Theoretical Photochemistry WiSe 2017/18

Transcription

1 Theoretical Photochemistry WiSe 2017/18 Lecture 7 Irene Burghardt (burghardt@chemie.uni-frankfurt.de) Theoretical Photochemistry 1

2 Topics 1. Photophysical Processes 2. The Born-Oppenheimer approximation 3. Wavepackets 4. Beyond Born-Oppenheimer non-adiabatic transitions 5. The Franck-Condon picture of electronic transitions 6. Interaction with light & what kind of spectroscopies? 7. Conical intersections 8. Examples: Ethene, Protonated Schiff Bases (Retinal), Azobenzene 9. Some electronic structure aspects 10. Dynamics: trajectories or wavefunctions? 11. Wavefunction propagation techniques 2

3 12. Trajectory surface hopping techniques 13. Non-linear optical spectroscopy: calculation of spectroscopic signals 14. Extended systems: Excitons, light-harvesting, etc. 15. Solvent/environmental effects 3

4 What do we need to calculate? (1) PES s from electronic structure calculations: calculate Potential Energy Surfaces (PES) calculate non-adiabatic couplings possibly diabatize (= transform to a diabatic representation) (2) Dynamics (once the PES s are known): wavepacket simulations on non-adiabatically coupled PES s or simplified descriptions of the dynamics: classical trajectories, Gaussian wavepackets (3) Spectroscopic signals: Franck-Condon wavepackets transition dipole moment, oscillator strength absorption spectra, nonlinear optical spectroscopy... 4

5 Interaction with the electromagnetic field Ĥ = Ĥ mol + Ĥ int (t) semiclassical equations for the interaction with the field: Ĥ int (t) = E(r, t)ˆµ = E(r, t) α ˆµ α = E(r, t) α q αˆr α back-of-the-envelope derivation: electromagnetic field E [N C 1 ] = force per unit charge e.g., force F x for a field in x direction: F x = α q αe x = (dv /dx) hence, the potential of interaction reads V = α q αxe x = α µ αe x 5

6 In general: Measure polarisation resulting from the interaction with an electromagnetic field ) P = P (1) + P (2) + P (3) +... = ɛ 0 (χ (1) E + χ (2) E 2 + χ (3) E χ (n) = nth order susceptibility 6

7 Absorption spectrum ( 1st-Order Polarization ) σ(ω I ) = 4π2 ω I 3 hc ψ E n µ ψg i 2 δ(ω I ω n ) n 7

8 Fermi s Golden Rule Transition probability between quantum states that are subject to a perturbation Ĥ(t) = Ĥ 0 + ˆV (t) e.g., ˆV (t) = ˆµ E 0 (e iωt + e iωt ) perturbation Transition rate between two states a b (in 2nd order perturbation theory): Γ a b = 2π h ψ(0) b ˆµ ψ(0) a 2 δ(e 0 b E0 a ± hω) where E 0 b E0 a = hω ba - resonance condition! 8

9 Perturbation Theory Brief Primer Ĥ = Ĥ 0 + λ ˆV Ĥ 0 = unperturbed Hamiltonian ˆV = perturbation Hamiltonian λ = small parameter Now expand ψ n = ψ (0) n E n = E (0) n + λψ(1) n + λ2 ψ (2) n λe(1) n + λ2 E (2) n +... and insert into the Schrödinger Equation Ĥψ n = E n ψ n... 9

10 Perturbation Theory Brief Primer, cont d... and insert into the Schrödinger Equation Ĥψ n = E n ψ n : (Ĥ 0 + λ ˆV )(ψ (0) n + λψ(1) n +...) = (E (0) n + λe(1) n + λ2 E (2) n +...)(ψ(0) n + λψ(1) n +...) Finally, sort according to contributions in λ 0, λ 1, λ 2 : (λ 0 )... Ĥ 0 ψ (0) n = E (0) n ψ(0) n (λ 1 )... Ĥ 0 ψ (1) n + ˆV ψ (0) n = E (0) n ψ(1) n (λ 2 )... Ĥ 0 ψ (2) n + ˆV ψ (1) n = E (0) n ψ(2) n + E(1) n ψ(0) n + E(1) n ψ(1) n + E(2) n ψ(2) n and obtain first-, second-order energies and wavefunctions... 10

11 Second-Order Perturbation Theory 11

12 Transition dipole moment & oscillator strength transition dipole moment: ψ E n µ ψg i µ EG ψ n ψ i with the electronic matrix element µ EG = E ˆµ G and the Franck- Condon factors S ni = ψ n ψ i oscillator strength: (dimensionless quantity which measures the total area under the absorption band σ(ω I ) = n ψ n µ ψ i 2 δ(ω I ω n )): 4π 2 ω I 3 hc f = ( 4πme ω EG 3e 2 h ) µ EG 2 12

13 Dipole allowed and forbidden transitions (a) carbonyl (C=O) group: π n transition is forbidden n O2p y ψ π = c χ(c2p x ) + cχ(o2p x ) π µ n c Op x µ Op y = 0 but: intensity borrowing possible (b) ethene: π π transition is allowed transition to π induces twisting 13

14 Another case: Benzene (D 6h ) three important bands in the UV: 185 nm symmetry-allowed, intense 200 nm symmetry-forbidden, weak 260 nm symmetry-forbidden, weak ground state: G( 1 A 1g ) electric dipole operator: A 2u (z) + E 1u (x, y) allowed transitions: E 1u 1 A 1g (185 nm) 1 A 2u 1 A 1g forbidden (but weak) transitions: 1 B 1u 1 A 1g (200 nm) 1 B 2u 1 A 1g (260 nm) 14

