Semantic search for patents retrieves documents by modeled meaning rather than by lexical overlap. A Boolean or keyword query matches on the surface form of a string: it returns patents whose text contains the specified tokens, combined through operators and often expanded with truncation, proximity, and classification filters. A semantic query matches on representation: each document and the query are encoded as high-dimensional vectors, and retrieval ranks documents by vector similarity, so patents whose claims and disclosures are conceptually related are returned even when they share no keywords with the query. This distinction is decisive in patent work because equivalent subject matter is routinely described in divergent vocabulary, and applicants draft claims with deliberately broad, idiosyncratic, or coined terminology to widen scope. When relevant patents use language the searcher did not anticipate, lexical retrieval fails to surface them, and those recall failures are the primary source of risk in prior art and freedom-to-operate analysis.

The stakes are set by scale. The World Intellectual Property Organization reported roughly 3.5 million patent applications filed worldwide in 2023,¹ and scientific literature has grown at approximately 8 to 9 percent per year across recent decades,² so the candidate space a searcher must cover exceeds what manually constructed Boolean queries can reliably span. Lexical retrieval trades recall for precision: it returns documents matching the exact terms and omits everything phrased differently. Evaluations in the patent-retrieval literature have found that keyword and Boolean logic disregard the syntax and semantics of technical language, which makes patent retrieval materially harder than general-domain information retrieval and leaves relevant documents unretrieved.³,⁴ In prior art and freedom-to-operate, the cost of a single missed document is high, because one overlooked reference can defeat a novelty position or expose a product to infringement liability.

Semantic search also underpins the shift toward agentic AI in R&D and IP. AI agents increasingly query patent and scientific corpora through an API and execute multi-step retrieval and analysis, and dense semantic retrieval is what allows an agent to locate relevant documents without a human hand-crafting Boolean strings. The Model Context Protocol (MCP), introduced by Anthropic in November 2024 and donated to the Linux Foundation's Agentic AI Foundation in December 2025, is now the common standard through which agents connect to external data.⁵ Semantic retrieval is the layer that grounds these agentic workflows in real documents.

How semantic search works

Semantic search rests on learned vector representations. A transformer-based language model, typically fine-tuned on patent and scientific text, encodes each document into a dense embedding, a fixed-length vector of several hundred dimensions whose geometry captures semantic relationships. The query is encoded by the same model, and relevance is scored by a similarity function, most commonly cosine similarity or inner product, over the shared vector space. Because encoding is done offline, retrieval at query time reduces to a nearest-neighbor search: the system returns the documents whose embeddings lie closest to the query embedding, which are the documents most similar in modeled meaning rather than in wording.

At corpus scale, exhaustive comparison against every vector is infeasible, so semantic search relies on approximate nearest-neighbor (ANN) indexing, using structures such as hierarchical navigable small-world graphs or inverted-file quantization to return near-optimal neighbors in sub-linear time. The retrieval model itself is usually a bi-encoder, which embeds query and document independently for speed. A second, more expensive cross-encoder is often applied as a reranking stage over the top candidates, jointly attending to query and document to refine the ordering. This retrieve-then-rerank architecture is standard because it combines the throughput of ANN retrieval with the precision of pairwise scoring.

Dense retrieval demonstrably outperforms lexical baselines. Dense passage retrieval improved top-20 retrieval accuracy by 9 to 19 percentage points over a strong keyword baseline in open-domain benchmarks,⁶ and patent-specific embedding models have been engineered for this domain: the European Patent Office's SEARCHFORMER uses siamese transformer encoders to produce semantic patent embeddings purpose-built for prior art search,⁷ and subsequent work has shown that few-shot fine-tuning and quantized embeddings can adapt these models efficiently to patent retrieval benchmarks.⁸ Three domain-specific factors make this adaptation necessary. Patents are long and structurally heterogeneous, so documents are segmented into passages, and claims are frequently indexed separately because claim language, not the abstract, defines legal scope. Patent corpora are multilingual, so cross-lingual embeddings allow a query in one language to retrieve prior art in another. And patents carry structured metadata, so production systems typically use hybrid retrieval, fusing dense semantic scores with lexical signals and Cooperative Patent Classification or International Patent Classification codes through rank-fusion methods to combine conceptual recall with exact-match precision.

Semantic search becomes substantially more powerful when the vector layer is combined with an explicit knowledge structure. An ontology that formalizes how technologies, materials, methods, and claims relate lets a platform interpret a query within a technology domain and cluster results by concept rather than surface wording. This combination, dense retrieval organized by an ontology, is what separates a modern patent-analytics platform from a keyword database with a search box: it improves recall by retrieving conceptually related work, and it improves interpretability by organizing results into a navigable conceptual structure. Retrieval quality in this setting is measured with recall-oriented metrics such as recall@k and mean average precision, which reflect the priority of finding all relevant documents rather than only the top few.

Where semantic search changes patent work

Prior art search. Prior art search is recall-bound, because the objective is to surface any earlier disclosure bearing on novelty or obviousness. Dense semantic retrieval surfaces prior art expressed in different terminology from the invention, including non-patent literature, which lexical search omits, producing a more complete novelty assessment.

Freedom-to-operate. Freedom-to-operate depends on identifying active, in-force claims that a product could read on, including claims drafted to cover a concept broadly. Semantic retrieval locates relevant claims independent of exact terminology and, combined with claim-level indexing, narrows the analysis to the independent claims that define infringement scope, reducing the coverage gaps that generate FTO risk.

White space analysis. White space analysis depends on clustering activity by concept to expose genuine gaps. Embedding-based clustering distinguishes real conceptual sparsity from apparent sparsity that is only an artifact of divergent terminology, so the identified white space reflects unclaimed technical territory rather than a vocabulary mismatch.

Technology and competitive intelligence. Characterizing a technology area requires connecting related work across patents and scientific literature. Encoding both sources in a shared vector space lets a platform retrieve and align conceptually related documents across them, supporting attribution of activity to technology areas and organizations.

Where Cypris fits

Cypris applies semantic search across a corpus of more than 500 million patents and scientific papers, organized through a proprietary R&D ontology. The ontology is what makes the dense retrieval interpretable: it maps how technologies, claims, and research relate, so retrieval is organized by concept rather than surface wording, and results are returned as a navigable conceptual structure rather than a flat ranked list. This lets Cypris surface conceptually relevant patents and papers for prior art, freedom-to-operate at the claim level, and white space analysis, closing the recall gaps that lexical search leaves. Cypris Q, the platform's agentic layer, lets teams run semantic queries conversationally and chain them into multi-step retrieval and analysis, and Agentic Monitoring keeps results current by tracking a technology area over time. Cypris provides enterprise API partnerships with OpenAI, Anthropic, and Google, so AI agents can execute semantic retrieval across the corpus programmatically, and it is built with enterprise-grade security. Cypris serves hundreds of enterprise customers across pharmaceuticals, chemicals, advanced materials, energy, and other regulated industries.

FAQ

What is semantic search for patents?

Semantic search for patents retrieves patents and scientific papers by modeled meaning rather than by exact keywords. It encodes documents and queries as dense vectors and ranks results by vector similarity, so conceptually related documents are returned even when they share no keywords. This closes the recall gaps that cause missed prior art and freedom-to-operate risk in Boolean search.

How is semantic search different from keyword search?

Semantic search differs from keyword search in what it matches. Keyword and Boolean search match the surface form of a query string, while semantic search matches learned vector representations of meaning. For patents this matters because equivalent subject matter is described in divergent vocabulary, and lexical search misses documents phrased differently or drafted with deliberately broad claim language.

What are embeddings in semantic patent search?

Embeddings in semantic patent search are dense vectors, produced by a transformer language model, that encode the meaning of a patent, a claim, a passage, or a query into a shared high-dimensional space. Similarity between embeddings, typically cosine similarity, measures conceptual relatedness. Retrieval returns the documents whose embeddings are nearest to the query embedding.

What is dense retrieval and how does it compare to BM25?

Dense retrieval encodes queries and documents as learned vectors and ranks by vector similarity, whereas BM25 is a sparse, term-frequency lexical method. Dense passage retrieval has been shown to improve top-20 retrieval accuracy by 9 to 19 percentage points over a strong lexical baseline. Production patent systems often combine the two in hybrid retrieval to gain both conceptual recall and exact-match precision.

Why does semantic search matter for prior art search?

Semantic search matters for prior art search because prior art is recall-bound, and relevant disclosures are frequently phrased differently from the invention or appear in non-patent literature. Lexical search omits these, leaving gaps in the novelty assessment. Dense semantic retrieval surfaces conceptually related disclosures regardless of wording, making the assessment more complete.

How does semantic search improve freedom-to-operate analysis?

Semantic search improves freedom-to-operate analysis by locating active claims a product could read on even when those claims use different terminology or cover a concept broadly. Combined with claim-level indexing, it focuses the analysis on the independent claims that define infringement scope. This reduces the coverage gaps that are the main source of FTO risk.

What is a retrieve-then-rerank pipeline?

A retrieve-then-rerank pipeline is a two-stage architecture. A fast bi-encoder retrieves a candidate set using approximate nearest-neighbor search, then a more expensive cross-encoder rescores the top candidates by jointly attending to the query and each document. This combines the throughput of vector retrieval with the precision of pairwise relevance scoring.

Does semantic search need an ontology?

Semantic search does not strictly require an ontology, but combining the two is substantially more powerful. An ontology formalizes how technologies relate, letting a platform interpret a query within a domain and cluster results by concept. Cypris combines dense semantic retrieval with a proprietary R&D ontology across more than 500 million patents and scientific papers for this reason.

How do AI agents use semantic search for patents?

AI agents use semantic search as the retrieval layer that lets them locate relevant patents and papers without a human writing Boolean strings, enabling autonomous multi-step analysis. Agents query the corpus through an API and use dense retrieval to ground their reasoning in real documents. Cypris offers enterprise API partnerships with OpenAI, Anthropic, and Google so agents can run semantic retrieval across its corpus.

Which teams benefit from semantic patent search?

Semantic patent search benefits R&D, innovation, and IP teams running prior art, freedom-to-operate, white space, and technology-intelligence searches, all of which depend on high-recall retrieval of conceptually relevant documents. It is most valuable in research-intensive industries such as pharmaceuticals, chemicals, advanced materials, and energy. Cypris serves hundreds of enterprise customers across these industries.

