Ir Spectroscopy 2021 Skoog

Download as pdf or txt
Download as pdf or txt
You are on page 1of 48

chapterSIXTEEN

An Introduction to
Infrared Spectrometry

T
he infrared (IR) region of the spectrum Table 16-1 gives the rough limits of each of the three IR regions.
­encompasses radiation with wavenumbers Measurements in the near-IR region are often made with
­photometers and spectrophotometers similar in design and
ranging from about 12,800 to 10 cm21 or
components to the instruments described in earlier chapters
­wavelengths from 0.78 to 1000 µm. Because of s­ imilar for ultraviolet (UV)-visible spectrometry. The most important
applications and instrumentation, the IR spectrum applications of this spectral region have been to the quantitative
is usually subdivided into three regions, the near-IR, analysis of industrial and agricultural materials and for process
the mid-IR, and the far-IR. The techniques and control. Applications of near-IR spectrometry are discussed in
Section 17D.
the ­applications of methods based on the three IR
­spectral regions differ considerably as discussed in Until the early 1980s, instruments for the mid-IR region
were largely of the dispersive type and used diffraction gratings.
this chapter. Since that time, however, mid-IR instrumentation has dramati-
cally changed so that now almost all new instruments are of the
Fourier transform (FT) type. Photometers based on interference
filters also find use for measuring the composition of gases and
atmospheric contaminants.
The appearance of relatively inexpensive FT spectrometers
in the last decade has markedly increased the number and type of
applications of mid-IR radiation. This increase has come about
because interferometric instruments can produce improvements
of an order of magnitude, or more, in signal-to-noise ratios and
detection limits over dispersive instruments. Before the appear-
ance of these instruments, the mid-IR spectral region was used
largely for qualitative organic analysis and structure determi-
nations based on absorption spectra. Now, in contrast, mid-IR
spectrometry is beginning to be used in addition for quantita-
tive analysis of complex samples by both absorption and emis-
sion spectrometry. Applications of this spectral region are also
beginning to appear for microscopic studies of surfaces, analysis
of solids by attenuated total reflectance and diffuse reflectance,
photoacoustic measurements, and other uses. Several of these
applications are described in Chapter 17.
In the past the far-IR region of the spectrum, although
potentially quite valuable, had limited use because of exper-
imental difficulties. The few sources of this type of radiation
are notoriously weak and are further attenuated by the need
for order-sorting filters that prevent radiation of higher grat-
ing orders from reaching the detector. FT spectrometers, with
their much higher throughput, largely alleviate this problem and
Throughout this chapter, this logo indicates
an ­opportunity for online self-study at make the far-IR spectral region much more accessible to scien-
www.tinyurl.com/skoogpia7, linking you to tists. Several applications of far-IR spectroscopy are described in
­interactive tutorials, simulations, and exercises. Section 17E.

389
390 Chapter 16 An Introduction to Infrared Spectrometry

Table 16-1 IR Spectral Regions


Region Wavelengths (l), µm Wavenumbers (n), cm21 Frequencies (n), Hz
Near 0.78 to 2.5 12800 to 4000 3.8 3 1014 to 1.2 3 1014
Middle 2.5 to 50 4000 to 200 1.2 3 1014 to 6.0 3 1012
Far 50 to 1000 200 to 10 6.0 3 1012 to 3.0 3 1011
Most used 2.5 to 15 4000 to 670 1.2 3 1014 to 2.0 3 1013

In this chapter, we first deal with the mechanisms of the as the abscissa because of the inconvenient size of the unit; that
absorption, emission, and reflection of IR radiation using is, a frequency scale of the plot in Figure 16-1 would extend
absorption spectroscopy as the basis for this discussion. We fol- from 1.2 3 1014 to 2.0 3 1013 Hz. Although the axis in terms
low this discussion with a description of the components of IR of wavenumbers is often referred to as a frequency axis, keep
instruments and how these are arranged in dispersive and non- in mind that this terminology is not strictly correct because the
dispersive instruments as well as in FT spectrometers.1 wavenumber n is only proportional to frequency n. The relation-
ships are given in Equation 16-1.
1 n 1 Hz 2
16A Theory of IR Absorption n 1 cm21 2 5
l 1 μm 2
3 10 4 1 μm/cm 2 5
c 1 cm/s 2
(16-1)
Spectrometry
Finally, note that the horizontal scale of Figure 16-1 changes
IR absorption, emission, and reflection spectra for molecular at 2000 cm21, with the units at higher wavenumbers being rep-
species can be rationalized by assuming that all arise from vari- resented by half the linear distance of those at lower wavenum-
ous changes in energy brought about by transitions of molecules bers. The expanded scale in the region from 2000 to 650 cm21
from one vibrational or rotational energy state to another. In this permits easier identification of the numerous IR bands that
section we use molecular absorption to illustrate the nature of appear in this region.
these transitions.
Dipole Moment Changes during
Vibrations and Rotations
16A-1 Introduction IR radiation is not energetic enough to bring about the kinds of
Figure 16-1 shows a typical output from a commercial IR electronic transitions that we have encountered in our discus-
­spectrophotometer. Although the y-axis is shown as linear in sions of UV and visible radiation. Absorption of IR radiation is
transmittance, modern computer-based spectrophotometers can thus confined largely to molecular species that have small energy
also produce spectra that are linear in absorbance. The abscissa differences between various vibrational and rotational states.
in this spectrum is linear in wavenumbers with units of recip-
rocal centimeters. A wavelength scale is also shown at the top To absorb IR radiation, a molecule must undergo a net
of the plot. Computer-based spectrophotometers can produce a change in dipole moment as it vibrates or rotates. Only under
variety of other spectral formats such as linear in wavelength, these circumstances can the alternating electric field of the
baseline corrected, and derivative and smoothed spectra. radiation interact with the molecule and cause changes in the
amplitude of one of its motions. For example, the charge distri-
A linear wavenumber scale is usually preferred in IR bution around a molecule such as hydrogen chloride is not sym-
s­ pectroscopy because of the direct proportionality between metric because the chlorine has a higher electron density than
this quantity and both energy and frequency. The fre- the hydrogen. Thus, hydrogen chloride has a significant dipole
quency of the absorbed radiation is, in turn, the molec- moment and is said to be polar. The dipole moment is deter-
ular vibrational frequency actually responsible for the mined by the magnitude of the charge difference and the dis-
absorption process. ­Frequency, however, is seldom if ever used tance between the two centers of charge. As a hydrogen chloride
molecule vibrates, a regular fluctuation in its dipole moment
occurs, and a field is established that can interact with the elec-
1
tric field associated with radiation. If the frequency of the radi-
For detailed treatments of IR spectrometry, see Introduction to E ­ xperimental
Infrared Spectroscopy: Fundamentals and Practical Methods, M. Tasumi, ation exactly matches a natural vibrational frequency of the
ed., Chichester, UK: Wiley, 2015; P. Larkin, Infrared and Raman Spectros- molecule, absorption of the radiation takes place that produces
copy: ­P rinciples and Spectral Interpretation, Waltham, MA: Elsevier, 2011; a change in the amplitude of the molecular vibration. Similarly,
N. B. ­Colthup, L. H. Daly, and S. E. Wiberley, Introduction to Infrared and Raman
the rotation of asymmetric molecules around their centers of
­Spectroscopy, 3rd ed., San Diego: Academic Press, 1990; C.-P. S. Hsu, in Handbook
of Instrumental Techniques for Analytical Chemistry, F. Settle, ed., Upper Saddle mass results in periodic dipole moment fluctuations that allow
River, NJ: Prentice Hall, 1997, Chap. 15. interaction with the radiation field.
16A Theory of IR Absorption Spectrometry 391

Wavelength, μ m
2.5 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 12.0 15.0
100

90

80

70
Transmittance, %

60

50

40

30

20

10

0
4000 3600 3200 2800 2400 2000 1800 1600 1400 1200 1000 800 650
Wavenumber, cm–1

Figure 16-1 IR absorption spectrum of a thin polystyrene film. Note the scale change on the x-axis at 2000 cm21.

No net change in dipole moment occurs during the define the number and nature of such vibrations and relate these
v­ ibration or rotation of homonuclear species such as O2, N2, or to energies of absorption. An analysis of this kind becomes diffi-
Cl2. As a result, such compounds cannot absorb IR radiation. cult if not impossible for molecules made up of many atoms. Not
With the exception of a few compounds of this type, all other only do large molecules have a large number of vibrating cen-
molecular species absorb IR radiation. ters, but also interactions among several centers can occur and
must be taken into account for a complete analysis.
Rotational Transitions
Vibrations fall into the basic categories of stretching and
The energy required to cause a change in rotational level is
bending. A stretching vibration involves a continuous change
quite small and corresponds to radiation of n # 100 cm 21
in the interatomic distance along the axis of the bond between
1 l . 100 μm 2 . Because rotational levels are quantized, absorp-
two atoms. Bending vibrations are characterized by a change in
tion by gases in this far-IR region is characterized by discrete,
the angle between two bonds and are of four types: scissoring,
well-defined lines. In liquids or solids, intramolecular collisions
rocking, wagging, and twisting. These are shown schematically
and interactions cause broadening of the lines into a continuum.
in Figure 16-2.
Vibrational-Rotational Transitions All of the vibration types shown in Figure 16-2 may be pos-
Vibrational energy levels are also quantized, and for most sible in a molecule containing more than two atoms. In addition,
­m olecules the energy differences between quantum states interaction or coupling of vibrations can occur if the vibrations
­correspond to the mid-IR region. The IR spectrum of a gas usu- involve bonds to a single central atom. The result of coupling is a
ally consists of a series of closely spaced lines, because there are change in the characteristics of the vibrations involved.
several rotational energy levels for each vibrational level. On the In the treatment that follows, we first consider isolated
other hand, rotation is highly restricted in liquids and solids; vibrations represented by a simple harmonic oscillator model.
in such samples, discrete vibrational-rotational lines disappear, Modifications to the theory of the harmonic oscillator, which
leaving only somewhat broadened vibrational bands. are needed to describe a molecular system, are taken up next.
Finally, the effects of vibrational interactions in molecular
Types of Molecular Vibrations
­systems are discussed.
The relative positions of atoms in a molecule are not fixed but
instead fluctuate continuously because of a multitude of differ-
ent types of vibrations and rotations about the bonds in the mol- 16A-2 Mechanical Model of a Stretching
ecule. For a simple diatomic or triatomic molecule, it is easy to
Vibration in a Diatomic Molecule
The characteristics of an atomic stretching vibration can be
Tutorial: Learn more about IR absorption at approximated by a mechanical model consisting of two masses
www.tinyurl.com/skoogpia7 connected by a spring. A disturbance of one of these masses
392 Chapter 16 An Introduction to Infrared Spectrometry

­restoring force F is proportional to the displacement (Hooke’s


law). That is,
F 5 2ky (16-2)
Symmetric Asymmetric where k is the force constant, which depends on the stiffness
(a) Stretching vibrations of the spring. The negative sign indicates that F is a restoring
force. This means that the direction of the force is opposite the
direction of the displacement. Thus, the force tends to restore
the mass to its original position.

Potential Energy of a Harmonic Oscillator


In-plane rocking In-plane scissoring The potential energy E of the mass and spring can be arbitrarily
assigned a value of zero when the mass is in its rest, or equilib-
rium, position. As the spring is compressed or stretched, how-
+ + +
– –
ever, the potential energy of this system increases by an amount
equal to the work required to displace the mass. If, for example,
– the mass is moved from some position y to y 1 dy, the work
and hence the change in potential energy dE is equal to the force
Out-of-plane wagging Out-of-plane twisting F times the distance dy. Thus,
(b) Bending vibrations dE 5 2F dy (16-3)
Figure 16-2 Types of molecular vibrations. Note that 1 indicates Combining Equations 16-3 and 16-2 yields
motion from the page toward the reader and 2 indicates motion
away from the reader. dE 5 ky dy
Integrating between the equilibrium position y 5 0 and y gives

3 dE 5 3 y dy
E y

along the axis of the spring results in a vibration called a simple


0 0
harmonic motion.
1 2
Let us first consider the vibration of a single mass attached to E5 ky (16-4)
2
a spring that is hung from an immovable object (see Figure 16-3a).
If the mass is displaced a distance y from its equilibrium The potential-energy curve for a simple harmonic oscilla-
­position by application of a force along the axis of the spring, the tion, derived from Equation 16-4, is a parabola, as depicted in

r1
1 2 1
r2
Dissociation energy
–A
2
m 0
y
Potential energy E

Potential energy E

+A
v=6
v=5
v=4
Energy level/
v=3
vibrational
v=2 quantum number
v=1
v=0
0 0
–A 0 +A Interatomic distance r
Displacement y

(a) (b)

Figure 16-3 Potential-energy diagrams. (a) Harmonic oscillator. (b) Curve 1, harmonic oscillator; curve 2, anharmonic motion.
16A Theory of IR Absorption Spectrometry 393

Figure 16-3a. Notice that the potential energy is a maximum Thus, the vibrational frequency for such a system is given by
when the spring is stretched or compressed to its maximum
amplitude A, and it decreases to zero at the equilibrium position. 1 k 1 k 1 m1 1 m 2 2
nm 5 5 (16-10)
2p Å m 2p Å m1m2
Vibrational Frequency
The motion of the mass as a function of time t can be deduced Molecular Vibrations
from classical mechanics as follows. Newton’s second law states The approximation is ordinarily made that the behavior of a
that molecular vibration is analogous to the mechanical model just
described. Thus, the frequency of the molecular vibration is cal-
F 5 ma culated from Equation 16-10 after substituting the masses of the
where m is the mass and a is its acceleration. But acceleration is two atoms for m1 and m2. The quantity k is the force constant of
the second derivative of distance with respect to time. Thus, the chemical bond, which is a measure of its stiffness.

d 2y
a5 16A-3 Quantum Treatment of Vibrations
dt 2
The equations of ordinary mechanics that we have used thus far
Substituting these expressions into Equation 16-2 gives do not completely describe the behavior of particles of atomic
dimensions. For example, the quantized nature of molecular
d 2y vibrational energies, and of other atomic and molecular energies
m 5 2ky (16-5)
dt 2 as well, does not appear in these equations. We may, however,
invoke the concept of the simple harmonic oscillator to develop
A solution to this equation must be a periodic function such
the wave equations of quantum mechanics. Solutions of these
that its second derivative is equal to the original function times
equations for potential energies have the form
2k/m. A suitable cosine relationship meets this requirement.
Thus, the instantaneous displacement of the mass at time t can 1 h k
be written as E 5 av 1 b (16-11)
2 2p Å m
y 5 A cos 2pnm t (16-6)
where h is Planck’s constant, and v is the vibrational quantum
where nm is the natural vibrational frequency and A is the max- number, which can take only positive integer values (including
imum amplitude of the motion. The second derivative of Equa- zero). Thus, in contrast to ordinary mechanics where vibrators
tion 16-6 with respect to time is can assume any potential energy, quantum mechanical vibrators
can take on only certain discrete energies.
d 2y
5 24p 2nm2 A cos 2pnm t (16-7) It is interesting to note that the factor !k/m/2p appears in
dt 2 both the classical and the quantum equations; by substituting
Substitution of Equations 16-6 and 16-7 into Equation 16-5 gives Equation 16-10 into 16-11, we find

4p 2nm2 m 1
E 5 av 1 bhnm (16-12)
A cos 2pnm t 5 A cos 2pnm t 2
k
The natural frequency of the oscillation is then where nm is the vibrational frequency of the classical model.2
We now assume that transitions in vibrational energy levels
1 k can be brought about by absorption of radiation, provided the
nm 5 (16-8)
2p Å m energy of the radiation exactly matches the difference in energy
levels DE between the vibrational quantum states and provided
where nm is the natural frequency of the mechanical oscillator.
also the vibration causes a change in dipole moment. This dif-
Although it depends on the force constant of the spring and the
ference is identical between any pair of adjacent levels, because v
mass of the attached body, the natural frequency is independent
in Equations 16-11 and 16-12 can assume only whole numbers;
of the energy imparted to the system; changes in energy merely
that is,
result in a change in the amplitude A of the vibration.
The equation just developed may be modified to describe h k
DE 5 hnm 5 (16-13)
the behavior of a system consisting of two masses m 1 and m2 2p Å m
connected by a spring. This is accomplished by substituting the
reduced mass m for the single mass m where 2
Unfortunately, the generally accepted symbol for the vibrational quantum num-
m1m2 ber v is similar in appearance to the Greek nu, n, which symbolizes frequency.
m5 (16-9) Thus, constant care must be exercised to avoid confusing the two in equations
m1 1 m 2 such as Equation 16-12.
394 Chapter 16 An Introduction to Infrared Spectrometry

At room temperature, the majority of molecules are in the


ground state v 5 0; thus, from Equation 16-12, Example 16-1

1 Calculate the approximate wavenumber and wavelength of


E0 5 hnm
2 the fundamental absorption due to the stretching vibration of
a carbonyl group CwO.
Promotion to the first excited state v 5 1 with energy
3
Solution
E1 5 hnm The mass of the carbon atom in kilograms is given by
2
requires radiation of energy 12 3 1023 kg/mol
m1 5 3 1 atom
6.0 3 10 23 atom/mol
3 1
a hnm 2 hnm b 5 hnm 5 2.0 3 10226 kg
2 2