15 PYP chromophore Electronic Structure of the PYP Chromophore A R T I C L E S Table1. CC2CalculatedPropertiesforthe π-π 1 /, π-π 2 /,n-π 1 /,n Ph-π 1 /,and π-arg52excitedstatesinthedifferentchromophore + Amino Acid(s) Supermolecular Complexes: Oscillator Strengths, f(au); Change in the Permanent Dipole Moment under Transition to the Excited State, µb (Debye) π π 1 / π π 2 / n π 1 / nph π 1 / π Arg52 molecular system/complex f µb f µb f µb f µb f µb pctm pctm < I pctm - +Arg < II I + Cys III I +Phe62+Phe < IV I + Thr50 + Val66 + Tyr V I + Tyr42 + Glu VI IV + Tyr42 + Glu VII VI + Cys Figure 3. (a-e) Patterns of the principal highest occupied and lowest unoccupied(virtual) molecular orbitals of the deprotonated chromophore (pctm - );(f)patternofthelowestunoccupiedmolecularorbitalofcomplex I(pCTM - +Arg52).Notethattheorbitalpatternsofthe πand π 1 / orbitals differfromthoseoftheneutralchromophore(seefigure1ofref18).in particular, the π 1 / orbital is no longer localized on the double bond conjugated with the aromatic ring. eration depend on the chromophore s environment. The n and n Ph MOsessentiallycorrespondtolonepairs,withthenorbital relating to the carbonyl oxygen lone pair with a contribution from the 3p atomic orbitals of sulfur, while the n Ph orbital corresponds to the lone pair of the phenolic oxygen. The results presented in the diagram were obtained throughout from CC2 calculations. For pctm, pctm -, and complex I, we also carried out EOM-CCSD calculations. This allowed us toassessthequalityofthecc2methodindescribingtheexcited states under consideration. A comparison of the CC2 and EOM- CCSD data obtained for those three systems shows that the CC2 method correctly describes the π-π 1 /, π-π 2 /, n-π 1 /, and π-arg52 excited states, with deviations from the EOM-CCSD energiesthatarenotlargerthanabout0.3evforthe π-π 1 / and π-π 2 / states, 0.1 ev for the n-π 1 / state, and 0.4 ev for the π-arg52 state. The EOM-CCSD energies are always higher than the CC2 values(explicit values of the excitation energies under consideration can be found in table SM2 of the Supporting Information). Then Ph -π 1 / excitedstateisaparticularcase: Here,theCC2 method was found to substantially underestimate the excitation energy,withadeviationofabout1evfromtheeom-ccsd result. A possible reason for this poor agreement is the more complicated electronic structure of the n Ph -π 1 / state and, in particular, the somewhat higher weight of doubly excited 15

16 Time-domain representation: Franck-Condon wavepacket excited-state wavepacket: φ i (0) = µ ba ψ i (0) 16

17 Franck-Condon transition wavepacket picture ground-state wavefunction: ψ (0) (x, t 0 ) = χ (0) G (x, t 0) G potential hν act with the dipole moment operator on ψ (0) : ˆµ ψ (0) = µ EG ( E G + G E ) ψ (0) = µ EG χ (0) G (x, t 0) E φ E (x, t 0 ) molecular coordinate φ E (x, t) = excited-state wavepacket non-stationary state! 17

18 Absorption spectrum: frequency and time-domain representation σ(ω I ) = 4π2 ω I 3 hc = 2πω I 3 hc ψ E n ˆµ ψg 0 2 δ(ω I ω n ) n dt φ E (0) φ E (t) e iωt excited-state wavepacket: φ E (0) = ˆµ ψ G 0 autocorrelation function: C(t) = φ E (0) φ E (t) The Fourier transform of φ E (0) φ E (t) yields the absorption spectrum 18

19 Translate Fermi s Golden Rule rate to time-dependent picture use Fourier transform relation δ(ω ω n ) = 1 2π dt e i(ω ω n)t such that σ = 4π2 ω I 1 3 hc 2π = 2πω I 3 hc = 2πω I 3 hc dt n dt n e i(ω ω n)t ψ G 0 ˆµ ψe n ψe n ˆµ ψg 0 e iωt ψ G 0 ˆµ e iω nt ψ E n S n0 dt e iωt φ E (0) φ E (t) 19

20 Time-frequency correspondence recurrences of the autocorrelation function generate the vibrational fine structure of the spectrum! 20

21 Emission spectra emission spectrum: very similar description but distinguish stimulated vs. spontaneous emission: Einstein coefficients A sp fi = (8πhν3 fi /c3 )Bfi st emission spectrum is red-shifted as compared to absorption 21

22 Second-Order Processes second-order perturbation theory: Ĥ = Ĥ 0 + λ ˆV (Ĥ 0 + λ ˆV )( ψ (0) + λ ψ (1) +...) = (E (0) n + λe(1) n + λ(2) E (2) n +...)( ψ(0) + λ ψ (1) +...) 22

23 Two-Photon Absorption (TPA) TPA is quadratically proportional to the intensity of the incident light localized excitation in a small volume: useful for materials & biosystems! 23

24 Raman Spectroscopy ( Inelastic Scattering of a photon by a molecule ) 24

25 Raman Spectroscopy (Off-resonant Raman Scattering) Kramers-Heisenberg-Dirac (2nd order perturbation) theory: I ab (ω) = = = χ b ˆµ m m ˆµ χ a E m a + hω E m + iɛ φ b m m φ a E m a + hω E m + iɛ dt e iωt φ b φ a (t) 2 0 = φ b R a ω R a ω = Raman wave function Lee, Heller, J. Chem. Phys. 71, 4777 (1979) I ab (ω) relates to polarizability! 25