Endnotes

1. World Intellectual Property Organization. World Intellectual Property Indicators (annual series). https://www.wipo.int/publications/
2. Bornmann, L. & Mutz, R. (2015). Growth rates of modern science: a bibliometric analysis based on the number of publications and cited references. Journal of the Association for Information Science and Technology. https://doi.org/10.1002/asi.23329
3. Zihayat, M. & Etwaroo, R. (2021). A non-factoid question answering system for prior art search. Expert Systems with Applications. https://doi.org/10.1016/j.eswa.2021.114910
4. Lupu, M., Piroi, F., Hanbury, A. & Zenz, V. (2011). CLEF-IP 2011: Retrieval in the Intellectual Property Domain.
5. Anthropic (2025). Donating the Model Context Protocol and establishing the Agentic AI Foundation. https://www.anthropic.com/news/donating-the-model-context-protocol-and-establishing-of-the-agentic-ai-foundation; Linux Foundation (2025). Formation of the Agentic AI Foundation. https://www.linuxfoundation.org/press/linux-foundation-announces-the-formation-of-the-agentic-ai-foundation
6. Karpukhin, V. et al. (2020). Dense Passage Retrieval for Open-Domain Question Answering. EMNLP. https://doi.org/10.18653/v1/2020.emnlp-main.550
7. Vowinckel, K. & Hähnke, V. D. (2023). SEARCHFORMER: Semantic patent embeddings by siamese transformers for prior art search. World Patent Information. https://doi.org/10.1016/j.wpi.2023.102192
8. Chikkamath, R. et al. (2025). Patent Retrieval with Few-Shot Fine-Tuning and Quantized Embeddings. https://doi.org/10.1145/3787279.3787295

‍

Direct air capture has moved from demonstration to first commercial plants, and its patent landscape is distinctive because DAC is not a single technology but a set of competing capture routes, each with its own chemistry and energy profile. DAC removes carbon dioxide directly from the atmosphere, where it is present at very low concentration, which makes the capture step energy-intensive and puts a premium on the materials and processes that do it efficiently. The routes divide into distinct regions of patenting: solid amine adsorption, in which amine groups grafted onto porous supports capture carbon dioxide at ambient temperature and release it with modest heat, suiting modular designs;¹ liquid alkaline absorption, in which a potassium or sodium hydroxide solution captures carbon dioxide and is regenerated through a high-temperature calcination loop, suiting large centralized plants; mineralization, which binds carbon dioxide to minerals at low cost but slow kinetics; and electrified regeneration, in which electricity rather than heat releases the captured carbon dioxide.²,³,⁴,⁵,⁶ Cutting across the routes are the sorbent and solvent chemistry, the air-contactor and module design, including structured-honeycomb contactors that lower pressure drop,⁷ the regeneration step, and sorbent durability. Because a viable plant depends on several of these layers, freedom-to-operate and white space analysis must span the routes and the layers together.

The landscape is being pulled forward by policy and by corporate demand. In the United States, the Department of Energy's Regional Direct Air Capture Hubs program, funded at roughly $3.5 billion under the Infrastructure Investment and Jobs Act, is standing up four commercial-scale hubs, each required to capture and store or utilize at least one million tonnes of carbon dioxide a year, alongside a network of feasibility-phase projects; the South Texas hub is among those that have advanced through award negotiations.⁸ The routes sit at different stages: solid amine systems are the most widely deployed in modular form, liquid alkaline systems operate at the largest single-plant scale, mineralization offers low capital cost at slow kinetics, and electrified regeneration is a fast-moving research and development area that the literature identifies as a promising near-term direction for solid-sorbent systems, spanning Joule-heated fiber sorbents,²,³ electrochemically mediated amine regeneration,⁴ redox-based electro-swing approaches,⁵ and microwave and induction heating.⁶,⁹ Across the Cypris corpus of more than 500 million patents and scientific papers, the DAC solid-sorbent set holds on the order of 887 families and grew from about 22 in 2021 to roughly 173 in 2024, with the most active assignees led by Robert Bosch, Siemens Energy, Shell, W.L. Gore, and X Development, and the United States, Germany, and China leading on geography; 2025 and 2026 counts are partial because of the publication lag.

The strategic question is which route and layer to back, and the white space sits where energy and durability are hardest. Sorbents with high carbon dioxide capacity, fast kinetics, and long cycle life are the central materials problem, and oxidative and humidity-driven degradation of amine sorbents is a persistent, high-value target,¹,¹⁰ with in-situ vapor and steam purges among the strategies that promote regeneration and stability.¹⁰ Regeneration is the other decisive layer, because it dominates energy use, so electrified and low-temperature regeneration that can run on renewable power is a fast-moving, comparatively open area,²,³,⁴,⁵,⁶ and cost-per-tonne remains a projection subject to techno-economic modeling.¹¹ Contactor and module engineering that cuts pressure drop and capital cost,⁷ and mineralization approaches that speed up kinetics, are further distinct layers. Reading the landscape by route, material, and step, and tracking both the patents and the underlying materials research, is what separates a crowded region from an open one.

Where the DAC white space is

Durable, high-capacity sorbents. Sorbents with high capacity, fast kinetics, and resistance to oxidative and humidity-driven degradation are the central materials problem and a high-value layer.¹,¹⁰

Electrified and low-temperature regeneration. Regeneration that uses electricity or low-grade heat, spanning Joule-heated, electrochemical, electro-swing, microwave, and induction approaches, is a fast-moving, comparatively open layer.²,³,⁴,⁵,⁶,⁹

Air-contactor and module engineering. Contactor and module designs, including structured honeycombs, that cut pressure drop, energy use, and capital cost are where deployment economics are decided.⁷

Mineralization kinetics. Approaches that speed up the low-cost but slow mineralization route are a distinct, differentiated area.

System and heat integration. Integrating DAC with low-grade waste heat, renewable power, and storage or utilization is a strategically important system layer.

How AI-powered landscape and white space analysis helps

Resolving a landscape that spans several capture routes, each with its own chemistry and energy profile, requires more than keyword search. AI-powered analysis addresses this with semantic search that clusters activity by route, material, and step across varied terminology, attribution that normalizes developer and academic filers to canonical entities, and continuous monitoring that keeps pace with a policy-driven surge. Because DAC advances appear in materials and separations literature before they are patented, reading both patents and literature gives the earliest signal of where low-energy, durable systems are emerging.

Where Cypris fits

Cypris runs patent landscape and white space analysis for multi-route materials fields such as direct air capture across a corpus of more than 500 million patents and scientific papers, organized through a proprietary R&D ontology. The ontology clusters activity by route, solid amine, liquid alkaline, mineralization, and electrified regeneration, and by layer, sorbent and solvent chemistry, contactor and module, and regeneration, and normalizes developer and academic filers to canonical entities, so a team can resolve which routes and layers are crowded and which remain open as white space, and can track new entrants as the field scales. Semantic search across patents and scientific literature connects filings to the underlying materials and separations research, which is where DAC advances appear first. Cypris Q, the platform's agentic layer, lets teams run landscape and white space analysis conversationally and chain the clustering, attribution, and gap analysis, and Agentic Monitoring tracks a defined route over time and flags new patents and papers as they publish. Cypris provides enterprise API partnerships with OpenAI, Anthropic, and Google, and is built with enterprise-grade security. Cypris serves hundreds of enterprise customers across pharmaceuticals, chemicals, advanced materials, energy, and other regulated industries.

FAQ

What is the direct air capture patent landscape? The direct air capture patent landscape is the set of patents covering technologies that remove carbon dioxide directly from ambient air. It divides across competing routes, solid amine adsorption, liquid alkaline absorption, mineralization, and electrified regeneration, plus sorbent, contactor, and regeneration IP. Each route is a distinct region of patenting.

Why is DAC energy-intensive? DAC is energy-intensive because carbon dioxide is present in air at very low concentration, so moving large volumes of air and then releasing the captured carbon dioxide from the sorbent or solvent takes significant energy, most of it in regeneration. Reducing that energy is the central challenge. It is why sorbent and regeneration IP is so valuable.

What routes does the DAC landscape cover? The landscape covers solid amine adsorption, which suits modular designs; liquid alkaline absorption, which suits large centralized plants; mineralization, which is low-cost but slow; and electrified regeneration, an emerging area. Each has distinct chemistry, energy needs, and maturity. Freedom-to-operate and white space analysis must treat them separately.

How large is the US DAC hubs program? The US Department of Energy's Regional Direct Air Capture Hubs program is funded at roughly $3.5 billion under the Infrastructure Investment and Jobs Act and is standing up four commercial-scale hubs, each required to capture and store or utilize at least one million tonnes of carbon dioxide a year. The South Texas hub is among those in award negotiations. It is the largest public DAC funding to date.

Where is the white space in DAC? The white space includes durable, high-capacity sorbents, electrified and low-temperature regeneration, air-contactor and module engineering, mineralization kinetics, and system and heat integration. The most deployed solid amine and liquid alkaline grounds are comparatively crowded. The most open, high-value opportunities are in sorbent durability and low-energy regeneration.

Why is regeneration such an important layer? Regeneration is important because releasing the captured carbon dioxide from the sorbent or solvent dominates a DAC plant's energy use and cost. Approaches that regenerate at lower temperature or with electricity, and can run on renewable power, directly improve economics and carbon balance. That makes regeneration a decisive, actively patented layer.

What software helps analyze the direct air capture patent landscape? Software for the DAC landscape should cluster activity by route and process layer, resolve developer and academic filers to canonical owners, search patents and scientific literature semantically, and monitor a policy-driven field continuously. Cypris does this across more than 500 million patents and scientific papers using a proprietary R&D ontology, semantic search, Cypris Q, and Agentic Monitoring.

Which teams use direct air capture patent landscape analysis? Direct air capture patent landscape analysis is used by R&D, innovation, IP, and strategy teams at carbon-removal, materials, chemicals, and energy companies, and their partners, as well as investors and policymakers. It informs which route to back, where to file, and where competitors are concentrated. Cypris serves hundreds of enterprise customers across advanced materials, energy, chemicals, and other regulated industries.

Endnotes

1. Sun, X., et al. (2023). Chemisorption and regeneration of amine-based CO2 sorbents in direct air capture. Materials Today Sustainability, 22. https://doi.org/10.1016/j.mtsust.2023.100453
2. Realff, M. J., Jones, C. W., Lively, R. P., et al. (2023). Sorbent-coated carbon fibers for direct air capture using electrically driven temperature swing adsorption. Joule, 7(7). https://doi.org/10.1016/j.joule.2023.05.016
3. Park, H., et al. (2025). Design of electrified fiber sorbents for direct air capture with electrically driven temperature-vacuum-swing adsorption. Advanced Materials, 37. https://doi.org/10.1002/adma.202504542
4. Stern, M. C., & Hatton, T. A. (2013). Bench-scale demonstration of CO2 capture with electrochemically mediated amine regeneration (EMAR). RSC Advances, 3. https://doi.org/10.1039/c3ra46774k
5. Kim, S., Kang, J. S., & Hatton, T. A. (2021). Redox-responsive sorbents and mediators for electrochemically based CO2 capture. Current Opinion in Green and Sustainable Chemistry, 30. https://doi.org/10.1016/j.cogsc.2021.100504
6. van der Wal, K., van Schagen, T. N., & Brilman, D. W. F. (2021). Through-flow microwave-based regenerator for sorbent-based direct air capture. Chemical Engineering Journal Advances, 8. https://doi.org/10.1016/j.ceja.2021.100187
7. Boger, T., et al. (2026). Steam-based vacuum-temperature-swing adsorption with honeycomb amine sorbents for direct air capture. Carbon Capture Science & Technology, 20. https://doi.org/10.1016/j.ccst.2026.100654
8. U.S. Department of Energy, Office of Clean Energy Demonstrations. Regional Direct Air Capture Hubs. https://www.energy.gov/oced/regional-direct-air-capture-hubs
9. Thompson, S., et al. (2024). Magnetic-nanoparticle-induced sorbent regeneration for direct air capture. AIChE Journal, 70. https://doi.org/10.1002/aic.18500
10. Webley, P. A., Li, S., Hu, X., et al. (2026). Promoting regeneration of PEI-impregnated sorbents through in-situ vapor purge for direct air capture. AIChE Journal, 72. https://doi.org/10.1002/aic.70571
11. Pirngruber, G. D., et al. (2013). Theoretical analysis of energy consumption of TSA post-combustion capture with solid sorbents. International Journal of Greenhouse Gas Control, 14. https://doi.org/10.1016/j.ijggc.2013.01.010

‍

An ontology is a formal, machine-readable specification of the concepts in a domain and the relationships among them. The term has a precise meaning in knowledge representation: an explicit specification of a conceptualization,¹ that is, a defined vocabulary of entity types, attributes, and relations, together with constraints on how they may be combined. This distinguishes an ontology from a flat taxonomy, which only arranges terms hierarchically; an ontology also encodes non-hierarchical relations, such as a material being used in a process or a method being applied to a claim. In R&D and patent intelligence, the ontology defines the domain schema: the technologies, materials, methods, claims, organizations, and research areas that matter, and the relationship types that connect them.²

A knowledge graph instantiates that schema over real data. It represents information as a graph of nodes and typed edges, commonly expressed as subject-predicate-object triples, linking specific patents, scientific papers, assignees, inventors, technologies, and materials as connected entities rather than isolated documents. Building the graph requires several engineering steps that determine its quality: named-entity recognition and relation extraction to convert unstructured patent and paper text into triples; entity resolution to normalize the many surface forms of an organization, inventor, or compound to a single canonical node; and provenance tracking so every assertion in the graph traces back to the source document that supports it. The result is a structure that can be queried declaratively, for example with a graph query language, and that supports multi-hop traversal, so a question can follow chains of relationships rather than matching a single string.

