Computational materials discovery offers a way to screen materials for useful optical and electronic properties before committing to experimental fabrication. In a set of first-principles studies using density functional theory (DFT), researchers examined two-dimensional materials, perovskite interfaces and magnetic compounds as candidates for photovoltaics, photocatalysis and spintronics. The work also shows how variables such as mechanical strain and layer structure can change material behavior—an important consideration when translating material concepts into future device designs.
Using DFT to evaluate candidate materials
DFT calculations estimate how electrons behave in a material from its atomic structure. In these studies, calculations were used to assess properties including structural stability, band gaps, light absorption, mechanical response and electronic structure. These are screening tools: a predicted property can help identify promising candidates, but calculations alone do not establish that a material can be manufactured reliably or will perform as a finished device.
Several of the investigated materials are atomically thin, or two-dimensional (2D). Their reduced thickness and electronic structure make them of interest for light-harvesting and electronic applications. The calculations considered whether layers could be separated from bulk crystals, whether predicted structures were dynamically stable, and how their properties changed with strain or layer count.
Computational materials discovery for photovoltaic candidates
Three groups of 2D compounds were evaluated as potential light absorbers: sodium phosphidostannate (NaSnP), palladium phosphorus sulfide (PdPS) and technetium dichalcogenides (TcS2 and TcSe2). Their calculated optical or electronic gaps were considered in relation to photovoltaic use, alongside their predicted stability and response to mechanical changes.
For monolayer NaSnP, the calculated cleavage energy was comparable to values reported for layered materials such as graphite and molybdenum disulfide, suggesting that exfoliation from bulk could be feasible. Hybrid-functional calculations gave a band gap close to the value considered desirable for solar-cell materials. The study also predicted that moderate compression could change the gap from indirect to direct, while greater compression could make the layer metallic. This illustrates why a material’s performance cannot be judged by composition alone: mechanical conditions may alter its electronic behavior.
The PdPS monolayer was calculated to have an optical gap of 1.65 eV and active edge sites for the hydrogen evolution reaction (HER), a reaction relevant to water splitting. Vibrational calculations indicated dynamic stability, and the study proposed mechanical cleavage as a possible route to obtaining the monolayer. These results make PdPS a computationally identified candidate for both light harvesting and catalysis, not a demonstrated commercial device material.
For TcS2 and TcSe2, the predicted monolayers showed band gaps and visible-light absorption considered promising for energy applications. The calculations also indicated that TcSe2 band-edge positions could support water splitting. Their electronic character was predicted to change under tensile or compressive strain, while bilayer TcS2 was found to have direct-gap character. Such layer- and strain-dependent behavior may inform which material configurations warrant further investigation.
Engineering perovskite interfaces and strain
Perovskite solar materials present a different challenge: their optical properties are attractive, but moisture-related stability is a concern. DFT and ab initio molecular dynamics simulations examined methylammonium lead triiodide (MAPbI3) with a graphene layer at the interface. The calculations predicted charge transfer between graphene and the perovskite, enhanced light absorption compared with pristine MAPbI3, and resistance to reaction with a water molecule in the simulated structure.
This is a modeled interface, not evidence of long-term device durability in operating conditions. Still, it suggests a design question for materials engineers: can a protective layer also influence charge separation and optical response? The answer depends on interface structure and must ultimately be tested experimentally.
Hydrostatic strain was also investigated as a means of tuning methylammonium lead and tin iodide perovskites. In the calculations, compression narrowed their band gaps and shifted the absorption edge toward lower-energy light. Strain was also predicted to reduce electron effective mass, a property associated with carrier transport. The findings identify mechanical state as a potential design variable, while leaving practical implementation and device-level outcomes to further study.
Dirac materials and spin-dependent behavior
The research extended beyond photovoltaics to electronic and spintronic materials. Structure-prediction calculations identified several dynamically stable boron hydride sheets. Some were predicted to host Dirac cones, where electronic bands cross with a linear dispersion; one structure showed a Dirac ring. These predicted features make the sheets of interest for nanoelectronics, although synthesis and experimental characterization remain necessary steps.
Calculations on manganese fluoride (MnF3) and rhombohedral lanthanum manganite (LaMnO3) identified spin-polarized Dirac behavior. In the predicted band structures, one spin orientation had Dirac-like crossings while the other behaved as a semiconductor with a large gap. This contrast is relevant to spintronics, which uses electron spin as well as charge, but the computational predictions do not by themselves demonstrate a working spintronic component.
What the findings mean for engineering
For engineers and technical decision-makers, the central lesson is that computational screening can help organize early-stage materials research around measurable questions:
- Is the proposed structure stable, and could it plausibly be separated into a thin layer?
- Does its calculated band structure support the intended optical or electronic function?
- Can strain, layer count or an interface alter that function in a useful way?
- Which predictions require experimental validation before device development?
Across these studies, DFT points to candidate materials and design levers rather than ready-to-deploy technologies. The practical value lies in narrowing the field of investigation and clarifying which structural and electronic properties should be tested next. For photovoltaic, catalytic and electronic applications, pairing computational predictions with fabrication and measurement will be essential to determine whether these promising concepts can meet real engineering requirements.