Lecture 6: Spectral Lineshapes

1. Background introduction
2. Types of line broadening ϕν
3. Voigt profiles ϕν(ν0)
ΔνN
4. Uses of quantitative lineshape
measurements ϕν(ν0)/2
ν0 ν
5. Working examples
A typical lineshape function
 1. Background introduction
 Beer’s Law
Recall: Gas
Io(ν) I(ν)
Collimated light @ ν
L

absorption coefficient @ ν, cm-1

T  I / I 0   I / I0  exp k L 
intensity or power @ ν spectral intensity @ ν

 
k , cm 1  S12 cm 1  s 1  s    h
S12 cm 1  s 1  

n1 B12 1  exp h / kT 
 c 
Line strength, ∫line kνdν 2 g 
 n1 A21  2 1  exp h / kT 
k  8  g1 
    ,    d  1
 k d
line

 The lineshape function
2
 1. Background introduction
 Alternate forms of ν, ϕ, S12
_ 1
 , s   c , cm 
1 c
 ν

 ϕ  , cm  c , s 
 S12  A common form of S

S12 , cm 2  S12 , cm 1s 1  / c k , cm 1  S12 cm 1  s 1  s 
 Another common form 2 1
Partial pressure
Notes:
n1
 
S12 , cm  2 / atm  S12 , cm  2 / Pi , atm  
S12 , cm s
cPi , atm
of absorber
1.
Pi , atm c n1 g2
n1  A 1  exp h / kT 
8 Pi , atm
2 21
 g1
ni kT / 1.013 10 6 dynes/cm2 atm
1  n1  6
 10
 S
12 
, cm 2 atm 1 Pi , atm 

kT  ni  Boltzmann fraction  * cm 1  molec 
  S ,  n,
2  i 3 
2. k , cm 1  S12 , cm 2 /atm Pi , atm  , cm   molec/cm  cm 
HITRAN database lists S*
Pi , atm  P, atm  i 
Mole fraction (cm/molec), usually at Tref = 296K 3
 1. Background introduction
 How are S12 and ϕ measured?

 High-resolution absorption experiments

Tν kν
Area=S12
1.0
 ln I / I 0 

0
ν0 ν ν0 ν
ϕ

k  Area=1
  Shape determined by

line
k d FWHM main broadening mechanism

ν0 ν
4
 2. Types of line broadening
 Brief overview

1. Natural broadening
 Result of finite radiative lifetime
Lorentzian
Homogeneous (affects 2. Collisional/pressure broadening
all molecules equally)
 Finite lifetime in quantum state owing to
collisions
Gaussian
Inhomogeneous 3. Doppler broadening
(affects certain class of  Thermal motion
molecule)

Lorentzian 4. Voigt profile

+ Gaussian  Convolution of 1-3

5
 2. Types of line broadening
 Natural line broadening

1. Heisenberg uncertainty principle: Eu tu  h / 2

u (upper level)
1
hν0 Aul 
u
l (lower level)
Decay rate

Eu  uncertainty in energy of u

tu   , the uncertainty in time of occupation of u
Eu  h u  h / 2  / t   rad    u  1 / 2 rad “lifetime” limited

2. In general 0 for ground state

1  1 1  (natural broadening)
 N   u   l    
2   u  l 
6
 2. Types of line broadening
 Natural line broadening
3. Typical values
 Electronic transitions:
 u ~ 10 8 s   N ~ 1.6  10 7 s 1
 N , cm 1   N / c  5  10  4 cm 1
 Vib-rot transitions
 u ~ 10 2 s   N ~ 16s 1 ,  N , cm 1  5  10 10 cm 1
 These are typically much smaller than ∆νD and ∆νC