This structure matters because patents and scientific literature become intelligence only when their relationships are made explicit. A ranked list of relevant documents does not state how a technology area is organized, which organizations are active, how research connects to patents, or where the graph is sparse. An ontology-backed knowledge graph makes those relationships first-class and queryable. A team can ask how two technologies relate, which body of research underpins a patent cluster, which assignees co-file in an area, or where a domain is unclaimed, and receive an answer computed over structured connections rather than assembled by reading.

The 2026 relevance is that structured knowledge is the most reliable way to ground generative AI. Large language models produce fluent output but can assert unsupported claims when they generate from parametric memory over unstructured text. Retrieval-augmented generation (RAG), which conditions a model's output on retrieved external evidence, was introduced to address this and improves factual accuracy on knowledge-intensive tasks.³,⁴ Graph retrieval-augmented generation (GraphRAG) extends RAG by retrieving connected subgraphs rather than isolated passages, so the model reasons over entities and their relationships and can answer questions that require traversing multiple hops.⁵,⁶ Grounding a system on an ontology-backed knowledge graph constrains its outputs to real, connected entities, which is essential for patent and R&D work where every conclusion must trace to actual patents and papers, and where retrieval quality directly governs the reliability of downstream generation.⁷ It is also what makes agentic workflows dependable: an agent reasoning over a structured, provenance-tracked graph produces results a team can verify against sources.

What an ontology and knowledge graph add to patent intelligence

Multi-hop reasoning over relationships. A knowledge graph answers relational and multi-hop questions, such as how two technologies connect through shared materials or which research a patent cluster builds on, rather than only returning documents that match a query string.

Concept-organized semantic search. Dense semantic retrieval returns conceptually relevant documents; the ontology organizes that retrieval within a domain schema, improving both recall and the interpretability of results by grouping them under defined concepts.

White space analysis. White space analysis depends on clustering activity by concept to expose genuine gaps. Clustering patents and papers over the ontology's relationship structure exposes real conceptual sparsity rather than gaps that are artifacts of divergent terminology.

Entity-resolved attribution and competitive intelligence. Entity resolution normalizes assignee and inventor variants to canonical nodes, which lets the graph attribute filings and research accurately and build co-assignee and citation networks rather than a document list.

Provenance-grounded AI. The ontology and knowledge graph give AI agents a structured, provenance-tracked foundation to reason over, which improves the accuracy of agentic analysis and makes its results traceable to the specific patents and papers that support them.

Where Cypris fits

Cypris organizes a corpus of more than 500 million patents and scientific papers through a proprietary R&D ontology. That ontology is the core of the platform: it defines how technologies, claims, materials, methods, and research relate, so Cypris reasons over an entity-resolved relationship structure rather than only matching keywords. This structure powers dense semantic retrieval organized by concept, white space analysis that exposes genuine conceptual gaps, and competitive intelligence that attributes activity to canonical organizations and technology areas. Cypris Q, the platform's agentic layer, reasons over this provenance-tracked foundation, which is what makes its multi-step analysis both reliable and traceable to real patents and papers, consistent with graph-grounded retrieval approaches. Agentic Monitoring tracks a technology area over time against the same structure. Cypris provides enterprise API partnerships with OpenAI, Anthropic, and Google, so AI agents can query the structured corpus programmatically, and it is built with enterprise-grade security. Cypris serves hundreds of enterprise customers across pharmaceuticals, chemicals, advanced materials, energy, and other regulated industries.

FAQ

What is an ontology in R&D and patent intelligence?

An ontology in R&D and patent intelligence is a formal, machine-readable specification of the concepts in the domain and the relationships among them, defined as an explicit specification of a conceptualization. It sets out the entity types, such as technologies, materials, methods, and claims, and the relations that connect them. This lets a platform reason over connections between patents and scientific literature rather than treating documents as isolated.

How is an ontology different from a taxonomy?

An ontology differs from a taxonomy in expressiveness. A taxonomy arranges terms in a hierarchy, while an ontology also encodes non-hierarchical, typed relationships and constraints, such as a material being used in a process. This richer structure is what allows multi-hop reasoning across patents and research rather than simple category lookup.

What is a knowledge graph for patents?

A knowledge graph for patents represents patents, scientific papers, assignees, inventors, technologies, and materials as nodes connected by typed edges, commonly expressed as subject-predicate-object triples. It applies an ontology's schema to real data so relationships are explicit and queryable. This turns a document collection into a structure that supports declarative queries and multi-hop traversal.

How is a knowledge graph built from patent text?

A knowledge graph is built from patent text through named-entity recognition and relation extraction to convert unstructured text into triples, entity resolution to normalize variant names to canonical nodes, and provenance tracking so each assertion links back to its source document. The quality of these steps determines the reliability of the graph. Poor entity resolution, for example, fragments an organization across many nodes and distorts attribution.

Why do knowledge graphs matter for AI in patent research?

Knowledge graphs matter for AI in patent research because they ground generative models on real, connected entities, which improves accuracy and traceability. A model generating from unstructured text alone can assert unsupported claims, whereas one conditioned on a provenance-tracked graph constrains its answers to actual patents and papers. This is essential where conclusions must be verifiable.

What is GraphRAG and how does it differ from standard RAG? GraphRAG is graph retrieval-augmented generation. Standard RAG retrieves isolated text passages to condition a model's output, while GraphRAG retrieves connected subgraphs, so the model reasons over entities and their relationships and can answer multi-hop questions. This suits patent intelligence, where questions often require traversing links between technologies, research, and organizations.

How does an ontology improve white space analysis?

An ontology improves white space analysis by clustering patents and papers over defined relationships rather than by exact keywords, which exposes genuine conceptual gaps instead of gaps that are only artifacts of differing terminology. Because the sparsity reflects the domain structure, the identified white space corresponds to unclaimed technical territory. Cypris organizes its corpus of more than 500 million patents and scientific papers through a proprietary R&D ontology for this purpose.

How do knowledge graphs reduce AI hallucination in patent work?

Knowledge graphs reduce AI hallucination in patent work by constraining a model's outputs to real, connected entities with tracked provenance rather than letting it generate from unstructured text. Retrieval-augmented approaches, and graph-based retrieval in particular, condition generation on retrieved evidence, which improves factual accuracy and lets conclusions be traced to sources. This makes results verifiable against the underlying patents and papers.

Is a knowledge graph the same as a vector database?

A knowledge graph is not the same as a vector database. A vector database supports semantic similarity search over embeddings, while a knowledge graph represents explicit, typed relationships between entities. They are complementary: dense retrieval finds relevant documents, and the graph structures how those documents and entities relate. Cypris combines semantic retrieval with a proprietary R&D ontology.

Which teams benefit from ontology-based patent intelligence?

‍Ontology-based patent intelligence benefits R&D, innovation, IP, and strategy teams that need to understand how technologies relate, attribute activity to organizations, and find genuine white space. It is most valuable in research-intensive industries such as pharmaceuticals, chemicals, advanced materials, and energy. Cypris serves hundreds of enterprise customers across these industries.

Endnotes

1. Gruber, T. R. (1993). A translation approach to portable ontology specifications. Knowledge Acquisition. https://doi.org/10.1006/knac.1993.1008
2. Gruber, T. R. (1995). Toward principles for the design of ontologies used for knowledge sharing. International Journal of Human-Computer Studies. https://doi.org/10.1006/ijhc.1995.1081
3. Lewis, P. et al. (2020). Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks. NeurIPS.
4. Gao, Y. et al. (2023). Retrieval-Augmented Generation for Large Language Models: A Survey. arXiv:2312.10997. https://doi.org/10.48550/arxiv.2312.10997
5. Procko, T. & Ochoa, O. (2024). Graph Retrieval-Augmented Generation for Large Language Models: A Survey. https://doi.org/10.1109/aixset62544.2024.00030
6. Han, S. et al. (2025). A Survey of Graph Retrieval-Augmented Generation for Customized Large Language Models. arXiv:2501.13958. https://doi.org/10.48550/arxiv.2501.13958
7. Chen, J. et al. (2024). Benchmarking Large Language Models in Retrieval-Augmented Generation. AAAI. https://doi.org/10.1609/aaai.v38i16.29728

‍

Patent filings are a leading indicator of competitor R&D direction, and the lead time is a structural consequence of how the patent system operates. An application is filed at its priority date, well before the corresponding product reaches the market, and under the standard 18-month publication rule reflected in USPTO practice and PCT Article 21,⁵ it is not published until roughly eighteen months after that priority date. The interval between when a competitor commits R&D and when the public can observe it is therefore built into the system. The International Energy Agency treats patenting as a leading indicator of technological change in its innovation analysis,¹ and the same logic holds across sectors: a competitor's published filings reveal committed R&D direction ahead of the market, and studies of the linkage between scientific publication and patenting document a measurable lag between the two that compounds the observable lead time.² For R&D and competitive intelligence teams, this makes patents one of the most reliable forward-looking competitive signals available.

Reading that signal well requires structured analysis rather than filing counts, and several technical steps determine its accuracy. First, the unit of analysis should be the patent family, not the individual document, because a single invention generates multiple applications across jurisdictions; counting documents rather than families overstates activity and double-counts international coverage. Second, filings must be located in the technology space using classification codes, principally the Cooperative Patent Classification and International Patent Classification systems, which assign standardized technology categories independent of the applicant's terminology. Third, activity must be attributed through assignee disambiguation, normalizing the many name variants, subsidiaries, and transliterations of an organization to a single canonical entity, because unresolved assignee names fragment a competitor's portfolio and distort the picture. Fourth, the analysis should read the trend over time rather than the latest counts, because the most recent eighteen-to-twenty-four months of data are systematically under-represented by publication lag, so apparent recent declines are usually artifacts rather than real slowdowns.

