Subdecoherent information encoding in a quantum-dot array

Devices using unique quantum-mechanical features can perform information processing in a much more efficient —or even unattainable— way than those relying just on classical physics. This fundamental discovery has stimulated in the last few years a big deal of work and scientific debates in the new born field of Quantum Computation. From a conceptual point of view these results represent a serious challenge to the time-honored notion of universal computational schemes independent of an underlying physical theory: information as well as computation are intrinsically physical. On the other hand, physical realizations of a Quantum Computer would result in tremendous practical advantages. The key ingredients which endow QC devices with computational capabilities that supersede their classical counterparts are basically: (i) the linear structure of their state space; (ii) the unitary character of their dynamical evolution; (iii) the tensorised form of multiparticle state spaces. The first two properties allow for a parallel processing of an arbitrary number of data sets, encoded in suitable quantum states. By resorting to quantum interference, between different computational branches, one can selectively amplify desired parts of the state vector in order to optimize the probability that a final (i.e., read-out) measurement will give us the information we were looking for. Point (iii) represents another striking departure from classicality: due to entanglement, combining different quantum systems results in an exponential growth of the available coding space; moreover, the tensor-product structure is at the very basis of many efficient quantum manipulations. Unfortunately, all this holds just for closed quantum systems. Real systems are unavoidably coupled with environmental (i.e., non computational) degrees of freedom. Such open character spoils points (i) and (ii) eventually turning quantum computing to classical. Different computational branches get entangled with different (quasiorthogonal) quantum states of the environment and their interference is then no longer observable. From a mathematical point of view, the relevant state space, given by density matrices, has now a convex structure and the allowed quantum dynamics is described by CP-maps. Initial pure preparations are typically corrupted on extremely short time-scales due to quantum-coherence loss that makes them mixed: the initial information irreversibly leaks out from the system into the huge number of uncontrollable degrees of freedom of the environment. This phenomenon —the so called decoherence problem in QC— represents the major obstacle for the experimental realization of any quantum-computing system. Other challenging requirements are of course given by the necessity of being able to perform on a system, with a well-defined state space, long coherent quantum manipulations (gating), precise quantum-state synthesis and detection as well.