time attributed to the lowest triplet T 1 state. In this case, a detailed study of spininduced ZFS and vibronic effects may result in a direct comparison with accurate
experimental findings. The second category of molecules represented by Re
(I) complexes is characterized by shorter lived signals following a cascade of
ultra-fast luminescence decays attributed to S n and T n states. The interpretation of
the processes underlying emissive properties has to be based on non-adiabatic
excited states dynamics, either within a full quantum method from first principle
or following surface hopping trajectories within a semi-classical approach. Nevertheless, a TD-DFT static approach based on the optimized structures of the key
emissive states is very useful for deciphering the mechanism as preamble to the
dynamical study. The in-between luminescent behaviors exemplified by flexible Pt
(II) square planar complexes have to be treated with care, some basic approximation such as Kasha rules being meaningless in that case as illustrated by the
applications reported in the present contribution.
The simulation of ISC processes by means of wave packet propagations on spinvibronic coupled multi-dimensional PES is out of reach for large transition metal
complexes. Pioneering studies based on recent methodological developments have
been able to reproduce with reasonable accuracy ultrafast luminescent time scales
observed in first-, second- and third-row transition metal complexes. Counterintuitive heavy atom effects on ISC kinetics observed experimentally have found
explanations in detailed theoretical analysis.
Whereas TD-DFT absorption spectra have been validated by a number of joined
experimental/theoretical studies or by comparison with accurate ab initio methods
in the past decade, we do not have the benefit of hindsight as far as the emissive
properties are concerned.
The interpretation of ultra-fast time-resolved spectroscopy outcomes is especially challenging because both spin-orbit and vibronic coupling effects have to be
considered. In the systems investigated so far involving mostly singlet and triplet
states, the standard level of approximation used for SOC assessment seems realistic.
For higher multiplicities or situations with large mixing between electronic states of
different multiplicities, a more refined treatment of spin-orbit interactions could be
mandatory. The newly developed perturbational treatment for generally applicable
high-level multireference methods is one useful approach [133].
The simulation and computation of ISC rates in pioneering applications are
based on different strategies: (1) non-adiabatic molecular dynamics where
TD-DFT is coupled to classical trajectory-based methods; (2) quantum dynamics
where both electronic and nuclear wave functions are treated exactly within a given
level of approximation; and (3) time-dependent non-radiative rate theory in the
multi-mode harmonic oscillator and Condon approximations, where the electronic
spin-orbit matrix elements depend linearly on the nuclear coordinates within a spinvibronic coupling scheme. As illustrated by the examples developed above for first, second- and third-row transition metal complexes, each approach has its strengths
and weaknesses. Several important aspects have to be considered in this expanding
field, namely the solvation dynamics, the variation of spin-vibronic couplings with
nuclear relaxation, the construction of realistic model Hamiltonians, and the
Absorption Spectroscopy, Emissive Properties, and Ultrafast Intersystem. . .
409
experimental findings. The second category of molecules represented by Re
(I) complexes is characterized by shorter lived signals following a cascade of
ultra-fast luminescence decays attributed to S n and T n states. The interpretation of
the processes underlying emissive properties has to be based on non-adiabatic
excited states dynamics, either within a full quantum method from first principle
or following surface hopping trajectories within a semi-classical approach. Nevertheless, a TD-DFT static approach based on the optimized structures of the key
emissive states is very useful for deciphering the mechanism as preamble to the
dynamical study. The in-between luminescent behaviors exemplified by flexible Pt
(II) square planar complexes have to be treated with care, some basic approximation such as Kasha rules being meaningless in that case as illustrated by the
applications reported in the present contribution.
The simulation of ISC processes by means of wave packet propagations on spinvibronic coupled multi-dimensional PES is out of reach for large transition metal
complexes. Pioneering studies based on recent methodological developments have
been able to reproduce with reasonable accuracy ultrafast luminescent time scales
observed in first-, second- and third-row transition metal complexes. Counterintuitive heavy atom effects on ISC kinetics observed experimentally have found
explanations in detailed theoretical analysis.
Whereas TD-DFT absorption spectra have been validated by a number of joined
experimental/theoretical studies or by comparison with accurate ab initio methods
in the past decade, we do not have the benefit of hindsight as far as the emissive
properties are concerned.
The interpretation of ultra-fast time-resolved spectroscopy outcomes is especially challenging because both spin-orbit and vibronic coupling effects have to be
considered. In the systems investigated so far involving mostly singlet and triplet
states, the standard level of approximation used for SOC assessment seems realistic.
For higher multiplicities or situations with large mixing between electronic states of
different multiplicities, a more refined treatment of spin-orbit interactions could be
mandatory. The newly developed perturbational treatment for generally applicable
high-level multireference methods is one useful approach [133].
The simulation and computation of ISC rates in pioneering applications are
based on different strategies: (1) non-adiabatic molecular dynamics where
TD-DFT is coupled to classical trajectory-based methods; (2) quantum dynamics
where both electronic and nuclear wave functions are treated exactly within a given
level of approximation; and (3) time-dependent non-radiative rate theory in the
multi-mode harmonic oscillator and Condon approximations, where the electronic
spin-orbit matrix elements depend linearly on the nuclear coordinates within a spinvibronic coupling scheme. As illustrated by the examples developed above for first, second- and third-row transition metal complexes, each approach has its strengths
and weaknesses. Several important aspects have to be considered in this expanding
field, namely the solvation dynamics, the variation of spin-vibronic couplings with
nuclear relaxation, the construction of realistic model Hamiltonians, and the
Absorption Spectroscopy, Emissive Properties, and Ultrafast Intersystem. . .
409