Two network structures add depth beyond volume. Forward and backward citation analysis situates a competitor's filings in the flow of prior art: backward citations reveal the foundations a filing builds on, and forward citations indicate influence and where a technology is being extended. Co-assignee and knowledge-search network analysis reveals partnerships, academic-industry pipelines, and the coupling between organizations, which shape a competitor's future direction; network-embedding methods over these structures are an established competitive-intelligence technique.³ Scientific literature strengthens the signal further, because research is published before it is patented and patents are filed before products ship, so combining the two sources extends the observable lead time; the scientific footprint within a competitor's filings can be traced through their non-patent references.⁴

What competitor filings reveal

Technology direction. The classification areas where a competitor is filing show where R&D is being committed, often well before those commitments appear in products.

Intensity and momentum. The distribution and rate of change of filing activity across technology areas indicate priorities, and shifts in filing momentum signal changes in strategy earlier than raw counts.

Adjacent moves. Filings in classifications adjacent to a competitor's current products can signal diversification or expansion before it is announced.

Research foundations. The non-patent references and scientific literature a competitor's filings build on show the research base behind their direction, and rising related research is an earlier signal still.

Collaboration structure. Co-assignee patterns and citation coupling reveal partnerships and academic-industry pipelines; network analysis of these relationships is an established competitive-intelligence method.³

How to read competitor R&D direction

Define the competitors and the technology space, scoping the latter with classification codes so the boundary is standardized and reproducible.

Resolve assignees to canonical entities and aggregate to the patent-family level, so activity is attributed accurately and international coverage is not double-counted.

Cluster filings by concept using semantic analysis over the classification and text, so related work groups together regardless of terminology.

Analyze filing momentum as a time series, discounting the most recent windows for publication lag, since direction is visible in trends rather than in the latest bar.

Connect filings to their non-patent references and to the scientific literature, to extend the lead time and expose the research foundations.

Monitor continuously, because competitor direction is revealed by how activity shifts, and continuous monitoring captures those shifts as they publish.

Where Cypris fits

Cypris supports competitive intelligence across a corpus of more than 500 million patents and scientific papers, organized through a proprietary R&D ontology. The ontology and its entity resolution are what turn filings into direction: they normalize assignees to canonical organizations, aggregate to the family level, and cluster activity by concept, so a team sees where a competitor is moving rather than a list of documents. Dense semantic search across patents and scientific literature connects filings to their research foundations, which extends the lead time on the signal, and citation and co-assignee structures expose collaboration and influence. Cypris Q, the platform's agentic layer, lets teams analyze competitor direction conversationally and chain the attribution, clustering, and time-series analysis. Agentic Monitoring is central to this use case: it tracks defined competitors and technology areas over time and flags new filings and research as they publish, so competitive intelligence is continuous rather than a one-time report. Cypris provides enterprise API partnerships with OpenAI, Anthropic, and Google, so AI agents can query the corpus programmatically, and it is built with enterprise-grade security. Cypris serves hundreds of enterprise customers across pharmaceuticals, chemicals, advanced materials, energy, and other regulated industries.

FAQ

How do patent filings reveal competitor R&D direction?

Patent filings reveal competitor R&D direction because an application is filed at its priority date, before the product ships, and is published only about eighteen months later under the standard publication rule. This built-in lag means published filings show committed R&D ahead of the market. Reading the direction requires attributing filings to competitors and technology areas and analyzing where activity concentrates and shifts.

What is the 18-month publication rule?

The 18-month publication rule is the standard practice, reflected in USPTO procedure and PCT Article 21, under which a patent application is published approximately eighteen months after its earliest priority date. It creates a predictable interval between filing and public visibility. It is also why the most recent windows of filing data are under-represented and should not be read as slowdowns.

Why analyze patent families instead of individual documents?

Analyzing patent families instead of individual documents avoids double-counting, because a single invention generates multiple applications across jurisdictions. Counting documents overstates activity and conflates international coverage with genuine volume. The family is the correct unit for measuring how much distinct R&D a competitor is committing.

What role do classification codes play?

Classification codes, principally the Cooperative Patent Classification and International Patent Classification systems, assign standardized technology categories to filings independent of the applicant's wording. They let an analyst locate and compare activity in a technology space reproducibly. This is more reliable than keyword filtering, which varies with drafting style.

Why is assignee disambiguation important?

Assignee disambiguation is important because organizations appear under many name variants, subsidiaries, and transliterations, and unresolved names fragment a competitor's portfolio across multiple entities. Normalizing these to a single canonical entity is what makes attribution and trend analysis accurate. Poor disambiguation systematically distorts competitive intelligence.

How do citation networks support competitive intelligence?

Citation networks support competitive intelligence by situating filings in the flow of prior art. Backward citations reveal the foundations a filing builds on, and forward citations indicate influence and where a technology is being extended. Co-assignee and knowledge-search network analysis additionally reveals partnerships and academic-industry pipelines.

Why combine patents with scientific literature?

Combining patents with scientific literature extends the observable lead time, because research is published before it is patented and patents precede products. Rising research associated with a competitor, followed by early filings, is an earlier and stronger signal than filings alone. The scientific footprint within filings can be traced through their non-patent references.

Why not just count competitor patent filings?

Counting filings alone is misleading because recent counts are depressed by publication lag and raw volume does not indicate direction. The informative signal is which classification areas activity concentrates in and how that distribution changes over time. Family-level aggregation, classification analysis, and time-series momentum are what reveal direction.

Why is continuous monitoring important for competitive intelligence? Continuous monitoring is important because competitor direction is revealed by how activity changes, which a one-time report cannot capture, and because new filings and research publish constantly. A shift in a competitor's focus is only visible if the area is tracked over time. Cypris uses Agentic Monitoring to track competitors and technology areas and flag new activity as it publishes.

Which teams read competitor R&D direction from patents?

Reading competitor R&D direction from patents is done by competitive intelligence, R&D, innovation, strategy, and corporate development teams that need forward-looking awareness of competitor moves. It is most valuable in research-intensive industries such as pharmaceuticals, chemicals, advanced materials, and energy. Cypris serves hundreds of enterprise customers across these industries.

Endnotes

1. International Energy Agency (2026). The State of Energy Innovation 2026. https://www.iea.org/reports/the-state-of-energy-innovation-2026
2. Fukuzawa, N. & Ida, T. (2015). Science linkages between scientific articles and patents for leading scientists in the life and medical sciences field. Scientometrics. https://doi.org/10.1007/s11192-015-1795-z
3. Yang, X. et al. (2024). Predicting patent transaction behaviour based on embedded features of knowledge search networks. Journal of Knowledge Management. https://doi.org/10.1108/jkm-12-2023-1220
4. Callaert, J., Grouwels, J. & Van Looy, B. (2011). Delineating the scientific footprint in technology: identifying scientific publications within non-patent references. Scientometrics. https://doi.org/10.1007/s11192-011-0573-9
5. World Intellectual Property Organization, PCT Article 21 (International Publication), and USPTO Manual of Patent Examining Procedure, on patent publication timing.

‍

Wide-bandgap power semiconductors have become one of the most strategically important and most litigated areas in electronics, and their patent landscape is distinctive because value and risk are spread across the full stack from crystal to module. Silicon carbide and gallium nitride switch faster, tolerate higher voltages and temperatures, and lose less energy than conventional silicon, which is why they are central to electric-vehicle drivetrains, fast charging, solar and grid power conversion, and the power delivery of AI data centers; peer-reviewed reviews document these comparative properties and the trade-offs between the two materials.¹,² The intellectual property divides across several regions, each with different owners and maturity: the substrate and bulk-crystal growth that produces the raw material; the epitaxy that grows the active layers, including gallium nitride on silicon; the device design, such as the transistor and diode structures, whose failure and reliability modes are a distinct engineering concern;³ the packaging and thermal-management technologies that manage heat and switching losses;² and the application-level integration into drivetrains and power systems, where gallium-nitride high-electron-mobility transistors are increasingly used.⁴ Because a competitive product depends on several of these layers, freedom-to-operate is a multi-layer analysis rather than a single clearance.

The landscape is defined by intense litigation. Established wide-bandgap developers with deep substrate and device portfolios have asserted their patents against newer entrants, and the disputes have moved through trade-enforcement bodies and the federal courts. In a US International Trade Commission investigation into certain semiconductor devices, a final initial determination issued on December 2, 2025 found a violation as to one asserted patent and no violation as to a second, with the Commission determining to review the decision in part, a proceeding that remained under Commission review as of early 2026.⁶ In parallel, a wide-bandgap developer has asserted foundational gallium-nitride and silicon-carbide patents against a competitor in US federal court, with the case active on the district-court docket.⁷ Parallel proceedings are underway in other jurisdictions. This pattern signals a maturing field in which foundational substrate, epitaxy, and device patents create real barriers, and in which a single infringement finding in a key market can reshape a competitor's access to it. The ownership picture is concentrated but contested: across the Cypris corpus of more than 500 million patents and scientific papers, the most active assignees in the gallium-nitride power-device set are incumbent integrated device manufacturers, led by Japanese and European firms such as Mitsubishi Electric, Fuji Electric, Infineon, Toshiba, and Rohm, together with foundries and a small number of US developers, of which one substrate-and-device specialist is the clearest pure-play; the set holds on the order of 80,000 families on an indicative basis and grew roughly 2.4 times between 2018 and 2024, with China and the United States leading on assignee geography, followed by Japan, Germany, and South Korea. Because applications publish about eighteen months after filing, the most recent device and packaging filings are under-represented (2025 counts are partial), so the current frontier is more active than granted-patent counts suggest.

The strategic question is which layer to own, and the answer differs by material. In silicon carbide, the substrate and bulk-crystal layer is a durable barrier because high-quality crystal growth is difficult and capital-intensive, so much of the defensible value sits upstream. In gallium nitride, where devices are often grown on silicon wafers, the contested ground is more in epitaxy, device architecture, and packaging, and the litigation has concentrated there. Across both, packaging and thermal management are rising in importance as switching speeds increase,² and ultra-wide-bandgap approaches are an emerging frontier beyond today's materials.⁵ Reading the landscape by material, layer, and owner, and tracking the live proceedings, is what separates a workable position from a blocked one.

What creates FTO risk in wide-bandgap power semiconductors

Substrate and crystal-growth claims. These cover bulk silicon-carbide crystal and wafer production, a capital-intensive, upstream layer that is a durable barrier in silicon carbide.

Epitaxy claims. These cover the growth of active layers, including gallium nitride on silicon, a heavily contested layer central to gallium-nitride litigation.

Device-design claims. These cover transistor and diode structures and their edge terminations and gate designs, whose reliability and failure modes are a frequent center of disputes.³

Packaging and module claims. These cover thermal management, interconnection, and module construction, a layer rising in importance as switching speeds increase.²

Application-integration claims. These cover integration into drivetrains, chargers, and power systems, so a device can be free at the component level and constrained in a specific application.⁴

How AI-powered landscape and FTO analysis helps

A multi-layer, cross-border, heavily litigated landscape is beyond manual clearance. AI-powered analysis addresses this with semantic search that retrieves relevant substrate, epitaxy, device, and packaging claims regardless of terminology, attribution that normalizes incumbent and challenger owners to canonical entities across jurisdictions, claim-level analysis that separates the layers, and continuous monitoring that tracks new filings and the live disputes. Because wide-bandgap advances appear in scientific literature before they are patented, reading both patents and literature gives earlier warning of where the field is extending.

