Quantum Control of Photo-isomerization: Applications of Coherent Control

Chemical reactions are at the heart of chemistry. The ability to control the outcome and yield of molecular processes at the quantum level has been a long-standing vision in the field. This dissertation presents theoretical investigations into coherent control, a strategy that exploits the phase properties of laser light to manipulate molecular quantum dynamics. The work focuses on applying this framework to control the isomerization and dissociation dynamics of small molecules using shaped infrared pulses. We simulate the quantum dynamics in the electronic ground state using the multi-configuration time-dependent Hartree method. Shaped pulses are determined by optimizing the phase profile via iterative stochastic and gradient-based learning algorithms, with emphasis on obtaining pulse shapes that are experimentally feasible.

Inspired by recent experimental work, we investigate the effectiveness of pulse shaping for controlling the laser-induced dissociation of the helium hydride ion (HeH⁺) under non-ionizing conditions. Spectral phase modulation, modeled after a spatial light modulator, is applied to transform-limited infrared pulses, with frequency-dependent phases optimized via Bayesian optimization to promote vibrational ladder climbing. In one-dimensional simulations, coherent control of this cascade excitation process is achieved with a quadratic chirped pulse, yielding near-complete dissociation. In full dimensions, however, ladder climbing is hindered by rotational motion, resulting in significantly reduced yields. Higher-order polynomial phase profiles provide only marginal improvements over quadratic chirping. We find that pulse shaping is most effective at low pulse energies and broad spectral bandwidths, while intense transform-limited pulses with narrow bandwidths maximize dissociation efficiency in the absence of shaping.

This methodology was extended to the study of the prototypical HCN/HNC isomerization reaction. A pump–dump scheme was employed in 3D quantum dynamics simulations to transfer ground-state HCN population into the HNC well. Shaped IR pulses with a linear temporal chirp were found to efficiently induce isomerization by exciting overtone bending-mode transitions through a vibrational ladder-climbing mechanism. We compared the convergence rate of stochastic and gradient-based algorithms for optimizing the pump–dump pulses. Our implementation of the gradient-based GOAT algorithm, integrated with MCTDH in a closed-loop control scheme, achieved the fastest convergence. Extending the model to 4D by including hydrogen rotation about the CN bond made fundamental vibrational transitions accessible. Rovibrational energy levels obtained with this 4D model were in excellent agreement with exact literature values. We investigated the relative efficacy of fundamental versus overtone excitation pathways, finding that fundamental transitions yield lower isomerization efficiency due to the additional sequential excitation steps required to overcome the isomerization barrier. Moreover, a reduced isomerization yield is found in the 4D model compared to the 3D results. Preliminary full-dimensional simulations were explored but remain computationally prohibitive with the current model, and represent a direction for future work.

These studies provide insight into the mechanistic understanding of coherent control with shaped infrared pulses, showing how rovibrational couplings can fundamentally limit controllability in realistic, multi-dimensional systems. The work establishes clear optimization principles, such as leveraging overtone transitions for efficient isomerization, and provides a robust theoretical framework for designing experiments aimed at controlling molecular dynamics with light.