LONGi Patent US12133398B2: The Ordered Layers Behind a Solar World Record
💡 TL;DR: Longi Green Energy Technology Co Ltd holds US patent US12133398B2 (granted October 29, 2024), which introduces two specialised ordered induction layers that make perovskite films crystallise cleanly on top of rough silicon surfaces. This engineering advance is a key building block explaining how LONGi broke through 35.5% power conversion efficiency in July 2026 - certified by ESTI and reported by pv-magazine. No single-junction silicon cell can ever reach that figure; perovskite-silicon tandems can.
What Patent US12133398B2 Actually Claims
The patent, simply titled "Tandem cell" and assigned to Longi Green Energy Technology Co Ltd, was filed at the USPTO on August 19, 2021 (claiming priority from August 20, 2020) and granted on October 29, 2024. The sole listed inventor is Chen Xu. In plain language, the patent solves a specific manufacturing problem: when you try to deposit a perovskite solar cell layer on top of a textured silicon wafer, the rough pyramid surface causes the perovskite to crystallise in a chaotic, defect-ridden way. Each defect is a trap where electrons and holes recombine before they can do useful work - and recombination bleeds efficiency.
The patent's answer is two thin ordered induction layers inserted between the silicon bottom cell and the perovskite top cell. The first layer sits between the silicon bottom cell and the hole transport layer. It uses rod-shaped molecules (such as BPTT) or metal oxides like zinc oxide that align upright through close packing, forming what the patent calls a "highly ordered geometric channel." That channel guides the hole transport layer above it to grow with higher crystallinity and fewer grain boundaries. The second layer sits between the hole transport layer and the perovskite absorber; it uses organic ammonium salts or inorganic lead compounds (lead oxide, lead iodide, lead bromide) that share chemical kinship with perovskite itself. That kinship means the perovskite crystallises with a far better structural match to the layer below. The result is a stack that behaves more like one coherent crystal than a disordered patchwork - and coherence means efficiency. This bridge between the two sublayers is what the rest of the efficiency gains depend on.
The Problem It Solves: Why Silicon's Rough Surface Is So Hard to Beat
Crystalline silicon solar cells are made by etching the wafer surface into microscopic pyramid shapes. Those pyramids scatter light, cutting reflection and lifting how much photon energy the cell captures. But those same pyramids make it extremely difficult to deposit a smooth, uniform perovskite layer on top - think of trying to lay a perfectly flat carpet over a room full of small mountains. Conventional deposition methods (spin coating, blade coating) produce films with uneven grain sizes and many defects on textured surfaces.
The Shockley-Queisser limit caps single-junction silicon at around 33.7% theoretical efficiency. A perovskite-silicon tandem, by stacking two absorbers that harvest different parts of the solar spectrum, can in theory exceed 43%. Closing that gap is primarily a materials and interface engineering challenge - and that challenge is precisely what US12133398B2 addresses. This is also why the patent matters for the field's trajectory, not just for LONGi.
What This Patent Depends On and What It Could Unlock
US12133398B2 sits inside a wider innovation loop. On the input side, it depends on high-quality heterojunction silicon bottom cells (which LONGi also leads in producing), high-purity perovskite absorber chemistry, and precision thin-film deposition equipment. On the output side, it enabled a series of efficiency records and laid the groundwork for LONGi's newer asymmetric self-assembled monolayer technology - named HTL201 - documented in their patent application CN117603266A (filed in China) and published in a 54-author Nature paper in July 2025. HTL201 targets a further problem: minimising steric hindrance at the buried perovskite interface. Together, ordered induction layers and the asymmetric SAM approach explain most of LONGi's efficiency trajectory from 33.9% (2023) to 35.5% (2026).
If this approach can be scaled to large-area deposition - still the main unsolved engineering challenge - it would push tandem module efficiencies well beyond what silicon alone can deliver, potentially displacing a significant share of today's silicon-only panel market. That threat is large enough that Oxford PV (UK, with over 400 granted perovskite patents) licensed its portfolio to First Solar in February 2026 for the US market, and to Trina Solar in 2025 for China. The IP race is already shaping which manufacturers will be allowed to produce the next generation of solar panels without paying licensing fees.
Why July 2026 Is a Turning Point for the Field
The 35.5% result - certified by the European Solar Test Installation (ESTI), operated by the European Commission's Joint Research Centre in Italy, and announced on July 14, 2026 - matters for three reasons. First, it is the first independently certified result to exceed 35% for a two-terminal tandem cell, a threshold that had been approached but not crossed. Second, it is part of an unbroken upward trajectory: LONGi moved from 34.85% (April 2025) to 35.5% in roughly 15 months - showing no plateau. Third, it narrows the gap between the leader and the nearest rival: JinkoSolar achieved 34.82% in June 2026, certified by the Shanghai Institute of Microsystem and Information Technology of the Chinese Academy of Sciences, using a different approach combining dual-layer passivation and gradient crystallisation kinetics. Under one percentage point separates the two.
Mass production is still years away. LONGi stated in mid-2026 that it had no active plan for commercial tandem module production, citing outstanding challenges in large-area deposition uniformity, encapsulation durability, long-term stability, and cost. But laboratory records lead commercial deployment by roughly five to ten years in solar history - and the laboratory record is moving fast.
Who Is Behind This and Who It Threatens
LONGi Green Energy Technology Co., Ltd. is headquartered in Xi'an, China and is the world's largest producer of monocrystalline silicon solar wafers. The 54-author Nature paper from July 2025 lists 40 authors from LONGi alone, with the remaining 14 from Soochow University, CNITECH (Chinese Academy of Sciences), Hong Kong Polytechnic University, Beijing University of Technology, and the University of Science and Technology of China. This is a company doing serious in-house basic research at industrial scale - a combination that is unusual and consequential.
