I N N 0 V A T I 0 N S A N D I D E A S

A SIMPLE STUDENT LABORATORY ON OSMOTIC FLOW,

OSMOTIC PRESSURE, AND THE REFLECTION COEFFICIENT

Joseph J. Feher and George D. Ford

Department of Physiology, Medical Co Zlege of Virginia, Richmond, Virgina 23298-0551

Downloaded from http://advan.physiology.org/ by 10.220.33.3 on May 28, 2017

Key words: artificial membranes; dialysis; sucrose; polyethylene glycol

The primary observations of osmosiswere originally filtration flow proportional to the pressure. Finally,
reported by Pfeffer in 1877 (5). Pfeffer made an he found that a closed cup containing a sucrose
artificial membrane in the walls of an unglazed solution would develop a pressure proportional to
porcelain cup by reacting copper salts with potas- the sucrose concentration. He defined the osmotic
sium ferrocyanide. He placed a sucrose solution pressure as the pressure applied to the solution side
inside the cup and water outside, and he found that that is necessary to stop the osmotic flow. This
water moved from the water side to the sucrose definition is restricted to osmosis across an ideal,
side. He then made three important observations. semipermeable membrane that allows water to pass
First, he found that the rate of flow was proportional but is impermeant to the solute. The pressure
to the sucrose concentration. Second, he observed defined in this way holds only at equilibrium, that is,
that a pressure applied inside the cup produced a when there is no net fluid movement. The magni-

1043 - 4046 / 95 - $3.00 - COPYRIGHT o 1995 THE AMERICAN PHYSIOLOGICAL SOCIETY

VOLUME 13 : NUMBER I -ADVANCES IN PHYSIOLOGY EDUCATION -JUNE 1995

SlO
 I N N 0 V A T I 0 N S A N D I D E A S

tude of the pressure is then given by van? Hoff’s law across membranes. It is the flow that is physiologi-
(9) cally important and results in cell swelling or shrink-
ing. The swelling or shrinking requires fluid move-
T = RTC (1) ment, which occurs at some quantitative rate. These
where ?r is the osmotic pressure, R is the gas concepts are not easy to convey and are often
constant, T is the absolute temperature, and C is the omitted from the usual laboratory exercises. At-
concentration of impermeant solute. tempts to teach osmosis range from using animated
films (2) to computer simulations (7) and weighing
Pfeffer’s result, that the flow is linearly related to the potato cubes with time (1). A recent laboratory
concentration, allows us to write the equation manual describes a simple “osmometer,” consisting
of a one-hole stopper inserted into dialysis tubing

Downloaded from http://advan.physiology.org/ by 10.220.33.3 on May 28, 2017

Q = -AL,AT
V (2) and secured with string (11). The open end of the
tubing is also tied, and then a l-ml pipette is
where Qv is the flow of volume across the membrane
inserted through the hole in the stopper. The
(in units of cm”/min), A is the area of the membrane
dialysis tubing thus forms a compartment that is
(in cm2), and L, is variously called the filtration
open to the air only through the l-ml pipette. The
coefficient, hydraulic conductivity, or hydraulic per-
dialysis tubing is then placed in a l-liter beaker of
meability. The minus sign in this equation indicates
water, and the dialysis tubing is filled with either 0,
that the water flow is from the region of low osmotic
15, or 30% sucrose solutions, and the position of
pressure (water, T = 0) to the region of high
fluid rise in the pipette is recorded with time. The
osmotic pressure (sucrose, r = RTC). In the absence
students then plot the volume of fluid movement
of solute, Pfeffer found that the filtration flow is
with time and obtain the flow rates. The results
proportional to the hydrostatic pressure (AP)
should confirm that the flow is twice as great with
Qv =A&AP (3 30% sucrose aswith 15% sucrose, in agreement with
van? Hoff’s law.
Results of experiments (4) and theoretical results
from irreversible thermodynamics show that the L, A variant of this laboratory exercise was the starting
in Eq. 2 is identical to the L, in Eq. 3. The net flow in point for the development of the exercise described
the presence of both a pressure and osmotic gradi- here. We found that it was difficult to seal the
ent is dialysis tubing successfully, the results were not
reproducible from group to group, and flow was
Q =AI,,(AP
V
- AT) (4)
often nonlinear because of the compliance of the
dialysis tubing. Because our goal was to offer an
In the case of nonideal membranes, which are exercise investigating osmotic flow, osmotic pres-
permeable to solute, the observed osmotic pressure sure, and the reflection coefficient, we concluded
is reduced because of the ability of the solute to that it was necessary to prevent the hydrostatic
penetrate the membrane, albeit less effectively than pressure developed by a vertical column of fluid
water. This has led to the notion of the reflection (thereby eliminating the problem of compliance of
coefficient (a), originally defined as the ratio of the the dialysis tubing compartment) and to accurately
observed osmotic pressure to the pressure calcu- control the surface area of the exposed dialysis
lated from the van? Hoff law for an ideal membrane. membrane. Thus the laboratory exercise described
In the case of the nonideal membrane, the flow herein was developed to give students a hands-on
equation is appreciation for the origin of osmotic pressure and
flow, and an understanding of the importance of the
Q =AL,(AP - aAn)
V (-5)
interaction of the solute and membrane in produc-
ing these quantities. In addition, the exercise was
From the foregoing discussion, it is clear that the meant to encourage a simple quantitative analysis of
osmotic pressure is an important concept that de- laboratory data that could be used to obtain mean-
scribes only one of the forces driving fluid flow ingful parameters describing osmosis, including the