The frequency of radiation n that will bring about this change Similarly, for oxygen,
is identical to the classical vibrational frequency of the bond nm. m2 5 1 16 3 1023 2 / 1 6.0 3 10 23 2 5 2.7 3 10226 kg
That is,
and the reduced mass m is given by (Equation 16-9)
h k
Eradiation 5 hn 5 DE 5 hnm 5 2.0 3 10226 kg 3 2.7 3 10226 kg
2p Å m
m5
1 2.0 1 2.7 2 3 10226 kg
or
5 1.1 3 10226 kg
1 k
n 5 nm 5 (16-14) As noted earlier, the force constant for the typical double
2p Å m
bond is about 1 3 103 N/m. Substituting this value and m into
If we wish to express the radiation in wavenumbers, we Equation 16-15 gives
­substitute Equation 6-3 and rearrange:
1 3 10 3 N/m
n 5 5.3 3 10212 s/cm
1 k k Å 1.1 3 10226 kg
n5 5 5.3 3 10212 (16-15)
2pc Å m Åm
5 1.6 3 10 3 cm21
where n is the wavenumber of an absorption maximum (cm21), The carbonyl stretching band is found experimentally to be in
k is the force constant for the bond in newtons per meter (N/m), the region of 1600 to 1800 cm21 (6.3 to 5.6 µm).
c is the velocity of light (cm s21), and m is the reduced mass (kg)
defined by Equation 16-9.3
Selection Rules
IR measurements in conjunction with Equation 16‑14 As given by Equations 16-12 and 16-13, the energy for a tran-
or 16-15 permit the evaluation of the force constants for vari- sition from energy level 1 to 2 or from level 2 to 3 should be
ous types of chemical bonds. Generally, k has been found to identical to that for the 0 to 1 transition. Furthermore, quantum
lie in the range between 3 3 102 and 8 3 102 N/m for most theory indicates that the only transitions that can take place are
single bonds, with 5 3 102 serving as a reasonable average those in which the vibrational quantum number changes by
value. Double and triple bonds are found by this same means unity; that is, the selection rule states that Dv 5 61. Because
to have force constants of about two and three times this value the vibrational levels are equally spaced for a harmonic oscilla-
1 1 3 103 and 1.5 3 103 N/m, respectively 2 . With these aver- tor, only a single absorption peak should be observed for a given
age experimental values, Equation 16-15 can be used to esti- molecular vibration. In addition to the Dv 5 61 selection rule,
mate the wavenumber of the fundamental absorption band, or there must be a change in dipole moment during the vibration.
the absorption due to the transition from the ground state to
the first excited state, for a variety of bond types. The following Anharmonic Oscillator
example demonstrates such a calculation. Thus far, we have considered the classical and quantum mechan-
ical treatments of the harmonic oscillator. The potential energy
of such a vibrator changes periodically as the distance between
the masses fluctuates (Figure 16-3a). From qualitative consid-
3
By definition, the newton has the units of N 5 kg m/s2 . Thus, !k/m has erations, however, this description of a molecular vibration
units of s21. appears imperfect. For example, as the two atoms approach one
16A Theory of IR Absorption Spectrometry 395

another, coulombic repulsion between the two nuclei produces a t­ ranslational motion requires three coordinates and thus this
force that acts in the same direction as the restoring force of the common motion requires three of the 3N degrees of freedom.
bond. For this reason, the potential energy can be expected to Another three degrees of freedom are needed to describe the
rise more rapidly than the harmonic approximation predicts. At rotation of the molecule as a whole. The remaining 3N 2 6
the other extreme of oscillation, a decrease in the restoring force, degrees of freedom involve interatomic motion and hence rep-
and thus the potential energy, occurs as the interatomic distance resent the number of possible vibrations within the molecule. A
approaches that at which dissociation of the atoms takes place. linear molecule is a special case because, by definition, all of the
In theory, the wave equations of quantum mechanics permit atoms lie on a single, straight line. Rotation about the bond axis
the derivation of more nearly correct potential-energy curves for is not possible, and two degrees of freedom suffice to describe
molecular vibrations. Unfortunately, however, the mathemat- rotational motion. Thus, the number of vibrations for a linear
ical complexity of these equations prevents their quantitative molecule is given by 3N 2 5. Each of the 3N 2 6 or 3N 2 5
application to all but simple diatomic systems. Qualitatively, the vibrations is called a normal mode.
curves take the anharmonic form shown as curve 2 in Figure For each normal mode of vibration, there is a potential-­
16-3b. Such curves depart from harmonic behavior by varying energy relationship such as that shown by the solid lines in
degrees, depending on the nature of the bond and the atoms Figure 16-3b. The same selection rules discussed earlier apply
involved. Note, however, that the harmonic and anharmonic for each of these relationships. In addition, to the extent that
curves are nearly alike at low potential energies. a vibration approximates harmonic behavior, the differences
Anharmonicity leads to deviations of two kinds. At between the energy levels of a given vibration are the same; that
higher quantum numbers, ΔE becomes smaller (see curve 2 in­ is, a single absorption peak appears for each vibration having a
Figure 16-3b), and the selection rule is not rigorously followed. change in dipole moment.
As a result, weaker transitions called overtones are sometimes Four factors tend to produce fewer experimental bands
observed. These transitions correspond to Dv 5 62 or 63. The than would be expected from the theoretical number of normal
frequencies of such overtone transitions are approximately two modes. Fewer absorption bands are found when (1) the symme-
or three times that of the fundamental frequency, and the inten- try of the molecules is such that no change in dipole moment
sities are lower than that of the fundamental. results from a particular vibration; (2) the energies of two or
Vibrational spectra are further complicated by two different more vibrations are identical or nearly identical; (3) the absorp-
vibrations in a molecule interacting to give absorptions at fre- tion intensity is so low as to be undetectable by ordinary means;
quencies that are approximately the sums or differences of their or (4) the vibrational energy is in a wavelength region beyond
fundamental frequencies. Again, the intensities of these sum and the range of the instrument used to measure absorption.
difference bands are generally low. Occasionally, more absorption bands are found than are
expected based on the number of normal modes. We have
already mentioned the overtone bands that occur at two or three
16A-4 Vibrational Modes times the frequency of the fundamental. In addition, combina-
It is ordinarily possible to deduce the number and kinds of vibra- tion bands are sometimes encountered when a photon excites
tions in simple diatomic and triatomic molecules and whether two vibrational modes simultaneously. The frequency of the
these vibrations will lead to absorption. Complex molecules may combination band is approximately the sum or difference of the
contain several types of atoms as well as bonds. For these mole- two fundamental frequencies. This phenomenon occurs when a
cules, the many types of possible vibrations give rise to IR spec- quantum of energy is absorbed by two bonds rather than one.
tra that are much more difficult to analyze or predict.
The number of possible vibrations in a polyatomic mole-
cule can be calculated as follows. Three coordinates are needed 16A-5 Vibrational Coupling
to locate a point in space. To fix N points requires a set of three The energy of a vibration, and thus the wavelength of the corre-
coordinates for each point, for a total of 3N. Each coordinate sponding absorption maximum, may be influenced by (or cou-
corresponds to one degree of freedom for one of the atoms in a pled with) other vibrators in the molecule. A number of factors
polyatomic molecule. For this reason, a molecule containing N influence the extent of such coupling.
atoms is said to have 3N degrees of freedom. 1. Strong coupling between stretching vibrations occurs
In defining the motion of a molecule, we need to consider only when there is an atom common to the two
(1) the motion of the entire molecule through space (that is, the vibrations.
translational motion of its center of gravity); (2) the rotational 2. Interaction between bending vibrations requires a
motion of the entire molecule around its center of gravity; and common bond between the vibrating groups.
(3) the motion of each of its atoms relative to the other atoms, 3. Coupling between a stretching and a bending vibra-
or in other words, its individual vibrations. Because all atoms tion can occur if the stretching bond forms one side
in the molecule move in concert through space, definition of of the angle that varies in the bending vibration.
396 Chapter 16 An Introduction to Infrared Spectrometry

4. Interaction is greatest when the coupled groups have 1 3 3 3 2 2 6 5 3 vibrational modes that take the following
individual energies that are nearly equal. forms:
5. Little or no interaction is observed between groups
separated by two or more bonds.
6. Coupling requires that the vibrations be of the same
symmetry species.4

As an example of coupling effects, let us consider the IR Symmetric stretching


spectrum of carbon dioxide. If no coupling occurred between
the two CwO bonds, an absorption band would be expected at
the same wavenumber as that for the CwO stretching vibration
in an aliphatic ketone (about 1700 cm21, or 6 µm; see Example
16-1). Experimentally, carbon dioxide exhibits two absorption
maxima, one at 2350 cm21 (4.3 µm) and the other at 667 cm21 Asymmetric stretching Scissoring
(15 µm).
Because the central atom is not in line with the other two, the
Carbon dioxide is a linear molecule and thus has
symmetric stretching vibration produces a change in dipole
1 3 3 3 2 2 5 5 4 normal modes. Two stretching vibrations are
moment and is thus IR active. For example, stretching peaks at
possible; furthermore, interaction between the two can occur
3657 and 3766 cm21 (2.74 and 2.66 µm) appear in the IR spec-
because the bonds involved are associated with a common car-
trum for the symmetric and asymmetric stretching vibrations of
bon atom. As can be seen, one of the coupled vibrations is sym-
the water molecule. There is only one component to the scis-
metric and the other is asymmetric.
soring vibration for this nonlinear molecule because motion
in the plane of the molecule constitutes a rotational degree of
freedom. For water, the bending vibration causes absorption at
Symmetric Asymmetric 1595 cm21 (6.27 µm). The difference in behavior of linear and
nonlinear triatomic molecules with two and three absorption
The symmetric vibration causes no change in dipole moment,
bands, respectively, illustrates how IR absorption spectroscopy
because the two oxygen atoms simultaneously move away from
can sometimes be used to deduce molecular shapes.
or toward the central carbon atom. Thus, the symmetric vibra-
tion is IR inactive. In the asymmetric vibration, one oxygen Coupling of vibrations is a common phenomenon. As a
moves away from the carbon atom as the carbon atom moves result, the position of an absorption band corresponding to a
toward the other oxygen. As a consequence, a net change in given organic functional group cannot be specified exactly. For
charge distribution occurs periodically, producing a change in example, the CiO stretching frequency in methanol is 1034 cm21
dipole moment, so absorption at 2350 cm21 results. (9.67 µm), in ethanol it is 1053 cm21 (9.50 µm), and in 2-butanol
it is 1105 cm21 (9.05 µm). These variations result from a cou-
The remaining two vibrational modes of carbon dioxide
pling of the CiO stretching with adjacent CiC ­stretching or
involve scissoring, as shown here.
CiH vibrations.
Although interaction effects may lead to uncertainties
in the identification of functional groups contained in a com-
– + – pound, it is this very effect that provides the unique features of
an IR absorption spectrum that are so important for the positive
identification of a specific compound.
These two bending vibrations are the resolved components at
90° to one another of the bending motion in all possible planes
around the bond axis. The two vibrations are identical in energy 16B IR Instrumentation
and thus produce a single absorption band at 667 cm21. Quan-
tum states that are identical, as these are, are said to be degenerate. Three types of instruments for IR absorption measurements are
It is of interest to compare the spectrum of carbon diox- commonly available: (1) dispersive spectrophotometers with
ide with that of a nonlinear, triatomic molecule such as water, a grating monochromator, (2) FT spectrometers employing an
sulfur dioxide, or nitrogen dioxide. These molecules have interferometer (Section 7I), and (3) nondispersive photometers
using a filter or an absorbing gas that are used for analysis of
atmospheric gases at specific wavelengths.
4
For a discussion of symmetry operations and symmetry species, see F. A. ­Cotton,
Until the 1980s, the most widely used instruments for
Chemical Applications of Group Theory, 3rd ed., New York: Wiley 1990;
R. S. Drago, Physical Methods for Chemists, 2nd ed., Philadelphia: Saunders, 1992; IR measurements were dispersive spectrophotometers. Now,
A. Vincent, Molecular Symmetry and Group Theory, Chichester: Wiley, 2001. ­however, this type of instrument has been largely displaced for
16B IR Instrumentation 397

mid- and far-IR measurements by FT spectrometers because Drive Mechanism. Requirements for satisfactory interfero-
of their speed, reliability, signal-to-noise advantage, and con- grams (and thus satisfactory spectra) are that the moving mirror
venience. Dispersive spectrometers are still used in the near-IR have constant velocity and that its position be precisely known
where they are often extensions of UV-visible instruments, but at many points along its travel. The planarity of the mirror must
many dedicated near-IR instruments are of the Fourier trans- also remain constant during its entire sweep of 10 cm or more.
form–IR (FTIR) type. In the far-IR region, where wavelengths range from 50
to 1000 µm (200 to 10 cm21), displacement of the mirror by a
16B-1 FT Spectrometers fraction of a wavelength, and accurate measurement of its posi-
The theoretical basis and the inherent advantages of FT and tion, can be accomplished by means of a motor-driven microm-
other multiplex instruments were discussed in some detail eter screw. A more precise and sophisticated mechanism is
in Section 7I, and the reader may find it worthwhile to review required for the mid- and near-IR regions, however. Here, the
that section before proceeding further. Two types of multiplex mirror mount is generally floated on an air bearing held within
instruments have been described for the IR region. In the FT close-fitting stainless steel sleeves (see Figure 16-4). The mount
spectrometer, coding is accomplished by splitting the source into is driven by a linear drive motor and an electromagnetic coil
two beams whose path lengths can be varied periodically to give similar to the voice coil in a loudspeaker; an increasing current
interference patterns. The FT is then used for data processing.5 in the coil drives the mirror at constant velocity. After reaching
The second is the Hadamard transform spectrometer, which is a its terminus, the mirror is returned rapidly to the starting point
dispersive instrument that uses a moving mask at the focal plane for the next sweep by a rapid reversal of the current. The length
of a monochromator for encoding the spectral data. Hadamard of travel varies from 1 to about 20 cm; the scan rates range from
transform IR instruments have not been widely adopted and 0.01 to 10 cm/s.
will, therefore, not be discussed further in this book.6 Two additional features of the mirror system are neces-
When FTIR spectrometers first appeared in the market- sary for successful operation. The first is a means of sampling
place, they were bulky, expensive (.$100,000), and required fre- the interferogram at precisely spaced retardation intervals. The
quent mechanical adjustments. For these reasons, their use was second is a method for determining exactly the zero retardation
limited to special applications where their unique characteristics point to permit signal averaging. If this point is not known pre-
(speed, high resolution, sensitivity, and unparalleled wavelength cisely, the signals from repetitive sweeps would not be in phase,
precision and accuracy) were essential. FT instruments have and averaging would then tend to degrade rather than improve
now been reduced to benchtop size and have become reliable the signal.
and easy to maintain. Furthermore, the price of simpler mod- The problem of precise signal sampling and signal aver-
els has been reduced to the point where they are competitive aging can be accomplished by using two or three interferom-
with all but the simplest dispersive instruments (,$15,000 and eters rather than one, with a single mirror mount holding the
more). For these reasons, FT instruments have largely displaced movable mirrors. Figure 16-4 is a schematic showing such an
dispersive instruments in the laboratory.7 arrangement. The components and radiation paths for each of
the three interferometer systems are indicated by the subscripts
Components of FT Instruments 1, 2, and 3, respectively. System 1 is the IR system that ultimately
The majority of commercially available FTIR instruments are provides an interferogram similar to that shown as curve A in
based on the Michelson interferometer, although other types Figure 16-5. System 2 is a so-called laser-fringe reference sys-
of optical systems are also encountered. We shall consider the tem, which provides sampling-interval information. It consists
Michelson design only, which is illustrated in Figure 7-43.8 of a He-Ne laser S2, an interferometric system including mirrors
MM2 and M2, a beamsplitter B2, and a transducer T2. The output
from this system is a cosine wave, as shown in C of Figure 16‑5.
5
For detailed discussions of FTIR spectroscopy, see B. C. Smith, Fundamentals This signal is converted electronically to the square-wave form
of Fourier Transform Infrared Spectroscopy, 2nd ed., Boca Raton, FL: CRC Press, shown in D; sampling begins or terminates at each successive
2011; P. R. Griffiths and J. A. deHaseth, Fourier Transform Infrared Spectroscopy,
zero crossing. The laser-fringe reference system gives a highly
2nd ed., New York: Wiley, 2007.
6
reproducible and regularly spaced sampling interval. In most
For a description of the Hadamard transform and Hadamard transform spec-
troscopy, see D. K. Graff, J. Chem. Educ., 1995, 72, 304, DOI: 10.1021/ed072p304;
instruments, the laser signal is also used to control the speed of
Fourier, Hadamard, and Hilbert Transforms in Chemistry, A. G. Marshall, ed., the mirror-drive system at a constant level.
New York: Plenum Press, 1982.
The third interferometer system, sometimes called the
7
For reviews of commercial FTIR spectrometers, see J. P. Smith and V. ­Hinson- white-light system, uses a tungsten source S3 and transducer T3
Smith, Anal. Chem., 2003, 75, 37A, DOI: 10.1021/ac031224d; R. Mukhopadhyay,
Anal. Chem., 2004, 76, 369A, DOI: 10.1021/ac041652z.
sensitive to visible radiation. Its mirror system is fixed to give a
8 zero retardation displaced to the left from that for the analytical
The Michelson interferometer was designed and built in 1891 by A. A. Michelson.
He was awarded the 1907 Nobel Prize in physics for the invention of signal (see interferogram B, Figure 16-5). Because the source is
interferometry. polychromatic, its power at zero retardation is much larger than
398 Chapter 16 An Introduction to Infrared Spectrometry

M3 M2 M1

T3
MM3 B3 B2
T2
MM2
IR
Linear drive Moving mirror source
motor assembly S1
MM1
B1

White-light
source
Air bearing S3

Laser
Laser To sample
source and T1
S2

Figure 16-4 Interferometers in an FTIR spectrometer. Subscript 1 defines the radiation path in the IR
­interferometer. Subscripts 2 and 3 refer to the laser and white-light interferometers, respectively. (Courtesy of
Thermo Fisher Scientific Inc., Waltham, MA.)

Zero
retardation
B

Signal sampling interval

D 0

Figure 16-5 Time-domain signals for the three interferometers


contained in many FTIR instruments. Curve A, IR signal; curve B,
white-light signal; curve C, laser-fringe reference signal; curve D,
square-wave electrical signal formed from the laser signal. (Adapted
from P. R. Griffiths, Anal. Chem., 1974, 46, 645A, DOI: 10.1021/
ac60343a756. Copyright 1974 American Chemical Society.)
16B IR Instrumentation 399

any signal before and after that point. Thus, this maximum is
used to trigger the start of data sampling for each sweep at a
highly reproducible point.
The triple interferometer system just described leads to
remarkable precision in determining spectral frequencies, which
significantly exceeds that realizable with conventional grating
instruments. This high reproducibility is particularly important
when many scans are to be averaged. Contemporary instruments,
such as the benchtop unit pictured in Figure 16-6, are able to
achieve excellent frequency precision with one or two interferom-
eters. In the instrument diagrammed in Figure 16-7, the interfer-
ometer is actually two parallel interferometers, one to modulate
the IR radiation from the source before it passes through the
sample and the second to modulate the red light from the He-Ne
Figure 16-6 Photo of a basic, benchtop FTIR spectrometer
laser to provide the reference signal for acquiring data from the
­suitable for student use. Spectra are recorded in a few seconds and IR detector. No white-light source is used, and the IR interfero-
­displayed on a laptop computer. Software is available for ­controlling gram is used to establish zero retardation. The maximum in the
the instrument, collecting and processing data, and analyzing IR interferogram is an excellent reference because this is the only
the results. An IR database is included for analysis and spectral point at which all wavelengths interfere constructively.
­interpretation. (Courtesy of Thermo Electron Corp., Madison, WI.)

