The shaping of light pulses coupled with feedback to alter their shape for optimum results, provides an experimental knob to control molecules. Such control can lead to improvement of the yield of a reaction in a way that is not possible by other means. The research has three aspects: (1) determining the optimal light pulse to control the three fundamental vibrational modes of a 3-atom system (e.g., CO2), (2) measuring the spectrum and time characteristics of the light pulse which gives the desired results, and (3) understanding theoretically the interaction between the light pulse and the molecular system. In the series of experiments to be done, the optimally shaped light pulses (with durations measured in femtoseconds) will be obtained via so called "adaptive" feedback. State-of-the-art imaging of the molecular system following interaction with a light pulse causes changes to the next light pulse via a computer running genetic or evolutionary algorithms which "learn" how to alter the amplitude of each mode of the light pulse to obtain a result closer to that which is desired. This process continues until the desired outcome is achieved.
The results of this work may contribute to transforming the field of molecular control because they will have an impact beyond the small systems to be studied. The approach could be used for extracting important information associated with inter- and intra-molecular motion and dynamics from much larger systems. Efforts to control vibrational modes of molecules will be one key to enabling the use of ultra-short, shaped pulses of light in the mid infrared range of the spectrum for the creation and manipulation of molecular qubits. A qubit, analogous to the bit of a conventional computer, is the fundamental unit of "quantum computer", which is a potentially powerful computing machine which scientists do not yet know how to build. This work may contribute to better understanding of one possible solution to this problem. Finally, the results of this work could enable new approaches to problems of concern to environmental stewardship and National Security. This project is being carried out at the University of Maryland and will be a training ground for graduate and undergraduate students as well as postdocs. The cross disciplinary nature of the program involves students with physics, chemistry and engineering backgrounds. A special effort is being made to attract investigators from underrepresented groups into our program.
Pulse shaping coupled with closed-loop feedback searches (e.g., genetic algorithm) provides an experimental knob to control dynamics, to select pathways through transient states and to enhance product yields in quantum systems not readily available by other means. Such searches provide optimal pulses, solutions causing a specific outcome. The initial impetus for this approach came from a desire to optimize product yield with lasers that was not possible by simply irradiating the sample. Attempts to design a pulse failed because of complex curve crossing and intramolecular energy exchange. The feedback approach essentially enables the system to "teach" the laser what field to use to avoid naturally occurring roadblocks and energy traps, i.e., the optimal field is learned. Optimal fields, however, are generally unintelligible; it is not obvious how they achieve their goal. In addition, it is well known that optimal pulses leading to the same outcome are typically not unique. These characteristics inhibit fully utilizing optimal pulses to learn more about the underlying physics or for processes like quantum computing where the field needs to be understood precisely. This is on the verge of changing because of new insight provided by this project. Focusing our effort on controlling the dynamics associated with strong-field dissociative ionization of carbon dioxide (CO2), we have shown that one of the fundamental elements of a control pulse is the relative phase between adjacent components of the pulse. Using the pulse with two peaks where the relative phase of the peaks was adjusted while maintaining a fixed pulse separation, we controlled the bending amplitude of CO2. At the same time, and 180o out of phase, the strength of the CO2 ionization was also controlled. At the minima, the signals are about the same as what are obtained with a single pulse. The conclusion we are tempted to draw is that the phase of the pulse can render the second pulse powerless. While more work is required to prove this, if it holds true, this could explain why optimal pulses are not unique -- depending on the phase some components are not participating in controlling the dynamics.
