Answers are grounded in this paper and cite their sources.
Solar-cell simulation is now routinely used to propose new device architectures, passivation routes and high-efficiency designs, yet the credibility of many simulation-based claims remains difficult to judge. A fitted J–V curve or an optimized power-conversion efficiency is often treated as validation, even when key assumptions—defect densities, contact barriers, recombination pathways, optical constants, ion motion, solver settings and fitted parameters—are weakly traceable. This review addresses that gap by treating simulation credibility itself as the object of analysis. Rather than ranking software packages or summarizing optimization trends, it reframes the problem around a more consequential issue: the evidence required for a solar-cell simulation to support a scientific claim. Evidence is synthesized across crystalline silicon, CdTe, CIGS, kesterite, perovskite, organic, dye-sensitized and tandem photovoltaics through six lenses: tool–physics fitness, parameter provenance, validation depth, identifiability and uncertainty, reproducible reporting, and benchmark consistency. The selected examples illustrate how simulated performance claims can exceed the evidence supplied to defend them; this purposive review does not estimate the frequency of such practice. To address this gap, the review develops a claim-specific credibility framework that separates exploratory screening, calibrated reproduction, validated explanation and predictive modelling. It further proposes reviewer-oriented requirements for traceable inputs, multi-observable validation, uncertainty-aware outputs, FAIR-style reporting, benchmark-device packages and a Simulation Credibility Index. The central message is that credible simulation should be judged by the strength of its evidence chain, not by the software name, curve-fit quality or predicted efficiency alone. The SCI is an optional ordinal evidence summary, not a probability or acceptance threshold; critical deficiencies override its mean. Quantitative assessment procedures and four documentary applications make the workflow inspectable, while improved peer-review agreement remains untested.
No takes yet. Share an insight, caveat, or question.
Khan et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: