Computer-Designed Solar Molecules Point to 28 Percent Efficiency

A team of computational chemists has designed a family of eight new organic molecules that, according to detailed quantum mechanical simulations, could push the power conversion efficiency of solar cells to a predicted 28 percent—a figure that would dwarf the 22 percent benchmark set by the experimental molecule on which the designs are based. The study, published in the Journal of the Saudi Chemical Society, was led by Muhammad Usman Khan of the University of Okara in Pakistan, together with colleagues at institutions in Pakistan, the United Arab Emirates, New Zealand and Saudi Arabia. Rather than synthesizing candidate after candidate in the laboratory, the researchers used density functional theory, or DFT, to screen molecular architectures on a computer first, a strategy that is rapidly becoming the fastest route to next-generation photovoltaic materials.

The starting point for the work was an experimentally validated molecule called ThPCyAc, synthesized in 2024 by Jianlin Chen and colleagues. ThPCyAc is a multifunctional dye molecule built around a triphenylamine skeleton, with a 2-cyanoacrylic acid group attached through a thiophene bridge. When used as a self-assembled hole transport layer in inverted perovskite solar cells, it achieved a power conversion efficiency above 20 percent. That real-world performance made it an ideal reference scaffold: the Pakistani-led team kept its donor core and thiophene pi-spacer intact and systematically swapped out the terminal acceptor group with eight different electron-withdrawing units, generating a new series labeled Thy1 through Thy8.

The logic behind this donor–pi–acceptor, or D–pi–A, architecture is central to modern organic photovoltaics. In such molecules, the donor fragment—here a combination of triphenylamine and carbazole—donates electron density, the thiophene bridge conducts it, and the acceptor unit pulls it in the opposite direction. This push-pull arrangement promotes intramolecular charge transfer, narrows the electronic band gap and broadens the absorption spectrum, allowing the material to harvest more of the sun’s output. By fine-tuning only the acceptor end, the designers could adjust the energy levels of the frontier molecular orbitals—the highest occupied molecular orbital, or HOMO, and the lowest unoccupied molecular orbital, or LUMO—without disturbing the parts of the molecule already proven to work.

Before any predictions could be trusted, the team had to validate their computational method. They benchmarked six different DFT functionals—B3LYP, CAM-B3LYP, M06, M062X, ωB97XD and MPW1PW91—each paired with the 6-31G(d,p) basis set, against the experimentally measured absorption maximum of the reference molecule in both the gas phase and dichloromethane solvent. The B3LYP functional reproduced the experimental values most closely, deviating by only 17 nanometers in the gas phase and 20 nanometers in solution, so it was selected for all subsequent calculations. Time-dependent DFT with a conductor-like polarizable continuum model then supplied the absorption spectra, while Marcus theory was applied to estimate the reorganization energies that govern how easily holes and electrons hop through the material.

The results were striking across the board. The designed molecules showed HOMO–LUMO energy gaps between 0.48 and 1.02 electron volts, far narrower than the 2.37 electron volt gap of the reference compound. Thy3, the standout performer, achieved the smallest gap of just 0.48 electron volts, thanks to a highly electron-withdrawing acceptor bearing carbonyl and malononitrile groups with extended conjugation. Narrower gaps translate directly into wider absorption: the new molecules were predicted to absorb light at wavelengths up to 799 nanometers in the gas phase and up to 837 nanometers in solution, compared with 397 and 546 nanometers for the reference. That red-shift pushes the materials deep into the near-infrared, a region most conventional solar absorbers leave untapped.

Charge dynamics looked equally promising. Exciton binding energies—the energy needed to tear a bound electron–hole pair apart—fell between 0.16 and 0.26 electron volts for the best candidates, lower than the reference values of 0.29 to 0.33 electron volts, meaning excitons should dissociate into free charges more readily. Transition density matrix analysis showed electron density flowing cleanly from the donor core through the thiophene bridge into the acceptor units, while density-of-states calculations confirmed that the HOMO levels were concentrated on the triphenylamine and carbazole donors and the LUMO levels on the acceptors—the textbook signature of efficient intramolecular charge transfer. Molecular electrostatic potential maps and natural population analysis reinforced the picture, revealing a clear separation of electropositive and electronegative regions across every designed molecule.