VOLUME 13 : NUMBER I - ADVANCES IN PHYSIOLOGY EDUCATION - JUNE 1995

Sll
 I N N 0 V A T I 0 N S A N D I D E A S

hydraulic conductivity and the reflection coefficient by simply tying it off, but then it would be difficult to
for particular solute-membrane pairs. produce a known area of membrane. Therefore the
membrane was sealed at both the top and bottom by
an 0 ring and clamp, where the 0 rings were seated
THE FLOW APPARATUS
in a machined groove. To save time in the labora-
The main problems in accurately measuring flow tory, the membranes were installed beforehand by
across membranes are to maintain a steady hydro- the instructor. Spectrapor membranes, 29 mm in
static pressure, expose a reproducible membrane diameter and of 3,500 and 1,000 mol wt cutoff
area to the solutions, and prevent leaks in the (MWCO), were used in all of the exercises. A hollow,
membrane. The device (Figs. l-3) surmounted truncated Delrin cone was used to assistin installing
these difficulties. Here a known area of dialysis the membranes and 0 rings (Fig. 2) and excess

Downloaded from http://advan.physiology.org/ by 10.220.33.3 on May 28, 2017

tubing separates a solution inside the tube from membrane was cut off. To test for leaks, the inner
pure water outside. The membrane could be sealed chamber was filled with water, and then a 60-ml

horizontal flow tube assembly fillingsyringe

-&.:.: magneticstir bar

membraneclamp

magneticstirrer

A B
FIG. 1.
Apparatus for measuring osmotic flow. The horizontal flow tube was calibrated in centimeters and was supported
in a level position by the lip of the beaker and a clamp on a ring stand. The membranes used were Spectrapor 1,000
or 3,500 mol wt cutoff (MWCO). The membranes were sealed using an 0 ring and a metal circular clamp at the top
and bottom of the tube. The 0 rings were seated in grooves machined in both the clamp and the plastic supports,
thereby providing a well-defined surface area for the membrane. Leaks were tested (A) by applying a pressure
using a 60-ml syringe. When no leaks were detected, the syringe was removed and replaced with the horizontal
flow tube. The water solution in the inner compartment was removed and replaced with test solution (B). After
filling, the filling syringe was removed, and the position of the solution in the horizontal tube was recorded at
regular time intervals (C).

VOLUME 13 : NL MBER I - ADVANCES IN PHYSIOLOGY EDUCATION - JUNE 1995

s12
 I N N 0 V A T I 0 N S A N D I D E A S

ASSEMBLER

TOP TOP TOP

$ DIA.

$ DIA. 0

DIA.

Downloaded from http://advan.physiology.org/ by 10.220.33.3 on May 28, 2017

FRONT SIDE FRONT

FRONT
FIG. 2.
Dimensioned drawings of the body, ring clamps, and assembler. The ends of the body were
fabricated from methyl methacrylate plastic. A, inside the s/W-diameter hole was an 0 ring,
PRP-568-012,3/S” ID x l/2“ OD x l/16” cross section. B, semicircular groove for 0 ring was
l/64“ deep and l/16” wide; 0 ring was PRP-568-018, 3/4” ID x 7/S” OD x l/16“ cross
section. C, 30” relief angle. D, rods connecting the ends were l/S” stainless steel. Ring clamps
were made from aluminum. 0 ring shown with the ring clamp is the PRP-568-018 0 ring as
described in B. The groove in the ring clamp to receive the 0 ring was 0.095“ wide and 0.035”
deep. All edges on the ring clamp were rounded to avoid cutting the dialysis tubing. Screws for
ring clamp assembly were 4-40 socket head cap, l/2“ long. The assembler was constructed
from acetal plastic (Delrin). The 30” angle and dimensions of the assembler were designed to
fit over either end of the body to aid in placing the dialysis tubing over the body.

