Embedded Files
As conventional silicon solar cells approach their theoretical efficiency limits, researchers are increasingly looking toward tandem solar cells, which combine multiple light-absorbing materials to capture more of the Sun's energy. One of the most promising designs combines a perovskite solar cell on top of a conventional silicon solar cell. Perovskites can absorb high-energy visible light particularly well, while silicon absorbs lower-energy wavelengths that pass through the perovskite. Together, the two materials can use a much larger portion of the solar spectrum than either material could alone.
However, simply stacking two solar cells does not guarantee high efficiency. Light must be carefully managed as it travels through the device, and the surfaces of the two materials must be compatible with one another. In this 2022 Nature Nanotechnology paper, Tockhorn et al. investigated how nanoscale optical structures could solve both problems.
The researchers developed a new surface architecture for monolithic perovskite–silicon tandems that simultaneously improves light trapping, perovskite film formation, and manufacturing yield. Their optimized device achieved a 29.80% independently certified power conversion efficiency, demonstrating how nanoscale engineering can improve both performance and manufacturability [1].
What Are Perovskite–Silicon Tandem Solar Cells?
Perovskites are a class of crystalline materials with a characteristic structure known as the ABX₃ structure. Their composition can be modified to tune their electronic and optical properties, making them particularly attractive for photovoltaics. Perovskites used for their photovoltaic properties are more specifically metal-halide perovskites because they are made of a combination of organic ions, metals, and halogens [3].
Figure 1: Methyl ammonium lead triiodide, or MAPbI3, one of the more common perovskites [3]
One of their most important advantages is that their bandgap can be tuned. This allows researchers to design a perovskite that preferentially absorbs portions of sunlight that are not efficiently utilized by silicon.
A conventional silicon solar cell has a single absorber. When sunlight enters the cell, silicon absorbs photons with enough energy to excite electrons across its bandgap. However, photons with energies significantly greater than silicon's bandgap lose their excess energy as heat. This represents a fundamental efficiency loss.
A tandem cell approaches the problem differently.
In a perovskite–silicon tandem, the higher-bandgap perovskite is placed above the silicon:
Sunlight → Perovskite → Silicon
The perovskite absorbs much of the higher-energy visible light, while lower-energy photons pass through it and are absorbed by the silicon underneath.
Because the two materials divide the solar spectrum between them, less energy is lost through thermalization [2]. This gives tandem cells a pathway beyond the efficiency ceiling of conventional single-junction silicon photovoltaics: as of 2026, “...perovskite solar cell efficiency increased from about 3% in 2009 to over 26% today on small area devices (about 0.1 cm2). Perovskite-silicon tandem cells have reached efficiencies of almost 34%” [3].
Figure 2: Perovskite-silicon tandem solar cell design by researchers from Helmholtz-Zentrum Berlin (HZB) and the university École polytechnique fédérale de Lausanne (EPFL) in Switzerland [5]
The paper focuses specifically on monolithic two-terminal perovskite–silicon tandem solar cells (PSTSCs), meaning that the perovskite and silicon subcells are integrated into a single device and electrically connected in series. This architecture is attractive because it can potentially be manufactured as a compact, integrated solar cell rather than as two independent devices mechanically stacked together. These PSTSCs have already achieved conversion efficiencies exceeding 31% as of 2022 [1].
However, integrating the two materials together has multiple challenges.
The Problem with Conventional Silicon Texturing
Silicon solar cells commonly use surface texturing to reduce reflection.
Figure 3: Graphical abstract from the article "Comparison of different approaches to texturing monocrystalline silicon wafers for solar cell applications" to highlight the necessity of texturing silicon wafers to produce high-efficiency silicon solar cells [4]
A flat silicon surface reflects a significant amount of incoming sunlight. One way to reduce this loss is to cover the surface with microscopic pyramids. Instead of bouncing directly away, incoming light encounters the angled surfaces and has more opportunities to enter the silicon.
These textures work extremely well for conventional silicon photovoltaics.
The problem appears when a perovskite layer is deposited on top of the textured silicon.
The pyramids used for silicon solar cells can be several micrometers tall and have relatively sharp features. A thin perovskite film deposited over such a surface has to conform to this complicated landscape.
This can cause problems with:
uneven perovskite thickness
incomplete coverage
defects and shunts
inconsistent film formation
lower device performance
reduced manufacturing yield
In other words, the surface that is excellent for trapping sunlight in silicon can be terrible for producing a uniform perovskite layer.