Beamsplitter
Interferometer
DO NOT STARE INTO LASER
OR VIEW DIRECTLY WITH
OPTICA L INSTRUMENTS

Mirror
Laser

IR source Mirror
Optical path

Mirror Sample

Mirror

Detector

Sample compartment

Figure 16-7 Diagram of a basic FTIR spectrometer. Radiation of all frequencies


from the IR source is reflected into the interferometer where it is modulated by the
moving mirror on the left. The modulated radiation is then reflected from the two
mirrors on the right through the sample in the compartment at the bottom. After
passing through the sample, the radiation falls on the transducer. A data-acquisition
system attached to the transducer records the signal and stores it in the memory of a
­computer as an interferogram. (Courtesy of Thermo Fisher Scientific, Waltham, MA.)
400 Chapter 16 An Introduction to Infrared Spectrometry

The benchtop system shown in Figure 16-6 is capable of A double-beam spectrometer is illustrated in Figure 16-9.
providing spectra with a resolution of better than 1 cm21. An The mirrors directing the interferometer beam through the
educational kit is available to aid in using the instrument in an sample and reference cells are oscillated rapidly compared to the
academic environment. The kit includes information on IR the- movement of the interferometer mirror so that sample and ref-
ory, FTIR instrumentation, and several experiments. erence information can be obtained at each mirror position. The
double-beam design compensates for source and detector drifts.
Instrument Designs
FTIR spectrometers can be single-beam or double-beam instru- Performance Characteristics
ments. Figure 16-8 shows the optics of a basic single-beam of Commercial Instruments
spectrometer, which sells in the range of $15,000 to $25,000. A number of instrument manufacturers offer several models of
A typical procedure for determining transmittance or absor- FTIR instruments. The least expensive has a range of 7800 to
bance with this type of instrument is to first obtain a reference 350 cm21 (1.3 to 29 µm) with a resolution of a few wavenum-
interferogram by scanning a reference (usually air) twenty or bers. This performance can be obtained with a scan time as brief
thirty times, coadding the data, and storing the results in the as 1 second. More expensive instruments with interchangeable
memory of the instrument computer (usually after transforming beamsplitters, sources, and transducers offer expanded fre-
it to the spectrum). A sample is then inserted in the radiation quency ranges and higher resolutions. For example, one instru-
path and the process repeated. The ratio of sample and reference ment is reported to produce spectra from the far-IR (10 cm21 or
spectral data is then computed to give the transmittance at var- 1000 µm) through the visible region to 50,000 cm21, or 200 nm.
ious frequencies. From this ratio the absorbance is calculated as
a function of wavenumber. Ordinarily, modern IR sources and
detectors are sufficiently stable so that reference spectra need to Simulation: Learn more about IR instrumentation
be obtained only occasionally. at www.tinyurl.com/skoogpia7

IR source

Interferometer
IR Movable
transducer mirror

Laser
Fixed detector
mirror

Beamsplitter

Mirrors with
center hole
for laser beam

Sample Laser
compartment

Figure 16-8 Single-beam FTIR spectrometer. In one arm of the interferometer, the IR source radiation travels through the
beamsplitter to the fixed mirror, back to the beamsplitter, and through the sample to the IR transducer. In the other arm, the
IR source radiation travels to the beamsplitter, is reflected to the movable mirror, and travels back through the beamsplitter
to the sample and to the transducer. When the two beams meet again at the beamsplitter, they can interfere with each other
if the phase difference (path difference) is appropriate. A plot of the signal versus mirror displacement is the interferogram.
The ­interferogram contains information about all the frequencies present. The spectrum, intensity versus wavenumber, is the
FT of the interferogram. It can be calculated with a computer from the signal versus mirror displacement. An empty sample
­compartment allows the reference spectrum to be calculated. Next, the sample is placed in the sample compartment and the sam-
ple s­ pectrum is obtained. The absorbance is then calculated at each wavenumber from the ratio of the sample intensity to
the reference intensity.
16B IR Instrumentation 401

IR source
Interferometer

Movable
mirror

Fixed
Sample mirror
compartment Beamsplitter
IR
transducer

M2
M1

Figure 16-9 Double-beam FTIR spectrometer. The beam emerging from the interferometer strikes mirror M1, which in one
­position directs the beam through the reference cell and in the other position directs it through the sample cell. Mirror M2,
which is synchronized to M1, alternately directs the reference beam and the sample beam to the transducer.

Resolutions for commercial instruments vary from 8 to less than FTIR instruments have advantages over dispersive
0.01 cm21. Several minutes are required to obtain a complete i­nstruments in all areas of chemistry, particularly for (1) very
spectrum at the highest resolution. high-­resolution work that is encountered with gaseous mix-
tures having complex spectra resulting from the superposition
Advantages of FT Spectrometers of vibrational and rotational bands, (2) the study of samples
Over most of the mid-IR spectral range, FT instruments have with high absorbances, (3) the study of substances with weak
signal-to-noise ratios that are better than those of a good-quality absorption bands (for example, the study of compounds that are
dispersive instrument, usually by more than an order of mag- chemisorbed on catalyst surfaces), (4) investigations requiring
nitude. The enhanced signal-to-noise ratio can, of course, be fast scanning such as kinetic studies or detection of chromato-
traded for rapid scanning, with good spectra being attainable graphic effluents, (5) collecting IR data from very small samples,
in a few seconds in most cases. Interferometric instruments are (6) obtaining reflection spectra, and (7) IR emission studies.
also characterized by high resolutions (,0.1 cm21) and highly
accurate and reproducible frequency determinations. The lat-
ter property is particularly helpful when spectra are to be sub- 16B-2 Dispersive Instruments
tracted for background correction. Although nearly all instruments produced today are FT sys-
Another advantage of FT instruments is that their optics tems, many dispersive spectrophotometers are still found in
provide a much larger energy throughput (one to two orders of laboratories. Dispersive IR spectrophotometers are generally
magnitude) than do dispersive instruments, which are limited double-beam, recording instruments, which use reflection grat-
in throughput by the need for narrow slit widths. The potential ings for dispersing radiation. As was pointed out in Section 13D-2,
gain in throughput, however, may be partially offset by the lower the double-beam design is less demanding with respect to the
sensitivity of the fast-response detector required for the interfer- performance of sources and detectors—an important character-
ometric measurements. Finally, it should be noted that the inter- istic because of the relatively low intensity of IR sources, the low
ferometer is free from the problem of stray radiation because sensitivity of IR transducers, and the resulting requirement for
each IR frequency is, in effect, chopped at a different frequency. large signal amplifications (see Section 16C).
402 Chapter 16 An Introduction to Infrared Spectrometry

4000 2500 Wavenumber, cm–1


5000 3000 2000 1500 1300 1100 1000 900 800 700 650
100
90 Double-beam mode
80
Transmittance, %

70
60
Single-beam
50 mode
40
30 H 2O
CO2
20
H 2O
10 CO2
0
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Wavelength, μm

Figure 16-10 Single- and double-beam spectra of atmospheric water vapor and CO2. In the lower, single-beam trace, the
a­ bsorption of atmospheric gases is apparent. The top, double-beam spectrum shows that the reference beam compensates
very well for this absorption and allows for a stable 100% T baseline to be obtained. (From J. D. Ingle Jr. and S. R. Crouch,
­Spectrochemical Analysis, p. 409, Upper Saddle River, NJ: Prentice-Hall, 1988. With permission.)

An additional reason for the general use of double-beam in which the radiant power of the reference beam is reduced,
instruments in the IR region is shown in Figure 16-10. The or attenuated, to match that of the beam passing through the
lower curve reveals that atmospheric water and carbon diox- sample. Attenuation is accomplished by imposing a device
ide absorb radiation in some important spectral regions and that removes a continuously variable fraction of the reference
can cause serious interference problems. The upper curve beam. The attenuator commonly takes the form of a comb, the
shows that the reference beam compensates nearly perfectly teeth of which are tapered so that there is a linear relationship
for absorption by both compounds. A stable 100%T baseline between the lateral movement of the comb and the decrease in
results. power of the beam. Movement of the comb occurs when a dif-
Generally, dispersive IR spectrophotometers incorporate ference in power of the two beams is sensed by the detection
a low-frequency chopper (five to thirty cycles per second) that system. Most dispersive IR instruments are older instruments
permits the detector to discriminate between the signal from that use mechanical recorders instead of computers. For such
the source and signals from extraneous radiation, such as IR instruments, the comb movement is synchronized with the
emission from various bodies surrounding the transducer. Low recorder pen so that its position gives a measure of the rela-
chopping rates are demanded by the slow response times of the tive power of the two beams and thus the transmittance of the
IR transducers used in most dispersive instruments. In gen- sample.
eral, the optical designs of dispersive instruments do not differ Note that three types of systems link the components of
greatly from the double-beam UV-visible spectrophotometers the instrument in Figure 16-11: (1) a radiation linkage indi-
discussed in the previous chapter except that the sample and cated by dashed lines, (2) a mechanical linkage shown by thick
reference compartments are always located between the source dark lines, and (3) an electrical connection shown by narrow
and the monochromator in IR instruments. This arrangement solid lines.
is possible because IR radiation, in contrast to UV-visible, is not Radiation from the source is split into two beams, half
sufficiently energetic to cause photochemical decomposition of passing into the sample-cell compartment and the other half
the sample. Placing the sample and reference before the mono- into the reference area. The reference beam then passes through
chromator, however, has the advantage that most scattered radi- the attenuator and on to a chopper. The chopper consists of a
ation and IR emission, generated within the cell compartment, motor-driven disk that alternately reflects the reference or trans-
is effectively removed by the monochromator and thus does not mits the sample beam into the monochromator. After dispersion
reach the transducer. by a prism or grating, the alternating beams fall on the trans-
Figure 16-11 shows schematically the arrangement of ducer and are converted to an electrical signal. The signal is
components in a typical IR spectrophotometer. Like many amplified and passed to the synchronous rectifier, a device that
inexpensive dispersive IR instruments, it is an optical null type, is mechanically or electrically coupled to the chopper to cause
16B IR Instrumentation 403

Synchronous
motor
Chart

Synchronous
motor
Attenuator
Reference

Grating
Source
Transducer
Chopper
Sample
Monochromator

Filter, Preamp
modulator, Synchronous
amplifier rectifier

Figure 16-11 Schematic diagram of a double-beam, dispersive


IR s­ pectrophotometer. The heavy black lines indicate mechanical linkages,
and the light lines indicate electrical connections. The radiation path is
­designated by dashed lines.

the rectifier switch and the beam leaving the chopper to change illustrates transmittance overshoot and rounded peaks in regions
simultaneously. If the two beams are identical in power, the sig- of low transmittance (1700 and 3000 cm21).
nal from the rectifier is constant direct current. If, on the other
hand, the two beams differ in power, a fluctuating, or ac, current
is produced, the phase of which is determined by which beam is
16B-3 Nondispersive Instruments
the more intense. The current from the rectifier is filtered and Several simple, rugged instruments have been designed for
further amplified to drive a synchronous motor in one direction quantitative IR analysis. Some are simple filter photometers,
or the other, depending on the phase of the input current. The whereas others use filter wedges in lieu of a dispersing element
synchronous motor is mechanically linked to both the attenu- to provide entire spectra. Finally, some gas analyzers use no
ator and the pen drive of the recorder and causes both to move wavelength-­selection device at all. Generally, these instruments
until a null balance is achieved. A second synchronous motor are less complex, more rugged, easier to maintain, and less
simultaneously drives the chart and varies the wavelength. Fre- expensive than the instruments we have previously described in
quently, a mechanical linkage between the wavelength and slit this chapter.
drives varies the slit width to keep the radiant power reaching
Filter Photometers
the detector approximately constant.
Figure 16-13 shows a schematic diagram of a portable, IR
The reference-beam attenuator system, such as the one just filter photometer designed for quantitative analysis of various
described, creates three limitations to the performance of disper- substances in the atmosphere. 9 Different models are avail-
sive IR instruments. First, the response of the attenuator system able that are factory calibrated for 1, 5, 30, or more than 100
always lags behind the transmittance changes, particularly in scan- gases. The instrument is computer controlled. For many com-
ning regions where the signal is changing most rapidly. Second, pounds, fixed bandpass filters of 1.8, 3.3, 3.6, 4.0, 4.2, 4.5, 4.7,
the momentum associated with both the mechanical attenuator 8, 11, and 14 µm can be used. A continuously variable filter,
and the recorder system may result in the pen drive overshooting which transmits in the range between about 7.7 and 14.1 µm
the true transmittance. Third, in regions where the transmittance
approaches zero, almost no radiation reaches the transducer, and
the exact null position cannot be established accurately. The result
is sluggish transducer response and rounded peaks. Figure 16-12 9
See P. A. Wilks, Amer. Lab., 1994 (12), 44; Proc. Control Qual., 1992, 3, 283.
404 Chapter 16 An Introduction to Infrared Spectrometry

Wavenumber, cm–1
4000 3000 2000 1500 1000 900 800

100

80
Transmittance, %

60

40

20

0
3 4 5 6 7 8 9 10 11 12 13 14
Wavelength, μm

Figure 16-12 IR spectrum of n-hexanal illustrating overshoot at low %T.

Sampling wand

Transducer

Variable-path-length cell

Filter

Chopper

Source

Pump

Figure 16-13 A portable IR photometer designed for gas analysis. The sample is introduced into the cell by means of a pump. The
path length can be changed by altering the number of mirror reflections. (Courtesy of Thermo Fisher Scientific, Waltham, MA.)
16B IR Instrumentation 405

Reference Sample mixture. The reference cell is a sealed container filled with a
source source nonabsorbing gas; as shown in the figure, the sample flows
through a second cell that is of similar length. The chopper
blade is so arranged that the beams from identical sources are
Chopper chopped simultaneously at the rate of about five times per sec-
ond. Selectivity is obtained by filling both compartments of the
sensor cell with the gas being analyzed, CO in this example.
Beam attenuator
The two chambers of the detector are separated by a thin, flex-
Sample ible, metal diaphragm that serves as one plate of a capacitor;
the second plate is contained in the sensor compartment on
Reference Sample the left.
cell cell In the absence of CO in the sample cell, the two sen-
sor chambers are heated equally by IR radiation from the two
sources. If the sample contains CO, however, the right-hand
Capacitor
beam is attenuated somewhat and the corresponding sensor
plate
Sensor chamber becomes cooler with respect to its reference coun-
Sensor cell filled terpart. As a result, the diaphragm moves to the right and the
with CO capacitance of the capacitor changes. This change in capacitance
Cell filled Flexible metal is sensed by the amplifier system. The amplifier output drives a
with CO To amplifier diaphragm servomotor that moves the beam attenuator into the reference
beam until the two compartments are again at the same tem-
Figure 16-14 A nondispersive IR photometer for monitoring CO.
perature. The instrument thus operates as a null balance device.
The chopper serves to provide an ac signal, which is less sensi-
tive to drift and 1/f noise.
(1300 to 710 cm21), can also be used for selecting alternative This type of instrument is highly selective because heating
wavelengths or for spectral scanning applications. The source of the sensor gas occurs only from that narrow portion of the
is a nichrome-wire filament, and the transducer is a pyroelec- spectrum of radiation absorbed by the CO in the sample. The
tric device (see Section 16C for descriptions of sources and device can be adapted to the determination of any IR-absorbing
transducers). gas (e.g., CO2, NO, or H2S).12
The gaseous sample is brought into the cell by means of a
battery-operated pump at a rate of 20 L/min. In the cell, three Filter Correlation Analyzers
gold-plated mirrors are used in a folded-path-length design. As shown in Figure 16-15, filter correlation analyzers use a rotat-
Path lengths of 0.5 m and 12.5 m may be selected. Detection ing gas filter through which the IR beam passes. The filter has
of many gases at sub‑parts-per-million levels, particularly with two compartments, one for the gas of interest and the other for a
the long-path-length setting, have been reported with this nonabsorbing gas such as nitrogen. When the gas of interest is in
­photometer. the beam, it selectively attenuates the IR source to produce a ref-
erence beam. The sample beam is produced by the transparent
A recently designed filter IR instrument has been reported
gas. Typically, the IR source is chopped at a fairly high frequency
for determination of CO2 in industrial settings. The estimated
(360 Hz) whereas the filter rotates at a fairly low frequency
limit of detection was 1 ppm.10
(30 Hz). A modulated signal is produced that is related to the
Photometers without Filters concentration of the analyte gas.
Photometers, which have no wavelength-restricting device, Filter correlation analyzers are available for such gases as
are widely used to monitor gas streams for a single compo- CO2 and CO. They can be set up to detect trace levels (,0.1
nent.11 Figure 16-14 shows a typical nondispersive instrument ppm) or higher amounts. The analyzers are calibrated by con-
designed to determine carbon monoxide (CO) in a gaseous structing a working curve by dilution of a gas standard.

10
See J. Hodgkinson et al., Sens. Actuators B, 2013, 186, 580, DOI: 10.1016/j. 12
For CO2 in seawater, see Marine Chemistry, 2006, 100, 24, DOI: 10.1016/j
snb.2013.06.006. .marchem.2005.10.009; for NO in cigarette smoke, see reference 11; for H2S in
11
For the determination of CO in cigarette smoke, see Appl. Spectrosc., 2006, 60, acidic alkylamine gas streams, see Hydrocarbon Processing, 2009, 88 (February,
272, DOI: 10.1366/000370206776342616. issue 2), 73.
406 Chapter 16 An Introduction to Infrared Spectrometry

Pressure
transducer

Electronics

Preamplifier

IR transducer

Sample
Valve

N2 CO2 Gas filter wheel


Bandpass filter
Purge Chopper

IR source Chopper motor

Exhaust

Capillary

Pump

Figure 16-15 An IR filter correlation analyzer for determining CO2. The sample is drawn into the sample cell by a pump. The
chopped IR source radiation (360 Hz) alternates between the N2 and CO2 sides of the filter wheel, which rotates at 30 Hz. The
CO2 side provides a reference beam that cannot be further attenuated by CO2 in the sample cell. The N2 side produces the sample
beam by allowing the IR radiation to pass through to the cell where it can be attenuated by CO2 in the sample. The modulated
amplitude of the chopped detector signal is related to the CO2 concentration in the sample. Other gases do not modulate
the detector signal because they absorb the reference and sample beams equally. (Courtesy of Thermo Fisher Scientific Inc.,
Waltham, MA.)