4. Lineshape function – “Lorentzian” – follows from Fourier transform

1  N / 2
ϕν   N 
   0 2   N / 2 2
ϕν(ν0)
ΔνN 2 1
Note: a) max    0  
ϕν(ν0)/2   N
ν0 ν b)    0   N / 2     0  / 2
7
 2. Types of line broadening
 Natural line broadening
Lineshape derivation from damped oscillator model (Ref. Demtröder)
 = Damping ratio x  x  02 x  0, 02  k / m xt   x0 exp t / 2 cos t   / 2 sin t 
x0   x0 , x 0   0   02   2 / 4
n  1/ 2
in units of s-1
Small damping (<<ω0) Amplitude of x(t) decrease 
x xt   x0 exp t / 2  cos 0t frequency of emitted radiation
x0 is no longer monochromatic
exp(–γt/2) |A(ω)|2
1 
xt    A expit d
2 2 0
t γ 1 
A    xt exp it dt
2 

x F ω0 ω x0  1 1 
   
 i  0    / 2 i   0    / 2 
x0 |A(ω)|2 8
I    A A*  , L  I / I 0

t 1  /2
L  0  
ω0 ω    0 2   / 22
8
 2. Types of line broadening
 Collision broadening
1. Also lifetime limited – time set by collision time interval

Effective area
A A
A   AB
2 .
v
B  B = optical collision diameter of B
Optical cross-
section
 A  B
 AB 
2
Z BA  # collision/s of a single B with all A For a mixture,
#/cc 8kT

 8kT 
 Z B   n A   AB
2

 n A   2
c    AB
 AB
AB A
 
8
m A mB  P  X A   AB
2

 AB   AB kT
m A  mB A

P, dynes/cm 2  1.013  10 6 P, atm 

9
 2. Types of line broadening
 Collision broadening
1. Also lifetime limited – time set by collision time interval
8kT 8
Z B   n A   AB
2
  P  X A   AB
2

A  AB A  AB kT
Since
 C , s 1  P, atm  X A 2 A
1  1 1  ZB
 C , s 
1    A
2γA = colli. halfwidth, i.e.,
2   
 coll ,upper  coll ,lower  FWHM per atm. pressure
8
 P, atm  X A   AB
2
 1.013  10 6  Notes:
A  AB kT  C , cm 1   C , s 1 / c
 
2 A ,s -1/atm
2 , cm 1 / atm  2 , s 1 / atm / c
1  C / 2
2. Lineshape function – Lorentzian   coll 
   0 2   C / 2 2
3. Crude approximation 2 T   2 300 / T 
300 n n=1/2 for hard sphere

 
cm-1/atm ≈0.1cm-1/atm 10
 2. Types of line broadening
 Collision broadening  C , cm 1  P, atm  X A 2 A
A
Example: Pressure broadening of CO with 2γA in cm /atm
-1

R(9) line of CO’s 2nd overtone, 50ppm in Air, 300K, 1.0atm

Species population: 77% N2, 20% O2, 2% H2O (85% humidity) 380ppm CO2

Species, A Mole Fraction, XA 2γCO-A (300K) cm-1/atm

N2 0.77 0.116
H2O 0.02 0.232
CO 50e-6 0.128
CO2 380e-6 0.146
O2 0.21 0.102


 C  P X N 2  2 CO  N 2  X H 2O  2 H 2O  N 2  X CO  2 CO CO
 X CO2  2 CO CO2  X O2  2 CO O2 
 0.115cm 1
11
 2. Types of line broadening
 Collision broadening
Some collisional broadening coefficients 2γ [cm-1/atm] in Ar and N2 at 300K
Species Wavelength [nm] Ar N2
Na 589 0.70 0.49
K 770 1.01 0.82
Rb 421 2.21 1.51
OH 306 0.09 0.10
NH 335 0.038
NO 225 0.50 0.58
NO 5300 0.09 0.12
CO 4700 0.09 0.11
HCN 3000 0.12 0.24

