Until recently, chemistry was a science of passive control. By adjusting external parameters such as temperature and pressure or even by introducing a catalyst into the reaction, the chemist strove to maximize the yield of the desired products.

Then 2 decades ago, a few chemists with some imagination and foresight proposed a new idea: quantum control. By altering the quantum-mechanical wave nature or wave functions of the reactants, we are now able to control, to an extent far greater than was previously possible, the final product. As Stuart Rice of the University of Chicago put it in a recent Nature article, this type of active control "meddles at the very heart of a reaction."

Quantum control sits at the intersection of chemistry, physics, and mathematics. Progress in this new field will depend upon significant contributions from all three and is required in both the theoretical and experimental domains. As such, this rapidly expanding field is accessible to a wide variety of researchers and students who are examining their future possibilities.

Scientific Community Initially Sceptical

Much of the chemistry community initially resisted the notion that quantum mechanics could have a major role in determining the outcomes of chemical reactions. It was widely held that these effects would cancel each other out because of the many particles (such as photons, electrons, protons, and neutrons) that are typically involved.

However, Paul Brumer of the University of Toronto and Moshe Shapiro of the Weizmann Institute of Science in Israel--the two theoretical pioneers of quantum control--persevered. And with recent advances in laser and matter wave technology, the science is finally at the stage at which experimentation can verify theory. "For decades we have stood by passively observing and analyzing chemical dynamics," explains Brumer. "We now have the tools and understanding to alter these reactions at their most fundamental level."

Methods of Quantum Control

Brumer and Shapiro base their theory around the most fundamental of all quantum-mechanical phenomena--the two-slit experiment. It basically works this way: By providing a chemical reaction with two coherent pathways leading to the same possible final products, quantum-mechanical-interference effects can be produced. By manipulating the properties--relative amplitudes and phases, in quantum-mechanical terms--of the two interfering pathways, it is often possible to achieve extensive control over the final products.

Experimentalists have been most successful with reactions involving photons, such as the photoionization and photodissociation of molecules. Here, two different laser frequencies set up the two interfering pathways.

A more challenging problem that the Brumer-Shapiro group has set its sights on is the control of reactions involving two molecules or at least an atom-molecule collision, whereby the final products can be different molecular species. In this case, the two pathways result from two different quantum states of the initial reactants. Whereas numerical calculations show that substantial control should be possible, an experimental design for this case is still pending.

Another method of implementing quantum control than the time-independent formalism of Brumer and Shapiro is to use pulsed laser beams and to vary the time between pulses. David Tannor of the Weizmann Institute and Stuart Rice were among the first to propose such a scheme.

For example, suitably timed pulses kick a chemical reaction into proceeding through different pathways at different times during the entire process. In essence, the first pulse excites the chemical system from its original potential energy surface onto another surface, thus bypassing a bottleneck in the original pathway because the reaction leaps onto another pathway at just the right time. The system proceeds along this second surface until eventually a second pulse knocks the reaction back to the original pathway, only further along. The outcomes of a number of different chemical reactions, such as ionization vs. dissociation of diatomic sodium, have employed this method.

Experimental Learning Algorithms

A third scheme, developed by Herschel Rabitz of Princeton University, is "learning control." The essence of this idea is that the results of a control experiment actually modify the original control parameters. As such, the new parameters then set up yet another control experiment, the results of which again modify the parameters, and so the cycle continues.

This closed-loop operation is in fact an experimental analogy to the computational method of optimization. Rather than using approximations in order to simulate a chemical reaction computationally, which can be quite constraining for complicated many-body reactions, the experimental version of optimization bypasses the need for any such approximations. It actually performs the experiment!

Experimentalists such as Philip Bucksbaum of the University of Michigan, Ann Arbor, and Thomas Weinacht of the State University of New York, Stony Brook, have been very successful at developing these learning algorithms and using them for quantum control. They have used this technique to precisely tailor the shape of an atomic electron's wave function, in effect engineering "designer wave functions."

However, a theoretical understanding of the final solution obtained through such learning algorithms has been stubbornly evasive. Such an understanding would, for example, be able to determine whether the learning algorithm truly has settled upon a global optimum solution or whether it has simply found a local minimum.

An Exciting Quantum Future

There is a multitude of other quantum control methods and schemes under development, as well as numerous other research groups actively involved in their study. They typically convene at the regular Gordon Conference on Quantum Control of Atomic and Molecular Motion.

Present-day applications of quantum control are most striking in the field of condensed-matter physics, where, by controlling the state of individual quantum dots, scientists can create very fast semiconducter optical switches. With the anticipated rapid improvement in laser technology, it is foreseeable that tailored light pulses will be able to create large molecules that, in turn, will revolutionize industries such as the pharmaceutical industry. There might also come a day when scientists will be able to design completely new biologically active molecules using this technology.

The field of quantum control also finds itself blending into other research fields that are forming a new vanguard in quantum physics and chemistry. Exploding fields such as Bose-Einstein condensation and quantum computation and information are making heavy use of quantum-control methods for their own purposes. Opportunities for graduate students and postdocs to work in the field of quantum control are thus available not only directly in the ways mentioned in this article, but also under the auspices of these other exciting research areas.