Monkey business in the aquarium

tion of the object wavefunction in amplitude and phase from the measured data. However, before we can make a measurement of a quantum state we have to prepare that state. In the case of the caesium atom this job was done by the first object pulse. Can we prepare any desired state, that is, a target state? This is an issue that arises in many disciplines — ranging from physical chemistry, through atomic physics, to quantum optics — and many ways of tackling it have been suggested. In the technique known as coherent control, one tries to find the appropriate sequence of laser pulses needed to break a specific chemical bond, and in quantum state engineering one attempts to prepare quantum states of the radiation field or vibrational modes of an ion in a trap. Both approaches are central to recent proposals for quantum computers. How can we prepare a prescribed wavepacket of an electron orbiting the nucleus? The tools are short laser pulses. We therefore have to find the right shape and sequence of laser pulses to steer the quantum state to the target state. Usually the Schrödinger equation is used to calculate the quantum state that results after applying a sequence of laser pulses. However, in the case of coherent control, we concentrate on the inverse problem: we start from the target state and a rather simple initial state, and try to find the laser pulses that would lead to the desired state. In their latest work, however, Weinacht et al. pursue a different approach and combine quantum state preparation and measurement with feedback. The first laser pulse (yellow in Fig. 1) prepares a trial wavefunction, ct (blue), and with the help of the second reference pulse (red) they reconstruct the trial state. The authors compare the reconstructed trial wavefunction ct to the desired wavefunction cd (green). When the two wavefunctions do not agree they readjust the dials on the pulse shaper so as to create a more appropriate laser pulse. This process is repeated until the desired wavefunction cd emerges — usually within two cycles of the feedback loop. The sculpting of wavepackets described by Weinacht et al., and the coherent control of electrons in a solid-state quantum well reported last year, are two impressive examples showing that the field of quantum control has entered a new era. It has broadened out beyond chemical reactivity and now allows encoding and decoding information on wavepackets. Moreover, the Weinacht procedure shows that it is possible to address and control individual quantum units. This feature is crucial for the implementation of various suggestions for quantum computation and coherent information processing and transfer. The long-held dream of coherent control and quantum state engineering has become reality. Wolfgang P. Schleich is in the Abteilung für Quantenphysik, Universität Ulm, Albert Einstein Allee 11, D-89069 Ulm, Germany. e-mail: [email protected]