16C IR Sources and Transducers The Nernst Glower


The Nernst glower is composed of rare earth oxides formed into
Instruments for measuring IR absorption all require a source a cylinder having a diameter of 1 to 3 mm and a length of 2 to
of continuous IR radiation and an IR transducer. The desir- 5 cm. Platinum leads are sealed to the ends of the cylinder to
able characteristics of these instrument components were listed permit electrical connections to what amounts to a resistive
in Sections 7B and 7E. In this section we describe sources and heating element. As current passes through the device, its tem-
transducers that are found in modern IR instruments. perature increases to between 1200 K and 2200 K. The Nernst
glower has a large negative temperature coefficient of electrical
16C-1 Sources resistance, and it must be heated externally to a dull red heat
IR sources consist of an inert solid that is heated electrically to before the current is large enough to maintain the desired tem-
a temperature between 1500 and 2200 K. These sources pro- perature. Because the resistance decreases with increasing tem-
duce continuum radiation approximating that of a blackbody perature, the source circuit must be designed to limit the current
(see Figure 6-22). The maximum radiant intensity at these tem- or the glower rapidly becomes so hot that it is destroyed.
peratures occurs between 5000 and 5900 cm21 (2 and 1.7 µm). Figure 16-16 shows the spectral output of a Nernst glower
At longer wavelengths, the intensity falls off smoothly until it is operated at approximately 2200 K. Note that the overall curve
about 1% of the maximum at 670 cm21 (15 µm). On the short shape is similar to that of a blackbody. The small peaks and
wavelength side, the decrease is much more rapid, and a similar depressions are a function of the chemical composition of the
reduction in intensity is observed at about 10,000 cm21 (1 µm). device.
16C IR Sources and Transducers 407

1 atmosphere. When an electric current passes through the


1000 vapor, an internal plasma source is formed that provides contin-
uum radiation in the far-IR region.
Energy, arbitrary units

100 The Tungsten Filament Lamp


An ordinary tungsten filament lamp is a convenient source for
the near-IR region of 4000 to 12,800 cm21 (2.5 to 0.78 µm).
10
The Carbon Dioxide Laser Source
A tunable carbon dioxide laser is used as an IR source for mon-
0 1 2 3 4 5 6 7 8 9 10 11 itoring the concentrations of certain atmospheric pollutants
l , μm and for determining absorbing species in aqueous solutions.13
A carbon dioxide laser produces a band of radiation in the range
Figure 16-16 Spectral distribution of energy from a Nernst glower of 900 to 1100 cm21 (11 to 9 µm), which consists of about 100
operated at approximately 2200 K. closely spaced discrete lines. As described in Section 7B-3, any
one of these lines can be chosen by tuning the laser. Although
The Globar the range of wavelengths available is limited, the region from
A Globar is a ceramic rod, usually silicon carbide or a composite 900 to 1100 cm21 is one particularly rich in absorption bands
material, electrically heated to more than 1000 K. Older disper- arising from the interactive stretching modes of CO2. Thus, this
sive instruments used Globars that were heated as high as 1500 K source is useful for quantitative determination of a number of
to provide sufficient intensity. The sensitivity of modern FTIR important species such as ammonia, butadiene, benzene, eth-
instruments has meant that lower source temperatures (1000 to anol, nitrogen dioxide, and trichloroethylene. An important
1300 K) can be used, resulting in longer source lifetimes. property of the laser source is the radiant power available in
­Globar sources for FTIR must be temperature controlled each line, which is several orders of magnitude greater than that
for high stability. Some simple sources in older instruments used of blackbody sources.
a fan to provide air cooling, but water cooling is more effective. Carbon dioxide lasers are widely used in remote-sensing
The Globar has the advantage of a positive temperature coeffi- applications such as light detection and ranging (lidar). The
cient of resistance that makes it possible to control temperature operating principle of lidar is similar to that of radar. The lidar
electronically. The most stable sources, therefore, use electronic system transmits radiation out to a target where it interacts
temperature control, which eliminates the inconvenience and with and is altered by the target. Some of the radiation is then
temperature fluctuations of circulating water. Electronic tem- reflected back to the lidar instrument where it is analyzed and
perature control is also said to increase the source lifetime. Spec- used to obtain information about the target. By means of lidar,
tral energies of the Globar and Nernst glower are similar, except distance, speed, rotation, chemical composition, and concentra-
in the region below 5 μm, where the Globar is significantly more tion of remote targets can be obtained.
intense. Spectral energies of the Globar and the Nernst glower
are comparable except in the region below 5 µm, where the Glo- Semiconductor IR Laser Sources
bar provides a significantly greater output. Tunable semiconductor lasers have also been used as sources
(see Section 7B-3 for a description of these devices) in the near-
Incandescent Wire Source and mid-IR regions. As an example, the Sample Analysis at
A source with somewhat lower intensity but with longer life Mars (SAM) experiment on the Mars rover Curiosity’s ­Science
than the Globar or Nernst glower is a tightly wound spiral of ­L aboratory uses a tunable diode laser spectrometer in the
nichrome wire heated to about 1100 K by an electrical current. 2.78 μm region to detect carbon, hydrogen, and oxygen in the
A rhodium wire heater sealed in a ceramic cylinder has similar ­Martian environment and to determine their isotopic ratios.
properties, although it is more expensive. Nichrome wire sources A quantum cascade tunable laser is used to detect methane
are less intense than many IR sources. However, the incandes- by scanning across several of its rotational lines in the 3.3 μm
cent wire source requires no cooling and is nearly maintenance region. Both lasers are used in Curiosity’s two-channel tunable
free. For this reason, the nichrome wire source is often used laser spectrometer (TLS) that can achieve a spectral resolution
where reliability is paramount, such as in process analyzers. of 0.00002 cm−1. Methane is determined by the ­difference in

The Mercury Arc


For the far-IR region of the spectrum (l . 50 µm), none of the
13
thermal sources just described provides sufficient radiant power See A. A. Demidov, in Introduction to Laser Spectroscopy, 2nd ed., D. L. Andrews
and A. A. Demidov, eds., New York: Kluwer Academic/Plenum Press, 2002;
for convenient detection. In this region, a high-pressure mer-
Z. Zelinger, M. Strizik, P. Kubat, and S. Civis, Anal. Chim. Acta, 2000, 422,
cury arc is often used. This device consists of a quartz-­jacketed 179, DOI: 10.1016/S0003-2670(00)01069-2; P. L. Meyer, M. W. Sigrist, Rev. Sci.
tube containing mercury vapor at a pressure greater than Instrum., 1990, 61, 1779, DOI: 10.1063/1.1141097.
408 Chapter 16 An Introduction to Infrared Spectrometry

the a­ bsorption signals obtained from a closed sample cell filled to protect the semiconductor from the atmosphere. Absorption
with the Mars atmosphere and the same cell evacuated by of radiation by these materials promotes nonconducting valence
means of a turbo-molecular pump. How much of the methane electrons to a higher energy-conducting state, thus decreasing
detected arises from a chemical source in the Martian atmo- the electrical resistance of the semiconductor. Typically, a pho-
sphere and how much comes from the rover itself remains an toconductor is placed in series with a voltage source and load
open question. resistor, and the voltage drop across the load resistor serves as a
measure of the power of the beam of radiation.

16C-2 IR Transducers A lead sulfide photoconductor is the most widely used


IR transducers are of three general types: (1) pyroelectric trans- transducer for the near-IR region of the spectrum from 10,000
ducers, (2) photoconducting transducers, and (3) thermal to 333 cm21 (1 to 3 µm). It can be operated at room temperature.
transducers. The first is found in photometers, many FTIR spec- For mid- and far-IR radiation, MCT photoconductor transduc-
trometers, and dispersive spectrophotometers. Photoconducting ers are used. They must be cooled with liquid nitrogen (77 K) to
transducers are also found in FTIR instruments. Thermal detec- minimize thermal noise. The long-wavelength cutoff, and many
tors are found in older dispersive instruments but are too slow to of the other properties of these transducers, depend on the ratio
be used in FTIR spectrometers. of the mercury telluride to cadmium telluride, which can be var-
ied continuously.
Pyroelectric Transducers The MCT transducer is faster and more sensitive than the
Pyroelectric transducers are constructed from single crystalline DTGS transducer discussed in the previous section. For this
wafers of pyroelectric materials, which are insulators (dielec- reason, the MCT transducer also finds widespread use in FTIR
tric materials) with very special thermal and electrical prop- spectrometers, particularly those requiring fast response times,
erties. Triglycine sulfate (NH2CH2COOH)3 ? H2SO4 (usually such as spectrometers interfaced to gas chromatographs and
­deuterated, DTGS, and often with a fraction of the glycines those intended for kinetic measurements.
replaced with alanine, DLaTGS), is the most important pyro-
electric material used for IR detection systems. Thermal Transducers
When an electric field is applied across any dielectric Thermal transducers, whose responses depend on the heating
material, polarization takes place, with the magnitude of the effect of radiation, are found in older dispersive spectrometers
polarization being a function of the dielectric constant of the for detection of all but the shortest IR wavelengths. With these
material. For most dielectrics, this induced polarization rapidly devices, the radiation is absorbed by a small blackbody and the
decays to zero when the external field is removed. Pyroelectric resultant temperature rise is measured. The radiant power level
substances, in contrast, retain a strong temperature-dependent from a spectrophotometer beam is minute (1027 to 1029 W),
polarization after removal of the field. Thus, by sandwiching so that the heat capacity of the absorbing element must be as
the pyroelectric crystal between two electrodes, one of which is small as possible if a detectable temperature change is to be pro-
IR transparent, a temperature-dependent capacitor is produced. duced. Under the best of circumstances, temperature changes
Changing its temperature by irradiating it with IR radiation are confined to a few thousandths of a kelvin.
alters the charge distribution across the crystal, which can be The problem of measuring IR radiation by thermal means
detected as a current in an external electrical circuit connecting is compounded by thermal noise from the surroundings. For
the two sides of the capacitor. The magnitude of this current is this reason, thermal transducers are housed in a vacuum and
proportional to the surface area of the crystal and the rate of carefully shielded from thermal radiation emitted by other
change of polarization with temperature. Pyroelectric crystals nearby objects. To further minimize the effects of extraneous
lose their residual polarization when they are heated to a tem- heat sources, the beam from the source is always chopped. In
perature called the Curie point. For triglycine sulfate, the Curie this way, the analyte signal, after transduction, has the frequency
point is 47°C. of the chopper and can be separated electronically from extrane-
Pyroelectric transducers exhibit response times that are ous noise signals, which are ordinarily broad band or vary only
fast enough to allow them to track the changes in the time-­ slowly with time.
domain signal from an interferometer. For this reason, many
FTIR spectrometers for the mid-IR region use this type of Thermocouples. In its simplest form, a thermocouple consists
transducer. of a pair of junctions formed when two pieces of a metal such as
bismuth are fused to each end of a dissimilar metal such as anti-
Photoconducting Transducers mony. A potential difference between the two junctions varies
IR photoconducting transducers consist of a thin film of a semi- with their difference in temperature.
conductor material, such as lead sulfide, mercury telluride–­ The transducer junction for IR radiation is formed from
cadmium telluride (MCT), or indium antimonide, deposited on a very fine wires or alternatively by evaporating the metals onto
nonconducting glass surface and sealed in an evacuated ­envelope a nonconducting support. In either case, the junction is usually
Questions and Problems 409

blackened (to improve its heat-absorbing capacity) and sealed in Bolometers. A bolometer is a type of resistance thermome-
an evacuated chamber with a window that is transparent to IR ter constructed of strips of metals, such as platinum or nickel,
radiation. The response time is typically about 30 ms. or from a semiconductor. This type of semiconductor device is
The reference junction, which is usually housed in the same sometimes called a thermistor. Semiconductor materials exhibit
chamber as the active junction, is designed to have a relatively a relatively large change in resistance as a function of tempera-
large heat capacity and is carefully shielded from the incident ture. The responsive element in a bolometer is kept small and
radiation. Because the analyte signal is chopped, only the dif- blackened to absorb the radiant heat. Bolometers are not so
ference in temperature between the two junctions is important; extensively used as other IR transducers for the mid-IR region.
therefore, the reference junction does not need to be maintained However, a germanium bolometer, operated at 1.5 K, is an
at constant temperature. To enhance sensitivity, several thermo- excellent transducer for radiation in the range of 5 to 400 cm21
couples may be connected in series to produce a thermopile. (2000 to 25 µm). The response time is a few milliseconds.

Questions and Problems


*Answers are provided at the end of the book for problems marked with an asterisk.
Problems with this icon are best solved using spreadsheets.

* 16-1 The IR spectrum of CO shows a vibrational absorption band centered at 2170 cm21.
What is the force constant for the CO bond?
(a)
At what wavenumber would the corresponding peak for 14CO occur?
(b)
* 16-2 Gaseous HCl exhibits an IR absorption at 2890 cm21 due to the hydrogen-chlorine stretching vibration.
Calculate the force constant for the bond.
(a)
Calculate the wavenumber of the absorption band for DCl assuming the force constant is the same as
(b)
that calculated in part (a).
16-3 Calculate the absorption frequency corresponding to the iCiH stretching vibration treating the group
as a simple diatomic CiH molecule with a force constant of k 5 5.0 3 102 N/m. Compare the calculated
value with the range found in correlation charts (such as the one shown in Figure 17-6). Repeat the
calculation for the deuterated bond.
16-4 The force constant of the NO bond is approximately 1.53 3 103 N/m. Treating the NO group as a simple
diatomic molecule, calculate the absorption frequency in Hz and the wavelength and wavenumber of the
fundamental absorption. How does this compare to the experimental value found for NO and NO dimers
by Varetti, E. L.; Pimentel, G. C., J. Chem. Phys., 1971, 55, 3813, DOI: 10.1063/1.1676666.
* 16-5 The wavelength of the fundamental OiH stretching vibration is about 3.0 µm. What is the approximate
wavenumber and wavelength of the first overtone band for the OiH stretch?
* 16-6 The wavelength of the first overtone of the NiH stretching vibration is about 1.5 µm. What is the
approximate wavenumber and wavelength of the fundamental band for the NiH stretch?
* 16-7 Sulfur dioxide is a nonlinear molecule. How many vibrational modes will this compound have? How many
IR absorption bands would sulfur dioxide be expected to have?
410 Chapter 16 An Introduction to Infrared Spectrometry

Questions and Problems (continued)


* 16-8 Indicate whether the following vibrations are active or inactive in the IR spectrum.

Molecule Motion
(a) CH3 CH3 C C stretching
(b) CH3 CCl3 C C stretching
(c) SO2 Symmetric stretching
(d) CH2 CH2 C H stretching:

H H
C C
H H

(e) CH2 CH2 C H stretching:

H H
C C

H H

(f) CH2 CH2 CH2 wag:


H H
C C

H H

(g) CH2 CH2 CH2 twist:


H H
C C

H H

16-9 What are the major advantages of an FTIR spectrometer compared with a dispersive instrument?
16-10 What length of mirror drive in an FTIR spectrometer would be required to provide a resolution of
(a) 0.010 cm21, (b) 0.50 cm21, and (c) 2.0 cm21?
* 16-11 It was stated that at room temperature (25°C) the majority of molecules are in the ground vibrational energy
level 1 v 5 0 2 .
Use the Boltzmann equation (Equation 8-1) to calculate the excited-state and ground-state population
(a)
ratios for HCl: N 1 v 5 1 2 /N 1 v 5 0 2 . The fundamental vibrational frequency of HCl occurs at 2885 cm21.
Use the results of part (a) to find N 1 v 5 2 2 /N 1 v 5 0 2
(b)
16-12 Why are nondispersive IR instruments often used for the determination of gases rather than dispersive
IR spectrometers?
16-13 The first FTIR instruments used three different interferometer systems. Briefly, describe how it has been
possible to simplify the optical systems in more contemporary instruments.
* 16-14 In a particular trace analysis via FTIR, a set of sixteen interferograms were collected. The signal-to-noise
ratio (S/N) associated with a particular spectral peak was approximately 4:1. How many interferograms
would have to be collected and averaged if the goal is to obtain an S/N 5 20:1?
Questions and Problems 411

* 16-15 Radiation leaves a source at a natural frequency of (a) 9.10 3 1013 Hz. What is the wavenumber
corresponding to this frequency? If a Michelson interferometer has a mirror velocity of 1.00 cm/s, what will
the frequency be at the transducer. Find the wavenumber at the source and frequency at the transducer for
source radiation of (b) 9.20 3 1013 Hz and (c) 9.30 3 1013 Hz.

Challenge Problem
16-16 (a) The IR spectrum of gaseous N2O shows three strong absorption bands at 2224 cm21, 1285 cm21, and
2089 cm21. In addition two quite weak bands are observed at 2563 cm21 and 2798 cm21. It is known
that N2O is a linear molecule, but assume it is not known whether the structure is NiNiO or
NiOiN. Use the IR data to decide between the two structures. What vibrations can be assigned to
the strong absorption bands? What are possible causes of the weak absorptions?
The IR spectrum of HCN shows three strong absorption bands at 3312 cm21, 2089 cm21, and 712 cm21.
(b)
From this information alone, can you deduce whether HCN is linear or nonlinear? Assuming that
HCN is linear, assign vibrations to the three absorption bands.
How many fundamental vibrational modes are expected for BF3? Which of these are expected to be IR
(c)
active? Why? Sketch the vibrations.
How many fundamental vibrational modes would you predict for (1) methane, (2) benzene,
(d)
(3) toluene, (4) ethylene, and (5) carbon tetrachloride?
chapterSEVENTEEN

Applications of Infrared
Spectrometry

I
nfrared (IR) absorption spectrometry is a As shown in Table 17-1, the applications of IR spectrometry fall
­versatile tool that is applied to the qualitative into three major categories based on the three IR spectral regions.
The most widely used region is the mid-IR, which extends from
and q­ uantitative determination of m­ olecular
about 670 to 4000 cm21 (2.5 to 14.9 µm). In this region, absorp-
species of all types. In this chapter we first focus tion, reflection, and emission spectra are used for both quali-
on the uses of mid-IR absorption and ­reflection tative and quantitative analysis. The near-IR region, from 4000
­spectrometry for structural investigations of to 14,000 cm21 (0.75 to 2.5 µm), also finds considerable use for
­molecular compounds, particularly organic c­ ompounds the routine quantitative determination of certain species, such
as water, carbon dioxide, sulfur, low-molecular-mass hydrocar-
and species of i­nterest in biochemistry. We then
bons, amine nitrogen, and other compounds of interest in agri-
examine several of the other applications of IR culture and in industry. These determinations are often based
­spectroscopy, including p­ hotoacoustic IR, near- and on diffuse-reflectance measurements of untreated solid or liquid
far-IR s­ pectroscopy, IR emission, and IR microscopy. samples or absorption studies of gases. The primary use of the
far-IR region (15 to 1000 µm) has been for the determination of
the structures of inorganic and m­ etal-organic species based on
absorption measurements.

