I have already told you about out activity in the field of electron dynamics and namely spin-flip dynamics. To wrap it briefly, we have looked what will happen to a system with the strong coupling of spin and orbital motion of electrons if we prepare a pure spin-state. One can imagine, e.g., a Fe2+ ion with four spin-up electrons forming a so-called quintet state. Now if such a system has a hole in the electronic core, which immediately causes strong spin-orbit coupling, an ultrafast spin-flip of an electron will occur leading to triplet final state with effectively only two spin-up electrons.
Well, sounds easy. However, such a situation is so far pure speculation, and we should ask ourselves, how to practically prepare this particular initial state? It is very special because it corresponds not to an eigenstate of a system but rather to a quantum superposition of such “natural” states. One can, of course, absorb light, but this light should also possess somewhat unusual characteristics. First of all, the light pulse should be very short. According to the uncertainty principle, the shorter is the pulse, the broader is it in energy range which it excites. And we exactly need excitation of lots of eigenstates to resemble the situation with an initially prepared core-excited pure-spin state.
Out of possible light sources, two candidates would potentially fit: free electron lasers and high harmonic generation (HHG) setups. They are able to produce light with energy that is high enough to excite core states and emerged pulses have temporal durations below 1 femtosecond (10-15 seconds). Free electron lasers provide single isolated pulses, whereas HHG usually gives periodic sequences of pulses. We have already discussed in previous publications, how the transition metal complex reacts on the isolated pulse. I have briefly described it in this blog.
In a recent publication:
we have looked how spin of the system will behave if we subject it to repeated sub-femtosecond pulses as resulting from the HHG source. The key difference to isolated pulses is that the yield of spin-fliped states rises in a stepwise manner after every subpulse in a train as can be seen from a figure. This makes this process to occur faster and more complete. Even more important is that due to stimulated emission the population is dumped from core-excited to spin-flipped valence states. Thus, via a core excitation and stimulated emission, and thus mediated by the strong spin-orbit coupling in the core state, an spin-flip which is faster than few fs can be triggered in the manifold of valence electronic state. Usually such a transition requires up to few hundreds of femtoseconds and might be immensely accelerated in this process. Such dumping also decreases the destructive influence of the Auger decay.
Finally, we have looked at the role of the decoherence caused by nuclear motions. Molecular vibrations have been treated at the level of a heat bath. Essentially, the electron dynamics studied in this work is that fast, that relatively slow vibrational motions do not much influence the result.
We envisage that this effect could be used for clocking ultrafast events. In this respect, it is of core-hole clock type but has a different nature. In case of spin flips, the characteristic timescale may be varied by changing the carrier frequency and bandwidth of the incoming radiation, thus adjusting the strength of the coupling and thereby determining essentially the measured time window.