You are on page 1of 43

Andrei Tokmakoff, MIT Department of Chemistry, 4/1/2009

p. 11-1

11. NONLINEAR SPECTROSCOPY 11.1. Introduction
Spectroscopy comes from the Latin “spectron” for spirit or ghost and the Greek “σκοπιεν” for to see. These roots are very telling, because in molecular spectroscopy you use light to interrogate matter, but you actually never see the molecules, only their influence on the light. Different spectroscopies give you different perspectives. This indirect contact with the microscopic targets means that the interpretation of spectroscopy in some manner requires a model, whether it is stated or not. Modeling and laboratory practice of spectroscopy are dependent on one another, and therefore a spectroscopy is only as useful as its ability to distinguish different models. The observables that we have to extract microscopic information in traditional spectroscopy are resonance frequencies, spectral amplitudes, and lineshapes. We can imagine studying these spectral features as a function of control variables for the light field (amplitude, frequency, polarization, phase, etc.) or for the sample (for instance a systematic variation of the physical properties of the sample). In complex systems, those in which there are many interacting degrees of freedom and in which spectra become congested or featureless, the interpretation of traditional spectra is plagued by a number of ambiguities. This is particularly the case for spectroscopy of disordered condensed phases, where spectroscopy is the primary tool for describing molecular structure, interactions and relaxation, kinetics and dynamics, and tremendous challenges exist on understanding the variation and dynamics of molecular structures. This is the reason for using nonlinear spectroscopy, in which multiple light-matter interactions can be used to correlate different spectral features and dissect complex spectra. We can resonantly drive one spectroscopic feature and see how another is influenced, or we can introduce time delays to see how properties change with time. Absorption or emission spectroscopies are referred to as linear spectroscopy, because they involve a weak light-matter interaction with one primary incident radiation field, and are typically presented through a single frequency axis. The ambiguities that arise when interpreting linear spectroscopy can be illustrated through two examples:

Andrei Tokmakoff, MIT Department of Chemistry, 4/1/2009

p. 11-2


Absorption spectrum with two peaks. Do these resonance arise from different, non-interacting molecules, or are these coupled quantum states of the same molecule? couplings directly). (One or cannot resolve spectral correlations


Broad lineshapes. Can you distinguish whether it is a homogeneous lineshape broadened by fast irreversible relaxation or an inhomogeneous lineshape arising from a static distribution of different frequencies? (Linear spectra cannot uniquely interpret linebroadening mechanism, or decompose heterogeneous behavior in the sample).



In the end effect linear spectroscopy does not offer systematic ways of attacking these types of problems. It also has little ability to interpret dynamics and relaxation. These issues take on more urgency in the condensed phase, when lineshapes become broad and spectra are congested.

Andrei Tokmakoff, MIT Department of Chemistry, 4/1/2009

p. 11-3

Nonlinear spectroscopy provides a way of resolving these scenarios because it uses multiple light fields with independent control over frequency or time-ordering in order to probe correlations between different spectral features. For instance, the above examples could be interpreted with the use of a double-resonance experiment that reveals how excitation at one frequency ω1 influences absorption at another frequency ω2. What is nonlinear spectroscopy? Linear spectroscopy commonly refers to light-matter interaction with one primary incident radiation field which is weak, and can be treated as a linear response between the incident light and the matter. From a quantum mechanical view of the light field, it is often conceived as a “one photon in/one photon out” measurement. Nonlinear spectroscopy is used to refer to cases that fall outside this view, including: (1) Watching the response of matter subjected to interactions with two or more independent incident fields, and (2) the case where linear response theory is inadequate for treating how the material behaves, as in the case of very intense incident radiation. If we work within the electric dipole Hamiltonian, nonlinear experiments can be expressed in terms of three or more transition matrix elements. The response of the matter in linear experiments will scale as μab

or μab μba , whereas in nonlinear experiments will take a

form such as μab μbc μca . Our approach to describing nonlinear spectroscopy will use the electric dipole Hamiltonian and a perturbation theory expansion of the dipole operator.

Andrei Tokmakoff, MIT Department of Chemistry, 4/1/2009

p. 11-4

We will specifically be dealing with the description of coherent nonlinear spectroscopy, which is the term used to describe the case where one or more input fields coherently act on the dipoles of the sample to generate a macroscopic oscillating polarization. This polarization acts as a source to radiate a signal that we detect in a well-defined direction. This class includes experiments such as pump-probes, transient gratings, photon echoes, and coherent Raman methods. However understanding these experiments allows one to rather quickly generalize to other techniques. Detection: Coherent
I coherent ∝ ∑ i μi

I spont . ∝ ∑ μi
i 2

Dipoles are driven coherently, and radiate with constructive interference in direction k sig Absorption

Dipoles radiate independently

Esig ∝ sin θ
Fluorescence, phosphorescence, Raman, and light scattering


Fluorescence-detected nonlinear spectroscopy, i.e. stimulated emission pumping, time-dependent Stokes shift

Pump-probe transient absorption, photon echoes, transient gratings, CARS, impulsive Raman scattering

plane wave was ∇2 E ( r . which we term the signal Esig . For the case of vibrational transitions where non-radiative relaxation is typically a picoseconds process and radiative relaxation is a μs or longer process.) In the electric dipole approximation. The description of coherent nonlinear spectroscopies is rooted in the calculation of the polarization. 11-5 Spontaneous and coherent signals are both emitted from all samples. From our earlier description of freely propagating electromagnetic waves. t ) − 2 1 ∂ E (r . c2 ∂t 2 (4) which gave a solution for a sinusoidal oscillating field with frequency ω propagating in the direction of the wavevector k. 4/1/2009 p. the polarization is one term in the current and charge densities that you put into Maxwell’s equations.t ) =0. the wave equation for a transverse. For electronic transitions in which dephasing is typically much faster than the radiative lifetime.Andrei Tokmakoff. and for a molecular system is expressed as a sum over the displacement of all charges for all molecules being interrogated by the light Sum over molecules: P ( r ) = ∑ μm δ ( r − Rm ) m (1) Sum over charges on molecules: μm ≡ ∑ qmα ( rmα − Rm ) α (2) In coherent spectroscopies. In the present case. t ) = ∂t 2 ∂t 2 c2 c2 (5) The polarization can be described by solutions of the form . coherently oscillating charge distribution P (ω ) = χ E (ω ) (3) as dictated by the susceptibility of the sample. the input fields E act to create a macroscopic. the relative amplitude of the two depend on the time-scale of dephasing within the sample. The polarization acts as a source to radiate a new electromagnetic field. P . the polarization acts as a source −an accelerated charge− and we can write ∇2 E ( r .t ) 4π ∂ P ( r . spontaneous emission is not observed. MIT Department of Chemistry. (Remember that an accelerated charge radiates an electric field. spontaneous emission is the dominant emission process. The polarization is a macroscopic collective dipole moment per unit volume. t ) − 2 2 1 ∂ E (r . however.

for which the radiated signal amplitude grows and becomes directional as it propagates through the sample. P is given by (6) and the radiation of the ensemble only constructively adds along k sig . k sig = ∑ ± kn n (7) (8) ωsig = ∑ ±ωn . t ) = Esig ( r . although there is a π/2 phase shift between the two. because in the sample the polarization is related to the gradient of the field. in a wave vector matched direction k sig . which we will discuss more later. The emitted signal Esig ( t ) = i 2πωs ⎛ Δkl ⎞ iΔkl / 2 l P ( t ) sinc ⎜ ⎟e nc ⎝ 2 ⎠ (10) Here we note the oscillating polarization is proportional to the signal field.1 for an ensemble of dipoles that have been coherently driven by external fields. P ( r . MIT Department of Chemistry. Δk is the wave-vector mismatch between the wavevector of the polarization ′ and the radiated field k sig . Esig ∝ i P . 4/1/2009 p. k sig For the purpose of our work. (9) This solution comes from solving (5) for a thin sample of length l. we obtain the polarization from the expectation value of the dipole operator P (t ) ⇒ μ (t ) was popularized by Mukamel. For the radiated field we obtain Esig ( r . 11-6 ′ ⋅ r − iωsig t ) + c.2 As before our Hamiltonian can generally be written as H = H0 + V (t ) (11) The treatment we will use for the spectroscopy is semi-classical. t ) = P ( t ) exp ( iksig (6) As we will discuss further later. Although a single dipole radiates as a sin θ field distribution relative to the displacement of the charge. Esig .c. t ) exp ( i ksig ⋅ r − iωsig t ) + c. n These relationships enforce momentum and energy conservation for the problem.Andrei Tokmakoff. The oscillating polarization radiates a coherent signal field.c. and follows the formalism that (12) . the wavevector and frequency of the polarization depend on the frequency and wave vector of incident fields.

