposure to Radon in Drinking Water at 1 pCi/L (37 Bqm−3 )a
Lifetime Riskb
Unit Risk
Exposure Pathway Male Female U.S. Population U.S. Population
Inhalation (ever-smokers) 1.15 × 10−6 7.40 × 10−7 9.62 × 10−7 1.31 × 10−8
Inhalation (never-smokers) 2.18 × 10−7 1.48 × 10−7 1.85 × 10−7 2.43 × 10−9
Inhalation (population) 7.77 × 10−7 4.44 × 10−7 5.92 × 10−7 7.90 × 10−9
Ingestion* (population) 5.55 × 10−8 8.51 × 10−8 7 × 10−8 9 × 10−10
Total risk (inhalation and ingestion) 8.14 × 10−7 5.18 × 10−7 6.66 × 10−7 8.89 × 10−9
*Breakdown of Ingestion Risks

Stomach 5.92 × 10−8
Colon 2.18 × 10−9
Liver 2.15 × 10−9
Lung 1.26 × 10−9
General tissue 2.92 × 10−9
Total ingestion 7 × 10−8
a
Data sources: Reference 2, pp. 16, 17; Reference 5, Table X11.3.
b
Annual unit risk can be obtained by dividing lifetime unit risks by the expected lifetime for each population category. Expected lifetime in years:
Ever-smokers = 73.7; never-smokers = 76.1; combined population = 74.9.
Table 2. Health Risks from Radon in Drinking Water

a. Estimated Annual U.S. Fatalities Due to Radon in ∗ Drinking Water
Health Risk Source of Estimate Central Estimate Range
Lung cancer NRC (2) 160 25–290

Lung cancer EPA (5) 148 70–263
Stomach cancer NRC (2) 23 5–50
Stomach cancer EPA (5) 18 3–38
b. Occurrence of Lung Cancer due to Radon by Exposure Pathway
Pathway Central Estimate Range
Outdoor air (2) 720 120–300

Indoor air (5) 18200 3,000–33,000
*Drinking Water (2) 160 25–290
*Drinking Water (5) 148 70–263
c. Comparative Health Risks of Radon in *Drinking Water
Annual Fatalities (Central Estimates) % of All Causes

Due to *Drinking
Health Risk All Causes *Drinking Water Water
Lung cancer (2) 160,000 160 0.10%

Stomach cancer (2) 14,000 18 0.13%
Data sources: NRC(2)page 18; EPA(5).

Fatalities based on 1998 data.
∗
Potable water delivered by a community water system (CWS).
intake, and smoking patterns among several factors to domestic water caused 148 lung cancer and 18 stomach
derive estimates of cancer fatalities in the United States. cancer fatalities annually. The EPA (5) also estimated
Table 2 summarizes the estimated U.S. fatalities due to (screening level) that an additional 36 fatalities annually
radon in domestic water relative to other sources. The NRC were from radon in water delivered outside residences, at
estimated that in 1998 radon from soil gas caused about schools, offices, etc.
18,200 lung cancer fatalities in the United States. The
NRC (2) also estimated that 160 lung cancer fatalities were
due to radon in domestic water. Notice that an estimated RISK MANAGEMENT
720 fatalities were from an unavoidable background level
of radon in outdoor air. The NRC (2) estimated that 23 The EPA (8) first proposed 300 pCi/L and 0 pCi/L as
stomach cancer fatalities were from ingesting radon in the maximum contaminant level (MCL) and MCL goal
drinking water. The EPA (5) estimated that radon in (MCLG), respectively, for radon in drinking water. Many
546 RADON IN WATER
criticized the proposed rule because it was not considered average costs of indoor radon screening for all homes and
cost-effective. In addition, a multimedia approach could mitigation cost for each home at or above the 4-pCi/L
be more cost-effective and achieve greater health risk indoor air action level). The EPA (5) estimated annual-
reduction. In 1996, as part of the amendments to the ized national costs at $408 million and monetized annual
SDWA, Congress directed the EPA to withdraw the benefits at $362 million if 100% of impacted CWS choose
proposed rule and re-propose using an updated risk to comply with the proposed MCL and assuming costs of
assessment provided by the NRC. Congress also directed capital of 7%. The estimated national annualized costs
the EPA to develop an alternate MCL (AMCL) and and monetized benefits are $60 million and $301 million,
guidelines for multimedia mitigation (MMM) programs respectively, under a scenario that assumes that 95% of
(MMMP). The NRC (2) defined the AMCL as the states implement MMMP and 90% of the CWS in each
concentration of radon in water that would cause an state comply with the AMCL. The EPA (5) prefers that
increase of radon in indoor air that is no greater than states adopt the AMCL approach and develop MMMP but
the level of radon naturally present in outdoor air. The the NRC (2) noted that noneconomic considerations, such
NRC (2) determined that the AMCL should be about as equity, could be deciding factors in whether to under-
4000 pCi/L based on the national average outdoor air take MMMP. The EPA (5) estimated that annual MCL
radon concentration of 0.4 pCi/L and the water to air compliance costs per household are about $270 and $10
transfer factor of 10,000. EPA is only to establish an AMCL for CWS serving 25 to 100 people and more than 100,000
if the MCL is more stringent than the water equivalent of respectively. The HRRCA shows that small CWS would
the national average of radon in outdoor air. incur most of the treatment costs but that the major-
The EPA (5) used NRC risk estimates (Table 1) to ity benefits would go to larger systems under the MCL
repropose the radon rule (MCL = 300 pCi/L; MCLG = approach. Some have argued that the EPA underestimated
0 pCi/L, AMCL = 4000 pCi/L). CWS whose radon levels the costs of compliance and overstated the benefits. Costs
are within the range of the MCL and the AMCL have could be higher because of unique local conditions and
the option of complying with either the MCL or the less environmental permitting. Hahn and Burnett (29) found
stringent AMCL. CWS that choose to comply with the that the proposed radon rule was unjustifiable based on
AMCL must participate in an approved and verifiable cost–benefit analyses.
MMMP that achieves at least an equivalent level of
risk reduction as treatment to the MCL. CWS may
TREATMENT TECHNOLOGIES
opt to develop individual MMMP in accordance with
EPA guidelines if the primacy state chooses not to
CWS may choose to comply with either the MCL or,
develop an MMMP. The EPA (5) proposed four criteria
under certain circumstances, with the AMCL. Treatment
required for approval of MMMP: involve the public in
of radon in domestic water will reduce the health risks
developing MMMP; include quantitative goals to measure
from both direct ingestion and inhalation through a
the effectiveness of MMMP; include implementation
decrement. Deb (30) reported on the decrements in indoor
plans, strategies and specific activities for achieving
radon level after water treatment in three New Hampshire
quantitative goals; and include plans for measuring and
communities. Deb (30) found that significant reduction in
reporting results. Examples of MMMP include mitigation
the indoor radon level, occurred only in homes served by
of radon levels in existing construction and use of radon-
the CWS where water (radon concentration >20,000 pCi/L
resistant techniques for new construction. CWS that
in water) was the major source of radon in indoor air.
choose the AMCL approach are not required to treat,
The EPA (5) identified aeration as the best available
but to participate in approved MMMP. The implemented
technology (BAT) for removing radon from domestic water
MMMP might not reduce the ingestion risk, but reduce
in large CWS. Article DW-32 (31) contains additional
the inhalation risks, though not necessarily in the
information regarding radon removal from drinking water.
same community, nor in the same homes. Some CWS
Aeration treatment methods such as a packed tower,
systems may opt to purchase water from other sources
diffused bubble, spray, slat tray, and cascade aeration
or consolidate with other CWS in lieu of treating or
are very efficient in removing radon from water because
implementing the MMMP.
of the volatility of radon gas. Liquid-phase granular
activated carbon (GAC) removes radon by adsorption
HEALTH RISK REDUCTION AND COST–BENEFIT but requires longer empty bed contact time (EBCT)
ANALYSES (HRRCA) ranging from 20 to 130 minutes, and its performance
depends on water quality (1). The EPA (5) identified
Radon is the first drinking water contaminant for which point-of-entry (POE) GAC as one of the affordable
the EPA (27) used the new cost–benefit analyses frame- small systems compliance technologies (SSCTs). Drago (1)
work. The HRCCA (27) included detailed cost–benefit reported removal efficiencies of 20% to greater than 99% at
analyses by water system size for each considered MCL 34 water treatment facilities and found that construction
and MMMP adoption. In estimating benefits, the EPA (27) costs for either BAT are site specific. Either treatment
used the value of $5.8 million (1997 dollars) as the value technology requires disinfection after treatment, which
of statistical life (VSL) for each avoided cancer fatality EPA believes, will reduce the microbial risk and provide
and willingness to pay (WTP) of $536,000 for nonfatal the additional benefit of complying with the proposed
cancers. The EPA (5) estimated that the cost for MMMP groundwater rule (32). The EPA (5,27) found that high
is $0.7 million per avoided lung cancer (the estimated levels of dissolved iron and manganese co-occurring
RADON IN WATER 547
with radon in water may complicate and increase radon • CWS may choose to comply with the AMCL rather
treatment costs. than the MCL. In some scenarios, health benefits
The EPA (5) reviewed the permitting and other from risk reduction might accrue to communities
environmental issues of either technology. Beta-gamma outside the service area of the CWS. Health benefits
radiation from radon progeny might pose concern to GAC might accrue to selected individuals rather than
operators (1). A major concern of aeration technologies uniform risk reduction to all consumers.
is off-gas emissions. Fisher et al. (33) found that radon
concentration in air could reach levels considered unsafe Ayotte et al. (35) estimated that a 0.05% reduction in the
for underground miners at some water treatment facilities prevalence of smoking would prevent as many deaths from
that aerate water as part of the treatment process. lung cancer as mitigation of indoor radon. The NRC (2)
The NRC (2) and EPA (5) recommended monitoring for noted ‘‘except in situations where concentrations of radon
radon in air around facilities treating groundwater for in water are very high, the reduction of radon in water
any contaminant and mitigation, if necessary, to reduce will generally not make a substantial reduction in the
exposure. The EPA (5) modeled risks from potential air total radon-related health risks to occupants of dwellings
emissions of radon off-gas from aeration facilities based served by the water supply.’’ The NRC (2) expressed
on data from 20 CWS whose radon levels ranged from similar concerns about the lack of data regarding the
1000 pCi/L to 16,000 pCi/L. The EPA (5) estimated a ‘‘overall effects of radon-resistant construction methods
maximum lifetime risk of 2 × 10−5 and a crude national on reducing concentrations of radon in indoor air radon
estimate of 0.09 cancer fatalities per year if all CWS concentrations.’’
comply with the MCL. An air dispersion modeling study
for the East Valley Water District, San Bernardino, BIBLIOGRAPHY
California, suggests that the PTA (Packed Tower Aeration)
downwind risk estimates could exceed the local regulatory 1. Drago, J.A. (1998). Critical Assessment of Radon Removal
risk threshold (34). The CDM (34) model predicted that a Systems for Drinking Water Supplies. American Water Works
maximum individual cancer risk at some locations near the Association Research Foundation, Denver, CO.
PTA could exceed 10−5 and 0.2 as the mean cancer burden 2. National Research Council. (1999). Risk Assessment of Radon
for the service area. Preliminary design parameters for in Drinking Water. National Academy Press, Washington,
PTA to mitigate risks from radon emissions were not DC.
feasible from design/construct perspectives, for example, 3. U.S. Geological Survey. (2000). The Geology of Radon. U.S.
stack height up to 200 feet (34). Geological Survey, Denver, CO.
4. United Nations Scientific Committee on the Effects of Atomic
Radiation. (1993). Report to the General Assembly. United
CONCLUSIONS
Nations, New York. Available: http://www.unscear.org/
reports/2000 1.html
The preferred risk management model for radon in
5. U.S. Environmental Protection Agency. (1999). National
domestic water differs from those of other regulated
Primary Drinking Water Regulations; Radon-222: Proposed
drinking water contaminants because Rule. 40 CFR Parts 141 and 142. Federal Register (November
2, 1999) 64(211): 59246–59378. [64 FR 59246]. U.S. Govt.
• Amendments to the SDWA in 1996 require the EPA Printing Office, Washington, DC.
Administrator to take into account the costs and
6. Cole, L.A. (1993). Element of Risk: The Politics of Radon.
benefits of control programs for radon from other AAAS Press, Washington, DC.
sources in setting the MCL.
7. World Health Organization. (2002). Radon and Health. Final
• The major source of radon in indoor air is not Information Sheet. October 2002. Available: http://www.
regulated. Domestic water, which contributes on who.int/ionizing radiation/env/Radon Info sheet.pdf
average less than 2% of radon in indoor air, would be 8. U.S. Environmental Protection Agency. (1991). National Pri-
regulated. Areas that have the highest indoor radon mary Drinking Water Regulations; Radionuclides: Proposed
level may not coincide with service areas of CWS that Rule. 40 CFR Parts 141 and 142. Federal Register (July 18,
have the highest level of radon in domestic water. 1991) 56(138): 33050–33127. [56 FR 33050]. Govt. Printing
• The majority of health risks is not from direct Office, Washington, DC.
ingestion of water, but rather from radon released 9. Makinde-Odusola, B. (2004). MW-66 drinking water quality
during domestic use. Drinking water standards are standards—United States. Encyclopedia of Water. John
Wiley & Sons, New York.
usually set based on risks primarily due to ingestion.
10. U.S. Environmental Protection Agency. (1999). Methods
• The majority of health risks are due to synergy
Occurrence, and Monitoring Document for Radon, Draft. U.S.
with voluntary lifestyle choice—smoking. An ongoing Environmental Protection Agency, Office of Ground Water
public health antismoking campaign may result in and Drinking Water.
significant reductions in the population of ever- 11. Zikovsky, L. and Chah, B. (1990). The lognormal distribution
smokers, thus reducing overall population unit risks. of radon concentration in ground water. Ground Water 28(5):
Though the EPA did not specifically categorize 673–676.
ever-smokers as a ‘‘sensitive subpopulation,’’ their 12. Field, R.W., Fisher, E.L., Valentine, R.L., and Kross, B.C.
inclusion in risk analyses and benefit analyses (1995). Radium-bearing pipe scale deposits: implications for
significantly increased the benefits of adopting national waterborne radon sampling methods. Am. J. Public
the rule. Health 85: 567–570.
548 SILICA IN NATURAL WATERS
13. Krewski, D. et al. (2002). A combined analysis of North- 31. DW-32. (2004). Methods of Reducing Radon in Drinking
American case control studies of residential radon and lung Water. Encyclopedia of Water. John Wiley & Sons, New York.
cancer: An update. Radiat. Res. 158(6): 785–790. 32. U.S. Environmental Protection Agency. (2000). National Pri-
14. National Council on Radiation Protection and Measurements. mary Drinking Water Regulations: Groundwater Rule; Pro-
(2001). Evaluation of the Linear -Nonthreshold Dose- posed Rule 40 CFR Parts 141 and 142. Federal Register (FR)
Response. Model for Ionizing Radiation NCRP Report No. 65(91): 30194–30274). Available: http://www.access.gpo.gov/
136. NCRP, Bethesda, MD. Available: http://www.ncrp.com/ su docs/aces/aces140.html (accessed Jan 2001).
rpt136.html. 33. Fisher, E.L., Fuortes, L.J., and Field, R.W. (1996). Occupa-
15. Lorimore, S.A. et al. (1998). Chromosomal instability in the tional exposure of water-plant operators to high concentra-
descendants of unirradiated surviving cells after a-particle tions of radon-222 Gas. J. Occup. Environ. Med. 38: 759–764.
irradiation. Proc. Natl. Acad. Sci. USA 95: 5730–5733. 34. CDM. (1999). Radon Removal Preliminary Cost Analysis and
16. Brenner, D.J. and Sachs, R.K. (2003). Domestic radon risks Risk Assessment. Prepared for the East Valley Water District,
may be dominated by bystander effects—but the risks are San Bernardino, CA.
unlikely to be greater than we thought. Health Phys. 85(1): 35. Ayotte, P. et al. (1998). Indoor exposure to 222 Rn: A public
103. health perspective. Health Phys. 75(3): 297–302.
17. Parsons, P.A. (2002). Radiation hormesis: Challenging LNT 36. Yamaoka et al. (2004). Study on biologic effects of radon and
theory via ecological and evolutionary considerations. Health thermal therapy on osteoarthritis. Journal of Pain 2004 5:
Phys. 82(4): 513. 20–25.
18. Cohen, B.L. (1995). Test of the linear-no threshold theory of 37. Van Pelt, W.R. (2003). Epidemiological associations among
radiation carcinogenesis for inhaled radon decay products. lung cancer, radon exposure and elevation above sea level–a
Health Phys. 68(2): 157–174. reassessment of Cohen’s county level radon study. Health
19. Puskin, J.S. (2003). Smoking as a confounder in ecologic Physics 85(4): 397–403.
correlations of cancer mortality rates with average county Note: All referenced URLs were valid as of July 2003.
radon levels. Health Phys. 84(4): 526–532.
20. National Research Council. (1999). Health Effects of Exposure
to Radon (BEIR VI). National Academy Press, Washington, SILICA IN NATURAL WATERS
DC.
21. Field, R.W. and Duport, P.J. (2003). Exposure to residen- ROBERT Y. NING
tial radon causes lung cancer. W.R. Hendee, Moderator, King Lee Technologies
Point/Counterpoint. Med. Phys. 30(4): 485–488. San Diego, California
22. Centers for Disease Control. (1995). Morbidity Mortality
Wkly. Rep., Cigarette smoking among adults—United States
1993. INTRODUCTION
23. Alavanja, M.C. (2002). Biologic damage resulting from
exposure to tobacco smoke and from radon: Implication for Silica (1–4) and silicates (5–7) are generic names given
preventive interventions. Oncogene 21(48): 7365–7375. to the family of silicon dioxide compounds derived from
24. U.S. Environmental Protection Agency. (1994). Report to the dehydration–polymerization of silicic acid [Si(OH)4 ].
the United States Congress on Radon in Drinking Water, Silicic acid, variously called orthosilicic acid, monosilicic
Multimedia Risk and Cost Assessment of Radon. Office of acid, or silicic acid monomer, is a weak tetravalent acid
Drinking Water, Washington, DC. that has pKa values of 9.9, 11.8, and 12. Although silicic
25. Hei, T.K. et al. (1997). Mutagenic effects of a single and an acid is prevalent in natural waters and is readily available
exact number of α particles in mammalian cells. Proc. Natl. from acidification of commercial crystalline sodium silicate
Acad. Sci. USA 94: 3765–3770. or metasilicate (8), it has a great propensity to polymerize
26. Bilban, M. and Vaupotic, J. (2001). Chromosome aberrations by dehydration to form Si–O–Si anhydride bonds:
study of pupils in high radon level elementary school. Health
Phys. 80(2): 157–163. n Si(OH)4 −−−→ (OH)3 Si–O–Si(OH)3 dimer
27. U.S. Environmental Protection Agency. (1999). Health
Risk Reduction and Cost Analysis for Radon in Drinking −−−→ Oligomers −−−→ Colloidal polymers
Water Regulations: Notice and Request for Comments and
−−−→ (SiO2 )n
Announcement of Stakeholder Meeting. 40 CFR Parts 141 and
142. Federal Register (February 26, 1999) 64(38): 9559–9599.
[64 FR 9559], U.S. Govt. Printing Office, Washington, DC. Thus silicon dioxide (SiO2 , silica) is the final and most
28. Mills, W.A. (1990). Risk assessment and control management
stable product in normal conditions of silico-oxygen acid
of radon in drinking water. In: Radon, Radium and Uranium polymerization. The molecule of silica represented by
in Drinking Water. C.R. Cothern and P.A. Rebers (Eds.). the formula SiO2 is polymeric and is more accurately
Lewis, Chelsea, MI, pp. 27–37. represented by the formula (SiO2 )n , where n is a
29. Hahn, R.W. and Burnett, J.K. (2001). The EPA’s Radon Rule: large number.
A Case Study in How Not to Regulate Risks. Regulatory The rate of silicic acid polymerization is strongly pH-
Analysis 01-01. AEI-Brookings Joint Center for Regulatory dependent and is influenced by silicic acid concentration.
Studies. Available: http://www.aei.brookings.org It is very fast in neutral and slightly alkaline solutions,
30. Deb, A.K. (1992). Contribution of Waterborne Radon to Home and extremely slow at low pH values of 2 to 3. Dilute
Air Quality. American Water Works Association Research polysilicate solutions that are undersaturated with respect
Foundation, Denver, CO. to amorphous silica depolymerize to monosilicic acid.
SILICA IN NATURAL WATERS 549
Silicates imply more complex forms of silica in which Seawater 1–10 ppm
silicic acid co-polymerizes with hydroxides of other metal Wells 2–60 ppm
elements covering almost the entire periodic table. This Wells in volcanic and oil fields 50–300 ppm
gives rise to the immense variety of rocks found on earth,
which are copolymers of silicic acid in ultimate dehydrated In river and seawater, silica is not found exclusively in
forms. In solution, silicic acid readily reacts with alcohols the colloidal state, but also in molecular solution, because
and other hydroxyl-group containing organic compounds its saturation concentrations are not attained. Despite the
to form anhydrides. This provides for a great diversity of propensity for monosilicic acid to polymerize, it has been
products and interactions in the biosphere. shown that silica in seawater is nearly exclusively in the
monosilicic acid form in concentrations in the range of
CLASSIFICATION <0.006 to 10.1 ppm (11, 12). The concentration is low at
the surface and increases with depth. Samples of lake and
Silica in natural waters can generally be classified in three river water contain only monosilicic acid; there was no
categories: reactive soluble, nonreactive soluble (colloidal), evidence of the presence of dimeric silicic acid (13). Silica
and nonreactive insoluble (particulate). Reactivity refers in ground waters also appears to be in low molecular
to staining by molybdate reagent in the colorimetric assay weight forms rather than the colloidal form (14). The
of dissolved silica. The rate of this staining reaction concentration of particulate silica in seawaters of several
is quantitatively reproducible only with low molecular oceans has been determined as 0.0006 to 0.43 ppm SiO2
weight polysilicic acids and varies inversely with the (12). It is thought that it is mainly of biogenic origin.
degree of polymerization of silicic acid (9). Monosilicic Colloidal silica is depolymerized by seawater, which is
acid reacts with molybdic acid under assay conditions considered at least one order of magnitude undersaturated
within 2 minutes, and disilicic acid reacts completely in with respect to amorphous silica.
about 10 minutes. Higher polysilicic acids require much
longer times for substantial but incomplete reactions. DISSOLUTION OF SILICATE ROCKS
After removal of suspended insoluble silica by filtration
using typically a 0.45 micron filter, total dissolved silica Pure silica such as quartz is completely dehydrated silicic
(reactive + colloidal) can be measured by atomic emission acid. Depolymerization by rehydration as represented
or absorption spectroscopy. The colloidal silica value is the below represents dissolution to form soluble silicic acid:
difference between the total and reactive silica. Normal
detection limits of reactive silica in 25 mL samples are SiO2 (s) + 2H2 O −−−→ Si(OH)4
0.1 mg/L. A greater sensitivity of 0.001 mg/L is possible in
100 mL sample sizes (9). The dissolution rate depends on temperature and pH, it is
Agglomerated silica and silicates can reach a size, minimum at pH = 2 to 3 and rises rising about 100-fold at
in solution, where they are no longer ions, but are pH = 11. Heat-accelerated hydrolysis would be expected
actually colloidal particles that can remain in suspension in volcanic regions.
indefinitely and are very difficult to remove by filtration. Silicic acid is also generated from the weathering of
silicate rocks. Silicates are basically dehydrated forms of
GEOCHEMISTRY metal ion salts of silicic acid, which can be decomposed
by carbonic acid. The following two reactions represent
Silica (SiO2 ) in nature occurs in the crystalline forms of important dissolution processes of silicates (1):
quartz, cristobalite, and tridymite, and also in the form
of amorphous silica and glass. In number of atoms as 2NaAlSi3 O8 + 2CO2 + 6H2 O −−−→ Al2 Si4 O10 (OH)2 + 2Na+
well as weight, silicon is exceeded only by oxygen in the
earth’s crust. More than 95% of the volume of the earth’s + 2HCO3 − + 2Si(OH)4
crust is composed of quartz and a small number of other KMgFe2 AlSi3 O10 (OH)2 + 1/2O2
rock-forming silicates.
Second only to carbon, silicon forms the largest number + 3CO2 + 11H2 O −−−→ Al(OH)3 + 2Fe(OH)3 + K+
of compounds with other elements. The large number + Mg2+ + 3HCO3 − + 3Si(OH)4
of carbon compounds is due to the fact that the bond
energies of the C–C, C–O, and C–H bonds are of about Field measurements of the weathering rate on a
equal magnitude, and they will therefore be formed with per square meter catchment area basis from different
about the same probability. In contrast, the bond energy geographic areas are remarkably similar: 10−2 to 10−1
of the Si–O bond is considerably higher than that of the moles m−2 yr−1 . It is estimated that 105 m2 surface area
Si–H bond and more than twice that of the Si–Si bond. As of mineral grains active in weathering is available per
a consequence, instead of the common C–C–C chains of m2 of geographic area and that only a few monolayers
carbon chemistry, chains of the type Si–O–Si–O are the of rock silicates are dissolved per year (1). The silica
skeletons of silicon chemistry (5). concentrations on porewater increase continuously from
Typical concentrations of dissolved silica in natural the earth’s surface to a depth of 3 meters. The silica
waters are: concentration decreases slightly with depth either by
Lakes 1–3 ppm precipitation or—under the influence of groundwater—by
Major rivers 3–15 ppm dilution. The total annual chemical load delivered from the
550 SILICA IN NATURAL WATERS
continents to the oceans is estimated at 39.3 × 1014 g of is stored in lakes and aquifers or is carried to the sea by
which SiO2 amounts to 4.26 × 1014 g (15). rivers and precipitated there as calcium and magnesium
carbonate minerals. The overall reactions shown here were
first formulated by Urey (18) in 1952.
DEPOSITION OF SILICA AND SILICATES
CO2 + CaSiO3 ⇐⇒ CaCO3 + SiO2
Most natural waters are undersaturated in amorphous
silica and silicates, but unusual conditions found in CO2 + MgSiO3 ⇐⇒ MgCO3 + SiO2
geothermal waters (16) and industrial concentration or
uses of waters do lead to deposition of silica and The reactions from left to right represent carbon dioxide
silicates. Biochemical extraction and deposition of silica absorption via weathering. The reactions from right to
by living organisms (see below) constitute a major left represent degassing due to thermal decarbonation
deposition pathway. of carbonates after burial at sufficient depths in the
Up to a pH of approximately 9, the solubility of earth’s crust. Added to the Urey reactions should also
amorphous silica (SiO2 ) is relatively constant. It has been be the organic carbon subcycle. It exerts an important
reported by various investigators as 100 to 150 ppm at additional control on the atmospheric carbon dioxide
25 ◦ C; the dissolved species is Si(OH)4 . Amorphous silica level. This involves the removal of carbon dioxide from
the atmosphere through photosynthesis by organisms, the
is metastable relative to quartz. The rate of crystallization
burial of organic matter in sediments, and the weathering
of quartz is so slow at ordinary temperatures that the
of organic matter (or its thermal decomposition) which
solubility of amorphous silica represents the upper limit
returns the carbon dioxide. The overall reaction is
of dissolved aqueous silica.
represented as:
The precipitation of amorphous silica and silicates
apparently proceeds in a series of steps. Polymerization by
6CO2 + 6H2 O ⇐⇒ (CH2 O)6 + 6O2
dehydration accompanied by cross-linking reactions and
aggregation by van der Waal forces leads to negatively
Together, these reactions regulate carbon dioxide and
charged silica colloids. Further aggregation eventually
silica on a geologic timescale.
leads to soft then hard gels.
At suitably high pHs, if Fe(OH)3 , Al(OH)3 , Ca(OH)2 ,
and Mg(OH)2 are present, these hydroxides readily BIOCHEMICAL ASPECTS
participate in the copolymerization resulting in the
incorporation of Fe, Al, Ca, and Mg into complex anhydrous The importance of silicon for life on the earth results
silicate structures. The strong tendency of magnesium primarily from the fact that the fertility of soil relies in
hydroxide to react with silica, even in the monomeric large part on the ability of clay minerals to absorb and
state, is shown by the fact that the addition of 300 ppm release water and several cations that are indispensable
of Mg(OH)2 to water will reduce the soluble silica content for plant nutrition. This process is fundamental to the life
from 42 to 0.1 ppm. The order of precipitability starting of plants and based on them, to the life of those animals
with the most readily precipitated metal is copper, zinc, which in turn feed upon plants. Data on the physiology of
manganese, cadmium, lead, nickel, silver, magnesium, silicon is limited, except that in a number of species, its
and calcium. Metals that react in acidic solutions with presence is needed for normal growth.
silicates are Fe3+ , Co2+ , Zn2+ , and Cu2+ (4). Aluminum Dissolved silica compounds play a substantial role in
seems to occupy a unique position in combining with the cells of living organisms. Large amounts of silica are
silica. Aluminosilicate gels (zeolites) can be prepared found in horsetail, rice, feather grass, reed and bamboo,
that have ion-exchange properties. A gel composition of where it contributes to the strength of the leaves and
5SiO2 :Al2 O3 :Na2 O has an exchange capacity of 6% of its stems, and in the skeletons of diatoms which consist of
weight of CaO, and almost all of the sodium ions are very pure SiO2 . Silicon compounds are vital components
exchangeable. Colloidal silica and silicates are active as in the metabolism of many bacteria, especially those
flocculants and precipitate with other natural colloidal that live in hot springs. The bacterium Proteus mirabilis
matter in rivers, lakes and the sea. even substitutes silicon for phosphorous in phospholipids.
Silicon is also present as a trace element in higher animals
and in humans, where it constitutes about 10 mg per kg
GEOCHEMICAL SILICA CYCLE live weight. It is essential in the cells of connective
tissue and is involved in the biosynthesis of collagen, the
Geologically, the silica cycle is intimately associated substance that forms hair and nails, and in the formation
with the carbonate cycle (see Carbonate in Natural of bony tissue.
Waters). The atmospheric CO2 concentration of about The enrichment of silicon in mineralized tissues of uni-
0.04% on a timescale of millions of years is controlled cellular organisms, such as diatoms and radiolaria, and a
mainly by a combination of degassing from the earth’s few metaozoa, such as sponges and gastropods is spectac-
interior balanced by the uptake of carbon dioxide ular (19). The mineral phase is always amorphous silica
via the weathering of calcium and magnesium silicate in various stages of hydration as SiO2 ·nH2 O. Individual
minerals (17,18). The carbon dioxide in water as carbonic diatom populations require a certain level of monosilicic
acid is converted to dissolved calcium and magnesium acid, and declining ambient silica concentrations in natu-
bicarbonate on land during weathering. This carbon form ral environments may influence the sequence of seasonal
SODIUM IN NATURAL WATERS 551
successions. The present day ocean contains a few ppm sodium; nearly all sodium compounds readily dissolve in
(parts per million) of dissolved silica; diatoms are often water, and it naturally leaches from rocks and soils.
cited as the primary agent of silica removal from the The sodium ion, the sixth most abundant metallic ion in
world’s ocean and keeping a low silicic acid level. Others the earth’s crust, is a natural constituent of both food and
favor inorganic extraction as the main regulating mech- water from several sources. Most water supplies contain
anism because common silicates take up dissolved silica less than 20 mg of sodium per liter, but in some countries,
from silica-rich seawater. The data suggest that the world levels can exceed 250 mg/liter. Saline intrusion, mineral
ocean was low in silicon for most of its history. deposits, seawater spray, sewage effluents, industrial
brines, chemicals used in paper processing, soap, glass,
and glass-wool industries, food processing, slaughtering
BIBLIOGRAPHY and meat packing industries and salt used in road
deicing all contribute significant quantities of sodium
1. Stumm, W. and Morgan, J.J. (1996). Aquatic Chemistry. 3rd to water. In addition, water treatment chemicals, such
Edn. John Wiley & Sons, New York. as sodium fluoride, sodium bicarbonate, and sodium
2. Iler, R.K. (1979). The Chemistry of Silica. John Wiley & Sons, hypochlorite can together result in sodium levels as high
New York. as 30 mg/liter. Domestic water softeners can give levels
3. Sosman, R.B. (1965). The Phases of Silica. University of over 300 mg/liter. A much more important source of
Microfilms International, Ann Arbor, MI (originally from sodium in fluvial waters is the input from leaching of
Rutgers University Press, New Brunswick, NJ). halite from sedimentary or evaporite deposits. More than
4. Iler, R.K. (1955). The Colloidal Chemistry of Silica and 40% of the sodium in fluvial waters comes from this
Silicates. Cornell University Press, Ithaca, NY. source. Sodium is also derived from the weathering of
5. Liebau, F. (1985). Structural Chemistry of Silicates. Springer- silicate minerals, principally sodic plagioclase feldspar.
Verlag, New York. This source accounts for more than 20% of the sodium
6. Babushkin, V.I., Matveyev, G.M., and Mchedlov-Petrossyan, in river water. It is also an important element in mica,
O.P. (1985). Thermodynamics of Silicates. Springer-Verlag, amphibole, and pyroxene.
New York. A Na+ restricted diet is recommended to patients
7. Eitel, W. (1954). The Physical Chemistry of the Silicates. suffering from hypertension or congenital heart disease;
University of Chicago Press, Chicago, IL. in such cases, the intake of high Na+ through drinking
8. Alexander, G.B. (1953). J. Amer. Chem. Soc. 75: 2887. water may prove critical (1). The current drinking water
9. Alexander, G.B. (1953). J. Amer. Chem. Soc. 75: 5655. standard for sodium is 20 milligrams per liter.
10. Kemmer, F.N. (Ed.). (1988). The Nalco Water Handbook. 2nd
Edn. McGraw-Hill, New York.
SODIUM IN NATURAL WATER
11. Isshiki, K., Sohrin, Y., and Nakayama, E. (1991). Marine
Chemistry 32: 1.
Groundwaters
12. Kido, K. and Nishimura. (1975). Deep-Sea Research 22: 323.
13. Okada, T. and Kuwamoto, T. (1985). Anal. Chem. 57: 258. The Na content of groundwaters is a function of
14. George, S., Steinberg, S.M., and Hodge, V. (2000). Chemo- (1) weathering of Na plagioclase from bedrock, followed
sphere 40: 57. by (2) exchange of Ca2+ for Na+ on the surfaces of newly
15. Matthess, G., Frimmel, F., Hirsch, P., Schulz, H.D., and formed clay minerals. Feth et al. (2) considered the weath-
Usdowski, H-E. (Eds.). (1992). Progress in Hydrogeochem- ering reactions which supplies Na+ to groundwater as:
istry. Springer-Verlag, New York.
16. Henley, R.W. (1983). Geothermics 12: 307. 3Na0.66 Ca0.33 Al1.33 Si2.66 O8 + 14H2 O + 4CO2
17. Berner, R.A. (1995). Am. J. Science 295: 491.
→ 2H4 Al2 Si2 O9 + 2Na+ + Ca2+ + 4H4 SiO4 + 4HCO−
3
18. Urey, H.C. (1952). The Planets: Their Origin and Develop-
ment. Yale Univ. Press, New Haven, CT, p. 242. (kaolinite)
19. Degens, E.T. (1976). Topics in Current Chemistry. Vol. 64,
Inorganic Biochemistry. Springer-Verlag, New York. and
6Na0.66 Ca0.33 Al1.33 Si2.66 O8 + 6CO2 + 18H2 O