Tockhorn et al. therefore faced a fundamental tradeoff where silicon wants a textured surface, but the perovskite wants a smoother surface.
Flattening the silicon would make perovskite deposition easier, but it would increase optical reflection and reduce the amount of sunlight entering the device.
The researchers' solution was to make the silicon textured but also make the texture small and smooth enough that the perovskite could still be deposited uniformly.
Nano-Optical Texturing
Instead of conventional micrometer-scale pyramids, Tockhorn et al. created a periodic nanoscale texture on the silicon surface.
The structure consisted of smooth, sinusoidal features arranged in a regular hexagonal pattern, with a periodicity of approximately 750 nanometers and a peak-to-valley height of roughly 300 nanometers.
This is dramatically smaller and smoother than conventional silicon pyramids.
The dimensions were deliberately chosen so that the perovskite layer–roughly 500-600 nm thick–could completely cover the underlying texture.
Figure 4: Nanotextured perovskite-silicon tandem solar cell design
A-C: Cross-sectional images of three solar cell designs (A is flat; B is nanotextured; and C is nanotextured with a rear dielectric layer (RDBL). D: Close-up image showing the nanoscale texture of the silicon surface. E: Photos of the completed solar cell, showing the blue active area on the front and the rear dielectric layer on the back [1]
Figure 5: Perovskite film formation and morphology
A-B: Perovskite films deposited on flat (a) and nanotextured (b) silicon surfaces. C: Percentage of solar cells successfully fabricated on each type of silicon surface. D-E: Photos showing how droplets of the perovskite solution spread on flat (d) and nanotextured (e) silicon surfaces. F: Comparison of perovskite crystal domain sizes on flat and nanotextured surfaces. G: Measurements showing the crystal orientation of the perovskite layers on the two types of surfaces [1]
Although Me-4PACz, a polymer material, helps improve the electrical performance of the solar cell, it can cause problems when the perovskite layer is deposited on a flat surface. The perovskite can form large holes or uncovered areas, which can prevent the solar cell from working properly. However, the Tockhorn et al. found that using a nanotextured silicon surface greatly reduced this problem.
When they deposited the perovskite onto the nanotextured cells, only 2 out of 45 devices developed visible holes. In comparison, 15 out of 30 flat devices developed holes. This shows that the nanotextured surface not only improves the optical properties of the solar cell, but also makes the manufacturing process much more reliable.
The silicon could still benefit from nanoscale optical engineering, while the perovskite was no longer forced to reproduce the sharp, irregular geometry of conventional silicon pyramids. This is the key innovation of the paper.
The nanotexture simultaneously serves two purposes: it manipulates light and improves the physical conditions for depositing the perovskite.
Why the Shape of the Surface Matters
The specific geometry of the texture determines how light interacts with the device.
When sunlight encounters the nanoscale surface, the gradual changes in refractive index and surface angle reduce reflection and alter the path of light through the tandem. More light can therefore enter the device and reach the appropriate absorber.
At the same time, because the features are relatively shallow and smooth, the perovskite layer can cover them without developing the severe irregularities associated with conventional pyramid textures.
This produced a surprising manufacturing benefit.
Tockhorn et al. found that the yield of functional perovskite devices increased from approximately 50% on planar reference structures to around 95% with the nanotextured architecture.
That means the innovation wasn't merely making the best individual solar cell better. It made the fabrication process considerably more reliable.
Rear Reflector
Tockhorn et al. also identified another source of efficiency loss which is the rear reflector.
After passing through the perovskite, some infrared light reaches the silicon. Ideally, any light that isn't absorbed immediately should be reflected back into the silicon for another chance to generate electricity.
Conventional designs use a transparent conducting oxide (TCO) together with a silver reflector. However, the TCO itself can absorb some of the infrared light.
Tockhorn et al. developed a different structure called a reflector with a dielectric buffer layer (RDBL).
Instead of placing the silver directly behind a relatively thick TCO, they inserted a layer of silicon dioxide (SiO₂) between them.
This allowed the TCO to be made much thinner, reducing parasitic absorption while the silver continued to provide strong reflection.