Luer lock plastic syringe was fitted to the opening moved and replaced with test solution, as shown in
for the flow tube (Fig. lA). This opening was sealed Fig. 1B. The test solution was added until it began to
with an 0 ring and fit the 60-ml syringe snugly. flow down the horizontal tube; the fill syringe was
Leaks were tested by adding a large pressure via the then removed and flow measurements were begun.
syringe. About one in four membranes leaked and During measurement of flow, the inlet aperture was
were replaced. In our experience, the leaks were left open to the atmosphere. Thus the pressure on
due not to unsuccessful sealsaround the 0 rings but the fluid in the horizontal calibrated tube was
to scratching the membrane and producing a hole maintained at atmospheric pressure while fluid
during assembly. After testing for leaks, the horizon- flowed down the tube. The tube was narrow enough
tal flow tube assembly was attached (Fig. IA and Fig. so that the moving water column had a vertical
3). The water in the inner compartment was re- meniscus that could easily be read on the scale.

VOLUME 13 : NUMBER 1 - ADVANCES IN PHYSIOLOGY EDUCATION - JUNE 1995

s13
 I N N 0 V A T I 0 N S A N D I D E A S

TOP

-
I/ 8

Downloaded from http://advan.physiology.org/ by 10.220.33.3 on May 28, 2017

FRONT SIDE
FIG. 3.
Dimensioned drawing of horizontal flow tube assembly. The right-angle
block was constructed of methyl methacrylate plastic. The scale on the
horizontal cylinder was made using Scala& scales (catalog no. 13375,
Bel-Art products, Pequannock, NJ).

Students were instructed to fill the dialysis sac until raised if desired (8). Stock solutions of 250 ml were
there was some fluid already in the tube, and to made, and - 55 ml filled the devices.
avoid air bubbles anywhere in the flow path because
these would interfere with the flow measurements.
DETERMINING FLOW RATES
Although flow began almost instantly, the students
were instructed to wait for up to 5 min before Students were assigned to one of four groups, each
beginning their measurements. of which had a corresponding set of test solutions
and a single test membrane. Each group was in-
THE SOLUTIONS structed to measure the position of the meniscus at
The solutes used in these experiments were urea, 5-min intervals and to record the result in tables
sucrose, and polyethylene glycol (PEG). Both urea provided in their handout. The students also wrote
and sucrose are readily available and solutions are their observations on the chalk board as they pro-
straightforward, but a note of caution is needed for gressed so that the data would be immediately
the PEG. PEG, with an average molecular weight of available to all students. Students plotted the dis-
3,350, was obtained from Sigma Chemical (cata- tance the meniscus traveled against the elapsed
logue no. P-3640). However, this is a weight-average time. The results of duplicate laboratory sessions
molecular weight. Osmotic experiments depend on performed in 1993 and 1994 for the l,OOO-MWCO
the number-average molecular weight, which is membrane are shown in Fig. 4. The flow was
smaller than the weight-averaged molecular weight routinely linear with time and was reproducible
for a polydisperse material. A number-average mo- from group to group and from year to year. Al-
lecular weight of 2,400 was used for this material. It though average numbers are used here, the data are
is probably unnecessary to explain this complication sufficiently reliable that single runs will provide
to students, but a discussion of the issue could be similarly quantitative and identically qualitative re-

VOLUME 13 : NUMBER I - ADVANCES IN PHYSIOLOGY EDUCATION - JUNE 1995

s14
 I N N 0 V A T I 0 N S A N D I D E A S

TABLE 1
Osmosis across a model membrane: flow calculation

Membrane A: Membrane B:
n 18 1,000 Mwco 3,500 Mwco
E Solution
2 16 Ax/& Q”9 Ax/& Qvy
0.075M PEG cm/min cm3/min cm/min cm3/min
30 14
/ Sucrose
2 12 I 0.25 M 0.462 0.0416 0.400 0.0360
0.50 M 0.930 0.0837 0.832 0.0749
0.75 M 1.328 0.1195 1.224 0.1102
/ Urea, 1 M 0.024 0.0022 0.024

