The architecture is deceptively simple in outline. The device follows a nip configuration, beginning with a transparent zinc oxide cathode 100 nanometers thick, followed by a 60-nanometer cadmium sulfide electron transport layer, a 500-nanometer active layer of the semiconducting polymer blend PTB7:PC71BM, an 80-nanometer zinc tungstate hole transport layer, and a gold back contact. What distinguishes the design is what happens inside the electron transport layer: instead of a flat film of CdS, the researchers patterned it into four parallel rectangular ridges, each 450 nanometers tall and 26 nanometers wide, with a grating period of 273.2 nanometers. These gratings penetrate the polymer absorber and act as continuous electron highways, shortening the distance carriers must travel before collection while simultaneously increasing the optical path length of light bouncing through the active layer.

Into the spaces between these gratings, the team embedded a double layer of gold nanospheres with radii ranging from 50 to 105 nanometers. The choice of gold is not arbitrary. When illuminated, the conduction electrons in each nanoparticle oscillate coherently at the particle surface — a phenomenon known as localized surface plasmon resonance, or LSPR. This resonance produces intense local electromagnetic fields and strong far-field scattering, both of which can funnel otherwise lost photons back into the absorber. Isolated gold nanospheres typically resonate around 520 nanometers, deep in the visible range, but the researchers showed that the strong near-field coupling between the two nanoparticle layers, combined with the high refractive index environment of the surrounding CdS grating, shifts and broadens the resonance into the near-infrared region between roughly 650 and 1100 nanometers — precisely where the PTB7:PC71BM polymer is weakest at absorbing on its own.

The bimodal spacing of the nanoparticles proved critical. Inter-particle distances vary from less than 70 nanometers up to 110 nanometers, deliberately combining strong near-field plasmonic coupling in closely packed regions with efficient far-field scattering from more widely separated particles. According to Mie scattering theory, the amount of light scattered by a nanoparticle scales with its size, which is why the team swept radii from 50 to 105 nanometers in their optimization. Their tailored parametric algorithm — chosen over machine learning approaches because the complex electromagnetic interactions of the plasmonic-grating hybrid exceed the predictive reach of models trained on conventional flat devices — identified configurations that produce peak absorption in the 730 to 860 nanometer band and sustained absorption out to 1000 nanometers. Finite-difference time-domain simulations revealed local field enhancements of up to 25 times at the nanoparticle surfaces, contributing an additional 6.77 milliamperes per square centimeter to the short-circuit current density.

Plasmonic nanoparticles are not without drawbacks, and the researchers confronted them directly. Metal nanoparticles can introduce parasitic absorption and serve as recombination centers, and previous experimental studies have reported performance losses of 15 to 30 percent from such effects. In the proposed architecture, however, the strategic placement of the double-layer nanoparticles between the CdS gratings, along with the sub-wavelength 273.2-nanometer grating period, suppresses parasitic losses to less than 2 percent. The gratings also screen trapped charges at the gold–polymer interface, reducing the interface recombination velocity from 100,000 to 1,000 centimeters per second, while the modeled surface trap density of 10^12 per square centimeter produces only a negligible 0.02-volt drop in open-circuit voltage. Because the grating period is sub-wavelength and gold coverage is discontinuous, propagating surface plasmon polaritons are not excited, keeping the enhancement firmly in the localized, beneficial regime.

The electrical results are striking. The short-circuit current density rises from 24.91 milliamperes per square centimeter for the grated CdS structure without nanoparticles to 31.68 milliamperes per square centimeter once the gold nanospheres are added. Open-circuit voltage reaches 1527 millivolts, a value the authors attribute to excellent energy level alignment among the grated CdS electron transport layer, the PTB7:PC71BM absorber, and the zinc tungstate hole transport layer, combined with suppressed non-radiative recombination. The conduction band offset of 0.15 electron-volts between CdS and PC71BM facilitates efficient electron injection while blocking hole back-transfer, and the small 0.15 to 0.25 electron-volt offset between the absorber’s HOMO level and the valence band of ZnWO4 provides a strong driving force for hole extraction. The fill factor holds at a remarkable 91 percent for both configurations, and the power conversion efficiency climbs from 34.66 percent to 44.02 percent.