Some collisional broadening coefficients 2γ [cm-1/atm] in Ar and N2 at 2000K

Species00 Wavelength [nm] Ar N2
NO 225 0.14 0.14
OH 306 0.034 0.04
NH 335 0.038

12
 2. Types of line broadening
 Doppler broadening
1. Moving molecules see different frequency (Doppler shift)

 app  act   act u / c  u /   app   act 1 u / c 

molec. velocity along beam path

2. Gaussian velocity distribution function (leads to Gaussian ϕ ν )

2 ln 2   2 ln 2  
2

    exp     0  
  D    D  

ϕ ν0
 Aside:
2kT ln 2
 D FWHM   2 0 Maxwellian velocity distribution
mc 2 1/ 2
 m   mU x2 
T f U x     exp  
 D FWHM   7.17  107 0  2kT   2kT 
M
g/mole of
emitter/absorber
13
 2. Types of line broadening
 Stark broadening
 Important in charged gases, i.e., plasmas.
 Coulomb forces perturb energy levels

 Types of instrument broadening

 Instruments have insufficient resolution
 Powerful lasers can perturb populations away from equilibrium
(saturation effect)
 Transit-time broadening

 Another type of lifetime-limited broadening is transit-time

broadening
T
r
a
n
s
i
t
t
i
m
e
 D /V
Laser beam
  transit V / D
V
Gas
for apparent broadening of an abs. line
D
Reference: Demtröder p.85-p.88
14
 2. Types of line broadening
 Examples
1st Example:
T = 300K, M = 30g/mole, P = 1atm
 Electronic transition  Vib-rot transition  λIR=10λvis
(λ=600nm, ν=5x1014s-1) (λ=6μm, ν=5x1013s-1)
 D  1.1 109 s 1 ~ 0.04cm 1  D ~ 1.1 108 s 1 ~ 0.004cm 1
 D   N ~ 10 7 s 1  C ~ 3  109 s 1  0.1cm 1
 C ~ 3  109 s 1  0.1cm 1   D   C
  D   C
2nd Example:
T = 2700K, M = 30g/mole, P = 1atm
 Electronic transition  Vib-rot transition
(λ=600nm, ν=5x1014s-1) (λ=6μm, ν=5x1013s-1)
 D ~ 0.11cm 1   C ~ 0.03cm 1  D ~ 0.01cm 1   C ~ 0.03cm 1
~T1/2 ~T-1/2 15
 2. Types of line broadening

 Conclusions

 Doppler broadening most significant at:

Low P, high T, small λ

 Collision broadening most significant at:

High P, low T, large λ

 Many conditions require consideration of both effects

Together  Voigt profile!

16
 3. Voigt Profiles
Voigt function
1. Dominant types of broadening
V a, w a=0
 Collision broadening
V   k (pure Doppler)
 
 Doppler broadening  D  0  k D  0 
2. Voigt profile a=1
3. Line-shifting mechanisms a=2

0.83
w  2 ln 2   0  /  D
 3. Voigt profiles
3.1. Dominant types of broadening
 Collision broadening review
1  C / 2
 C    Lorentzian form “lifetime limited”
   0    C / 2 
2 2

 
  C , s 1    X A 2 A P, atm of mixture 
 A 
mole fraction of A coll. width/atm for A as coll. partner,  1 / T
 Typical value of 2γA ~ 0.1cm-1/atm (or 0.3x1010s-1/atm)

1 8 1
2
 Aside: A , s1
 1.01310 
6 2
 
ABkT
AB
c T
if σAB is constant
0.1cm1 / atm 31010cm/s 0.31010s1 / atm
 A type of “Homogenous broadening”, i.e., same for all molecules of
absorbing species

18
 3. Voigt profiles
3.1. Dominant types of broadening
 Doppler broadening review

2 ln 2   2 ln 2 
2

     exp     0   Gaussian form
  D    D  

  
  0 
1/ 2
 2kT ln 2 
  D , s 1  2 2   0  7 . 17  10 7
0 T / M
 mc 
FWHM g/mole of absorber/emitter