SODIUM IN NATURAL WATERS → 3Na0.66 Al2.66 Si3.33 O10 (OH)2 + 2Ca2+ + 2Na+
NITISH PRIYADARSHI + 6H4 SiO4 + 6HCO3 − (montmorillonite)
Ranchi University
Ranchi, Jharkhand, India These theoretical calculations apply well to the actual
minerals in the rocks and soils of the area studied
(western U.S.)
INTRODUCTION Garrels (3) suggested that the Na+ content of ground-
water would rise due to weathering of plagioclase
Sodium is a light alkali metal which floats on water in feldspars. If the system is closed, the waters eventually
its pure state. However, it is very chemically active and is proceed through kaolinite stability. If calcite is precipitat-
rarely found in its pure state. All natural waters contain ing, this montmorillonite will be sodic and hold the Na+
552 SODIUM IN NATURAL WATERS
content of the water relatively steady. If calcite is not pre- and potassium. Sodium was calculated by difference by
cipitating, Ca2+ will be used by the montmorillonite, and subtracting the sums of the chemical equivalents of the
the Na+ content of the groundwater will continue to rise. negative and positive ions. This procedure was followed
Low concentrations of Na+ ions are associated with the because Dittmar was unable to achieve satisfactory direct
minerals, kaolinite and quartz (4). determinations for sodium. He estimated 30.61% sodium
Sathyanarayanan and Periakali (5), in their study in sea water.
of Na concentration in groundwater in Ultrabasic and In the valuable treatise, The Data of Geochemistry,
Peninsular Gneissic Rocks of Tamil Nadu State of India, (1924), F. W. Clarke calculated that the 300,000,000 cubic
suggested that weathering of plagioclase feldspar and miles of oceanic waters owed their salinity to 4,800,000
biotite in the Peninsular gneiss could be the source of cubic miles of dissolved salts which had been derived from
Na in groundwater. the igneous rocks of the earth’s crust. He confined his
In short, the Na+ content of groundwater is a func- computations to the sodium (Na) percentages, reckoning
tion of weatherable minerals, pH, reaction time versus an average of 1.14% in seawater and 2.90% of this element
drainage time, the supply of Ca2+ to form Ca montmoril- in igneous rocks. The average concentration of sodium
lonite, sluggish drainage conditions, intensive irrigation in seawater is 1.1 × 107 g/L (14). Earlier, Goldberg (15)
practices, extensive use of chemical fertilizers, inadequate estimated 10,500 ppm Na+ in seawater. Calculations of
sanitary facilities, and precipitation dominance control- Garrels and Thompson (16) indicated that 99% of the Na
ling water chemistry (6–9). In addition, groundwater can in seawater exists as uncomplexed Na+ .
contain large amounts of Na+ from the dissolution of evap-
orites, from saltwater intrusion, and smaller amounts from
cyclic sea salts in original precipitation. TREATMENT
Rivers Sodium may be removed from a water supply by ion

exchange. The ion exchange resin in the water softener is
The concentration of elements in river water varies with enriched with calcium and removes the sodium from the
time at any single sampling site as a function of discharge, water. Other demineralization techniques, such as reverse
tributary supply, and groundwater discharge. In areas osmosis, distillation, electrodialysis and freeze–thaw
where rivers are draining evaporite deposits, the Na processes, should be effective for removing sodium from
content can be very high. For example, streams in the water. If a large supply of water is required, it may be
Brazos River basin (Texas) have as much as 99,800 ppm necessary, if feasible, to blend it with another water source
Na and an average of 35,900 ppm (10). low in sodium.
The unweighted average Na content of 639 rivers and
√ In addition to water treatment, another method of
streams is 38.6 ± 5.9 ppm (s/ n). The range is 0.1 to removing sodium from drinking water would require
1.540 ppm, and the median is 6.1 ppm Na (11,12). The the construction or reconstruction of a water well. This
mean content may be used as a worldwide estimate, frequently involves installing additional casing beyond
but the population is actually polymodal with a mean the depth normally required by regulations.
for humid and temperate weathering zones, a mean for
arid climates, and an extremely high mean for rivers
draining evaporite deposits (13). Comparison shows that BIBLIOGRAPHY
river water and seawater are opposite in character for
Na concentration. In seawater, Na>Mg>Ca; in average 1. Holden, W.S. (1970). Water Treatment and Examination.
river water, Ca>Na>Mg (14). Livingstone’s (11) weighted J & Churchill, London.
mean Na+ content of world rivers was 6.3 ppm. 2. Feth, J.H., Roberson, C.E., and Polzer, W.L. (1964). Sources
of Mineral Constituents in Water from Granite Rocks of the
Lake Waters Sierra Nevada. U.S. Geological Survey Water Supply Papers,
1535-I, pp. 1–70.
The sodium content of lake water is a function of 3. Garrels, R.M. (1967). Genesis of some groundwaters from
river and groundwater supply followed by increase due igneous rocks. In: Researches in Geochemistry, 2nd Edn.
to evaporation where this process exceeds input. The P.H. Abelson (Ed.). John Wiley & Sons, New York, pp.
evaporation may reach a stage of Na mineral precipitation, 405–420.
which occurs at a high level of Na+ . 4. Garrels, R.M. (1976). A survey of low temperature water-
mineral relations. In: Interpretation of Environmental
Seawater Isotope and Hydrochemical Data in Groundwater Hydrology.
International Atomic Energy Agency, Vienna, Austria.
The sodium released into solution during weathering is 5. Sathyanarayanan, M. and Periakali, P. (2003). Geochemistry
transported by rivers to seas/oceans and represents a of groundwater in Ultrabasic and Peninsular gneissic rocks,
primary cyclic reservoir. Early knowledge of the average Salem District, Tamil Nadu. J. Geol. Soc. India 62(1): 63–73.
composition of seawater was provided by Dittmar in 1884; 6. Gibbs, R.J. (1970). Mechanisms controlling world water
he carefully analyzed 77 water samples, representative of chemistry. Science. 170: 1088–1090.
all oceans and taken both from the surface and from the 7. Prakasa Rao, J. (1997). Environmental Evaluation of Ground-
depths. These samples had been collected on the voyage water Quality of a Developing Urban Area of Andhra
around the world of H.M.S. Challenger (1872–1876). Pradesh, India. Unpublished. Ph.D. Thesis, Andhra Univer-
He determined halides, sulfate, magnesium, calcium, sity, Visakhapatnam, p. 104.
SOFT WATER 553
8. Subba Rao, N. (1998). Groundwater quality in crystalline 4. Temporary hardness: Equals alkalinity, if alkalinity
terrain of Guntur District, Andhra Pradesh. Visakha Sci. J. is less than total hardness
2: 51–54.
5. Permanent hardness: The amount of hardness
9. Ballukrya, P.N. and Ravi, R. (1999) Characterization of
greater than temporary hardness
groundwater in the unconfined aquifers of Chennai city,
India: Part I- Hydrogeochemistry. J. Geol. Soc. India 54: 6. Noncarbonate hardness: Permanent hardness
1–11.
10. Baker, R.C., Hughes, L.S., and Yeast, I.D. (1964). Natural Of the six different forms of hardness, the most important
Sources of Salinity in the Brazos River, Texas. U.S. Geological and the one that is commonly known as hardness,
Survey Water Supply Papers. 1669-cc, pp. 1–81. is total hardness. It is calculated by the following
11. Livingstone, D.A. (1963) Chemical Composition of Rivers simplified equation:
and Lakes. U.S. Geological Survey Professor. Papers 440-G,
pp. 1–64.
Hardness = 2.5 Ca + 4.1 Mg (units in mg/l)
12. Anderson, P.W. and George, J.R. (1966). Water-Quality Char-
acteristics of New Jersey Streams. U.S. Geological Survey
Water Supply Papers 1819-G, pp. 1–48.
Note that there are equations to calculate each of the six
different forms of hardness.
13. Wedepohl, K.H. (1970). Handbook of Geochemistry. Vol. II/2,
Springer-Verlag, Berlin.
Example: The Ca and Mg concentrations in a water sample
are 20 and 15 mg/l, respectively. What is its hardness?
14. Mason, B. and Moore, C.B. (1985). Principles of Geochemistry,
4th Edn. Wiley Eastern Limited, New Delhi, p. 242.
15. Goldberg, E.D. (1957). Biogeochemistry of trace metals. In: Hardness = 2.5 × 20 + 4.1 × 15 = 111.5 mg/l
Ecology, vol.1. Treatise on Marine Ecology and Paleocology.
J.W. Hedgpeth (Ed.). Geological Society America Member 67, As stated, each of the six forms of hardness can be
pp. 345–358. calculated by analyzing a water sample for the relevant
16. Garrels, R.M. and Thompson, M.E. (1962). A chemical model ion(s), followed by using the relevant equation. However,
for sea water at 25 ◦ C and one atmosphere total pressure. Am. there are techniques to measure total hardness directly
J. Sci. 260: 57–66. such as (1) a standard solution of soap, that was in use for
many years (2); and (2) Schwarzenbach et al’s. method of
colorimetric titration (5), which is sometime referred to as
SOFT WATER the EDTA method.
The author believes that earlier, measuring hardness
GHOLAM A. KAZEMI
was a common and an inseparable practice in water
Shahrood University of
analysis. To agree with this statement, one needs
Technology
Shahrood, Iran to look at a relatively old handbook and read, ‘‘The
common denominator of the majority of water problems is
hardness’’ (2, p. 4.1, first published in 1979). Somehow, in
INTRODUCTION contrast, recent hydrochemical analysis of water samples
does not usually include a measurement of hardness.
For instance, there is not a single report of hardness
At the outset, it should be pointed out that soft and
in the hundreds of hydrochemistry papers presented at
hard waters are two sides of a coin. Hence, the respective
the International Association of Hydrogeolgists (IAH)
articles, this one and the Hard Water article are alike in
conferences in Prague (2003) and in Munich (2001) and
many aspects. In defining soft water, it is necessary first to
in the recent past (1997–2001) annual symposia of the
describe hardness, a closely related chemical property of
Geological Society of Iran. This occurred, perhaps, because
water. Hardness is defined by some as a property of water
other aspects of water chemistry and water biology,
that prevents lather formation with soap and produces
such as nutrients, heavy metals, bacteriological species,
scale in hot water systems, (1); by others, as the soap
chlorinated hydrocarbons, and many contaminants have
consuming capacity of water (2); and in broader terms
become so widespread that they are considered more
by Freeze and Cherry (3), as the ‘‘metallic ion content of
important and more hazardous than hardness. This is to
water which reacts with sodium soaps to produce solid
say that rejecting water for drinking or other uses, these
soaps or scummy residue and which reacts with negative
days, is many times more likely to be due to its harmful
ions, when the water is evaporated in boilers, to produce
DDT level than because of its high hardness value. One
solid boiler scale.’’
other possible reason is that hardness and soft water topics
The ions which result in, or produce, hardness include
are important only in some narrow fields of water, e.g., hot
Mg+2 , Ca+2 , Sr+2 , Fe+2 , and Mn+2 , and to a lesser extent,
water boilers.
Ba+2 and Zn+2 and other divalent ions. The first two of
The second point is that industrialization, urbaniza-
these, are the most influential. There are different forms
tion, agricultural practices, and most environmentally
of hardness (4):
unfriendly activities are not likely to result in an increase
in the hardness of water resources because these activities
1. Total hardness: Ca and Mg expressed as CaCO3
do not usually generate hardness-producing ions such as
2. Calcium hardness: Ca expressed as CaCO3 Ca2+ or Mg2+ , though, they may produce other trivial (in
3. Magnesium hardness: Mg expressed as CaCO3 terms of hardness) ions such as Mn2+ and Fe2+ .
554 SOFT WATER
The author, here, wishes to draw the reader’s attention Manning (9), discredits soft water and states that very
to one other point and pose a question as well. He believes soft water when used for bathing feels slick on the body,
that ‘‘softness’’ would have had an identical application as though the soap is still there. He also explains that
and a similar meaning to hardness if it would have been most people prefer slightly hard water because of its taste
there. Therefore, it is interesting to speculate as to what and its ability to wash the soap off the body. Soft waters
was the rationale for use of the term ‘‘hardness’’, not are not very good for irrigation either because they may
softness. One reason might be that hard water is generally lead to a lower infiltration rate compared with slightly
easier to notice because of its unwanted and disappointing hard water (9). In contrast to hard water, soft water
characteristics and our predecessors looked for something produces a foam or lather easily but does not produce
to describe it. In other words, the term hardness was scale in hot water pipes, heaters, boilers, and other high
created to explain easily noticeable hard water, not to water temperature units. Therefore, using soft water is
describe soft water. This is why there is no such practice an economically worthwhile practice because it leads to
as water hardening, changing soft water to hard water, in a decrease in soap consumption and lowers fuel costs for
contrast to the well-established exercise of water softening boilers due to good heat conduction.
which changes hard waters to soft waters. The recommended water for domestic use should have
a hardness of 80 mg/l; i.e., it should be soft (10). This is
DEFINITION AND SPECIFICATIONS OF SOFT WATER why water softening is a common practice where the water
supply has a hardness greater than 80–100 mg/l (3). Soft
Todd (6, p. 282) reports that Hippocrates (460–354 B.C.) water is also required for some industrial processes such
was the first to use the terms hard and soft for water, when as food equipment washing, confectionery, food canning
in his treatises on public hygiene, he wrote, ‘‘consider the and freezing, food processing, and laundering (6).
waters which the inhabitants use, whether they be marshy
and soft, or hard and running from elevated and rocky
SOFT WATER AND HEALTH
situations, and then if saltish and unfit for cooking. . . for
water contributes much to health.’’ However, it is probably
There is a general perception among the public that
safe to say that hard and soft water in Hippocrates’
soft water is healthier to drink. Studies, though not
writings, are likely to mean different things from what
we today call hard and soft water. satisfactorily conclusive, suggest the opposite, especially
The most frequently used classification system for with regard to artificially softened water. For instance,
water based on hardness is that of Sawyer and Tebbutt (1995) (1) reports that there is some statistical
McCarty (7), presented in Table 1. There is another evidence to suggest that artificially softened waters may
slightly different classification system by Hem (8), which increase the incidence of some forms of heart disease.
is referred to in some publications as the classification of Also, a recent study in Taiwan (11) shows that there is a
the U.S. Geological Survey (9). significant negative relationship between drinking water
The definition of soft water varies slightly among hardness and colon cancer mortality, a finding considered
researchers. Based on Table 1, soft water is defined as important for the Taiwanese water industry and human
water whose hardness is less than 60 or 75 mg CaCO3 /l. health. All these confirm Muss’s (12) observation that in
This means that waters with less than 15 mg Ca2+ /l and the United States, death rates from heart and circulatory
10 mg Mg2+ /l are soft; higher concentrations in waters diseases are lower in states where public water supplies
belong to the other categories such as hard water or very are harder. Furthermore, artificially softened water may
hard water. However, some researchers argue that there taste salty because its sodium content increases in the
is no exact definition for determining soft or hard water softening process.
because it depends, at least partially, on what one is
accustomed to (9). What is soft water for one person, may AVAILABILITY OF NATURALLY SOFT WATER
be moderately hard for another and vice versa. The very
hard water category in Table 1 can be grouped into the Soft water may exist naturally or may be produced
hard water category as it has the specifications of hard artificially, as stated before. Atmospheric precipitations
water. It is, however, unclear to which category one can (snow and rain) are soft waters. Similar to these are
assign moderately hard water, if hard water and soft water waters that emanate from upland catchments with an
are the only categories available. average total hardness of 10 mg/l (1). In contrast, most
groundwaters, excluding those that occur in fractured
Table 1. Different Types of Water igneous rocks, are hard. All ground water and surface
Based on Hardness water that pass through, or are in contact with, carbonate
and other Ca–Mg bearing rocks are hard. Lowland rivers
Water Type Hardness mg CaCO3 /l
with a hardness of 200 mg/l (1) are also hard. Therefore,
Soft below 75a Below 60b only a limited percentage of freshwater resources of the
Moderately hard 75–150a 60–120b world, groundwaters and river waters, is soft.
Hard 150–300a 120–180b Examples of naturally occurring soft waters include the
Very hard Over 300a Over 180b Vilnius well fields groundwater (intermorainic and allu-
a
Reference 7. vial deposit aquifer), Vilnius District, Lithuania, that have
b
Reference 8. unbelievably low hardness values of 3.9–8.5 mg/l (13),
SOLUBILITY OF CHEMICALS IN WATER 555
groundwater in Akwa Ibom State of Nigeria (fluvio- ketones, ethers, aldehydes, acids, acetates, amines,
volcanic rocks) with a hardness of 1.2–65 mg/l (Akujieze, monochloroalkanes, and monobromoalkanes. The new
2003) (14), and shallow basement groundwater of Lithem- correlation provides reliable solubility values down to
bwe dambo, Malawi, with 1–14 mg/l of calcium and very low concentrations (parts per million range). The
1–7 mg/l of magnesium (15). correlation is based on the boiling point of the compound.
The correlation and experimental data are in favorable
BIBLIOGRAPHY agreement. The results are useful in process engineering
for wastewater minimization.
1. Tebbutt, T.H.Y. (1995). Principles of Water Quality Control.
Reprint of 4th Edn. Butterworth Heinemann, Oxford, UK.
IMPORTANCE OF WATER SOLUBILITY
2. Kemmer, F.N. (1988). The NALCO Water Handbook, 2nd
Edn. McGraw-Hill, New York.
The solubility of compounds in water is very important.
3. Freeze, R.A. and Cherry, J.A. (1979). Groundwater. Prentice-
This importance will increase in the future because of
Hall, Englewood Cliffs, NJ.
heath, safety, and environmental considerations. The
4. Hounslow, A.W. (1995). Water Quality Data: Analysis and
following brief discussion illustrates the importance even
Interpretation. Lewis, New York.
at very low concentrations.
5. McCoy, J.W. (1981). The Chemical Treatment of Boiler. Water
For health involving human exposure to substances
Chemical Publishing, New York.
in air, the threshold limit value (TLV) for sec-butanol in
6. Todd, D.K. (1980). Groundwater Hydrology, 2nd Edn. John
air is given as 100 ppm (parts per million, vol) by the
Wiley & Sons, New York.
Occupational Safety and Health Act (1). A concentration
7. Sawyer, C.N. and McCarthy, P.L. (1967). Chemistry for
Sanitary Engineers, 2nd Edn. McGraw-Hill, New York.
of only 0.0005 mol fraction of sec-butanol in water provides
about 236 ppm of sec-butanol in air at the air–water
8. Hem, J.D. (1985). Study and Interpretation of the Chemical
Characteristics of Natural Waters, 3rd Edn. U.S. Geological
interface. This concentration of 236 ppm in air exceeds
Survey Water Supply Paper 2254, Denver, CO. the threshold limit value of 100 ppm.
9. Manning, J.C. (1997). Applied Principles of Hydrology.
For safety, the lower explosion limit (LEL) for sec-
Prentice Hall, Upper Saddle River, NJ. butanol in air is given as 1.7% by Yaws (2). A concentration
10. Bell, F.G. (1998). Environmental Geology, Principles and
of only 0.04 mol fraction of sec-butanol in water provides
Practice. Blackwell Science, Cambridge, UK. about 1.9% of sec-butanol in air at the air–water interface.
11. Yang, C.Y. and Hung, C.F. (1998). Colon cancer mortality
This concentration of 1.9% in air exceeds the lower
and total hardness levels in Taiwan’s drinking water. Arch. explosion limit of 1.7%.
Environ. Contamination Toxicol. 35(1): 148–151. For environmental, consider a spill of sec-butanol in
12. Muss, D.L. (1962). Relationship between water quality and contact with water. The water will become saturated with
deaths from cardiovascular disease. J. Am. Water Works sec-butanol. At saturation, the solubility of sec-butanol
Assoc. 54: 1371–1378. in water is about 0.181 weight fraction or 0.0508 mol
13. Klimas, A.A. (1995). Impacts of urbanization and protection of fraction (3). This concentration of only 0.0508 mol fraction
water resources in the Vilnius District, Lithuania. Hydrogeol. at saturation will provide about 24,000 ppm or 2.5%
J. 3(1): 24–35. of sec-butanol in air at the air–water interface. This
14. Akujieze, C.N., Coker, S.J.L., and Oteze, G.E. (2003). Ground- concentration of 24,000 ppm or 2.4% greatly exceeds both
water in Nigeria—a millennium experience—distribution, the threshold limit value of 100 ppm and the lower
practice, problems and solutions. Hydrogeol. J. 11(2): explosion limit of 1.7%.
259–274. This brief discussion (using sec-butanol as an example)
15. Chilton, P.J. and Foster, S.D. (1995). Hydrogeological char- indicates that very low concentrations of a chemical in
acterization and water-supply potential of basement aquifers water can provide concentrations in air at the air–water
in tropical Africa. Hydrogeol. J. 3(1): 36–49. interface that exceed the threshold limit value for human
exposure and the lower explosion limit for flammability.
SOLUBILITY OF CHEMICALS IN WATER

CORRELATION FOR WATER SOLUBILITY
CARL L. YAWS
PRASAD K. NARASIMHAN In earlier work by Yaws and co-workers (2), the water
HELEN H. LOU solubility of hydrocarbons was correlated as a function of
Lamar University the boiling point of the compound. In this article, it was
Beaumont, Texas determined that the boiling point method is also applicable
to the correlation of the water solubility of chemicals:
RALPH W. PIKE
Louisiana State University log10 (S) = A + B TB + C TB 2 + D TB 3 (1)
Baton Rouge, Louisiana
where S = solubility in water at 25 ◦ C, ppm
by weight,
Correlations are presented for the solubility of chemicals in TB = boiling point of compound, K
water. A wide variety of compounds are covered—alcohols, A, B, C, and D = regression coefficients
556 SOLUBILITY OF CHEMICALS IN WATER
Table 1. Solubility in Watera
log10 S = A + B TB + C TB 2 + DTB 3 (S, ppm(wt), TB , K)
Chemical Family A B C D TBMIN TBMAX Formula Example Compound
Alcohols 45.6398 −2.3859E−01 4.8739E−04 −3.7160E−07 370 620 Cn H2n+2 O n-Pentanol

