Publications

Due to their efficient molecular design, nonfullerene acceptors (NFAs) have significantly advanced organic photovoltaics (OPVs). However, the lack of models to screen and evaluate candidate NFAs based on the resulting device performance has impeded the rapid development of high-performance molecules. This work introduces a computational framework utilizing a kinetic Monte Carlo (kMC) model to derive device parameters from molecular properties computed through first principles. By analyzing the quantum chemical properties of diverse dimeric conformers, we estimate the relative probabilities of microscopic processes such as charge separation, recombination, and transport along with charge transfer state formation in the active layer of OPVs. These probabilities set up a random walk of charge carriers in a grid with disordered molecular sites, allowing us to track their average behavior and calculate key device parameters. Our model consistently predicts measured device parameters, including the short-circuit current and open-circuit voltage, for OPVs with diverse NFAs with high accuracy. Additionally, we applied the model to evaluate donor–acceptor combinations of known compounds and newly designed NFA molecules, identifying high-performing pairs. This model offers a computationally efficient approach for designing novel NFA molecules and optimizing the OPV performance.

Graph neural networks (GNN) have been demonstrated to correlate molecular structure with properties, enabling rapid evaluation of molecules for a given application. Molecular properties, including ground and excited states, are crucial to analyzing molecular behavior. However, while attention-based mechanisms and pooling methods have been optimized to accurately predict specific properties, no versatile models can predict diverse molecular properties. Here, we present graph neural networks that predict a wide range of properties with high accuracy. Model performance is high regardless of dataset size and origin. Further, we demonstrate an implementation of hierarchical pooling enabling high-accuracy prediction of excited state properties by effectively weighing aspects of features that correlate better with target properties. We show that graph attention networks consistently outperform convolution networks and linear regression, particularly for small dataset sizes. The graph attention model is more accurate than previous message-passing neural networks developed for the prediction of diverse molecular properties. Hence, the model is an efficient tool for screening and designing molecules for applications that require tuning multiple molecular properties.

Interlayers are known to enhance the performance of organic devices by reducing contact resistance, however, the details of the mechanism are uncertain. Models have correlated properties of interlayers to their extent of reduction of contact resistance, but a universal parameter correlating the microscopic phenomenon to device characteristics is yet to be established. Here, we demonstrate that the energy-level modulation at the interface of interlayer functionalized electrode and organic semiconductor, combined with the charge transfer integral between them, determines the extent of the reduction of contact resistance. Moreover, the rate of charge transfer calculated from these quantities is demonstrated to be a universal parameter predicting the characteristics of devices with functionalized electrodes, regardless of the nature of the semiconductor (p - or n-type). These observations explain the mechanism of interlayers and provide a computational model capable of selecting interlayers leading to high-performing devices.

Contact resistance (R_C) in organic devices originates from a mismatch in energy levels between injecting electrodes and organic semiconductors (OSCs). However, the microscopic effects governing charge transfer between electrodes and the OSCs have not been analyzed in detail. We fabricated transistors with different OSCs (PTB7, PCDTBT, and PTB7–Th) and electrodes (MoO₃, Au, and Ag) and measured their contact resistance. Regardless of the electrodes, devices with PTB7–Th exhibit the lowest values of R_C. To explain the trends observed, first-principles computations were performed on contact interfaces based on the projector operator diabatization method. Our results revealed that differences in energy levels and the electronic couplings between OSCs’ highest occupied molecular orbitals and vacant states on the electrodes influence device R_C. Further, based on values obtained from the first-principles, the rate of charge transfer between OSCs and electrodes is calculated and found to correlate strongly with trends in R_C for devices with different OSCs. We thus show that device R_C is governed by the feasibility of charge transfer at the contact interface and hence determined by energy levels and electronic coupling among orbitals and states located on OSCs and electrodes.

Photoexcitation of noble metal nanoparticles creates surface plasmons which further decay to form energetic charge carriers. These charge carriers can initiate and/or accelerate various chemical processes at nanoparticle surfaces, although the efficiency of such processes remains low as a large fraction of these carriers recombine before they can reach the reaction sites. Thus efficient utilization of these charge carriers requires designing nanostructures that promote the separation of charges and their transport toward the reaction sites. Here we demonstrate that covalently bound surface-coating ligands with suitable orbital alignment can provide electron transport channels boosting hot electron extraction from a gold nanostructure leading to a huge enhancement in the rate of hydrogen evolution reaction (HER) under NIR excitation. A (p)Br-Ph-SH substituted gold nanoprism (AuTP) substrate produced ∼4500 fold more hydrogen compared to a pristine AuTP substrate under 808 nm excitation. Further experimental and theoretical studies on a series of substituted benzene-thiol bound AuTP substrates showed that the extent of the ligand-mediated HER enhancement depends not only on the polarity of the ligand but on the interfacial orbitals interactions.