To describe this. and the electromagnetic fields V(t) are treated classically and take the standard form V (t ) = −μ ⋅ E (13) The fields only act to drive transitions between quantum states of the system. χ ′′ . Linear absorption spectroscopy Absorption is the simplest example of a coherent spectroscopy. Then we can write the measured intensity after the sample as . which we previously found was proportional to the imaginary part of the susceptibility. that is a signal field of the same frequency propagates in the direction of the transmitted excitation field. In the semi-classical picture. 4/1/2009 p. For one input field. all of the relevant information is in R ( t ) or χ (ω ) . 3 We consider one monochromatic field incident on the sample that resonantly drives dipoles in the sample to create a polarization. which subsequently re-radiate a signal field (free induction decay). so that A ≈ δ I and I in ≈ I out . For the perturbative case. In practice. P ( t ) = ∫ dτ R (τ ) E ( t − τ ) 0 ∞ (14) (15) P (ω ) = χ (ω ) E (ω ) Let’s begin with a frequency-domain description of the absorption spectrum. 11-7 where the material system is described by H0 and treated quantum mechanically. the polarization induced by the electromagnetic field radiates a signal field that is out-of-phase with the transmitted light. let’s take the change of intensity δ I = I in − I out to be small. an absorption spectrum is measured by characterizing the frequencydependent transmission decrease on adding the sample A = − log I out I in . nth-order perturbation theory will be used to describe the nonlinear signal derived from interacting with n electromagnetic fields. We take the interaction with the fields to be sufficiently weak that we can treat the problem with perturbation theory.Andrei Tokmakoff. Thus. MIT Department of Chemistry. the energy and momentum conservation conditions dictate that ωin = ωsig and kin = k sig .

Andrei Tokmakoff. 4/1/2009 p. h −∞ ⎣ Substituting eq. or more precisely. ⎣ μ I ( t ′ ) . i R (τ ) = θ (τ ) ( C (τ ) − C * (τ ) ) h (21) ( ) (20) So the linear response function is the sum of two correlation functions. the . B ] . we can calculate the polarization in the density matrix picture as P ( t ) = Tr μ I ( t ) ρ I(1) ( t ) ( ) (17) where the first-order expansion of the density matrix is i t ρ I(1) = − ∫ dt1 ⎡ VI ( t1 ) . and made use of the identity ⎡ ⎣ A. (14) and should recover the relationships to the dipole autocorrelation function that we discussed previously. MIT Department of Chemistry. Now comparing to eq. ρeq ⎤ P ( t ) = Tr ⎜ μ I ( t ) ∫ dt ′ ⎡ ⎟ ⎣ ⎦ −∞ h ⎝ ⎠ −i t = ∫ dt ′ E ( t ′ ) Tr μ I ( t ) ⎡ . A time-domain approach to absorption draws on eq. ρeq ⎤ ⎦. we see. as expected i ⎡ μ I (τ ) . (14). C ]⎤ ⎦=⎡ ⎣[ A. We see that as a result of the phase shift between the polarization and the radiated field that the absorbance is proportional to χ ′′ : δ I = 2 χ ′′ I in . (13) we find i t ⎛ ⎞ − μ I ( t ′ ) E ( t ′ ) . μ I ( 0 ) ⎤ ⎦ ρeq h 0 (18) ( ) (19) ( ) In the last line. [ B. C ⎤ ⎦ . Equating P ( t ) with μ ( t ) . ρeq ⎤ ⎦ h −∞ i ∞ = + ∫ dτ E ( t − τ ) Tr ⎡ ⎣ μ I (τ ) . we switched variables to the time interval τ = t − t ′ . 11-8 2 2 I out = Eout + Esig = Eout + ( iP ) 2 2 2 = Eout + i χ Ein ≈ Ein + i χ Ein = Ein 1 + i ( χ ′ + i χ ′′ ) 2 (16) I out = I in − δ I = I in (1 − 2 χ ′′ + K) ⇒ Here we have made use of the assumption that Ein >> χ . μ I ( 0 ) ⎦ ⎤ ρeq R (τ ) = θ (τ ) Tr ⎣ h imaginary part of the dipole correlation function.

⎣ −∞ −∞ −∞ ⎣ t (26) Let’s examine the second-order polarization in order to describe the nonlinear response function. We calculate P from the density matrix P ( t ) = Tr ( μ I ( t ) ρ I ( t ) ) = Tr μ I ρ I( 0) + Tr μ I ρ I(1) ( t ) + Tr μ I ρ I( 2 ) ( t ) + K As we wrote earlier. we will calculate the polarization arising from interactions with multiple fields. ⎡VI ( tn −1 ) . 11-9 C (τ ) = Tr ( μ I (τ ) μ I ( 0 ) ρ eq ) C*(τ ) = Tr ( μ I (τ ) ρ eq μ I ( 0 ) ) (22) Also. ρeq ⎤ ⎦ K⎤ ⎣K . i. Nonlinear Polarization For nonlinear spectroscopy. MIT Department of Chemistry. We will use a perturbative expansion of P in powers of the incoming fields P ( t ) = P ( 0) + P (1) + P ( 2 ) + P (3) + L n 2 (23) where P ( ) refers to the polarization arising from n incident light fields. 4/1/2009 p.Andrei Tokmakoff. ρ eq ⎤ ⎦ h ∫−∞ ⎣ 2 (25) ρ I( 2) = ⎜ − ⎟ h ⎛ i⎞ ⎝ ⎠ ∫ 2 ⎡VI ( t2 ) . ⎡ ⎦⎦⎦ .. ⎡ ⎣VI ( t1 ) . the polarization is directly proportional to the response function. when we use an impulsive driving potential to induce a free induction decay. Earlier we stated that we could write the second-order nonlinear response arise from two time-ordered interactions with external potentials in the form . P ( ) and higher are the nonlinear terms. ρ I( n ) is the nth order expansion of the density matrix ( ) ( ) ( ) (24) ρ ( 0) = ρeq ρ I(1) = − i t dt1 ⎡VI ( t1 ) . ρeq ⎦ ⎤⎤ dt2 ∫ dt1 ⎣ ⎦ −∞ −∞ t t ρI (n) ⎛ i⎞ =⎜− ⎟ ⎝ h⎠ n ∫ tn t2 ⎤⎤ dtn ∫ dtn −1 K ∫ dt1 ⎡VI ( tn ) .e. which can be Fourier transformed to obtain the absorption lineshape. as we would expect. ⎣ ⎡VI ( t1 ) . So. E ( t − τ ) = E0δ ( t − τ ) .

i. MIT Department of Chemistry. These correlation functions differ by whether sequential operators act on the bra or ket side of ρ when enforcing the timeordering. Comparison of eqs. and enforced the time-ordering t1 ≤ t2 . the nonlinear polarization is directly proportional to the response function. μ I (τ n −1 + τ n + Lτ 1 ) ⎤ ⎦ { } (30) We see that in general the nonlinear response functions are sums of correlation functions. for impulsive interactions. .Kτ n ) = ⎜ ⎟ θ (τ 1 ) θ (τ 2 )Kθ (τ n ) ⎝h⎠ (n) n ⎡ ⎡K ⎡ ⎤ μ I ( 0 ) ⎤ ρ eq × Tr ⎣ ⎦ . (27) and (28) allows us to state that the second order nonlinear response function is ⎛i⎞ ⎡⎡ ⎤ ρ eq R (τ 1 . μ I (τ 1 ) ⎤ ⎦ . μ I (τ 1 ) ⎤ ⎦ . Since the bra and ket sides represent conjugate wavefunctions. these correlation functions will contain coherences with differing phase relationships during subsequent timeintervals. 2 ⎧ ⎫ t2 ⎪ ⎪ ⎛ i⎞ t ⎤ ⎡ ⎤ . μ I ( t2 ) ⎤ ⎦ . Similar exercises to the linear and second order response can be used to show that the nonlinear response function to arbitrary order R ( ) is n ⎛i⎞ R (τ 1 .τ 2 ) = ⎜ ⎟ θ (τ 1 ) θ (τ 2 ) Tr ⎣ ⎣ μ I (τ 1 + τ 2 ) . . μ I ( t1 ) ⎤ ⎣⎡ ⎦ ρ eq −∞ t2 ∞ { } } (28) 2 0 ⎡⎡ ⎤ ρ eq dτ 2 ∫ dτ 1 E2 ( t − τ 2 ) E1 ( t − τ 2 − τ 1 ) Tr ⎣ ⎣ μ I (τ 1 + τ 2 ) . ρ P ( t ) = Tr ⎨ μ I ( t ) ⎜ − ⎟ ∫ dt2 ∫ dt1 ⎡ V t V t ( ) ( ) ⎬ 2 1 I I eq ⎣ ⎦ ⎣ ⎦ −∞ −∞ ⎝ h⎠ ⎪ ⎪ ⎩ ⎭ ( ) ( 2) ⎛i⎞ =⎜ ⎟ ⎝h⎠ ⎛i⎞ =⎜ ⎟ ⎝h⎠ 2 ∫ ∫ t −∞ ∞ dt2 ∫ dt1 E2 ( t2 ) E1 ( t1 ) Tr ⎡ ⎣ μ I ( t ) . Substituting as we did in the linear case. 4/1/2009 ∞ ∞ p. μI ( 0 )⎦ ⎝h⎠ ( 2) 2 { } (29) Again. delta function light pulses. μI ( 0)⎦ 0 { In the last line we switched variables to the time-intervals t1 = t − τ 1 − τ 2 and t2 = t − τ 2 .e.Andrei Tokmakoff.K ⎦ ⎣ ⎣ μ I (τ n + τ n −1 + K + τ 1 ) . and the nth order response has 2n correlation functions contributing. 11-10 P( 2) ( t ) = ∫0 dτ 2 ∫0 dτ 1 R( 2) (τ 2 .τ 2 .τ 1 ) E1 ( t − τ 2 − τ 1 ) E2 ( t − τ 2 ) (27) We can compare this result to what we obtain from P ( 2) ( t ) = Tr μ I ( t ) ρ I( 2) ( t ) .

