MIT Department of Chemistry�5.74, Spring 2004: Introductory Quantum Mechanics II Instructor: Prof. Robert Field 5.74 RWF Lecture #10 Supplement10S – 1 Stationary Phase for Vibration-Electronic Spectra Consider two electronic potential energy curves R V′′′′(R) V″″″″(R) ∆∆∆∆Te ′E vib ′′E vib ′′Re ′Re E The classical Franck-Condon principle specifies ∆R = 0, ∆P = 0. If ∆P = 0, then the kinetic energy, KE(R), must also be unchanged upon excitation RKE ′()= E ′− V ′()R vib R KE ′′()= E ′′ − V ′′()R vib thus E ′− E ′′ = V ′()− V ′′()R Rvib vib which can only be satisfied at special values of R, each of which is a stationary phase point. The choice of transition frequency from a given vibrational level of either the lower or upper electronic state determines the stationary phase point. This is the R-value (or R-values) at which the transition occurs and the region where a piece of the initial state wavefunction is transferred onto the final state potential.5.74 RWF Lecture #10 Supplement 10S – 2 The transition frequency hν =∆Te + E ′− E ′′ =∆Te + V ′()− V ′′()R Rvib vib hν R R−∆Te = V ′()− V ′′()= V ′(Rsp )− V ′′(Rsp ) specified by the experimentalist satisfied at Rsp Rsp may be swept through a small region of the initial state vibrational wavefunction by systematic variation of the center frequency of the probe laser, ν. As Rsp sweeps through lobes of the initial state wavefunction, the transition amplitude increases and decreases. The maximum transition amplitude is obtained when Rsp is at the maximum of the outermost lobe of the vibrational wavefunction (because the outer turning point is always softer than the inner turning point). A similar Franck-Condon-like principle applies to Rydberg-Rydberg transitions (1 ll +1)Vr() =− +l r 2µr2 µ≈ m ≡ 1e l +1Vl r() − V () = +1 r l 2r  1 1 n′2 − n′′2  ∆Enn′′′= hc ℜn′′2 − n′2 = hcℜ ′′2nn′2 In Rydberg units hcℜ = 1 n′ = n″ + m 2n′2 − n′′2 n′′2 + 2mn ′′ + m − n′′2 = 2 2 ′′2nn′2 n′′ (n′′ + 2n′′m + m2 )5.74 RWF Lecture #10 Supplement 10S – 3 Two useful limits: ∆E(1) m >> n″ (excitation from low-n″ initial state) n′n″ = hcℜn″–2 ∆E(2) m << n″ (transition between high Rydberg states) n′n″ = hcℜ 2mn″–3 Stationary phase requires ()= l + 1 ′′′= Vl r()− Vl∆En n +1 rsp sp 2rsp / l( + 1)n′′ 2 12 Case (1) rsp = hcℜ / l( + 1)n′′ 3 12rCase (2) sp =mhcℜ Case (1) shows that rsp is small (inside the core) when n″ is small and m is large. High-n wavepackets are launched from inside the core.let n″= 1, l = 0, rsp = 1a0 (1a0 = 0.529Å). Case (2) implies that rsp can be very large. The rsp lies near the outer turning point of the lower state. n″ = 4, l = 3, m = 1 (5g ← 4f) r = 16a0 spYou should verify that there is no stationary phase point for n = 5, l = 2 ← n = 4, l = 3! All upward (i.e. ∆n > 0) Rydberg-Rydberg transitions have a strong ∆l = +1 propensity rule. By adjusting n″ and n′ (via choice of initial state and center transition frequency), one can build a Rydberg wavepacket that looks like the inner or outer lobe of the initial n″,l″ eigenstate. Simple physical pictures are based on this!5.74 RWF Lecture #10 Supplement 10S – 4 Heller’s Fractionation Index This provides a simple measure of how strongly a single bright state gets diluted into a bath of dark states. Three cases: 1) You have an Heff and can look at all of the eigenvectors. ψ B 0()()0 1 ≤≤Suppose is bright and iN} are dark. The eigenvectors are{ψ Di 2 12/N j0() ()ψ j =∑ cij + 1−∑ ψ B 0ciψ Di  44ii =1 12443β j The relative intensities of transitions into the N + 1 mixed eigenstates ψj〉 are given by β2.jHeller’s fractionation index, for the Bright state, B, is N +1 fB =∑ β4 j . j =1 If all eigenstates have equal shares of the Bright state character (strong-mixing limit) β  = (N + 1)–1/2 j−12 ]4 terms insum fB = [(N + 1) / (N + 1)=(N + 1)−1 which implies that the single Bright state is delocalized into N + 1 states, N of which are dark. Alternatively, you can ask how fractionated a single eigenstate is with respect to the zero-order basis in which it is expressed.  Nj 4  cifj =∑()+β4 ji =15.74 RWF Lecture #10 Supplement 10S – 5 This fractionation index depends on the choice of basis set, just as the Bright state fractionation index is determined by the specific basis set to which the Bright state belongs. 2. An eigenstate-resolved spectrum.Heller’s fractionation index is, where Iiis the relative intensity of the i-th line, N ∑ Ii 2 f = i= 1 2. N   ∑ Ii i= 1 As for the eigenvector fractionation, if each line has equal intensity, Ii = 1/N, and N ∑ N− 2 (− 2 )NN 1f = i= 1 = = N ⋅ N− 1 2 ( NN− 1) 2 N . ∑ i= 1 So we get the same result for the maximally fractionated (small f) limit, regardless of whether we look at eigenvectors or eigenstate-resolved transition intensities.5.74 RWF Lecture #10 Supplement 10S – 6 3. A continuous spectrum.For simplicity, consider I(ν) as a single rectangular feature, I Imax ← ∆ν 0 → ν 2∫ Idν 2 ∆ν 1f = 2 = Imax = (∫ Idν) (Imax ∆ν)2 ∆ν . The stronger the mixing of the Bright state with the continuum bath, the broader is the spectrum. This can be converted to an eigenstate dilution factor merely by assuming that the density of states, ρ(ν), is independent of frequency. # states Nρ = = .δν ∆ν “state”width of one Thus, in the limit of maximum fractionation within ∆ν, 1f = = ρ N calculated ∆ν ρtrue = Nf observed = N ∆ν5.74 RWF Lecture #10 Supplement 10S – 7 The fractionation index tells us what fraction of symmetry and energetically accessible state or phase space is accessed by our pluck. 1 Fraction of space accessed = Nf when f = 1/N, the fraction accessed = 1 (all of it!) when f = 1, the fraction accessed =