Speaker
Description
Betavoltaic energy conversion uses a semiconductor device to convert beta radiation (high-energy electrons) from a radioisotope to generate electrical power. Current challenges include designing semiconductor structures that minimize backscattered electrons, non-radiative carrier recombination, and maximize minority carrier diffusion lengths. Diamond is a promising material to tackle these challenges. In this work, we numerically modeled a diamond betavoltaic cell to convert beta radiation into electricity. We optimized the thickness and doping of the diamond structure to maximize the output power.
Our numerical model uses Monte Carlo simulations to quantify energy deposition profiles as a function of depth in diamond. A custom software converts the energy density profile into an electron-hole pair generation rate profile. The generation rate profiles are sent to the software Sentaurus to assess the diamond betavoltaic cell performance under beta radiation by solving Poisson, drift, and diffusion equations. The model includes series and parallel resistances, doping-dependent carrier mobility, Schottky-Read-Hall recombination, and Auger recombination. Validation against experimental data demonstrates that our numerical model reproduces betavoltaic conversion mechanisms with high fidelity. We use a custom high-dimensionality particle swarm optimization algorithm to find optimal layer thicknesses and doping for maximum power output.
Results show that diamond is suited to maximize energy absorption through reduced backscattered electrons, due to its lower atomic number and lower density compared to other semiconductors. Diamond can absorb 5% more energy than SiC, 18% more than GaN, and 25% more than GaAs. Optimal diamond structure enables a carrier collection that exceeds 95% over a wide range of beta particle energies. High carrier collection can be explained by large minority carrier diffusion lengths, which also limit bulk recombination. The primary carrier-collection losses are attributed to surface recombination, which is exacerbated by the absence of a wider-bandgap passivating semiconductor on the device surface. Optimal devices have a conversion efficiency of 30%, outperforming GaAs betavoltaic cells by an order of magnitude. A tolerance study demonstrates that significant variations in thickness and doping induce a maximum power reduction of at most 5%.
| Keyword-1 | Diamond |
|---|---|
| Keyword-2 | Betavoltaics |
| Keyword-3 | Numerical modeling |