Ketones 45.2000 −2.3859E−01 4.8739E−04 −3.7160E−07 350 590 Cn H2n O Diethyl ketone
Ethers 11.4000 −2.2000E−02 0.0000E+00 0.0000E+00 300 560 Cn H2n+2 O Dipropyl ether
Aldehydes 12.4200 −2.2000E−02 0.0000E+00 0.0000E+00 320 560 Cn H2n O 1-Hexanal
Acids 14.3000 −2.2000E−02 0.0000E+00 0.0000E+00 440 560 Cn H2n O2 n-Hexanoic acid
Acetates 13.0000 −2.3382E−02 0.0000E+00 0.0000E+00 330 600 Cn H2n O2 Ethyl acetate
Primary amines 11.4300 −1.8000E−02 0.0000E+00 0.0000E+00 400 560 Cn H2n+3 N n-Hexyl amine
Secondary amines 11.3600 −1.8000E−02 0.0000E+00 0.0000E+00 360 560 Cn H2n+3 N Dipropyl amine
Tertiary amines 11.8200 −2.0000E−02 0.0000E+00 0.0000E+00 360 560 Cn H2n+3 N Triethyl amine
Monochloroalkanes −7.4500 1.0050E−01 −2.7288E−04 1.9987E−07 300 590 Cn H2n+1 Cl 2-Chlorobutane
Monobromoalkanes −7.1700 1.0050E−01 −2.7288E−04 1.9987E−07 300 590 Cn H2n+1 Br 1-Bromobutane
a
S-solubility in water at 25 C, parts per million by weight, ppm(wt).
A, B, C, and D-regression coefficients.
TB -boiling point temperature of compound, K.
TBMAX -maximum temperature, K.
TBMIN -minimum temperature, K.
The range of boiling point depends on the chemical family. Data-alcohols Equation
The coefficients (A, B, C, and D) are given in Table 1 1000000
for a wide variety of chemicals—alcohols, ketones, ethers,
100000
aldehydes, acids, acetates, amines, monochloroalkanes,
Solubility in water, ppm (wt)
10000
and monobromoalkanes. The table also provides the
boiling point range (TBMIN , TBMAX ) for each chemical 1000
family. The correlation is applicable to compounds whose 100
boiling points are in this range. The correlation should not 10
be used for compounds whose boiling points are outside 1
the boiling point range.
0.1
The coefficients were determined from regression of
0.01
available data. In preparing the correlation, a literature
search was conducted to identify data source publica- 0.001
tions (2–13). The excellent compilations by Horvath (5–6); 0.0001
Howard and Meylan (7); Mackay, Shiu, and Ma (9); Ver- 300 350 400 450 500 550 600 650
schueren (11); Yalkowsky (3); and Yaws (2) were used Boiling point, K
extensively. The publications were screened and copies Figure 1. Solubility of alcohols in water.
of appropriate data were made. These data were then
keyed in to the computer to provide a database for which
experimental data are available. The database also served Data-alcohols Data-ketones
as a basis to check the accuracy of the correlation. Equation Equation
Graphs of water solubility versus boiling point are 1000000
shown in Figs. 1–11 for the chemical families. The graphs
100000
indicate favorable agreement of correlation values and
experimental data. 10000

For ketones, the graphs disclose that more data are 1000
available for alcohols. The graphs, also disclose that the
100
data for ketones appear to be below and approximately
parallel to the more extensive data for alcohols. This 10
suggests that the solubility of ketones is lower and 1
approximately parallel to the water solubility of alcohols. 0.1
For the other oxygenated compounds (ethers, aldehy-
des, acids, and acetates), the graphs disclose that the data 0.01
for oxygenated compounds appear to be above and approx- 0.001
imately parallel to the data for benzenes. This suggests 0.0001
that the solubility of oxygenated compounds is higher and 300 350 400 450 500 550 600
approximately parallel to the water solubility of benzenes. Boiling point, K
For the nitrogen compounds (primary amines, sec-
Figure 2. Solubility of ketones in water.
ondary amines, and tertiary amines), the graphs disclose
SOLUBILITY OF CHEMICALS IN WATER 557
Equation Data-ethers Equation Equation

Equation Data-benzenes Data-acids Data-benzenes
100000 1000000
10000 100000

1000 10000
1000
100
100
10
10
1
1
0.1 0.1
0.01 0.01
0.001 0.001
300 350 400 450 500 550 600 300 350 400 450 500 550 600 650
Boiling point, K Boiling point, K
Figure 3. Solubility of ethers in water. Figure 5. Solubility of acids in water.
Data-acetates Data-benzenes
Equation Data-aldehtdes
Equation Equation
Equation Data-benzenes
1000000 1000000
100000 100000
10000
10000
1000
1000
100
10 100
1 10
0.1
1
0.01
0.1
0.001 300 350 400 450 500 550 600
300 350 400 450 500 550 600 Boiling point, K
Boiling point, K
Figure 6. Solubility of acetates in water.
Figure 4. Solubility of aldehydes in water.
Equation Equation
that the data for amines appear to be above and approx- Data-amines-primary Data-benzenes
imately parallel to the data for benzenes. This suggests 100000
that the solubility of amines is higher and approximately
10000
parallel to the water solubility of benzenes.

For monochloro and monobromoalkanes, the graphs 1000
disclose that the data appear to be lower and approxi-
100
mately parallel to the data for benzenes. This suggests
that the solubility of monochloro- and monobromoalkanes 10
is lower and approximately parallel to the water solubility 1
of benzenes.
0.1
0.01
EXAMPLES
0.001
300 350 400 450 500 550 600
The correlations for water solubility are useful in Boiling point, K
engineering applications involving process wastewater.
Examples are shown below. Figure 7. Solubility of primary amines in water.
558 SOLUBILITY OF CHEMICALS IN WATER
Equation Equation Monobromo alkanes Benzenes

Data-amines-secondary Data-benzenes Equation Equation
100000 100000
10000 10000

1000 1000
100 100
10
10
1
1
0.1
0.1
0.01
0.01
0.001
0.001 300 350 400 450 500 550 600
300 350 400 450 500 550 600 Boiling point, K
Boiling point, K
Figure 11. Solubility of monobromoalkanes in water.
Figure 8. Solubility of secondary amines in water.
Example. A chemical spill of dodecanol occurs into
a body of water at ambient conditions. Estimate the
Equation Equation concentration in the water at saturation.
Data-amines-tertiary Data-benzenes The correlation for alcohols maybe used to determine
100000 the solubility in water. Substitution of the coefficients and
boiling point of dodecanol in the correlation equation yields
10000
1000 log10 (S) = A + B TB + C TB 2 + D TB 3
100
= 45.6398 − 2.3859E−01(535.00)
+ 4.8739E−04(535.00)2 − 3.7160E−07(535.00)3
10
S = 3.93 ppm (wt).
1
The calculated value and data compare favorably
0.1
(3.93 vs. 4.00; deviation = 0.07/5 = 1.8%).
0.01
Example. A chemical spill of diethyl ketone occurs into
0.001 a body of water at ambient conditions. Estimate the
300 350 400 450 500 550 600 concentration in the water at saturation.
Boiling point, K The correlation for ketones maybe used to determine
the solubility in water. Substitution of the coefficients
Figure 9. Solubility of tertiary amines in water.
and boiling point of diethyl ketone in the correlation
equation yields
Monochloro alkanes Benzenes log10 (S) = A + B TB + C TB 2 + D TB 3
Equation Equation
100000 = 45.2000 − 2.3859E−01 (375.14)
10000 + 4.8739E−04(375.14)2 − 3.7160E−07 (375.14)3
1000 S = 45, 257 ppm (wt).

100
10 (45,257 vs. 48,235; deviation = 1,758/48,235 = 3.6%).
1 Acknowledgment
0.1 The Texas Hazardous Waste Research Center (Lamar University,
Beaumont, Texas) provided partial support for this work.
0.01
0.001 BIBLIOGRAPHY
300 350 400 450 500 550 600
Boiling point, K 1. Guide to Occupational Exposure Values. (1997). American
Conference of Governmental Industrial Hygienists, Cincin-
Figure 10. Solubility of monochloroalkanes in water. nati, OH.
SOLUBILITY OF HYDROCARBONS IN SALT WATER 559
2. Yaws, C.L. (1999). Chemical Properties Handbook. McGraw- concentration:

Hill, New York.
3. Yalkowsky, S.H. (1990–2004). Aquasol Database. University log10 S = A + B X + C X 2 (1)
of Arizona, Tucson, AZ.
4. Hine, J. and Mukherjee, P.K. (1975). J. Org. Chem. 40: 292. where S = solubility in salt water at 25 ◦ C,
5. Horvath, A.L. and Getzen, F.W. (1999). J. Phys. Chem. Ref. parts per million by weight, ppm(wt)
Data 28(3): 649–777. X = concentration of salt (NaCl) in water,
6. Horvath, A.L. (1982). Halogenated Hydrocarbons: Solubility- parts per million by weight, ppm(wt)
Miscibility with Water. Marcel Dekker, New York. A, B, and C = regression coefficients
7. Howard, P.H. and Meylan, W.M. (Eds.). (1997). Handbook of
The coefficients (A, B, and C) are given in Table 1.
Physical Properties of Organic Chemicals. CRC Press, Boca
Raton, FL.
The values in the table were determined from regression
of the data sources (1–10) for water solubility. The
8. Liquid–Liquid Equilibrium Data Collection. (1979). Dechema
Chemistry Data Series, Frankfurt/Main, Germany, Vol. V,
compilations by Price (7) and Sutton and Calder (8) were
part 1. especially helpful. The values presented are applicable to
9. Mackay, D., Shiu, W.Y., and Ma, K.C. (1992, 1992, 1993,
a wide variety of hydrocarbons (alkanes, cyclopentanes,
1995, 1997). Illustrated Handbook of Physical-Chemical cyclohexanes, benzenes (no and single substitution), and
Properties and Environmental Fate for Organic Chemicals. benzenes (double and triple substitution). The range
Vols. 1–5, Lewis, New York. of application for the correlations is about C5 to
10. Suzuki, T. (1991). J. Computer-Aided Molecular Design 5: C16 compounds.
149. Graphs of water solubility are shown in Figs. 1–3 for
11. Verschueren, K. (1996, 2001). Handbook of Environmental representative hydrocarbons (pentane, methylcyclopen-
Data on Organic Chemicals, 3rd and 4th Edn. Van Nostrand tane, and benzene). In the figures, solubility values are
Reinhold, New York. plotted at salt concentrations ranging from water without
12. Donald, M., Shiu, W-Y., and Ma, K-C. (1995). Illustrated salt to water saturated with salt. The graphs discloses
Handbook of Physical-Chemical Properties and Environmen- favorable agreement of correlation values and experimen-
tal Fate for Organic Chemicals. Vol. 4, Lewis, New York. tal data.
13. (2001). Handbook of Environmental Data on Organic
Chemicals, 4th Edn. John Wiley & Sons, New York. Examples
The correlations for solubility are useful in engineering
applications involving hydrocarbons in contact with water
SOLUBILITY OF HYDROCARBONS IN SALT containing salt. Examples are shown below.
WATER
Example 1. A chemical spill of n-pentane occurs in
CARL L. YAWS a body of saltwater [salt = 10,000 ppm(wt)] at ambient
PRASAD K. NARASIMHAN conditions. Estimate the concentration in the saltwater.
Lamar University The values in Table 1 are used for the coefficients.
Beaumont, Texas Introducing the boiling point of n-pentane to calculate A
and using the tabulated values for B and C provides
INTRODUCTION A = −17.7916 + 1.7865E−01 TB − 5.0255E−04

TB 2 + 4.1218E−07 TB 3
Physical and thermodynamic property data are required
to design and operate industrial processes. In particular, = −17.7916 + 1.7865E−01(309.22)
the water solubility of hydrocarbons in the sea and ocean
is becoming increasingly important because of more and − 5.0255E−04(309.22)2 + 4.1218E−07(309.22)3
more stringent regulations regarding the environment. = 1.5850
In this article, results are presented for the solubility
of hydrocarbons in salt water. The results are applicable B = −4.5956E−06
to the complete range of salt concentrations, including C = 2.2978E−12
water without salt to water saturated with salt. The
results are intended for use in engineering applications. Substitution of these coefficients and the salt con-
As an example of such usage, solubility values from the centration in the correlation equation for solubility in
correlation are useful in determining the distribution saltwater yields
of a hydrocarbon spill upon contact with seawater.
Solubility values at other salt concentrations may also log10 S = A + B X + C X 2
be ascertained.
= 1.5850 − 4.5956E−06(10,000)
CORRELATION OF SALT WATER SOLUBILITY + 2.2978E−12(10,000)2

= 1.5393
The correlation of the solubility of hydrocarbons in salt
water containing salt is based on a series expansion in salt S = 34.62 ppm (wt)
560 SOLUBILITY OF HYDROCARBONS IN SALT WATER
Table 1. Solubility in Salt Watera

log10 S = A + BX + CX 2 S-solubility, ppm (wt), X(salt) − ppm(wt)
Chemical Family A, B, and C Formula Example Compound
Alkanes A = −17.7916 + 1.7865E−01TB − Cn H2n+2 n-pentane

5.0255E−04 TB 2 + 4.1218E−07 TB 3
B = −4.5956E−06
C = 2.2978E−12
Cyclopentanes A = 9.55 − 2.2987E−02TB Cn H2n Methyl cyclopentane
B = −4.5987E−06
C = 2.2993E−12
Cyclohexanes A = 9.75 − 2.2987E−02TB Cn H2n Cyclohexane
B = −4.5987E−06
C = 2.2993E−12
Benzenes (no and single A = 5.725 + 3.6778E−03 TB − Cn+6 H2n+6 Benzene
substitution) 3.0346E−05 TB 2
B = −3.2714E−06
C = 1.6357E−16
Benzenes (double and A = 5.375 + 3.6778E−03 TB − Cn+6 H2n+6 Benzene
triple substitution) 3.0346E−05 TB 2
B = −3.2714E−06
C = 1.6357E−16
a
S-solubility in salt water at 25 C, parts per million by weight, ppm(wt).
X-salt (NaCl) concentration in water, ppm(wt).
Salt concentration range is X = 0 to 358,700 (saturated).
Salt in seawater is 34,472 ppm(wt).
A, B, and C—regression coefficients.
TB —boiling point of compound, K.
TBMAX —maximum temperature, K = 560.
TBMIN —minimum temperature, K = 300.
Correlation should not be used for compounds having boiling points outside TB MAX and TB MIN.
Range of application for correlation is about C5 to C16 .
Data Equation Data Equation

45 45
40 40
Solubility in salt water, ppm(wt)
35 35
30 30
25 25
20 20
15 15
10 10
5 5
0 0
0 50 100 150 200 250 300 350 400 0 50 100 150 200 250 300 350 400
Salt concentration in water, thousands of ppm(wt) Salt concentration in water, thousands of ppm(wt)
Figure 1. Solubility of n-pentane in salt water. Figure 2. Solubility of methyl cyclopentane in salt water.
using the tabulated values for B and C provides

The calculated value and data compare favorably (34.62
vs. 34.50; deviation = 0.12/34.50 = 0.3%). A = 5.725 + 3.6778E−03 TB − 3.0346E−05 TB 2
= 5.725 + 3.6778E−03 (353.31)
Example 2. A chemical spill of benzene occurs in
seawater [salt = 34,472 ppm(wt)] at ambient conditions. − 3.0346E−05 (353.31)2
Estimate the concentration in the seawater.
= 3.2364
The values in Table 1 are used for the coefficients.
Introducing the boiling point of benzene to calculate A and B = −3.2714E−06
SOLUBILITY OF HYDROCARBONS AND SULFUR COMPOUNDS IN WATER 561
Data Equation 9. Verschueren, K. (1996, 2001). Handbook of Environmental

Data on Organic Chemicals, 3rd and 4th Edn. Van Nostrand
2000
Reinhold, New York.
1800
10. Yalkowsky, S.H. (1990–2004). Aquasol Database. University

1600 of Arizona, Tucson, AZ.
1400 11. Yaws, C.L. (1999). Chemical Properties Handbook. McGraw-
1200 Hill, New York.
1000
800 SOLUBILITY OF HYDROCARBONS AND
600 SULFUR COMPOUNDS IN WATER
400
200
CARL L. YAWS
PRASAD K. NARASIMHAN
0
0 50 100 150 200 250 300 350 400 HELEN H. LOU
Salt concentration in water, thousands of ppm(wt) Lamar University
Beaumont, Texas
Figure 3. Solubility of benzene in salt water.
RALPH W. PIKE
Louisiana State University
C = 1.6357E−16. Baton Rouge, Louisiana
Substitution of these coefficients and the salt con-

centration in the correlation equation for solubility in Correlations are presented for the solubility of hydrocar-
saltwater yields bons and organic sulfur compounds in water. A wide vari-
ety of compounds are covered—alkanes, olefins, diolefins,
log10 S = A + B X + C X 2 acetylenes, cyclopentanes, cyclohexanes, benzenes, mer-
captans, thiophenes, and sulfides. The new correlation
= 3.2364 − 3.2714E−06(34,472) provides reliable solubility values down to very low con-
centrations (parts per million range). The correlation is
+ 1.6357E−16(34,472)2
based on the boiling point of the compound. Correlation
= 3.1236 and experimental data are in favorable agreement. The
results are useful in process engineering for wastewater
S = 1329.24 ppm (wt) minimization.

(1329.24 vs. 1391.00; deviation = 61.76/1391.00 = 4.4%). IMPORTANCE OF WATER SOLUBILITY
The solubility of compounds in water is very important.

Acknowledgment
This importance will increase in the future because of
The Texas Hazardous Waste Research Center (Lamar University,
Beaumont, Texas) provided partial support for this work.
heath, safety, and environmental considerations. The
following brief discussion illustrates the importance even
at very low concentrations.
BIBLIOGRAPHY For health involving human exposure to substances
in air, the threshold limit value (TLV) for n-pentane in
1. Donald, M., Shiu, W-Y., and Ma, K-C. (1995). Illustrated air is given as 600 ppm (parts per million, vol) by the
Handbook of Physical-Chemical Properties and Environmen- Occupational Safety and Health Act (1). A concentration of
tal Fate for Organic Chemicals. Vol. 4, Lewis, New York. only 0.0000001 mol fraction of n-pentane in water provides
2. (2001). Handbook of Environmental Data on Organic about 7000 ppm of n-pentane in air at the air–water
Chemicals, 4th Edn. John Wiley & Sons, New York. interface. This concentration of 7000 ppm in air exceeds
3. Hine, J. and Mukherjee, P.K. (1975). J. Org. Chem. 40: 292. the threshold limit value of 600 ppm.
4. Howard, P.H. and Meylan, W.M. (Eds.). (1997). Handbook of For safety, the lower explosion limit (LEL) for n-
Physical Properties of Organic Chemicals. CRC Press, Boca pentane in air is given as 1.4% by Yaws (2). A
Raton, FL. concentration of only 0.000001 mol fraction of n-pentane
5. Mackay, D., Shiu, W-Y., and Ma, K-C. (1992, 1992, 1993, in water provides about 7% of n-pentane in air at the
1995, 1997). Illustrated Handbook of Physical-Chemical air–water interface. This concentration of 7% in air
Properties and Environmental Fate for Organic Chemicals. exceeds the lower explosion limit of 1.7%.
Vols. 1–5, Lewis, New York. For environmental, consider a spill of n-pentane in
6. Price, L.C. (1976). Bull. Am. Assoc. Pet. Geologists 60(2): 213. contact with water. The water will become saturated. At
7. Sutton, C. and Calder, J.A. (1974). Environ. Sci. Technol. 8(7): saturation, the solubility of n-pentane in water is about
654. 0.0000385 weight fraction or 0.00000916 mol fraction,
8. Suzuki, T. (1991). J. Comput.-Aided Molecular Design 5: 149. as given by Yalkowsky (3). This concentration of only
562 SOLUBILITY OF HYDROCARBONS AND SULFUR COMPOUNDS IN WATER
0.00000916 mol fraction at saturation will provide about search was conducted to identify data source publica-
644,000 ppm or 64.4% of n-pentane in air at the air–water tions (2–10). The excellent compilations by Howard and
interface. This concentration of 644,000 ppm or 64.46% Meylan (5); Mackay, Shiu, and Ma (6); Verschueren (8);
greatly exceeds both the threshold limit value of 600 ppm Yalkowsky (3); and Yaws (2) were used extensively. The
and the lower explosion limit of 1.4%. publications were screened and copies of appropriate data
This brief discussion (using n-pentane as an example) were made. These data were then keyed in to the com-
indicates that very low concentrations (ppm or less) of puter to provide a database for which experimental data
compounds in water can provide concentrations in air at are available. The database also served as a basis to check
the air–water interface that exceed the threshold limit the accuracy of the correlation.
value for human exposure and the lower explosion limit Graphs of water solubility versus boiling point are
for flammability. shown in Figs. 1–8 for the compound families. The graphs
indicate favorable agreement of correlation values and
experimental data.
CORRELATION FOR WATER SOLUBILITY
For olefins and diolefins, the graphs disclose that more
In earlier work by Yaws and coworkers (2), the water data are available for alkanes. The graphs also disclose
solubility of various chemical types was correlated as a that the data for olefins and diolefins appear to be above
function of the boiling point of the compound. In this and approximately parallel to the more extensive data for
article it was determined that the boiling point method alkanes. This suggests that the solubility of olefins and
is also applicable to correlation of the water solubility of diolefins is higher and approximately parallel to the water
hydrocarbons and organic sulfur compounds: solubility of alkanes.
log10 (S) = A + BTB + CTB 2 + DTB 3 (1) data-alkanes equation

◦
100
where S = solubility in water at 25 C, ppm by
weight,
10
TB = boiling point of compound, K
A, B, C, and D = regression coefficients 1
The range for boiling point is 298 K to about 560 K.
0.1
The coefficients (A, B, C, and D) are given in Table 1
for a wide variety of hydrocarbons and sulfur com-
0.01
pounds—alkanes, olefins, diolefins, acetylenes, cyclopen-
tanes, cyclohexanes, benzenes, mercaptans, thiophenes,
0.001
and sulfides. The table also provides the boiling point
range (TBMIN , TBMAX ) for which the correlation is appli-
0.0001
cable. The correlation should not be used for compounds
250 300 350 400 450 500 550 600
whose boiling points are outside the boiling point range.
Boiling point, K
The coefficients were determined from regression of
available data. In preparing the correlation, a literature Figure 1. Solubility of alkanes in water.
Table 1. Solubility in Watera
log10 S = A + BTB + CTB + DTB 3

2
(S - ppm(wt), TB − K)
Chemical family A B C D TBMIN TBMAX Formula Example Compound
Alkanes −17.6520 1.77811E−01 −5.00907E−04 4.11124E−07 298 560 Cn H2n+2 n-Hexane

Olefins −17.0300 1.77811E−01 −5.00907E−04 4.11124E−07 298 560 Cn H2n 1-Hexene
Diolefins −16.5610 1.77811E−01 −5.00907E−04 4.11124E−07 298 560 Cn H2n−2 1,5-Hexadiene
Acetylenes −15.8350 1.77811E−01 −5.00907E−04 4.11124E−07 298 560 Cn+2 H2n+2 1-Hexyne
Cyclohexanes −16.7000 1.77811E−01 −5.00907E−04 4.11124E−07 298 560 Cn H2n Methyl cyclohexane
Cyclopentanes −16.9000 1.77811E−01 −5.00907E−04 4.11124E−07 298 560 Cn H2n Methyl cyclopentane
Benzenes (single −24.0080 2.21196E−01 −5.55632E−04 4.18830E−07 298 560 Cn+6 H2n+6 Toluene
substitution)
Benzenes (multiple −23.6500 2.21196E−01 −5.55632E−04 4.18830E−07 298 560 Cn+6 H2n+6 p-Xylene
substitution)
Mercaptans −6.9000 1.00500E−01 −2.72880E−04 1.99870E−07 298 560 Cn H2n+2 S Propyl mercaptan
Thiophenes −6.8500 1.00500E−01 −2.72880E−04 1.99870E−07 298 560 Cn H2n−4 S 2-Methyl thiophene
Sulfides −6.5390 1.00500E−01 −2.72880E−04 1.99870E−07 298 560 Cn H2n+2 S Ethyl methyl sulfide
a
S = solubility in water at 25 C, parts per million by weight, ppm(wt).
A, B, C, and D = regression coefficients.
TB = boiling point of compound, K.
TBMAX -maximum temperature, K.
TBMIN -minimum temperature, K.
SOLUBILITY OF HYDROCARBONS AND SULFUR COMPOUNDS IN WATER 563
data-cyclopentanes equation Data-alkanes Data-diolefins

1000.0000 Equation Equation
1000
100.0000
100

10.0000
10
1.0000 1
0.1000 0.1
0.01
0.0100
0.001
0.0010
250 300 350 400 450 500 550 600 0.0001
Boiling point, K 250 300 350 400 450 500 550 600
Boiling point, K
Figure 2. Solubility of cyclopentanes in water.
Figure 5. Solubility of diolefins in water.
data-benzenes equation
10000
Equation Equation
Data-mercaptans Data-benzenes
1000 100000
10000
100
1000
10 100
10
1 1
0.1
0.1
250 300 350 400 450 500 550 600 0.01
Boiling point, K
0.001
Figure 3. Solubility of benzenes in water. 250 300 350 400 450 500 550 600
Boiling point, K
Data-alkanes Figure 6. Solubility of mercaptans in water.

Data-olefins
Equation Equation
1000
Equation Data-benzenes
100 Data-thiophenes Equation
10000
10
1000
1
100
0.1
0.01 10
0.001 1
0.0001 0.1
250 300 350 400 450 500 550 600
Boiling point, K 0.01
250 300 350 400 450 500 550 600
Figure 4. Solubility of olefins in water.
Boiling point, K
Figure 7. Solubility of thiophenes in water.

