Detection and Control of Coherent Processes on a Femtosecond Timescale
Many lightinduced chemical reactions like isomerization, fragmentation, and electrocyclic reactions are completed within just a few femtoseconds (1fs=10-15 seconds). Processes like these take place in many rapid biochemical and biophysical reactions and their dynamics are clearly dominated by the coherent wavecharacter of the system because of the short timescales. Owing to the developement of short tuneable laserpulses these fast reactions have become experimentally observable (leading to the 1999 Nobel Prize in Chemistry for A. Zewail).
To understand the underlying mechanisms not only experiments with excellent time resolution (below 20fs) but also theoretical research is necesseray. Our approach to describe wave phenomena in complex systems like this concentrates on those molecular motions that are active on the femtosecond timescale only, thus reducing 50 and more vibrational degrees of freedom to a lower dimensional subspace. This reduction to no more than two to four active coordinates enables us to exactly simulate the system's quantum mechanical dynamics and even to include interactions with the laser beam.
Characteristic for a large class of ultrafast reactions, that we are interested in, is the appearance of conical intersections. These are conically shaped reaction channels that, on the one hand, promote ultrafast transitions between potential energy surfaces of different electronic states and that are, on the other hand, branching points for different products. We are particularily interested in the delicate interplay of electronic and nuclear motion that drives the molecule towards these reactive centers and how this process might be influenced externally.
Our future goal is to control the resulting product ratios by intelligently modulated femtosecond laserpulses, since in principle quantumdynamical processes can be influenced with the help of optimized laserfields due to constructive and destructive interference. We already demonstrated successfully strategies for efficient lasercontrol of chemical reactions such as photoisomerization, SEP-processes and IR-excitation.
The methods already in use reach from quantum chemical computations of electronic structure to quantum dynamical wavepacket propagation to density matrix propagation including environmental influences. These simulations are performed on massively parallelized computers such as the IBM p690 Regatta since they need considerable computational power. Apart from those methods we develop new theoretical concepts, that will hopefully enable us to easily extract from many degrees of freedom of complex molecular systems all relevant reaction surfaces as well as the corresponding Hamiltonian. These modes should then be coupled to the remaining ones within a very promising new quantum mechanical concept that is as reliable as the density matrix formalism yet more attractive for numerical application. This scheme should make the quantum dynamical simulation of high dimensional ultrafast chemical reactions very realistic and efficient. Apart from this, we try to develop realistic concepts for the experimental detection of reactions and their control in real time on the basis of our theoretical investigations.
For further information take a look at the Nobel Prize webpage or read Ahmed H. Zewail's feature article in Journal of Physical Chemistry A 104 (2000) 5660-5694.