Infrared Spectroscopy

© 1999 Roger Rowlett

 Infrared Spectroscopy
• Near IR—780-2500 nm (12,800-4000 cm–1)
• Mid IR—2500-50,000 nm (4000-200 cm–1)
• Far IR—50-1000 mm (200-10 cm–1)
• Mid IR most useful (chemical fingerprinting)
• These aboroptions represent vibrational and
rotational transitions
• Change in dipole moment required during
vibration for IR absorption to occur
© 1999 Roger Rowlett
Typical Mid-IR Spectrum

© 1999 Roger Rowlett

 Types of
Molecular
Vibrations

© 1999 Roger Rowlett

 Vibrational Modes
• Linear molecules: 3N – 5 modes
• Nonlinear molecules: 3N – 6 modes
• Polyatomic molecules may exhibit
coupling which will alter predicted
frequencies:
– stretching-stretching
Note: strong coupling requires
– bending-bending interaction with a common atom
– stretching-bending
© 1999 Roger Rowlett
 Carbon Dioxide
• Stretching :

2330 cm–1

• Bending (scissoring, degenerate):

667 cm–1 667 cm–1

© 1999 Roger Rowlett
Water

© 1999 Roger Rowlett

 Quantitative Aspects
• Frequency of molecular vibration,  (s–1)
depends on mass of atoms and strength
of bond: 1 k

2 
• The force constant, k, is about 500 N/m
for a typical single bond
• The reduced mass, (kg), is defined
m1m2

m1  m2
© 1999 Roger Rowlett
 IR / Vibrational spectroscopy
• Ev = (v + 1/2) (h/2) (k/)
1/2

• v = 0, 1, 2, 3, …
Selection rules:
v = 1 & p must change during vibration
Let e = wavenumber of transition then “energy”:
v = (v + 1/2) e
• Untrue for real molecules since parabolic potential does
not allow for bond breaking.
v = (v + 1/2) e - (v + 1/2)2 e xe
– where xe is the anharmonicity constant
© 1999 Roger Rowlett
 Differences?
• Energy levels unequally
spaced, converging at high
energy. The amount of
distortion increases with
increasing energy. 6 6
• All transitions are no longer 5 5
the same
 v > 1 are allowed 4 4

E 3 3
–fundamental 01
2 2
–overtone 02
–hot band 12 1 1

0 0
0

© 1999 Roger Rowlett

 Example
• HCl has a fundamental band at 2,885.9 cm-1, and an
overtone at 5,668.1cm-1.
Calculate e and the anharmonicity constant xe.
v = (v + 1/2) e - (v + 1/2)2 e xe
2 = (2 + 1/2) e - (2 + 1/2)2 e xe
1 = (1 + 1/2) e - (1 + 1/2)2 e xe
0 = (0 + 1/2) e - (0 + 1/2)2 e xe
2 - 0 = 2e - 6e xe= 5,668.1
1 - 0 = e - 2e xe= 2,885.9
e = 2,989.6 cm-1 e xe = 51.9 cm-1 xe = 0.0174
© 1999 Roger Rowlett
 High resolution infrared
Ev = (v + 1/2) (h/2) (k/m)1/2
v = (v + 1/2) 
EJ = J(J+1) (h2/8I)
J = J(J + 1) v
Vibrational + rotational energy changes
 v,J = (v + 1/2) J(J + 1) v
• Selection rule: v=+1, J=1
– Rotational energy change must accompany a
vibrational energy change.
© 1999 Roger Rowlett
Vibrational + rotational changes in the IR
J '= 3

'
V IB R A TIO N A L
J=2 EX C ITED STA TE
J '= 1
v = 1 , J '= 0

J=3

V IB R A T IO N A L
J=2
G R O U N D STA TE
J=1
v= 0 , J= 0

© 1999 Roger Rowlett

 Hi-resolution spectrum of HCl

• Above the “gap”; J = +1

• Below the “gap”: J = –1
• Intensities mirror populations of starting levels

© 1999 Roger Rowlett

 Example: HBr
Lines at … 2590.95, 2575.19, 2542.25, 2525.09, ... cm-
1

• Difference is roughly 15 except between 2nd & 3rd

where it is double this. Hence, missing transition lies
around 2560 cm-1.
So 2575 is (v=0,J=0)  (v=1,J=1) & 2590 is (v=0,J=1)  (v=1,J=2)
So 2542 is (v=0,J=1)  (v=1,J=0) & 2525 is (v=0,J=2)  (v=1,J=1)
(2575.19 - 2525.25) = 6B0 B0=8.35 cm-1
(2590.95 - 2542.25) = 6B1 B1=8.12 cm-1
• Missing transition at 2542.25 + 2B0 = 2558.95 cm-1
© 1999 Roger Rowlett
Vibrational Spectroscopy

