One of the first experiments where the chemical reaction could be followed directly in the time domain has been performed in 1987 in our research group (at that time at the TUM) [1], [2]. In these experiments we followed the initial reaction dynamics of the photosynthetic protein Bacteriorhodopsin (BR). The initial reaction of BR is a light induced photoisomerisation of the chromophore retinal from the all-trans to the 13 - cis form (See Fig.1). The development of femtosecond light pulses in the early 1980's [3] allowed for the first time to study fast photochemical reactions in real time. In 1985, the latest part of the initial photoisomerisation reaction - the transfer to the ground state photo product - could be detected by optical techniques [4], [5], [6]. The most important question at that time treated the nature of the reactive motions preceding this transfer to the groundstate photoproduct. These reactive motions can be discussed within the frame of a potential energy diagram of the system where the energy of the different electronic states (in our case of the retinal molecule) is plotted as a function of a nuclear coordinate (see Fig.2). For the photoisomerisation of retinal, the reaction coordinate could be assigned - at least in part - to the rotation around the C13 - C14 double bond. The presented picture is a cut through the multidimensional energy landscape of the polyatomic molecule where we present only one of the many degrees of freedom (the minimum energy path). In this picture the excitation of bacteriorhodopsin by light promotes the molecule to the excited electronic state potential surface without any change of nuclear geometry. From this so-called Franck-Condon region the system moves along the S1 potential curve to the minimum. From this part of the potential curve the back reaction to the ground state takes place.

Fig.1: The relevant configurations of the retinal molecule

During the motion on the potential surface the distance between S1 and S0 surface is strongly changing. As a consequence one would expect a continuous change of the frequency of emission from the excited electronic state. These spectral changes can be well observed experimentally.

Fig.2: Potential energy diagram explaining the femtochemical S1-reaction of bacteriorhodopsin		Fig.3: Schematic of the set-up used for the femtosecond pump-probe experiment.

In the experiments be performed in 1987, we studied absorption changes and stimulated emission of bacteriorhodopsin on the 100 fs time scale. We excited the sample at 620 nm (Fig.3) and probed the transmission changes at different wavelengths (photon energies) throughout the near infrared (see arrows in Fig.4). In these experiments, Fig.5, we could find a fast decay of stimulated emission with 200 fs at high energies of the probing photons. At lower photon energies (larger wavelengths) we observed predominantly the decay of the S1 state with 500 fs. Within the model of the potential surfaces (see Fig.2) we can relate the 200 fs reaction to a motion of the retinal molecule away from the Franck-Condon region to a geometry from where the remaining isomerizational motion and the transition to the ground state can take place. The experimental data show clearly that the important initial part of the femtochemical reaction of the retinal molecule could be followed in time by the "slow motion" technique of ultrafast spectroscopy. In parallel to our experiments published in Chemical Physics Letters [1], other investigations had been performed by the group of R. Mathies and C. Shank [7]. These experiments were focused to the Franck-Condon region and the motion of the retinal out of this region. The data gave very interesting information complementary to our experiment, which presents the motion towards the minimum of the potential surface.

Fig.4: Absorption and fluorescence spectrum of bacteriorhodopsin.		Fig.5: Absorption changes recorded during the S1-reaction of bacteriorhodopsin.

More recent investigations by different research groups confirmed the basic results of the early papers. They also gave interesting additional information for a better understanding of the molecular processes during the S1 motion as a multidimensional process.

In conclusion: The experiments on bacteriorhodopsin have been the first where photochemical reactions could be observed in 'real time' on the excited state potential energy surface. Later on, fascinating experiments on photo dissociation of small molecules have displayed the dissociative motion on the S1 potential surface and oscillatory features during ultrafast chemical reactions (see: [8], [9] and The Nobel Price for Chemistry 1999, www.nobel.se/announcement-99/chemistry99.html )

If you are interested in doing research in the field of femtochemistry and bacteriorhodopsin please contact Wolfgang Zinth. Currently we are expanding this research area and can offer different diploma thesis (Diplomarbeiten) on ultrafast spectroscopy on bacteriorhodopsin and other retinal containing proteins. We are also offering positions for students (Werkstudenten) for IR-spectroscopy on these proteins.

References