Computational Materials Diagnostics and Optimization of Photoelectrochemical Devices
LaboratoryLawrence Livermore National Laboratory (LLNL)
Capability ExpertJoel Varley, Tadashi Ogitsu
ClassComputational Tools and Modeling
Node Readiness Category1: Low-Temperature Electrolysis (LTE)
1: Photoelectrochemical (PEC)
2: Solar Thermochemical (STCH)
This capability provides a computational procedure for diagnosing sources of discrepancy between idealized PEC device behavior and observed performance, as well as identifying optimal synthesis and processing conditions for component materials. Based on ab-initio density functional theory (DFT) simulations using advanced hybrid functionals, our procedure examines the optoelectronic consequences of point defects and how doping and alloying can be leveraged to optimize specific properties such as electronic conductivity, band gap and resulting optical absorption, and band edge positions that facilitate charge transport. We also provide thermodynamic stability and defect analyses for evaluating the practicality of candidate materials. We have applied this methodology to optimize the following components relevant to PEC devices: (1) optical absorber layers, (2) buffers and window layers for charge transport and electrochemical stabilization, and (3) catalysts. Our approach is highly complementary to experimental characterization techniques including photoemission and absorption spectroscopies and provides a material design strategy that is more comprehensive than conventional idealized modeling. The experts are funded by EERE/FCTO to apply this methodology to chalcopyrite-based PEC devices, and have years of experience in defect analysis and materials screening.
Simulation cost (as well as required amount of human labor) depends on the complexity of the compound material (e.g., number of elements in the alloy). Computations do not explicitly consider kinetics.
This capability goes beyond conventional idealized bulk calculations to consider materials compatibility at the device level, and can be used as a diagnostic and engineering tool for identifying suitable synthesis and processing conditions. The Materials Science Division at LLNL has spent the past several years building a specialized focus on "real" materials modeling, which includes development of this computational diagnostic capability for a wide range of applications. The NNSA high-performance computing facilities at LLNL can facilitate simulations of a far broader and more complex materials space than would be achieved using conventional supercomputers.
The general procedure and computational techniques are mature and ready for immediate application. The experts are currently involved in an EERE/FCTO project understanding defect chemistry and engineering buried interfaces in PEC materials. The simulations can make use of leadership-class computing facilities at LLNL.
The capability can be used to guide synthesis of PEC materials towards satisfying performance, durability, and cost targets. It can inform thermodynamic stability conditions and diagnose deviations of material properties due to defects favored under a given synthesis condition, and can help to identify compatible materials combinations at the device level.
Calculated deep levels of native defects in the buffer layer in combination with calculated band alignment in a conventional CIGSe thin-film solar cell stack (a) and a calculated phase diagram for the earth-abundant absorber material Cu2ZnSnS4 showing the region of stability for Cu-rich conditions (b).
1. J. B. Varley and V. Lordi, "Intermixing at the absorber-buffer layer interface in thin-film solar cells: The electronic effects of point defects in Cu(In,Ga)(Se,S)2 and Cu2ZnSn(Se,S)4 devices," J. Appl. Phys. 116, 063505 (2014).
2. H. Peelaers, D. Steiauf, J. B. Varley, and A. Janotti, "(InxGa1− x)2O3 alloys for transparent electronics," Phys. Rev. B 92, 085206 (2015).
3. H. H. Hansen, J. B. Varley, A. A. Peterson, and J. K. Norskov, "Understanding trends in the electrocatalytic activity of metals and enzymes for CO2 reduction to CO," J. Phys. Chem. Lett. 4, 388 (2013).