© 1999 Roger Rowlett

© 1999 Roger Rowlett
© 1999 Roger Rowlett
© 1999 Roger Rowlett
© 1999 Roger Rowlett
Line ν (cm-1) Δν (cm-1) Line ν (cm-1) Δν (cm-1)

P(1) 2139.43 R(0) 2147.08

3.88 3.78
P(2) 2135.55 R(1) 2150.86
3.92 3.73
P(3) 2131.63 R(2) 2154.59
3.95 3.72
P(4) 2127.68 R(3) 2158.31
3.98 3.66
P(5) 2123.70 R(4) 2161.97

© 1999 Roger Rowlett

 Instrumentation

• Source
• Wavelength selector
• Sample holder
• Detector

© 1999 Roger Rowlett

 Sources
• Tungsten lamp
• Nernst glower
• Globar
• Nichrome coil
• Mercury arc lamp
• Carbon dioxide laser

© 1999 Roger Rowlett

 Sources
• Tungsten lamp
– Near IR only
• Nernst glower
• Globar
• Nichrome coil
• Mercury arc lamp
• Carbon dioxide laser

© 1999 Roger Rowlett

 Sources
• Tungsten lamp
• Nernst glower
– A 2-5 cm rod, 1-3 mm in diameter, composed of oxides of Zr, Y,
Th
– Electrically heated to 1200-2000K, emits mid-IR
– Current-limiting required (R/T < 0)
• Globar
• Nichrome coil
• Mercury arc lamp
• Carbon dioxide laser

© 1999 Roger Rowlett

 Sources
• Tungsten lamp
• Nernst glower
• Globar
– Silicon carbide rod, 6 x 50 mm
– Electrically heated to 1600K, emits mid-IR; better than glower at
<5 m
– Self-starting (R/T > 0)
• Nichrome coil
• Mercury arc lamp
• Carbon dioxide laser

© 1999 Roger Rowlett

 Sources
• Tungsten lamp
• Nernst glower
• Globar
• Nichrome coil
– Electrically heated to 1100K, emits mid-IR
• Mercury arc lamp
• Carbon dioxide laser

© 1999 Roger Rowlett

 Sources
• Tungsten lamp
• Nernst glower
• Globar
• Nichrome coil
• Mercury arc lamp
– High pressure (> 1 atm), emits far IR
• Carbon dioxide laser

© 1999 Roger Rowlett

 Sources
• Tungsten lamp
• Nernst glower
• Globar
• Nichrome coil
• Mercury arc lamp
• Carbon dioxide laser
– Many discrete lines from 9-11 m
– Very intense

© 1999 Roger Rowlett

 Wavelength Selection

• Interference filters, wedges

• Prisms
– Quartz (0.8-3 m) near IR
– LiF (1-5 m) near IR
– NaCl (2.5-20 m) mid-IR
– KBr, CsBr (15-40 m) far-IR
• Reflection gratings (Al coated)

© 1999 Roger Rowlett

 Detectors

• Thermopiles
• Thermistors
• Photoconductors
• Pyroelectric detectors
• Golay detector

© 1999 Roger Rowlett

Thermopiles
• Thermocouples
• Typical junctions
include Pt-Ag, Sb-Bi
• Sensitive to
temperatures
changes of 1 K

© 1999 Roger Rowlett

 Thermistors and
Photoconductors
• Thermistors
– Temperature-sensitive resistors
– Typically have very large R/T values
• Photoconductors
– Temperature-sensitive resistors
– PbS is typical
– Near IR only

© 1999 Roger Rowlett

 Pyroelectric Detectors

• Temperature sensitive capacitors

– Triglycine sulfate (TGS)
– Lithium tantalate (LiTaO3)
– Lithium niobate (LiNbO3)
• Materials have large (dipole)/T
• Response times of 1 ms or less (fast for
IR detectors)
© 1999 Roger Rowlett
 Golay Detector
• Essentially a Xe gas thermometer
• Expensive, not better than other detectors in
near to mid IR
• Very useful in far IR

