Theoretical Design of Robust Quantum Processes: Exciton Transfer and Quantum Computation
Sawaya, Nicolas Per Dane
MetadataShow full item record
CitationSawaya, Nicolas Per Dane. 2018. Theoretical Design of Robust Quantum Processes: Exciton Transfer and Quantum Computation. Doctoral dissertation, Harvard University, Graduate School of Arts & Sciences.
AbstractThis dissertation is split into two broad topics: the study of exciton transfer in a variety of organic systems, and the classical numerical simulation of quantum algorithms for chemistry. In both parts, the effects of environmental noise played an integral role. For the latter portion, a main focus was to model the implementation of a quantum algorithm on noisy quantum hardware. For the exciton studies, environmental noise was a defining characteristic of the dynamical process being studied. In both of these research areas, advanced numerical methods were often implemented for use on parallel hardware.
The first excitonic system we present is the chlorosome, a large photosynthetic antennae found in some species of green sulfur bacteria, containing tens of thousands of molecules. Using a method capable of simulating massive open quantum systems, we modeled the dynamics of an exciton's migration across this unusual supramolecular object. The chlorosome's unique concentric-roll structure leads to some anomalous effects with respect to initial conditions, and the study is a demonstration that these numerical methods can be directly applied to other very large excitonic simulations.
In a theoretical-experimental effort, our next system is a rationally designed DNA-based excitonic system. We designed a coherent "building block'' in which we demonstrated that an exciton can delocalize across multiple pseudoisocyanine molecules, which are non-covalently attached to the DNA backbone. With a combination of computational and experimental evidence, we show that these building blocks can be used to create complex DNA structures with somewhat tunable excitonic properties.
In the final excitonics project discussed in this dissertation, we designed a novel ``exciton gate.'' The design is similar in principle to a traditional transistor, and is the first excitonic transistor that can be activated by external light or via an auxiliary flow of excitons. This concept provides a method for a user to regulate the magnitude and direction of excitonic current in a chromophoric system. We show how to overcome the two potential challenges: short-lived second excited states and errors in the precision of orienting molecules relative to each other. Additionally, we show how a universal set of binary logic gates can be constructed, leading to universal classical computation. This new invention may lead to new developments in bioimaging, complex directed photocatalyis, phototherapeutics, and/or chemical sensing.
In the second portion of this dissertation, we developed two parallel software packages for the simulation of quantum algorithms. Such simulation packages are essential for the numerical study of quantum algorithms, since a viable large-scale quantum computer has not yet been built. The first package is called the Quantum High-Performance Software Testing Environment (qHiPSTER), which simulates the full Hilbert space of the quantum circuit's state. We simulated 40 qubits, which was the largest simulation of its kind at the time.
The second quantum computation software package, the Quantum TensOR Contraction Handler (qTorch), implements a tensor network contraction algorithm. Using the Quantum Approximate Optimization Algorithm (QAOA)/Max-Cut and the Hubbard model as test cases, we found heuristic trends that help a user determine which circuits will benefit most from a tensor network simulation approach.
Finally, we investigated the effects of noise on chemical state preparation in quantum algorithms. Chemical state preparation is necessary for both of the main quantum algorithmic approaches to chemistry: the variational quantum eigensolver and the quantum phase estimation algorithm. Trends were elucidated with respect to mapping types (Jordan-Wigner versus Bravyi-Kitaev), noise types, and chemical properties.
We end with a short discussion of the future of rational excitonic design and classical simulation of quantum phenomena.
Citable link to this pagehttp://nrs.harvard.edu/urn-3:HUL.InstRepos:42014961
- FAS Theses and Dissertations