Where Cypris fits

Cypris runs patent landscape and freedom-to-operate analysis for multi-layer, litigated fields such as wide-bandgap power semiconductors across a corpus of more than 500 million patents and scientific papers, organized through a proprietary R&D ontology. The ontology clusters the landscape by material and layer, substrate, epitaxy, device, packaging, and application, and normalizes owners to canonical entities across jurisdictions, so a team sees how rights are distributed between incumbents and challengers rather than a flat list. Semantic search across patents and scientific literature surfaces relevant claims regardless of terminology and connects filings to the underlying materials and device research, which is where next-generation structures emerge first. Cypris Q, the platform's agentic layer, lets teams run landscape and FTO analysis conversationally and chain the attribution, clustering, and claim-level analysis across layers, and Agentic Monitoring tracks the landscape over time and flags new filings and developments as they publish. Cypris provides enterprise API partnerships with OpenAI, Anthropic, and Google, and is built with enterprise-grade security. Cypris serves hundreds of enterprise customers across pharmaceuticals, chemicals, advanced materials, energy, and other regulated industries.

FAQ

Why are wide-bandgap power semiconductors a patent hotspot? Wide-bandgap power semiconductors are a patent hotspot because silicon carbide and gallium nitride enable more efficient power electronics for electric vehicles, fast charging, renewables, and AI data centers, creating a large and fast-growing market. Value and risk are spread across substrate, epitaxy, device, and packaging layers. That breadth, plus intense competition, has produced heavy litigation.

What layers does the SiC and GaN landscape cover? The landscape covers substrate and bulk-crystal growth, epitaxy, device design, packaging and modules, and application integration. In silicon carbide the substrate layer is a durable upstream barrier, while in gallium nitride the contested ground is more in epitaxy, device design, and packaging. Freedom-to-operate must span the relevant layers for each material.

Why is this field so heavily litigated? The field is heavily litigated because foundational substrate, epitaxy, and device patents create real barriers, incumbents hold deep portfolios, and fast-growing challengers are building their own. Disputes have moved through the US International Trade Commission and the federal courts, with a December 2025 ITC determination finding a violation as to one patent and none as to another, and a separate district-court case over foundational gallium-nitride and silicon-carbide patents. A single ruling in a key market can reshape competitive access.

Where is the white space in wide-bandgap semiconductors? The white space includes packaging and thermal management as switching speeds rise, device architectures that design around crowded structures, gallium-nitride epitaxy and integration approaches, and application-level integration into drivetrains and power systems. The silicon-carbide substrate layer is a durable barrier held by incumbents. The higher-value opportunities are in packaging, device design, and integration.

How do silicon carbide and gallium nitride differ in the patent picture? They differ because silicon carbide value concentrates upstream, in difficult, capital-intensive crystal growth, while gallium nitride, often grown on silicon, concentrates contested IP in epitaxy, device architecture, and packaging. The litigation patterns reflect this. Freedom-to-operate strategy should therefore be tailored to the material.

Why does wide-bandgap analysis need scientific literature? Wide-bandgap analysis needs scientific literature because materials, device, and packaging advances appear in research before they are patented, so the literature gives the earliest signal. Analyzing patents alone gives a lagging view. Cypris analyzes both across more than 500 million patents and scientific papers.

What software helps analyze the SiC and GaN patent landscape? Software for the wide-bandgap landscape should cluster activity by material and layer, resolve incumbent and challenger owners to canonical entities across jurisdictions, search patents and scientific literature semantically, and monitor active litigation and new filings continuously. Cypris does this across more than 500 million patents and scientific papers using a proprietary R&D ontology, semantic search, Cypris Q, and Agentic Monitoring.

Which teams use wide-bandgap patent landscape analysis? Wide-bandgap patent landscape analysis is used by R&D, IP, and strategy teams at semiconductor, automotive, power-electronics, and energy companies, as well as investors assessing the sector. Because the field is litigated across the full stack and across borders, structured analysis is essential. Cypris serves hundreds of enterprise customers across advanced materials, energy, and other research-intensive industries.

Endnotes

1. Stanley, C., Herzog, S., Viewegh, M., Biggerstaff, T., et al. (2026). Wide bandgap semiconductors for power electronics: comparative properties, applications, and reliability of GaN and SiC devices. Hardware, 4(1). https://doi.org/10.3390/hardware4010006
2. Kim, J., Bae, S., Han, S., & Park, S. (2025). Thermal management of wide-bandgap power semiconductors: strategies and challenges in SiC and GaN power devices. Electronics, 14(21), 4193. https://doi.org/10.3390/electronics14214193
3. Romano, G., Imburgia, A., Ala, G., Rizzo, G., et al. (2025). Comprehensive review of wide-bandgap devices: SiC MOSFET and its failure modes affecting reliability. Physchem, 5(1). https://doi.org/10.3390/physchem5010010
4. Rusli, M., Jarndal, A., & Hamza, K. H. (2026). GaN HEMTs for electric vehicle power electronics: device architectures, reliability and next-generation wide-bandgap opportunities. Energies, 19(7), 1752. https://doi.org/10.3390/en19071752
5. Adekunle, A. (2025). Review of ultra wide bandgap GaN-based HEMTs for high-efficiency power conversion. International Journal of Future Engineering Innovations, 2(3). https://doi.org/10.54660/ijfei.2025.2.3.77-83
6. U.S. International Trade Commission. Certain semiconductor devices and products containing the same, Investigation No. 337-TA-1414 (Final Initial Determination, Dec. 2, 2025; Commission review in part). Federal Register / public-inspection record. https://public-inspection.federalregister.gov/2026-02297.pdf
7. Wolfspeed, Inc. v. Navitas Semiconductor Corporation, U.S. District Court for the District of Delaware, No. 1:24-cv-01038 (docket). https://www.courtlistener.com/docket/69457003/parties/wolfspeed-inc-v-navitas-semiconductor-corporation

‍

Small modular reactors and advanced nuclear designs have moved from concept toward licensing and deployment, and their patent landscape is distinctive because it spans several competing reactor families and a fuel supply chain that must be built alongside them. Where conventional nuclear plants are large, bespoke, and light-water-cooled, SMRs and advanced reactors are smaller, factory-built, and often use novel coolants, moderators, and fuels to achieve passive safety and flexible operation. The field divides into distinct regions of patenting: the reactor concepts themselves, including high-temperature gas-cooled reactors, molten-salt reactors, sodium-cooled fast reactors, integral light-water SMRs, and microreactors; the core, coolant, and moderator designs within each, from tristructural isotropic coated-particle fuel in gas-cooled reactors to fluoride fuel salts in molten-salt reactors;¹,²,³ the advanced fuels, especially TRISO particles and the high-assay low-enriched uranium, or HALEU, they require; the passive safety systems and control approaches that shut a reactor down without active intervention;⁵,⁶ and the modular manufacturing and construction methods that make factory production possible.⁴ Alongside these sit non-electric applications, from industrial heat and hydrogen to powering data centers. Because a viable design depends on several of these layers, and on a fuel supply that does not yet exist at scale, freedom-to-operate and white space analysis must span reactor, fuel, and manufacturing together.

The landscape is being reshaped by licensing progress and by a surge in demand. Regulators have modernized the framework for licensing advanced reactors, adopting a risk-informed, performance-based, and technology-inclusive rule intended to accommodate the range of SMR and advanced designs,⁷ while government programs are standing up a domestic supply of HALEU, the enriched fuel that many advanced reactors need and that has not been commercially produced at scale. Demand has accelerated sharply as technology companies contract for nuclear power to run data centers, drawing large investment. The competitive picture spans reactor developers pursuing different coolant and fuel choices, specialized fuel fabricators building TRISO and HALEU capacity, and established nuclear suppliers. This shows in the record: across the Cypris corpus of more than 500 million patents and scientific papers, the nuclear-anchored SMR, microreactor, TRISO, molten-salt, and HALEU set holds on the order of 2,017 families and grew from about 87 in 2020 to roughly 250 in 2024, with the most active assignees mixing large academic and institutional filers, led by Chinese universities and institutes, with reactor developers and fuel fabricators such as TerraPower, Mitsubishi Heavy Industries, BWXT, and Westinghouse and the US Department of Energy, and China ahead of the United States and South Korea on geography; a broad, unfiltered "microreactor" query returns far more families but is contaminated by unrelated chemical-microreactor art, so the nuclear-anchored figure is the defensible one. Because applications publish about eighteen months after filing, the most recent reactor and fuel filings are under-represented (2025 and 2026 counts are partial), so the current frontier is more active than granted-patent counts suggest.

The strategic question is which reactor family and layer to back, and the white space sits where technology and supply chain intersect. Advanced fuel, TRISO fabrication and the HALEU supply chain, is a critical, comparatively concentrated layer where scaling and cost are the barriers, so fabrication methods and fuel designs carry high value.¹ Coolant and materials that survive high temperatures and long fuel cycles, across gas, salt, and metal-cooled designs, passive safety systems, modular manufacturing and construction that lower cost and schedule, and non-electric applications such as process heat, hydrogen, and dispatchable power for data centers are all distinct, contested layers.²,³ Reading the landscape by reactor family, layer, and owner, and tracking both the patents and the underlying nuclear-engineering research, is what separates a crowded region from an open one.

Where the advanced-nuclear white space is

Advanced fuel and HALEU supply. TRISO fabrication methods and the high-assay low-enriched uranium supply chain are a critical, concentrated layer where scaling and cost are the barriers.¹

Coolant, moderator, and materials. Coolants and materials that survive high temperatures and long fuel cycles, across gas, salt, and metal-cooled designs, are a distinct, high-value layer.²,³

Passive safety systems. Systems and control approaches that shut a reactor down without active intervention are a differentiating capability central to SMR value.⁵,⁶

Modular manufacturing and construction. Factory-built modules and construction methods that lower cost and schedule are where the economic case is decided.⁴

Non-electric applications. Industrial heat, hydrogen production, and dispatchable power for data centers open distinct application and integration IP.

How AI-powered landscape and white space analysis helps

Resolving a landscape that spans several reactor families, a fuel supply chain, and manufacturing requires more than keyword search. AI-powered analysis addresses this with semantic search that clusters activity by reactor family, layer, and application across varied terminology, attribution that normalizes developer, fuel-fabricator, and supplier filers to canonical entities, and continuous monitoring that keeps pace with a fast-moving field. Because nuclear advances appear in scientific and engineering literature before they are patented, reading both patents and literature gives the earliest signal of where the frontier is moving.

Where Cypris fits

Cypris runs patent landscape and white space analysis for engineering-intensive fields such as small modular reactors and advanced nuclear across a corpus of more than 500 million patents and scientific papers, organized through a proprietary R&D ontology. The ontology clusters activity by reactor family, gas-cooled, molten-salt, sodium-cooled, light-water, and microreactor, and by layer, reactor design, advanced fuel, passive safety, and manufacturing, and normalizes developer, fuel-fabricator, and supplier filers to canonical entities, so a team can resolve which families and layers are crowded and which remain open as white space, and can separate genuine advanced-nuclear filings from unrelated art that shares terminology. Semantic search across patents and scientific literature connects filings to the underlying nuclear-engineering research, which is where advances appear first. Cypris Q, the platform's agentic layer, lets teams run landscape and white space analysis conversationally and chain the clustering, attribution, and gap analysis, and Agentic Monitoring tracks a defined family over time and flags new patents and papers as they publish. Cypris provides enterprise API partnerships with OpenAI, Anthropic, and Google, and is built with enterprise-grade security. Cypris serves hundreds of enterprise customers across pharmaceuticals, chemicals, advanced materials, energy, and other regulated industries.