For sulfur compounds (mercaptans, thiopheolefins, and
diolefins), the graphs disclose that more data are available
for benzenes. The graphs also, disclose that the data approximately parallel to the more extensive data for
for sulfur compounds appear to be slightly above and benzenes. This suggests that the solubility of sulfur
564 SORPTION KINETICS
Equation Equation The calculated value and data compare favorably

Data-benzenes Data-sulfides (524.68 vs 542.4; deviation = 17.72/542.4 = 3.3%).
100000
Acknowledgments
10000 The Texas Hazardous Waste Research Center (Lamar University,
1000 Beaumont, Texas) provided partial support for this work. Portions
of this material appeared in Oil & Gas Journal 89; 79 (April 8,
100 1991), Chem. Eng., 100; 122 (Oct., 1993), Oil & Gas Journal 89:
86 (Sept. 16, 1991), Oil & Gas Journal 95 (35): 80 (Aug. 29, 1994),
10
and Chem. Eng. 110: 60 (Aug., 2003). These portions are reprinted
1 by special permission.
0.1
BIBLIOGRAPHY
0.01
0.001 1. Guide to Occupational Exposure Values. (1997). American

250 300 350 400 450 500 550 600 Conference of Governmental Industrial Hygienists, Cincin-
Boiling point, K nati, OH.
2. Yaws, C.L. (1999). Chemical Properties Handbook. McGraw-
Figure 8. Solubility of sulfides in water. Hill, New York.
3. Yalkowsky, S.H. (1990–2004). Aquasol Database. University
compounds is higher and approximately parallel to the of Arizona, Tucson, AZ.
water solubility of benzenes. 4. Hine, J. and Mukherjee, P.K. (1975). J. Org. Chem. 40: 292.
5. Howard, P.H. and Meylan, W.M. (Eds.). (1997). Handbook of
Examples Physical Properties of Organic Chemicals. CRC Press, Boca
Raton, FL.
The correlations for water solubility are useful in
6. Mackay, D., Shiu, W-Y., and Ma, K-C. (1992, 1992, 1993,
engineering applications involving process wastewater. 1995, 1997). Illustrated Handbook of Physical-Chemical
Examples are shown below. Properties and Environmental Fate for Organic Chemicals.
Vols. 1–5, Lewis, New York.
Example 1. A chemical spill of n-hexane occurs into 7. Suzuki, T. (1991). J. Comput.-Aided Molecular Design 5: 149.
a body of water at ambient conditions. Estimate the 8. Verschueren, K. (1996, 2001). Handbook of Environmental
concentration in the water at saturation. Data on Organic Chemicals, 3rd and 4th Edn. Van Nostrand
The correlation for alkanes maybe used to determine Reinhold, New York.
the solubility in water. Substitution of the coefficients and 9. Donald, M., Shiu, W-Y., and Ma, K-C. (1995). Illustrated
boiling point of n-hexane in the correlation equation yields: Handbook of Physical–Chemical Properties and Environmen-
tal Fate for Organic Chemicals. Vol. 4, Lewis Publishers, New
log10 (S) = A + BTB + CTB 2 + DTB 3 York.
10. (2001). Handbook of Environmental Data on Organic
= −17.6520 + 1.77811E−01(341.88)
Chemicals, 4th Edn. John Wiley & Sons, New York.
− 5.00907E−04(341.88)2
+ 4.11124E−07(341.88)3
SORPTION KINETICS
S = 10.46 ppm (wt)
TH.D. KARAPANTSIOS
The calculated value and data compare favorably (10.46 M.X. LOUKIDOU
vs 9.47; deviation = 0.99/9.47 = 10.4%). K.A. MATIS
Aristotle University
Example 2. A chemical spill of toluene occurs into a body Thessaloniki, Greece
of water at ambient conditions. Estimate the concentration
in the water at saturation.
The correlation for benzenes (single substitution) INTRODUCTION
maybe used to determine the solubility in water.
Substitution of the coefficients and boiling point of toluene Sorption is, by definition, a general term describing
in the correlation equation yields the attachment of charged species (such as toxic metal
ions) from a solution to a coexisting solid surface.
log10 (S) = A + BTB + CTB 2 + DTB 3 Sorption kinetics may be controlled by several independent
processes that can act in series or in parallel (see Fig. 1).
= −24.0080 + 2.21196E−01 ∗ (383.78)
These processes fall in one of the following general
− 5.55632E−04 ∗ (383.78)2 categories:
+ 4.18830E−07 ∗ (383.78)3 1. bulk diffusion,
S = 524.68 ppm (wt) 2. external mass transfer (film diffusion),
SORPTION KINETICS 565
where q is the specific metal uptake (mg of metal

per g of sorbent), C is the metal bulk concentration
Intraparticle diffusion (mg/L) and X is the sorbent feed per unit volume of
solution (g/L). Subscripts o and t denote conditions at
the beginning and any other instant of the process,
Reaction respectively. The superscript s denotes conditions at
Mass transfer the sorbent interface. For adequately long adsorption
times: qt ≈ qe and Cst ≈ Cse ≈ Ce , where the subscript e
denotes equilibrium conditions. The known Langmuir and
Freundlich adsorption models have usually been employed
Figure 1. Illustration of the mechanism during sorption to to describe equilibrium metal uptake:
describe suitably and model the respective process kinetics.
qmax bCe
qe = (2a)
3. chemical reaction (chemisorption), and 1 + bCe
4. intraparticle diffusion. qe = KF C1/n
e (2b)
Kinetic analyses allow estimating sorption rates and also qmax (mg/g), b (L/mg), and KF [mg [g (mg/L)−1/n ]−1 ] and n
lead to suitable rate expressions characteristic of possible (dimensionless) are constants.
reaction mechanisms (1–3). To compare measurements from various experiments
At sufficiently high agitation speed in a reaction vessel, for a kinetics investigation, it is necessary to introduce a
the bulk diffusion step can be safely ignored because dimensionless degree of conversion. Thus, by normalizing
then sorption onto sorbent particles is decoupled from the remaining ion concentration, Ct , with respect to some
mass transfer in the bulk mixture. Apart from that, reference value, an index of sorption is defined. Taking
it is quite common that more than one process can advantage of the values of Ct before the onset of sorption,
contribute to system performance at the same time. In this Co , and for completed sorption, C∞ = Ce , the following
case, the extensive interrelationships between the various degree of conversion is proposed:
equations make the overall kinetic model exceedingly
complicated to evaluate. A rather simplifying approach Co − C t
α= (3)
to circumvent this problem is to assume that each one of C o − Ce
the co-current processes dominate over the others (i.e., the
rate-controlling step) at specific time regimes of the process Kinetic studies customarily use the basic conversion
and then study them independently (4); this approach was rate equation:
also adopted in this article. dα
= k(T)f (α) (4)
Many studies engaged so far to examine sorption dt
phenomena involved analysis of batch experiments where
where f (α) is a conversion-dependent function and k(T) is
data were sampled at even time intervals over the entire
the reaction rate constant. Sorption in heterogeneous
course of the process. As a result, fast changing kinetic
liquid–solid systems has usually been considered to follow
data characteristic of the phenomena just after the onset of
Arrhenius kinetics, which means that the temperature
sorption could not be accurately depicted on an adequately
dependence of the rate constant k may be described by the
short timescale. Thus, a primary objective of this article
well-known expression:
is to investigate the kinetic mechanism of metal sorption
on sorbent particles, putting more emphasis on samples
k(T) = A exp(−E/RT) (5)
collected at short times after the initiation of the process,
where the major part of the adsorption occurs. However, to apply the model in such systems, the
To identify the most appropriate mechanism for a morphology of the reactive surface and the final sorption
process, several models must be checked for suitability capacity must not vary with temperature, conditions
and consistency in a broad range of system parameters.
commonly overlooked by many workers (6).
The model selection criteria proposed by Ho et al. (5) for
The example used in this article deals with an
sorption of pollutants in aqueous systems were used
investigation of the removal and depletion of cadmium
herein as a general guideline. According to this, several
and chromates from aqueous solutions by biosorption
reaction-based and diffusion-based models were tested in
on Aeromonas caviae (the sorbent). This microorganism
simulating our data. The finally chosen kinetic models are
is often present in groundwater and generally in
those that fit the data closely and also represent reasonable
aquatic environments. Biosorption, the uptake of heavy
sorption mechanisms.
metals by dead biomass, has gained credibility during
recent years as it offers a technically feasible and
EQUILIBRIUM AND KINETIC MODELING economical approach (7–10); the latter author illustrated
that adsorption is different from ion exchange. Below,
An overall mass balance of the sorbate across the sorbent evidence is also provided that the system examined is a
surface can be written as complex process. More than one sorption model were often
(Co − Cst ) reported to describe a case study correctly (11). Attention
X qt + Cst = Co ⇔ qt = (1) should be paid to any possible experimental error.
X
The Langmuir isotherm equation was often found Equations 7 and 8 provided quite a suitable description
to fit quite satisfactorily the experimental data (12,13). of data for advancing time (see Fig. 2). It is noteworthy
Experimental adsorption isotherms of metal cations or that both models adequately capture the rapid rate of
oxyanions, obtained with different sorbent concentrations, adsorption during the first minutes of the experiments.
temperatures, etc. are usually presented as figures and This implies that the metal uptake by the sorbent is a
then, tables display the results of fitting the Langmuir satisfactory rate-controlling parameter under a second-
and/or the Freundlich model to the data obtained; the order reaction mechanism. The numerically best fit values
values of the respective correlation coefficient (r2 ), if high, of the rate parameters of these equations were displayed in
are often used to demonstrate a satisfactory description. a table (16). The predicted equilibrium sorption capacities
The assumptions involved in these models were well were quite close to the experimental values for both
described, among others, by Volesky (9). The Langmuir models. Nevertheless, the rate constant of the pseudo-
isotherms obtained during the equilibrium study lend second-order model, km , is monotonically correlated with
further support to the notion of a monolayer (14). changes in the biomass load and in the bulk concentration,
On the other hand, the Freundlich isotherm is heavily features that have been encountered in the past in
used in industry. This model is often accurate for higher biosorption (18).
concentrations, but is inaccurate at low concentrations due In contrast, the rate constant of the Ritchie second-
to its power-law nature. The equilibrium properties do not order equation, k2 , fluctuated, beyond any physical
change proportionally with the sorbent addition, possibly reasoning. In addition, Equation 8 exhibited better fitting
due to an aggregation of solids at higher loads and hence statistics. Despite the goodness of fit for sorption at 40 and
reducing the effective adsorption area (15). 60 ◦ C, the reaction rate constant of both models varied
From the chemical reaction category (chemisorption), randomly with temperature. Preliminary calculations,
the best fit for the data sets of this study is achieved
by second-order chemical reactions (16). The solution of
the standard second-order reaction, based on constant (a) 7
stoichiometry of one metal ion per binding site, is (e.g., 4):
6
Co
Ct = (6)
Co 5
1− exp(−k1 Ce t)
Ce
4
q, mg/g
X :1 g/L
where k1 is the reaction rate constant [L*(mg−1 of
metal)*min−1 ]. This adsorption model has been very 3 X :2 g/L
effective in describing the kinetics of adsorption of gases X :0.5 g/L
on solids. Nevertheless, when the data were plotted, it 2 T :40°C
was shown that Eq. 6 clearly fails to capture the steep
concentration gradient of the early removal stage (16). T :60°C
1
This was a direct indication that adsorption on solids
from a liquid phase is a process different from adsorption 0
from a gas phase, where traditionally the remaining bulk 0 20 40 60 80 100 120 140
concentration dictates the kinetics. t, min
If the rate of sorption depends not on bulk concentration (b) 6
but on uptake by the sorbent, this can be described by the
so-called Ritchie second-order equation according to which X :1 g/L
5
one metal ion occupies two binding sites (17): X :0.5 g/L

1 4 X :2 g/L
qt = qe 1 − (7)
1 + k2 t T :40°C
t /q
3 T :60°C
where k2 is the reaction rate constant (min−1 ). When in the
above treatment, qe does not dictate sorbate uptake, then a
pseudo-second-order rate expression is more appropriate: 2
t 1 1 1
= + t (8)
qt km q2m qm
0
where km is the reaction rate constant [g of sorbent*(mg−1 0 20 40 60 80 100 120
of metal)*min−1 ] and qm is a numerically determined t, min
parameter which, under ideal second-order rate control, Figure 2. Comparison of experimental removal curves vs.
corresponds to qe . In the literature (5), various other theoretical predictions based on (a) the Ritchie second-order
kinetic equations have been attempted: zero, first (forward equation (at an initial cadmium concentration of 5 mg/L) and
or reversible)-order, Langmuir–Hinshelwood, and the (b) the pseudo-second-order equation (at an initial cadmium
Elovich-type. concentration of 50 mg/L). Reprinted from Reference 16.
SORPTION KINETICS 567
using the Arrhenius model between two temperatures concentration at the sorbent surface is zero at all times.
every time, always gave activation energies below However, this is not true, particularly where a significant
10 kJ/mol, which is far less than expected for reaction- quantity of sorbate is adsorbed rapidly at the beginning
controlled sorption processes. The morphological changes of the process. A more realistic model should consider
in the biomass (sorbent) surface at different temperatures instead that a rapid equilibrium is established between
and the dependence of sorption capacity on temperature the sorbate at the interface and that present on the sorbent
may be blamed for this irregularity. surface (21), and this concept is adopted in this article.
If one combines the mass balance across the sorbent
surface, Equation 1, the Langmuir adsorption isotherm,
DIFFUSION KINETICS AND MODELING
Equation 2a, and the rate of change equation in the bulk
concentration, one ends up with (21)
For intraparticle diffusion, Crank (19) proposed a model
that takes into account the continuously decreasing bulk
dCt
concentration due to sorbate removal. This gives rise to a = −Km S(Ct − Cst ) (11)
time-dependent boundary condition for the concentration dt

at the surface of the sorbent particle. The solution of the dCst Km S
= [(Ct − Cst )(1 + bCst )2 ] (12)
diffusion equation for such a boundary condition and a dt Xqmax b
concentration independent diffusivity is
∞ where Km is the external mass transfer coefficient (m/s)
exp(−ξ p2n t)
α =1−6 (9) and S is the specific surface area of the sorbent particles
9/(1 − ) + (1 − )p2n per unit volume of the reactor (m2 /m3 ). Using the following
n=1
dimensionless variables: C∗ = Ct /Co , C∗s = Cst /Cso , and t∗ =
where pn is given by the nonzero roots of t/τ , where Cso is determined from Eqs. 1 and 2a and τ is
the total adsorption time, Eqs. 11 and 12 convert to
3pn
tan(pn ) = (10)
3 + p2n /(1 − ) dC∗
= −Km Sτ (C∗ − C∗s ) (13)
dt∗
and ≡ (Co − C∞ )/Co is the fraction of metal ultimately
dC∗s Km Sτ
adsorbed by the sorbent. = (C∗ − C∗s )(1 + bCo C∗s )2 (14)
Equation 9 can be solved numerically to determine ξ , dt∗ Xqmax b
the effective diffusional time constant, which for particle
(also called micropore) diffusion control equals Dc /R2c ; which is a system of two first-order ordinary differential
Dc and Rc are the intraparticle diffusion coefficient equations that must be solved simultaneously. The initial
(m2 /s) and mean particle radius (m), respectively. The conditions are C∗ = 1 and C∗s = 0 at t∗ = 0. Note that
same expression is the solution of the diffusion equation normalizing Cst with respect to Cso —and not Co , as Puranik
for (macro)pore diffusion control but only where the et al. (21) did —and also t with respect to τ , improved the
equilibrium isotherm is linear (qe ≈ K*Ce ) for the stability and convergence characteristics of the solution
concentration range under investigation. Then ξ equals markedly due to comparable spreading of all variables
(Dp /R2p )/[1 + (1 − εp )K/εp ], where εp Dp /[1 + (1 − εp )K] is over the computational domain.
the effective macropore diffusivity, K is the equilibrium Equations 9–10 and 13–14 are solved numerically to
constant, εp is the void volume fraction, and Rp is the determine ξ and Km S, respectively. The nonlinear numer-
radius of the particle (14). ical regression to fit experimental data to those equations
The experimental procedure in this article is based is performed by the Levenberg–Marquardt method, which
on measuring the remaining metal concentration in the gradually shifts the search for the minimum of the sum
bulk, so is always above 0.5. For greater than about of the errors squared (SSE), from the steepest descent to
0.1, the effect of diminishing bulk concentration becomes quadratic minimization—Gauss–Newton (22):
significant, and under these conditions, the assumption of
constant metal concentration at the surface of the sorbent (qexp,i − qcal,i )2
leads to an erroneously high diffusivity. Provided that SSE = (15)
the concentration step employed in a sorption experiment i
q2exp,i
corresponds to a linear section of the equilibrium isotherm,
the constant K can be replaced by the local slope of the Figure 3a presents the results of fitting Eq. 9 to biosorp-
isotherm, dqe /dCe . This is approximately valid for the tion data obtained with different initial concentrations,
concentrations employed in this work, as can be seen solids loads, and temperatures. It is apparent that, despite
from the experimental isotherms. Our interest is chiefly some scatter in measurements, the finite volume diffusion
focused on the short time region ( Dc,p t/R2c,p < 0.2) where model can fairly well describe the entire range of data, also
Eq. 9 converges slowly, so at least 200 terms are used in including the steep concentration gradient at short times.
the summation to achieve satisfactory accuracy. The values of , ξ , and the computed values of Dc were
External mass transfer has been customarily analyzed displayed in a table (14). Such behavior has been custom-
in literature by adopting a pseudo-first-order reaction ary as a consequence of the decreasing slope of a nonlinear
model (5,20). This approach assumes that the sorbate equilibrium curve, example, a Langmuir isotherm, which
(a) end of sorption (for α higher than 0.9), a much slower

process, intraparticle diffusion, gradually becomes the
1
rate-controlling step. Yet, this is a regime of no practical
significance.
The computed mass transfer rate constant, Km S, was
0.8
presented in a table (14). As regards the effect of changing
the solids load, it appears that Km S is virtually not
0.6 affected. To assess the statistical significance of the
determination, a derivative time series analysis has been
a
performed where dα/dt is plotted versus t and then,

0.4 Co = 5 mg/L, s = 0.5 g/L, T = 20°C a Hanning low-pass filter was applied to flatten out
the signal undulations until dα/dt versus α became
Co = 5 mg/L, s = 1 g/L, T = 20°C
a reasonably smooth curve (23). Next, the integrated
0.2 Co = 5 mg/L, s = 2 g/L, T = 20°C smoothed signal α(t) was processed as before and resulted
Diffusion model in Km S values closely comparable to the values obtained
from the unfiltered data, with no preferential trend
0 in biomass load. This is an additional positive sign
0 20 40 60 80 100 120 140
t, min
that external mass transfer may be the predominant
mechanism of the sorption process after the initial fast
(b) 1.2 chromium removal.
As a conclusion in the example thoroughly studied, we
1 may say that the analysis conducted has not been capable
of providing strong evidence in favor of any of the examined
mechanisms because several diverse kinetic models were
0.8
successful in fitting the experimental data. Perhaps,
the questioning of surface kinetics by Levenspiel (24) is
0.6 known, concluding that it is good enough to use the
a
simplest available correlating rate expression.

0.4 Co = 5 mg/L, s = 1 g/L, T = 20°C
Co = 5 mg/L, s = 1 g/L, T = 40°C NOMENCLATURE
0.2 Co = 5 mg/L, s = 1 g/L, T = 60°C
External mass transfer q: specific metal uptake (mg/g)
0 C: metal bulk concentration (mg/L)
0 0.2 0.4 0.6 0.8 1 1.2 X: sorbent loading per unit volume of solution (g/L)
t qmax :
Langmuir constant (mg/g)
Figure 3. Experimental degree of conversion, α, vs. predictions b: Langmuir constant (L/mg)
based on the solution of (a) the diffusion equation, for various KF : Freundlich constant {mg [g (mg/L)−1/n ]−1 }
adsorbent loads and (b) the mass transfer equation, for various n: Freundlich constant (dimensionless)
temperatures (both at an initial chromium concentration of E: apparent activation energy (J/mol)
5 mg/L). Reprinted with permission from Reference 14; copyright
A: preexponential factor (min−1 )
(2004) American Chemical Society.
n: apparent reaction order
R: gas constant (8.3136 J/mol/K)
T: temperature (K)
causes the diffusivity to increase rapidly as concentration
t: time (min)
increases (11).
k: rate constant
A way to check further on the possibility of a pore
k1 : second-order chemical reaction rate constant
diffusion-controlled mechanism is to perform desorption
[L∗ (mg−1 of metal)∗ min−1 ]
kinetic tests with adsorbent previously used for sorption.
k2 : second-order Ritchie reaction rate constant (min−1 )
For a nonporous sorbent particle, transport of solute inside
km : pseudo-second-order reaction rate constant [g of
the particle may be neglected, and it can be assumed that
sorbent∗ (mg−1 of metal)∗ min−1 ]
biosorption occurs mainly at the particle surface. This
Dc : intraparticle diffusion coefficient (m2 /s)
idea can be effectively extended to cases of relatively large
Rc : mean particle radius (m)
macropores, where the metal ions may have ready access
Rp : radius of a particle
to react with internal surface sites.
Km : external mass transfer coefficient (m/s)
It was argued that, at the beginning of the process,
S: specific surface area of sorbent particles per unit
chromium is sorbed according to quite a fast and highly
volume of the reactor (m2 /m3 )
favorable chemical mechanism, such as ion exchange, but
soon external film diffusion comes into play (14). On this
Greek Letters
account, if one ignores the very first minute of sorption,
the remaining curves were fitted pretty well by the model α: degree of conversion
Eqs. 13 and 14. Figure 3b displays these results. Near the : fraction of metal ultimately adsorbed by the sorbent
SOUND IN WATER 569
ξ: numerically determined parameter 19. Crank, J. (1975). Mathematics of Diffusion. Oxford University
εp : void volume fraction Press, London.
20. Dzul Erosa, M.S., Saucedo, T.I., Navarro Mendoza, R., Avila
Rodriquez, M., and Guibal, E. (2001). Cadmium sorption
BIBLIOGRAPHY on chitosan sorbents: Kinetic and equilibrium studies.
Hydrometallurgy 61: 157–167.
1. Tien, C. (1994). Adsorption Calculations and Modelling. 21. Puranik, P.R., Modak, J.M., and Paknikar, K.M. (1999). A
Butterworth-Heinemann, Boston. comparative study of the mass transfer kinetics of metal
biosorption by microbial biomass. Hydrometallurgy 52:
2. Yiacoumi, S. and Tien, C. (1995). Kinetics of Metal Ion
189–197.
Adsorption from Aqueous Solutions. Kluwer Academic,
Boston. 22. Bates, D.M. and Watts, D.G. (1988). Nonlinear Regression
2+
Analysis and its Applications. John Wiley & Sons, New York.
3. Xu, Y. and Schwartz, F.W. (1994). Sorption of Zn and
Cd2+ on hydroxyapatite surfaces. Environ. Sci. Technol. 28: 23. Bendar, J.S. and Piersol, A.G. (1986). Random Data: Analysis
1472–1480. and Measurement Procedures. John Wiley & Sons, New York.
4. Hill, C.G. (1977). An Introduction to Chemical Engineering 24. Levenspiel, O. (1999). Chemical Reaction Engineering, 3rd
Kinetics and Reactor Design. John Wiley & Sons, New York, Edn. John Wiley & Sons, New York, p. 380.
pp. 24–166.
5. Ho, Y.S., Ng, J.C.Y., and McKay, G. (2000). Kinetics of
pollutant sorption by biosorbents: review. Sep. Purific. SOUND IN WATER
Methods 29: 189–232.
D.L. MARRIN
6. Smith, J.M. (1981). Chemical Engineering Kinetics, 3rd Edn.
McGraw-Hill, Singapore, pp. 37–92, 636–645. Hanalei, Hawaii
7. Aksu, Z. (2001). Equilibrium and kinetic modeling of
cadmium(II) biosorption by C. vulgaris in a batch system:
Effect of temperature. Sep. Purific. Tech. 21: 285–294. As sound travels through an elastic medium such as water,
8. Butter, T.J., Evison, L.M., Hancock, I.C., Holland, F.C., a wave is created that imparts energy to the individual
Matis, K.A., Philipson, A., Sheikh, A.I., and Zouboulis, A.I. molecules—causing them to compress and then to relax
(1998). The removal and recovery of cadmium from dilute along the path of the wave. Because the medium is elastic,
aqueous solutions by biosorption and electrolysis at labora- the distance separating individual water molecules is
tory scale. Water Res. 32: 400–406. permitted to vary just slightly, so that the propagation
9. Volesky, B. (1999). Biosorption of Heavy Metals. CRC Press, of sound may be defined as a periodic variation in
Boca Raton, FL, Chap. 6. Available: http://ww2.mcgill.ca/ pressure that is transmitted via adjacent molecules. Sound
biosorption/publication/index.htm. emanating from a point source in water usually (but not
10. Zouboulis, A.I., Matis, K.A., and Hancock, I.C. (1997). Bio- always) results in a three-dimensional waveform that
sorption of metals from dilute aqueous solutions. Sep. Purif. spreads spherically through the medium (1). Similarly,
Methods 26: 255–295. a planar source of sound in water can be considered
11. Smith, E.H. (1996). Uptake of heavy metals in batch multiple point sources that create a wave consisting of
systems by a recycled iron-bearing material. Water Res. 30: all the individual waves.
2424–2434. Sound waves are longitudinal, meaning that molecules
12. Mandjiny, S., Zouboulis, A.I., and Matis, K.A. (1995). Re- (e.g., water or air) comprising the elastic medium move
moval of cadmium from dilute solutions by hydroxyapatite. in a direction parallel to the propagating wave. Because
Part I. Sorption studies. Sep. Sci. Tech. 30: 2963–2978. water has a significantly higher density than air, water
13. Kefala, M.I., Zouboulis, A.I., and Matis, K.A. (1999). Biosorp- transmits sound more rapidly and efficiently than air. The
tion of cadmium ions by Actinomycetes and separation by water molecules oscillate, or move back and forth over
flotation. Environ. Pollut. 104: 283–293.
a very short longitudinal distance; hence, sound waves
14. Loukidou, M.X., Karapantsios, Th.D., Zouboulis, A.I., and are classified as mechanical. By contrast, light waves and
Matis, K.A. (2004). Diffusion kinetic study of chromium(VI)
microwaves are classified as electromagnetic and do not
biosorption by Aeromonas caviae. Ind. Eng. Chem. Res. 43:
1748–1755.
involve the movement of water molecules for propagation.
The speed of sound in water is about 1500 meters per
15. Esposito, A., Pagnanelli, F., Lodi, A., Solisio, C., and Vegliò,
F. (2001). Biosorption of heavy metals by Sphaerotilus
second or 3300 miles per hour (i.e., approximately five
natans: An equilibrium study at different pH and biomass times greater than that in air), which varies slightly with
concentration. Hydrometallurgy 60: 129–141. temperature, pressure, and salinity.
16. Loukidou, M.X., Karapantsios, Th.D., Zouboulis, A.I., and
Matis, K.A. (2003). Cadmium(II) biosorption by Aeromonas
caviae: Kinetic modelling. 15th Intl. Biohydrometallurgy
FREQUENCY AND AMPLITUDE
Symp., Athens.
17. Ritchie, A.G. (1977). Alternative to the Elovich equation for The frequency of a sound wave is the number of waves
the kinetics of adsorption of gases on solids. J. Chem. Soc.- that pass by a given point during a specified period of
Farad. Trans. 73: 1650–1653. time. A single wave consists of a cycle that includes
18. Cheung, C.W., Porter, J.F., and McKay, G. (2001). Sorption both compression (squeezing together) and rarefaction
kinetic analysis for the removal of cadmium ions from (rebounding apart); the distance over which a cycle is
effluents using bone char. Water Res. 35: 605–612. continually repeated is known as the wavelength. The
570 SOUND IN WATER
period of time over which waves are normally counted is from a straight path) (5). Ambient noise in a shipping
one second; therefore, sound frequencies are expressed as channel can exceed that outside a channel by as much as
cycles per second or hertz. Whereas high-frequency sounds 45 decibels (re 1 µPa), but noise in a channel is attenuated
possess short wavelengths, low-frequency sounds possess about 100 decibels from that measured directly beneath
long wavelengths. Frequency ranges and estimated ships (2).
intensity levels for common underwater sounds are Due to the vertical stratification of oceans, sound
listed in Table 1. Frequency-dependent differences in behaves differently in turbulent shallow waters from
underwater sound attenuation result in low-pitched the way it does in either more isothermal middepth
sounds that travel farther than high-pitched sounds waters or waters immediately overlying the bottom (see
generated at similar intensities. Table 2). Sound transmission at the surface is highly
The amplitude of a sound wave is the maximum dependent on local conditions (e.g., wind, precipitation,
longitudinal displacement (i.e., during one cycle of swells, bubbles) and is usually restricted to low-frequency
compression and rarefaction) of water molecules relative sounds. However, most of the ocean lies between shallow
to their resting, or equilibrium, position. The intensity and bottom waters, creating a relatively thick layer that
of a sound wave is proportional to its amplitude and is transmits a spectrum of sound waves that is far less
often quantified in units of decibels. The decibel scale is attenuated than that in either shallower or deeper waters.
based on a ratio between the intensities (expressed as Both humans and whales have taken advantage of this
a power) of a measured sound and a reference sound sofar channel to broadcast signals and songs, respectively,
that corresponds to a standard pressure. Many of the over vast oceanic distances (6). Sounds generated within
man-made sounds listed in Table 1 are actually more the sofar channel itself or at shallower depths alongside
intense than naturally produced sounds, contributing, at seamounts are entrained and transmitted within this
least in part, to the controversy regarding the possible deep isothermal layer as a result of many contributing
effects of anthropogenic noise on marine organisms. Sound physical processes.
intensities are commonly presented for a distance of only 1
meter from the source and, as such, are rarely indicative of UNDERWATER DETECTION
ambient underwater noise (e.g., the cumulative intensity
of sounds measured at greater distances from their Detecting and identifying an underwater sound from a
respective sources). particular source is often difficult due to the aforemen-
tioned effects of the environment and to the myriad of
FATE OF OCEANIC SOUND background noises from which any single sound must be
distinguished. Sea sounds are detected by hydrophones
The propagation of sounds in the ocean is a rather com- (i.e., the underwater equivalent of a microphone) that are
plex matter owing to differences in water density as a often positioned in specific arrays, thus permitting the rel-
function of depth, latitude, seafloor topography, and many ative position of a sound source to be determined. This type
other factors. Changes in seawater temperature, salinity, of passive detection may be used to locate oceanic precip-
and pressure contribute to changes in density that, in itation, breaking surface waves, underwater explosions,
turn, affect sound waves traveling through the medium. undersea volcanoes, shipping traffic, and marine organ-
Generally, sound is attenuated by spreading (proportional isms that emit audible sounds (e.g., crustaceans, fishes,
to distance), reflection (due to solid structures or bound- mammals) (7).
aries between water of different densities), scattering (due Active detection techniques, such as sonar, are routinely
to rough surfaces), absorption (conversion of acoustic to used to locate underwater objects because the objects
thermal energy), and refraction (deflection of sound waves themselves need not emit sound and because the results
Table 1. Approximate Ranges for the Frequencies and Intensity Levels of Common Underwater Sounds. Reported Ranges
were Complied from Data Presented in Various Sourcesa
Intensity Level, Decibels
Sound Source Frequency, Hertz re 1 µPa at 1 mb
Ship engines and propellers 10–5000 160–190