Figure 6: Implementing a rear RDBL for high-perfomance nanotextured PSTSCs
A-B: Diagrams showing the layers of the solar cell with a standard rear reflector (a) and a rear dielectric layer (b). C: Map showing how different silicon dioxide and transparent conductive oxide thicknesses affect the current produced by the silicon layer. D: Map showing power losses for different grid designs and conductive oxide thicknesses, comparing the new rear dielectric design with the standard reflector. E: Measurements showing how efficiently the perovskite and silicon layers absorb light and generate electricity in the two cell designs. F: Current–voltage curve of the nanotextured solar cell with a rear dielectric layer [1]
However, making the TCO thinner creates another problem: electrical resistance increases.
Tockhorn et al. found that making the transparent conducting layer thinner reduced overall energy losses. Although a thinner layer conducts electricity less effectively, it also absorbs less sunlight, allowing more sunlight to reach the silicon.. In this case, the extra light reaching the silicon more than made up for the increased electrical resistance.
The researchers therefore had to optimize the spacing of the silver electrical contacts. Too much metal would block sunlight; too little would increase electrical resistance.
This illustrates an important principle in solar-cell engineering: improving one component can create a new problem somewhere else. Tockhorn et al. therefore optimized the entire optical and electrical architecture together, rather than treating each layer independently.
Results
The combination of the nanotextured front surface and optimized rear reflector produced substantial improvements.
The nanotexture reduced optical reflection while also improving the performance of the perovskite layer. Tockhorn et al. observed an approximately 15 mV increase in open-circuit voltage compared with the planar reference.
The rear reflector increased the silicon photocurrent from approximately 19.4 to 19.7 mA cm⁻² by reducing infrared absorption in the rear layers.
Most importantly, the final tandem reached a 29.80% power conversion efficiency, with [1]:
Open-circuit voltage: 1.92 V
Short-circuit current density: 19.56 mA cm⁻²
Fill factor: 79.4%
Active area: approximately 1 cm²
The efficiency was independently certified by Fraunhofer ISE CalLab.
But perhaps the most interesting result for scaling was the improvement in fabrication yield. The nanotextured architecture increased the yield of functioning perovskite devices from roughly 50% to 95%.
That distinction matters because a technology intended for widespread deployment cannot be judged solely by the efficiency of its best laboratory sample.
Why Nano-Optical Engineering Matters
Tockhorn et al. demonstrated that nanoscale structures can solve multiple problems simultaneously. The same surface architecture that improves the way sunlight enters the solar cell also makes it easier to manufacture a uniform perovskite layer.
This is particularly important for perovskite–silicon tandems because their two materials have fundamentally different requirements. Silicon benefits from aggressive surface texturing for light trapping, while perovskite deposition requires a comparatively controlled surface.
More broadly, the paper demonstrates an important idea for the development of sustainable technologies: innovation does not always mean inventing a completely new material. Sometimes the breakthrough comes from changing the way existing materials are structured and integrated.
By engineering surfaces at the scale of hundreds of nanometers, Tockhorn et al. were able to improve optical efficiency, electrical performance, and manufacturing reliability at the same time.
As perovskite–silicon tandems continue moving toward commercial production, approaches like this could help address one of photovoltaics' biggest challenges: developing solar cells that are not only highly efficient in the laboratory, but also reproducible and manufacturable at scale.
References
Tockhorn, P., Sutter, J., Cruz, A. et al. Nano-optical designs for high-efficiency monolithic perovskite–silicon tandem solar cells. Nat. Nanotechnol. 17, 1214–1221 (2022). https://doi.org/10.1038/s41565-022-01228-8
Son, Y., Lim, J., Le, A. K., Kim, B.-S., Song, S., & Kim, H. (2026). A review of perovskite/Si tandem solar cells: internal and external components toward high efficiency, long-term durability, and commercialization. Materials Chemistry Frontiers, 10(1), 21–51. https://doi.org/10.1039/d5qm00637f
Perovskite Solar Cells. (2025). In Energy.gov. https://www.energy.gov/cmei/systems/perovskite-solar-cells
Han, S., Chu, M., Pham, D. P., Suresh Kumar Dhungel, & Yi, J. (2024). Comparison of different approaches to texturing monocrystalline silicon wafers for solar cell applications. Surface Science, 748, 122540–122540. https://doi.org/10.1016/j.susc.2024.122540
Perovskite/silicon tandem solar cell achieves record efficiency | Perovskite-Info. (2015). In Perovskite-info.com. https://www.perovskite-info.com/perovskitesilicon-tandem-solar-cell-achieves-record-efficiency
Page updated
Google Sites
Report abuse