The reason FeS$_2$ is not a good PV absorber

FeS$_2$ is representative of an ideal earth-abundant candidate absorber for thin film PV, because of its appropriate band gap, high absorption coefficient and good electron/hole mobility. Yet, despite $\sim$15 years of research, the promise of FeS$_2$ as an absorber layer has been unrealized, manifesting as a low open circuit voltage which has been attributed to E$_f$ pinning arising from bulk sulfur vacancies. Our first-principles calculations and experimental thermogravimetirc analyses, however, show that S vacancies and other point defects have rather high formation energies. Hence, they are unlikely to form and pin E$_f$. We find that the widely observed S deficiency in FeS$_2$ is accommodated by phase-coexistence of a few Fe$_{1-x}$S compounds, rather than S vacancies. These minority phases are metallic and detrimental for PV. We find select ternary Fe sulfides do not have thermodynamically-mandated phase-coexistence like FeS$_2$, yet they retain optimal band gaps and high absorption strengths comparable to FeS$_2$. These properties and associated surface-defect calculations will be discussed.