Downloaded from http://advan.physiology.org/ by 10.220.33.3 on May 28, 2017

0.0022
I 0.0375M PEG PEG
0.0375 M 0.308 0.0277 0.374 0.0337
a
0.075 M 0.810 0.0729 0.748 0.0673
/ /
MWCO, molecular weight cutoff; PEG, polyethylene glycol. The
position of meniscus was plotted vs. time to obtain velocity of
1M UREA column movement or slope (Ax/At), which was multiplied by
-+ cross-sectional area (0.09 cm*) to obtain flow (QJ. Data were
calculated from averages of 4 separate determinations obtained
by students.

TIME (minutes)
FIG. 4. THINKING INTERLUDE
Position of meniscus in the flow tube related to
elapsed time. Position at time 0 was subtracted so that At this point, the students can be asked several
all flows would begin on the origin. Flow is shown for questions:
0.75 M sucrose (a), 0.50 M sucrose (A), 0.25 M sucrose
(0), 0.075 M polyethylene glycol (PEG, A), 0.0375 M
PEG (a), and 1 M urea (+). The membrane was 1) Is there a relationship between the osmotic flow
Spectrapor 1,000 MWCO. Values are means +: SE of 4 and the concentration of sucrose? If so, write an
separate determinations. equation for this relationship.

2) Is there a relationship between the osmotic flow

sults. The data in Fig. 4 show that the flow increased and the concentration of PEG? If so, write an
with concentration for a given solute. Similar results equation for this relationship.
were obtained for the 3,500-MWCO membrane.
3) Is there a relationship between the osmotic flow
The slope of the lines in Fig. 4, in units of cm/min, and the concentration of urea? If so, write an
can be converted to the flow (in units of cm”/min) equation for this relationship.
by multiplying by the cross-sectional area of the tube
(0.09 cm2). The students were provided with the 4) Are the relations between flow and concentration
cross-sectional area and instructed to calculate the different for sucrose, PEG, and urea? Why?
flows, record them in a table (Table l), and plot
the flow against the concentration. The results for In this section, the instructors helped the students
the l,OOO-MWCO membrane are shown in Fig. 5. realize that their observations were consistent with
The data clearly show the linear relationship be- the relation
tween flow and concentration (in mol/l) and that
this relationship was different for urea, sucrose, and Q = -h$pmAc
V (61
PEG. Similar qualitative results were obtained for where Qv is the flow (measured in cm”/min), A is the
the 3,500-MWCO membrane. area (in cm2), L, is the hydraulic conductivity or

VOLUME 13 : NUMBER 1 - ADVANCES IN PHYSIOLOGY EDUCATION - JUNE 1995

s15
 I N N 0 V A T I 0 N S A N D I D E A S

.13

.I2
r outlet tube was replaced
and the membrane
with a four-way
was mechanically
with stainless steel mesh. This allowed
stopcock,
reinforced
the develop-
ment of a significant hydrostatic pressure difference
.I1 across the membrane. Water was infused using a
Harvard syringe pump at a constant and known rate.
.I0 The pressure in the inside compartment rose, driv-
ing fluid out through the dialysis membrane. Eventu-
.09
ally a steady state was reached in which the rate of
infusion of water into the inner compartment was
.08
-n exactly equal to the rate of water exit through the