FAQ

What is the small modular reactor patent landscape? The small modular reactor patent landscape is the set of patents covering factory-built, smaller nuclear reactors and advanced designs. It spans reactor families, high-temperature gas-cooled, molten-salt, sodium-cooled fast, light-water, and microreactors, plus advanced fuels, passive safety, and modular manufacturing. Each is a distinct region of patenting.

What are the main advanced reactor families? The main families are high-temperature gas-cooled reactors, molten-salt reactors, sodium-cooled fast reactors, integral light-water SMRs, and microreactors. They differ in coolant, moderator, fuel, and operating temperature. The choice of family shapes both the technical and the freedom-to-operate picture.

Why are TRISO and HALEU fuels important? TRISO and HALEU fuels are important because many advanced reactors depend on TRISO coated particles, which resist very high temperatures, and on high-assay low-enriched uranium, which is not yet commercially produced at scale. Building this fuel supply chain is a critical enabler. Fabrication methods and fuel designs are therefore a concentrated, high-value layer.

Why is data-center demand accelerating advanced nuclear? Data-center demand is accelerating advanced nuclear because technology companies need large amounts of reliable, carbon-free power, and have begun contracting for SMR and advanced-reactor output. This has drawn substantial investment and sharpened competition. It has also expanded interest in non-electric and dispatchable applications.

How does advanced-reactor licensing work? Advanced-reactor licensing in the United States distinguishes several steps, including design approval or certification, a construction permit, and an operating license, and regulators have adopted a modernized, risk-informed, technology-inclusive framework to accommodate advanced designs. A construction permit is not an operating license. Claims about a given project's status should be tied to the specific regulatory record.

Where is the white space in advanced nuclear? The white space includes advanced fuel and HALEU supply, coolant, moderator, and materials, passive safety systems, modular manufacturing and construction, and non-electric applications. The reactor-design and fuel layers are the most active. The most open, high-value opportunities are in fuel, materials, and manufacturing.

What software helps analyze the small modular reactor patent landscape? Software for the advanced-nuclear landscape should cluster activity by reactor family, layer, and application, resolve developer, fuel-fabricator, and supplier filers to canonical owners, separate genuine advanced-nuclear filings from unrelated art, search patents and scientific literature semantically, and monitor a fast-moving field continuously. Cypris does this across more than 500 million patents and scientific papers using a proprietary R&D ontology, semantic search, Cypris Q, and Agentic Monitoring.

Which teams use advanced-nuclear patent landscape analysis? Advanced-nuclear patent landscape analysis is used by R&D, IP, and strategy teams at reactor developers, fuel fabricators, nuclear suppliers, utilities, and technology buyers, as well as investors assessing the sector. Because the field spans reactor, fuel, and manufacturing layers, structured analysis is essential. Cypris serves hundreds of enterprise customers across energy and other research-intensive industries.

Endnotes

1. International Atomic Energy Agency (2025). Coated particle fuels for high-temperature gas-cooled small modular reactors. https://doi.org/10.61092/iaea.zyya-k9sd
2. Powers, J. J., & Gehin, J. C. (2016). Liquid-fuel molten salt reactors for thorium utilization. Nuclear Technology, 194(2). https://doi.org/10.13182/nt15-124
3. Grimes, W. R. (1970). Molten-salt reactor chemistry. Nuclear Applications and Technology, 8(2). https://doi.org/10.13182/nt70-a28621
4. Hannah, B. (2026). Technical assessment of molten salt reactor small modular reactors: design and licensing considerations. https://doi.org/10.5281/zenodo.20087242
5. Zarei, M. (2020). State feedback control of power in a small modular reactor. Annals of Nuclear Energy, 144. https://doi.org/10.1016/j.anucene.2020.107743
6. Diniz, R., et al. (2023). Reactivity calculation in molten salt reactors with an inverse kinetics model. Annals of Nuclear Energy, 190. https://doi.org/10.1016/j.anucene.2023.110130
7. U.S. Nuclear Regulatory Commission (2026). 10 CFR Part 53 — risk-informed, technology-inclusive regulatory framework for advanced nuclear reactors (final rule). https://www.ecfr.gov/current/title-10/chapter-I/part-53

Perovskite-silicon tandem solar cells are the leading path to higher photovoltaic efficiency, and their patent landscape has become unusually central to competition because the field is commercializing through licensing as much as through manufacturing. A tandem cell places a wide-bandgap perovskite layer on top of a conventional silicon cell, so the two absorb different parts of the solar spectrum and the stack converts more sunlight than either alone, surpassing the single-junction limit that constrains standard silicon.¹,² The theoretical ceiling for a silicon-based tandem is about 43.2 percent, far above the roughly 33 percent limit of a single-junction silicon cell, which is what makes the architecture so attractive.³ The intellectual property divides across several regions, each with different owners and maturity: the perovskite compositions and their stability chemistry; the passivation and interface layers that raise efficiency and lifetime; the tandem device architecture, including the recombination layers that join the sub-cells; the texturing and deposition processes used to build the stack; and the encapsulation and manufacturing that make a durable module. Because a working tandem depends on all of these, freedom-to-operate and white space analysis must span the full stack.

The landscape is being shaped by patents and cross-licensing in real time. Certified efficiencies have climbed steeply: the current certified perovskite/silicon tandem record stands at 34.85 percent, achieved by LONGi and certified by the US National Renewable Energy Laboratory in 2025, and peer-reviewed work now describes certified perovskite/silicon efficiencies approaching 35 percent, with the live record register maintained on the NREL Best Research-Cell Efficiency Chart.⁴,⁵,⁶ These are laboratory cell records rather than commercial-module ratings, and translating them to industry-compatible cells and full modules is a distinct challenge the field is actively working through.¹⁰ Multi-junction routes are advancing in parallel, with triple-junction perovskite/perovskite/silicon devices exceeding 30 percent.⁷ Holders of strong foundational portfolios have begun licensing their technology to large manufacturers, signaling that IP position, not only manufacturing capacity, will determine who benefits from the transition. That structure is visible in the patent record: across the Cypris corpus of more than 500 million patents and scientific papers, the perovskite-tandem space is a comparatively small, fast-moving set of a few thousand de-duplicated families that stepped up sharply in 2025, and its most active assignees mix national laboratories such as CEA and CNRS, the perovskite specialist Oxford PV, and large silicon-module manufacturers, with China, the United States, South Korea, and France the leading jurisdictions; because assignee names are not fully canonicalized, manufacturer totals are best read as indicative. Because applications publish about eighteen months after filing, the most recent composition and process filings are under-represented, so the current frontier is more active than granted-patent counts suggest.

The strategic question is which layer to own, and the white space sits where durability is hardest to achieve. Perovskite stability under heat, humidity, and light remains the central unsolved problem, and the degradation mechanisms and stabilization techniques that address it are an area of intense, well-mapped research,⁸ as are the barrier and encapsulation designs that protect the cell over its service life.⁹ Compositions, passivation chemistries, and encapsulation that extend lifetime therefore carry high, defensible value, while tandem architecture and light-management texturing are contested and improving quickly, and scalable deposition and manufacturing are where laboratory records must survive the move to gigawatt production. Reading the landscape by composition, layer, and process, and tracking both the patents and the underlying materials research, is what separates a crowded region from an open one.

Where the perovskite tandem white space is

Stability and encapsulation. Compositions, passivation, and encapsulation that keep efficiency under heat, humidity, and light are the central unsolved problem and the highest-value, still-open target.⁸,⁹

Wide-bandgap perovskite compositions. Formulations tuned for the top cell that resist phase segregation are a contested, fast-moving composition layer.

Tandem architecture. Recombination layers, interconnection, and two-terminal versus four-terminal designs are a distinct device-engineering layer.⁷

Texturing and light management. Surface texturing and optical designs that maximize capture across the stack are an active process-IP area.

Scalable deposition and manufacturing. Moving high-efficiency processes from small cells to gigawatt-scale modules is where cost is decided and where durable process IP concentrates.¹⁰

How AI-powered landscape and white space analysis helps

Resolving a device landscape that spans compositions, interface and architecture layers, and manufacturing processes, across institutions and regions moving at different speeds, requires more than keyword search. AI-powered analysis addresses this with semantic search that clusters activity by layer and process across varied terminology, attribution that normalizes academic and commercial filers to canonical entities and tracks the licensing structure, and continuous monitoring that keeps pace with a fast-commercializing field. Because perovskite advances appear in scientific literature before they are patented, reading both patents and literature gives the earliest signal of where the frontier is moving.

Where Cypris fits

Cypris runs patent landscape and white space analysis for fast-commercializing energy fields such as perovskite tandem solar across a corpus of more than 500 million patents and scientific papers, organized through a proprietary R&D ontology. The ontology clusters activity by layer, perovskite composition, passivation and interfaces, tandem architecture, texturing and deposition, and encapsulation and manufacturing, and normalizes filers to canonical entities, so a team can resolve which layers are crowded and which remain open as white space, and can see the licensing structure clearly rather than as a flat list. Semantic search across patents and scientific literature connects filings to the underlying materials and device research, which is where perovskite advances appear first. Cypris Q, the platform's agentic layer, lets teams run landscape and white space analysis conversationally and chain the clustering, attribution, and gap analysis, and Agentic Monitoring tracks a defined layer over time and flags new patents and papers as they publish. Cypris provides enterprise API partnerships with OpenAI, Anthropic, and Google, and is built with enterprise-grade security. Cypris serves hundreds of enterprise customers across pharmaceuticals, chemicals, advanced materials, energy, and other regulated industries.

FAQ

What is the perovskite tandem solar patent landscape? The perovskite tandem solar patent landscape is the set of patents covering perovskite-silicon and all-perovskite tandem cells that exceed the single-junction efficiency limit. It divides across perovskite compositions and stability, passivation and interfaces, tandem architecture, texturing and deposition, and encapsulation and manufacturing. Each is a distinct region with different owners and maturity.

Why is IP so central to perovskite tandem solar? IP is central because the field is commercializing through licensing as much as through manufacturing, with holders of strong foundational portfolios licensing their technology to large manufacturers. Foundational process and architecture patents are concentrated among a few institutions and companies. That makes licensing and freedom-to-operate, not only production capacity, decisive.

What layers does the perovskite tandem landscape cover? The landscape covers perovskite composition and stability chemistry, passivation and interface layers, tandem device architecture including recombination layers, texturing and deposition processes, and encapsulation and manufacturing. A working tandem depends on all of them. Freedom-to-operate and white space analysis must span the full stack.

Where is the white space in perovskite tandem solar? The white space sits where durability is hardest: stability and encapsulation, wide-bandgap compositions that resist phase segregation, tandem architecture, light-management texturing, and scalable deposition. Stability under heat, humidity, and light is the central unsolved problem. The highest-value, most defensible positions are in lifetime and manufacturability.

Why is stability the key problem in the patent record? Stability is the key problem because perovskites can degrade under heat, humidity, and light, so the compositions, passivation, and encapsulation that extend lifetime are where the most valuable and defensible IP concentrates. Efficiency records matter, but durable modules require solving stability. The patent record reflects intense activity in these layers.

Why does perovskite analysis need scientific literature? Perovskite analysis needs scientific literature because new compositions, passivation chemistries, and device architectures appear in materials research before they are patented, so the literature gives the earliest signal. Analyzing patents alone gives a lagging view. Cypris analyzes both across more than 500 million patents and scientific papers.

