Improving energy conversion efficiencies in various systems, such as organic solar cells, is a crucial goal in energy research. Organic solar cells, with their low cost, environmental friendliness, and processability, have seen significant efficiency improvements, exceeding 18% in some non-fullerene acceptor-based devices. These cells function through exciton dissociation into charge-transfer (CT) states at donor-acceptor interfaces, followed by dissociation into free charge carriers. Early optimization focused on materials with small optical and large photovoltaic gaps (|IED − EAA|), aiming for broader solar spectrum harvesting and higher Voc. However, this simplified view neglected the broadening of energy levels and the influence of charge-carrier mobility and recombination. The development of low-bandgap non-fullerene acceptors marked a turning point. Research revealed the importance of long-range electrostatic interactions of charges with surrounding molecular quadrupoles, influencing the energy landscape of electrons, holes, and CT states. These interactions contribute to ionization energy (IE) and electron affinity (EA), directly affecting Voc. The challenge lies in balancing electrostatic effects, which promote efficient CT state dissociation, with the undesired reduction of the photovoltaic gap and consequently, Voc. This paper uses tailored material systems to separate the influence of morphology, molecular structure, and device architecture on microscopic energetics, aiming to optimize solar cell efficiency through control of interfacial electrostatics.

The literature review extensively covers previous research on organic solar cells, highlighting the progress in power conversion efficiency achieved through material optimization and the understanding of underlying energy transfer mechanisms. It discusses the evolution of optimization strategies, starting from a focus on materials with small optical and large photovoltaic gaps and progressing to incorporate more sophisticated models that consider the effects of energetic disorder, charge-carrier mobility, and non-geminate recombination. The review also highlights the significance of recent developments in low-bandgap non-fullerene acceptors (NFAs) and the growing understanding of the role of long-range electrostatic interactions in determining the energetic landscape of charge carriers and CT states. It cites numerous studies demonstrating how these interactions influence ionization energy (IE), electron affinity (EA), and the energy offset between CT and charge-separated states, directly impacting the open-circuit voltage (Voc) of organic solar cells. The reviewed works emphasize the necessity of balancing these electrostatic contributions to optimize overall cell efficiency.

The study employed two experimental approaches. The first involved planar heterojunction solar cells with a constant C60 acceptor phase and varying donor phases (mixtures of ZnPc, F4ZnPc, and FZnPc). Fluorination allowed precise tuning of the electrostatic field in the intermixed crystalline phase, enabling continuous adjustment of IEs and EAs while maintaining crystalline morphology. The second approach involved incorporating a mixed donor:acceptor interlayer between pure donor and acceptor layers to control interface roughness. Microscopic simulations, performed using atomistically resolved models of mixed ZnPc:F4ZnPc and ZnPc:FZnPc films (10 nm thick), modeled the long-range effects on energy levels using periodic boundary conditions. The simulations determined perturbative corrections to hole energy levels and calculated ionization energies. Sensitive EQEpy measurements were conducted using a chopped light source, monochromator, and lock-in amplifier to analyze photocurrent, determining EQEpy by dividing the photocurrent by the incoming photon flux. Electroluminescence measurements, using an Andor spectrometer with cooled Si and InGaAs detectors, were obtained at various injection currents. The low-energy tail of the EQEpy spectrum was fitted with a Marcus equation to extract charge-transfer state energy (ECT), relaxation energy (λ), and oscillator strength (f). Device fabrication involved thermal evaporation of organic materials onto cleaned ITO substrates, layer-by-layer deposition of BPAPF (hole transporting layer), the active layer blend, a ZnPc:C60 interlayer (with varying thickness for roughness control), C60, BPhen (electron contact), and Al. Current-voltage characteristics were measured using a SMU under standard testing conditions. The VOTCA package was used for simulations.

The study found a linear relationship between ionization energy (IE) and the fraction of ZnPc in ZnPc:F4ZnPc and ZnPc:FZnPc mixtures, attributed to the superposition of quadrupolar fields. The photovoltaic gap showed a similar linear dependence, directly inherited from the composition-dependent IE. Interfacial electrostatic bias for holes and electrons (Bh, Be) was quantified, reflecting the electrostatic asymmetry of the interface. A negative bias (B < 0) for F4ZnPc at specific compositions correlated with fill factor deterioration, suggesting trap formation. The energy of interfacial CT states (ECT) was modeled, incorporating effective electron-hole binding energy and interfacial electrostatic bias. Interface roughness significantly impacted ECT, with greater roughness influencing the electrostatic bias's impact on CT energy. A single parameter, ρ = 2h/δ (where h is the width of the intermixed region and δ the interlayer spacing), captured the roughness effect. The composition-dependent ECT measurements best fitted with ρ = 1, suggesting approximately two intermixed molecular layers. The open-circuit voltage (Voc) was modeled, considering chemical potentials, charge-carrier densities, and energetic disorder. The observed increase in Voc with increased roughness in ZnPc/C60 and FZnPc/C60 systems (where B > 0) was explained by improved charge push-out forces. Conversely, Voc decrease was observed for F4ZnPc/C60 (B < 0). The experimental Voc and ECT data, shown as a function of composition and interfacial roughness (controlled by interlayer thickness), supported the model's predictions, confirming the significant impact of interfacial roughness on Voc. Table 1 summarized detailed experimental measurements (Jsc, Voc, FF, PCE, ECT) for various systems and interface roughnesses.

The findings demonstrate a clear correlation between interfacial roughness, electrostatic bias, and the open-circuit voltage (Voc) of organic solar cells. The model successfully explains how nanoscale roughness at the donor-acceptor interface can either enhance or hinder Voc, depending on the magnitude and sign of the electrostatic bias. The results highlight the importance of considering interfacial morphology in addition to bulk material properties when designing high-efficiency organic solar cells. The ability to control Voc through morphology offers a new pathway for optimizing device performance, potentially surpassing limitations imposed by the inherent bulk properties of individual materials. The observation that roughness can compensate for negative electrostatic effects presents significant implications for material selection and device architecture in organic photovoltaics.

This research provides a comprehensive model linking interfacial morphology, electrostatic interactions, and open-circuit voltage in organic solar cells. The results show that interfacial roughness significantly influences Voc, offering a new design principle for optimizing device performance. Future work could focus on exploring diverse material combinations and interface engineering techniques to further enhance Voc and overall efficiency in organic solar cells.

The model utilizes certain approximations, such as assuming constant electron-hole binding energy and using bulk values to approximate solid-state contributions to ionization energy at the interface. While the model successfully captures the major trends, these approximations may limit its precision in specific scenarios. The experiments focused on specific small molecule systems; further studies are needed to determine the generality of the findings across broader classes of organic materials and device architectures. The study also focuses on the impact of morphology on Voc while acknowledging the potential correlation between changes in Voc and Jsc. Further investigation into disentangling these effects would refine the understanding of the overall performance optimization.