When paired with the standard fullerene acceptor PC70BM, the designed donors delivered open-circuit voltages between 0.40 and 0.54 volts, with Thy1 and Thy3 reaching 0.54 volts, above the reference molecule’s 0.47 volts. Fill factors—a measure of how much of a cell’s theoretical maximum power it actually delivers—were calculated at an exceptionally high 85 to 91 percent. Combining these figures with a short-circuit current density of 24.87 milliamperes per square centimeter taken from the reference molecule’s experimental data, the Scharber model yielded predicted power conversion efficiencies of up to 28 percent for Thy3. The authors are careful to note that these are theoretical upper-bound estimates: the model assumes ideal morphology and ignores recombination losses, charge trapping and interfacial resistance, and the fixed short-circuit current is an approximation. The numbers are best read as comparative indicators of relative potential, not guaranteed device performance.

Reorganization energies, which quantify how much a molecule must distort when it gains or loses an electron, painted a nuanced picture. The designed molecules showed hole reorganization energies of 0.0077 to 0.0092 atomic units and electron reorganization energies of 0.0058 to 0.1122 atomic units, with several candidates, including Thy2, Thy5, Thy6 and Thy7, beating the reference for electron mobility. Quantum chemical reactivity indices told a consistent story: the new molecules are softer and more chemically reactive than the reference, with higher ionization potentials and electron affinities confirming their role as donors. When Thy3 was blended with PC70BM in a simulated donor–acceptor interface, the HOMO density sat entirely on the donor and the LUMO density entirely on the acceptor, demonstrating complete charge transfer between the two molecules.

Why does this matter beyond one laboratory’s calculations? Perovskite solar cells have stormed the photovoltaic world because they combine strong light absorption, cheap processing and flexible fabrication, but their efficiency and stability hinge critically on the hole transport material that ferries positive charges out of the device. The workhorse material, spiro-OMeTAD, requires chemical doping that degrades stability, and few dopant-free alternatives have matched its performance. Triphenylamine-based small molecules offer a way out: they dissolve readily in organic solvents, resist oxidation, transport holes efficiently and remain stable over the long term. The Thy series, and Thy3 in particular, fits that profile while adding the deep near-infrared absorption that could lift current generation well beyond existing designs.

The study’s broader lesson is methodological as much as material. By anchoring a computational screen to an experimentally validated scaffold and validating the theoretical level against measured absorption data, the researchers produced predictions that synthetic chemists can pursue with reasonable confidence. The authors recommend Thy3—copolymerized into a donor material—for future solar cell development, citing its narrow band gap, favorable exciton binding energy, high predicted fill factor and strong charge mobility. If laboratory synthesis and device fabrication bear out even a fraction of the predicted gains, molecules designed entirely on a computer could soon be carrying the current in some of the most efficient perovskite solar cells ever built, a quiet vindication of the idea that the fastest path to better solar energy may begin not on a lab bench but in the equations of quantum chemistry.

Subject of Research: DFT-based rational design of triphenylamine–benzothiophene donor molecules for perovskite and organic solar cells

Article Title: Rational design of triphenylamine–benzothiophene-based donor materials for enhanced photovoltaic performance in solar cells: a DFT study

Article References: Rational design of triphenylamine–benzothiophene-based donor materials for enhanced photovoltaic performance in solar cells: a DFT study. (n.d.). https://doi.org/10.1007/s44442-026-00092-8

Image Credits: AI Generated

DOI: 10.1007/s44442-026-00092-8

Keywords: perovskite solar cells, density functional theory, triphenylamine, benzothiophene, hole transport materials, power conversion efficiency, donor–pi–acceptor, charge transfer, exciton binding energy, TD-DFT, organic photovoltaics, molecular engineering