What software helps analyze the perovskite tandem solar patent landscape? Software for the perovskite tandem landscape should cluster activity by device layer and process, resolve academic and commercial filers and the licensing structure to canonical owners, search patents and scientific literature semantically, and monitor a fast-commercializing field continuously. Cypris does this across more than 500 million patents and scientific papers using a proprietary R&D ontology, semantic search, Cypris Q, and Agentic Monitoring.

Which teams use perovskite tandem patent landscape analysis? Perovskite tandem patent landscape analysis is used by R&D, innovation, IP, and strategy teams at solar manufacturers, materials developers, and equipment makers, as well as investors and research institutions. It informs which layer to back, where to file, where to license, and where competitors are concentrated. Cypris serves hundreds of enterprise customers across energy, advanced materials, chemicals, and other regulated industries.

Endnotes

1. Zhang, F., Shi, Y., & Berry, J. J. (2024). Perovskite/silicon tandem solar cells: insights and outlooks. ACS Energy Letters, 9(4). https://doi.org/10.1021/acsenergylett.4c00172
2. Zheng, X., Tan, H., et al. (2023). Efficient perovskite/silicon tandem solar cells on industrially compatible textured silicon. Advanced Materials, 35(10). https://doi.org/10.1002/adma.202207883
3. Schubert, M. C., Glunz, S. W., et al. (2025). Elucidating the efficiency limit of silicon-based monolithic tandem cells through the combination of Auger and Shockley-Queisser limits. EES Solar. https://doi.org/10.1039/d5el00085h
4. Chen, W., Mularso, K. T., Jung, H. S., & Jo, B., et al. (2026). Strategies toward maximizing power conversion efficiency in all-perovskite tandem solar cells. Solar RRL. https://doi.org/10.1002/solr.202500911
5. LONGi (2025, April 16). LONGi breaks world record for crystalline silicon-perovskite tandem solar cell efficiency (34.85%). https://www.longi.com/en/news/silicon-perovskite-tandem-solar-cells-new-world-efficiency
6. National Renewable Energy Laboratory. Best research-cell efficiency chart. https://www.nrel.gov/pv/cell-efficiency
7. Aydin, E., Xu, L., De Wolf, S., et al. (2024). Four-terminal perovskite/perovskite/silicon triple-junction tandem solar cells with over 30% power conversion efficiency. ACS Energy Letters, 9(8). https://doi.org/10.1021/acsenergylett.4c01292
8. Ahn, N., & Choi, M. (2023). Towards long-term stable perovskite solar cells: degradation mechanisms and stabilization techniques. Advanced Science, 10(35). https://doi.org/10.1002/advs.202306110
9. Yang, Z., Liu, Z., Chen, W., et al. (2020). Barrier designs in perovskite solar cells for long-term stability. Advanced Energy Materials, 10(26). https://doi.org/10.1002/aenm.202001610
10. Jost, M., et al. (2021). 27.9% efficient monolithic perovskite/silicon tandem solar cells on industry-compatible bottom cells. Solar RRL, 5(6). https://doi.org/10.1002/solr.202100244

‍

Metal-organic frameworks have moved from a laboratory curiosity to a commercial materials platform, and their patent landscape is being staked out just as the field reaches scale. A MOF is a porous crystalline material built by linking metal nodes with organic linkers into an ordered framework, producing extraordinarily high surface areas and pores that can be tuned for a target molecule; more than 20,000 distinct MOFs had already been reported by the early 2010s, and the reticular chemistry behind them has continued to mature.¹,² Recognition by the 2025 Nobel Prize in Chemistry, awarded to Susumu Kitagawa, Richard Robson, and Omar Yaghi for the development of metal-organic frameworks, underscored the field's arrival and named applications from carbon-dioxide capture and toxic-gas storage to water harvesting, catalysis, and the separation of per- and polyfluoroalkyl substances from water.³ The intellectual property now divides across three broad regions: the specific framework compositions and structures themselves; the synthesis, shaping, and manufacturing processes that turn a powder into a usable, scalable product; and the application-level systems that integrate a MOF into a working device. Because MOFs serve many functions, freedom-to-operate and white space analysis must span all of them.

The commercial tipping point is reshaping the patent picture. After years in which scale-up and cost were the barriers, industrial-scale production of MOFs for carbon capture has begun, and a wave of startups is pursuing modular capture systems that are easier to scale than incumbent solvent processes, alongside chemical majors moving into manufacturing. Direct air capture of carbon dioxide has been demonstrated in purpose-designed frameworks from the laboratory through pilot scale, distinct from higher-concentration point-source capture.⁶ The scale of activity is large: across the Cypris corpus of more than 500 million patents and scientific papers, the MOF and reticular-chemistry space holds well over 100,000 de-duplicated families and has sustained high-volume filing since around 2018, with the assignee base led by academic institutions and chemical majors such as Sinopec, BASF, and ExxonMobil also prominent, while China accounts for the large majority of families, ahead of the United States, Japan, Germany, and South Korea. This shift moves value from the bare framework composition, where foundational academic estates are concentrated, toward the synthesis, shaping, and system-integration layers. Because applications publish about eighteen months after filing, the most recent synthesis and application filings are under-represented, so the current frontier is more active than granted-patent counts suggest.

The landscape divides by application, and the white space sits where a MOF must survive real conditions at low cost. Water stability, cycling durability, and inexpensive, scalable synthesis are the recurring bottlenecks; the criteria for a high-performance water-adsorbing framework, pore size and shape, hydrophilicity, and stability, are well defined but hard to meet at once.⁴ Carbon capture, both point-source and direct air capture, is the most active application and the one drawing the most new entrants; gas separation, increasingly through mixed-matrix membranes,⁷ and gas adsorption and storage⁸ are established; atmospheric water harvesting has advanced from concept toward passive devices,⁵ and newer uses such as direct lithium extraction and contaminant removal are earlier and less crowded. Underlying all of them is the reticular-design principle that lets chemists build frameworks to order for a target function.⁹ Reading the landscape by composition, synthesis route, and application is what separates a crowded region from an open one.

Where the MOF white space is

Water-stable, low-cost frameworks. MOFs that keep performance under humidity and real operating conditions, made by inexpensive routes, are the central bottleneck and a high-value, still-open target.⁴

Scalable synthesis and shaping. Converting powders into pellets, monoliths, and coatings by manufacturable processes is where deployment is decided and where hard-to-design-around process IP concentrates.

Direct air capture sorbents. MOFs tuned for capturing dilute atmospheric carbon dioxide are an active, high-value frontier distinct from point-source capture.⁶

Non-carbon separations. Direct lithium extraction, contaminant and per- and polyfluoroalkyl-substance removal, and other selective separations are earlier and less crowded application layers.

System integration. Contactors, modules, and regeneration systems that turn a MOF into a working unit are a distinct engineering layer separate from the framework chemistry.

How AI-powered landscape and white space analysis helps

Resolving a materials platform that spans many framework chemistries, synthesis routes, and application areas requires more than keyword search. AI-powered analysis addresses this with semantic search that clusters activity by composition, synthesis route, and application across varied terminology, attribution that normalizes academic and commercial filers to canonical entities and tracks new entrants, and continuous monitoring that keeps pace with a field reaching commercial scale. Because MOF advances appear in scientific literature before they are patented, reading both patents and literature gives the earliest signal of where deployable materials are emerging.

Where Cypris fits

Cypris runs patent landscape and white space analysis for materials platforms such as metal-organic frameworks across a corpus of more than 500 million patents and scientific papers, organized through a proprietary R&D ontology. The ontology clusters activity by framework composition, by synthesis and shaping route, and by application, carbon capture, gas separation and storage, water harvesting, catalysis, and mineral recovery, and normalizes filers to canonical entities, so a team can resolve which compositions, routes, and applications are crowded and which remain open as white space, and can track new entrants as the field scales. Semantic search across patents and scientific literature connects filings to the underlying materials chemistry research, which is where MOF advances appear first. Cypris Q, the platform's agentic layer, lets teams run landscape and white space analysis conversationally and chain the clustering, attribution, and gap analysis, and Agentic Monitoring tracks a defined application over time and flags new patents and papers as they publish. Cypris provides enterprise API partnerships with OpenAI, Anthropic, and Google, and is built with enterprise-grade security. Cypris serves hundreds of enterprise customers across pharmaceuticals, chemicals, advanced materials, energy, and other regulated industries.

FAQ

What is the metal-organic framework patent landscape? The metal-organic framework patent landscape is the set of patents covering MOFs, porous crystalline materials built from metal nodes and organic linkers, and their uses. It divides across framework compositions, synthesis and shaping processes, and application-level systems. Each application, from carbon capture to gas storage, is a distinct region of the landscape.

Why are MOFs a patenting hotspot now? MOFs are a patenting hotspot now because the field has reached a commercial tipping point, with industrial-scale production beginning for carbon capture and a wave of startups pursuing modular systems. Recognition by the 2025 Nobel Prize in Chemistry has further raised the field's profile. That shift is concentrating new filings in synthesis and application layers.

What application areas does the MOF landscape cover? The MOF landscape covers carbon capture from flue gas and directly from air, gas separation and storage, water harvesting, catalysis, sensing, drug delivery, and recovery of critical minerals such as lithium. Each demands different framework and system properties. Freedom-to-operate and white space analysis must span all of them.

Where is the white space in MOFs? The white space sits where a MOF must survive real conditions cheaply: water-stable, low-cost frameworks and scalable synthesis and shaping are the central bottlenecks, and direct air capture sorbents, non-carbon separations, and system integration are less crowded. The bare framework composition is where foundational estates concentrate. The higher-value opportunities are in deployability.

How is MOF value shifting from composition to manufacturing? MOF value is shifting because, as the field scales, the barrier moves from discovering a framework to making it durable and affordable at volume. Foundational composition estates are concentrated among a few academic groups, while synthesis, shaping, and system-integration IP is where deployment is now decided. That is where much of the defensible, hard-to-design-around value sits.

Why does MOF analysis need scientific literature? MOF analysis needs scientific literature because new frameworks, synthesis routes, and application concepts appear in materials research before they are patented, so the literature gives the earliest signal. Analyzing patents alone gives a lagging view. Cypris analyzes both across more than 500 million patents and scientific papers.

What software helps analyze the MOF patent landscape? Software for the MOF landscape should cluster activity by framework composition, synthesis route, and application, resolve academic and commercial filers to canonical owners, search patents and scientific literature semantically, and monitor a scaling field continuously. Cypris does this across more than 500 million patents and scientific papers using a proprietary R&D ontology, semantic search, Cypris Q, and Agentic Monitoring.

Which teams use MOF patent landscape analysis? MOF patent landscape analysis is used by R&D, innovation, IP, and strategy teams at chemicals, materials, carbon-capture, and energy companies, as well as investors and research institutions. It informs where to invest, where to file, and where competitors are concentrated. Cypris serves hundreds of enterprise customers across chemicals, advanced materials, energy, and other regulated industries.