 Typical value
 D   600nm, M  30   0.35  1010 s 1 3000K   0.12cm 1
 0.11010 s 1 300K   0.03cm 1
 This is a type of “Inhomogenous broadening”, i.e., depends on
specific velocity class of molecule

19
 3. Voigt profiles
3.1. Dominant types of broadening
 Comparison of ϕD and ϕC (for same ∆ν(FWHM))
 Both have same area (unity)
 Peak heights
2 ln 2
  0 Dopp   0.94 /  D
  D
2 1
  0 coll   0.637 /  C
  C
for  C /  D  1
  0 Dopp  1.48  0 coll
 Gaussian: higher near peak
 Lorentzian: higher in wings
 Some exceptions/improved models
 Collision narrowing (low-pressure phenomenon)
 Galatry profiles, others, with additional parameters
 Stark broadening  Plasma phenomenon
Ready to combine Doppler & collision broadening; done via Voigt profile 20
 3.2. Voigt profile
 Physical argument
The physical argument employed in establishing the Voigt profile is that the
effects of Doppler & collision broadening are decoupled. Thus we argue that
every point on a collision-broadened lineshape is further broadened by
Doppler effects.

Convolution: V     D  * C     D u C   u du


 1

 2 ln 2   
2

 C / 2 2 ln 2
V        exp       d
  
  C / 2     0        D    D   
2 2
 

V   
2 ln 2  a  
exp  y 2 dy 

  D   a 2  w  y 2  V a, w

 
 D 0  the “Voigt function” (V≤1)

where a  ln 2    C   N  /  D  ln 2 C /  D

w  2 ln 2   0  /  D
y  2 ln 2 /  D integrated out 
21
 3.2. Voigt profile

V   
2 ln 2  a  
exp  y 2 dy   


  D   a 2  w  y 2  V a , w 


 
 D 0  the “Voigt function” (V≤1)

Notes: V a, w a=0

V   k (pure Doppler)
1. V     D  0 V a, w , so that  
 D  0  k D  0 
2. k  k0V a, w a=1
k D  0  , the line-center
Spec. abs. coeff. spec. abs. coeff. for a=2
Doppler broadening
Recall: k  S
0.83
 
3. V a,0   exp a erfca 
2
w  2 ln 2   0  /  D
V  0   D  0   expa 2 erfca  a=1: expa erfca   0.43
2

a=2: expa erfca   0.257

2

22
 3.2. Voigt profile
 Voigt table

23
 3.2. Voigt profile
 Procedure
 Given: T, M, ν0, P, σ, or 2γ
 Desire: ϕ(ν)

1. Compute: ∆νD and ϕD(ν0)

2. Compute: ∆νC
3. Compute: a  ln 2 C /  D
4. Pick w, enter table (for a) and obtain k / k D  0    /  D  0 
5. Solve for ν – ν0 (and hence ν) for that w
6. Results: ϕ(ν) vs ν – ν0

24
 3.2. Voigt profile
 Procedure
 Given: T, M, ν0, P, σ, or 2γ
 Desire: ϕ(ν)

1. Compute: ∆νD and ϕD(ν0)

2. Compute: ∆νC
3. Compute: a  ln 2 C /  D
4. Pick w, enter table (for a) and obtain k / k D  0    /  D  0 
5. Solve for ν – ν0 (and hence ν) for that w
6. Results: ϕ(ν) vs ν – ν0

 Refinements
 Galatry profiles (collision narrowing)
 Berman profiles (speed-dependent broadening)

25
 3. Voigt profiles
3.3. Line shifting mechanisms
 Pressure shift of absorption lines
 Interaction between two collision partners can have a perturbing
effect on the intermolecular potential of the molecule
 differences in the energy level spacings
 pressure shift
Notes:
 S  P  X A A
A 1. While 2γ>0, δ can be + or –
cm-1/atm
M 2. E.g., average values for IR H2O
T 
 A T    A T0  0  spectra: δ = –0.017cm-1/atm, M=0.96
T 

