Ninth International Conference on Mars 2019 (LPI Contrib. No. 2089) 6339.pdf

SOLAR CELL PRODUCTION CAPACITY ON THE MARTIAN SURFACE. A.J. Abel,†,1,2 A. J. Berliner,†,1,3 M. Mirkovic,1,4 W. D. Collins,5,6 A. P. Arkin,1,3,7 and D. S. Clark1,2,8 †equal contribution 1Center for the Utilization of Biological Engineering in Space (CUBES), 2Department of Chemical and Biomolecular Engineering, University of California, Berkeley, CA 94720, USA, 3Department of Bioengineering, University of California, Berkeley, CA 94720, USA, 4Department of Electrical Engineering and Computer Sciences, University of California, Berkeley, CA 94720, USA, 5Climate and Ecosystems Sciences Division, Lawrence Berkeley National Laboratory, One Cyclotron Road, Berkeley, CA 94720, USA, 6Department of Earth and Planetary Sciences, University of California, Berkeley, CA 94720, USA, 7Environmental Ge- nomics and Systems Biology Division, Lawrence Berkeley National Laboratory, One Cyclotron Road, Berkeley, 94720, CA, USA, 8Molecular Biophysics and Integrated Bioimaging Division, Lawrence Berkeley National Laboratory, One Cyclotron Road, Berke- ley, CA 94720, USA

Introduction: Photovoltaic (PV) and photoelectro- calculated according to the assumptions of the detailed chemical (PEC) devices could be used to produce elec- balance model: all photons with energy greater than the tricity and commodity chemicals necessary to sustain absorption threshold (band gap) are absorbed, the quasi- human life on Mars using in situ resources. Because life Fermi level separation is constant and equal to V across support systems require robust operation, understanding the device (infinite carrier mobility), and only carriers device performance is critical to determining the feasi- recombine only via radiative recombination [4]. For bility of solar-powered manufacturing at candidate PEC devices, the solar cell must achieve a minimum landing sites. Solar cell device production limits are not voltage greater than Eredox + VO, where Eredox is the redox well characterized on the Martian surface, mainly due to potential of the chemical reaction driven by the cell, and differences in the surface temperature and solar inten- VO is the overvoltage losses associated with electrode sity and spectrum from typical conditions on Earth or in kinetics and solution polarization in the electrochemical space. Here, we integrate relevant climate data from the reactor. Mars Climate Database [1] with a radiative transfer Sunlight incident on the surface (Psun) is mediated model, libRadtran [2], to predict spectrally-resolved so- by atmospheric particles including gases, ice, and dust. lar flux across the Martian surface over the course of a Using the Mars Climate Database, we track the concen- Martian year. We use this result to inform detailed bal- tration and particle size of these species across the Mar- ance calculations for PV and PEC devices producing tian surface for a variety of climate scenarios, including electricity, hydrogen, ammonia, and acetic acid. These average and heavy dust conditions corresponding to a products were selected to represent simple precursor planet-wide dust storm. We calculate the wavelength- chemicals necessary to support power systems, agricul- resolved optical depth, given by: ture, and manufacturing. We determine optimal solar (3) 휏(휆) = 휏푔(휆) + 휏푖(휆) + 휏푑(휆) device performance and best-possible production rates using libRadtran software which considers both absorp- across the Martian surface and provide design guide- tion due to molecular composition and scattering events lines for semiconductor band gap engineering. Finally, approximated by Rayleigh or Mie equations depending we compare production capacity to estimated demand on the effective particle size. from established reference mission architectures to de- Results: Figure 1 shows the and limiting efficiency termine the solar cell array size necessary to support a of a two-gap tandem PV (a) and PEC (b) device using six-person mission. Our results compare favorably to the average solar flux at noon at Jezero Crater over the both established and speculative power technologies. course of an average Martian year. The PEC device is Methods: We use a detailed balance model [3] to calculated assuming H2 production from water splitting calculate the power conversion efficiency of single gap with 700 mV overvoltage, which is typical for state-of- and two- and three-gap tandem solar conversion de- the-art systems [5]. Optimal band gaps are 5–10% from vices. Power conversion efficiency is written as: those for terrestrial systems [4,5], emphasizing the im- 퐽∙푉 (1) 휂 = portance of responsive design to Mars. 푃푉 푃 푠푢푛 We determine optimal band gaps for best-possible for a PV device, and: 퐽∙퐸푟푒푑표푥 power and chemical production by integrating device (2) 휂푃퐸퐶 = 푃푠푢푛 performance over the course of a Martian year, and for a PEC device, where J is the current density, V is the compare this result to demand estimated from estab- operating voltage, and Psun is the solar intensity incident lished six-person reference mission architectures. From on the device surface. The current density, J, is this analysis, we show that solar cells can readily Ninth International Conference on Mars 2019 (LPI Contrib. No. 2089) 6339.pdf

support life support systems for human missions to Mars. For example, at Jezero Crater, we estimate that a ~2800 m2 solar array (7–8 tons) would meet demand with an additional 50% tolerance for power, represent- ing only 2–8% of the total estimated mass required for a human mission to Mars.

Figure 1. Power conversion efficiency for PV (A) and PEC (B) devices. The PEC device produces H2 in a wa- ter splitting configuration with 700 mV overvoltage. We use the solar flux at noon at Jezero Crater averaged over the course of a year as the reference solar spectrum.

References: [1] Forget F. et al. (1999) J. Geophys. Res. 104, 24155-24176. [2] Emde C. et al. (2016) Geo- scientific Model Devel. 9, 1647-1672. [3] Shockley W. and Queisser H. J. (1961) J. Appl. Phys. 32, 510-519. [4] Hanna M. C. and Nozik A. J. (2006) J. Appl. Phys. 100, 074510. [5] Doscher H. et al. (2014) Energy Envi- ron. Sci. 7, 2951-2956.