17A Mid-IR Absorption


Spectrometry
Mid-IR absorption and reflection spectrometry are major tools
for determining the structure of organic and biochemical spe-
cies. In this section we examine mid-IR absorption applications.
Section 17B is devoted to mid‑IR reflectance measurements.1

17A-1 Sample Handling


As we have seen in earlier chapters, ultraviolet (UV) and
­visible molecular spectra are obtained most conveniently from
dilute solutions of the analyte. Absorbance measurements in
the ­optimal range are obtained by suitably adjusting either the

1
For further reading see R. M. Silverstein, F. X. Webster, D. Kiemle, and
D. L. Bryce, Spectrometric Identification of Organic Compounds, 8th ed., Chap. 2,
New York: Wiley, 2015; Introduction to Experimental Infrared Spectroscopy: Fun-
damentals and Practical Methods, M. Tasumi, ed., Chichester, UK: Wiley, 2015;
Throughout this chapter, this logo indicates P. Larkin, Infrared and Raman Spectroscopy: Principles and Spectral Interpretation,
an ­opportunity for online self-study at Waltham, MA: Elsevier, 2011; B. Schrader, Infrared and Raman Spectroscopy, New
www.tinyurl.com/skoogpia7, linking you to York: VCH, 1995; N. B. Colthup, L. H. Daly, and S. E. Wiberley, Introduction to
­interactive tutorials, ­simulations, and exercises. Infrared and Raman Spectroscopy, 3rd ed., San Diego: Academic Press, 1990.

412
17A Mid-IR Absorption Spectrometry 413

Table 17-1 Major Applications of IR Spectrometry


Spectral Regions Measurement Type Kind of Analysis Applicable Samples
Near-IR Diffuse reflectance Quantitative Solid or liquid commercial materials
Absorption Quantitative Gaseous mixtures
Mid-IR Absorption Qualitative Pure solid, liquid, or gases
Quantitative Complex liquid, solid, or gaseous mixtures
Chromatographic Complex liquid, solid, or gaseous mixtures
Far-IR Reflectance Qualitative Pure solids or liquids
Emission Quantitative Atmospheric samples
Absorption Qualitative Pure inorganic or organometallic species

concentration or the cell length. Unfortunately, this approach Solutions


is often not applicable for IR spectroscopy because no good When feasible, a convenient way of obtaining IR spectra is
solvents are transparent throughout the region of interest. As on solutions prepared to contain a known concentration of
a result, sample handling is frequently the most difficult and sample, as is generally done in UV-visible spectrometry. This
time-consuming part of an IR spectrometric analysis.2 In this technique is somewhat limited in its applications, however, by
section we outline some of the common techniques for prepara- the availability of solvents transparent over significant regions
tion of samples for IR absorption measurements. in the IR.
Gases Solvents. Figure 17-1 lists several common solvents used for IR
The spectrum of a low-boiling-point liquid or gas can be studies of organic compounds. This figure illustrates that no sin-
obtained by permitting the sample to expand into an evacuated gle solvent is transparent throughout the entire mid-IR region.
cylindrical cell equipped with suitable windows. For this pur-
pose, a variety of cylindrical cells are available with path lengths Water and the alcohols are difficult to use as solvents in IR
that range from a few centimeters to 10 m or more. The longer spectrometry. Water shows several strong absorption bands in
path lengths are obtained in compact cells by providing reflect- the IR region, as can be seen in Figure 17-2. The spectrum of
ing internal surfaces, so that the beam makes numerous passes water is shown along with the spectrum of an aqueous solution
through the sample before exiting the cell (see Figure 16-13). of aspirin. The computer-calculated difference spectrum reveals
the spectrum of the water-soluble aspirin. Water and alcohols
also attack alkali-metal halides, the most common materials
2
See Handbook of Spectroscopy, G. Gauglitz and D. S. Moore, eds, Vol. 1, used for cell windows. Hence, water-insoluble window materi-
Section 1, Weinheim, Germany: Wiley-VCH, 2014; Practical Sampling Tech-
als, such as barium fluoride, must be used with such solvents.
niques for ­Infrared Analysis, P. B. Coleman, ed., Boca Raton, FL: CRC Press, 1993;
T. J. Porro and S. C. Pattacini, Spectroscopy, 1993, 8 (7), 40; Ibid., 8 (8), 39; Care must also be taken to dry the solvents shown in Figure 17-1
A. L. Smith, Applied Infrared Spectroscopy, New York: Wiley, 1979, Chap. 4. before use with typical cells.

Wavenumber, cm–1
5000 2500 1667 1250 1000 833 714 625
Carbon disulfide
Carbon tetrachloride
Tetrachloroethylene
Chloroform
Dimethylformamide
Dioxane
Cyclohexane
Benzene
2 4 6 8 10 12 14 16
Wavelength, μm

FIGURE 17-1 Solvents for IR spectroscopy. Horizontal lines indicate


useful regions.
414 Chapter 17 Applications of Infrared Spectrometry

expensive. Even with care, however, their surfaces eventually


MAX 5 100.00T
become fogged because of absorption of moisture. Polishing
0.015-mm BaF2 cell with a buffing powder can, in many cases, return them to their
Percent transmittance

original condition.
Difference
For samples that are wet or aqueous, calcium and barium
fluoride are suitable, although neither transmits well throughout
Water the entire mid-IR region. Silver bromide is often used, although

Water-
soluble
aspirin
MIN 5 0.00T Table 17-2 Common IR Window Materials
4000 3500 3000 2500 2000 1500 1000 500
Wavenumber, cm21
Applicable Water Solubility,
Window Range, g/100 g H2O,
FIGURE 17-2 IR spectra of water and of an aqueous solution Material cm21 20°C
­containing water-soluble aspirin. The top spectrum shows the
­computer-generated difference spectrum obtained by s­ ubtracting Sodium chloride 40,000–625 36.0
the water spectrum from that of the solution. (From Practical
Potassium bromide 40,000–385 65.2
­Sampling Techniques for Infrared Analysis, P. B. Coleman, ed., Boca
Raton, FL: CRC Press, 1993. With permission.) Potassium chloride 40,000–500 34.7
Cesium iodide 40,000–200 160.0
Fused silica 50,000–2,500 Insoluble
Cells. Because of the tendency for solvents to absorb IR radia- Calcium fluoride 50,000–1,100 1.51 3 1023
tion, IR liquid cells are ordinarily much narrower (0.01 to 1 mm)
Barium fluoride 50,000–770 0.12 (25°C)
than those used in the UV and visible regions. Often, relatively
high sample concentrations (from 0.1% to 10%) are required Thallium bromide 16,600–250 ,0.0476
because of the short path lengths and the low molar absorptiv- iodide, KRS-5
ities of IR bands. Liquid cells are frequently designed for easy Silver bromide 20,000–285 1.2 3 1027
disassembly and use Teflon spacers to allow variation in path
Zinc sulfide, 10,000–715 Insoluble
length (see Figure 17-3). Fixed-path-length cells can be filled or
Irtran-2
emptied with a hypodermic syringe.
Zinc selenide, 10,000–515 Insoluble
A variety of window materials are available as listed in
Irtran-4
Table 17-2. Window materials are often chosen on the basis of
cost, range of transparency, solubility in the solvent, and reactiv- Polyethylene 625–30 Insoluble
ity with the sample or solvent. Sodium chloride and potassium From Practical Sampling Techniques for Infrared Analysis, P. B. Coleman, ed., Boca
bromide windows are most commonly used and are the least Raton, FL: CRC Press, 1993.

Inlet tubes
Sample
Back
plate

Window
Neoprene
gasket Window
Spacer
Neoprene
gasket
Front plate Quick-acting
nuts (4)

FIGURE 17-3 Expanded view of a demountable IR cell for liquid samples.


Teflon spacers ranging from 0.015 to 1 mm thick are available. (Courtesy of
Perkin-Elmer Corp., Norwalk, CT.)
17A Mid-IR Absorption Spectrometry 415

As shown in Figure 17-4b, the number of interference fringes


DN between two known wavelengths l1 and l2 are counted and
introduced into the equation
2000 1400 900
cm – 1 2b 2b
(a)
DN 5 2 5 2bn1 2 2bn2
l1 l2

or
Wavelength, cm –1 DN
3200 2800 2400 2000
b5 (17-2)
2 1 n1 2 n2 2
100
Δ N = 12 It should be noted that interference fringes are ordinarily not
seen when a cell is filled with liquid because the refractive
90
index of most liquids approaches that of the window mate-
Transmittance, %

rial and reflection is minimized (Equation 6-15). On the other


80 hand, interference can be observed between 2800 and 2000
ΔN cm 21 in the spectrum shown in Figure 16-1. The sample is
b=
2( –1 – –2 ) a sheet of polystyrene, which has a refractive index consid-
70
12
erably different from that of air. For this reason, significant
= = 0.0052 cm reflection occurs at the two interfaces of the sheet. Equation
2(3230 – 2080)
60 17-2 is often used to calculate the thickness of thin polymer
films.
(b)
Liquids
FIGURE 17-4 Determination of the path length of IR cells from When the amount of a liquid sample is small or when a suit-
interference fringes. Fringes are observed with an empty cell in the able solvent is unavailable, it is common practice to obtain
light path (a). In (b) an illustration of the calculation of the path spectra on the pure (neat) liquid. The thickness of a very thin
length is given.
film of liquid presents a sufficiently short path length to pro-
duce satisfactory spectra. Commonly, a drop of the neat liquid
is squeezed between two rock-salt plates to give a layer 0.015
mm thick or less. The two plates, held together by capillary
it is more expensive, is photosensitive, and is easily scratched. action, are then placed in the beam path. This technique does
The Irtran materials are also resistant to chemical attack and are not give particularly reproducible transmittance data, but
water insoluble. They are somewhat expensive and have high the resulting spectra are usually satisfactory for qualitative
indexes of refraction (2.2 to 2.4), which can lead to interference ­investigations.
fringes because of mismatches with the sample index of refrac-
tion (often around 1.4 to 1.6). Solids
Figure 17-4 shows how the thickness b of very narrow Most organic compounds exhibit numerous absorption bands
IR cells can be determined by the interference fringe method. throughout the mid-IR region, and finding a solvent that does
The interference pattern of Figure 17‑4a is obtained by insert- not have overlapping bands is often impossible. Because of this,
ing an empty cell in the light path of the sample beam and spectra are often obtained on dispersions of the solid in a liq-
recording the transmittance against air in the reference beam. uid or solid matrix. Generally, in these techniques, the solid
The reflected radiation from the cell walls interferes with the sample must be ground until its particle size is much less than
transmitted beam to produce the interference maxima and the wavelength of the radiation to avoid the effects of scattered
minima. The maxima occur when the radiation reflected radiation.
from the two internal surfaces of the cell has traveled a dis-
Pelleting. One of the most popular techniques for handling
tance that is an integral multiple N of the wavelength of the
solid samples has been KBr pelleting (other alkali-metal halides
radiation transmitted without reflection. Constructive inter-
have also been used). Halide salts have the property of cold flow,
ference then occurs whenever the wavelength is equal to
in which they have glasslike transparent or translucent proper-
2b/N. That is,
ties when sufficient pressure is applied to the finely powdered
2b materials. In using this technique, a milligram or less of the
5 l (17-1)
N finely ground sample is intimately mixed with about 100 mg of
416 Chapter 17 Applications of Infrared Spectrometry

dried, spectroscopic grade, potassium bromide powder. Mixing the same kinds of information. Reflectance and photoacoustic
can be carried out with a mortar and pestle or, better, in a small methods are discussed in Sections 17B and 17C.
ball mill. The mixture is then pressed in a special die at 10,000 to
15,000 pounds per square inch to yield a transparent disk. Best
results are obtained if the disk is formed in a vacuum to elimi-
17A-2 Qualitative Analysis
nate occluded air. The disk is then held in the instrument beam The general use of mid-IR spectroscopy by chemists for iden-
for spectroscopic examination. The resulting spectra frequently tifying organic compounds began in the late 1950s with the
exhibit bands at 3450 and 1640 cm21 (2.9 and 6.1 µm) due to appearance of inexpensive and easy-to-use double-beam record-
absorbed moisture. ing spectrophotometers that produce spectra in the region of
5000 to 670 cm21 (2 to 15 µm). The appearance of this type of
A good pellet spectrum should have a relatively flat instrument (as well as nuclear magnetic resonance and mass
(non-sloping) baseline with the intensities of the strongest spectrometers) revolutionized the way chemists identify organic,
bands in the 5–15%T (0.7–1.3 A) range. If the strongest bands inorganic, and biological species. The introduction of these tools
show smaller absorbances, this can be corrected by repeating dramatically reduced the time required to perform a qualitative
the formation of the pellet with a smaller amount of KBr. If the analysis or structural determination.
absorbances of the strongest bands are too high (e.g., totally
absorbing), the pellet can be broken into smaller fragments, and Figure 17-5 shows four typical spectra obtained with an
one of these can be repressed with more KBr. inexpensive double-beam instrument. Identification of an
organic compound from a spectrum of this kind is a two-step
With many compounds, KBr pelleting produces excellent process. The first step involves determining the functional
spectra that appear in many spectral libraries. Being ionic, KBr groups present by examining the group frequency region, which
transmits throughout most of the IR region with a lower cutoff of encompasses radiation from about 3600 cm21 to approximately
about 400 cm21. Ion exchange can occur with some samples such 1250 cm21 (see Figure 17-5). The second step involves a detailed
as amine hydrochlorides or inorganic salts. With the former, comparison of the spectrum of the unknown with the spectra
bands of the amine hydrobromide are often found. Polymor- of pure compounds that contain all of the functional groups
phism can also occur because of the forces involved in grinding found in the first step. The fingerprint region, from 1200 to
and pressing the pellet. These can convert one polymorph into 600 cm21 (Figure 17-5) is particularly useful here because small
another. Although KBr is the most frequently used pelleting salt, differences in the structure and constitution of a molecule result
materials such as CsI and CsBr are sometimes used. Cesium in significant changes in the appearance and distribution of
iodide has greater transparency at low frequencies than KBr. absorption bands in this region. Consequently, a close match
between two spectra in the fingerprint region (as well as others)
Mulls. IR spectra of solids that are not soluble in an IR‑trans- constitutes almost certain evidence that the two compounds are
parent solvent or are not conveniently pelleted in KBr are often identical.
obtained by dispersing the analyte in a mineral oil or a fluori-
nated hydrocarbon mull. To obtain good mull spectra, the sam- Group Frequencies
ple must be ground to an average particle size that is lower than We have noted that the approximate frequency (or wavenum-
about one-fifth of the shortest wavelength to be used. For many ber) at which an organic functional group, such as CwO, CwC,
mid-IR spectra, the starting point of the scan is 4000–5000 CiH, C{C, or OiH, absorbs IR radiation can be calculated
cm21. This corresponds to wavelengths of 2–2.5 μm. Hence, from the masses of the atoms and the force constant of the bond
grinding to a particle size lower than 0.4–0.5 μm is needed. between them (Equations 16-14 and 16-15). These frequencies,
Larger particles scatter radiation in proportion to the fourth called group frequencies, are seldom totally invariant because of
power of the radiation frequency and give rise to sloping base- interactions with other vibrations associated with one or both
lines. Often 20–30 mg samples are used for grinding, but sam- of the atoms composing the group. On the other hand, such
ples as small as 5–10 mg can be used. Grinding should result in interaction effects are ordinarily small; as a result, a range of
an even, smooth, glossy sample in the mortar. Mulls are then frequencies can be assigned within which it is highly probable
formed by adding a small drop of a heavy hydrocarbon (usu- that the absorption maximum for a given functional group will
ally Nujol™) and continuing the grinding process until a smooth be found. Table 17-3 lists group frequencies for several common
paste, similar in appearance to toothpaste or cold cream, results. functional groups. A more detailed presentation of group fre-
If hydrocarbon bands are likely to interfere, Fluorolube ™ , a quencies is found in the correlation chart shown in Figure 17-6.3
halogenated polymer, can be used instead. In either case, the Note that, although most group frequencies fall in the range of
resulting mull is then examined as a film between flat salt plates. 3600 to 1250 cm21, a few fall in the fingerprint region. These
include the CiOiC stretching vibration at about 1200 cm21
Other Methods for Solids. The IR behavior of solids can also and the CiCl stretching vibration at 700 to 800 cm21.
be obtained by casting a thin film of the sample as is done with
many polymers. For some samples, spectra can be acquired by 3
Other correlation charts can be found in R. M. Silverstein, F. X. Webster,
reflectance techniques and by the photoacoustic method. These D. Kiemle, and D. L. Bryce, Spectrometric Identification of Organic Compounds,
spectra are often similar to absorption spectra and provide 8th ed., Chap. 2, New York: Wiley, 2015.
17A Mid-IR Absorption Spectrometry 417

100 FIGURE 17-5 The group frequency region of the


mid-IR (,3600–1250 cm21) is used to identify
80
common functional groups. The fingerprint
region (~1200–600 cm21) is used to identify
Transmittance (%)

C H bend compounds. (Adapted from J. C. Gilbert and


60 S. F. Martin, Experimental Organic Chemistry:
CH 3 CH CHCH 3 A Miniscale and Microscale Approach, 6th ed.,
40 Figures 8.11 (a) and (b), 10.11 (c), and 14.7 (d),
CH 3 CH 3 Boston: Cengage Learning, 2016.)
C H stretch
20

0
4000 3600 3200 2800 2400 2000 1800 1600 1400 1200 1000 800 600
Wavenumber (cm–1)
(a)
100

80
C H bend
Transmittance (%)

60
CH3CH CHCH3
40
CH3 H
C H stretch
20

0
4000 3600 3200 2800 2400 2000 1800 1600 1400 1200 1000 800 600
Wavenumber (cm–1)
(b)
100

80
Transmittance (%)

60

CH 3CHCH 2CH(CH 3)2


40
OH C H bend
O H stretch
20

C H stretch
0
4000 3600 3200 2800 2400 2000 1800 1600 1400 1200 1000 800 600
Wavenumber (cm–1)
(c)
100

80
Transmittance (%)

60

Cl
40
(CH3)2CCH 2CH3
C Cl stretch
20
C H bend
C H stretch
0
4000 3600 3200 2800 2400 2000 1800 1600 1400 1200 1000 800 600
Wavenumber (cm–1)
(d)
Group frequency region Fingerprint region
418 Chapter 17 Applications of Infrared Spectrometry

TABLE 17-3 Abbreviated Table of Group Frequencies for Organic Functional Groups
Frequency
Bond Type of Compound Range, cm21 Intensity
CiH Alkanes 2850–2970 Strong
1340–1470 Strong
CiH Alkenes 3010–3095 Medium
 675–995 Strong
CiH Alkynes 1 iC{CiH 2 3300 Strong
CiH Aromatic rings 3010–3100 Medium
 690–900 Strong
OiH Monomeric alcohols, phenols 3590–3650 Variable
Hydrogen-bonded alcohols, phenols 3200–3600 Variable, sometimes broad
Monomeric carboxylic acids 3500–3650 Medium
Hydrogen-bonded carboxylic acids 2500–2700 Broad
NiH Amines, amides 3300–3500 Medium
CwC Alkenes 1610–1680 Variable
CwC Aromatic rings 1500–1600 Variable
C{C Alkynes 2100–2260 Variable
CiN Amines, amides 1180–1360 Strong
C{N Nitriles 2210–2280 Strong
CiO Alcohols, ethers, carboxylic acids, esters 1050–1300 Strong
CwO Aldehydes, ketones, carboxylic acids, esters 1690–1760 Strong
NO2 Nitro compounds 1500–1570 Strong
1300–1370 Strong