Different correlation functions are distinguished by the order that they act on bra or ket. † = Tr ( μ U 0 (τ 2 ) μ U 0 (τ 1 ) μ ρ eqU 0 (τ 1 )U 0† (τ 2 ) U 0† (τ 1 ) U 0† (τ 2 ) U 0 (τ 2 ) ) (1) (2) (3) (4) (5) KET/KET interaction At each point of interaction with the external potential.Andrei Tokmakoff. (29). Q1 = Tr μ G 2 1 eq ( ) (31) Term 2: Q2 = Tr ( μ I ( 0 ) μ I (τ 1 + τ 2 ) μ I (τ 1 ) ρeq ) = Tr ( μ I (τ 1 + τ 2 ) μ I (τ 1 ) ρ eq μ I ( 0 ) ) BRA/KET interaction For the remaining terms we note that the bra side interaction is the complex conjugate of ket side. let’s look at R ( ) and enforce the time-ordering: 2 Term 1 in eq. (29): Q1 = Tr ( μ I (τ 1 + τ 2 ) μ I (τ 1 ) μ I ( 0 ) ρ eq ) † = Tr U 0 (τ 1 + τ 2 ) μ U 0 (τ 1 + τ 2 )U 0† (τ 1 ) μ U 0 (τ 1 ) μ ρeq ( ) dipole acts on ket of ρeq evolve under H0 during τ1. the dipole operator acted on ket side of ρ . We only count the interactions with the incident fields. and the convention is that the final operator that we use prior to the trace acts on the ket side. So the term Q1 is a ket/ket interaction. An alternate way of expressing this correlation function is in terms of the time-propagator ˆ ( t ) ρ = U a b U † . Remembering for the density matrix. MIT Department of Chemistry. we can identify only two independent terms: . this allows Q1 to be written as ˆ (τ ) μ G ˆ (τ ) μ ρ . 11-11 To see more specifically what a specific term in these nested commutators refers to. 4/1/2009 p. Multiply by μ and take trace. Evolve during τ2. dipole acts on ket. a superoperator defined through: G ab 0 0 the time-ordering. so of the four terms in eq.

φ (τ ) = ωabτ . Any subsequent field that acts on this state must account for time-dependent overlap of these wavepackets with other target states. R3 = bra / bra / ket . μ I (τ 1 + τ 2 ) ⎤ ⎦ . R2 = bra / ket / bra . ⎛i⎞ R (τ 1 .τ 2 .τ 2 .τ 3 ) = ⎜ ⎟ θ (τ 3 ) θ (τ 2 ) θ (τ 1 ) ∑ ⎡ ⎣ Rα (τ 3 .τ 3 ) = ⎜ ⎟ θ (τ 3 ) θ (τ 2 ) θ (τ 1 ) Tr ⎡ ⎣ μ I (τ 1 + τ 2 + τ 3 ) . μI ( 0 )⎤ ⎦ ρeq ⎝h⎠ ( 3) 3 { } (35) 4 ⎛i⎞ * R (τ 1 .τ 2 . In the .τ 2 . then one must start to carefully watch the relative phase that coherences acquire during different consecutive time-periods. the third-order nonlinear response described the most widely used class of nonlinear spectroscopies. 11-12 Q1 ⇒ ket / ket Q1* ⇒ bra / bra Q2 ⇒ ket / bra * Q2 = bra / ket . This is a general observation. consider wavepacket evolution: light interaction can impart positive or negative momentum ( ± kin ) to the evolution of the wavepacket. and R4 ⇒ ket / bra / bra . 4/1/2009 p. you really only need to calculate 2n−1 correlation functions. MIT Department of Chemistry. μ I (τ 1 ) ⎤ ⎦ .τ 2 ) − Qα (τ 1 . Third-Order Response Since R(2) orientationally averages to zero for isotropic systems.Andrei Tokmakoff. So for R(2) we write 2 ⎛i⎞ * R ( 2) = ⎜ ⎟ θ (τ 1 ) θ (τ 2 ) ∑ ⎡ Qα (τ 1 .τ 1 ) − Rα (τ 3 . The sum describes the evolution including possible interference effects between different interaction pathways.τ 2 )⎤ ⎣ ⎦ ⎝h⎠ α =1 2 (32) where Q1 = Tr ⎡ ⎣ μ I (τ 1 + τ 2 ) μ I (τ 1 ) μ I ( 0 ) ρ eq ⎤ ⎦ Q2 = Tr ⎡ ⎣ μ I (τ 1 ) μ I (τ 1 + τ 2 ) μ I ( 0 ) ρeq ⎤ ⎦ . The different terms in the nonlinear response function account for all of the permutations of interactions and the phase acquired by these coherences involved. which influences the direction of propagation and the phase of motion relative to other states. To illustrate.τ 1 ) ⎤ ⎦ h ⎝ ⎠ α =1 ( 3) 3 (36) Here the convention for the time-ordered interactions with the density matrix is R1 = ket / ket / ket . For R ( n ) . and more than one time period during which coherences can evolve. (33) (34) So what is the difference in these correlation functions? Once there is more than one excitation field.

Ei ( t − t0 ) = Ei δ ( t − t0 ) .b . c . MIT Department of Chemistry. μ I (τ n−1 + τ n + Lτ 1 ) ⎤ ⎦ Here the interactions of the light and matter are expressed in terms of a sequence the of consecutive For timeintervals τ 1 Kτ n prior to observing system.Kτ n ) E1 ( t − τ n − L − τ 1 )L En ( t − τ n ) n ∞ ∞ (38) ⎛i⎞ R ( n ) (τ 1 .Andrei Tokmakoff. The radiated nonlinear signal field is proportional to the nonlinear polarization: P( n) ( t ) = ∫0 dτ n L ∫0 dτ1 R( n) (τ 1.Kτ n ) = ⎜ ⎟ θ (τ 1 ) θ (τ 2 )Kθ (τ n ) ⎝h⎠ ⎡ ⎡K ⎡ ⎤ μ I ( 0 ) ⎤ ρeq × Tr ⎣ ⎦ .b .b .d): R1 = a . delta-function interactions. (40) .b. t ) E1 L En .Kτ n−1.c.τ 2 . d Summary: General Expressions for nth Order Nonlinearity For an nth-order nonlinear signal.b . Counting the radiated signal there are n+1 fields involved (n+1 light-matter interactions).τ 2 . d R4 = a . 4/1/2009 p. d R3 = a . there are n interactions with the incident electric field or fields that give rise to the radiated signal.K ⎦ ⎣ ⎣ μ I (τ n + τ n−1 + K + τ 1 ) .τ 2 . 11-13 eigenstate representation. c . c . c . so that nth order spectroscopy is also at times referred to as (n+1)wave mixing. d ∑ ∑ ∑ ∑ pa μad (τ 1 + τ 2 + τ 3 ) μdc (τ 1 + τ 2 ) μcb (τ 1 ) μba ( 0 ) pa μad ( 0 ) μdc (τ 1 + τ 2 ) μcb (τ 1 + τ 2 + τ 3 ) μba (τ 1 ) (37) pa μad ( 0 ) μdc (τ 1 ) μcb (τ 1 + τ 2 + τ 3 ) μba (τ 1 + τ 2 ) pa μad (τ 1 ) μdc (τ 1 + τ 2 ) μcb (τ 1 + τ 2 + τ 3 ) μba ( 0 ) R2 = a . the polarization and response function are directly proportional P( n) { } (39) ( t ) = R( n) (τ 1. the individual correlation functions can be explicitly written in terms of a sum over all possible intermediate states (a.