External quantum efficiency measurements from the simulations underscore where the gains come from. In the near-infrared band between 800 and 1070 nanometers, the nanoparticle-enhanced structure reaches an EQE of approximately 62 percent at 930 nanometers, compared with a mere 3 percent for the grated structure without gold. Enhancements also appear across the 300 to 430 nanometer and 530 to 790 nanometer bands, driven by the increased contact area between the CdS gratings and the polymer absorber. Overall absorption in the 850 to 1150 nanometer range improves by more than 50 percent relative to the initial grated design. The team validated their model by comparing predicted parameters against experimental literature values for PTB7:PC71BM devices, acknowledging that the simulated efficiency represents a theoretical ceiling achievable only under ultra-low defect conditions that remain challenging to realize at scale.

Fabrication feasibility was a central consideration throughout the study. The authors propose a hybrid manufacturing route in which the zinc oxide, PTB7:PC71BM, and zinc tungstate layers are deposited by slot-die coating — a scalable, roll-to-roll-compatible technique in which ink forms a controlled meniscus between a slot die and the substrate, governed by the interplay of viscous, gravitational, and shear forces. The CdS grating would be patterned by nanoimprint lithography, while the double-layer gold nanospheres could be realized through controlled electrodeposition or colloidal spin-coating. Recent demonstrations of similarly processed devices reaching 21.87 percent efficiency suggest the pathway to experimental validation is realistic, though the authors caution that annealing temperatures, process tolerances, and yield analysis will require experimental optimization.

Thermal stability remains an open challenge. The mismatch in thermal expansion coefficients among the layers is relatively small, but localized plasmonic heating around the gold nanoparticles and the inherent sensitivity of the organic active layer could induce mechanical stress and long-term degradation under prolonged solar exposure at operating temperatures of 50 to 70 degrees Celsius. The researchers propose thin encapsulation layers of PMMA or glass, buffer layers with matched thermal expansion, and — in future work — replacing the fullerene-based PTB7:PC71BM absorber with more thermally robust non-fullerene acceptors. They also note that the design deliberately avoids machine learning optimization for now, arguing that full-wave electromagnetic modeling combined with drift–diffusion electrical simulation offers physically transparent insight into plasmonic hot-spot distribution and recombination suppression that black-box models cannot yet provide, though machine learning-assisted global optimization is planned for future refinements.

The broader significance of the work lies in its demonstration that photonic and plasmonic strategies, often studied in isolation, can be engineered to reinforce one another. By coupling the diffractive light-trapping of a periodic CdS grating with the resonant field concentration of a bimodal gold nanosphere array, the design extracts useful current from a spectral region that most single-junction cells abandon. Whether the simulated 44.02 percent efficiency can survive contact with real-world fabrication imperfections is the question that will define the next chapter of this research, but the numerical evidence presented offers a concrete, physically grounded roadmap — one that could reshape expectations for what polymer solar cells, with their low-cost solution processing and mechanical flexibility, might ultimately deliver in the global transition away from fossil fuels.

Subject of Research: Plasmonic enhancement of polymer solar cells using gold nanosphere arrays and grated CdS electron transport layers

Article Title: Achieving higher efficiency in solar panels using gold nanosphere arrays in structure of Grated CdS and advanced interfacial layers

Article References: Kianjo, S., & Nooramin, A. (2026). Achieving higher efficiency in solar panels using gold nanosphere arrays in structure of Grated CdS and advanced interfacial layers. Results in Optics, 25, Article 101140. https://doi.org/10.1016/j.rio.2026.101140

Image Credits: AI Generated

DOI: 10.1016/j.rio.2026.101140

Keywords: solar cells, gold nanoparticles, plasmonics, LSPR, CdS grating, PTB7:PC71BM, FDTD simulation, near-infrared absorption, power conversion efficiency, zinc tungstate, slot-die coating, nanoimprint lithography