Study On The Performance Of GaN-based Betavoltaic Battery

In this paper,Monte Carlo software and COMSOL Multiphysics software are used to study the ⁶³Ni GaN based betavoltaic battery（BV battery）.The research content focuses on the self-absorption effect of radioactive sources,the transport behavior of carriers in semiconductor devices and the transport process ofβparticles in semiconductor materials,the influencing factors of nuclear battery output performance and the optimization of battery structure parameters.Firstly,the self-absorption law of ⁶³Ni was studied.In this paper,MCNP is used to simulate theβspectrum,surface emission power density and surface emission activity density of 100 m Ci ⁶³Ni.The calculation results show that as the thickness increases,the surface emission power density and activity density increase first and then reach saturation,and the peak position of theβenergy spectrum gradually moves to the high energy region.The optimal ⁶³Ni source thickness is determined to be 2μm.Furthermore,the variation trend of ⁶³Ni source energy spectrum,surface emission energy density and activity density with time is analyzed.When the time increases from 0 to 200 years,the probability ofβparticle emission decreases,and the surface emission power density and surface emission activity density also decrease exponentially with time.Secondly,the output parameters of GaN based p-n junction BV battery are calculated and its structure is optimized.The variation of total energy deposition power density and cumulative energy deposition rate of ⁶³Ni source in GaN material with depth was determined by MCNP software.The optimization model of GaN based p-n junction BV battery is proposed,and the influence of its structural parameters（doping concentration and junction depth）on the output performance of the BV battery is analyzed.The calculation results show that the lower the doping concentration in the p-region,the higher the doping concentration in the n-region,which will lead to higher open-circuit voltage.The smaller junction depth is more beneficial for the separation and collection of electron-hole pairs.Furthermore,the variation of total energy deposition power density and cumulative energy deposition rate with time is simulated,and the formula of the maximum output power density of the battery with time is further fitted,which can be used to predict the output power of ⁶³Ni GaN based p-n junction BV nuclear battery within 200 years.Taking 50 years as an example,the output power of the battery is predicted to reach 0.243μW·cm^-2.This has guiding significance for the actual service life of nuclear batteries.Finally,the finite element analysis software COMSOL Multiphysics was used to compare the output performance of GaN based BV batteries with p-n junction and p-i-n junction under the irradiation of 100 m Ci ⁶³Ni source,and the structural parameters of the two different cells were optimized.MCNP software is used to simulate ⁶³Ni source in GaN.The energy deposition distribution in the semiconductor is used to calculate the electron-hole pair generation rate,which is further used as an input parameter for subsequent COMSOL Multiphysics simulation.The output performance with the thickness of each region is analyzed.And the change of doping concentration.Under the same structural parameters,the p-i-n junction BV battery has a larger width of the depletion layer due to the intrinsic GaN layer,and the maximum output power is 33.4%higher than that of the p-n junction.The maximum output power density of the p-i-n junction BV battery is 0.167μW·cm^-2.Then,the structural parameters of these two structures in this study were optimized.The maximum output power density of the optimized p-n junction BV battery can be increased by 64%compared with that before optimization.The maximum output power density of the optimized p-i-n junction BV battery is 21.6%higher than that before optimization.The p-i-n junction is a promising transducer structure.