2. Principles of Nonlinear Optical Spectroscopy. (Oxford University Press.θ .Andrei Tokmakoff. S. 1995). Classical Electrodynamics. See Jackson. the index of refraction n (ω ) . 11-14 1. 4πε 0 r (written in spherical coordinates). New York. 3. The radiation pattern in the far field for the electric field emitted by a dipole aligned along the z axis is p k 2 sin θ E ( r . 4/1/2009 p. t ) = − 0 sin ( k ⋅ r − ωt ) . Mukamel. φ . and the absorption coefficient κ (ω ) : ε (ω ) = 1 + 4πχ (ω ) % (ω ) = n (ω ) + iκ (ω ) ε (ω ) = n . MIT Department of Chemistry. Remember the following relationships of the susceptibility with the complex dielectric constant ε (ω ) .

given the specifics of the quantum system we are studying and the details of the incident radiation. but also differ macroscopically in terms of the frequency and wavevector of the emitted radiation. is represented by a final wavy line connecting dipole coupled states. DIAGRAMMATIC PERTURBATION THEORY In practice. It is also a bookkeeping method that allows us to keep track of the contributions of the incident fields to the frequency and wavevector of the nonlinear polarization. The trace taken at the end following the action of the final dipole operator. The different terms in the response function can lead to an array of different nonlinear signals that vary not only microscopically by the time-evolution of the molecular system. DPT allows us to interpret the microscopic origin of a signal with a particular frequency and wavevector of detection.e. These multiply nested terms are difficult to understand when faced the numerous light-matter interactions.3. the signal emission. showing repeated interaction of ρ with the fields followed by time-propagation under H 0 . The diagrams show an explicit series of absorption or stimulated emission events induced by the incident fields which appear as action of the dipole operator on the bra or ket side of the density matrix. . It uses a series of simple diagrams to represent the evolution of the density matrix for H0.Andrei Tokmakoff. Feynman and ladder diagrams. the nonlinear response functions as written above provide little insight into what the molecular origin of particular nonlinear signals is. the first step in drawing a diagram is to identify the states of H0 that will be interrogated by the light-fields. which can be used to understand the microscopic information content of particular experiments. Diagrammatic perturbation theory (DPT) is a simplified way of keeping track of the contributions to a particular nonlinear signal given a particular set of states in H0 that are probed in an experiment. From a practical sense. They also symbolize the coherence or population state in which the density matrix evolves during a given time interval. each of which has certain advantages and disadvantages. For both types of diagrams. There are two types of diagrams we will discuss. 11-15 11. which can take on huge range of permutations when performing experiments on a system with multiple quantum states. It provides a shorthand form of the correlation functions contributing to a particular nonlinear signal. 4/30/2009 p. i. MIT Department of Chemistry.

. 2. Multiple states arranged vertically by energy. 4. 4. Action on the left line is action on the ket. and for noting absorption and emission events. MIT Department of Chemistry. Emission is an outward pointing arrow. Lines intersecting diagram represent field interaction. Arrows connecting levels indicate resonant interactions. Ladder Diagrams1 Ladder diagrams are helpful for describing experiments on multistate systems and/or with multiple frequencies. however. whereas the right line is bra. Time-evolution is upward. 1. Absorption is an upward arrow and emission is downward. Time propagates to right. Free propagation under H0 between interactions. 4/30/2009 p. 3. 3. 2. whereas a dotted line is action on the bra.Andrei Tokmakoff. System evolves freely under H0 between interactions. Double line represents ket and bra side of ρ . 11-16 Feynman Diagrams Feynman diagrams are the easiest way of tracking the state of coherences in different time periods. but the state of the density matrix is not always obvious. and density matrix element for that period is often explicitly written. 1. They naturally lend themselves to a description of interactions in terms of the eigenstates of H0. Absorption is designated through an inward pointing arrow. it is difficult to immediately see the state of the system during a given time interval. A solid line is used to indicate action on the ket.

Remember that action on the bra is the complex conjugate of ket and absorption is complex conjugate of stimulated emission. convention: ket side) b a a b ˆ an μba ⋅ε . 11-17 For each light-matter interactions represented in a diagram. which are used to describe the frequency and wavevector of the radiated signal. to ( n) contribution to R k sig & ωsig (μ ⎡ ⎤ ba ⋅ En ) exp ⎣ik n ⋅ r − iωn t ⎦ b b a En Stimulated Emission * n a μba ⋅ ˆ εn +k n + ωn ( μba ⋅ E ) exp ⎡ ⎣ −ikn ⋅ r + iωn t ⎤ ⎦ * En b a a b μba ⋅ ˆ εn −k n − ωn BRA SIDE Absorption (μ * ba * ⋅ En ) exp ⎡ ⎣ −ikn ⋅ r + iωn t ⎤ ⎦ b a * En b a ˆ μ* ba ⋅ ε n −k n − ωn Stimulated Emission (μ * ba ⋅ En ) exp ⎡ ⎣ikn ⋅ r − iωnt ⎤ ⎦ b a En a b ˆ μ* ba ⋅ ε n +k n + ωn SIGNAL EMISSION: (Final trace. MIT Department of Chemistry. The action of the final dipole operator must return you to a diagonal element to contribute to the signal. Interaction KET SIDE Absorption Diagrammatic Representation contrib. A table summarizing these interactions contributing to a diagram is below. either through absorption or stimulated emission.2 Each interaction adds input electric field factors to the polarization. Each light-matter interaction acts on one side of ρ .Andrei Tokmakoff. there is an understanding of how this action contributes to the response function and the final nonlinear polarization state. Each interaction adds a dipole matrix element μij that describes the interaction amplitude and any orientational effects. 4/30/2009 p.

11-18 Once you have written down the relevant diagrams. . and for the final signal emission. Coherences and populations are propagated by assigning the damping constant Γ ab to propagation of the ρ ab element: ˆ (τ ) ρ = exp [ −iω τ − Γ τ ] ρ . There is only one independent correlation function in the linear response function. G ab ab ab ab (1) * = Gba . 5) The radiated signal will have frequency ωsig = ∑ ωi and wave vector k sig = ∑ ki i i Example: Linear Response for a Two-Level System Let’s consider the diagrammatic approach to the linear absorption problem. it is valuable to include the effects of relaxation of the system eigenstates in the time-evolution using a simple phenomenological approach. 2) Read off products of transition dipole moments for interactions with the incident fields. being careful to identify all permutations of interactions of your system states with the fields relevant to your signal. We can then recognize Γii =1 T1 as the population Note Γ ab = Γba and Gab relaxation rate for state i and Γij = 1 T2 the dephasing rate for the coherence ρij . It is convenient to write the correlation function as a product of several factors for each event during the series of interactions: 1) Start with a factor pn signifying the probability of occupying the initial state. As a starting point for understanding an experiment. some correlation functions are subtracted from others. 3) Multiply by terms that describe the propagation under H 0 between interactions. the correlation functions contributing to the material response and the frequency and wavevector of the signal field can be readily obtained. 4) Multiply by a factor of ( −1) where n is the number of bra side interactions. 4/30/2009 p. typically a Boltzmann factor.Andrei Tokmakoff. using a two-level system with a lower level a and upper level b. This factor n accounts for the fact that in evaluating the nested commutator. MIT Department of Chemistry.

MIT Department of Chemistry. 11-19 C ( t ) = Tr ⎡ ⎣ μ ( t ) μ ( 0 ) ρ eq ⎤ ⎦ ˆ ⎤ = Tr ⎡ ⎣ μ G ( t ) μ ρeq ⎦ (2) This does not need to be known before starting. Working from bottom up: (1) (2) (3) (4) (5) − iωba t −Γba t ⎤ [ μab ] C ( t ) = pa [ μba ] ⎡ ⎣e ⎦ = pa μba e−iωbat −Γbat 2 More sophisticated ways of treating the time-evolution under H0 in step (3) could take the form of some of our earlier treatments of the absorption lineshape: . The product of incident fields indicates: E1 e−iω1t +ik1⋅r ⇒ P (t ) e − iωsig t +iksig ⋅r (3) (4) so that ωsig = ω1 ksig = k . we can begin by describing the signal characteristics in terms of the induced polarization. (1) Start in ρaa (add factor of pa when reading). (3) Propagate under H0: −iω τ −Γbaτ Gab (τ ) = e ba . but is useful to consider. Next we can write down the correlation function for this term. and which we take to be a ket side interaction. With this diagram. As expected the signal will radiate with the same frequency and in the same direction as the incoming beam. 4/30/2009 p.Andrei Tokmakoff. We can now draw two diagrams: (4) Act on ket with μ and take trace. which must be absorption. (2) Act on ket with μ(0) to create ρba. Linear response only allows for one input field interaction. The system will be taken to start in the ground state ρaa. since it should be recovered in the end.

and initial population in states a and b: However. which makes it useful for studies of interfaces and chiral systems. the permutations of unique diagrams are as follows: . If we only have population in the ground state at equilibrium and if there are only resonant interactions allowed. and they are generally taken to be understood through one diagram. The signal generation requires a lack of inversion symmetry. 4/30/2009 p. Diagram ii is just the complex conjugate of eq. 11-20 ˆ (τ ) ρ ~ ρ exp [ −iω τ ] F (τ ) G ab ab ab = ρ ab exp ⎡ ⎣ −iωabτ − g ( t ) ⎤ ⎦ (6) Note that one could draw four possible permutations of the linear diagram when considering bra and ket side interactions. MIT Department of Chemistry. h (8) Example: Second-Order Response for a Three-Level System The second-order response is the simplest nonlinear case. let’s show how one would diagrammatically evaluate the second order response for a very specific system pictured at right.Andrei Tokmakoff. there is no new dynamical content in these extra diagrams. (5) so adding this signal contribution gives: C ( t ) − C *( t ) = 2i pa μba sin(ωba t ) e−Γbat . However. 2 (7) Accounting for the thermally excited population initially in b leads to the expected two-level system response function that depends on the population difference R (t ) = 2 2 ( pa − pb ) μba sin(ωbat ) e−Γbat . but in molecular spectroscopy is less commonly used than third-order measurements.