Downloaded from http://advan.physiology.org/ by 10.220.33.3 on May 28, 2017

dialysis membrane. This occurred at some pressure
difference between the inner and outer compart-
ments. This pressure was measured by a pressure
transducer connected to a Grass polygraph. The
plot of flow against pressure for the 3,500-MWCO
iz and l,OOO-MWCO Spectrapor membranes is shown
.04 in Fig. 7. The slopes of the lines give A X L, (in units
of cm3 . mmHg- l. min- l) . The area of the membrane
.03 was 90.5 cm2, but this variable always appeared in
conjunction with L,, and the product A X L, was the
.02
parameter of interest.
.Ol
CALCULATION OF o, THE REFLECTION
0 COEFFICIENT
0 0.2 0.4 0.6 0.8 The students were asked to calculate the reflection
CONCENTRATION (M) coefficients for urea, sucrose, and PEG for both the
FIG. 5. l,OOO-MWCO and 3,500-MWCO membranes by use
Osmotic flow in cd/min as a function of the concen- of the flow rates they had calculated and shown in
tration of sucrose (0) or PEG (A) for the 1,000~MWCO Table 1 and the value ofA x L, provided to them in
Spectrapor membrane. the handout. The results of the calculations (Table
2) allow several conclusions. First, the reflection
filtration coefficient, (T is the reflection coefficient, R coefficient, 0, appears to be independent of the
is the gas constant, T is the temperature, and AC is concentration of solute. Second, u depends on the
the concentration difference across the membrane. solute, being smallest for urea, larger for sucrose,
To calculate 0 from data of flow and RTAC, it was
and larger still for PEG. Third, u depends on the
necessary to first determine A x L,.
membrane, being smaller in every case (urea, su-
crose, and PEG) for the 3,500-MWCO membrane
DETERMINATION OF A x Lp
than for the l,OOO-MWCO membrane. These obser-
The pressure-driven flow of water across the model vations make it clear that the phenomenon of
membrane was determined using the device shown osmotic flow depends on the characteristics of the
in Fig. 6. The entire device was submerged in water, membrane and of the solute.
so water was in both the inner and outer compart-
ments and there was no osmotic pressure differ- The mechanistic interpretation of (T depends on the
ence. In this case, the net water flow is given byI@. 3 kind of membrane involved in the osmosis. The
(3). The device used to measure A X L, was identical microporous membrane is perhaps easier for stu-
to that described earlier in Fig. 1, except that the dents to visualize. In this model, the reflection

VOLUME 13 : NUMBER I - ADVANCES IN PHYSIOLOGY EDUCATION - JUNE 1995

s16
 I N N 0 V A T I 0 N S A N D I D E A S

from syringe pump

d-

Downloaded from http://advan.physiology.org/ by 10.220.33.3 on May 28, 2017

four-way stopcock

“0” ring

/
to pressure transducer

/
spring clamps -----+ membrane

\ stainless steel mesh

\ /
FIG. 6.
Apparatus for measuring the area x filtration coefficient (A x Lp) for the
dialysis membranes. Apparatus was identical to flow device except that the
horizontal flow tube assembly was replaced with a four-way stopcock to
allow flow into the inner chamber via a syringe pump and simultaneous
measurement of pressure by a pressure transducer. Pressure increased
until a steady-state value was reached in which influx of water via the
syringe pump was exactly balanced by pressure-driven flow of water
across the dialysis membrane into the surrounding solution.

coefficient is viewed as being due to a steric hin- case, 0 < (T < 1.0, and the observed osmotic
drance of solute entry into a cylindrical pore ori- pressure is diminished. Various equations have been
ented perpendicular to the surface of the mem- derived to relate (Tto the geometric size and shape
brane. Because a solute molecule is larger than the of hypothetical pores within membranes (3,6, 10).
pore, then the solute cannot enter the pore, and all
collisions of the solute with the membrane result in
DEMONSTRATION OF THE MAGNITUDE OF
reflection of the solute molecule. The membrane is
OSMOTIC PRESSURE
impermeant, and u = 1 .O for this membrane. When
the solute is smaller than the pore, it can travel In a separate demonstration, the device shown in
across the membrane by going through the pore. Fig. 6 was filled with 1 M sucrose and placed in a
Because the solute molecules are larger than the beaker of pure water (Fig. 8). The outlet tube was
solvent water molecules, they are reflected back closed to the outside air but connected by a fluid-
from the membrane more often than water. In this filled cannula leading to a pressure transducer

VOLUME 13 : NUMBER I - ADVANCES IN PHYSIOLOGY EDUCATION - JUNE 1995

s17
 I N N 0 V A T I 0 N S A N D I D E A S

TABLE 2
Osmosis across a model membrane: calculation of u
.04
1 ,ooo-Mwco 3,500~Mwco
Membrane Membrane
T = RTC,
Concentration
mmHg
c 03 Qv, u Qw u
‘c l
cm”/min cm3/min
.-

c9
Sucrose
E 0.25 M 4,636 0.0416 0.153 0.0360 0.072
-o_

s . 02 0.50 M 9,272 0.0837 0.154 0.0749 0.075

0.75 M 13,908 0.1195 0.147 0.1102 0.073
0

Downloaded from http://advan.physiology.org/ by 10.220.33.3 on May 28, 2017

Urea, 1M 18,544 0.0022 0.002 0.0022 0.001
ii! PEG
0.0375 M 696 0.0277 0.680 0.0337 0.449
.01 0.075 M 1,391 0.0729 0.896 0.0748 0.497