Navigation and profiling sonars 100–3000 180–230
Explosive devices/air guns 1000–17,000 190–260
Military surveillance sonars 1000–10,000 190–235
Icebreaking/drilling operations 20–1000 100–150
Whale songs and moans 10–8000 120–190
Dolphin clicks and whistles 500–25,000 100–180
Cetacean echolocation 10,000–150,000 130–230
Snapping shrimp colony 2000–15,000 180–190
Lightning strikes/undersea volcanoes/earthquakes 0.1–20,000 Up to 260
a
Reference 2–4.
b
Intensity levels are presented in units of decibels relative to a reference pressure of 1 micropascal at a distance of 1 meter from the sound source. Sound
level attenuation is a function of distance from the source, frequency range, and various environmental factors.
SOUND IN WATER 571
Table 2. Approximate Depths and Representative Sound Speeds for Various Water Layers in the Ocean. Reported Ranges
were Complied from Data Presented in Various Sourcesa
Depth, Speed,
Vertical Zone Meters m/s Comments
Overlying air NAb 330 Significantly lower acoustic impedance than water; minimal
transfer of sound energy
Surface water layer 0 to 50 1515 Influenced by weather (e.g., wind, rain, temp.) and turbulence
that affect sound waves
Seasonal thermal layer 50 to 800 1485–1505 Sound channel (mostly for low frequencies); limited vertical
thickness; refracts sound waves originating shallower and
deeper
Deep isothermal layer 800 to 4000+c 1485–1525 Optimal sound channel (for all frequencies); extensive vertical
thickness and horizontal extent; minimal sound attenuation
Bottom water layer 4000+ to seafloorc 1525 Influenced by seafloor topographic features that reflect and
scatter sound waves
Underlying rock NAb 2000–6000 Higher acoustic impedance than water; moderate transfer of
sound energy
a
References 2, 6, and 7.
b
Not applicable.
c
Indicates that the transition between deep isothermal layers and bottom water layers may exceed 4000 meters, depending on the exact location within an
ocean basin.
are generally more interpretable. The most basic type of on the water itself, creates infrasound at frequencies of
sonar is echolocation, whereby a sound pulse (acoustic 1 to 5 hertz (10). Anthropogenic infrasound in the ocean
energy) is introduced into the water. Depending on the is created primarily via shipping traffic, low-frequency
sonic wavelength and the size of the objects, a portion of active sonar (LFAS), and acoustic thermography. A
that acoustic energy is reflected back to the source (e.g., thermography technique known as ATOC generates tones
a ship) where it is detected, transduced, and analyzed. down to about 1 hertz, whereas LFAS sweeps multiple
The distance, movement, size, and even shape of some tones of 100 to 500 hertz across each other and produces
objects may be determined using sophisticated sonars (e.g., resulting tones as low as 0.1 hertz (2). Both ATOC and
Doppler, multibeam) and data analysis techniques (8). LFAS generate peak intensity sounds that are greater
Several marine organisms—most notably cetaceans—are than those of ships.
believed to use echolocation as a means of finding or
assessing prey, predators, conspecifics, seabed features,
and environmental conditions (2). BIBLIOGRAPHY
ULTRASOUND AND INFRASOUND 1. Leighton, T.G. (1994). The Acoustic Bubble. Academic, San
Diego, Chap. 1.
Most of the underwater sounds discussed thus far lie 2. Stocker, M. (2003). Ocean bio-acoustics and noise pollution.
within the sonic range of humans (i.e., frequencies of 20 J. Acoust. Ecol. 4: 16–29.
to 20,000 hertz), but water also transmits mechanical 3. Science of Sound in the Sea. (2004). Office of Marine Programs.
waves of ultrasonic (>20,000 hertz) and infrasonic (<20 University of Rhode Island, HTML format.
hertz) frequencies. Dolphins emit whistles at frequencies 4. Greene, C.R. and Moore, S.E. (1995). Man-made noise. In:
as high as 30,000 hertz and complex echolocation clicks Marine Mammals and Noise. W.J. Richardson et al. (Eds.).
at frequencies exceeding 300,000 hertz (8). Obviously, Academic, San Diego, Chap. 6.
the upper limit of hearing of dolphins far exceeds that 5. Haines, G. (1974). Sound Underwater. Crane Russak, New
of humans. In addition to marine organisms, ambient York, pp. 22–25.
oceanic noise associated with bubbles (i.e., trapped gases), 6. Payne, R. (1995). Among Whales. Scribner, New York,
surface winds, waves, and sea spray can reach ultrasonic Appendix.
frequencies. Man-made ultrasounds include everything 7. Medwin, H. and Clay, C.S. (1998). Fundamentals of Acousti-
from complex signals for communication to guided waves cal Oceanography. Academic, San Diego, Chap. 1 and 10.
for inspecting underwater pipelines (9). 8. Thompson, D.H. and Richardson, W.J. (1995). Marine mam-
The moans of baleen whales are the most common mal sounds. In: Marine Mammals and Noise. W.J. Richardson
source of biologically produced infrasound, which has been et al. (Eds.). Academic, San Diego, Chap. 7.
documented down to about 10 hertz (8). By contrast, 9. Na, W.B. and Kundu, T. (2002). Underwater pipeline inspec-
a number of geologic and meteorologic events (e.g., tion using guided waves. J. Pressure Vessel Technol. 124:
volcanoes, earthquakes, sea ice cracking, hurricanes) 196–200.
produce infrasonic waves that propagate through the 10. Garcés, M., Hetzer, C., Marrifield, M., Willis, M., and
ocean at frequencies of approximately 0.1 to 10 hertz. Aucan, J. (2003). Observations of surf infrasound in Hawai’i.
Even the breaking of large surf, either onto the shore or Geophys. Res. Lett. 30: 2264–2266.
572 WATER ON THE SPACE STATION
WATER ON THE SPACE STATION installments will explore air management, thermal control
and fire suppression—in short, all of the things that will
PATRICK L. BARRY make the Space Station comfortable and safe.
TONY PHILLIPS
NASA MAKING A SPLASH IN SPACE
Before recycling can begin, there has to be some water to

Rationing and recycling will be an essential part of start with.
life on the International Space Station. In this article, ‘‘We have plenty of water on the Space Station now,’’
Science@NASA explores where the crew will get their says Jim Reuter, leader of the ECLSS group at the
water and how they will (re)use it. Marshall Space Flight Center. ‘‘The Russian module Zarya
November 2, 2000—Future astronauts poised to blast is packed with contingency water containers (CWCs) that
off for an extended stay on the International Space Station were carried over from the Space Shuttle during assembly
(ISS) might first consider dashing to the restroom for a missions earlier this year. They look like duffle bags and
quick splash at the lavatory, or better yet, a luxurious each one holds about 90 lbs.’’
hot shower. Once on board the ISS, spacefarers are in ‘‘But it’s expensive to ferry water from Earth,’’ he added.
for a steady diet of sponge baths using water distilled ‘‘We have to recycle. There’s already a Russian-built water
from—among other places—their crewmates breath! processor in orbit that collects humidity from the air. Here
If you’re squeamish, read no farther, because the at Marshall we’re building a regenerative system that will
crew will eventually include lab rodents—and they’ll be be able to recycle almost every drop of water on the station
breathing, too. All of the denizens of the space station lose and support a crew of seven with minimal resupplies.’’
water when they exhale or sweat. Such vapors add to the The ECLSS Water Recycling System (WRS), developed
ambient cabin humidity, which is eventually condensed at the MSFC, will reclaim waste waters from the Space
and returned to the general water supply. Shuttle’s fuel cells, from urine, from oral hygiene and
Sometimes it’s better not to think about where your hand washing, and by condensing humidity from the air.
next glass of water is coming from! Without such careful recycling 40,000 pounds per year
Rationing and recycling will be an essential part of of water from Earth would be required to resupply a
daily life on the ISS. In orbit, where Earth’s natural life minimum of four crewmembers for the life of the station.
support system is missing, the Space Station itself has to Not even research animals are excused from
provide abundant power, clean water, and breathable air the program.
at the right temperature and humidity—24 hours a day, ‘‘Lab animals on the ISS breath and urinate, too, and
7 days a week, indefinitely. Nothing can go to waste. we plan to reclaim their waste products along with the
In this article, the first of a series about the practical
challenges of living in space, Science@NASA will examine
how the Space Station’s Environmental Control and Life
Support System (ECLSS), under continuing development
at the Marshall Space Flight Center, will help astronauts
use and re-use their precious supplies of water. Future
This article is a US Government work and, as such, is in the Shuttle pilot Terry Wilcutt with 7 contingency water containers
public domain in the United States of America. destined for the space station Mir
WATER ON THE SPACE STATION 573
crew’s. A full complement of 72 rats would equal about one for the ECLSS project at MSFC. ECLSS depends on
human in terms of water reclamation,’’ says Layne Carter, machines—not microbes—because, ‘‘if a machine breaks,
a water-processing specialist at the MSFC. you can fix it.’’
It might sound disgusting, but water leaving the space The water purification machines on the ISS will cleanse
station’s purification machines will be cleaner than what wastewater in a three-step process.
most of us drink on Earth. The first step is a filter that removes particles and
‘‘The water that we generate is much cleaner than debris. Then the water passes through the ‘‘multi-filtration
anything you’ll ever get out of any tap in the United beds,’’ which contain substances that remove organic and
States,’’ says Carter. ‘‘We certainly do a much more inorganic impurities. And finally, the ‘‘catalytic oxidation
aggressive treatment process (than municipal waste water reactor’’ removes volatile organic compounds and kills
treatment plants). We have practically ultra-pure water bacteria and viruses.
by the time our water’s finished.’’
EVERY DROP COUNTS
MIMICKING MOTHER EARTH Once the water is purified, astronauts will do everything
possible to use it efficiently. ‘‘On the ground, people flick
On Earth, water that passes through animals’ bodies is on the faucet and they probably waste a couple of liters
made fresh again by natural processes. Microbes in the soil of water just because it’s free and the water pressure is
break down urea and convert it to a form that plants can high,’’ notes Carter.
absorb and use to build new plant tissue. The granular ‘‘On the ISS, the water pressure will be about half what
soil also acts as a physical filter. Bits of clay cling to you might experience in a typical household,’’ Carter said.
nutrients in urine electrostatically, purifying the water ‘‘We don’t use faucets on the ISS, we use a wash cloth. It’s
and providing nutrients for plants. much more efficient. If you’re an astronaut, you’ll wet the
Water excreted by animals also evaporates into the wash cloth with a spray nozzle and then use the cloth to
atmosphere and rains back down to the Earth as fresh wash your hands.’’
water—a natural form of distillation. On the space station, people will wash their hands with
Water purification machines on the ISS partly mimic less than one-tenth the water that people typically use on
these processes, but they do not rely on microbes or any Earth. Instead of consuming 50 liters to take a shower,
other living things. which is typical on Earth, denizens of the ISS will use less
‘‘While you try to mimic what’s happening on than 4 liters to bathe.
Earth—which is so complicated if you really think Even with intense conservation and recycling efforts,
about it—we have to use systems that we can control the Space Station will gradually lose water because of
100 percent,’’ said Monsi Roman, chief microbiologist inefficiencies in the life support system.
When water evaporates from the ocean and surface waters, it leaves behind impurities. In the
absence of air pollution, nearly pure water falls back to the ground as precipitation
574 STRONTIUM ISOTOPES IN WATER AND ROCK
‘‘We will always need resupply, because none of the ‘‘It takes processes that are slightly more efficient than
water reprocessing technology that is available right now we have developed for the space station to do that,’’ Edeen
for space flight . . . is 100 percent efficient. So there’s said. ‘‘Those are the next generation water processing
always some minimal loss,’’ said Marybeth Edeen, deputy systems. Those are being developed now, but they’re not
assistant manager of environmental control and life ready for space flight yet.’’
support at NASA’s Johnson Space Center. The ECLSS life support system will join the Space
Water is lost by the Space Station in several ways: Station as part of Node 3, which is scheduled to launch in
the water recycling systems produce a small amount of October 2005. Until then, the environment inside the ISS
unusable brine; the oxygen-generating system consumes will be maintained primarily by life support systems on
water; air that’s lost in the air locks takes humidity with the Russian Zvezda Service Module.
it; and the CO2 removal systems leach some water out of
the air, to name a few.
WEB LINKS
Lost water will be replaced by carrying it over from the
Shuttle or from the Russian Progress rocket. The Shuttle International Space Station—NASA’s Web page for the Interna-
produces water as its fuel cells combine hydrogen and tional Space Station
oxygen to create electricity, and the Progress rocket can Wheels in the Sky—Science@NASA article about humanity’s
be outfitted to carry large containers of water. dreams of a space station from the science fiction fantasies
NASA scientists will continue to look for ways to of the Nineteenth Century to Wernher von Braun’s catalytic
improve the life support systems of the Space Station, vision in the 1950s
reducing water losses and finding ways to reuse other Advanced Life Support Web Page—from the Johnson Space
waste products. If the water recycling systems can be Flight Center
improved to an efficiency of greater than about 95 percent, Environmental Control and Life Support Systems—describes
then the water contained in the Station’s food supply the life support systems being developed at Marshall Space
would be enough to replace the lost water, Edeen said. Flight Center
STRONTIUM ISOTOPES IN WATER AND ROCK

GHOLAM A. KAZEMI
Shahrood University of
Technology
Shahrood, Iran
INTRODUCTION
A brief summary of strontium and strontium isotopes,

their concentrations in natural materials, and their
evolution in the geologic system is given in the following
paragraph. These introductory notes have been taken
from a benchmark paper by Capo et al. (1). Anyone
intending to deal with strontium isotopes should consult
that paper as well as the paper by Clow et al., (2).
Textbooks such as those by Faure (3) and Faure (4) provide
One of the ‘‘nodes’’ that will become a part of the Space Station. excellent coverage of strontium isotopes in rocks. Being
The ECLSS life support equipment will be housed in Node 3, an encyclopedia of water, the emphasis here is more on
which is scheduled to be attached to the station in October 2005 strontium isotopes in water.
Strontium has four naturally occurring isotopes, whose
approximate abundances are 84 Sr, 0.56%; 86 Sr, 9.87%;
87
Sr, 7.04%; and 88 Sr, 82.53%. All four isotopes are
stable, although the fraction of 87 Sr varies due to
radioactive decay of 87 Rb. The ionic radius of strontium
(1.18 Å) is similar to that of calcium (1.00 Å), and,
hence, Sr substitutes for Ca in minerals, including
plagioclase feldspar, apatite, sulfates such as gypsum
and anhydrite, and carbonates (calcite, dolomite, and
especially aragonite). The average concentration of Sr in
soils is 240 ppm but may fall below 10 ppm or exceed
1000 ppm in some cases. Weathering of bed rock or
sediments can be a significant source of strontium in
soil. Soil parent materials often have distinct Sr isotopic
signatures. The average concentration of Sr in crustal
Sign up for our EXPRESS SCIENCE NEWS delivery rocks is 370 ppm but can vary from 1 ppm in some
STRONTIUM ISOTOPES IN WATER AND ROCK 575
ultramafic rocks to several percent in some aragonitic (9) to investigate the effect of carbonate dissolution
corals. If Rb and Sr are incorporated into a mineral on the chemical composition of groundwater (11); and
or rock at its formation and the system remains closed (10) to indicate the influence of seawater on fresh
with respect to those elements, then the amount of groundwater (12).
87
Sr increases over time as radioactive 87 Rb decays; the
amounts of 84 Sr, 86 Sr, and 88 Sr remain constant. Therefore, Analysis of Sr Isotopes in a water sample
older rocks in general have higher 87 Sr/86 Sr ratios than The strontium isotopes of a water sample are measured
younger ones that have the same initial Rb/Sr ratio. under clean laboratory conditions using a mass spectrom-
Over geologic time, rocks of a given age, composed of eter, such as a Sector 54 type (13), a Finnigan Mat 261
minerals that have a high Rb/Sr ratio (e.g., granites in multicollector model (2), or a VG354 thermal ionization
the continental crust), develop a higher 87 Sr/86 Sr than brand (9). To analyze water for a Sr isotope ratio using a
rocks that have a lower Rb/Sr ratio (e.g., oceanic basalts). Sector 54 Type Mass Spectrometer, an initial aliquot of
Thus, 87 Sr/86 Sr ratios in geologic materials are indicators the sample is placed in a microwave oven to be evaporated
of both age and geochemical origin. The strontium in rocks to dryness. For water that has a total Sr content of more
is released by weathering, cycled through vegetation and than 0.5 ppm, the aliquot is 2 mL, and evaporation time
animals, and eventually enters the oceans, primarily via is about 1 hour; for water that has low concentrations
rivers. Minerals within a crystalline rock or sediment such as rainwater, 500 mL is required which takes much
generally have variable Rb/Sr ratios and therefore a range longer to evaporate to dryness. Then 3 mL of 6 molar HCl
of 87 Sr/86 Sr ratios. River water has a Sr concentration that is added to the residue to dissolve it and left to stabilize
varies from ∼6 to 800 ppb and averages ∼60 ppb. The Sr for a few hours. The following steps are then taken one
concentration in seawater is about 8 ppm. Precipitation after another (13):
near ocean basins generally has a strontium isotopic
composition similar to that of seawater (0.709). However, 1. Add 1 mL of 2.5 molar HCl, and evaporate to
the concentration of Sr in rainfall is generally lower dryness by heating slowly.
than that in seawater by several orders of magnitude 2. Add 1 mL of 6 molar HCl, and leave to dissolve.
(usually <1 ppb). Strontium leaves the oceans, the largest
3. Transfer to centrifuge vessels, and centrifuge for
reservoir of dissolved Sr, primarily by deposition in marine
around 5 minutes.
carbonate. A small amount of Sr is also transferred directly
from the oceans to the atmosphere and hence transferred 4. Pour 1 mL of the sample into the column (the
to the continents in precipitation. column detail is AG W50-X 8).
5. Add 1 mL of 2.5 molar HCl to the column.
SR ISOTOPES IN WATER AND ROCK: SOURCES 6. Add 1 mL of 2.5 molar HCl to the column.
AND IMPORTANCE 7. add 1 +30 mL of 2.5 molar HCl to the column in
two stages, respectively.
The Sr isotopic ratio in natural water reflects the 8. Add 10 mL of 2.5 molar HCl to the column.
contribution made by various minerals as they dissolve
in water. Rivers are the major provider of strontium 9. Collect Sr by placing a container beneath the
to the oceans and define, to a great extent, the marine column and letting it drain.
87
Sr/86 Sr ratio. Strontium in rivers is mainly derived from 10. Add 1 mL of HNO3 , and slowly heat to dryness.
two end members: limestone and evaporites as one end 11. Add 1 mL of HNO3 , and slowly heat to dryness.
member (high Sr and low 87 Sr/86 Sr ratio of 0.706–0.709) 12. Add 1 mL of Milli-Q water to the sample.
and silicate rocks as the other end member that are low
13. Place a small volume of the sample in the
in Sr and have a higher 87 Sr/86 Sr ratio. Processes such as
holding filament.
atmospheric deposition, cation exchange, and biological
cycling, too, influence the Sr isotope ratio in water. 14. Heat sample slowly.
However, unless there have been recent anthropogenic 15. Add a small volume of phosphoric acid to the sample
or natural disturbances, annual gains or losses of Sr from to facilitate ionization.
cation exchange and biological pool should be minimal. 16. Heat the sample rapidly to a very high temperature
Sr isotopes in water samples are used (1) as a (to burn) to remove any Rb.
chemical tracer to determine the sources of atmospheric 17. Place the sample in the sample holder.
contributions to soil cation reservoirs over time (1);
18. Transfer sample/s to the mass spectrometer.
(2) to determine the contribution of various components
to the water balance of a lake (5); (3) to distinguish 19. Start running sample with mass spectrometer.
weathering reactions from cation exchange processes (6);
Analysis of Sr Isotopes in a rock sample
(4) to investigate the origin of brine, brine migration, and
the relationship between brine and diagenetic phases (7); Rocks Other Than Limestone. To analyze a rock sample
(5) to characterize weathering reactions and identify for Sr isotopes ratio using a Sector 54 Type Mass
sources of dissolved Ca in stream water (2); (6) as a tracer Spectrometer under clean laboratory condition, first it
of cation sources in stream water (8); (7) to correlate has to be ground to a 1 micron size. After grinding, for
and date marine sequences of late Neogene age (9); all rock types but limestone, the following steps should be
(8) to identify the origin of salts in groundwater (10); undertaken one after another:
576 STRONTIUM ISOTOPES IN WATER AND ROCK
1. 0.1 gram of sample to accuracy of 0.0001 gram standard delta notation (δ 87 Sr) relative to NBS standard
is selected. 987. To calculate δ 87 Sr, the following formula is used:
2. 2 mL of HF acid and 10 drops of HNO3 are added 87
( Sr/86 Sr)sample
to each sample. δ 87 Sr = 87 − 1 × 105
( Sr/86 Sr)seawater
3. Samples are heated to dryness.
4. 2 mL of HF is again added to each sample. where (87 Sr/86 Sr)sample is the measured ratio of 87 Sr/86 Sr in
5. 10 drops of HClO4 are added to each sample, and a sample that has been normalized to 86 Sr/88 Sr = 0.1194
sample is placed on a hot plate at 150 ◦ C. and (87 Sr/86 Sr)seawater is the ratio of 87 Sr/86 Sr in modern
seawater which is about 0.709172 (9). In most analysis,
6. 5 drops of HClO4 are added to each sample.
the precision of the 87 Sr/86 Sr ratios is ±0.0001 or better,
7. 5 more drops of HClO4 are added to each sample. and higher precision is of little value because seasonal
8. Samples are placed on a hot plate at 180 ◦ C. variations in 87 Sr/86 Sr ratios in rivers are commonly in the
9. Samples are then placed on a hot plate at 130 ◦ C third figure (14).
on the next day, and 1 mL 2.5 M HCl is added to In a majority of cases, the Sr isotopic composition of
each sample. a water or rock sample is a mixture of two or more end
10. Samples are transferred to centrifuge tube to be members. To calculate the relative proportion of each
centrifuged for 5 minutes. end member of a two component mixture, the following
equation applies:
11. From this step forward, all the steps are exactly
the same as those for water samples from Step (87 Sr/total Sr)m = x(87 Sr/total Sr)a + (1 − x)(87 Sr/total Sr)b
5 onward. (1)
where the subscript ‘‘m’’ is mixture, ‘‘a’’ is the first
Limestone. Due to its high solubility compared to other end member, ‘‘b’’ is the second end member, x is the
rock types, limestone is prepared for analysis using proportion of end member a, and (1−x) is the proportion
different techniques and the following steps: of end member b. To solve this equation, 87 Sr/total Sr can
be calculated from the equation given by (Reference 2)
1. 0.1 mg of sample is selected. as follows:
2. 1 mL of 1 M acetic acid is added to the sample.
87
Sr/total Sr = (87 Sr/86 Sr)/(3 +87 Sr/86 Sr) (2)
3. 2 mL of 1 M acetic acid is added.
4. After reaction subsides, sample is centrifuged, and Example (from Reference 13). In Table 1, the isotopic
the residue is decanted to a clean beaker and heated composition of groundwater (GW) is considered a mixture
to dryness at 150 ◦ C. of soil and rainwater isotopic compositions. The relative
5. 3 mL of 6.25 M HCl is added and evaporated to proportion of Sr from rainfall and soil to groundwater
dryness, and 1 mL more is added again and heated in Piezometer 40, P40, is computed here. In Equation 1,
to dryness. ‘‘m’’ = groundwater; ‘‘a’’ = rainfall; and ‘‘b’’ = soil; x =
6. 1 mL of 2.5 M HCl is added. proportion of end member a, rainfall; and 1−x = proportion
of end member b, soil. To calculate the proportion of
7. The sample is then loaded into columns (like other
each end member, first the 87 Sr/total Sr ratio (Eq. 2) must
rock types).
be computed:
8. The Sr is then collected and dried.
9. The columns are flushed and cleaned.
87
Sr/total Sr for rainfall = (0.7089)/(3 + 0.7089)
10. The dry sample (Sr) is again dissolved with 1 mL = 0.191135
of 2.5 M HCl and loaded into the columns again. 87 total
Sr/ Sr for soil = (0.72068)/(3 + 0.72068)
In contrast to the other samples, carbonates have
a higher calcium content. Therefore, they need to = 0.19368
be cleaned up twice before they are loaded into the 87 total
Sr/ Sr for groundwater = (0.7105)/(3 + 0.7105)
mass spectrometer.
= 0.191091
11. The rest of the procedure is like that for other
rock types. Substituting these values in Eq. 1, the relative contribu-
tion of rainfall to the Sr budget of groundwater in P40 is
REPORTING THE RESULTS OF SR ISOTOPIC ANALYSIS calculated as 86%, and only 14% of the Sr is from soils.
Case Studies
Before reporting the results of the strontium isotopic anal-
ysis, they should first be corrected for mass fractionation Kazemi (13) and Kazemi and Milne-Home (15), through
to 88 Sr/86 Sr = 8.37521. Also, the value of isotopic ratio a strontium isotope study of rainwater, surface water,
(88 Sr/86 Sr) of the NBS (National Bureau of Standards) 987 groundwater (shallow and deep), soil, and rocks of a
Sr metal used as a reference by the spectrometer should be salinized watershed in Eastern Australia (Buckinbah
included in the report. Measured strontium isotopic val- Creek watershed) found that the ratio of 87 Sr/86 Sr of
ues are expressed as both ratios (ratio of 87 Sr to 86 Sr) and shallow groundwater was similar to that of soil and
STRONTIUM ISOTOPES IN WATER AND ROCK 577
Table 1. The 87 Sr/86 Sr of Various Materials in the Buckinbah Creek Watershed in Eastern Australiaa
GW (Shallow) GW (Deep) Surface Water Rocksb Soils
B7 0.71197 B4 0.7084 S6 0.7099 Kabadah Fmc 0.70555 P45-Top 0.709

