Coherent Control

The bond-selective control of a chemical reaction has been a longstanding goal of modern chemical physics. Early attempts using selective laser excitation were thwarted by fast intramolecular energy redistribution. Now ultrafast laser pulses, optical pulse shaping, and feed-back algorithms have been successfully combined in a number of laboratories to control bond dissociation reactions in simple isolated molecules. In our work these techniques are extended to control elementary solution phase reactions.

A genetic algorithm and optical pulse shaper was used to find ultrafast laser pulses that optimize the multi-photon ring-opening of cyclohexadiene (CHD). The GA used the absorption difference between Z-HT and CHD for feedback and optimization.

UV spectroscopy shows that more photoproduct was created by the tailored laser pulse than by the transform-limited pulse.

The reaction is coherently driven by a laser pulse with negatively chirped phase (inset).

Fitness functions optimized on the increase in absorption between 260 and 280 nm. In some experiments a penalty was added for absorption between 280 nm and 300 nm (arising from solvent fragmentation rather than ring-opening. In addition, the probe delay was varied to limit the contribution of background absorption to the fitness function.

The time delay between the pump and probe had a significant influence on the nature of the optimal solution found by the GA. For a 1 ns delay the optimal pulse is near transform limit. When the delay is increased to 12 ns the optimal pulse is no longer transform limited. The longer time delay minimizes the importance of background signal due to solvent fragmentation, and optimizes the ring-opening reaction. Two representative optimal pulses are shown below.

Many pulses were tested in these experments. Phase of each pulse may be fit with a Taylor’s series expansion. The quadratic and cubic terms (linear and quadratic chirp) are important parameters

For a probe delayed by 1 ns from the pump pulse the optimal pulses cluster near the transform limit. When the probe is delayed by 12 ns the optimal pulses are not transform-limited. A combination of negative linear chirp and positive quadratic chirp optimize the three-photon ring-opening reaction. (1 = best pulse, 0 = worst pulse)