Several group frequencies are identified in the four spectra The Fingerprint Region
in Figure 17-5. All four spectra contain an absorption band at Small differences in the structure and constitution of a mole-
2900 to 3000 cm21, which corresponds to a CiH stretching cule result in significant changes in the distribution of absorp-
vibration and generally indicates the presence of one or more tion maxima in this region of the spectrum that extends from
alkane groups (see Table 17-3). The two bands at about 1375 about 1200 to 600 cm21 (8 to 14 µm). The fingerprint region is
cm21 and 1450 cm21 are also characteristic group frequencies thus well suited for identifying compounds based on spectral
for CiH groups and result from bending vibrations in the mol- comparisons. Most single bonds give rise to absorption bands
ecule. The spectrum in Figure 17-5c illustrates the group fre- at these frequencies; because their energies are about the same,
quency for an OiH stretching vibration at about 3200 cm21 as strong interaction occurs between neighboring bonds. The
well as the alkane group frequencies (these bands are also pres- absorption bands are thus composites of these various interac-
ent in the spectrum for n-hexanal shown in Figure 16-12). tions and depend on the overall skeletal structure of the mol-
Finally, the characteristic group frequency for a CiCl ecule. Exact interpretation of spectra in this region is seldom
stretching vibration is shown at about 800 cm21 in Figure 17-5d. possible because of the complexity of the spectra. However, for
identification purposes, the richness of spectral features is a
Group frequencies and correlation charts permit intelli-
distinct advantage. Many compounds show unique absorption
gent guesses as to what functional groups are likely to be present
bands in this region, which is very useful for final identification.
or absent in a molecule. Ordinarily, it is impossible to identify
unambiguously either the sources of all of the bands in a given Figure 17-5a and b illustrates the unique character of IR
spectrum or the exact identity of the molecule. Instead, group spectra, particularly in the fingerprint region. The two mole-
frequencies and correlation charts serve as a starting point in the cules shown differ by just one methyl group, yet the two spectra
identification process. differ dramatically in the fingerprint region.
17A Mid-IR Absorption Spectrometry 419

As shown in Figure 17-6 (pages 420–421), a number of inor- lyte and that of each potential match can be shown simultaneously
ganic groups such as sulfate, phosphate, nitrate, and carbonate absorb on the computer display for comparison, as shown in Figure 17-7.
IR radiation in the fingerprint region (,1200 cm21 or .8.3 µm). In 1980 the Sadtler Standard IR Collection and the Sadtler
Commercial IR Collection became available as software
Limitations to the Use of Correlation Charts
­packages. Currently, this library contains more than 220,000
The unambiguous establishment of the identity or the structure
spectra.7 Included are vapor-phase spectra and condensed-phase
of a compound is seldom possible from correlation charts alone.
spectra of pure compounds. Individual databases in application
Uncertainties frequently arise from overlapping group frequen-
areas such as polymers, industrial products, pure organic com-
cies, spectral variations as a function of the physical state of the
pounds, forensic sciences, and the environment are available.
sample (that is, whether it is a solution, a mull, in a pelleted
Several manufacturers of Fourier transform IR (FTIR) instru-
form, and so forth), and instrumental limitations.
ments have now incorporated these packages into their instru-
In using group frequencies, it is essential that the entire ment computers, thus creating instantly available IR libraries of
spectrum, rather than a small isolated portion, be considered hundreds of thousands of compounds.
and interrelated. Interpretation based on one part of the spec-
The Sadtler algorithm consists of a search system in which
trum should be confirmed or rejected by studying other regions.
the spectrum of the unknown compound is first coded according
To summarize, correlation charts serve only as a guide for to the location of its strongest absorption peak; then each addi-
further and more careful study. Several excellent monographs tional strong band (%T , 60%) in ten regions 200 cm21 wide from
describe the absorption characteristics of functional groups in 4000 to 2000 cm21 are coded by their location. Finally, the strong
detail.4 A study of these characteristics, as well as the other phys- bands in seventeen regions 100 cm21 wide from 2100 to 400 cm21
ical properties of the sample, may permit unambiguous identifi- are coded in a similar way. The compounds in the library are
cation. IR spectroscopy, when used in conjunction with mass coded in this same way. The data are organized by the location of
spectrometry, nuclear magnetic resonance, and elemental analy- the strongest band with only those compounds having the same
sis, usually leads to positive identification of a species. strongest band being considered in any sample identification. This
procedure is rapid and produces a list of potential matches within
Collections of Spectra
a short period. For example, searching a base library of 50,000
As just noted, correlation charts seldom suffice for the positive compounds by this procedure requires only a few seconds.
identification of an organic compound from its IR spectrum.
However, several catalogs of IR spectra are available that assist In addition to search systems, artificial intelligence-based
in qualitative identification by providing comparison spectra computer programs seek to determine the structure or substruc-
for a large number of pure compounds.5 Manually searching ture from spectral profiles. Some of these programs use data
large catalogs of spectra is slow and tedious. For this reason, from different types of instruments (nuclear magnetic resonance,
­computer-based search systems are widely used. mass, and IR spectrometers, etc.) to determine ­structures.8

Computer Search Systems


17A-3 Quantitative Applications
Virtually all IR instrument manufacturers now offer computer
Quantitative IR absorption methods differ somewhat from
search systems to aid in identifying compounds from stored IR
UV-visible molecular spectroscopic methods because of the
spectral data.6 The position and relative magnitudes of peaks in
greater complexity of the spectra, the narrowness of the absorp-
the spectrum of the analyte are determined and stored in memory
tion bands, and the instrumental limitations of IR instruments.
to give a peak profile, which can be compared with profiles of pure
Quantitative data obtained with older dispersive IR instruments
compounds stored on disk, CD-ROM, or a website. The computer
were generally significantly inferior in quality to data obtained
then matches profiles and prints a list of compounds having spec-
with UV-visible spectrophotometers. Modern FTIR instru-
tra similar to that of the analyte. Usually, the spectrum of the ana-
ments, however, achieve markedly better precision and accuracy
than their dispersive predecessors. Meticulous attention to detail
4
R. M. Silverstein, F. X. Webster, D. Kiemle, and D. L. Bryce, Spectrometric Identi- is, however, essential for obtaining good-quality results.9
fication of Organic Compounds, 8th ed., New York: Wiley, 2015; B. Schrader, Infra-
red and Raman Spectroscopy, New York: VCH, 1995; N. B. Colthup, L. H. Daly,
and S. E. Wiberley, Introduction to Infrared and Raman Spectroscopy, 3rd ed., San Tutorial: Learn more about IR spectral interpretation
Diego: Academic Press, 1990.
and identification at www.tinyurl.com/skoogpia7
5
See Informatics/Sadtler Group, Bio-Rad Laboratories, Inc., Philadelphia, PA;
C. J. Pouchert, The Aldrich Library of FT-IR Spectra, Milwaukee, WI: Aldrich
Chemical Co., 1997; Thermo Fisher Scientific-Spectral Libraries (http://www 7
.thermoscientific.com/en/products/spectral-libraries.html). See note 5.
8
6
See E. Pretsch, G. Toth, M. E. Monk, and M. Badertscher, Computer Aided See, for example, M. E. Munk, J. Chem. Inf. Comput. Sci., 1998, 38, 997, DOI:
Structure Elucidation: Spectra Interpretation and Structure Generation, New 10.1021/ci980083r.
York: VCH-Wiley, 2003; B. C. Smith, Infrared Spectral Interpretation: A System- 9
For a discussion of quantitative FTIR spectroscopy, see B. C. Smith, Fundamen-
atic Approach, Boca Raton, FL: CRC Press, 1998, Chap. 9; W. O. George and tals of Fourier Transform Infrared Spectroscopy, 2nd ed., Boca Raton, FL: CRC
H. Willis, Computer Methods in UV, Visible and IR Spectroscopy, New York: Springer- Press, 2011; P. R. Griffiths and J. A. deHasech, Fourier Transform Infrared Spec-
Verlag, 1990. trometry, 2nd ed., New York: Wiley, 2007.
420 Chapter 17 Applications of Infrared Spectrometry
FIGURE 17-6 Correlation chart. (From N. Colthup, J. Opt. Soc. Am., 1950, 40, 397.)
17A Mid-IR Absorption Spectrometry 421
422 Chapter 17 Applications of Infrared Spectrometry

Absorbance Unknown

US000022 Benzene

3500 3000 2500 2000 1500 1000 500


Wavenumber, cm –1

FIGURE 17-7 Plot of an unknown spectrum and the best match from a computer search
report. (Courtesy of Informatics/Sadtler Group, Bio-Rad Laboratories, Inc., Philadelphia, PA.)

Deviations from Beer’s Law measurements in the IR region because of the difficulty in
With IR radiation, instrumental deviations from Beer’s law are obtaining cells whose transmission characteristics are iden-
more common than with UV and visible wavelengths because tical. Most IR cells have very short path lengths, which are
IR absorption bands are relatively narrow. Furthermore, with difficult to duplicate exactly. In addition, the cell windows
dispersive instruments, the low intensity of sources and low are readily attacked by contaminants in the atmosphere and
sensitivities of transducers in this region require the use of the solvent. Because of this, their transmission characteristics
relatively wide monochromator slit widths; thus, the band- change continually with use. For these reasons, a reference
widths used are frequently of the same order of magnitude as absorber is often dispensed with entirely in IR work, and the
the widths of absorption peaks. We have pointed out (Section intensity of the radiation passing through the sample is simply
13B-2) that this combination of circumstances usually leads to a compared with that of the unobstructed beam. In either case,
nonlinear relationship between absorbance and concentration. the resulting transmittance is ordinarily less than 100%, even
As discussed in Section 16B-1, FTIR instruments have better in regions of the spectrum where no absorption by the sample
performance characteristics. Hence, Beer’s law deviations are occurs (see Figure 17-5).
not quite as serious as with dispersive instruments. In either For quantitative work, it is necessary to correct for the scat-
case, quantitative analysis usually requires that empirical cali- tering and absorption by the solvent and the cell. Two methods
bration curves be used. are used. In the so-called cell-in-cell-out procedure, spectra of
the pure solvent and the analyte solution are obtained succes-
Absorbance Measurement sively with respect to the unobstructed reference beam. The
Matched absorption cells for solvent and solution are ordinarily same cell is used for both measurements. The transmittance of
used in the UV and visible regions, and the measured absor- each solution versus the reference beam is then determined at an
bance is then found from the relation absorption maximum of the analyte. These transmittances can
Psolvent be written as
A 5 log
Psolution T0 5 P0/Pr
The use of the solvent in a matched cell as a reference absorber and
has the advantage of largely canceling out the effects of radi-
Ts 5 P/Pr
ation losses due to reflection at the various interfaces, scat-
tering and absorption by the solvent, and absorption by the where Pr is the power of the unobstructed beam and T0 and
container windows. This technique is seldom practical for Ts are the transmittances of the solvent and analyte solution,
17A Mid-IR Absorption Spectrometry 423

Detector signal amplitude

Time
(a)

1.0

0.8
Transmittance

0.6

0.4

0.2

0.0
4000 3600 3200 2800 2400 2000 1600 1200 800 400
Wavenumber, cm–1
(b)

FIGURE 17-8 (a) Interferogram obtained from a typical FTIR spectrometer for methylene chloride. The plot shows detector signal
output as a function of time or displacement of the moving mirror of the interferometer. (b) IR spectrum of methylene chloride
produced by the Fourier transformation of the data in (a). Note that the Fourier transform takes signal intensity collected as a
function of time and produces transmittance as a function of frequency after subtraction of a background interferogram and
proper scaling.

respectively, against this reference. If Pr remains constant during An alternative way of obtaining P 0 and T for a single
the two measurements, then the transmittance of the sample absorption band is the baseline method. For an instrument that
with respect to the solvent can be obtained by division of the two displays transmittance, the solvent transmittance is assumed to
equations. That is, be constant or at least to change linearly between the shoulders
of the absorption peak, as illustrated in Figure 17-9. The quan-
T 5 Ts/T0 5 P/P0
tities T0 and Ts are then obtained as shown in the figure. For
With modern FTIR spectrometers, the reference interfer- direct absorbance readout, as is shown in Figure 17-7, the absor-
ogram is obtained with no sample in the sample cell. Then, the bance is assumed to be constant or to change linearly under the
sample is placed in the cell, and a second interferogram obtained. absorption band. The peak absorbance is then obtained by sub-
Figure 17-8a shows an interferogram collected using an FTIR tracting the baseline absorbance.
spectrometer with methylene chloride, CH2Cl2, in the sample cell.
The Fourier transform is then applied to the two interferograms Typical Applications
to compute the IR spectra of the reference and the sample. The With the exception of homonuclear molecules, all organic and
ratio of the two spectra can then be computed to produce an IR inorganic molecular species absorb radiation in the IR region.
spectrum of the analyte such as the one illustrated in Figure 17-8b. IR spectrophotometry thus offers the potential for determining
424 Chapter 17 Applications of Infrared Spectrometry

100 cyclohexane solvent is shown in Figure 17-10. Useful absorp-


tion bands for determination of the individual ­c ompounds
occur at 13.47, 13.01, 12.58, and 14.36 µm, ­respectively. Unfor-
Transmittance, %

tunately, however, the absorbance of a mixture at any one of


T0 P these wavelengths is not entirely determined by the concen-
50 A = log = log 0 tration of just one component because of overlapping absorp-
Ts P T0
tion bands. Thus, molar absorptivities for each of the four
compounds must be determined at the four wavelengths.
Then, four simultaneous equations can be written that permit
Ts the calculation of the concentration of each species from four
0
Wavelength
absorbance measurements (see Section 14D-2). Alternatively,
chemometric techniques, such as factor analytical methods,10
FIGURE 17-9 The baseline method for determining the absorbance can use entire spectral regions to determine the individual
of an absorption maximum. components. Such methods can be used even when the rela-
tionship between absorbance and concentration is nonlinear
(as frequently occurs in the IR region).
an unusually large number of substances. Moreover, the unique-
Determination of Air Contaminants. The recent prolif-
ness of an IR spectrum leads to a degree of specificity that is
eration of government regulations with respect to atmospheric
matched or exceeded by relatively few other analytical methods.
contaminants has demanded the development of sensitive,
This specificity has found particular application to analysis of
rapid, and highly specific methods for a variety of chemical
mixtures of closely related organic compounds. Two examples
compounds. IR absorption procedures appear to meet this need
that typify these applications follow.
better than any other single analytical tool.
Analysis of a Mixture of Aromatic Hydrocarbons. A Table 17-4 demonstrates the potential of IR spectroscopy
typical application of quantitative IR spectroscopy involves the for the analysis of mixtures of gases. The standard sample of
resolution of C8H10 isomers in a mixture that includes o-xylene,
m-xylene, p-xylene, and ethylbenzene. The IR absorption spec-
tra of the individual components in the range of 12 to 15 µm in 10
E. R. Malinowski, Factor Analysis in Chemistry, 3rd ed., New York: Wiley, 2002.

100
o-Xylene m-Xylene p-Xylene
Transmittance, %

50

0
12 13 14 15 12 13 14 15 12 13 14 15
l, μm l, μm l, μm
100
Ethylbenzene
Transmittance, %

Cyclohexane
solvent
50

0
12 13 14 15 12 13 14 15
l, μm l, μm

FIGURE 17-10 Spectra of C8H10 isomers in cyclohexane.


17B Mid-IR Reflection Spectrometry 425

Table 17-4 An Example of IR Determinations of such a large number of absorbing compounds, we expect over-
Air Contaminants lapping absorption bands. However, the method can provide a
moderately high degree of selectivity.
Concn, Found, Relative
Contaminant ppm ppm Error, % Disadvantages and Limitations
to Quantitative IR Methods
Carbon monoxide  50.0 49.1 1.8
There are several disadvantages to quantitative analysis by IR
Methyl ethyl ketone 100.0 98.3 1.7 spectrometry. Among these are the frequent nonadherence
Methanol 100.0 99.0 1.0 to Beer’s law and the complexity of spectra. The richness of
Ethylene oxide  50.0 49.9 0.2 spectral features enhances the probability of overlapping
absorption bands. In addition, with older dispersive instru-
Chloroform 100.0 99.5 0.5
ments, the narrowness of bands and the effects of stray radia-
Courtesy of Thermo Fisher Scientific, Inc. tion make absorbance measurements critically dependent on
the slit width and the wavelength setting. Finally, the narrow-
path-length cells required for many analyses are inconvenient
air containing five species in known concentration was ana- to use and may lead to significant analytical uncertainties. For
lyzed with a computerized version of the instrument shown in these reasons, the analytical errors associated with a quanti-
­Figure 16-13; a 20-m gas cell was used. The data were printed tative IR analysis often cannot be reduced to the level asso-
out within a minute or two after sample introduction. ciated with UV and visible methods, even with considerable
Table 17-5 shows potential applications of IR filter photom- care and effort.
eters (such as that shown in Figure 16-13) for the quantitative
determination of various chemicals in the atmosphere for the
purpose of assuring compliance with Occupational Safety and 17B Mid-IR Reflection
Health Administration (OSHA) regulations.
Spectrometry
Of the more than 400 chemicals for which maximum tol-
erable limits have been set by OSHA, more than half appear to IR reflection spectrometry has found a number of applications,
have absorption characteristics suitable for determination by particularly for dealing with solid samples that are difficult to
means of IR filter photometers or spectrophotometers. With handle, such as polymer films and fibers, foods, rubber-based

Table 17-5 IR Determination of Gases for OSHA Compliance


Permissible Exposure Minimum Detectable
Compound Limit, ppm* Wavelength, µm Concentration, ppm†
Carbon disulfide  2  4.54 0.5
b-Chloroprene 10 11.4 4
Diborane  0.1  3.9 0.05
Ethylenediamine 10 13.0 0.4

Hydrogen cyanide  4.7  3.04 0.4
Methyl mercaptan  0.5  3.38 0.4
Nitrobenzene  1 11.8 0.2
Pyridine  5 14.2 0.2
Sulfur dioxide  2  8.6 0.5
Vinyl chloride  1 10.9 0.3
Courtesy of Thermo Fisher Scientific, Inc.
*Cal/OSHA exposure limits for 8-h weighted average as of 26-April-2013. See http://tinyurl.com/ksbnhbm.

For a 20.25-m cell.