Interlayers at electrode interfaces have been shown to reduce contact resistance in organic devices. However, there still needs to be more clarity regarding the role of microscopic properties of interlayer functionalized interfaces on device behavior. Here, we show that the impact of functionalized electrodes on device characteristics can be predicted by a few critical computationally derived parameters representing the interface charge distribution and orbital interactions. The significant influences of interfacial orbital interactions and charge distribution over device and interface properties are exhibited. Accordingly, a function is developed based on these parameters that capture their effect on the interface resistance. A strong correlation is observed, such that enhanced orbital interactions and reduced charge separation at the interface correspond to low resistance regardless of the individual molecules utilized as the interlayer. The charge distribution and orbital interactions vary with the molecular structure of the interlayer, allowing the tuning of device characteristics. Hence, the proposed function serves as a guideline for molecular design and selection for interlayers in organic devices.

Mono metal (Ni, Co)-substituted (in) and supported (on) CeO₂ catalysts were prepared by using solution combustion synthesis and formaldehyde reduction methods. The catalysts were completely characterized by both bulk and surface techniques. Both supported and substituted catalysts show distinct differences in the dry reforming of methane (DRM) activity. Co-substituted CeO₂ showed the highest stability under the DRM reaction conditions at 800 °C. Detailed kinetic investigations were also carried out to estimate the apparent activation energy. Carbon deposition on the spent catalysts was investigated by thermal gravimetric analysis (TGA) and TEM which shows that the deactivation is due to the presence of amorphous and graphitic carbon. Transient studies on a mass spectrometer indicate that the prominence of the reaction CO₂ + C → 2CO is responsible for the catalyst's stability. Surface lattice oxygen reactivity is a vital factor in catalytic stability and its action decides the reaction steps. DFT further verifies that the energy of vacancy formation is significantly lower in Co-substituted CeO₂ as compared to Ni-substituted CeO₂. This confirms that the Co-substituted catalyst favors oxidation due to higher availability of surface oxygen, while in contrast Ni hinders oxidation by decreasing the availability of surface oxygen for the reaction.

Coating electrodes with self-assembled monolayer (SAM) of polar molecules is known to reduce contact resistance (RC) in organic field effect transistors (OFETs). It is shown that the behavior of SAM in OFETs can be explained by considering a mechanism of interfacial doping in organic semiconductor by electrodes due to charge transfer during interface formation. The enhancement is analyzed in performance of pentacene OFETs with Cu electrodes, by coating Cu with SAM of pentafluorobenzenethiol or perfluorodecanethiol. It is found that application of either SAM leads to an increase in work function of Cu surface. However, work function shift due to SAM does not correlate with trends in RC and OFET drain current. Further, first principle calculations reveal a notable difference in delocalization of frontal orbitals with either SAM, an indicator of the difference in ease of charge transfer across interface. Based on the mechanism of interfacial doping, a semiconductor physics model is developed for estimating interface doping and injection barrier, and for predicting the consequent device characteristics. It is believed that the model and methodology developed in this study can be utilized beyond the SAM and semiconductor system used here.

Charge transport properties of common donor copolymers in organic photovoltaics, poly({4,8-bis[(2-ethylhexyl)oxy]benzo[1,2-b:4,5-b′] dithiophene-2,6-diyl}{3-fluoro-2-[(2-ethylhexyl)carbonyl]thieno[3,4-b] thiophenediyl}) (PTB7) and poly([2,6′-4,8-di(5-ethylhexylthienyl)benzo[1,2- b;3,3-b]dithiophene]{3-fluoro-2[(2-ethylhexyl) carbonyl]thieno[3,4-b] thiophenediyl}) (PTB7-Th), with molecular structures differing only in the pendant group, are studied. This is the first report of field-effect transistor mobility (μ_FET) of PTB7-Th (0.14 cm2 V−1 s−1) and the highest μ_FET for PTB7 (0.01 cm2 V−1 s−1). μ_FET of PTB7-Th is found to be almost one order of magnitude higher than PTB7. To understand the influence of molecular structure on charge transport, hole reorganization energy (λh) is calculated from first-principles. λh of PTB7-Th (≈150 meV) is found to be lower than PTB7 (≈346 meV). Further, the ratio of hopping rate versus square of charge transfer integral calculated from Marcus theory using λh for these systems is found to indicate a higher rate of hole transfer across dimers or homojunction interface for PTB7-Th. These results are supplemented by experimentally determined λ using bulk-heterojunction organic solar cells, where λPTB7-Th≈200 meV and λPTB7≈310 meV follow a similar trend. The effective hole-mobility estimation from BHJ devices correlates well with these λ values. This study provides understanding of charge transport properties via reorganization energy, as a function of pendant group without altering the backbone of the chains.