Femtochemistry is a mature field which now moves beyond observing molecular motion, towards laser control of chemical reactions. In early molecular control experiments success meant simply increasing the reaction yield. However the shape of the optimal light pulses and the particular path taken by a molecule can reveal a lot about the inner workings of chemical reactions. Thus the emphasis has shifted towards understanding control mechanisms. A second important development is progress beyond control of prototype systems towards very large, biologically relevant molecules. Finally, coherent control methods can be used to enable selective spectroscopy. Specially crafted laser pulses can enhance the reaction branching ratio for hard-to-reach, low yield states. These states can then be spectroscopically analyzed with good signal to background ratio. All three of these developments are emphasized in this work.

This PhD thesis is split in four sections. Chapter I introduces the problem, outlines the experiments, and draws the scientific background. Chapter II presents the laser setup and the methods used for developing an ultrashort, shaped, visible laser source.

Chapter III describes the first coherent control experiment where vibrational coherences in the laser dye LD690 have been selectively excited in the ground or first excited electronic state by means of spectral tuning and linear phase chirp. Blue-tuning of the excitation pulses coupled with positive chirp significantly boosts the excited state vibrations, well beyond what chirp alone can do. The enhanced selectivity enabled the retrieval of the excited state vibrational frequency and dephasing time without spurious ground state contributions.

Chapter IV presents a second experiment using a learning algorithm to maximize the yield of retinal isomerization in bacteriorhodopsin. The algorithm finds that very short, intense laser pulses increase the isomerization yield by 50% compared to long, low-intensity pulses. This solution unfolds a pathway which includes higher excited electronic states, not accessible under sunlight illumination. Arguably nature over-designed the molecule, allowing for this very efficient reaction pathway. Bacteriorhodopsin is very promising for bio-photonic applications and the results reported here have the potential to advance these endeavors.