Intersection of Quantum Computing and Machine Learning

Quantum computing is an emerging paradigm of computing for computationally intensive problems which are currently intractable on classical computing platforms. It is believed that within the next decade, quantum computing will have disruptive impact in many areas of development including materials design, data science and machine learning, health, climate science, and combinatorial optimisation. Among these, machine learning is widely considered as the recipient of early quantum advantage. 

In this project, we are interested to explore the intersection of quantum computing and machine learning. We believe that classical machine learning could help in design and characterisation of quantum processors, in particular learning quantum noise in-situ and performing error mitigation by innovative quantum control. On the other hand, quantum machine learning algorithms can help to develop new methods with superior computational efficiency and/or enhanced accuracy and robustness. 

Quantum Hardware Design and Scalability

In this project, we focus on emerging topics of research in quantum technologies namely integrated spin qubit devices, spin-photon interfaces and single photo emitters and detectors. We develop and apply high-end computational methods to tackle key challenges both at the level of fundamental qubit material design as well as at the quantum device level where qubit control, couplings and scale-up protocols are important parameters of interest. Our group has extensive expertise in spin qubit design based on silicon material system which is one of the most widely used material system in the microelectronic industry. Silicon has been a workhorse for computing devices over the last many decades and spin qubits in silicon could leverage from many decades of advancements in fabrication and device design. Solid state spin qubits in silicon are one of the frontrunner candidates for quantum computing devices and related technologies due to the associated very long coherence times (of the order of seconds!) and promising pathways for scalability to large-scale quantum architectures which are crucial to implement error correction. 

We are one of the leading groups in Australia and have made significant contributions to advance the Australian national quantum hardware development effort through the ARC Center of Excellence CQC2T platform. Our team has developed a multi-million-atom electronic structure tool which is capable of providing unprecedented understanding of solid-state qubit spin physics. Our work in the last few years includes new proposals for the design of high-fidelity two-qubit quantum gates, the world’s first spatial metrology of spin qubits with single atom precision, spatially resolved mapping of coupled qubit interactions, a framework for autonomous characterisation of qubits by machine learning and a proposal for exchange-based scalable quantum computer architecture. Building upon this extensive experience of working with solid-state spin environment and expertise in design of quantum computing devices, this project will continue the development of new protocols for the implementation of quantum control, spin-photon coupling and scalable quantum computer architectures.

Metal-Organic Frameworks (Custom-build 2D Materials)

Two-dimensional metal-organic frameworks (2D-MOFs) are highly versatile material system which can be custom-build for the next generation technologies such as organic topological insulators for dissipationless electronics and 2D magnetic materials. One of the key goals of this project is to investigate long-range magnetic ordering in 2D-MOF structures. Elucidating such magnetic properties can be important for understanding possible nontrivial topological electronic properties of these systems, since the latter can result from magnetic phenomena but also be the cause of well-defined spin structure. The presence of such spin structure in 2D materials provides an ideal platform for the realisation of 2D magnets, with applications including in sensing and hard-disk data storage. 2D-MOF structures can be engineered to implement 2D magnets by the selection of metal ions and the selection of structural morphologies which is the subject of our ongoing research. 

Due to the availability of a large number of building blocks (molecules and metal atoms) as well as several possibilities to assemble 2D materials by bottom-up approach, the 2D-MOFs form a tremendously large design space which is impractical to explore via direct quantum mechanical simulations based on DFT theory. Alternatively, a machine learning framework, carefully trained by carrying out the rigorous theoretical analysis of a set of selected materials, can rapidly screen a large database of unknown materials to reliably identify those of the same character at only a fraction of the computational cost. In this project, our team is building a machine learning framework which will utilize advanced techniques such as quantitative structure-property relationship (QSPR) to establish a correlation between the electronic and spin properties of 2D-MOFs with their composition, structure, and symmetry. To enable accurate and fast learning, the QSPR analysis will be conducted by employing a variety of advanced algorithms such as linear regression analysis, decision tree regression, and non-linear support vector machines. The detailed insights obtained by QSPR will train an artificial neural network for high-throughput screening and computer-aided design of 2D-MOF TIs. 

Machine Learning Design of Materials and Devices

Discovery of custom-designed nanomaterials is often an expensive pursuit, which typically relies on multiple trial and error experiments, followed by energy draining manufacturing processes. Even with access to supercomputers, traditional DFT or  tight-binding methods takes many CPU hours to execute even a few set of simulations, leaving the exploration of a vast parameter space nearly a formidable task. Contrarily, machine learning approaches could provide reliable, robust, autonomous, and time efficient discovery of new materials at only a fractional computational cost. In this project, our vision is to integrate state-of-the-art multi-million-atom electronic structure and transport calculations with advanced machine learning tools which can form a powerful theoretical framework capable of designing nanomaterials with target functionalities. Our focus is to advance emerging photonic and quantum technologies which are anticipated to drive Industry 4.0 revolution.

Self-Assembled Quantum Dots, Quantum Dot Molecules, Quantum Dot Stacks

Self-assembled In(Ga)As/GaAs quantum dots are a promising solid-state system, and are widely employed for the design of a variety of optoelectronic devices and quantum information applications. Based on multi-million-atom simulations, we provide an in-depth understanding of their electronic and optical properties, and perform engineering of related geometry parameters for implementation of devices with tailored functionalities. We have investigated both single quantum dots, as well as large stacks of strongly-coupled quantum dots.