Osmotic pressure of solutions was calculated as T = RTC, where

r;~ is pressure, R is the gas constant (0.082 liter atmemol-l .OK-‘),
T is absolute temperature (298”Ic), C is the concentration of
0 impermeant solute, and 1 atm is 760 mmHg. Reflection coeffi-
0 100 200 300 400 500 600 700 cient (D) was calculated as the ratio of observed flow rates to what
should be observed given a hydrostatic pressure of T; that is, if
PRESSURE (mmHg) Qv = A (area) .L,, (filtration coefficient) T-AT, then 0 = &/(A-
FIG. 7. LP . AT). A -Lp (0.0000585 for l,OOO-MWCO membrane and
0.000108 for 3,500-MWCO membrane) values were determined
Pressure-driven flow in cms/min against steady-state
separately (cm3. mmHgl . min -‘>.
pressure for 1,000~MWCO (0) and 3,500~MWCO (A)
Spectrapor dialysis membranes. Slope of line is A x L,
in units of cm3 l rnin-l l mmHg-?
7) Explain the difference in the time to produce
steady-state pressure compared with steady-state
connected to a Grass polygraph. The demonstration flow.
was set up a few minutes before the laboratory, so
that the students could see the full development of In the case of the flow experiments, the steady-state
the pressure. The pressure increased nearly linearly flow was established almost immediately, because
to - 700 mmHg and then increased more slowly all of the flow went down the calibrated tube and
(Fig. 9). none went to expanding the compartment. In the
case of the pressure demonstration, the osmotic
flow caused an expansion of the inner compart-
The purpose of this demonstration was twofold, to ment, which developed a pressure because of this
demonstrate that the osmotic flow could produce a expansion. Because the pressure results from this
real pressure of considerable magnitude and to expansion, considerable fluid transfer was required
illustrate that the development of pressure in the for the pressure generation. Here the students were
bulk solution required time, whereas the develop- introduced to the concept of compliance, C =
ment of osmotic flow at no pressure did not. The AV/AP, the change in volume per unit change in
students were asked three questions in this regard: pressure. The rapid development of steady-state
osmotic flow is due to the fact that the pressure
difference driving the flow is only within the pores
5) How long did it take to establish steady-state flow
of the thin dialysis membrane, and the compliance
for 0.75 M sucrose?
of this structure is small. The compliance of the
whole inner compartment and dialysis membrane,
6) How long did it take to establish steady-state however, is large, so a large volume must be trans-
pressure with 1 M sucrose? ferred to produce the pressure. In both cases,

VOLUME 13 : NUMBER 1 - ADVANCES IN PHYSIOLOGY EDUCATION - JUNE 1995

S18
 I N N 0 V A T I 0 N S A N D I D E A S

to pressure
7
transducer

+----- beaker

- stainless steel mesh

Downloaded from http://advan.physiology.org/ by 10.220.33.3 on May 28, 2017

magnetic stirrer

FIG. 8.
Setup for measuring the pressure developed by osmotic flow. After the inner
compartment was filled with 1 M sucrose, the device was closed to air and
inserted in a beaker full of water. The hose connected the inner compart-
ment to a P23XL pressure transducer.

however, the pressure driving flow within the mem- semipermeable membrane, one which does not let
brane itself was established rapidly. solute pass at all, the observed pressure at equilib-
rium would be the osmotic pressure. Because the
KEY POINTS membranes used in this laboratory exercise are not
This laboratory exercise illustrated the following key ideal semipermeable membranes, the pressure ob-
points for the students. served at equilibrium would be the effective osmotic
pressure.
1) A membrane separating a solution from pure
water will result in fluid movement from the pure 3) Osmotic flow depends on the concentration of
water to the solution side. This is the principal solute (Figs. 4 and 5), because the effective osmotic
observation of osmosis, but it is not an explanation pressure across the membrane is proportional to
of it. concentration (the van? Hoff Law), and the greater
the osmotic pressure, the greater the flow.
2) This fluid movement can generate a pressure,
which in this exercise was > 1,000 mmHg. The 4) Osmotic flow also depends on the kind of solute
pressure that would be generated at equilibrium is (Fig. 5). This is due to the different values of u for
the observed osmotic pressure. For a perfectly each solute-membrane pair. Thus CTfor PEG was