Endnotes

1. O'Keeffe, M., Cordova, K. E., Furukawa, H., & Yaghi, O. M. (2013). The chemistry and applications of metal-organic frameworks. Science, 341(6149). https://doi.org/10.1126/science.1230444
2. Li, H., Rampal, N., & Yaghi, O. M. (2025). Reticular chemistry: past, present, and future. Molecular Frontiers Journal. https://doi.org/10.1142/s2529732525300034
3. Royal Swedish Academy of Sciences (2025, October 8). The Nobel Prize in Chemistry 2025 [press release]. https://www.nobelprize.org/prizes/chemistry/2025/press-release/
4. Furukawa, H., Queen, W. L., Yaghi, O. M., et al. (2014). Water adsorption in porous metal-organic frameworks and related materials. Journal of the American Chemical Society, 136(11). https://doi.org/10.1021/ja500330a
5. Diercks, C. S., Kalmutzki, M. J., & Yaghi, O. M. (2018). Metal-organic frameworks for water harvesting from air. Advanced Materials, 30(37). https://doi.org/10.1002/adma.201704304
6. Yao, M.-S., et al. (2023). Direct air capture of CO2 in designed metal-organic frameworks at lab and pilot scale. Carbon Capture Science & Technology, 8. https://doi.org/10.1016/j.ccst.2023.100145
7. Chai, M., Hou, J., & Chen, R. (2023). Metal-organic framework-based mixed matrix membranes for gas separation: recent advances and opportunities. Carbon Capture Science & Technology, 8. https://doi.org/10.1016/j.ccst.2023.100130
8. Sculley, J., Yu, J., Zhou, H.-C., et al. (2011). Carbon dioxide capture-related gas adsorption and separation in metal-organic frameworks. Coordination Chemistry Reviews, 255(15-16). https://doi.org/10.1016/j.ccr.2011.02.012
9. Chen, Z., Kirlikovali, K. O., Li, P., & Farha, O. K. (2022). Reticular chemistry for highly porous metal-organic frameworks: the chemistry and applications. Accounts of Chemical Research, 55(4). https://doi.org/10.1021/acs.accounts.1c00707

‍

Lipid nanoparticles are the delivery system that made mRNA medicines practical, and their patent landscape is distinctive because the delivery layer, rather than the therapeutic payload, is frequently the binding freedom-to-operate constraint. An LNP is built from four carefully selected lipid components, an ionizable lipid that carries the nucleic acid and enables its release inside the cell, a helper phospholipid, cholesterol, and a PEG-lipid that stabilizes the particle, combined in specific molar ratios and manufactured by a defined process.¹ The ionizable lipid is the primary determinant of potency, protonating in the acidic endosome to release the cargo, which is why it is the most heavily engineered and contested element,² and the lipid molar ratio is a first-order formulation variable that developers optimize through statistical design-of-experiments screens.³ Each of these elements can be claimed independently, and the ionizable lipid and the molar-ratio composition are the most heavily contested, so freedom-to-operate for an mRNA vaccine, an RNA therapeutic, or a gene-editing product delivered by LNP is a layered analysis across many owners rather than a single clearance of the drug substance.

The landscape is dense, multi-owner, and among the most litigated in biotechnology. The foundational LNP work traces to a small set of academic and company lineages, and rights have been licensed to many developers, so a single product can implicate several estates at once. The stakes are large: in March 2026, Genevant Sciences and Arbutus Biopharma reached a global settlement with Moderna resolving their lipid-nanoparticle patent dispute for up to $2.25 billion, comprising a $950 million upfront payment and a further $1.3 billion contingent on a pending appellate ruling over a government-use defense.⁴,⁵,⁶ Multiple parallel lipid-nanoparticle suits remain pending across US, European, and Canadian forums, and outcomes have turned on the specific patents asserted rather than on any single view of the technology. The concentration of rights is visible in the patent record: across the Cypris corpus of more than 500 million patents and scientific papers, the LNP and ionizable-lipid space holds on the order of 29,400 de-duplicated families, with filings inflecting sharply during the COVID-19 period, roughly tripling between 2020 and 2023, and the most active assignees, led by mRNA and RNA-therapeutics developers, mapping onto the same entities visible in the litigation; the United States leads on geography, followed by China, with a notable Canadian share reflecting the field's foundational lipid lineage. Because applications publish about eighteen months after filing, the newest lipid, targeting, and process filings are under-represented, so the current frontier is more active than granted-patent counts suggest.

The practical consequence is that delivery IP now shapes the economics of the entire RNA field. A developer typically needs freedom to operate on the ionizable lipid and the composition, plus the formulation and manufacturing process, and that can mean licensing from or designing around several holders. The durable value is concentrating in novel ionizable lipids, where iterative and structure-activity design continues to yield new, patentable chemistries,⁷ down to fine distinctions such as lipid isomerism that measurably change performance,⁸ in compositions that fall outside the contested molar-ratio claims, in targeting chemistries that reach tissues beyond the liver, and in manufacturing processes. Reading the landscape by lipid, layer, and owner, and tracking the live proceedings, is what separates a workable position from a blocked one.

What creates FTO risk in LNP delivery

Ionizable lipid claims. These cover the structures that carry and release the nucleic acid, the most heavily contested layer and the frequent center of litigation.²

Molar-ratio and composition claims. These cover the specific percentage ranges of the four lipid components, a layer that can block a formulation independently of the individual lipids.³

PEG-lipid and helper-lipid claims. These cover the stabilizing and structural lipids, a distinct and separately owned layer.

Formulation and manufacturing claims. These cover the process by which LNPs are assembled at scale, where practical, hard-to-design-around barriers concentrate.

Targeting and application claims. These cover tissue-targeting chemistries and specific cargo applications, so a delivery system can be free for one use and blocked for another.

How AI-powered landscape and FTO analysis helps

A dense, multi-owner, heavily litigated delivery landscape is beyond manual clearance. AI-powered analysis addresses this with semantic search that retrieves relevant ionizable-lipid, composition, PEG-lipid, formulation, and targeting claims regardless of terminology, attribution that resolves the many company and academic owners to canonical entities and captures the license chains, claim-level analysis that separates the layers, and continuous monitoring that tracks new filings and the live disputes. Because delivery advances appear in scientific literature before they are patented, reading both patents and literature gives earlier warning of where the field is extending.

Where Cypris fits

Cypris runs patent landscape and freedom-to-operate analysis for dense, contested fields such as LNP delivery across a corpus of more than 500 million patents and scientific papers, organized through a proprietary R&D ontology. The ontology clusters the landscape by layer, ionizable lipid, composition, PEG-lipid, formulation, and targeting, and normalizes company and academic owners to canonical entities, so a team sees how rights are distributed across the web of holders rather than a flat list. Semantic search across patents and scientific literature surfaces relevant claims regardless of terminology and connects filings to the underlying chemistry research, which is where novel lipids and targeting approaches emerge first. Cypris Q, the platform's agentic layer, lets teams run landscape and FTO analysis conversationally and chain the attribution, clustering, and claim-level analysis across layers, and Agentic Monitoring tracks the landscape over time and flags new filings and developments as they publish. Cypris provides enterprise API partnerships with OpenAI, Anthropic, and Google, and is built with enterprise-grade security. Cypris serves hundreds of enterprise customers across pharmaceuticals, chemicals, advanced materials, energy, and other regulated industries.

FAQ

Why is freedom-to-operate hard for lipid nanoparticles? Freedom-to-operate is hard for lipid nanoparticles because an LNP is built from four lipid components combined in specific ratios by a specific process, each independently patentable and held across many owners. The ionizable lipid and molar-ratio composition are especially contested. FTO must be assessed layer by layer across multiple estates, often for a delivery system rather than the drug itself.

Why is LNP the binding constraint for RNA products? LNP is frequently the binding constraint because delivery, not the nucleic acid payload, is the hardest part of an RNA medicine, and the delivery IP is densely held. A product can clear its therapeutic sequence and still be blocked on the lipid or the composition. That is why delivery litigation has been so consequential.

What claim types create FTO risk in LNP delivery? Five claim types create FTO risk: ionizable-lipid claims, molar-ratio and composition claims, PEG-lipid and helper-lipid claims, formulation and manufacturing claims, and targeting and application claims. Each covers a distinct layer and can independently block a product. Ionizable lipids and molar ratios are the most litigated.

Why has LNP patent litigation been so significant? LNP patent litigation has been significant because the technology enabled a very large market, and rights are held across several estates traceable to a few foundational lineages. Disputes over ionizable lipids, molar ratios, and formulation have produced high-value cases and settlements across jurisdictions, including a multi-billion-dollar 2026 settlement between Genevant and Arbutus and Moderna. Outcomes turn on the specific patents asserted rather than a single view of the technology.

Where is the white space in LNP delivery? The white space sits in novel ionizable lipids, compositions outside the contested molar-ratio claims, targeting chemistries that reach tissues beyond the liver, non-PEG stabilization, and manufacturing processes. The core lipid and composition ground is crowded and litigated. The durable, defensible value is in these newer chemistry and process layers.

Why does LNP analysis need scientific literature? LNP analysis needs scientific literature because new lipids, targeting chemistries, and formulation advances appear in research before they are patented, so the literature gives the earliest signal. Analyzing patents alone gives a lagging view. Cypris analyzes both across more than 500 million patents and scientific papers.

What software helps analyze the LNP delivery patent landscape? Software for the LNP delivery landscape should resolve the many company and academic owners and license chains to canonical entities, cluster the ionizable-lipid, composition, formulation, and targeting layers, search patents and scientific literature semantically, and monitor active litigation and new filings continuously. Cypris does this across more than 500 million patents and scientific papers using a proprietary R&D ontology, semantic search, Cypris Q, and Agentic Monitoring.

Which teams need LNP patent landscape and FTO analysis? LNP patent landscape and FTO analysis is needed by R&D, IP, and business-development teams at mRNA, RNA-therapeutic, vaccine, and gene-editing companies, as well as investors assessing RNA assets. Because delivery is often the binding constraint, structured analysis is essential. Cypris serves hundreds of enterprise customers across pharmaceuticals and other research-intensive industries.

Endnotes

1. Liu, S., Zhang, X., Zhang, Y., & Gao, Y. (2024). Principles of lipid nanoparticle design for mRNA delivery. BMEMat. https://doi.org/10.1002/bmm2.12116
2. Han, X., Tang, X., & Zhang, Y. (2023). Ionizable lipid nanoparticles for mRNA delivery. Advanced NanoBiomed Research, 3. https://doi.org/10.1002/anbr.202300006
3. Fenton, O. S., Anderson, D. G., et al. (2015). Optimization of lipid nanoparticle formulations for mRNA delivery in vivo with fractional factorial and definitive screening designs. Nano Letters, 15(11). https://doi.org/10.1021/acs.nanolett.5b02497
4. Genevant Sciences & Arbutus Biopharma (2026, March 3). Genevant Sciences and Arbutus Biopharma announce $2.25 billion global settlement with Moderna. https://www.genevant.com/genevant-sciences-and-arbutus-biopharma-announce-2-25-billion-global-settlement-with-moderna
5. Roivant Sciences (2026). Settlement disclosure (Exhibit 99.1), U.S. Securities and Exchange Commission. https://www.sec.gov/Archives/edgar/data/1635088/000114036126007548/ef20067067_ex99-1.htm
6. Arbutus Biopharma (2026, March 3). Form 8-K. https://investor.arbutusbio.com/static-files/f6868345-37b9-4bd3-9ba3-799e754e6ce1
7. Manning, A. M., Khan, O. F., et al. (2023). Iterative design of ionizable lipids for intramuscular mRNA delivery. Journal of the American Chemical Society, 145(4). https://doi.org/10.1021/jacs.2c10670
8. Zuo, T., He, Z., Li, Z., et al. (2026). Unraveling the role of ionizable lipid isomerism in modulating lipid nanoparticles for mRNA delivery. Journal of the American Chemical Society. https://doi.org/10.1021/jacs.5c20438