The early steps in the photocycle of the bacterial proton pump proteorhodopsin (PR) were analyzed by ultrafast pump/probe spectroscopy to compare the rate of retinal isomerization at alkaline and acidic pH values. At pH 9, the functionally important primary proton acceptor (Asp97, pK(a) = 7.7) is negatively charged; consequently, a reaction cycle analogous to the archaeal bacteriorhodopsin (BR) is observed. The excited electronic state of PR displays a pronounced biphasic decay with time constants of 400 fs and 8 ps. At pH 6 where Asp97 is protonated a similar biphasic decay is observed, although it is significantly slower (700 fs and 15 ps). The results indicate, in agreement to similar findings in other retinal proteins, that also in PR the charge distribution within the chromophore binding pocket is a major determinant for the rate and the efficiency of the primary reaction.

The early steps (<1 ns) in the photocycle of the detergent solubilized proton pump proteorhodopsin are analyzed by ultrafast spectroscopic techniques. A comparison to the first primary events in reconstituted proteorhodopsin as well as to the well known archaeal proton pump bacteriorhodopsin is given. A dynamic Stokes shift observed in fs-time-resolved fluorescence experiments allows a direct observation of early motions on the excited state potential energy surface. The initial dynamics is dominated by sequentially emerging stretching (<150 fs) and torsional (approximately 300 fs) modes of the retinal. The different protonation states of the primary proton acceptor Asp-97 drastically affect the reaction rate and the overall quantum efficiencies of the isomerization reactions, mainly evidenced for time scales above 1 ps. However, no major influence on the fast time scales (approximately 150 fs) could be seen, indicating that the movement out of the Franck-Condon region is fairly robust to electrostatic changes in the retinal binding pocket. Based on fs-time-resolved absorption and fluorescence spectra, ground and exited state contributions can be disentangled and allow to construct a reaction model that consistently explains pH-dependent effects in solubilized and reconstituted proteorhodopsin.

The dependence of the interfacial electron transfer in alizarin-sensitized TiO2 nanoparticles on the sample pH has been examined via transient absorbance spectroscopy in the visible spectral region (443-763 nm). Excitation of the alizarin/TiO2 system with visible pump pulses (lambdaexc = 500 nm) leads to a very fast electron injection (tauinj < 100 fs) over a wide pH range. Back electron transfer shows complicated multiphasic kinetics and strongly depends on the acidity of the solution. The strong dependence of back-electron-transfer dynamics on the ambient pH value is explained by a Nernstian-type change in the semiconductor band energy. Indeed, a variation of pH values over 7 units leads to a approximately 0.42 eV change of the conduction band edge position (i.e., the nominal free energy of the electron in the electrode). Assuming a pH-independent redox potential of the dye, this change was sufficient to push the system to a condition where direct photoinitiated electron injection to intraband gap surface states could be investigated. The existence of an electron-transfer pathway via surface trap states is supported by the similarity of the observed back-electron-transfer kinetics of alizarin/TiO2 at pH 9 and alizarin/ZrO2 reported in earlier work (J. Phys. Chem. B 2000, 104, 8995), where the conduction band edge is approximately 1 eV above the excited state of the dye. The influence of surface trap states on interfacial electron transfer has been studied, and a detailed analysis of their population, depopulation, and relaxation kinetics is performed. Therefore, alizarin adsorbed on the surface of TiO2 nanoparticles is an ideally suited system, where pH-dependent investigations allow a detailed study of the electron dynamics in trap states of TiO2 nanoparticles.