Design and growth of wide-bandgap III-V solar cells on silicon by molecular beam epitaxy

Li, Brian Deyuan

Permalink

https://hdl.handle.net/2142/127178 Copy

Description

Title

Design and growth of wide-bandgap III-V solar cells on silicon by molecular beam epitaxy

Author(s)

Li, Brian Deyuan

Issue Date

2024-11-12

Director of Research (if dissertation) or Advisor (if thesis)

Lee, Minjoo

Doctoral Committee Chair(s)

Lee, Minjoo

Committee Member(s)

• Choquette, Kent
• Dallesasse, John
• Kim, Kyekyoon

Department of Study

Electrical & Computer Eng

Discipline

Electrical & Computer Engr

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

• Iii-v
• Mbe
• Gainp
• Gaasp
• Solar

Language

eng

Abstract

Solar energy based on photovoltaic (PV) solar cells is a fast-growing and promising technology to replace fossil fuels in meeting our global electricity demand, as well as a primary energy source for space applications. Silicon (Si) solar cells currently dominate PV production due to their high efficiency (27.1% record under 1-sun illumination) and low cost, but are rapidly reaching their efficiency limit of 29.4%, with little room for further improvement. Increased efficiency can enable lower total PV systems costs by reducing the need for components such as electrical and structural equipment, maintenance, etc. To-date, the highest efficiency solar cells are multi-junction solar cells (MJSC) based on III-V semiconductors with multiple sub-cells stacked from lowest to highest bandgap, with record efficiencies of 39.5% under 1-sun and 47.6% under concentrated light. Unfortunately, III-V MJSCs have orders-of-magnitude greater cost than Si cells, and a major contributor to their cost is growth on lattice-matched GaAs or Ge substrates. A promising approach for increased efficiency at low cost is to grow III-V sub-cells on a low-cost Si substrate for the bottom cell (III-V/Si), which has a theoretical efficiency of > 37% for two-junction cells and > 41% for triple-junction cells. The main drawback of epitaxial growth of III-V/Si MJSCs is the significant lattice mismatch (3-4%) of III-V semiconductors on Si for the materials of interest, causing the formation of extended defects, known as threading dislocations, that propagate through the III-V sub-cells and degrade performance. To date, the highest-efficiency III-V/Si MJSCs have obtained 1-sun efficiencies of 23.4-25.9%, below the record efficiency of a Si cell by itself. Significant research efforts in III-V-on-Si growth have enabled III-V buffers with threading dislocation density (TDD) of ~1.0×107 cm-2 and lower, compared to non-optimized TDD of > 1×109 cm-2. However, even at these lower TDD values, significant research in III-V solar cell growth and design is still needed to improve the efficiency of III-V/Si devices beyond Si alone. A particularly important requirement is to maximize the collection of photo-generated carriers, as described by quantum efficiency, which requires minimizing unwanted recombination losses in the solar cell bulk absorber and at the front and rear surfaces. In this dissertation, I focus on the design, growth and characterization of III-V sub-cells both lattice-matched on GaAs and lattice-mismatched on Si, with the goal of producing cells with superior efficiency for future III-V/Si MJSCs. All devices in this work were grown by molecular beam epitaxy (MBE), and the sub-cells of interest were 1.9 eV Ga0.51In0.49P cells and 1.7 eV GaAs0.77P0.23 cells, which have been used in all record epitaxial III-V/Si MJSCs to-date. For GaInP cells, I first report the use of Tellurium (Te) n-type dopant to obtain superior cell efficiency over Si doping, which can be a complementary or alternative strategy to our existing post-growth annealing approach to improve the quality of MBE-grown phosphides. I then performed a detailed comparative study of GaInP on Si cells with a traditional front-junction design (primarily p-type absorber) and the alternative rear-junction design (primarily n-type absorber) and showed higher material quality for the latter design that should enable improved efficiency of GaInP sub-cells on Si. For GaAsP cells, I performed a detailed design study, but faced significant challenges to further improvement, particularly in quantum efficiency, due to limited material quality in the p-type absorber layer. However, a follow-up study of AlGaAsP distributed Bragg reflectors (DBR) yielded promising results for an alternative path to high carrier collection by rear-surface reflection of photons into a thinned GaAsP cell, with a simulated ~5% boost in efficiency.

Graduation Semester

2024-12

Type of Resource

Thesis

Handle URL

https://hdl.handle.net/2142/127178

Copyright and License Information

Copyright 2024 Brian Li

Owning Collections