P19 0.7058 B3 0.708 S2 0.7087 Wansey Fmd 0.70813 P41 0.7373
P37 0.71 B2 0.7083 S5 0.7076 CV 0.72547 P42-Top 0.7135
P45 0.7078 B1 0.7084 Kabadah Fme 0.7051 P40 0.72068
P40 0.7105 Burrawong Ls 0.70812 P40B 0.72298
B5 0.7099 0.71047 0.720692
P41 0.7169 Rain 0.7089
P39 0.7092 Leaf 0.7115
P42 0.7125
P15 0.71023
a
Source: References 13.
b
CV: Canowindra Volcanics, Burrawong Ls: Burrawong limestone, Fm: Formation.
c
Rock type- shale.
d
Rock type- volcanoclastic sandstone.
e
Rock type- volcanoclastic.
rainwater (Table 1). They showed that between 20 to They found that the ratios of 87 Sr/86 Sr of Lake Fryxell,
90% of the Sr budget of shallow groundwater and surface Lake Hoare, and Lake Bonney were 0.70895, 0.71057, and
waters is derived from rainwater. They further concluded 0.71187–0.71204, respectively. They also found that in
that the deep groundwater is the source of baseflow at the Lake Fryxell and Lake Hoare, the 87 Sr/86 Sr became less
watershed outlet (due to identical 87 Sr/86 Sr ratios of deep radiogenic with depth, whereas in Lake Bonney, the ratios
groundwater and baseflow). increase slightly with depth.
Hunt et al. (6) used water and strontium isotopes to Sr isotopes together with isotopes of Nd and Pb of
investigate the hydrology of a natural and a constructed Cenozoic basalts in eastern Australia were studied by
wetland. They showed that the porewater from the natural Zhang et al. (19) to show that the isotopic signatures
wetland had a 87 Sr content ranging from 0.00 to 0.86 per of the Pacific Ocean midoceanic ridge basalts (MORB)
mil (87 Sr/86 Sr = 0.71026 to 0.71087) and the 87 Sr of pore type characterize the lava-field basalts in southeastern
water from the constructed wetland ranged from −2.35 Australia, whereas the mantle isotopic signatures of
to +1.24 per mil (87 Sr/86 Sr = 0.70859 to 0.71114). They Indian Ocean MORB type characterize younger basalts
attributed this difference to the presence of a thick peat from northeastern Australia.
layer in the natural wetland.
The 87 Sr/86 Sr ratio of 32 brine samples from the Upper BIBLIOGRAPHY
Jurassic Smackover Formation in southern Arkansas was 1. Capo, R.C., Stewart, B.W., and Chadwick, O.A. (1998). Stron-
measured at 0.7071–0.7101, more radiogenic than the Sr tium isotopes as tracers of ecosystem processes: Theory and
in Late Jurassic seawater because of the contribution of methods. Geoderma 82: 197–225.
Sr from detrital sources (7). 2. Clow, D.W., Alisa Mast, M., Bullen, T.D., and Turk, J.T.
The linear negative correlation between the Sr and (1997). Strontium 87/strontium 86 as a tracer of min-
87
Sr/86 Sr ratio (when plotted against each other) has eral weathering reactions and calcium sources in an
been used by Faure (3), Dogramaci et al. (16), and Singh alpine/subalpine watershed, Loch Vale, Colorado. Water
et al. (17) as proof of the mixing of two waterbodies. Resour. Res. 33(6): 1335–1351.
Similarly, the negative correlation between the molar 3. Faure, G. (1986). Principles of Isotope Geology. John Wiley &
ratio of Mg/Ca and Sr isotopes is an indication of Sons, New York.
the incongruent dissolution of calcitic minerals in the 4. Faure, G. (2001). Origin of Igneous Rocks: The Isotopic
aquifer (11). Incongruent dissolution of carbonates occurs Evidence. Springer, Heidelberg.
progressively through two mechanisms: 5. Clauer, N., Chauduhuri, S., Toulkeridis, T., and Blanc, G.
(2000). Fluctuations of caspian sea level: Beyond climatic
1. Dissolution of calcitic minerals, causing the addition variations. Geology 28(11): 1015–1018.
of Sr to groundwater and modification of the 6. Hunt, R.J., Bullen, T.D., Krabbenhoft, D.P., and Kendall, C.
87
Sr/86 Sr ratio; and (1998). Using stable isotopes of water and strontium to
investigate the hydrology of a natural and a constructed
2. Reprecipitation of calcitic minerals, thereby remov- wetland. Groundwater 36(3): 434–443.
ing some of the Sr from the groundwater without 7. Stueber, A.M., Pushkar, P., and Hetherington, E.A. (1984). A
changing the Sr isotope ratio, approaching eventu- strontium isotopic study of Smackover brines and associated
ally the ratio in calcitic minerals (11). solids, southern Arkansas. Geochim. Cosmochim. Acta 48:
1637–1649.
A large data set of Sr isotopic ratios of the rivers of the 8. Miller, E.K., Blum, J.D., and Friedland, A.J. (1993). Deter-
Ganges, Orinoco, and Amazon basins is found in Palmer mining of soil exchangeable-cation loss and weathering rates
and Edmond (14). using Sr isotopes. Nature 362: 438–441.
The study by Lyons et al. (18) focuses on the strontium 9. Hodell, D.A., Mead, G.A., and Mueller, P.A. (1990). Variation
isotopic signature of some streams and lakes in Antarctica. in the strontium isotopic composition of seawater (8 Ma to
578 TECHNETIUM IN WATER
present): Implications for chemical weathering rates and cyclotron by E. Lawrence and then sent to Perrier and
dissolved fluxes to the ocean. Chem. Geol. 80: 291–307. Segre. Technetium was the first element to be produced
10. Morgan, K. and Jankowski, J. (2003). Origin of salts in a artificially.
fractured bedrock aquifer in the Central West Region, NSW, Twenty-two isotopes of technetium whose masses range
Australia. Proc. Int. Conf. Groundwater Fractured Rocks, from 90 to 111 are known, and all are radioactive (Table 1).
Prague, Krasny-Hrkal-Bruthans (Eds.). IHP-VI, series on
Technetium has three long-lived radioactive isotopes:
groundwater no. 7, pp. 221–222.
technetium-97 [97 Tc] (half-life {T1/2 } = 2.6 × 106 years),
11. Dogramaci, S.S., Herczeg, A.L., and Bone, Y. (1998). 87 Sr/86 Sr 98
Tc (T1/2 = 4.2 × 106 years), and 99 Tc (T1/2 = 2.13 × 105
in Groundwater as Indicators of Carbonate Dissolution.
years). An important isotope is 95m Tc (‘‘m’’ stands for
Proc. 9th Int. Symp. on water–rock Interaction, Taupo, New
Zealand, March–April, 1998, G.B. Arehart and J.R. Hulston metastable state) [T1/2 = 61 days], which is used in tracer
(Eds.). A.A. Balkema, Rotterdam, pp. 211–214. work. However, the most useful isotope of technetium
12. Jorgensen, N.O. and Heinemeier, J. (2003). Isotope studies is 99m Tc (T1/2 = 6.01 hours), which is used in many
of saline groundwater in a sandstone aquifer on Bornholm, medical radioisotope tests because of its short half-life,
Denmark. Proc. Int. Conf. Groundwater Fractured Rocks, easy detection of the gamma ray it emits, and the ability
Prague, Krasny-Hrkal-Bruthans (Eds.). IHP-VI, series on of technetium to bind chemically to many biologically
groundwater no. 7, pp. 205–206. active molecules.
13. Kazemi, G.A. (1999). Groundwater factors in the management Technetium-99 is a weak, beta-emitting radionuclide,
of dryland and stream salinity in the upper Macquarie Valley, 0.292 million-electron volts (MeV), produced during
New South Wales, Australia. Unpublished Ph.D. Thesis, nuclear fission, which decays to the stable isotope
University of Technology, Sydney. ruthenium-99. It has a specific activity of 3.78 × 1010
14. Palmer, M.R. and Edmond, J.M. (1992). Controls over the transmutations per minute per gram, which corresponds
strontium isotope composition of river water. Geochim. to 17 microcuries per milligram (µCi/mg) (4). Technetium
Cosmochim. Acta 56: 2099–2111.
metal is a silver-gray color, tarnishes slowly in the
15. Kazemi, G.A. and Milne-Home, W.A. (1999). The application air, and has properties similar to those of rhenium.
of strontium, oxygen 18 and deuterium isotopes in the
However, as mentioned earlier, technetium does not occur
estimation of the solute budget and hydrology of a small
naturally.
watershed in Eastern Australia. Eos Trans. AGU 80(17)
Spring Meet. Suppl. S139, 1999. However, 99 Tc in terrestrial material has been evalu-
ated since its discovery. In 1961, 99 Tc was isolated and
16. Dogramaci, S.S., Herczeg, A.L., and Bone, Y. (1998). Chemi-
cal evolution of groundwater in the Murray Basin: Geochem- identified in African pitchblende (a uranium-rich ore) in
istry and mass transfer modelling. Abstract No. 49, Geol. Soc. extremely minute quantities (5). It was theorized that
99
Aust., 14th Aust. Geol. Convention. Tc was produced as a result of the spontaneous fission of
17. Singh, K.S. et al. (1998). Chemical and strontium, oxygen, uranium-238 (238 U).
and carbon isotopic compositions of carbonates from the Technetium-99 is produced primarily through anthro-
Lesser Himalaya: Implications to the strontium isotope pogenic processes such as detonation of nuclear weapons,
composition of the source waters of the Ganga, Ghaghara, and reprocessing of spent nuclear reactor fuel rods contain-
the Indus rivers. Geochim. Cosmochim. Acta 62: 743–755. ing uranium, and disposal of nuclear waste. Reprocessing
18. Lyons, W.B. et al. (2002). Strontium isotopic signatures of the involves dissolving the spent nuclear fuel in acid to sepa-
streams and lakes of Taylor Valley, Southern Victoria Land, rate the potentially reusable components of the spent fuel
Antarctica: Chemical weathering in a polar climate. Aquatic
Geochem. 8(2): 75–95.
19. Zhang, M., O’Reilly, S.Y., and Chen, D. (1999). Location of Table 1. Isotopes, Half-Lives, and Decay Processes for
Pacific and Indian mid-ocean ridge-type mantle in two time Technetiuma
slices: Evidence from Pb, Sr, and Nd isotopes for Cenozoic
Australian basalts. Geology 27(1): 39–42. Type of Type of
Nuclide Half-Lifeb Decay Nuclide Half-Life Decay
91
Tc 3.2 m β+, γ 99
Tc 2.13 × 105 y B−
TECHNETIUM IN WATER 92
Tc 4.4 m β+, γ 100
Tc 15.8 s B− , γ
93m
Tc 43.5 m ε, γ 101
Tc 14 m B− , γ
JAY CLAUSEN
93
Tc 2.7 h ε, β + , γ 102m
Tc 4.3 m B− , γ
94
Tc 53 m β+, γ 102
Tc 6.3 m B− , γ
AMEC Earth and 94
Tc 4.9 h ε, β + , γ 103
Tc 50 s B− , γ
Environmental, Inc. 95m
Tc 60 d ε, β + , γ 104
Tc 18.0 m B− , γ
Westford, Massachusetts 95
Tc 20 h ε, γ 105
Tc 7.6 m B− , γ
96m
Tc 52 m ε, γ 106
Tc 36 s B− , γ
INTRODUCTION 96
Tc 4.3 d ε, γ 107
Tc 21 s B− , γ
97m
Tc 91 d γ 108
Tc 5 s B− , γ
Technetium, element 43, was predicted on the basis of its 97
Tc 2.6 × 106 y ε 109
Tc 1 s B−
location in the periodic table. Technetium was erroneously 98
Tc 4.2 × 106 y B− , γ 110
Tc 0.83 s B− , γ
reported as having been discovered in 1925 by Noddack 99m
Tc 6 h γ , B− 111
Tc ? ?
and Tacke (1), who called it masurium. However, Element a
Reference 3.
43 was actually discovered in Italy (2). It was found in b
d: days; h: hours; m: minutes; s: seconds; y: years.
a sample of molybdenum, which had been bombarded c +
β : positron; β − : beta particle; ε: internal electron conversion; γ :
with deuterons in the University of California—Berkeley gamma ray.
(the uranium and plutonium) from the waste component Table 2. Thermodynamic
(the fission products). The fission products, which are of Data for Technetiuma
no use, must be disposed of properly. In a 1000-megawatt G0 f
nuclear reactor (thermal fission yield of approximately Speciesb kcal/gfwc
6%), approximately 28 g of 99 Tc is produced daily from
the slow neutron-induced fission of 235 U or plutonium-239 TcOH(s) −56.1
Tc(OH)2 (s) −110.2
[239 Pu] (6). The long half-life and mobility of 99 Tc makes
Tc3 O4 (s) −206.3
it an important element to the nuclear fuel industry in TcO4 (s) − −149.1
terms of its potential environmental impact. This is a par- TcS2 (s) −51.0
ticular concern for designing underground nuclear waste TcO2 (s) −88.3
storage facilities because 99 Tc is the most mobile of all
a
radionuclides present. Reference 19.
b
s: solid.
There are no known uses of technetium other than c
kcal/gfw: kilocalories per gram
99m
Tc for medical purposes. Technetium has supercon- formula weight
ducting properties and is a good corrosion preventer (7).
However, its unstable nature and production of radioactive
particles negate its practical use. 1.2
1 PO =
ANALYSIS 0.8 TcO2 2 1 bar
0.6
Analysis for 99 Tc typically consists of methyl ethyl ketone 0.4 TcO4−
Eh (v)
(MEK) extraction followed by liquid scintillation counting 0.2
using United States Environmental Protection Agency Tc3O4
0 Tc
(USEPA) Method 900.0. The typical quantification limit
−0.2
is 1 becquerel per liter (Bq/L), which is equivalent to 25
−0.4 Tc(OH)2 PH2 =
picocuries per liter (pCi/L). 1 ba r
−0.6
Chemistry −0.8
0 2 4 6 8 10 12 14
The oxidation states most commonly found for Tc are99 pH
Tc0 , Tc+4 , and Tc+7 . The most stable chemical species Figure 1. Eh -pH diagram for the Tc–O–H system at a concen-
of 99 Tc in aqueous solutions is the pertechnetate anion, tration of 90 µg/L under standard temperature and pressure.
TcO4 − , in the Tc+7 valence state, which is highly mobile
and nonreactive (8–12). However, 99 Tc is very sensitive
to changes in redox conditions (13). Under reducing The solubility of sodium pertechnetate is reportedly 11
conditions, 99 Tc can be precipitated by various cations moles per liter [mol/L] (13). The solubility of 99 Tc in a
to form anhydrous technetium dioxide, TcO2 , (as Tc+4 ) reducing environment varies from 1.32 × 10−8 to 2.56 ×
or technetium heptoxide, Tc2 O7 , or other insoluble oxides 10−9 mol/L and is controlled by the Tc species present (14).
and hydroxides such as Tc2 S7 or TcS2 or incorporated into Redox thus is a key variable affecting solubility (8). The
minerals (14–17): solubility of technetium under reducing conditions in
carbonate groundwater is pH dependent. At pH levels
TcO4 − + 4H+ + 3e− = TcO2 + 2H2 O; E0 = +0.738 volts between 5 and 7.5, the solubility of Tc+4 is less than
(1) 10−8 mol/L. Pilkington (22) suggests that the solubility of
In the presence of oxygen, the Tc+4 oxide or oxyhydroxide TcO2 is closer to 10−7 mol/L under reducing conditions.
can be reoxidized (14,18): At higher pHs of 8 to 11, solubility increases, and the
dominant ionic complex is TcO(OH)3 CO− 3 .
[TcO2 (s) or TcO(OH2 (s) ] → TcO4 − (2) Equally important are the fate-and-transport proper-
ties of 99 Tc. Under aerobic soil conditions, 99 Tc behaves as
The pertechnetate anion also forms complexes with an inert tracer and readily passes through soil unless
organic compounds containing –C –O and –COO func- there is a high content of organic matter or metal
tional groups, especially under reducing conditions. The oxide (10,23–30). As Gu and Dowlen (31) report, this is
thermodynamic data for technetium (Table 2) can be used not surprising because soil mineral surfaces are nega-
to develop a phase diagram to identify which species are tively charged as is TcO4 − and therefore the two repel
likely to be present at a given set of Eh and pH conditions each other. Once introduced to surface water or ground-
(Fig. 1). water, it can be transported significant distances due to
The pertechnetate ion is mobile under oxidizing its recalcitrant nature and lack of sorption. Given the
conditions (Fig. 1). Bidoglio et al. (20,21) found in column concern over 99 Tc mobility, extensive studies have been
experiments that the oxidized species persists, even under conducted in various geologic media to assess suitable
slightly reducing or anoxic conditions. Technetium can nuclear waste disposal storage facilities. Adsorption coef-
undergo a number of surface reactions on various minerals ficients for various minerals are presented in Sheppard
under reduced conditions. et al. (23), Wildung et al. (27), Palmer and Meyer (32), and
Strickert et al. (33). Sheppard et al. (23) indicate that soil National Laboratory where radioactive liquids generated
redox conditions and organic matter content are the key from routine operations were disposed of in shallow waste
variables influencing 99 Tc mobility in soil. pits (44). Nearby seeps have 99 Tc activity levels of up to
1185 pCi/L [3,200 Bq/L].
Blaylock and Frank (45) conducted a study in which
REGULATORY STANDARDS
TcO4 − was added to a freshwater pond. Their study
indicated that 84% of the TcO4 − added remained in
The USEPA has set an effective dose equivalent of 4
solution; very little was taken up by biota or by settling out
millirem (mrem)/year for exposure, which they incorrectly
into sediments. The vast majority of the 99 Tc was removed
calculated as 900 pCi/L (36 Bq/L). The correct activity
through drainage, indicating that the ultimate reservoir
equivalent to 4 mrem/year is 3790 pCi/L [152 Bq/L] (34).
for technetium released into the environment is the ocean.
The U. S. Department of Energy (DOE) Order 5400.5 sets
Similarly, in a study of the Rhone River, it was found that
a permissible ingested-water limit of 100 mrem/year or
the 99 Tc activity varied from 2.5 to 53 × 10−3 pCi/L (0.1 to
100,000 pCi/L (4000 Bq/L), whereas the U. S. Nuclear
2.1 × 10−3 Bq/L) in seven samples collected in 1985 and
Regulatory Agency (NRC) has set a value of 50 mrem/year
1986 (46).
or 60,000 pCi/L (2400 Bq/L). The NRC has set the
Finally, significant quantities of 99 Tc laden effluent
maximum activity of 99 Tc that can be released to a sanitary
water were released to a settling lagoon as well as to
sewer at 500 mrem/year or 600,000 pCi/L (24,000 Bq/L).
several small streams surrounding the PGDP. Activity
The proposed action level for groundwater cleanup at the
levels as high as 100,000 pCi/L (4000 Bq/L) were routinely
Paducah Gaseous Diffusion Plant (PGDP) site is 25 pCi/L
released to surface water. Ferrous iron sulfate was used
(1 Bq/L), which is the quantification limit.
to precipitate metals in the lagoon before discharge
into a nearby stream. The addition of this material
Rainwater
shifted the redox state of the lagoon such that TcO4 −
Technetium has been released to the atmosphere in the was precipitated from solution. The likely form of the
form of Tc2 O7 and HTcO4 as a consequence of aboveground technetium precipitate was Tc2 S7 . In addition, the iron-
detonation of nuclear weapons through the fission of rich clay and organic-rich stream sediments retained much
235
U and 239 Pu. The nuclear industry also releases of the 99 Tc, limiting its downgradient migration from the
TcF6 in addition to Tc2 O7 and HTcO4 during uranium lagoon. Therefore, the 99 Tc migrated only several thousand
enrichment. During a nuclear detonation, elemental 99 Tc feet from the point of release. However, downgradient of
burns to form Tc2 O7 , which upon cooling and contact the site, groundwater discharges to the surface water
with water vapor forms HTcO4 . The activity of 99 Tc streams and 99 Tc-rich water has been measured in several
measured in rainwater reportedly ranges from 8.2 × 10−8 seeps (47).
to 1.0 × 10−6 µg/L (35,36). Work by Leon (37) and Attrep
et al. (36) suggest the concentration of 99 Tc in rainwater Ground Water
has increased from the early 1960s through the 1980s.
The most extensive groundwater distribution of 99 Tc is
Both papers suggest that the source of this increase in
99 at the PGDP where a 2- 21 mile long plume of 99 Tc
Tc in rainwater is from the reprocessing of spent nuclear
contaminated groundwater is commingled with the solvent
fuel rather than nuclear weapons testing.
trichloroethene. The 99 Tc was introduced to the site
during the 1970s when fissioned uranium was reprocessed.
Surface Water 99
Tc became entrained in processing equipment at the
Technetium-99 is a surface water and groundwater PGDP, which was later cleaned with organic solvents and
contaminant at a number of DOE sites such as the PGDP released to the environment. Gu et al. (48) determined
in Paducah, Kentucky; the Portsmouth Gaseous Diffusion conclusively that the species of 99 Tc mobile in groundwater
Plant in Piketon, Ohio; Oak Ridge National Laboratory at the PGDP was the pertechnetate anion, TcO4 − . This
Y-12 Plant, Savannah River Laboratory in Aiken, South later was supported through the development of phase
Carolina; the Pacific Northwest National Laboratory 200 diagrams using site-specific Eh and pH groundwater
west area; and the Nevada Test Site (30,38–40). data (38,39). In addition, Gu and Dowlen (31) and Gu
Currently, the largest annual release of 99 Tc to the et al. (48) conclusively ruled out colloidal and natural
environment occurs at the two nuclear fuel-reprocessing organic matter facilitated transport as viable mechanisms
facilities in Sellafield, United Kingdom. Technetium-99 of 99 Tc mobilization. Previous studies of PGDP, other sites,
laden effluent water from these facilities is discharged and laboratory studies assumed that TcO4 − was the form
directly into the Irish Sea. of 99 Tc in most environmental settings. The estimated
Contamination has been detected as far west as partitioning coefficient of 99 Tc exceeded 27,000 milliliter
Svalbard and along the Norwegian coastline, distances per gram [mL/g] (48). Technetium-99 activities as high as
of several hundred miles. The total discharge from these 179 Bq/L (4800 pCi/L) were observed in the groundwater
two plants amounts to between 70 and 90 tera (TBq)/year plume (38,39). The estimated mass of 99 Tc dissolved in the
of 99 Tc (41). groundwater plume was 4 to 6 lb, which equates to an
Techentium-99 is reported in the Savannah River as activity of 1.0 to 1.6 × 1012 Bq/L (28 to 42 × 1012 pCi)
a result of discharges from the U. S. Department of of 99 Tc.
Energy Savannah River Plant located near Aiken, South At the Hanford Site in Washington, 99 Tc has been
Carolina (42,43). Technetium-99 is present at Oak Ridge found in groundwater in the Hanford 200 West Area
where spent nuclear fuel was reprocessed and then the BIBLIOGRAPHY
waste material was stored in underground tanks (30).
Technetium-99 levels between 900 and 27,000 pCi/L (36 1. Noddack, W. and Tacke, I. (1925). Die Ekamangane. Natur-
and 1080 Bq/L) have been found in groundwater. wissenschaften 13: 567.
Technetium-99 activity in shallow groundwater at the 2. Perrier, C. and Segre, E. (1937). Nature (London). 140: 193.
Portsmouth Gaseous Diffusion Plant has been observed 3. Seelmann-Eggebert, W., Pfennig, G., and Munzel, H. (1974).
at the 1000 pCi/L (40 Bq/L) level (49). The causes of Chart of the Nuclides. Gersbach Verlag, Munich, Germany.
contamination are similar to those discussed for the 4. Schwochau, K. (1981). The analytical chemistry of tech-
Paducah site. However, because the Portsmouth site netium. Top. Curr. Chem. 96: 109–147.
received feed material from the Paducah site, the quantity 5. Kenna, B.T. and Kuroda, P.K. (1961). Technetium 99. J.
of 99 Tc introduced into the site was much lower. Inorg. Nucl. Chem. Lett. 23: 142.
A unique study was conducted by the Lawrence 6. De Alleluia, T.B. (1983). Tc: Technetium, System No. 69. In:
Livermore National Laboratory at the Nevada Test site Gmelin Handbook of Chemistry. Supplement Vol. 1, 8th Edn.
to assess the distribution of radionuclides in caverns Springer-Verlag, Berlin.
produced by underground nuclear testing. The work by 7. Cartledge, G.H. (1973). Twenty-year inhibition of corrosion
Buddemeier et al. (50) and Schroeder et al. (40) found that by the pertechnetate ion. Corrosion 29: 361–362.
99
Tc was present in groundwater associated with two 8. Eriksen, T.E., Ndalamba, P., Bruno, J., and Caceci, M.
separate nuclear weapons tests. The 99 Tc appeared to (1992). The solubility of TcO2 ·nH2 O in neutral to alkaline
be associated and transported with colloidal material, but solutions under constant pCO2 . Radiochimica Acta 58/59:
its mobility was reduced upon contact with zones of high 67–70.
levels of iron oxyhydroxides. 9. Bruhl, H., Bock, W., Trapp, C., and Winkler, A. (1989). Aus-
There are several known nongovernmental sites breitungsverhalten von Technetium (Selen) in geochemisch
containing 99 Tc. One such facility is the Fields Brook unterschiedlichen Grundwasserleitern. Institut fur Ange-
National Priority Listed (NPL) site in Ashtabula, Ohio. andte Geologie Freie Unversitat Berlin. Final report 02 U
The site was formerly a storage site for liquid waste, 5695 A9.
which contained 99 Tc as well as other radionuclides and 10. Vandergraaf, T.T., Ticknor, K.V., and George, I.M. (1984).
solvents. Technetium-99 activities as high as 100,000 Reactions Between Technetium in Solution and Iron-
pCi/L (4000 Bq/L) have been detected in groundwater. Containing Minerals Under Oxic and Anoxic Conditions.
A second site is the uranium fuel plant in Hematite, Atomic Energy of Canada. Limited Report AECL-1957.
Montana. Recycled uranium, containing 99 Tc, was used to Pinawa, Manitoba, Canada, pp. 25–43.
fabricate metallic fuel rods. Low-levels of 99 Tc have been 11. Paquette, J., Reid, J.A.K., and Rosinger, E.L.J. (1980).
observed in groundwater. Review of Technetium Behavior in Relation to Nuclear Waste
Disposal. Whiteshell Nuclear Research Establishment. Tech-
nical Record TR-25. Pinawa, Manitoba, Canada.
REMEDIATION 12. Allard, B., Kigatsi, H., and Torstenfelt, B. (1979). Tech-
netium: Reduction and sorption in granitic bedrock.
The technologies explored for remediating 99 Tc in water Radiochem. Radioanalytical Lett. 37: 223–230.
include ion exchange, liquid–liquid extraction, precipi- 13. Gu, B. and Schulz, R.K. (1991). Anion Retention in Soil: Pos-
tation with various forms of iron, redox manipulation, sible Application to Reduce Migration of Buried Technetium
the FORAGER sponge, granular activated carbon (GAC), and Iodine. U.S. Nuclear Regulatory Commission, Washing-
and natural attenuation (10,16,18,31–33,48,51–65). The ton, DC.
report by Bostick et al. (53) provides an excellent reference 14. Meyer, R.E., Arnold, W.D., and Case, F.I. (1987). The Solu-
list of the many studies conducted on 99 Tc removal from bility of Electrodeposited Tc(IV) Oxides. Oak Ridge National
aqueous waste streams. At the Portsmouth and Paducah Laboratory, Oak Ridge, TN.
Gaseous Diffusion plants, ion exchange is being used for 15. Lieser, K.H. and Bauscher, C.H. (1987). Technetium in the
treating 99 Tc-contaminated groundwater. Although, Gu hydrosphere and in the geosphere 1. Chemistry of technetium
et al. (31) found that GAC was superior to ion exchange and iron in natural waters and influence of the redox
due to its low cost, efficiency, and sorption capacity, GAC potential on the sorption of technetium. Radiochimica Acta
has not been employed at these sites. 42: 205–213.
16. Lee, S.K. and Bondietti, E.A. (1983). Technetium behavior in
sulfide and ferrous iron solutions. Mater. Res. Soc. Symp.
CONCLUSION Proc. 15: 315–321.
17. Wildung, R.E., Routson, R.C., Serne, R.J., and Garland, T.R.
Technetium is an element not typically found in nature (1974). Pertechnetate, Iodide, and Methyl Iodide Retention
and has environmental consequence at facilities that by Surface Soils. In Pacific Northwest Laboratory Annual
enrich or reprocess uranium fissioned material. These Report for 1974 to the USAEC Division of Biomedical
facilities are typically government operated. The high and Environmental Research. B.E. Vaughan (Ed.). Pacific
solubility of TcO4 − and its stability across a wide spectrum Northwest National Laboratory, Richland, WA, pp. 37–40.
of redox conditions enables the ion to be extremely 18. Meyer, R.E., Arnold, W.D., and Case, F.I. (1984). Valence
mobile once introduced into water. As a consequence, at Effects on the Sorption of Nuclide on Rocks and Minerals.
the few sites where 99 Tc has been introduced into the Oak Ridge National Laboratory, Oak Ridge, TN.
natural environment, it has produced very large plumes 19. Rard, J.A. (1983). Critical Review of the Chemistry and
of contaminated surface and groundwater. Thermodynamics of Technetium and Some of its Inorganic
Compounds and Aqueous Species. Lawrence Livermore 39. Clausen, J.L., Richards, W.L., Korte, N.E., and Liang, L.
National Laboratory, Livermore, CA. (1995). ORNL/MMES research into remedial applications of
20. Bidoglio, G., Avogadro, A., and De Plano, A. (1985). Influence zero-valence metals, 3: Removal of TCE, cis-1,2-DCE, vinyl
of redox environments on the geochemical behavior of chloride and technetium. Am. Chem. Soc. Symp. Reductive
radionuclides. Mater. Res. Symp. Proc. 50: 709–716. Dehalogenation Proc., April 2–7, Anaheim, CA, pp. 755–758.
21. Bidoglio, G., Chatt, A., De Plano, A., and Zorn, F. (1983). 40. Schroeder, N.S., Morgan, D., Rokop, D.J., and Fabryka-
Investigation on technetium species in ground water and Martin, J. (1992). Migration of Technetium-99 in the Alluvial
migration through clay columns. J. Radioanalytical Chem. Aquifer at the Nevada Test Site. Los Alamos National
79: 153–164. Laboratory, LA-UR-92-463, Los Alamos, NM.
22. Pilkington, N.J. (1990). The solubility of technetium in the 41. Brown, J., Kolstad, A.K., Lind, B., Rudjord, A.L., and
near-field environment of a radioactive waste repository. J. Strand, P. (1998). Technetium-99 contamination in the North
Less-Common Met. 161: 203–212. Sea and in Norwegian Coastal areas 1996 and 1997.
23. Sheppard, S.C., Sheppard, M.I., and Evenden, W.G. (1990). A Stralevern Report. 1998:3. Oslo, Norway.
novel method used to examine variation in Tc sorption among 42. Turcotte, M.S. (1982). Environmental Behavior of Technetium-
34 soils, aerated and anoxic. J. Environ. Radioactivity 11: 99. Savannah River Plant and Laboratory Report DP-1644.
215–233. Aiken, SC.
24. Henrot, J., Garten, C.T., Jr., and Blaylock, B.G. (1989). 43. Anderson, T.J. (1979). Development of chemical isolation and
Behavior of Technetium in Soil: Sorption-Desorption Pro- concentration techniques for Tc-99 analysis by resin—bead
cesses. Oak Ridge National Laboratory, Oak Ridge, TN. mass spectrometry. In: Radioelement Analysis. Savannah
25. Schulte, E.H. and Scoppa, P. (1987). Sources and behavior River Laboratory, Aiken, SC.
of technetium in the environment. Sci. Total Environ. 64:
44. Olsen, C.R. et al. (1986). Geochemical and environmental
163–179.
processes affecting radionuclide migration from a formerly
26. Wolfrum, C. and Bunzl, K. (1986). Sorption and desorption used seepage trench. Geochim. Cosmochim. Acta 50: 593–607.
of technetium by humic substances under oxic and anoxic
45. Blaylock, B.G. and Frank, M.L. (1981). Bioaccumulation and
conditions. J. Radioanal. Nucl. Chem. 99: 315–323.
distribution of Tc-95m in an experimental freshwater pond.
27. Wildung, R.E., Garland, T.R., McFadden, K.M., and Cowan, In: Int. Symp. Migration Terrestrial Environ. Long-Lived
C.E. (1986). Technetium sorption in surface soils. In: Tech-
Radionuclides Nucl. Fuel Cycle. Savannah River Laboratory,
netium in the Environment. G. Desmet and C. Myttenaere
Aiken, SC.
(Eds.). Elsevier Applied Science, London, pp. 115–129.
46. Ballestra, S. et al. (1987). J. Radiochem. 115: 51.
28. Wildung, R.E., Garland, T.R., and Cataldo, D.A. (1977).
Accumulation of technetium by plants. Health Phys. 32: 47. Fryar, A. and Lasage, D.M. (2001). Technetium-99 as an
314–317. in-situ tracer of ground-water/stream interactions and
organic co-contaminant fate. Geol. Soc. Am. Annu. Meet.,
29. Bondietti, E.A. and Francis, C.W. (1979). Geologic migration
potentials of technetium-99 and neptunium-237. Science 203: November 5–8, Boston, MA.
1337–1340. 48. Gu, B., Clausen, J.L., McDonald, J.A., and McCarthy, J.F.
30. Brown, D.J. (1967). Migration Characteristics of Radionu- (1994). Assessment of the Influences of Groundwater Colloids
clides Through Sediments Underlying the Hanford Reserva- on the Migration of 99 Tc at the PGDP Site, Paducah, KY. Oak
tion. In: Disposal of Radioactive Wastes in to the Ground. Ridge National Laboratory, Oak Ridge, TN.
International Atomic Energy Agency Symposium. Vienna, 49. Muck, M.T., Korte, N.E., Liang, L., Marsh, D., Zutman, J.L.,
Austria. and Clausen, J.L. (1995). Technology Screening Report
31. Gu, B., Dowlen, K.E., and Clausen, J.L. (1996). Efficient for Tc Removal from Groundwater Portsmouth Uranium
separation and recovery of technetium-99 from contaminated Enrichment Plant, Piketon, Ohio. Martin Marietta Energy
groundwater. Efficient Sep. 6: 123–132. Systems, Inc., Piketon, OH.
32. Palmer, D.A. and Meyer, R.E. (1981). Adsorption of tech- 50. Buddemeier, R.W., Finkel, R.C., Marsh, K.V., Ruggieri, M.R.,
netium on selected inorganic ion-exchange materials and on a Rego, J.H., and Silva, R.J. (1992). Radionuclide migration
range of naturally occurring minerals under oxic conditions. at the Cheshire and Cambric event sites. In: Hydrology
J. Inorg. Nucl. Chem. 43: 2979–2984. and Radionuclide Migration Program 1989 Progress Report.
33. Strickert, R.G., Friedman, A.M., and Fried, S. (1980). The K.V. Marsh (Ed.). Lawrence Livermore National Laboratory,
sorption of technetium and iodine radioisotopes by various Albuquerque, NM, pp. 5–20.
minerals. Nucl. Technol. 49: 253–266. 51. Clausen, J.L., Sturchio, N.C., Heraty, L., Abrajano, T., and
34. Davenport, D. (1994). Interoffice Memorandum to L. Stinton. Huang, L. (1997). Evaluation of Natural Attenuation Pro-
Martine Marietta Energy Systems, Piketon, OH. cesses for Trichloroethene and Technetium-99 in the North-
35. Erhardt, K.C. and Attrep, M., Jr. (1978). Technetium-99 in west and Northeast Plumes at the Paducah Gaseous Diffusion
the atmosphere. Environ. Sci. Technol. 12: 55–57. Plant Paducah, Kentucky. Lockheed Martin Energy Systems,
36. Attrep, M., Jr., Enochs, J.A., and Broz, L.D. (1971). Atmo- Inc., Kevil, KY.
spheric technetium-99. Environ. Sci. Technol. 5: 344–345. 52. Clausen, J.L. and Richards, W.L. (1994). Treatability Studies
37. Leon, M. (1989). Determination and levels of 99 Tc in Using Iron Filings to Remediate Trichloroethene and
environmental and biological samples. In: Technetium in the Technetium. Martin Marietta Energy Systems. KY/ER-51.
Environment. G. Desmet and C. Myttenaere (Eds.). Elsevier Paducah, KY.
Applied Science, London, pp. 171–178. 53. Bostick, W.D., Osberne, P.E., Del Cul, G.D., and Sim-
38. Clausen, J.L., Zutman, J.L., Pickering, D.A., and Farrow, mons, D.W. (1995). Treatment of Aqueous Solutions Con-
N.D. (1995). Final Report on Drive-Point Profiling of the taminated with Technetium-99 Originating from Uranium
Northwest Plume and Analysis of Related Data. Lockheed Enrichment Activities: Final Report. Martin Marietta Energy
Martin Energy Systems, Inc., Kevil, KY. Systems, Inc., Oak Ridge, TN.
WATER—NATURE’S MAGICIAN 583
54. Ishanullah. (1994). Optimization of various factors for the