combination Q2 − Q2 . 11-21 From the frequency conservation conditions. a Q2 shown at right has a downward transition –emission– as the final interaction. The complex conjugate of iv. Reading in a time-ordered manner. we can write the correlation function corresponding to this diagram as C2 = Tr ⎡ ⎣ μ (τ ) ρeq μ ( 0 ) ⎤ ⎦ 1 ˆ (τ ) μ G ˆ (τ ) ρ μ ∗ . = ( −1) μbc G cb ca ab aa ba 2 1 (9) = − pa μab μbc μca e − iωabτ1 −Γabτ1 e − iωcbτ 2 −Γcbτ 2 Note that a literal interpretation of the final trace in diagram iv would imply an absorption event – an upward transition from b to c. it should be clear that process i is a sum-frequency signal for the incident fields. What does this have to do with radiating a signal? On the one hand it is important to remember that a diagram is just mathematical shorthand. 4/30/2009 p. The ∗ ultimately describes the observable. MIT Department of Chemistry.Andrei Tokmakoff. and that one can’t distinguish absorption and emission in the final action of the dipole operator prior to taking a trace. with it in the response function. The other thing to remember is that such a diagram always has a complex conjugate associated ∗ ket/bra term. To better interpret what these diagrams refer to let’s look at iii. whereas diagrams ii-iv refer to difference frequency schemes.

4/30/2009 p. Note that the most efficient wavevector matching here would be when fields 1 and 2 are collinear. the signal would be phase-matched such that the signal is radiated closest to beam 2.Andrei Tokmakoff. the length of the vectors will be scaled by the resonance frequencies. . When the two incident fields are crossed as a slight angle. 11-22 Now. consider the wavevector matching conditions for the second order signal iii. Remembering that the magnitude of the wavevector is k = ω c = 2π λ . MIT Department of Chemistry.

Let’s write out the diagrams and correlation functions for a two-level system starting in ρ aa . 11-23 Third-Order Nonlinear Spectroscopy Now let’s look at examples of diagrammatic perturbation theory applied to third-order nonlinear spectroscopy. These 3 experiments are described by some or all of the eight correlation functions contributing to R ( ) : ⎛i⎞ R ( 3) = ⎜ ⎟ ⎝h⎠ 3 ⎡ ∑ ⎣ Rα − Rα ⎤ ⎦ α * =1 4 (10) The diagrams and corresponding response first requires that we specify the system eigenstates. where the dipole operator couples b and a . Third-order nonlinearities describe the majority of coherent nonlinear experiments that are used including pump-probe experiments. 4/30/2009 p. which allows us discuss a number of examples of third-order spectroscopy is a two-level system. photon echoes. R1 ket/ket/ket R2 bra/ket/bra R3 bra/bra/ket R4 ket/bra/bra b a τ3 b a b b a b b a a a b a b b E3 a a E2 b a a a τ2 τ1 a b a a b a a a a a E1 b a +ω1 − ω2 + ω3 ksig = + k1 − k2 + k3 −ω1 + ω2 + ω3 −k1 + k2 + k3 −ω1 + ω2 + ω3 −k1 + k2 + k3 +ω1 − ω2 + ω3 + k1 − k2 + k3 .Andrei Tokmakoff. coherent anti-Stokes Raman spectroscopy (CARS). and degenerate four wave mixing (4WM). The simplest case. MIT Department of Chemistry. transient gratings.

(13) Now. the phase acquired continuously in τ1 and τ3. let’s compare this to the response obtained from R4 . For that reason the term is called “rephasing. 11-24 As an example. these are obtained from the product of the incident field contributions E1 E2 E3 = E1* e + iω1t −ik1 ⋅r ⇒ E1* E2 E3 e ( )( E e 2 − iω2t + ik2 ⋅r )(E e 3 + iω3t −ik3 ⋅r3 ) −ωsig t + iksig ⋅r (12) ∴ ωsig 2 = − ω1 + ω2 + ω3 ksig 2 = − k1 + k2 + k3 . whereas the R4 term evolves in the same coherence state during both periods: Coherences in τ 1 and τ 3 R4 R2 b a → b a a b → b a Phase acquired in τ 1 and τ 3 e e − iωba (τ1 +τ 3 ) − iωba (τ1 −τ 3 ) The R2 term has the property of time-reversal: the phase acquired during τ1 is reversed in τ3. The R2 term evolves in conjugate coherences during the τ1 and τ3 periods. These we obtain R4 = pa μab exp ⎡ ⎣ −iωba (τ 3 + τ 1 ) − Γba (τ 1 + τ 3 ) − Γbb (τ 2 ) ⎤ ⎦ 4 (14) (15) ωsig 4 = + ω1 − ω2 + ω3 ksig 4 = + k1 − k2 + k3 Note that R2 and R4 terms are identical. and this term is called “non- . For R4 . This term is important for understanding photon echo experiments and contributes to pump-probe and degenerate four-wave mixing experiments. let’s write out the correlation function for R2 obtained from the diagram above.Andrei Tokmakoff. MIT Department of Chemistry. except for the phase acquired during the initial period: exp [iφ ] = exp [ ±iωbaτ 1 ] . − iωabτ1 −Γ abτ1 * ⎤ ( μba ) e −iωbbτ 2 −Γbbτ 2 R2 = ( −1) pa ( μba )⎡ ⎣e ⎦ 2 4 ( )(μ * ab − iω τ −Γ )⎡ ⎣e ba 3 baτ 3 ⎤ ( μab ) ⎦ = pa μab exp ⎡ ⎣ −iωba (τ 3 − τ 1 ) − Γba (τ 1 + τ 3 ) − Γbb (τ 2 ) ⎤ ⎦ (11) The diagrams show how the input field contributions dictate the signal field frequency and wave( 3) vector. 4/30/2009 p. Recognizing the dependence of Esig ~ P ( 3) ~ R2 ( E1 E2 E3 ) .” Rephasing signals are selected in photon echo experiments and are used to distinguish line broadening mechanisms and study spectral diffusion.

The nonrephasing pair R1 and R4 both scatter in the + k1 − k2 + k3 direction and are labeled as SII.Andrei Tokmakoff. Our findings for the four independent correlation functions are summarized below. Similarly. they are also referred to as SI.” Analysis of R1 and R3 reveals that these terms are non-rephasing and rephasing. but differ by the τ2 population. ωsig SI SII rephasing non-rephasing R2 R3 R1 R4 −ω1 + ω2 + ω3 −ω1 + ω2 + ω3 +ω1 − ω2 + ω3 +ω1 − ω2 + ω3 k sig −k1 + k2 + k3 −k1 + k2 + k3 + k1 − k2 + k3 + k1 − k2 + k3 τ2 population excited state ground state ground state excited state . phase acquired φ e 0 − iωbaτ1 e − iωbaτ 3 e + iωbaτ 3 non-rephasing τ1 τ3 rephasing P (t ) t1 t2 t3 t For the present case of a third-order spectroscopy applied to a two-level system. 4/30/2009 p. and would therefore both contribute equally to a given detection geometry. MIT Department of Chemistry. respectively. The two terms differ in which population state they propagate during the τ2 variable. we observe that the two rephasing functions R2 and R3 have the same emission frequency and wavevector. Since the rephasing pair R2 and R3 both contribute equally to a signal scattered in the −k1 + k2 + k3 direction. these terms could be separated by frequency or wavevector (see appendix). 11-25 rephasing. the nonrephasing functions R1 and R4 each have the same emission frequency and wavevector. For transitions between more than two system states.