© 1999 Roger Rowlett

 Sample Handling
• Gases
– 10 cm to 120 m pathlengths (may utilize multiple reflection)
• Liquids—thin (0.1 mm) layers used
– Water or ROH are no-no’s (Why?)
– 10% in CCl4 (4000-1333 cm–1)
– 10% in CS2 (1333-650 cm –1)
– CHCl3, CH3CN, acetone for more polar substances
– neat liquids
• Evaporated Films

© 1999 Roger Rowlett

Liquid IR Cell

© 1999 Roger Rowlett

 More Sample Handling
• Mulls (pastes)
– Finely ground solid and mulling agent fashioned
into a thin film
– Mineral oil, Nujol™, Fluorolubes used as
mulling agents
• Pellets
– Finely ground solid (1-100 g) + KBr
– Pressed into a transparent disk at 60,000 to
100,000 psi

© 1999 Roger Rowlett

Multiple Internal Reflectance
(For spectra of surfaces)

© 1999 Roger Rowlett

Dispersive IR Spectrophotometer

• Monochromator
after sample
• Low frequency
chopper (10 min–1)
• Works on nulling
principle: reference
beam attenuated to
match sample beam

© 1999 Roger Rowlett

Michelson Interferometer
• Difference in path
length,
 = 2(M-F)
• If mirror is moved
at velocity vm, and
time to move /2
= ,
• Then detector
output frequency
f = 1/ = 2vm /
• Note:
 2v m 
f  
 c 
© 1999 Roger Rowlett
 Why is this important?
• Consider, a typical mid-IR frequency is 1.5 
1013 s–1, impossibly fast for any detector to
directly follow
• But at typical mirror speeds, vm = 0.01 - 10
cm/s, the scaling factor 2vm/c is (6.7-6700) 
10–13
• Thus, the corresponding interferometer output
frequency is 10-10,000 Hz, within the range of
pyroelectric detectors

© 1999 Roger Rowlett

 Detector output for:
1/ = 400 cm-1 vm = 0.01 cm/s

© 1999 Roger Rowlett

 Detector output for:
1/ = 600 cm-1 vm = 0.01 cm/s

© 1999 Roger Rowlett

 Detector output for:
1/ = 400 and 600 cm-1 vm = 0.01 cm/s

© 1999 Roger Rowlett

 So what can we do with this?
• All frequencies in the original IR beam are
“encoded” at a frequency at the detector
determined by the scaling factor 2vm/c
• Thus there is sufficient information to
reconstruct the entire spectrum if only…
• …we could extract the intensity of each
frequency in the spectrum from the complex
signal received by the detector

© 1999 Roger Rowlett

 Fourier Transforms,
Oversimplified
• How we “decode” the complex
interferogram
1. Multiply interferogram by cos2ft (a cosine
wave of frequency f)
2. Integrate result
3. Repeat for all desired f values
• Mathematically speaking:

P    P t   cos 2ftdt
 © 1999 Roger Rowlett
 What you get
• If a signal “encoded” in the interferogram
matches the frequency f, the integral will be
large
• If a signal “encoded” in the interferogram
does not match the frequency f, the integral
will be small
• A plot of the integral values vs. f yields an
absorption spectrum (Voila!)

© 1999 Roger Rowlett

 Advantages of FT
Spectroscopy
• Jaquinot (throughput) advantage
– Few optical elements to attenuate or scatter
radiation  more radiant power reaches the
detector,  S/N
• Felgett (multiplex) advantage
– Entire spectrum collected rapidly at once
– Allows study of transient species
– Allows signal averaging by adding together
spectra
© 1999 Roger Rowlett
 Signal Averaging
• The magnitude of the signal increases as the number
of scans, N
• The magnitude of noise increases as the square root
of the number of scans, N
• Therefore, S/N increases as N
• Signal averaging enables the study of dilute solutions
• Consider:
– dispersive instrument: 600 s/spectrum
– FT instrument: 1 s/spectrum
– In 600 s an FT instrument can improve S/N by 600  25-fold

© 1999 Roger Rowlett

FT-IR Spectrophotometer

© 1999 Roger Rowlett