The companies most threatened by this trajectory are those manufacturing conventional 22-24% efficiency silicon modules who have not invested in perovskite IP. The companies best positioned are those with both IP portfolios and manufacturing scale: LONGi, Oxford PV, JinkoSolar, Trina Solar, and First Solar (via Oxford PV licensing). In Southeast Asia, where solar installation capacity is growing rapidly, the patent landscape in tandem technology will directly determine who can manufacture the next generation of panels - and at what cost.
| Detail | Value |
|---|---|
| Patent Number | US12133398B2 |
| Title | Tandem cell |
| Assignee | Longi Green Energy Technology Co Ltd |
| Inventor | Chen Xu |
| Priority Date | August 20, 2020 |
| Filing Date (US) | August 19, 2021 |
| Grant Date | October 29, 2024 |
| Jurisdiction | United States (USPTO) |
| Core Claim | Two ordered induction layers enabling high-crystallinity perovskite on textured silicon |
A Note on IP Strategy: Why Patent Translation Is Central Here
LONGi's US12133398B2 is a Chinese-origin invention filed into the US system. Oxford PV files from the UK into US, EP, and PCT systems. JinkoSolar files primarily in China with PCT extensions. For any manufacturer, government agency, or research institution in Vietnam or Southeast Asia wanting to license these technologies, challenge a claim in an infringement dispute, or file a response to an invalidity action, the technical language of the patent claims must be translated accurately. Terms like "ordered induction layer," "Van der Waals interaction," "weak epitaxial effect," "organic ammonium salt," or "rod-shaped molecular material" are precise claim elements where a single mistranslation can shift the scope of protection. Professional patent translation and technical translation are not a formality in this field; they are a legal instrument.
So What Does It Mean for Us?
The 35.5% world record confirms that perovskite-silicon tandem technology is real, reproducible, and still improving. The patent US12133398B2 is one verified engineering brick in that structure. What remains is the industrial leap from certified laboratory cells to durable, affordable commercial modules - and that leap will take years and several more generations of IP.
For energy planners, the signal is clear: solar efficiency is not plateauing. The economic assumptions behind large-scale solar deployments (cost per watt, required land area) will keep improving. For IP professionals and patent attorneys working in Asia-Pacific, the tandem solar space is generating a dense and growing body of cross-jurisdictional patent filings that will require expert IP translation and English to Vietnamese technical translation as these filings move into Southeast Asian markets. For technology companies thinking about localisation, the software and systems that control these solar cells - inverters, energy management platforms, monitoring tools - are also targets for technology localization into Vietnamese and other regional languages.
FAQ
What makes a perovskite-silicon tandem cell more efficient than a regular silicon cell?
A regular silicon cell captures photons within a limited energy range. A tandem stacks two absorbers: the perovskite top layer captures high-energy (blue) photons, and the silicon bottom layer captures lower-energy (red) photons. Together they extract energy from a wider portion of the solar spectrum, which is why the theoretical efficiency limit rises from about 33.7% (silicon alone) to over 43%.
What exactly is an ordered induction layer in LONGi's patent?
It is a thin molecular or metal-oxide layer deposited between the silicon bottom cell and the perovskite top cell. Rod-shaped molecules align upright and close-packed, forming a template that guides the perovskite film to crystallise in a more ordered structure with fewer defects. Fewer defects mean less energy lost to electron-hole recombination, which translates directly into higher cell efficiency.
Is the 35.5% efficiency commercially available today?
No. The 35.5% result is a certified laboratory measurement on a small-area device. LONGi stated in mid-2026 that it had no active commercial production plan for tandem cells, citing unresolved challenges in large-area deposition, encapsulation, long-term stability, and manufacturing cost. Commercial tandem modules are likely still several years from market at scale.
Who are the key competitors in the perovskite-silicon tandem space?
LONGi (China) currently holds the world efficiency record at 35.5%. JinkoSolar (China) achieved 34.82% in June 2026. Oxford PV (UK) has the largest granted patent portfolio in perovskite PV (400+ patents) and has licensed its technology to First Solar (US market, February 2026) and Trina Solar (China market, 2025). Hanwha Q CELLS and other major manufacturers are also active.
Why does patent translation matter for solar technology IP?
Solar patents are filed across multiple jurisdictions - China, the US, Europe (EPO), and WIPO PCT. When a company in Vietnam or Southeast Asia wants to license one of these patents, challenge a claim, or defend against infringement, the technical claim language must be translated accurately into Vietnamese. Terms like "ordered induction layer" or "weak epitaxial effect" are legal claim elements - mistranslating them can change the scope of protection. This is precisely why professional patent translation services are essential in cross-border IP strategy.
Sources
Google Patents: US12133398B2 "Tandem cell" - Longi Green Energy (granted Oct 2024)
pv-magazine: LONGi 35.5% world record (Jul 2026)
pv-magazine: JinkoSolar 34.82% (Jun 2026)
Solar Power World: First Solar - Oxford PV license (Feb 2026)
About the author
Dao Huy (Lucas) is a professional translator specialising in English, Chinese, and French into Vietnamese, with over 7 years of experience in technical translation, patent translation, and intellectual property documentation. The rapid expansion of perovskite solar IP across US, EP, and PCT jurisdictions is generating a growing demand for precise IP translation for licensing negotiations, patent validity proceedings, and regulatory filings in Vietnamese-speaking markets.
If your organisation needs Vietnamese translations of solar technology patents, engineering documents, or IP filings, Dao Huy offers professional patent translation and technology localization services. Request a quote at daohuy.com.
Written by Dao Huy (Lucas), Vietnamese translator & localization specialist (EN · ZH · FR → Vietnamese). See translation services →