χ (3) (ω1 . we have to permute over all possible time orderings. ω3 ) = ⎜ ⎟ 6⎝h⎠ 3 abcd ∑ p α∑ ⎡ ⎣ χα − χα ⎤ ⎦ * a =1 4 (19) Here a is the initial state and the sum is over all possible intermediate states. In the frequency domain. Multiply by resonance denominator terms that describe the propagation under H 0 .4 Given a set of diagrams. ω2 . MIT Department of Chemistry. The radiated signal will have frequency ωsig = ∑ ωi and wavevector k sig = ∑ ki . Also. (17) to response functions that use phenomenological time-propagators of the form eq. we obtain ˆ (τ ) ρ ⇒ G m ab 1 ( Ωm − ωba ) − iΓba . 4/30/2009 p. n n 1 1 ∞ ∞ 0 0 Here the Fourier transform conjugate variables Ωm to the time-interval τ m are the sum over all frequencies for the incident field interactions up to the period for which you are evolving: Ωm = ∑ ωi i =1 m (18) For instance. if we apply eq. ω3 ) E1 E2 E3 where (16) (17) χ ( n ) = ∫ dτ n eiΩ τ L ∫ dτ 1 eiΩ τ R ( n ) (τ 1 . ω2 . Most generally.τ 2 . multiply by a factor of ( −1) for n bra side interactions. i i . n As for the time domain. ω3 ) : P (3) (ωsig ) = χ (3) (ωsig .Andrei Tokmakoff. (18). the conjugate variable for the third time-interval of a + k1 − k2 + k3 experiment is the sum over the three preceding incident frequencies Ω3 = ω1 − ω2 + ω3 . ω2 . (20) 3) 4) Ωm is defined in eq. ω1 . as a result of the 3!=6 permutations of the time-ordering of the input fields. χ(3) is a sum over many correlation functions and includes a sum over states: χ ( 3) 1⎛i ⎞ (ω1 . to describe frequency domain experiments. In general. the eight terms in R (3) lead to 48 terms for χ ( 3) . (1).Kτ n ) . 11-26 Frequency Domain Representation3 A Fourier-Laplace transform of P ( 3) ( t ) with respect to the time intervals allows us to obtain an expression for the third order nonlinear susceptibility. we can write the nonlinear susceptibility directly as follows: 1) 2) Read off products of light-matter interaction factors.

11-27 As an example.Andrei Tokmakoff. consider the term for R2 applied to a two-level system that we wrote in the time domain in eq. The full frequency domain response is a sum over multiple terms like these. (11) χ 2 = μba = μba 4 ( −1) ( −1) 1 ⋅ ⋅ ωab − ( −ω1 ) − iΓ ab ωbb − (ω2 − ω1 ) − iΓbb ωba − (ω3 + ω2 − ω1 ) − iΓba ω1 − ωba − iΓba − (ω2 − ω1 ) − iΓbb − (ω3 + ω2 − ω1 − ωba ) − iΓba 1 ⋅ 1 ⋅ 1 (21) 4 The terms are written from a diagram with each interaction and propagation adding a resonant denominator term (here reading left to right). MIT Department of Chemistry. . 4/30/2009 p.

These are examples of correlation functions within R(3) for a four-level system representative of vibronic transitions accompanying an electronic excitation. 4/30/2009 p. The R1 term is more properly written ( μba ⋅ ε Note that the product of transition dipole matrix elements obtained from the sequence of interactions can always be re-written in the cyclically invariant form μab μbc μcd μda .Andrei Tokmakoff. as relevant to resonance Raman spectroscopy. Note that these diagrams present only one example of multiple permutations that must be considered given a particular time-sequence of incident fields that may have variable frequency. 11-28 Appendix: Third-order diagrams for a four-level system The third order response function can describe interaction with up to four eigenstates of the system Hamiltonian. . MIT Department of Chemistry. orientational effects. R1 ket / ket / ket R2 bra/ket/bra R3 bra/bra/ket R4 ket/bra/bra d a τ3 d c d b d c a c b a b b E3 c a E2 b a a a τ2 τ1 a b a b a a b a a a a a E1 d d d d b c a b c a b c a b c a * * μba μda μcd μcb * * * * μba μcb μdc μad μba μda μcb μcd μba μcb μda μcd ωsig = +ω1 − ω2 + ω3 −ω1 + ω2 + ω3 −ω1 + ω2 + ω3 +ω1 − ω2 + ω3 = ωba + ωcb + ωdc = ωda −ωba + ωda − ωcb = ωdc −ωba − ωcb + ωda = ωdc ωba − ωda − ωcd = ωbc The signal frequency comes from summing all incident resonance frequencies accounting for the sign of the excitation. The products of transition matrix elements are written in a time-ordered fashion without the projection onto the incident field polarization needed to properly account for ˆ1 )( μcb ⋅ ε ˆ2 )( μdc ⋅ ε ˆ3 )( μad ⋅ ε ˆan ) . This is one further manifestation of closed loops formed by the sequence of interactions.

when the incident fields are resonant with the fundamental vibrational transition. MIT Department of Chemistry. 11-29 Appendix: Third-order diagrams for a vibration The third-order nonlinear response functions for infrared vibrational spectroscopy are often applied to a weakly anharmonic vibration. we generally consider diagrams involving the system eigenstates v = 0. and which include v=0-1 and v=1-2 resonances. For high frequency vibrations in which only the v = 0 state is initially populated. Then.Andrei Tokmakoff. there are three distinct signal contributions: Signal k sig Diagrams and Transition Dipole Scaling R/NR SI −k1 + k2 + k3 rephasing μ10 4 μ10 4 μ10 μ21 2 2 SII + k1 − k2 + k3 non-rephasing μ10 4 μ10 4 μ10 μ21 2 2 SIII + k1 + k2 − k3 non-rephasing μ10 μ21 2 2 μ10 μ21 2 2 . 1 and 2. 4/30/2009 p.

2. R. T. Pure Appl. 4. 2 μ10 ≠ μ21 . MIT Department of Chemistry. we can see that the signal contributions should destructively interfere and vanish. Infrared and Raman Spectr. which can be seen by counting the number of bra side interactions. IEEE J. The are obtained by projecting the matrix element prior to the final trace onto the ˆan . Some deviation from harmonic behavior is required to observe a signal. This projected onto the incident electric field polarization ε leads to a nonlinear polarization that can have x. A complete expression for the third order susceptibility χ ( 3) -perturbative and diagramatic approaches. 151 (1978). 12. 59 (1993). Bloembergen. the transition dipole moment must be ˆ leading to the terms in the table. N. & Lynch. Lineshapes in coherent resonant Raman scattering. resonance Raman.Andrei Tokmakoff. 1. See also. such as vibrational anharmonicity ω10 ≠ ω21. “A unified view of Raman. Phys. To properly account for all orientational factors. y. 179 (1985). and z polarization components in the lab frame. D. Chem. and have emission frequencies of ω10 or ω21. Lee and A. These two types of contributions have opposite signs. Prior. 171. and fluorescence spectroscopy (and their analogues in two-photon absorption). or level-dependent . for harmonic oscillators. Therefore. which have ω10 = ω21 and 2 μ10 = μ21 . Response functions and susceptibilities for multiresonant nonlinear optical spectroscopy: Perturbative computer algebra solution including feeding. Phys. Indian J. H. 4/30/2009 p. Albrecht. 16. This is a manifestation of the finding that harmonic systems display no nonlinear response. Dick. 11-30 Note that for the SI and SII signals there are two types of contributions: two diagrams in which all interactions are with the v=0-1 transition (fundamental) and one diagram in which there are two interactions with v=0-1 and two with v=1-2 (the overtone). electrical anharmonicity damping Γ10 ≠ Γ21 or Γ00 ≠ Γ11 .. B. Y. 37 (1984). desired analyzer axis ε 3. Lotem. QE-20. Quantum Electron. C.” Adv.

4. with the emphasis on trying to connect the particular signals with their microscopic origin.b.Andrei Tokmakoff. 4/18/2009 p. Let’s start by discussing how one can distinguish a rephasing signal from a non-rephasing signal. We choose a box geometry. 11-31 11. Generally. Here is a strategy for describing a particular experiment: 1) Start with the wavevector and frequency of the signal field of interest. However you can use the properties of the incident and detected fields to help with selectivity. This section will discuss a number of these methods. Selecting signals by wavevector The question that arises is how to select particular contributions to the signal. 2) (a) Time-domain: Define a time-ordering along the incident wavevectors or (b) Frequency domain: Define the frequencies along the incident wavevectors. including the common pump-probe experiment. it will not be possible to uniquely select particular diagrams. THIRD-ORDER NONLINEAR SPECTROSCOPIES Third-order nonlinear spectroscopies are the most widely used class of nonlinear methods. as shown. Consider two degenerate third-order experiments (ω1 = ω2 = ω3 = ωsig) which are distinguished by the signal wave-vector for a particular time-ordering. and the frequency and wavevector of the signal. where the three incident fields (a. incident from three corners of the box. (Note that the color in these figures is not meant to represent the frequency of the incident fields –which are all the kc ka kb sample k sig k sig = + ka − kb + kc c b a top-down view k sig + ka − kb + kc . keeping only resonant terms (rotating wave approximation). frequency and time-ordering of the input fields.c) are crossed in the sample. MIT Department of Chemistry. The approach here is meant to be practical. This approach can be used for describing any experiment in terms of the wave-vector. use ladder diagrams to determine which correlation functions yield signals that pass through your filter/monochromator. 3) Sum up diagrams for correlation functions that will scatter into the wave-vector matched direction. In the frequency domain.