separation of technetium using anion-exchange resins. Sep.
Sci. Technol. 29: 239–247.
55. Del Cul, G.D., Bostick, W.D., Trotter, D.R., and Osberne, P.E.
(1993). Technetium-99 removal from process solutions and
contaminated groundwater. Sep. Sci. Technol. 28: 551–564.
56. Kawasaki, M., Amori, T., and Hasegawa, K. (1993). Adsorp-
tion of pertechnetate on an anion exchange resin. Radiochim-
ica Acta 63: 53–56.
57. Takeuchi, M., Tanaka, S., and Yamawaki, M. (1993). Solvent
extraction of Tc(vii) and Re(vii) from nitric acid solution
with the mixture of CMPO and TBP. Radiochimica Acta 63:
97–100.
58. Yamagishi, I. and Kubota, M. (1993). Recovery of technetium
with active carbon column in partitioning process of high-level
liquid waste. J. Nucl. Sci. Technol. 30: 717–719.
59. Karalova, Z.K., Lavrinovich, E.A., and Myasoedov, B.F.
(1991). Separation of technetium from accompanying
elements in basic solutions by extraction with
triphenyltetrazolium chloride. Sov. Radiochem. 33: 266–269.
60. Chen, Q.J., Dahlgaard, H., Hansen, H.J.M., and Aarkrog, A.
(1990). Determination of Tc-99 in environmental samples by
anion exchange and liquid extraction at controlled valency.
Anal. Chim. Acta 228: 163–167.
61. Emery, M.F. and Lim, C.K. (1989). Separation of cationic
technetium-99m amine complexes on porous graphiti carbon.
J. Chromatogr. 479: 212–215. • Canada has about 7% of the world’s renewable
62. Bostick, W.D. and Evans-Brown, B.S. (1988). Sorptive freshwater supply, compared with 18% for Brazil,
Removal of Technetium from Heavy Metals Sludge Filtrate 9% for China, and 8% for the United States.
Containing Nitrate Ion. Oak Ridge National Laboratory, Oak
Ridge, TN. • The ice we skate on in winter is water in its solid
form. Unlike most substances, which are densest in
63. Yamagishi, I. and Kubota, M. (1989). Separation of tech-
netium with active carbon. J. Nucl. Sci. Technol. 26: their solid state, ice is less dense than water and
1038–1044. thus floats. If this were not the case, lakes and rivers
64. Walton, F.B., Paquette, J., Ross, J.P.M., and Lawrence, W.E. would freeze from the bottom up. Fish could not
(1986). Tc(IV) and Tc(VII) interactions with oxyhydroxides. survive, and it is unlikely that rivers and lakes in
Nucl. Chem. Waste Manage. 6: 121–126. northern countries would ever completely thaw.
65. Fried, S., Friedman, A.M., Cohen, D., Hines, J.J., and Strick- • Water vapour forms a kind of global ‘‘blanket’’ which
ert, R.G. (1978). The Migration of Radioactive Processing helps to keep the earth warm. Heat radiated from
Wastes in Selected Rocks. Argonne National Laboratory, the sun-warmed surface of the earth is absorbed and
Argonne, IL. held by the vapour.
WATER’S MAGICAL PROPERTIES

WATER—NATURE’S MAGICIAN
Water molecules are attracted to each other, creating
hydrogen bonds. These strong bonds determine almost
Environment Canada
every physical property of water and many of its chemical
properties too.
Water is the solvent, the medium and the participant
in most of the chemical reactions occurring in our
Boiling and Freezing
environment.
Water is at once simple and complex. A water molecule Pure water at sea level boils at 100 ◦ C and freezes at
itself is simple, made up of three atoms: two hydrogen and 0 ◦ C. At higher elevations (lower atmospheric pressures)
one oxygen, H2 0. The configuration of these building blocks water’s boiling temperature decreases. This is why it takes
produces a molecule with almost magical properties. longer to boil an egg at higher altitudes. The temperature
does not get high enough to cook the egg properly. If a
WATER—VISIBLE AND INVISIBLE substance is dissolved in water, then the freezing point is
lowered. That is why we spread salt on streets in winter
On earth, water is found as a liquid, as a solid (ice) or as a to prevent ice formation.
gas (water vapour).
Thermal Properties
This article is produced by Environment Canada and is Water absorbs or releases more heat than many
reproduced with permission. substances for each degree of temperature increase or
584 WATER—NATURE’S MAGICIAN
decrease. Because of this, it is widely used for cooling

and for transferring heat in thermal and chemical
Did You Know
processes.
Differences in temperature between lakes and rivers • Raindrops are not tear-shaped. Scientists, using
and the surrounding air may have a variety of effects. For high-speed cameras, have discovered that rain-
example, local fog or mist is likely to occur if a lake cools the drops resemble the shape of a small ham-
surrounding air enough to cause saturation; consequently burger bun.
small water droplets are suspended in the air • About 70% of the human body is water.
Large bodies of water, such as the oceans or the Great • Life on earth probably originated in water.
Lakes, have a profound influence on climate. They are the
• More than half of the world’s animal and plant
world’s great heat reservoirs and heat exchangers and the
species live in the water.
source of much of the moisture that falls as rain and snow
over adjacent land masses. • Almost 75% of the earth is covered in water.
When water is colder than the air, precipitation is • The human body needs 2 liters of water a day
curbed, winds are reduced, and fog banks are formed. in our climate; we can last only a few days
without water.
• Most of our food is water: tomatoes (95%), spinach
Surface Tension
(91%), milk (90%), apples (85%), potatoes (80%),
Surface tension is a measure of the strength of the water’s beef (61%), hot dogs (56%).
surface film. The attraction between the water molecules
creates a strong film, which among other common liquids
is only surpassed by that of mercury. THE SUN-POWERED CYCLE
This surface tension permits water to hold up
substances heavier and denser than itself. A steel needle
carefully placed on the surface of a glass of water will The endless circulation of water from the atmosphere
float. Some aquatic insects such as the water strider rely to the earth and its return to the atmosphere through
on surface tension to walk on water. condensation, precipitation, evaporation and transpiration
Surface tension is essential for the transfer of energy is called the hydrologic cycle.
from wind to water to create waves. Waves are necessary Heating of the ocean water by the sun is the key
for rapid oxygen diffusion in lakes and seas. process that keeps the hydrologic cycle in motion. Water
evaporates, then falls as precipitation in the form of rain,
hail, snow, sleet, drizzle or fog. On its way to earth some
Molecules in Motion precipitation may evaporate or, when it falls over land, be
intercepted by vegetation before reaching the ground. The
Water molecules as well as binding to each other,
cycle continues in three different ways:
bind to many other substances such as glass, cotton,
plant tissues, and soils. This is called adhesion. For
example, in a thin glass tube, when the molecules • Evaporation/Transpiration— On average, as much
at the edge reach for and adhere to the molecules as 40% of precipitation in Canada is evaporated or
of glass just above them, they at the same time tow transpired.
other water molecules along with them. The water • Percolation Into the Ground— Water moves down-
surface, in turn, pulls the entire body of water to a ward through cracks and pores in soil and rocks to
new level until the downward force of gravity is too the water table. Water can move back up by capil-
great to be overcome. This process is called capillary lary action or it can move vertically or horizontally
action.
under the earth’s surface until it re-enters a surface
Thus water readily wets many materials. Capillary water system.
action allows a paper towel or a sponge to be used to soak
up spilled water. Without this property, the nutrients • Surface Runoff— Water runs overland into nearby
needed by plants and trees would remain in the soil. streams and lakes; the steeper the land and the less
porous the soil, the greater the runoff. Overland flow
is particularly visible in urban areas. Rivers join
The Universal Solvent each other and eventually form one major river that
carries all of the subbasins’ runoff into the ocean.
An extraordinary property of water is its ability to dissolve
other substances. There is hardly a substance known
which has not been identified in solution in the earth’s Although the hydrologic cycle balances what goes up with
waters. Were it not for the solvent property of water, life what comes down, one phase of the cycle is ‘‘frozen’’ in
could not exist because water transfers nutrients vital to the colder regions during the winter season. During the
life in animals and plants. Canadian winter, for example, most of the precipitation is
A drop of rain water falling through the air dissolves simply stored as snow or ice on the ground. Later, during
atmospheric gases. When rain reaches the earth, it affects the spring melt, huge quantities of water are released
the quality of the land, lakes and rivers. quickly, which results in heavy spring runoff and flooding.
FREEZING AND SUPERCOOLING OF WATER 585
FRESHWATER SERIES A-1 Water samples often contain other substances that
organize water molecules in an ice-like pattern and that
cause freezing to be initiated at a higher temperature than
Note: A resource guide, entitled Let’s Not Take Water For
would otherwise be the case. Such substances are known as
Granted, is available to help classroom teachers of grades
ice nucleators, and freezing catalyzed by an ice nucleator is
5–7 use the information from the Water Fact Sheets.
referred to as heterogeneous freezing. Ice nucleator activity
is seen in various types of crystals and dust particles,
FREEZING AND SUPERCOOLING OF WATER and biological organisms may synthesize nucleating
lipoproteins, which are physiologically important as ice
KARL ERIK ZACHARIASSEN nucleators, and which may have a high nucleating activity.
Norwegian University of Certain bacteria produce ice nucleators that can nucleate
Science and Technology ice formations at above −1 ◦ C.
Trondheim, Norway
FACTORS AFFECTING FREEZING

Water in the crystalline solid state is known as ice, and
for pure water at normal atmospheric pressure, ice is the The equilibrium freezing point is influenced by factors
physically stable form at temperatures below 0 ◦ C. such as solute concentration (Fig. 1) and hydrostatic
Ice has a specific gravity of 0.92 g/ml, which is about 8% pressure (Fig. 2). The osmolal melting point depression is
below the corresponding value of water, reflecting the fact 1.86 ◦ C/Osm, whereas pressure depresses melting points
that water expands by about 8% when it enters the solid by about 1 ◦ C/100 bar.
state. The specific heat capacity of ice is 0.502 cal/g.deg, The effect of these parameters on supercooling points
which is about half the corresponding value of water. The depends on whether nucleation is homogeneous or
amount of heat liberated during the solidification (heat of heterogeneous. Homogeneous nucleation temperature is
fusion) is 79.71 cal/g. depressed 1.5–3 times more than melting points by
The transition of water from the liquid to the solid increased osmolality and pressure, whereas, at least
state is known as freezing, and pure water has a so- for the more active nucleators, nucleation temperature
called freezing point of 0 ◦ C, which does not imply that depression seems to be strictly equivalent to the
pure water will freeze when cooled to this temperature. A corresponding melting point depression.
small sample of pure water can be cooled to as much as Nucleation temperature of purified water samples (no
−40 ◦ C before it freezes spontaneously. For this reason, the potent ice nucleators) also depends on sample volume
term freezing point is misleading, and it is often replaced as shown by Bigg in 1953 (1) (Fig. 3). When potent
by the more precise terms melting point or equilibrium ice nucleators are present, the nucleation temperature
freezing point. depends only on the activity of the most active nucleator,
A substance that exists in the liquid state at a and it does not matter if this nucleator occurs in a large or
temperature below the equilibrium freezing point is said to small water volume.
be supercooled or undercooled, and the temperature where
freezing is initiated is referred to as the supercooling point
or the temperature of crystallization. 0
Melting point
When a sample of supercooled water freezes, the
released heat of fusion will immediately heat the system −5
to the melting point. As a result of the high heat of fusion
of water, the melting point temperature will be reached
−10
following the freezing of only a small fraction of the sample. Nucleator-induced
From this point, the freezing can proceed only as fast as the supercooling point
−15
Temperature (°C)
released heat of fusion can be removed from the system.

When pure water freezes, the temperature will remain at
0 ◦ C until all water is frozen. −20
Initiation of freezing requires that the water sample
contains an ice-like structure, which can act as a
−25
nucleus for ice formation, from which ice can grow.
Liquid water is assumed to contain aggregates of water
molecules organized in an ice-like pattern, but which −30
usually are too unstable to act as nucleators. As the Supercooling point without
potent nucleators
temperature drops, these aggregates become more stable −35
and increase in size. On sufficient cooling, one of these
aggregates will eventually reach the size critical for
ice nucleation, and freezing will be initiated from this −40
0 500 1000 1500 2000
aggregate. When nucleation is initiated in this manner
Osmolality (mOsm)
from the water molecules themselves, it is known as
homogeneous nucleation. Figure 1.
586 CHEMICAL PRECIPITATION
0 Melting point
−2 BRADLEY A. STRIEBIG
Temperature (°C)
−4 Gonzaga University
−6 Spokane, Washington
−8
−10
−12 Chemical precipitation is the removal of dissolved and
−14 Nucleation temperature of
suspended solids from aqueous solutions through chemical
−16 different ice nucleators means. Chemical precipitation includes the removal of
−18 dissolved solids by altering the aqueous chemistry to
0 100 200 300 400 500 600 700 800 900 100011001200 make the dissolved solids less soluble. It also includes
Pressure (bar) removing suspended solids by increasing their size in a
process referred to as chemical flocculation. The larger the
Figure 2.
particle size of the suspended solids, the easier it is to
remove them, usually through gravimetric settling in a
0 process called clarification.
−5
USES OF CHEMICAL PRECIPITATION
Nucleation temperature (°C)
−10
Chemical precipitation processes are utilized in water
−15
Curve of Bigg (1953) and wastewater treatment due to their effectiveness for
−20 pollutant removal with relatively low capital investment.
The chemical reaction rates are typically very fast
−25
compared to biological reactions, so residence times in
−30 treatment processes or the reactor size may be significantly
smaller than in other treatment techniques. Chemical
−35
precipitation processes can also process very toxic
−40 wastewaters. However, chemical precipitation processes
often require the on-going addition of chemicals to
−45 control the pH and polymers to increase the size of
−9 −8 −7 −6 −5 −4 −3 −2 −1 0 1 2 3 4 5
Log microlitre water volume
the particles. The residuals collected from chemical
precipitation processes may also be quite toxic and
Figure 3. expensive to dispose of as sludge. Chemical precipitation
is an effective tool for water treatment, and the most
common uses are briefly described.
FREEZING OF SOLUTIONS
Industrial wastewaters often require pretreatment
When a solution freezes, the solutes are not incorporated before being released to the sewer or receiving waters.
in the ice structure, but remain in the liquid fraction High strength industrial wastewaters, such as those used
surrounding the ice. Here, they become concentrated in electroplating, contain high concentrations of dissolved
precisely until the solution reaches a melting point equal metals. These metals can be removed from the wastewater
to the actual temperature. In this situation, vapor pressure by altering the pH of the water to reduce the solubility of
equilibrium will exist between the solution and the ice, and the metal ions. The insoluble fractions of the metal ions
no further freezing will occur until the released heat of form solid particles or precipitates, which are removed
fusion is removed from the system. As the heat is removed through filtration or gravimetric clarification. This allows
and the freezing proceeds, a gradual reduction of the industrial wastewaters to significantly reduce their metals
temperature of the system will occur as the fluid fraction concentrations in order to meet RCRA standards and also
becomes more concentrated and the melting temperature reduce the toxicity of the wastewater.
of the fluid fraction drops. The solute concentration will Municipal wastewaters may contain dissolved ions and
increase until the eutectic point is reached. At the eutectic a large amount of cell mass from biological treatment
point, water and solutes freeze together. processes. The focus of chemical precipitation on municipal
wastewaters is to remove the cell mass or suspended solids
prior to and after biological treatment. The suspended
BIBLIOGRAPHY
solids are removed by adding chemical polymers, which
1. Bigg, E.K. (1953). The supercooling of water. Proc. Roy. Soc. attach to the existing suspended solids and help combine
London. B 66: 688–694. the individual particles into larger particles called flocs.
This process, called coagulation, is accomplished by the
READING LIST chemical destabilization of particles and the subsequent
aggregation. The process of gentle mixing and increasing
Zachariassen, K.E. and Kristiansen, E. (2000). Ice nucleation and the size of the suspended solids particles is called
antinucleation in nature. A review. Cryobiology 41: 257–279. flocculation. The large floc particles are removed through
CHEMICAL PRECIPITATION 587
gravimetric settling in a process called clarification. commonly estimated from measurements of the total
Thus a large fraction of the suspended solids and dissolved solids (TDS) and the electrical conductivity:
also the remaining BOD (in the form of cell mass)
Langelier approximation I = 2.5 × 10−5 (TDS)[mg/L]
can be removed with the aid of chemical precipitation.
Russell approximation I = 1.6 × 10−5 (conductivity)
Chemical precipitation is also commonly used when
[µ mho]
phosphorus removal is required at municipal wastewater
treatment plants. One estimate of the activity coefficient is the Debye–
Industrial process water and recycled wastewater use Hückel limiting law for solutions with ionic strength less
chemical precipitation to remove hardness and dissolved than 0.005 mol/L (I < 0.005 M):
salts from water. This allows a higher quality of water
needed for specific industrial processes. The limits of γA ∼
= −Az2 I1/2
hardness and salt removal are governed by the solubility
where A is a function of the dielectric constant of the
of the dissolved ions in solution. Therefore, chemical
solution. For water at 25 ◦ C, A = 0.51. The term z is the
precipitation has limited uses in industries that require
ionic charge of the ion of interest. For more concentrated
highly purified water. However, chemical precipitation can
solutions (I > 0.005 M), other estimates for the activity
be used as an inexpensive pretreatment step prior to ion
coefficient, such as the extended Debye–Hückel model
exchange or reverse osmosis processes, which produce very
and the Davies approximation can be used.
high-quality water.
The potential to precipitate multiple ions from solution
can be approximated from the activity of the ions in
DISSOLVED SOLIDS solution and the solubility product. Often, the pH of
the solution is used as a ‘‘master variable.’’ Solids lines
Dissolved solids are defined as the solid residuals in water
on a graph of this type indicate the solids equilibrium
that pass through a filter with a nominal pore size of
concentration. For most hydroxide species, a minimum
2.0 µm. These solids include dissolved ions and small
soluble concentration appears between a pH of 8 and 12.
colloidal solids. The dissolved ions in solution can be
Similar graphical representations can be made for sulfide
predicted based on the solubility product constant for
and carbonate species. This minimum solubility can serve
the dissolved species. The solubility product constant,
as a starting point for estimating and optimizing chemical
or solubility product, is an equilibrium constant that
precipitation of various ions in solution.
describes the reaction by which ions dissolve and solids
The removal of hexavalent chromium (Cr6+ ) from solu-
precipitate:
tion is one illustration of the potential of chemical pre-
Az By(s) −−
−−
y+
−− zA + yB
z−
cipitation processes to reduce the toxicity of wastewater.
For ions A and B, the generic equilibrium constant is Hexavalent chromium can be converted to the less toxic
trivalent form by reaction at low pH with a reducing agent
{Ay+ }z {Bz− }y such as ferrous bisulfite. Figure 1 shows a solution con-
Ks =
{Az By }(s) taining hexavalent chromium, which when dissolved in
water results in a yellow color. Figure 2 shows the change
Many salts will form and precipitate as either hydroxides, in color associated with the change in the oxidation state of
with the hydroxide ion (OH− ), or carbonates, with the the chromium from the 6+ state to the 3+ state. When the
carbonate ion (CO3 2− ). The concentration of these ions pH of the solution is raised to 8 or 9, the trivalent chrome
can often be predicted through an equilibrium analysis becomes much less soluble in solution than the hexavalent
with the aid of electron or pH diagrams. Since the activity form of chromium as shown in Fig. 3. The resulting solids
of the solid substance is defined as 1, and the solubility precipitate out of solution to form a concentrated sludge
product is a constant value, the activity of the anion of as shown in Fig. 4.
interest (A) can be predicted in the aqueous solution of
interest. Solubility constants for a wide array of dissolved SUSPENDED SOLIDS
ions of interest in water and wastewater are found in the
referenced literature (1–3). Suspended solids are composed of floating matter,
In a dilute solution under ideal conditions the activity settleable matter, and large colloidal matter. This matter
of the ions of interest can be closely approximated by their
concentration:
{A} ≈ [A]
However, many wastewaters have high concentrations

of dissolved ions and do not behave ideally. An activity
coefficient, γA , must be used to establish an accurate
relationship between the activity of a compound in solution
and the concentration:
{A} ≈ γA [A]
The activity coefficient can be estimated from the ionic

strength of the solution. The ionic strength is also Figure 1. Hexavalent chromium dissolved in water.
Chemical precipitation or coagulation of the suspended

particles involves the destabilization of colloidal particles.
Coagulation is usually followed by flocculation, which is
the addition of a chemical, typically an organic polymer,
to increase particle aggregation.
Suspended solids in water often have a charge that acts
at the surface of the particle. Oil droplets, gas bubbles,
and other inert substances may have a negative charge
Figure 2. Reduction of hexavalent chromium to trivalent due to the preferential absorption of anions. Cellular
chromium. materials such as microorganisms and proteins develop
a charge through the ionization of carboxyl and amino
acids functional groups. These charges create a force or
zeta potential around the particles, which prevent them
from increasing in size through aggregation as illustrated
in Fig. 5. Coagulants are added to overcome these surface
forces by the addition of potential determining ions or
electrolytes. Once these charges have been neutralized,
flocculants can be added to increase the speed of
aggregation (Fig. 6).
Flocculant polymers that are anionic and nonionic
become attached to the particles. These long particles
increase the rate of aggregation and become intertwined
Particle
Particle
Repulsive
forces
Figure 3. Formation of insoluble chromium oxide salts.
Particle
Figure 5. Repulsive forces acting on the surface of sus-

pended particles.
Particle
Particle
Slow mixing
Figure 4. Precipitation of solid insoluble chromium into a brings particles
sludge blanket. Coagulants overcome into contact
repulsive forces with one another
may be organic or inorganic in nature. In municipal

wastewater treatment plants, a significant fraction of
the suspended solids consist of biological cell mass. The Particle
amount of suspended solids in water is an important
parameter in determining water quality. Suspended solids
are typically removed through a process of coagulation, Figure 6. Chemical coagulants overcome repulsive forces and
flocculation, and gravimetric settling (or clarification). the solution is mixed.
ANTIMONY IN AQUATIC SYSTEMS 589
samples. Although antimony was already known to the

ancients, it is still often overlooked, both as an element of
environmental concern and as a subject for study, probably
because of its lower abundance and the relative insolubility
Particle
of most of its compounds. This is reflected in the poor
Particle standard of existing data on the behavior of antimony
in natural waters. However, interest in the study of this
element seems to be growing and an increasing number of
papers on the subject are being published. A useful series
of comprehensive reviews on antimony in the environment
has recently been published (1,2).
Antimony has no known biological function and, like
Flocculants arsenic, is toxic. Trivalent species are reported to be more
increase toxic than pentavalent forms (3). The U.S. Environmental
Particle
the rate of Protection Agency (EPA) (4) and the Commission of the
aggregation European Communities (5) both consider antimony and
and particle stability Floc its compounds to be pollutants of prime interest. The U.S.
EPA drinking water standards are: MCLG (maximum
Figure 7. Flocculants increase the rate of aggregation to form
contaminant level goal) and MCL (maximum contaminant
large flocs.
level), both 6 µg/L (6). The European Union’s maximum
admissible concentration of antimony in drinking water is
during flocculation. This results in a significant increase 5 µg/L (7). Antimony is on the list of hazardous substances
in the particle size and mass (Fig. 7). The large particles under the Basel convention on the restriction of the
called flocs are stabilized by the polymers, which prevent transfer of hazardous waste across borders (8).
them from breaking up during the mixing process. The
large flocs can then be removed with gravimetric settling ABUNDANCE
in the clarification process. The resultant solids collected
in the clarifier are the wastewater solids, or sludge. The relative abundance of antimony in different systems
Chemical precipitation by coagulation and flocculation is shown in Table 1. According to Goldschmidt’s classifica-
has been used in the past to enhance removal of suspended tion, antimony is a strong chalcophile element of which the
solids and BOD. These processes are widely used to most common natural form is the sulfide mineral stibnite,
improve the performance of primary clarifiers, in basic Sb2 S3 . Antimony is occasionally found as a native metal or
physical–chemical processing of industrial wastewaters, as an intermediate compound with arsenic. More than 113
for phosphorus removal, and for heavy metal removal. different sulfide, antimonide, arsenide, and telluride min-
erals have been identified. Sulfosalts represent the most
BIBLIOGRAPHY common antimony-bearing minerals, usually in some com-
bination with silver, copper, arsenic, mercury, and lead.
1. Tchobanoglous, G., Burton, F.L., and Stensel, H.D. (2003).
More than 35 oxide, hydroxide, and oxyhalide mineral
Wastewater Engineering Treatment and Reuse, 4th Edn.
Metcalf and Eddy, Inc/McGraw Hill, Boston, pp. 42–52, species are known. Antimony is also found in seven sili-
475–517. cate minerals, all of which are rare (9). Antimony is also a
2. Stumm, W. and Morgan, J.J. (1996). Aquatic Chemistry common component of coal and petroleum.
Chemical Equilibria and Rates in Natural Waters, 3rd Edn. Antimony is present in the aquatic environment as a
John Wiley & Sons, Hoboken, NJ, pp. 252–398. result of rock weathering, soil runoff, and anthropogenic
3. Jensen, J.N. (2003). A Problem-Solving Approach to Aquatic activities. Typically, concentrations of dissolved antimony
Chemistry. John Wiley & Sons, Hoboken, NJ, pp. 418–450. in unpolluted waters are less than 1 µg/L. The mean
antimony concentration in surface marine waters is 184 ±
45 ng/L (1). However, in the proximity of anthropogenic
ANTIMONY IN AQUATIC SYSTEMS sources, concentrations can reach up to 100 times
natural levels.
MONTSERRAT FILELLA Airborne supply to aquatic and terrestrial systems
University of Geneva is important for the environmental fate of antimony
Geneva, Switzerland in systems located far from direct sources of pollution.
Atmospheric emission values for antimony are shown in
Table 2. The values in this table should be viewed with
INTRODUCTION some caution because the data used comes from relatively
old studies. For instance, antimony production was
Antimony is a naturally occurring element. It belongs to considered to be only 80,000 metric tons/yr (142,000 metric
group 15 of the periodic table of the elements. Antimony tons in 2003) and, more importantly, the uses of antimony
can exist in a variety of oxidation states (−III, 0, III, have changed significantly over the past few years. Pacyna
V) but it is mainly found in two oxidation states (III and Pacyna (15) estimated natural and anthropogenic
and V) in environmental, biological, and geochemical emissions into the atmosphere to be 2400 tons/yr and
590 ANTIMONY IN AQUATIC SYSTEMS
Table 1. Antimony Geochemical element have exceeded natural fluxes for more than 2000
Abundance years. The present day enrichment factor (relative to the
Material Sb Abundancea element/Sc ratios of typical crust rocks) is on the order of
70 times (as compared to 20 for As and 130 for Pb) (18).
Cosmic abundance 0.246b The chronology and intensity of the antimony enrichments
Chondrites 0.1
in peat are similar in many ways to those of lead, which
Mean crustal average 0.2
Upper continental crust 0.31 indicates that the extent of the human impact on the
Lower continental crust 0.30 geochemical cycle of antimony is comparable to that of
Terrestrial abundance 0.7 lead (16).
Basic rocks (basalt) 0.15 The production and use of antimony have steadily
Granitic rocks 0.2 increased. World reserves of antimony are in excess of
Sedimentary rocks 2 million tons and are located principally in China,
Shales 1–2 South Africa, Bolivia, Russia, and Tajikistan (19). The
Carbonates 0.2
world’s mines produced 142,000 tons of antimony in
Deep sea clays 1
2003 (19). Table 3 shows the main uses of primary
a
All values in ppm (mg/g), except for the antimony in the United States in 2003, which can be
cosmic abundance. seen as representative of the use of antimony across the
b
Atomic abundance relative to Si (=1,000,000)
deduced from the following mean composition
developed world. Uses of antimony have changed over
of meteorites: silicate = 100, sulfide = 7, metal = the years. Traditionally, it was used in lead–antimony
10.6 parts per weight. alloys. Bulk-secondary antimony could be recovered as
Source: Adapted from Reference 1. antimonial lead, most of which was regenerated and then
consumed by the battery industry. Today, the main use of
antimony is as Sb2 O3 in flame-retardants, which have
1600 tons/yr, respectively. Very recently, Shotyk and applications in, among others, children’s clothing and
co-workers (16) dismissed all previous values of natural toys and aircraft and automobile seat covers. Antimony
antimony releases as being grossly overestimated. Based trioxide cannot be recycled and is released into the
on antimony concentrations in ombrotrophic peat from environment.
Switzerland and Scotland, these authors suggested that
90–154 metric tons/yr would be more appropriate values.
USES Table 3. Reported Industrial