Short-term exposure limit: 15-min time-weighted average that shall not be exceeded at any time during the work day.
426 Chapter 17 Applications of Infrared Spectrometry

materials, agriculture products, and many others. 11 Mid-IR Ellipsoidal mirror


reflection spectra, although not identical to the corresponding
absorption spectra, are similar in general appearance and pro-
vide the same information as do their absorption counterparts.
Reflectance spectra can be used for both qualitative and quantita-
tive analysis. Most instrument manufacturers now offer adapters
that fit into the cell compartments of IR absorption instruments
and make it possible to obtain reflection spectra readily.
Sample
17B-1 Types of Reflection cup
From To
Reflected radiation is of four types: specular reflection, diffuse interferometer detector
reflection, internal reflection, and attenuated total reflectance
(ATR). Specular reflection occurs when the reflecting medium is FIGURE 17-11 A diffuse-reflectance attachment for an FTIR
a smooth polished surface. Here, the angle of reflection is iden- spectrometer.
tical to the incident angle of the radiation. If the surface is made
up of an IR absorber, the relative intensity of reflection is less for and they are randomly oriented, radiation is reflected in all
wavelengths that are absorbed than for wavelengths that are not. ­directions. The intensity of the reflected radiation is roughly
Thus, a plot of reflectance R, which is the fraction of the inci- independent of the viewing angle.
dent radiant energy reflected, versus wavelength or wavenumber
A number of models have been developed to describe
provides a spectrum for a compound similar in general appear-
in quantitative terms the intensity of diffuse reflected radia-
ance to a transmission spectrum for the species. Specular reflec-
tion. The most widely used of these models was developed by
tion spectra find some use for examining and characterizing the
Kubelka and Munk.13 Fuller and Griffiths in their discussion
smooth surfaces of solids and coated solids but are not as widely
of this model show that the relative reflectance intensity for a
used as diffuse- and total-reflection spectra. We will, therefore,
­powder f 1 Rr` 2 is given by14
focus on the latter two types of spectra.
1 1 2 Rr` 2 2 k
f 1 Rr` 2 5 5
17B-2 Diffuse-Reflectance Spectrometry 2Rr` s
Diffuse-reflectance IR Fourier transform spectrometry where Rr` is the ratio of the reflected intensity of the sample to
(DRIFTS) is an effective way of directly obtaining IR spectra on that of a nonabsorbing standard, such as finely ground potas-
powdered samples with a minimum of sample preparation.12 In sium chloride. The quantity k is the molar absorption coefficient
addition to saving time in sample preparation, it permits con- of the analyte, and s is a scattering coefficient. For a diluted sam-
ventional IR spectral data to be gathered on samples not appre- ple, k is related to the molar absorptivity ε and the molar con-
ciably altered from their original state. The widespread use of centration of the analyte c by the relationship
diffuse-reflectance measurements had to await the general
k 5 2.303ec
availability of FTIR instruments in the mid-1970s because the
intensity of radiation reflected from powders is too low to be Reflectance spectra then consist of a plot of f 1 Rr` 2 versus wave-
measured at medium resolution and adequate signal-to-noise number (see Figure 17-12b).
ratios with dispersive instruments.
Instrumentation
Diffuse reflection is a complex process that occurs when
Currently, most manufacturers of FTIR instruments offer adapt-
a beam of radiation strikes the surface of a finely divided pow-
ers that fit in cell compartments and permit diffuse-reflectance
der. With this type of sample, specular reflection occurs at each
measurements. Figure 17-11 illustrates one type of adapter.
plane surface. However, because there are many of these s­ urfaces
The collimated beam from the interferometer is directed to an
ellipsoidal mirror and then to the sample. The sample is usually
11
See F. M. Mirabella, ed., Modern Techniques in Applied Molecular Spectroscopy, ground and mixed with KBr or KCl as a diluent. The mixture is
New York: Wiley, 1998; G. Kortum, Reflectance Spectroscopy, New York: Springer, then placed in a sample cup 3–4 mm deep and about 10–15 mm
1969; N. J. Harrick, Internal Reflection Spectroscopy, New York: Wiley, 1967; for in diameter. A complex combination of reflection, absorption,
the theory of reflectance spectroscopy, see B. Hapke, Theory of Reflectance and
and scattering occurs before the beam is directed to the detector.
Emittance Spectroscopy, 2nd ed., Cambridge: Cambridge University Press, 2012.
12
For additional information, see A. Drochner and G. H. Vogel, in Methods in
Physical Chemistry, 2nd ed., R. Schäfer and P. C. Schmidt, eds., Ch. 14, Somerset, 13
P. Kubelka and E. Munk, Tech. Phys., 1931, 12, 593; P. Kubelka, J. Opt. Soc. Am.,
NJ: Wiley, 2012; M. Milosevic, S. L. Berets, Appl. Spectrosc. Rev., 2002, 37, 347,
1948, 38, 448, DOI: 10.1364/josa.38.000448.
DOI: 10.1081/asr-120016081; P. R. Griffiths and M. P. Fuller in Advances in Infra-
14
red and Raman Spectroscopy, R. J. H. Clark and R. E. Hester, eds., Vol. 9, Chap. 2, M. P. Fuller and P. R. Griffiths, Anal. Chem., 1978, 50, 1906, DOI: 10.1021/
London: Heydon and Sons, 1982. ac50035a045.
17B Mid-IR Reflection Spectrometry 427

beam reflected increases as the angle of incidence becomes


larger; beyond a certain critical angle, reflection is com-
N
1.00 H plete. It has been shown both theoretically and experimen-
tally that during the reflection process the beam penetrates
Absorbance

carbazole
a small distance into the less dense medium before reflection
0.75
occurs. The depth of penetration, which varies from a frac-
tion of a wavelength up to several wavelengths, depends on
0.50 the wavelength, the index of refraction of the two materials,
and the angle of the beam with respect to the interface. The
2000 1500 1000 penetrating radiation is called the evanescent wave. At wave-
(a) lengths where the less dense medium absorbs the evanescent
radiation, attenuation of the beam occurs, which is known as
attenuated total reflectance, or ATR. The resulting ATR spec-
trum resembles that of a conventional IR spectrum with some
differences.
0.30
Instrumentation
f (R′∞)

0.20 Figure 17-13 shows an apparatus for ATR measurements. As can


be seen from the upper figure, the sample (here, a solid) is placed
0.10 on opposite sides of a transparent crystalline material of high
refractive index. By proper adjustment of the incident angle, the
2000 1500 1000
radiation undergoes multiple internal reflections before passing
Wavenumber, cm–1 from the crystal to the detector. Absorption and attenuation take
(b) place at each of these reflections.
Figure 17-13b is an optical diagram of an adapter that fits
FIGURE 17-12 Comparison of the absorption spectrum
(a) for ­carbazole with its diffuse-reflectance spectrum (b). into the cell area of most IR spectrometers and permits ATR
measurements. Cells for liquid samples are also available.
To obtain a spectrum with a single-beam instrument, the ATR Spectra
signal for the sample is first stored. A reference signal with ATR spectra are similar but not identical to ordinary absorption
a good reflector, such as finely ground KBr or KCl, is then spectra. In general, although the same bands are observed, their
recorded in place of the sample. The ratio of these signals is then relative intensities differ. With ATR spectra, the absorbances,
taken to give the reflectance. although dependent on the angle of incidence, are independent
Comparison of Absorption and Reflection Spectra of sample thickness, because the radiation penetrates only a few
micrometers into the sample. The effective penetration depth dp
Figure 17-12 compares the conventional IR absorption spec-
depends on the wavelength of the beam, the refractive indexes of
trum for carbazole obtained by means of a KBr pellet with the
the crystal and the sample, and the beam angle. The penetration
diffuse-reflectance spectrum of a 5% finely ground mixture of
depth can be calculated from
carbazole in potassium chloride. Note that the peak locations are
the same in the two spectra but that the relative peak heights dif- lc
fer considerably. The differences are typical, with minor peaks dp 5 1/2
2p3sin u 2 1 ns /nc 2 24
2

generally appearing larger in reflection spectra.


where lc is the wavelength in the crystal (l/nc), θ is the angle of
incidence, and ns and nc are the refractive indexes of the sam-
17B-3 ATR Spectrometry ple and crystal, respectively. Note that the effective p
­ enetration
Internal-reflection spectroscopy is a technique for obtaining IR depth can be changed by changing the crystal material, the
spectra of samples that are difficult to deal with, such as solids of angle of incidence, or both. It is possible to obtain a depth
limited solubility, films, threads, pastes, adhesives, and ­powders.15 profile of a surface using ATR spectroscopy. In practice, a multi-­
reflection crystal with a 45° angle can accommodate most rou-
Principles
tine ­samples.
When a beam of radiation passes from a more dense to a less
dense medium, reflection occurs. The fraction of the incident One of the major advantages of ATR spectroscopy is that
absorption spectra are readily obtainable on a wide variety of
15 sample types with a minimum of preparation. Threads, yarns,
See F. M. Mirabella, ed., Modern Techniques in Applied Molecular Spectroscopy,
New York: Wiley, 1998; G. Kortum, Reflectance Spectroscopy, New York: Springer, fabrics, and fibers can be studied by pressing the samples against
1969; N. J. Harrick, Internal Reflection Spectroscopy, New York: Wiley, 1967. the dense crystal. Pastes, powders, or suspensions can be ­handled
428 Chapter 17 Applications of Infrared Spectrometry

From interferometer To detector


or monochromator Sample
Solid with high Modulated Mirror
Sample refractive index beam
from FTIR
(a)
Window
To
From interferometer detector
or monochromator
Microphone

Sample

(b) FIGURE 17-14 Diagram of a photoacoustic attachment for an FTIR


spectrometer.
FIGURE 17-13 ATR apparatus. In (a) a solid sample is shown
mounted on an internal reflection crystal of high refractive index.
The materials used as ATR crystals include KRS-5, AgCl, Ge, Si, gas, which in turn results in regular pressure fluctuations within
and the Irtran materials. Solid samples can be pressed against the
the chamber. If the chopping rate lies in the acoustical frequency
­crystal to obtain optical contact. In (b), a typical attachment for
ATR is shown. With many attachments, the internal reflection plate
range, these pulses of pressure can be detected by a sensitive
can be positioned in the holder to provide several incident angles. microphone. The photoacoustic effect has been used since the
turn of the century for the analysis of absorbing gases and has
recently taken on new importance for this purpose with the
in a similar way. Aqueous solutions can also be accommodated advent of tunable IR lasers as sources. Of greater importance,
provided the crystal is not water soluble. There are even ATR however, has been the use of the photoacoustic effect for obtain-
flow cells available. ATR spectroscopy has been applied to many ing absorption spectra of solids and turbid liquids.
substances, such as polymers, rubbers, and other solids. It is of
interest that the resulting spectra are free from the interference
fringes mentioned previously. 17C-2 Photoacoustic IR Spectra
Photoacoustic IR spectroscopy saw limited application before
The spectra obtained with ATR methods can differ from IR
the advent of FTIR instruments. Now, several manufacturers
absorption spectra because of distortions that occur near strong
make photoacoustic accessories for FTIR instruments. In pho-
absorption bands where the sample refractive index may change
toacoustic measurements, the sample is placed in a small sam-
rapidly. Also, the orientation of the sample on the ATR crystal
ple cup within the photoacoustic attachment as illustrated in
can influence band shapes and relative intensities. However, the
Figure 17‑14. The photoacoustic spectrometer chamber is filled
ATR band intensity is usually proportional to concentration so
with a high-thermal-conductivity gas such as helium or nitrogen
that quantitative measurements can be made.
and placed in the FTIR sample compartment. The mirror shown
deflects the modulated beam onto the sample. Absorption of
the IR beam by the sample can result in nonradiative decay of
17C Photoacoustic IR Spectroscopy the excited vibrational states of the sample molecules. This can
Photoacoustic spectroscopy (PAS) provides a way to obtain UV, transfer heat to the surface of the sample and result in the gen-
visible, and IR absorption spectra of solids, semisolids, or turbid eration of a modulated acoustic wave in the gas inside the cham-
liquids.16 Acquisition of spectra for these materials by ordinary ber. A very sensitive microphone then detects the acoustic wave.
methods is usually difficult at best and often impossible because Photoacoustic spectra are normally plotted in a format sim-
of light scattering and reflection. ilar to absorption spectra as shown in Figure 17-15. The spectra
are usually plotted as a ratio to a background scan of a totally
absorbing material such as carbon black. Although the frequen-
17C-1 The Photoacoustic Effect cies of transitions are the same as in absorption spectra, relative
PAS is based on a light absorption effect that was first investi- intensities depend on wavelength and modulation frequency.
gated in the 1880s by Alexander Graham Bell and others. This
Photoacoustic spectra can be obtained on samples with
effect is observed when a gas in a closed cell is irradiated with a
essentially no sample preparation. The only requirement is that
chopped beam of radiation of a wavelength that is absorbed by
the sample must fit within the sample cup. In addition to solids
the gas. The absorbed radiation causes periodic heating of the
and turbid liquids, photoacoustic methods have been used for
detecting the components of mixtures separated by thin-layer
16
See K. H. Michaelian, Photoacoustic Infrared Spectroscopy, 2nd ed., Weinheim, and high-performance liquid chromatography and for monitor-
Germany: Wiley-VCH, 2010. ing the concentrations of gaseous pollutants in the atmosphere.
17D Near-IR Spectroscopy 429

in Section 7C-3. Speed and ruggedness are the main advantages of


AOTF devices. Fourier transform spectrometers are also available
Intensity

commercially for NIR spectrometry. The advantages normally


associated with FT instruments in the mid-IR region, such as high
(a) Pulverized medium-
volatility coal sample throughput and resolution, are less applicable in the NIR region.
However, the wavelength reproducibility and signal-to-noise-­
(b) Extruded polymer pellet ratio characteristics are still major advantages of FT systems.
Intensity

Most NIR spectrometers use tungsten-halogen lamps with


quartz windows. Cells for absorption measurements are usu-
ally quartz or fused silica transparent up to about 3000 nm. Cell
4000 3500 3000 2500 2000 1500 1000 500 lengths vary from 0.1 to 10 cm. Detectors range from PbS and
Wavenumber, cm21 PbSe photoconductors to InSb and InAs photodiodes. Array
FIGURE 17-15 Photoacoustic IR spectra of (a) a pulverized coal detectors, such as InGaAs detectors, have also become available
sample and (b) an extruded polymer pellet. (From Practical Sampling for the region. Several commercial UV-visible spectrophotome-
Techniques for Infrared Analysis, P. B. Coleman, ed., Boca Raton, FL: ters are designed to operate from 180 to 2500 nm and can thus
CRC Press, 1993.) be used to obtain NIR spectra.
Several solvents are used for NIR studies. Some of these
are listed in Figure 17-16. Note that only carbon tetrachloride
17D Near-IR Spectroscopy and carbon disulfide are transparent throughout the entire NIR
region.
The near-IR (NIR) region of the spectrum extends from the
upper wavelength end of the visible region at about 770 nm
to 2500 nm (13,000 to 4000 cm21).17 Absorption bands in this 17D-2 Data Processing in NIR spectrometry
region are overtones or combinations (Section 16A-4) of fun- NIR spectral bands are normally broad and often overlapping.
damental stretching vibrations that occur in the region of 3000 There are rarely clean spectral bands that allow simple correla-
to 1700 cm21. The bonds involved are usually CiH, NiH, tion with analyte concentration. Instead, multivariate calibration
and OiH. Because the bands are overtones or combinations, techniques are used.19 Most commonly, partial least squares,
their molar absorptivities are low and detection limits are on the principal components regression, and artificial neural networks
order of 0.1%.18 are used. Such calibration involves development of a calibration
model through obtaining results on a “training set” that includes
In contrast to mid-IR spectrometry, the most important
as many of the conditions encountered in the samples as possi-
uses of NIR radiation are for the routine quantitative determi-
ble. Problems and pitfalls in developing the models have been
nation of species, such as water, proteins, low-molecular-mass
discussed.20
hydrocarbons, and fats, in products of the agricultural, food,
petroleum, and chemical industries. Both diffuse-reflection and The manufacturers of NIR instruments include software
transmission measurements are used, although diffuse reflec- packages for developing calibration models. In addition, third-
tance is by far the more widely used. party software for multivariate calibration is readily available.

17D-1 Instrumentation 17D-3 Applications of NIR Absorption


Four different types of instruments are available for the NIR Spectrometry
region. Grating instruments are similar to those used for UV-­ In contrast to mid-IR spectroscopy, NIR absorption spec-
visible absorption spectroscopy. There are also discrete filter tra are less useful for identification and more useful for quan-
instruments usually containing filter wheels for selecting differ- titative analysis of compounds containing functional groups
ent wavelengths. These are less flexible than other instruments made up of hydrogen bonded to carbon, nitrogen, and oxygen.
but useful for fixed, well-characterized samples. In addition,
­acousto-optic tunable filter (AOTF) instruments are available.
Accousto-optic filters and filter-based instruments are discussed 19
For a discussion of multivariate techniques, see K. R. Beebe, R. J. Pell, and
M. B. Seasholtz, Chemometrics: A Practical Guide, Chap. 5, New York: Wiley,
1998; H. Martens and T. Naes, Multivariate Calibration, New York: Wiley, 1989;
R. G. Brereton, Applied Chemometrics for Scientists, Chap. 6, Chichester, UK:
17
In the literature of NIR spectrometry, the abscissa is usually wavelength in Wiley, 2007; K. Varmuza and P. Fitzmoser, Introduction to Multivariate Statistics
­ anometers or micrometers in contrast to mid-IR spectra in which the abscissa is
n in ­Chemometrics, Chap. 4, Boca Raton, FL: CRC Press, 2009.
wavenumber in units of (cm)21. 20
For example, see M. A. Arnold, J. J. Burmeister, and G. W. Small, Anal. Chem.,
18
For a general reference on NIR, see Handbook of Near-Infrared Analysis, 3rd ed., 1998, 70, 1773, DOI: 10.1021/ac9710801; L. Zhang, G. W. Small, and M. A.
D. A. Burns and E. W. Ciurczak, eds., Boca Raton, FL: Taylor & Francis, 2008. Arnold, Anal. Chem., 2003, 75, 5905, DOI: 10.1021/ac034495x.
430 Chapter 17 Applications of Infrared Spectrometry

Wavelength, μm
1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0

Carbon tetrachloride

Carbon disulfide

Methylene chloride

Dioxane

Heptane or dioxane

Benzene

Acetonitrile

Dimethyl sulfoxide

FIGURE 17-16 Some useful solvents for NIR spectroscopy. Solid lines indicate
­satisfactory transparency for use with 1-cm cells.