4/18/2009 p. The efficiency of radiating the signal field falls of with the ′ . MIT Department of Chemistry. and c = 3. we see that we can select non-rephasing signals R1 and R4 by setting the time ordering of pulses such that a = 1. b = 2. whereas the magnitude of the signal wavevector ksig frequency (energy conservation). only its direction varies. 11-32 same – but rather is just there to distinguish them for the picture). comparing the wavevector matching condition for this signal with those predicted by the third-order Feynman diagrams. .10). then the rephasing and non-rephasing signals will be radiated as shown below: NR E3 E2 E1 sample k sig + k1 −k2 + k3 R k sig = − k1 + k 2 + k3 NR k sig = + k1 − k2 + k3 R k sig + k3 R 3 2 NR 1 + k2 − k1 In this case the wave-vector matching for the rephasing signal is imperfect. (The colors in the figure do not represent frequency. but just serve to distinguish the beams). Vector addition of the contributing terms from the incident fields indicates that the signal ksig = + ka − kb + kc will be radiated in the direction of the last corner of the box when observed after the sample. 1. Now. Since the frequencies are the same. b = 1. The vector sum of the incident fields ksig dictates the direction of propagation of the radiated signal (momentum ′ is dictated by the radiated conservation). the length of the wavevector k = 2π n λ is equal for each field. The rephasing signals R2 and R3 are selected with the time-ordering a = 2.Andrei Tokmakoff. and c = 3. b = 2. If we set the time ordering to be a = 1. Alternatively. as Esig ( t ) ∝ P ( t ) sinc ( Δkl 2 ) where l is the path length wave-vector mismatch Δk = ksig − ksig (see eq. and c = 3. we can recognize that both signals can be observed by simultaneously detecting signals in two different directions.

and time-scales for energy gap fluctuations. Now look at the experiment in which two pulses are crossed to generate a signal in the direction k sig = 2k2 − k1 (5) This signal is a special case of the signal ( k3 + k2 − k1 ) where the second and third interactions are both derived from the same beam. we can set 2 2 g ( t ) = Γba t + 1 2Δ t . (1) For an inhomogeneous distribution. MIT Department of Chemistry. we could average the homogeneous response. 11-33 Photon Echo The photon echo experiment is most commonly used to distinguish static and dynamic linebroadening. that is a convolution of a homogeneous line shape with width Γ with a static inhomogeneous distribution of width Δ. ⎟ ⎠ (3) Equivalently. Remember that linear spectroscopy cannot distinguish the two: R (τ ) = μab e 2 −iωabτ − g (τ ) − c. the absorption spectrum is a broad Gaussian lineshape centered at the mean frequency ωba which just reflects the static distribution Δ rather than the dynamics in Γ. . Both non-rephasing diagrams contribute here. since a convolution in the frequency domain is a product in the time domain. with an inhomogeneous distribution R = ∫ d ωab G (ωab ) R (ωab ) (2) 2 which we take to be Gaussian ⎛ (ω − ω ba ba G (ωba ) = exp ⎜ − 2 ⎜ 2Δ ⎝ ) ⎞ ⎟. (4) So for the case that Δ > Γ .Andrei Tokmakoff. 4/18/2009 p. The rephasing character of R2 and R3 allows you to separate homogeneous and inhomogeneous broadening.c. τ 2 = 0 and R2 = R3. To demonstrate this let’s describe a photon echo experiment for an inhomogeneous lineshape. The nonlinear signal can be obtained by integrating the homogeneous response. but since both second and third interactions are coincident. g ( t ) = Γba t .

. This rephasing enhancement is called (τ . The broad distribution of frequencies rapidly dephases during τ1.Andrei Tokmakoff. 11-34 Two-pulse photon echo R( 3) (ωab ) = μab pa e 4 −iωab (τ1 −τ 3 ) −Γab (τ1 +τ 3 ) e (6) over the inhomogeneous distribution as in eq. and that an intensity level measurement doubles the decay rate of the polarization. 4/18/2009 p. the only source of relaxation of the polarization amplitude at τ1 = τ3 is Γ ab .τ 3 ) 2 (8) In the inhomogeneous limit ( Δ >> Γ ab ). MIT Department of Chemistry. For a given pulse separation τ (setting τ1=τ). e ( 1 3 ) ≈ δ (τ 1 − τ 3 ) . In practice. leading to a large constructive enhancement an echo. the signal is observed with a integrating intensity-level detector placed into the signal scattering direction. At this point inhomogeneity is removed and only the homogeneous dephasing is measured. The factor of four in the decay rate reflects the fact that damping of the initial coherence evolves over two periods τ1 + τ3 = 2τ. we calculated the integrated signal intensity radiated from the sample during τ3 as I sig (τ ) = Esig ∝ ∫ dτ 3 P ( 2 −∞ ∞ 3) of the polarization at τ1=τ3. i. This leads to R ( ) = μ ab pa e 3 4 − i ωab (τ1 −τ 3 ) e −Γ ab (τ1 +τ 3 ) 2 − τ −τ Δ 2 /2 e ( 1 3) (7) 2 − τ −τ Δ 2 /2 For Δ >> Γ ab .e. (2). (9) In this case. we find I sig (τ ) ∝ μ ab e 8 −4 Γ abτ . R (3) is sharply peaked at τ 1 = τ 3 . but is rephased (or refocused) during τ3.

If the beams are crossed at an angle 2θ ˆ sin θ ) ˆ cos θ + x ka = ka ( z ˆ sin θ ) ˆ cos θ − x kb = kb ( z with (11) (10) ka = kb = 2π n λ .Andrei Tokmakoff. but is uniquely suited for looking at optical excitations with well-defined spatial period. so you cannot distinguish which field interacts first. That is the signal depends on R1+R2+R3+R4. Consider the terms contributing to the polarization that arise from the first two interactions. 11-35 Transient Grating The transient grating is a third-order technique used for characterizing numerous relaxation processes. For two time-coincident pulses of the same frequency. The first two pulses are set time-coincident. Therefore. the signal will have contributions both from k sig = k1 − k2 + k3 and ksig = −k1 + k2 + k3 .c. the first two fields have an excitation profile in the sample ⎤ Ea Eb = Ea Eb exp ⎡ ⎣ −i (ωa − ωb ) t + i ( ka − kb ) ⋅ r ⎦ + c. (12) the excitation of the sample is a spatial varying interference pattern along the transverse direction . MIT Department of Chemistry. 4/18/2009 p.

β = sin θ = λ η This spatially varying field pattern is called a grating. we should think of excitation with this pulse pair leading to a periodic spatial variation of the complex index of refraction of the medium. MIT Department of Chemistry. Thus. (15) Absorption images this pattern into the sample. creating a spatial pattern of excited and ground state molecules. Nonresonant scattering processes (Raleigh and Brillouin scattering) can create a spatial modulation in the real index or refraction. (8). Most commonly. leading to scattering of a signal out of the sample at an angle −θ.Andrei Tokmakoff. 4/18/2009 p. 11-36 Ea Eb = Ea Eb exp ⎡ ⎣i β ⋅ x ⎤ ⎦ + c. (13) The grating wavevector is β = k1 − k2 4π n 2π . whereas subsequent relaxation can lead to heating a periodic temperature profile (a thermal grating). we measure the intensity of the scattered light. the transient grating signal will be sensitive to any processes which act to wash out the spatial modulation of the grating pattern: • Population relaxation leads to a decrease in the grating amplitude. I sig (τ ) ∝ exp [ −2Γbbτ ] (16) . where the wavevector matching conditions are equivalent to the constructive interference of scattered waves at the Bragg angle off a diffraction grating.c. as given in eq. and has a fringe spacing (14) η= λ 2n sin θ . A time-delayed probe beam can scatter off this grating. More generally. Absorption can create an excited state grating. observed as a decrease in diffraction efficiency. For ω1 = ω2 = ω3 = ωsig this the diffraction condition is incidence of k3 at an angle θ.

MIT Department of Chemistry. 4/18/2009 p. 11-37 • ˆ acts to wash out the fringe pattern. For a diffusion Thermal or mass diffusion along x constant D the decay of diffraction efficiency is 2 I sig (τ ) ∝ exp ⎡ ⎣ −2 β Dτ ⎤ ⎦ (17) • Rapid heating by the excitation pulses can launch counter propagating acoustic waves ˆ . which can modulate the diffracted beam at a frequency dictated by the period along x for which sound propagates over the fringe spacing in the sample. .Andrei Tokmakoff.

and the using the theoretical formalism of nonlinear spectroscopy often unnecessary to interpret the experiment. and is most commonly used to follow population relaxation. Two pulses separated by a delay τ are crossed in a sample: a pump pulse and a time-delayed probe pulse. Described as a third-order coherent nonlinear spectroscopy. so terms that contribute to scattering along the probe are k sig = ± k1 m k2 + k3 .Andrei Tokmakoff. 11-38 Pump-Probe The pump-probe or transient absorption experiment is perhaps the most widely used third-order nonlinear experiment. the time-ordering of pumpinteractions cannot be distinguished. i. the pump-probe can be thought of as the limit of the transient grating experiment in the limit of zero grating wavevector (θ and β → 0). the signal is radiated collinear to the transmitted probe field. all correlation functions R1 to R4 . The pump pulse E pu creates a non-equilibrium state. The detector observes the intensity of the transmitted probe and nonlinear signal I= 2 nc E′ pr + Esig . or wavepacket dynamics and quantum beats. chemical kinetics.e. so the wavevector matching condition is ksig = + k pu − k pu + k pr = k pr . 4π (18) . There are two interactions with the pump field and the third interaction is with the probe. and the time-dependent changes in the sample are characterized by the probe-pulse E pr through the pump-induced intensity change on the transmitted probe. In fact. 4/18/2009 p. ΔI. Similar to the transient grating. MIT Department of Chemistry. The principle is quite simple. It can be used to follow many types of time-dependent relaxation processes and chemical dynamics.