Consumption (20) of Primary Antimony in
The first recorded use of antimony as a pigment to make the United States in 2002 (Metric Tons of
mascara is found in an Egyptian papyrus dating from 1600 Antimony Content)a
B.C. According to Emsley (17), the first uses of antimony Uses Sb Amountb
date back even further; part of a 5000-year-old vase now
Metal products
in the Louvre is made of almost pure antimony. A peat
Antimonial lead 910
core from an ombrotrophic Swiss bog revealed significant
Bearing metal and bearings 43
enrichments of antimony dating back to Roman times, Solder 85
which indicates that the anthropogenic fluxes of this Ammunition, cable 1,370
covering, castings, sheet
and pipe, and type metal
Table 2. Global Emissionsa of Antimony into the Total 2,410
Atmosphere, 103 metric tons/yr Nonmetal products
Ammunition primers W
Sb Anthropogenic Sb
Ceramics and glass 487
Natural Sources Emission Sources Emission
Pigments 597
Windborne dust 0.8 Energy production 1.3 Plastics 532
(includes industrial (coal, oil, gas) Fireworks and rubber products 488
sources of dust) Total 2,100
Sea salt spray 0.6 Mining 0.1 Flameretardants
Volcanic activity 0.7 Smelting and 1.4 Adhesives W
refining Plastics 3,680
Forest fires 0.2 Waste incineration 0.7 Rubber 126
Biogenic sources 0.3 Textiles 150
Total 2.6 Total 3.5 Paper and pigments 767
Total 4,720
a
Adapted from Reference 1. Original estimations come from References Grand Total 9,230
10–13. A completely different estimation, which is often cited in the
a
literature, was published by Lantzy and Mackenzie in 1979 (14). They Secondary antimonial lead was 5600 metric tons in
estimated natural and anthropogenic emissions of antimony into the 2003.
atmosphere to be 0.98 × 103 metric tons/yr and 38 × 103 metric tons/yr, b
W = Withheld to avoid disclosing company propri-
respectively. etary data.
ANTIMONY SOLUTION CHEMISTRY 1.3
Both Sb(III) and Sb(V) ions hydrolyze easily in aqueous

solution, thus making it difficult to keep antimony ions 1.0
O
stable in solution except in highly acidic media. Since most 2 (g
)
of the techniques used for the study of solution equilibria
(solubility measurements, potentiometry, voltammetry, Sb(OH)2+
spectroscopy) necessitate the use of macroamounts of
reactants, relatively little information is available on 0.5
the chemical species of antimony that exist in aqueous Sb(OH)6−
solution, and on their stability constants in particular.
Eh (V)
Moreover, most of the published studies are relatively old.
Only the system antimony–sulfur has been the object of Sb(OH)4−
recent studies, mainly in relation to the understanding of 0.0 Sb(OH)3
Sb2S3(s)
hydrothermal systems.
The following main points can be made regard-
ing the antimony inorganic chemistry in solution (2):
(1) According to thermodynamic data, antimony should H (SbS2)−
2 (g
)
exist as Sb(V) in oxic systems and as Sb(III) in anoxic −0.5
ones. (2) At environmental concentration levels, at the
pH range commonly found in natural waters and in the
absence of sulfur, antimony exists as a soluble species
regardless of its oxidation state. Antimony(V) is present
as a negative species in solution [Sb(OH)6 − , also repre- −0.9
0 2 4 6 8 10 12
sented as SbO3 − ] and Sb(III) as a neutral species [Sb(OH)3 ,
pH
also often quoted as HSbO2 ]. (3) It is not clear whether
Sb(V) and Sb(III) form any chloride-containing species in Figure 1. Eh–pH diagram of antimony in the Sb–S–H2 O
seawater, but hydroxychloro species would be the most system at a dissolved antimony concentration of 10−8 mol/L and
a dissolved sulfur concentration of 10−3 mol/L. Adapted from
probable entities. (4) Antimony oxides exhibit low water
Reference 2.
solubility, which may affect the fate of antimony in soils
and water, particularly since today antimony is mostly
used as Sb2 O3 (s) in flame-retardant applications. More- oxidize the trivalent antimony to pentavalent antimony.
over, it is thought that, when released into the atmosphere Several studies have found a correlation between the
as an aerosol, antimony is oxidized to antimony trioxide adsorption of antimony in soil and oxic sediments and
through reactions with atmospheric oxidants. (5) In anoxic the presence of manganese and iron oxides (22,23).
systems, and in the presence of sulfur, antimony forms
insoluble stibnite, Sb2 S3 (s), and soluble SbS2 2− , depending ANALYTICAL METHODS
on the pH. The formation of polymerized species at natural
concentration levels (e.g., Sb2 S4 2− , Sb2 S5 4− , Sb4 Sb7 2− ) is The total concentrations of antimony in aqueous and
not proved. The formation of an Sb(V)-containing species solid environmental samples can readily be determined
(SbS4 3− ) has been observed by EXAFS. using current analytical techniques, primarily the result
Eh–pH diagrams can illustrate the redox behavior of of the development of microwave digestion techniques
elements like antimony. Figure 1 shows one such diagram for solid samples and NAA and plasma-based analyte
for the system Sb–S–H2 O at environmentally realistic detection systems. Both ICP-AES and ICP-MS techniques
concentrations of antimony and dissolved sulfur. are in widespread use, although ICP-MS has less spectral
A comprehensive, thermodynamically consistent anti- interference and a lower detection limit (ng/L). An
mony database of equilibrium constants of antimony with excellent critical review of the methodologies currently
inorganic ligands has recently been published (21). available for the environmental analysis of antimony
Barely any data have been reported on Sb(V) binding has recently been published (24). However, determining
by low molecular weight organic ligands. More extensive antimony speciation still presents a significant analytical
studies have been carried out on the binding of Sb(III), challenge (25). Recent data are almost exclusively based
which behaves like a borderline metal. The extent of on hydride generation methods and often only oxidation
antimony binding by natural organic matter (NOM) is state data are obtained. A lack of stability in the antimony
a point of some debate. A few studies report significant species being investigated, particularly in trivalent forms,
Sb–NOM interactions, but most do not, and observed throughout the entire analytical process may result in an
environmental behavior seems to suggest that Sb–NOM overestimation of pentavalent species concentrations.
interactions are not relevant in most aquatic systems. Determining methylated antimony species is hampered
Hydrous oxides of manganese and iron have been by severe molecular rearrangements originating from the
reported to significantly sorb Sb(III) but very few demethylation of trimethylantimony during the analysis.
systematic studies have been carried out on antimony This problem has been the subject of extensive literary
sorption by pure metal oxides. These oxides will also discussion but a solution has yet to be found (26).
ANTIMONY CONCENTRATIONS IN THE DIFFERENT fresh waters, groundwaters, and rain waters. This con-
AQUATIC COMPARTMENTS trasts with thermodynamic equilibrium predictions, which
suggest that Sb(V) concentration should exceed Sb(III)
The distribution and behavior of antimony in freshwater by several orders of magnitude. For thermodynamically
systems have not been studied extensively. Published unstable species to be present in water, mechanisms for
concentrations range from a few nanograms/liter to a few their production and slow rates of interconversion must
micrograms/liter depending on location. They reflect the exist. The presence of Sb(III) has often been attributed
wide range of physical and chemical conditions existing in to phytoplankton activity. However, although possible, a
freshwater bodies and are very sensitive to the proximity detailed analysis of published data has so far failed to pro-
of sources of pollution. duce conclusive proof to validate this hypothesis. Recently,
The concentration of antimony in ocean surface waters the photochemical reduction of Sb(V) to Sb(III) has been
is about 200 ng/L. The behavior of antimony in the open suggested as an alternative to the phytoplankton source.
seas is not considered to be highly reactive. According The true speciation of antimony under anoxic conditions
to some authors, it behaves as a conservative element remains unclear. According to thermodynamic calcula-
and to others as a mildly scavenged element with surface tions, in the absence of oxygen, antimony should only be
(atmospheric) input. The comparison between antimony present in the trivalent form. However, the occurrence of
contents in deep waters of different oceans indicates pentavalent forms of antimony has been reported in dif-
that no antimony accumulation is produced during deep- ferent anoxic systems. A variety of mechanisms have been
water oceanic circulation (27). Antimony concentrations suggested to explain these results, including delivery of
seem to reflect regional geochemical characteristics in Sb(V) on sinking detritus from oxic waters, formation of
coastal environments. Antimony behavior in estuaries thiocomplexes by the pentavalent element, and advection
varies somewhat according to the characteristics of the of surface waters containing high concentrations of anti-
estuary. In some cases, it behaves conservatively and in monate. All these mechanisms must be accompanied by
others it shows a midestuary maximum. Behavior has relatively slow rates of reduction.
even been seen to vary in the same estuary from one year
to another. Methylated Species
Antimony concentrations in sediments are on the
order of a few micrograms/gram of sediment and rarely Methylated species have been detected in surface sea-
exceed 1 mg/g. Higher concentrations are directly related waters, in estuaries, and in some rivers. They usually
to anthropogenic sources, mainly proximity to smelting account for 10% or less of the total dissolved antimony. The
plants and outfalls of sewage and fertilizer facilities. monomethyl species is more abundant than the dimethyl
form and close to the surface values tend to be higher.
Monomethyl, dimethyl, trimethyl, and triethyl antimony
ANTIMONY SPECIATION IN NATURAL WATERS derivatives have been detected in sediments from rivers
and harbors. However, the identification of methylated
Today it is widely accepted that an understanding antimony species has been questioned because some of
of biogeochemical processes requires knowledge of the these compounds undergo rearrangement reactions when
chemical forms, or species, that are present in the natural subjected to the hydride generation procedures typically
environment. Despite this, the speciation of antimony in used for aquatic speciation.
natural aquatic systems has yet to form the subject of The organoantimony compounds detected by Dodd
many in-depth studies. et al. (26) in a freshwater plant (Potamogetan pectinatus)
and a stibnolipid isolated from a marine diatom by Ben-
Dissolved Versus Particulate Fractions son (29) are, as yet, the only organoantimony compounds
Weak interaction between anionic species and suspended with unequivocal biological origin to be detected in aquatic
particles negatively charged in natural waters is to be environments.
expected and it is generally accepted that antimony is
present as ‘‘dissolved’’ in aquatic systems. Experimental KINETIC ISSUES
data (2) exist to support this. Fractionation of the so-called
dissolved phase into several classes by ultrafiltration
As mentioned, the biogeochemical cycle of antimony is
showed that most of the dissolved antimony is present
more complicated than equilibrium thermodynamic calcu-
in the smallest molecular weight region (2,28).
lations predict (1). Several authors have cited the kinetic
stabilization of thermodynamically unstable species to
Redox
explain their presence in natural systems. Several pro-
Antimony occurs in two oxidation states in natural waters cesses may influence oxidation and reduction rates, includ-
and, thus, its behavior and toxicity can be affected by ing chemical and photochemical transformations and bio-
changes in the redox status of the aquatic environment. logical mediated reactions. It is well known that, when a
Antimony speciation studies have focused mainly on the redox process is accompanied by hydrolysis reactions, as
separation and identification of Sb(III) and Sb(V) species. is likely to be the case for antimony, the overall process
Antimony (V) is the predominant species found in oxy- can be much slower than expected. At present, kinetic
genated systems but thermodynamically unstable Sb(III) information for redox reactions of antimony in natural
has also been detected in different oxic marine water, waters is limited and rate constants are, for the most
part, unknown. Cutter estimated an overall Sb(III) oxida- Chlorella vulgaris, isolated from an arsenic-polluted envi-
tion rate of 0.008 d−1 in the 65–85 m layer of the Black ronment, excreted 40% Sb(V) and 60% Sb(III) on exposure
Sea from depth profiles (30). However, antimony residence to Sb(III), suggesting that a change in oxidation state is
times in seawater are probably longer because this rate used by this alga as a detoxifying mechanism (38).
included all forms of removal; that is, Sb(III) oxidation and
also Sb(III) scavenging by sedimenting particles. More- BIBLIOGRAPHY
over, it was calculated at the oxic–suboxic interface of
the Black Sea, where the presence of manganese and iron 1. Filella, M., Belzile, N., and Chen, Y-W. (2002). Antimony in
oxides is likely to increase the Sb(III) oxidation rate (31). the environment: a review focused on natural waters. I.
Indeed, Belzile and co-workers (32), in a study on the oxi- Occurrence. Earth Sci. Rev. 57: 125–168.
dation of Sb(III) in the presence of natural and synthetic 2. Filella, M., Belzile, N., and Chen, Y-W. (2002). Antimony in
iron and manganese oxyhydroxides, showed that Sb(III) the environment: a review focused on natural waters. II.
Relevant solution chemistry. Earth Sci. Rev. 59: 265–285.
can be rapidly oxidized to Sb(V) by both compounds. Iron
3. Bencze, K. (1994). Antimony. In: Handbook on Metals in
oxides also proved to play an important role in antimony
Clinical and Analytical Chemistry. H.G. Seiler, A. Sigel, and
redox chemistry in lake sediments (23). However, iron
H. Sigel (Eds.). Marcel Dekker, New York, pp. 227–236.
and manganese oxides are not likely to be present in
4. United States Environmental Protection Agency. (1979).
oxic waters in significant quantities and thus are unlikely Water Related Fate of the 129 Priority Pollutants. Vol. 1,
to influence antimony chemistry in oxic systems. It has EP-440/4-79-029A, U.S. EPA, Washington, DC.
recently been shown that Sb(III) oxidation by hydrogen 5. Council of the European Communities. (1976). Council
peroxide and iodate is strongly pH dependent: the neutral Directive 76/464/EEC of 4 May 1976 on pollution caused by
Sb(OH)3 species is unreactive to hydrogen peroxide and certain dangerous substances discharged into the aquatic
the presence of Sb(OH)4 − is needed for the oxidation to environment of the Community. Official Journal L 129,
occur (33,34). 18/05/1976, pp. 23–29.
The addition of organic substances such as tartaric, 6. United States Environmental Protection Agency. (1999).
lactic, citric, and ascorbic acids to natural or synthetic National Primary Drinking Water Standards. Doc. 810-F-
solutions has been shown to have a stabilizing effect 94-001, U.S. EPA, Washington, DC.
on Sb(III). Solutions of Sb(III) prepared from potassium 7. Council of the European Union. (1998). Council Directive
antimonyl tartrate have been reported to remain stable 98/83/EC of 3 November 1998 on the quality of water intended
for long periods. The organic matter present in natural for human consumption. Official Journal L 330, 05/12/1998,
pp. 32–54.
waters probably has a similar stabilizing effect. Ligand
stabilization has been used to explain the apparent 8. United Nations Environmental Program. (1999). Basel
Convention on the Control of Transboundary Movements of
presence of Fe(II) in oxic solutions that contain relatively
Hazardous Wastes and Their Disposal (with amended Annex
high levels of organic matter. I and two additional Annexes VIII and IX, adopted at the
Little information is available on Sb(V) reduction fourth meeting of the Conference of the Parties in 1998), SBC
kinetics in anoxic media. The reduction rate for Sb(V) in No. 99/001. UNEP, Geneva.
water at the bottom of the Black Sea was estimated to be 9. Lueth, V.W. (1999). Antimony: element and geochem-
1.1 × 10−6 d−1 (35). istry. In: Encyclopedia of Geochemistry. C.P. Marshall and
R.W. Fairbridge (Eds.). Kluwer Academic Publishers, Dor-
drecht, The Netherlands, pp. 15–16.
ANTIMONY AND AQUATIC BIOTA 10. Nriagu, J.O. and Pacyna, J.M. (1988). Quantitative assess-
ment of worldwide contamination of air, water and soils by
Reported concentrations for antimony in fresh water and trace metals. Nature 333: 134–139.
marine algae range from 0.02 to 1 µg/g dry weight (1). 11. Nriagu, J.O. (1989). A global assessment of natural sources
Although algae bioaccumulation and detoxifying mecha- of atmospheric trace metals. Nature 338: 47–49.
nisms may play a key role in antimony redox speciation 12. Nriagu, J.O. (1990). Global metal pollution. Poisoning the
in natural waters, few studies have been carried out. biosphere? Environment 32: 7–11, 28–33.
Kantin (36) collected three marine algae near San Diego 13. Maeda, S. (1994). Safety and environmental effects. In:
Bay in the United States and Sb(V) was the dominant The Chemistry of Organic Arsenic, Antimony and Bismuth
Compounds. S. Patai (Ed.). Wiley, Hoboken, NJ, pp. 725–759.
species in all of them. Only Sargassum sp. contained
up to 30% of Sb(III), thus demonstrating an ability to 14. Lantzy, R.J. and Mackenzie, F.T. (1979). Atmospheric trace
metals: global cycles and assessment of man’s impact.
form reduced antimony species. Andreae and Froelich (37)
Geochim. Cosmochim. Acta 43: 511–525.
analyzed the chemical speciation of antimony in phyto-
15. Pacyna, J.M. and Pacyna, E.G. (2001). An assessment of
plankton (largely diatoms) from the eastern North Pacific
global and regional emissions of trace metals to the
and found that the trivalent species formed a significant atmosphere from anthropogenic sources worldwide. Environ.
part of the inorganic antimony. Both studies have since Rev. 9: 169–298.
been used to justify the frequently cited biological origin 16. Shotyk, W. et al. (1996). Two thousand years of atmospheric
of the Sb(III) found in oxic waters. Although no detailed arsenic, antimony, and lead deposition recorded in an
studies exist, antimonate is expected to behave differently ombrotrophic peat bog profile, Jura Mountains, Switzerland.
in bioavailability than arsenate and phosphate because Earth Planet. Sci. Lett. 145: E1–E7.
of the weaker Lewis acidity and larger ionic radius of 17. Emsley, J. (2001). Nature’s Building Blocks. An A–Z Guide
the antimonate ion in solution. More recently, the algae to the Elements. Oxford University Press, New York.
18. Shotyk, W., Krachler, M., and Chen, B. (2004). Antimony in flow-field-flow fractionation; multi-element determination of
recent, ombrotrophic peat from Switzerland and Scotland: colloidal size distributions in a natural water sample. Anal.
comparison with natural background values (5,320 to Chim. Acta. in press.
8,02014 C yr BP) and implications for the global Sb cycle. 29. Benson, A. (1988). Antimony metabolites in marine algae. In:
Global Biogeochem. Cycles 18: GB1016. The Biological Alkylation of Heavy Metals. P.J. Craig and
19. Carlin, J.F., Jr. (2004). Antimony. In: U.S. Geological Survey F. Glocking (Eds.). Special Publication, Vol. 66, Royal Society
Mineral Commodity Summaries. of Chemistry, London, pp. 135–137.
20. Carlin, J.F. (2003). U.S. Geological Survey Minerals Year- 30. Cutter, G.A. (1992). Kinetic controls on metalloid speciation
book. in seawater. Mar. Chem. 40: 65–80.
21. Filella, M. and May, P.M. (2003). Computer simulation of 31. Cutter, L.S., Cutter, G.A., and San Diego-McGlone, M.L.C.
the low-molecular-weight inorganic species distribution of (1991). Simultaneous determination of inorganic arsenic
antimony(III) and antimony(V) in natural waters. Geochim. and antimony species in natural waters using selective
Cosmochim. Acta 67: 4013–4031. hydride generation with gas chromatography/photoionization
22. Müller, B. et al. (2002). As, Sb, Mo, and other elements in detection. Anal. Chem. 63: 1138–1142.
sedimentary Fe/Mn layers of Lake Baikal. Environ. Sci. 32. Belzile, N., Chen, Y-W., and Wang, Z. (2001). Oxidation of
Technol. 36: 411–420. antimony(III) by amorphous iron and manganese oxyhydrox-
23. Chen, Y-W., Deng, T-L., Filella, M., and Belzile, N. (2003). ides. Chem. Geol. 174: 379–387.
Distribution and early diagenesis of antimony species in 33. Quentel, F., Filella, M., Elleouet, C., and Madec, C-L. (2004).
sediments and porewaters of freshwater lakes. Environ. Sci. Kinetic studies on Sb(III) oxidation by hydrogen peroxide
Technol. 37: 1163–1168. in natural waters. Environ. Sci. Technol. 38: 2843–
24. Nash, M.J., Maskall, J.E., and Hill, S.J. (2000). Methodolo- 2848.
gies for determination of antimony in terrestrial environmen- 34. Quentel, F., Filella, M., Elleouet, C., and Madec, C-L. (2004).
tal samples. J. Environ. Monit. 2: 97–109. Sb(III) oxidation by iodate in seawater: a cautionary tale.
25. Krachler, M., Emons, H., and Zheng, J. (2001). Speciation of (Submitted).
antimony for the 21st century: promises and pitfalls. Trends 35. Cutter, G.A. (1991). Dissolved arsenic and antimony in the
Anal. Chem. 20: 79–90. Black Sea. Deep-Sea Res. 38: S825–S843.
26. Dodd, X. et al. (1996). Antimony speciation in freshwater 36. Kantin, R. (1983). Chemical speciation of antimony in marine
plant extracts by using hydride generation–gas chromatog- algae. Limnol. Oceanogr. 28: 165–168.
raphy–mass spectrometry. Analyst. 121: 223–228. 37. Andreae, M.O. and Froelich, P.N. (1984). Arsenic, antimony,
27. Cutter, G.A. and Cutter, L.S. (1998). Metalloids in the high and germanium biogeochemistry in the Baltic Sea. Tellus 36:
latitude North Atlantic Ocean: sources and internal cycling. 101–117.
Mar. Chem. 61: 25–36. 38. Maeda, S. et al. (1997). Bioaccumulation of antimony by
28. Stolpe, X., Hassellöv, M., Andersson, K., and Turner, D. Chlorella vulgaris and the association mode of antimony
(2004). High resolution ICPMS as an on-line detector for in the cell. Appl. Organomet. Chem. 11: 393–396.

The Mirage of The H Economy: Previous Page

Uploaded by

Copyright:

Available Formats

The Mirage of The H Economy: Previous Page

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Mirage of The H Economy: Previous Page

Uploaded by

Copyright:

Available Formats

Previous Page

THE MIRAGE OF THE H2 ECONOMY 477

Table 1. Notable Characteristics of the Hydrogen Ion, H+a

Type Cation + 8H+ + 4SO4 2−

≡ FeOH + Cu+ = ≡ FeOCu + H+ HCl(g) → H+ (aq) + Cl− (aq) (1)

the prototype of all acid–base chemistry in an aque-

H2 O(l) + CH3 COOH(aq) ⇔ H3 O+ (aq) + CH3 COO− (aq)

Water Capillary rise

H O Figure 1. Unsaturated zone definition and active processes (1).

To characterize soil moisture and infiltration, the

Figure 3. Capillarity and adsorption combine to produce suction Sandy soil

H Empirical. Empirical infiltration models generally uti-

5 where S is a parameter called sorptivity, t is time

properties. In this model, the infiltration rate approaches a

0 Infiltration rates can be measured at a point using

Table 2. Properties of Adsorbents Employed for Inorganic Ion Removal (2–6)

Surface area (m2 /g) 1100–1300 575 1142 1689 700

91.8 xxx xxx

Pb(II) Adsorption adsorbent with respect to one adsorbate. This phenomenon

20 Hg(II) Table 3. Thermodynamic Parameters for the Mercury

293 60 −987.714 +25708.82 +91.11444

Table 4. Selective Removal of Heavy Metal Ions by Ion Exchangers

Pb (II) Leawtit TP 207 (24) Chelating ion exchanger with

Table 5. Summary Table of Maximum Reported Adsorption Capacities (mg/g)

Bark (41–44) 32 19.45 400 182

16 recommended dietary allowance for iron in healthy

10 issue. It can impart an unpleasant metallic taste to

As a consequence, the sample may develop a reddish- Fe2+ + CO3 2− 

Figure 2. pE–pH diagram for the iron–

The presence of large concentrations of dissolved iron is

GHOLAM A. KAZEMI (Ion/Cl)sample

(note that the temperature is in K). S

h Figure 2. Mariotte bottle used to keep a constant water level

a then it can be used to determine how much water has been

1.4 analysis (13). It is generally accepted that sorption of an

angstrom and the diameter of the hydrated metal ions is

that adsorption of benazolin on F400 was through the 25

Copper uptake, mmol/g

20 carbon had a low oxygen content of 2.72% compared

1. Environment Agency. (2002). The Annual Report of the NETWORK BEHAVIORS

Hydrogen atom BIOLOGICAL STRUCTURE AND FUNCTION

water’s covalent bonds vibrate at about 1014 hertz (one- H

IN SITU CHEMICAL MONITORING is to continuously renew the reagent at the sensor

NITROGEN to be mineralized in the soil by a chain of microbial

ammonium ions. For example, at 15 ◦ C, ammonia gas Analytical Terminology

Nitrification AQUATIC NITROGEN

be life threatening for children less than 3 years old.

UNITS FOR OSMOTIC CONCENTRATIONS As soon as a chemical is spilled it becomes a pollutant,

degradability in the various environmental compartments. n = number of moles of the chemical

Direct contact Public area

Vapours Vapours Vapours

BIOAVAILABILITY direct or indirect and biological or chemical (14), although,

Figure 2. (a) Schematic of a genetically modified

PHYSICAL PROPERTIES number is used in calculating friction losses, pipe sizing,

Table 1. Physical Properties of Water in U.S. Customary Unitsa

32 62.42 1.940 287 3.746 1.931 0.00518 0.09

Table 2. Physical Properties of Water in SI Unitsa

0 9.805 999.8 1.98 1.781 1.785 0.0765 0.61

As a consequence, the sample may develop a reddish- Fe2+ + CO3 2−

Gf 0 (kJ·mol−1 ) 276b 13b