Avoiding Irreversible Dynamics in Quantum Systems

Devices that exploit laws of quantum physics offer revolutionary advances in computation and communication. However, building such devices presents an enormous challenge, since it would require technologies that go far beyond current capabilities. One of the main obstacles to building a quantum computer and devices needed for quantum communication is decoherence or noise that originates from the interaction between a quantum system and its environment, and which leads to the destruction of the fragile quantum information. Encoding into decoherence-free subspaces (DFS) provides an important strategy for combating decoherence effects in quantum systems and constitutes the focus of my dissertation. The theory of DFS relies on the existence of certain symmetries in the decoherence process, which allow some states of a quantum system to be completely decoupled from the environment and thus to experience no decoherence. In this thesis I describe various approaches to DFS that are developed in the current literature. Although the general idea behind various approaches to DFS is the same, I show that different mathematical definitions of DFS actually have different physical meaning. I provide a rigorous definition of DFS for every approach, explaining its physical meaning and relation to other definitions. I also examine the theory of DFS for Markovian systems. These are systems for which the environment has no memory, i.e., any change in the environment affects the quantum system instantaneously. Examples of such systems include many systems in quantum optics that have been proposed for implementation of a quantum computer, such as atomic and molecular gases, trapped ions, and quantum dots. Here I develop a rigorous theory that provides necessary and sufficient conditions for the existence of DFS. This theory allows us to identify a special new class of DFS that was not known before. Under particular circumstances, dynamics of a quantum system can connive together with the interactions between the system and its environment in a special way to reduce decoherence. This property is used to discover new DFS that rely on rather counterintuitive phenomenon, which I call an "incoherent generation of coherences." I also provide examples of physical systems that support such states. These DFS can be used to suppress & coherence, but may not be sufficient for performing full quantum computation. I also explore the possibility of physically generating the DFS that are useful for quantum computation. For quantum computation we need to preserve at least two quantum states to encode the quantum analogue of classical bits. Here I aim to generate DFS in a system composed from a large collection of atoms or molecules and I need to determine how one should position atoms or molecules in 3D space so that the overall system possesses a DFS with at least two states (i.e., non-trivial DFS). I show that for many Markovian systems, non-trivial DFS can exist only when particles are located in exactly the same position in space. This, of course, is not possible in the real world. For these systems, I also show that states in DFS are states with infinite lifetime. However, for all practical applications we just need long-lived states. Thus in reality, we do just need to bring quantum particles close together to generate an imperfect DFS, i.e. a collection of long-lived states. This can be achieved, for example, for atoms within a single molecule.