International Journal of Energy Research, Vol.43, No.9, 4370-4389, 2019
Planar and textured surface optimization for a tritium-based betavoltaic nuclear battery
Remote, terrestrial, and space sensors require sources that have high enough power and energy densities for continuous operation for multiple decades. Conventional chemical sources have lower energy densities and lifetimes of 10 to 15 years depending on environmental conditions. Betavoltaic (beta V) nuclear batteries using beta(-)-emitting radioisotopes possess energy densities approximately 1000 times greater than conventional chemical sources. Their electrical power density (P-e,P-vol in W/cm(3)) in a given volume is a function of beta(-)-flux surface power density )" open="(">P beta-, surface interface type between radioisotope and transducer, beta(-) range, and transducer thickness and conversion efficiency (eta(s)). Tritium is the most viable beta(-)-emitting radioisotope because of its commercial availability, low biotoxicity, half-life, and low energy, which minimizes the penetration depth and damage of transducer. To maximize P-e,P-vol, tritium in solid or liquid form must be used in the beta V nuclear battery. A Monte Carlo source model using MCNP6 was developed to maximize the P-e,P-vol of a tritium-based beta V nuclear battery. First, a planar coupling configuration with different tritiated compounds (ie, titanium tritide and tritiated nitroxide) and a semiconductor transducer (4H-SiC) with thicknesses of 1 and 100 mu m were modeled. The results showed that beta(-)-source efficiency (eta(beta)), which is the percentage of energy deposited in the transducer, decreased as the tritiated compound's mass density increased. The highest P-e,P-vol was dependent on a combination of characteristics: specific activity (A(m) in Ci/g), mass density, and 4H-SiC layer thickness. The tritiated nitroxide with the highest A(m) at 2372 Ci/g produced the highest P-e,P-vol at 2.46 mW/cm(3). Second, a 3-D coupling configuration was modelled to increase surface interfacing between the radioisotope source and textured transducer surface. 3-D coupling configuration increased the percentage of energy deposited into the transducer because of more surface interfacing between the transducer and source in the same volume. The tritiated nitroxide was selected as the radioisotope source coupled with five different textured surface feature types. The P-e,P-vol as a function of textured surface feature and gap, where the radioisotope is located, width was calculated for 1- and 100-mu m 4H-SiC layer thicknesses. Results showed that eta(beta) increased compared with planar coupling configuration (ie, approximately 56.2% increase over planar with cylindrical hole array) except with the rectangular pillar array. Still, the rectangular pillar array produced the highest P-e,P-vol at 4.54 mW/cm(3) with an increasing factor of 2.29 compared with the planar coupling configuration.