(19) If we work under conditions of a weak signal relative to the transmitted probe E ′ pr >> Esig . or the imaginary part of the index of refraction. In this case. the optical Kerr effect.c.τ 3 ) = P (3) (τ . such as nonresonant Raman. In addition to the fully resonant processes. These can be described through a variety of nonresonant interactions. (21) the measured pump-probe signal is proportional to the imaginary part of the polarization ⎤ ′ ( 3) ΔI (τ ) = 2ωsig l Im ⎡ ⎣ E pr P (τ ) ⎦ . or the second hyperpolarizability of the sample. The measured signal is typically the differential intensity on the probe field with and without the pump field present: ΔI (τ ) = 2 nc ′ E′ pr + Esig (τ ) − E pr 4π { 2 }.⎦ 4π ⎣ .τ 3 )e − iωsigτ 3 (3) +⎡ ⎣ P (τ . Esig (τ ) = i 2πωsig l nc P (3) (τ ) . (19) is dominated by the cross term ΔI (τ ) ≈ nc ∗ ⎡ E′ ⎤ pr Esig (τ ) + c.τ 3 ) ⎤ ⎦ sin(ωsigτ 3 ) (23) . we can describe the timedevelopment of the polarization and radiated signal field as P (3) (τ . considered earlier. the nonlinear changes in absorption of the transmitted probe field are related to the imaginary part of the susceptibility. nc ∗ ⎤ = Re ⎡ pr Esig (τ ) ⎦ ⎣ E′ 2π (20) So the pump-probe signal is directly proportional to the nonlinear response. Dichroic and Birefringent Response (22) which is also proportional to the correlation functions derived from the resonant diagrams we In analogy to what we observed earlier for linear spectroscopy.Andrei Tokmakoff. then the differential intensity in eq.τ 3 ) ⎤ = 2 Re ⎣ ⎦ cos(ωsigτ 3 ) + 2 Im ⎡ ⎣ P (τ . 4/18/2009 p.τ 3 ) ⎤ ⎦ e ∗ iωsigτ 3 (3) ⎡ P (3) (τ . Since the signal field is related to the nonlinear polarization through a π/2 phase shift. coherent Raleigh or Brillouin scattering. MIT Department of Chemistry. 11-39 E′ pr is the transmitted probe field corrected for linear propagation through the sample. it is also possible for the pump field to induce nonresonant changes to the polarization that modulate the real part of the index of refraction.

τ 3 ) sin(ωsigτ 3 ) + Edic (τ . ϕ ) ≈ nc ′ ⎡ E′ ⎤ pr (τ ) Ebir (τ )sin(ϕ ) + E pr (τ ) Edic (τ ) cos(ϕ ) ⎦ 2π ⎣ (27) (28) The birefringent and dichroic response of the molecular system can now be observed for phases of φ = π/2.τ 3 ) cos(ωsigτ 3 ) ( Re ⎡⎣ P ( 3) ⎡ ( 3) ⎤ (τ . 3π/2… and φ = 0.τ 3 ) ⎦ cos(ωsigτ 3 ) ) (24) Here the signal is expressed as a sum of two contributions. As before the imaginary part. the pump-probe can be interpreted as an incoherent . 4/18/2009 p. Incoherent pump-probe experiments What information does the pump-probe experiment contain? Since the time delay we control is the second time interval τ2. So that the (25) ΔI (τ ) ≈ nc ⎡ E′ ⎤ pr (τ ) Edic (τ ) ⎦ 2π ⎣ (26) Because the signal is in-quadrature with the polarization (π/2 phase shift). referred to as the birefringent (Ebir) and dichroic (Edic) responses. the diagrams for a two level system indicate that these measure population relaxation: ΔI (τ ) ∝ μ ab e −Γbbτ 4 (29) In fact measuring population changes and relaxation are the most common use of this experiment. π… . the absorptive or dichroic response is in-phase with the transmitted probe. When dephasing is very rapid.τ 3 ) ⎤ ⎦ sin(ωsigτ 3 ) + Im ⎣ P (τ . In this scheme. respectively. MIT Department of Chemistry. whereas the birefringent part is not observed.Andrei Tokmakoff. the transmitted probe is ′ E′ pr (τ 3 ) = E pr (τ 3 ) cos(ω prτ 3 ) . I (τ . 11-40 Esig (τ 3 ) = 4πωsig l nc = Ebir (τ . If we allow for the phase of the probe field to be controlled. for instance through a quarter-wave plate before the sample. or dichroic response. then we can write ′ E′ pr (τ 3 . ϕ ) = E pr (τ 3 ) cos(ω prτ 3 + ϕ ) . whereas the birefringent response corresponds to the real part of the nonlinear polarization and represents the phase-shift or retardance of the signal field induced by the sample. describes the sample-induced amplitude variation in the signal field.

so that Γaa = Γbb. If population relaxation from the excited state is through an intermediate. These diagrams reflect the equal signal contributions from the ground state bleach (loss of ground state population) and stimulated emission from the excited state. P (ω pr . which mirror our earlier description of electronic spectroscopy for an electronic transition coupled to nuclear motion. then the incoherent pump-probe signal is related to the joint probability of exciting the system at ωpu and detecting at ω pr after waiting a time τ.τ . There are two equivalent ways of describing the experiment. If we consider the case of pump-probe experiments on electronic states where ω pu = ω pr . the signal decays as a results of population relaxation of the initially excited state.Andrei Tokmakoff. MIT Department of Chemistry. For direct relaxation from excited to ground state the loss of population in the excited state Γbb is the same as the refilling of the hole in the ground state Γaa. When the resonance frequencies of the pump and probe fields are different. The two-level system diagrams indicate that the evolution in τ2 is differentiated by evolution in the ground or excited state. ω pu ) . Coherent pump-probe experiments Ultrafast pump-probe measurements on the time-scale of vibrational dephasing operate in a coherent regime where wavepackets prepared by the pump-pulse modulate the probe intensity. Both wavepackets will contribute to the signal. then the pump-probe decay will reflect equal contributions from both processes. or chemical kinetics. 4/18/2009 p. and the differential intensity (or absorption) change is proportional to the change of population of the states observed by the probe field. our description of the pump-probe from Feynmann diagrams indicates that the pump-pulse creates excitations both on the excited state and ground state. For the case where the pump and probe frequencies are the same. 11-41 experiment. redistribution. which can be described by coupled first-order rate equations. This provides a mechanism for studying the dynamics of excited electronic states with coupled vibrations and photoinitiated chemical reaction dynamics. . The pump-induced population changes in the probe states can be described by rate equations that describe the population relaxation.

a two level system is inadequate to capture the wavepacket dynamics. n . 4/18/2009 p. describe the spectroscopy in terms of the four-level system diagrams given earlier. R1 E3 c a a a R2 τ2 d b a a E2 E1 ΔI (τ ) Quantum Beats Population Relaxation d d b c a b c a e −Γbbτ 2 τ e − iωcaτ 2 −Γcaτ 2 e − iωdbτ 2 −Γdbτ 2 . For an underdamped wavepacket these coherences are observed as quantum beats on the pump-probe signal. 11-42 The first is to describe the spectroscopy in terms of the eigenstates of H0. and the wavepacket dynamics are captured by HSB. we see that the R2 and R4 terms describe the evolution of coherences in the excited electronic state. whereas the R1 and R3 terms describe the ground state wave packet.Andrei Tokmakoff. In addition to the population relaxation terms. MIT Department of Chemistry. The second draws on the energy gap Hamiltonian to describe the spectroscopy as two electronic levels HS that interact with the vibrational degrees of freedom HB. For the eigenstate description. Instead. e.

c. . The same processes contribute to Optical Kerr Effect Experiments and Impulsive Stimulated Raman Scattering.Andrei Tokmakoff. The CARS experiment is similar to a linear experiment in which the lineshape is determined by the Fourier transform of C (τ ) = α (τ ) α ( 0 ) . 11-43 CARS (Coherent Anti-Stokes Raman Scattering) Used to drive ground state vibrations with optical pulses or cw fields.c. A second interaction with the pump frequency lead to a signal that radiates at the antiStokes frequency: ωsig = 2ωP − ωS and the signal is observed background-free next to the transmitted pump field: ksig = 2kP − kS . The experiment is described by R1 to R4. The first field is the “pump” and the second is the “Stokes” field. 4/18/2009 p. • • • Two fields. = α eg e − iωegτ −Γegτ α ge + c. with a frequency difference equal to a vibrational transition energy. are used to excite the vibration. MIT Department of Chemistry. and the polarization is R ( ) = μev′ μv′g e 3 −iωegτ −Γegτ μ gv μve + c.