Such ­compounds can often be determined with accuracies and in finely ground solids.21 The most widespread use of this tech-
­precisions equivalent to UV-visible spectroscopy, which are nique has been for the determination of protein, moisture,
usually much better than those of mid-IR spectroscopy. Some starch, oil, lipids, and cellulose in agricultural products such
applications include the determination of water in a variety of as grains and oilseeds.22 For example, Canada sells all its wheat
samples, including glycerol, hydrazine, organic films, and fuming based on guaranteed protein content, so to certify the grain, the
nitric acid; the quantitative determination of phenols, alcohols, Canadian Grain Commission formerly ran more than 600,000
organic acids, and hydroperoxides based on the first overtone of Kjeldahl protein determinations annually. It was estimated in
the OiH stretching vibration that absorbs radiation at about 1984 that 80% to 90% of all Canadian grain was analyzed for
7100 cm21 (1.4 µm); and the determination of esters, ketones, protein by NIR reflectance spectroscopy at a savings of more
and carboxylic acids based on their absorption in the region of than $500,000 a year in analytical costs.23
3300 to 3600 cm21 (2.8 to 3.0 µm). The absorption in this last In NIR reflectance spectroscopy, the finely ground solid
case is the first overtone of the carbonyl stretching vibration. sample is irradiated with one or more narrow bands of r­ adiation
NIR spectrophotometry is also a valuable tool for identifi- ranging in wavelength from 1 to 2.5 µm or 10,000 to 4000 cm21.
cation and determination of primary and secondary amines in Diffuse reflectance occurs in which the radiation penetrates
the presence of tertiary amines in mixtures. The analyses are the surface layer of the particles, excites vibrational modes of
generally carried out in carbon tetrachloride solutions in 10-cm the analyte molecule, and is then scattered in all directions.
cells. Primary amines are determined directly by measurement A reflectance spectrum is thus produced that is dependent on
of the absorbance of a combination NiH stretching band the composition of the sample. A typical reflectance spectrum
at about 5000 cm21 (2.0 µm); neither secondary nor tertiary for a sample of wheat is shown in Figure 17-17. The ordinate in
amines absorb radiation in this region. Primary and secondary this case is the logarithm of the reciprocal of reflectance R, where
amines have several overlapping absorption bands in the 3300 R is the ratio of the intensity of radiation reflected from the sam-
to 10,000 cm21 (1 to 3 µm) region due to various NiH stretch- ple to the reflectance from a standard reflector, such as finely
ing vibrations and their overtones. On the other hand, tertiary ground barium sulfate or magnesium oxide. The reflectance
amines can have no such bands. Thus, one of these bands gives
the secondary amine concentration after correction for the
absorption by the primary amine.
21
See, J. M. Soriano-Disla, L. J. Janik, R. A. Viscarra Rossel, L. M. Macdonald,
M. J. McLaughlin, Appl. Spectrosc. Rev., 2014, 49, 139, DOI: 10.1080/
17D-4 Applications of NIR 05704928.2013.811081.
Reflectance Spectrometry 22
H. Cen and Y. He., Trends Food Sci. Technol., 2007, 18, 72, DOI: 10.1016/
NIR reflectance spectroscopy has become a most important j.tifs.2006.09.003.
tool for the routine quantitative determination of ­constituents 23
S. A. Borman, Anal. Chem., 1984, 56, 933A, DOI: 10.1021/ac00272a727.
17F IR Emission Spectroscopy 431

s­ ubstituted benzene derivatives, which generally show several


absorption bands. The spectra are frequently quite specific and
useful for identifying a particular compound. There are also
characteristic group frequencies in the far-IR region.
1.0 Pure rotational absorption by gases is observed in the
far-IR region, provided the molecules have permanent dipole
log 1/R

0.7
moments. Examples include H2O, O3, HCl, and AsH3. Absorp-
tion by water is troublesome; elimination of its interference
requires evacuation or at least purging of the spectrometer.
0.4
Prior to the advent of Fourier transform spectrometers,
experimental difficulties made it difficult to obtain good far-IR
0.1 spectra. The sources available in this region are quite low in
1400 1800 2200 2600 intensities. Order-sorting filters are needed with grating instru-
Wavelength, nm ments in this region to minimize radiation diffracted from
FIGURE 17-17 Diffuse-reflectance NIR spectrum for a sample higher grating orders. These reduced the already low through-
of wheat. puts of dispersive spectrometers. It is no surprise that Fourier
transform spectrometry was first applied in this spectral region.
The high throughput of interferometers and the relatively low
band at 1940 nm is a water band used for moisture determina- mechanical tolerances required combine to make it fairly simple
tions. The band at about 2100 nm is, in fact, two overlapping to obtain FTIR spectra in the far-IR region with good signal-to-
bands, one for starch and the other for protein. By making mea- noise ratios.
surements at several wavelengths in this region, the concentra-
tions of each of these components can be determined.
Instruments for diffuse-reflectance measurements are avail- 17F IR Emission Spectroscopy
able from commercial sources. Some of these use several inter-
molecules that absorb IR radiation are also capable of emitting
ference filters to provide narrow bands of radiation. Others are
characteristic IR wavelengths when they are heated. The prin-
equipped with grating monochromators. Ordinarily, reflectance
cipal deterrent to the analytical application of this phenome-
measurements are made at two or more wavelengths for each
non has been the poor signal-to-noise characteristics of the IR
analyte species being determined.
emission signal, particularly when the sample is at a temperature
The great advantage of NIR reflectance methods is their only slightly higher than its surroundings. With interferometry,
speed and the simplicity of sample preparation. Once method however, interesting and useful applications have appeared in
development has been completed, analysis of solid samples for the literature.
several species can be completed in a few minutes. Accuracies
An early example of the application of IR emission spec-
and precisions of 1% to 2% relative are regularly reported.
troscopy is found in a paper that describes the use of a Fourier
transform spectrometer for the identification of microgram
quantities of pesticides.24 Samples were prepared by dissolving
17E Far-IR Spectroscopy them in a suitable solvent followed by evaporation on a NaCl or
The far-IR region is particularly useful for inorganic studies KBr plate. The plate was then heated electrically near the spec-
because absorption due to stretching and bending vibrations trometer entrance. Pesticides such as DDT, malathion, and diel-
of bonds between metal atoms and both inorganic and organic drin were identified in amounts as low as 1 to 10 µg.
ligands generally occurs at frequencies lower than 650 cm21 Equally interesting has been the use of interferometry for
(.15 µm). For example, heavy-metal iodides generally absorb the remote detection of components emitted from industrial
radiation in the region below 100 cm21, and the bromides and smoke stacks. In one of these applications, an interferometer was
chlorides have bands at higher frequencies. Absorption frequen- mounted on an 8-inch reflecting telescope.25 With the telescope
cies for metal-organic bonds ordinarily depend on both the focused on the plume from an industrial plant, CO2 and SO2
metal atom and the organic portion of the species. were readily detected at a distance of several hundred feet.
Far-IR studies of inorganic solids have also provided use-
ful information about lattice energies of crystals and transition
energies of semiconducting materials. 24
J. Coleman and M. J. D. Low, Spectrochim. Acta, 1966, 22, 1293, DOI:
Molecules composed only of light atoms absorb far‑IR 10.1016/0371-1951(66)80034-6.
radiation if they have skeletal bending modes that involve more 25
M. J. D. Low and F. K. Clancy, Env. Sci. Technol., 1967, 1, 73, DOI: 10.1021/
than two atoms other than hydrogen. Important examples are es60001a007.
432 Chapter 17 Applications of Infrared Spectrometry

Wavenumber, cm21 that reduce the size of the IR beam to about that of the sam-
1500 1000 500 ple. The optical microscope is used to visually locate the sample
particle or spot to be studied with the IR beam. The IR source
1
is an ordinary Fourier transform spectrometer and not a grat-
ing instrument because of the much greater sensitivity of FTIR
instruments. The detector is usually a liquid-nitrogen-cooled
0.95 mercury telluride–cadmium telluride photoconductive device
Emissivity

that is more sensitive than other types of IR detectors.


Bound water IR microscopy has been applied to identification of polymer
0.9
contaminants, imperfections in polymer films, and examination
of layers of laminated polymer sheets. The IR microscope has
Mini-TES
Orbiting MGS/ TES proven quite useful in forensic science.27 Forensic applications
0.85 include identification of micro-samples of fibers, paint, and
6 10 24 explosives. IR microscopy has also been applied in the textile
Wavelength, μm industry to characterize single fibers.
FIGURE 17-18 Spectra from Mars rover Spirit showing an In a general sense, IR microspectroscopy can be divided
­unidentified mineral containing bound water in its crystal into sample mapping or sample imaging applications.28 In sam-
­structure. Minerals such as zeolites and gypsum are possible candi- ple mapping, the spectrum of a region of sample is taken, and
dates. Mini-TES 5 mini thermal-emission spectrometer, MGS 5 then another region is moved into the measurement zone and
Mars Global Surveyor. (Courtesy of NASA/JPL/Arizona State a new spectrum obtained. This sequential acquisition of spec-
University.)
tra is repeated until the entire area of interest of the sample has
been mapped. In contrast, imaging is accomplished by focusing
The Mars Global Surveyor launched from Cape Canaveral in an image of the sample on an array detector and taking a digi-
November 1996. It reached Mars orbit in 1997. One of five instru- tal picture. The intensity of the radiation passing through each
ments on board was an IR emission spectrometer, called a ther- sample region is measured simultaneously by the array detec-
mal emission spectrometer. The spacecraft completed its mapping tor. The mapping/imaging distinction can be somewhat blurry
mission in 2001, providing measurements of the Martian surface since some commercial instruments use an array of detectors to
and atmosphere. The Mars rover Spirit had a mini thermal-emis- acquire a line map prior to repositioning the sample and repeat-
sion spectrometer on board capable of indicating the composi- ing the process several times.
tion of nearby soils and rocks. Figure 17-18 shows an IR emission In hyperspectral imaging, images at multiple wavelength
spectrum taken in early 2004 that showed evidence of an uniden- regions are acquired simultaneously with a two-dimensional
tified water-containing mineral. Thermal emission spectra have array detector. Both the mid-IR and the near-IR regions have
also indicated the presence of carbonates and other hydrated min- been used, although the instrumentation is different in the two
erals. These may have been produced by long-standing bodies of spectral regions. Raman spectroscopy (see Chapter 18) is also
water. Spirit stopped communicating in March 2010; efforts to widely used in hyperspectral imaging applications. These imag-
resume transmissions were concluded in May 2011. ing techniques are often referred to as chemical imaging methods.
The Rover Environmental Monitoring Station (REMS) Biomedical applications of IR imaging and mapping include
aboard the Mars Rover Curiosity contains a suite of IR sensors examination of tissue samples and imaging of individual cells.29
located on one of its booms. These measure the IR radiation IR microspectroscopy has been applied to the examination of
emitted by the ground and atmosphere, which are used to cal- agricultural, plant, and food materials as well as to polymers
culate temperatures. and pharmaceuticals.30 Although many of these techniques are
fairly new, we will undoubtedly see many more applications in
the near future.
17G IR Microscopy and Imaging
IR microspectrometry is used for obtaining absorption or reflec- 27
See Infrared and Raman Spectroscopy in Forensic Science, J. M. Chalmers,
tion spectra of species in samples having physical dimensions in H. G. M. Edwards, and M. D. Hairgreaves, eds., Chichester, UK: Wiley, 2012.
the range of 10 to 500 µm.26 IR microscopes, which were first 28
P. R. Griffiths and E. V. Miseo in Infrared and Raman Imaging, 2nd ed., Chap. 1,
introduced by several instrument manufacturers in the 1980s, R. Salzer and H. W. Siesler, eds., Weinheim, Germany: Wiley-VCH, 2014.
generally consist of two microscopes, one an ordinary optical 29
C. Krafft and J. Popp, Chap. 3, S. Gamsjaeger, R. Mendelsohn, K. Klaushofer, and
microscope and the other an IR device with reflection optics E. P. Paschalis, Chap. 4, M. Diem et al., Chap. 5, in Infrared and Raman Imaging,
2nd ed., R. Salzer and H. W. Siesler, eds., Weinheim, Germany: Wiley-VCH, 2014.
30
H. Schulz, A. Krahmer, A. Naumann, and G. Gudi, Chap. 6, V. Bellon-Maurel
26
See Practical Guide to Infrared Microspectroscopy, H. J. Humecki, ed., New York: and N. Gorretta, Chap. 7, E. Grotheer et al. in Infrared and Raman Spectroscopic
Marcel Dekker, 1995; J. Katon, Infrared Microspectroscopy in Modern Techniques Imaging, 2nd ed., R. Salzer and H. W. Siesler, eds., Weinheim, Germany: Wiley-
in Molecular Spectroscopy, F. Mirabella, ed., New York: Wiley, 1998. VCH, 2014.
Questions and Problems 433

Questions and Problems


*Answers are provided at the end of the book for problems marked with an asterisk.
Problems with this icon are best solved using spreadsheets.

17-1 Cyclohexanone exhibits its strongest IR absorption band at 5.86 µm, and at this wavelength there is a linear
relationship between absorbance and concentration.
Identify the part of the molecule responsible for the absorbance at this wavelength.
(a)
Suggest a solvent that would be suitable for a quantitative analysis of cyclohexanone at this wavelength.
(b)
A solution of cyclohexanone (4.0 mg/mL) in the solvent selected in part (b) exhibits a blank-corrected
(c)
absorbance of 0.800 in a cell with a path length of 0.025 mm. What is the detection limit for this com-
pound under these conditions if the noise associated with the spectrum of the solvent is 0.001 absor-
bance units?
* 17-2 The spectrum in Figure 17-19 was obtained for a liquid with the empirical formula of C3H6O. Identify the
compound.
17-3 The spectrum in Figure 17-20 is that of a high-boiling-point liquid having the empirical formula C9H10O.
Identify the compound as closely as possible.
* 17-4 The spectrum in Figure 17-21 is for an acrid-smelling liquid that boils at 50°C and has a molecular mass of
about 56. What is the compound? What impurity is clearly present?
* 17-5 The spectrum in Figure 17-22 is that of a nitrogen-containing substance that boils at 97°C. What is the
compound?
17-6 Why are quantitative analytical methods based on NIR radiation often more precise and accurate than
methods based on mid-IR radiation?
17-7 Discuss the origins of photoacoustic IR spectra. What are the usual types of samples for photoacoustic
measurements?
17-8 Why do IR spectra seldom show regions at which the transmittance is 100%?
17-9 A quantitative analysis of methyl iso-butyl ketone (MIBK) was performed by measuring absorbances at
1745 cm21. Solutions of differing MIBK concentration gave the absorbances shown in the table.
Concentration of MIBK, % A
1.00 0.12
3.00 0.28
5.00 0.42
7.00 0.56
9.00 0.68
Samples 1, 2, and 3 containing an unknown amount of MIBK gave the absorbances that follow:
Sample A
1 0.49
2 0.74
3 0.34
Determine the equation of the calibration curve and the concentrations of the three samples.
* 17-10 An empty cell showed 20 interference fringes in the wavelength range of 6.0 to 12.2 µm. Calculate the path
length of the cell.
* 17-11 An empty cell exhibited 17.5 interference fringes in the region of 1050 to 1580 cm21. What was the path
length of the cell?
* 17-12 Estimate the thickness of the polystyrene film that yielded the spectrum shown in Figure 16-1.
434 Chapter 17 Applications of Infrared Spectrometry

Questions and Problems (continued)

Challenge Problem
17-13 (a) Consider an ATR experiment with a sample having a refractive index of 1.03 at 2000 cm21. The ATR
crystal was AgCl with a refractive index of 2.00 at this wavelength. For an angle of incidence of 45°,
what is the effective penetration depth of the evanescent wave? How would the penetration depth
change if the angle were changed to 60°?
For the same experiment in part (a) and a 60° incidence angle, find the penetration depths for sample
(b)
refractive indexes varying from 1.00 to 1.70 in steps of 0.10. Plot penetration depth as a function of
refractive index. Determine the refractive index for which the penetration depth becomes zero. What
happens at this point?
For a sample with a refractive index of 1.37 at 2000 cm21 and an incidence angle of 45°, plot the pene-
(c)
tration depth versus the ATR crystal refractive index. Vary the crystal refractive index over the range
of 2.00 to 4.00 in steps of 0.25. Which crystal, AgCl (nc 5 2.00) or Ge (nc 5 4.00), gives the smaller
penetration depth? Why?
An aqueous solution with a refractive index of 1.003 is measured with an ATR crystal with a refractive
(d)
index of 2.8. The incidence angle is 45°. What is the effective penetration depth at 3000 cm21,
2000 cm21, and 1000 cm21? Is absorption by the aqueous solvent as much of a problem in ATR as in
normal IR absorption measurements? Why or why not?
Work by S. Ekgasit and H. Ishida (S. Ekgasit and H. Ishida, Appl. Spectrosc., 1996, 50, 1187; Appl.
(e)
Spectrosc., 1997, 51, 461) describes a new method to obtain a depth profile of a sample surface using
ATR spectroscopy. Describe the principles of this new approach. Why are spectra taken with different
degrees of polarization? What is the complex index of refraction?
Wavelength, μm
2.5 3 4 5 6 7 8 9 10 11 12 13 14 15
100

80 0.1

60
0.3

40
0.5

Transmittance, %
20 0.7
0.9
1.5
0 ∞
4000 3500 3000 2500 2000 1900 1800 1700 1600 1500 1400 1300 1200 1100 1000 900 800 700 650
Wavenumber, cm –1

FIGURE 17-19 See Problem 17-2. (Spectrum courtesy of Thermodynamics Research Center Data Project, Texas A&M University,
College Station, Texas.)

Wavelength, μm
2.5 3 4 5 6 7 8 9 10 11 12 13 14 15
100

80 0.1

60
0.3

40
0.5

Transmittance, %
20 0.7
0.9
1.5
0 ∞
4000 3500 3000 2500 2000 1900 1800 1700 1600 1500 1400 1300 1200 1100 1000 900 800 700 650
Wavenumber, cm –1

FIGURE 17-20 See Problem 17-3. (Spectrum courtesy of Thermodynamics Research Center Data Project, Texas A&M University,
College Station, Texas.)
Questions and Problems 435
Wavenumber, cm –1
20000 10000 5000 4000 3000 2500 2000 1800 1600 1400 1200 1100 1000 950 900 850 800 750 700 650

0.0

0.2

0.4

Transmittance, %
0.6
0.8
436 Chapter 17 Applications of Infrared Spectrometry

1.0
1.5

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Wavelength, μm

FIGURE 17-21 See Problem 17-4. (Spectrum courtesy of Thermodynamics Research Center Data Project, Texas A&M University,
College Station, Texas.)

Wavelength, μm
Questions and Problems (continued)

2.5 3 4 5 6 7 8 9 10 11 12 13 14 15
100

80 0.1

60
0.3

40
0.5

Transmittance, %
20 0.7
0.9
1.5
0 ∞
4000 3500 3000 2500 2000 1900 1800 1700 1600 1500 1400 1300 1200 1100 1000 900 800 700 650
Wavenumber, cm –1

FIGURE 17-22 See Problem 17-5. (Spectrum courtesy of Thermodynamics Research Center Data Project, Texas A&M University,
College Station, Texas.)

You might also like