 Doppler shift 
v kν
Abs. line for static sample
Io(ν0) θ
δν = shift in frequency
u = |v|cosθ required to excite this
transition!
   0 u / c 
ν0 
26
 4. Uses of quantitative lineshape measurements
 Species concentration and pressure
 Integrated absorbance area
Ai
 Xj 
Ai     d  Si PX j L Si PL


Line strength of Species mole Pathlength Ai

P
the transition Pressure fraction Si X j L
 Temperature
 FWHM of lineshape gives T in Doppler-limited applications
 Two-line technique with non-negligible pressure broadening

S T , 1  S T0 , 1    hc  "  1 1  T sensitivity:

R 
S T , 2  S T0 , 2 

exp    E1  E2" 
   1 dR  
hc  E1"  E2" 
  k   T T0  % / K     2
100
R dT  k  T
hc "
k

E1  E2"  Large ∆E" for higher sensitivity;
T Absorbance: 0.1<α<2.3
S T  hc E1  E2
ln R  ln 2 0 
"
 "
 Tradeoff between acceptable
S1 T0  k T0 absorbance and T sensitivity.
27
 4. Uses of quantitative lineshape measurements
 Examples
 1st Example: Spectrally resolved absorption of sodium (Na) in a heated cell
λ = 589nm, T = 1600K, P = 1atm What is PNa?

1) Find k 0   1 / L  ln I / I 0 
0 ln 2 C ln 2 0.21
a   1.75
2) Find   0   D 0.10
3) Find Pi Interpolate Voigt table

 C  P  2 1600K  V a, w  V 1.75,0  0.2852

300
 P  2 300K   0.21cm 1 2 ln 2 2 ln 2
1600 D  0     9.39cm
 D  0.10 
589 10 cm  16978cm
7 1 1
  0   D  0 V 0
1/ 2
 9.39  0.2852  2.68cm
  7.17 10 16978cm 
7  1600  1

D
23   k
Solve for Pi using Pi 
 0.10cm 1 S  0 
Could also have solved for T from lineshape data 28
 4. Uses of quantitative lineshape measurements
 Examples
 2nd Example: Atomic H velocity
LIF (Laser Induced Fluorescence) in an arcjet thruster is used to
measure the Doppler shift of atomic hydrogen at 656nm.

Doppler shift: δν = 0.70cm-1

Use line position to infer velocity
The corresponding velocity component is found
c 3 108 m/s  0.70cm 1
u   13800m/s
0 15.232cm 1

LIF detector

u
Incident laser light, variable ν
Supersonic arcjet exhaust
29
 5. Working examples - 1
 CW laser strategies for multi-parameter measurements of high-
speed flows containing NO

Schematic for NO
LIF experiments

30
 5. Working examples - 1
 CW laser strategies for multi-parameter measurements of high-
speed flows containing NO

31
 5. Working examples - 2
 TDL mass flux sensor
 Full-scale aero-engine inlet

32
 5. Working examples - 2
 TDL mass flux sensor
 Sensor tests in Pratt and Whitney engine inlet

 Bellmouth installed on inlet of commercial engine (Airbus 318)

 Sensor hardware remotely operated in control room
 TDL beams mounted in engine bellmouth

33
 5. Working examples - 2
 TDL mass flux sensor
 P & W mass flux versus TDL sensor measurements

 TDL data agrees well (1.2% in V and 1.5% in ρ) w/ test stand instrumentation
 Flow model employed to account for non-uniformities
 Success in non-uniform flow suggest other potential applications

34
Next: Electronic Spectra of Diatomics

 Term Symbols, Molecular Models

 Rigid Rotor, Symmetric Top
 Hund’s Cases
 Quantitative Absorption