VOLUME 13 : NUMBER 1 - ADVANCES IN PHYSIOLOGY EDUCATION - JUNE 1995

s19
 I N N 0 V A T I 0 N S A N D I D E A S

1100 flow than on the pressure. The students are able,

1000 - within a 2-h period, to determine parameters, such
900 -
asthe reflection coefficient, which stress the interac-
800 -
tion of both solute and membranes in generating
700 -
the osmotic phenomenon.
600 -
500 -
The devices described here were designed and fabricated in
400 - collaboration with Mr. Tom Gentry in the Department of Biomedi-
300 - cal Engineering at the Medical College of Virginia, Virginia
200 - Commonwealth University. The devices can be custom-made by
100 - contract with the Department of Biomedical Engineering. Arrange-
0 I I I ments can be made through the authors.

Downloaded from http://advan.physiology.org/ by 10.220.33.3 on May 28, 2017

I I
0 20 40 60 80 100 120 Address for reprint requests: J. Feher, Dept. of Physiology,
TIME (minutes) Medical College of Virginia, PO Box 980551, Richmond, VA
23298-0551.
FIG. 9.
Pressure caused by osmotic flow against elapsed time
of the flow. The pressure was recorded in the device Received 6 June 1994; accepted in final form 1 December 1994.
shown in Fig. 8.
References
highest, (Tfor sucrose was next, and u for urea was 1. Barrett, D. R. B. Osmosis and surface area to volume rati0.J.
near zero. Biol. Ed. 18: 273-274, 1984.
2. Johnstone, A. H., and N. A. Mahmoud. Pupils’ response to
a model for water transp0rt.J. Biol. Ed. 15: 203-208, 1981.
5) The reflection coefficient, U, also depends on the 3. Hobbie, R. K. Intermediate Physics for Medicine and
characteristics of the membrane. The (Twas always Biology. New York: Wiley, 1978, p. 126-132.
larger for the membrane containing small pores (the 4. Meschia, G., and I. Setnikar. Experimental study of osmosis
1,OOO-MWCO membrane). through a collodion membrane. J. Gen. Physiol. 42: 429-
444,1958.
5. Pfeffer, W. Osmotiche Untersuchungen. Studien zur Zell-
6) An equation can be written that describes fluid mechanik. Leipzig, Germany: Wilhelm Engelmann, 1877.
movement produced by both hydrostatic pressure (Translated by G. R. Kepner and E. J. Stadelmann. In:
(the kind of pressure produced by the weight of a Osmotic Investigations. Studies on Cell Mechanics. New
column of fluid) and osmotic pressure. This equa- York: Van Nostrand Reinhold, 1985.)
6. Renkin, E. M. Filtration, diffusion and molecular sieving
tion is
through porous cellulose membranes. J. Gen. Physiol. 38:
Qv = AL,(AP - CaAr) 225--243,1954.
7. Soto, E., T. Castaneda, and M. Martin. Osmosis: a microcom-
puter laboratory teacher. Int. J. Biomed. Comput. 32: 85-93,
1993.
Here the 2 indicates that the osmotic contributions
8. Tanford, C. Physical Chemistry of Macromolecules. New
of all solutes on both sides of the membrane need to York: Wiley, 1961, p. 145-150.
be considered to obtain the total effective osmotic 9. Van? Hoff, J. H. Die Rolle des Osmotischen Druckes in der
pressure difference across the membrane. Analogie Zwischen Losungen und Gasen.2. Physik. Chemie
1: 481-493, 1887. (Translated by G. L. Blackshear. In: Cell
Membrane Permeability and Transport, edited by G. R.
SUMMARY
Kepner. Stroudsburg, PA: Dowden, Hutchinson, and Ross,
The laboratory exercise described here provides a 1979.)
10. Villars, F. M., and G. B. Benedek. Physics with Illustrative
practical series of experiments that novice students
Examplesfrom Medicine and Biology. Reading, MA: Addison-
can reliably perform within a 2-h laboratory period. Wesley, 1974, vol. 2.
The exercise provides a confirmation of van? Hoff s 11. Woodring, J., and T. H. Dietz. Laboratory Manual for
law while placing more emphasis on the osmotic Human Physiology. New York: Saunders, 1992.

VOLUME 13 : NUMBER 1 - ADVANCES IN PHYSIOLOGY EDUCATION - JUNE 1995

s20