Photovoltaic (PV) technology has been studied in U.S. government-funded laboratories for decades and much of that work has been published. That makes Department of Energy (DOE) and National Renewable Energy Laboratory (NREL) documents a natural place to look when testing whether a solar patent claim is valid.
This article explains when such research can invalidate a claim as anticipatory prior art, what evidence is needed and how the choice of forum affects the strategy. The scenarios are hypothetical and illustrative. Nothing here says any particular patent is invalid. It is general information, not legal advice and the procedural landscape described in Section 8 has been changing quickly.
1. Why DOE and NREL literature is a rich source
- Long history and volume. NREL’s predecessor, the Solar Energy Research Institute, was renamed NREL in 1991. Decades of work on cell materials, device architectures, modules, reliability and manufacturing processes sit in its archive.
- Publication culture. National laboratories publish technical reports, conference papers, journal articles and presentations as a normal part of their mission and DOE-funded work is generally reported to DOE’s Office of Scientific and Technical Information (OSTI) and searchable there.
- Extended network. DOE funds universities and companies, whose final and subcontract reports add to the public record beyond the laboratory’s own staff.
- Breadth of subject matter. The literature covers cell structures (crystalline silicon, thin-film CdTe and CIGS, III-V multijunctions, perovskites and tandems), module packaging and encapsulation, degradation mechanisms, bifacial and tracking systems, testing methods and modeling tools.
- Dating evidence tends to exist. Report numbers, OSTI records, DOIs and conference proceedings often carry clear dates.
Two cautions follow. Not every DOE or NREL document is public and one document rarely discloses everything a claim requires. Both points come up repeatedly below.
2. The legal framework
Which version of § 102 applies
| AIA (effective filing date on or after March 16, 2013) | Pre-AIA (earlier filings) | |
| Prior art publications | § 102(a)(1): described in a printed publication (or patented, in public use, on sale, or otherwise available to the public) before the effective filing date | § 102(a): before the applicant’s invention date; § 102(b): more than one year before filing |
| Antedating | Not available, but limited grace period exceptions for the inventor’s own disclosures within one year (§ 102(b)(1)) | A patentee can try to “swear behind” a § 102(a) reference by proving earlier invention, but not a § 102(b) reference |
| Other patents and applications | § 102(a)(2): earlier-filed patents and published applications naming another inventor | § 102(e), § 102(g) |
Because solar patents often have long lives and many were filed before 2013, check which regime governs before evaluating any reference. A DOE report from 2004 might be a strong § 102(b) reference against a 2006 filing, but only a potentially antedatable § 102(a) reference against a filing made a few months after the report.
What anticipation requires
Anticipation requires that a single reference disclose every element of the claim, arranged as in the claim, in a way that would let a skilled person practice the invention. Three points matter especially for research literature:
- Arrangement matters. A paper that discusses a layer, a dopant and a contact in different contexts does not anticipate a claim to those elements in a particular stack. That is a job for obviousness.
- Enablement is presumed for prior art. A published prior art reference is generally presumed enabling and the patentee bears the burden of showing it isn’t. Lab-scale work can be enabling even if it was never commercialized. The reference needn’t be practical or cost-effective, only enabling.
- Ranges and inherency. A reference disclosing a specific value inside a claimed range can anticipate. Overlapping ranges can too under some circumstances. Inherent disclosure counts if the claimed feature necessarily results, but “may” or “probably” results is not enough.
3. Is the DOE or NREL document a “printed publication”?
The key test is public accessibility: whether the document was disseminated or otherwise made available so that a person interested in the field, exercising reasonable diligence, could locate it.
Courts have used this test across several settings. Some illustrative decisions:
- In re Klopfenstein (Fed. Cir. 2004) treated a poster displayed at a conference as a printed publication after weighing how long it was displayed, how many people saw it, whether copies were made and how sophisticated the audience was.
- In re Cronyn (Fed. Cir. 1989) held that student theses on file in a library were not publicly accessible because they weren’t indexed in a way that would let a researcher find them.
- Massachusetts Institute of Technology v. AB Fortia (Fed. Cir. 1985) found a conference paper publicly accessible when it was distributed to interested persons without restriction.
- Blue Calypso v. Groupon (Fed. Cir. 2016) confirmed that the inquiry is fact-specific and turns on whether the reference was actually made available.
Applying the test to DOE and NREL materials
Facts that support public accessibility:
- The report appears in OSTI or in NREL’s public publications listing with a date
- It has a report number, DOI, or catalog record
- It was published in conference proceedings, journals, or a widely distributed series
- Archived copies (for example, web archive captures) show the document online on a specific date
- Library or repository records show the document was indexed and searchable
Facts that create risk:
- Limited-distribution reports. Some documents carry restricted distribution statements, or are internal or “official use” materials.
- Protected data. Under the federal technology transfer statute, information from a Cooperative Research and Development Agreement (CRADA) that would be a trade secret may be withheld from disclosure for a limited period (up to five years). Similar protections can apply to some other collaborations.
- Draft or preprint status. An unpublished draft is not public merely because it exists.
- Undated web pages. A webpage without reliable dating evidence is a weak reference.
- Poster or oral presentations with no distributed copy may need testimony under the Klopfenstein factors.
- Conference proceedings vs. presentation dates. The relevant date is when the material became accessible, which may differ from the date of the talk.
The practical lesson is to treat “it’s from a government lab” as a starting point, not as proof that the document is public and dated.
4. Beyond reports: other DOE and NREL-related prior art
- DOE and NREL patents and published applications are prior art as patents and, under the AIA, as earlier-filed applications under § 102(a)(2) (as of their filing date, subject to the statutory exceptions).
- Software and data releases. Publicly released modeling tools and datasets can disclose methods, though their status as printed publications depends on how they were distributed.
- Test standards, protocols and methodology papers can disclose measurement methods that patents later claim.
- Public use or on-sale activity, such as field deployments or commercial sale of modules developed under DOE programs, can qualify as prior art in district court. They are not available in inter partes review, which is limited to patents and printed publications.
5. Anticipation is the narrow path and obviousness is the wide one
In real disputes, DOE and NREL research is often more useful as a building block for obviousness than as a single anticipating reference. Under KSR v. Teleflex (2007), a claim can be obvious from a combination of references when a skilled person would have had reason to combine them with a reasonable expectation of success. Government-lab literature is good at supplying:
- a known structure or process step,
- a motivation (efficiency gains, degradation problems, cost reduction) and
- evidence of a reasonable expectation of success (measured results).
Where anticipation can work:
- The claim is drawn narrowly around a specific device stack, process, or parameter range that a single report discloses.
- The report is a detailed experimental paper with a full description of structure and fabrication.
Common patentee responses:
- Missing element or wrong arrangement. The reference doesn’t disclose one limitation, or discloses the elements separately.
- Not enabled. The reference reports a result without the details needed to reproduce it.
- Performance limitations. A claim reciting a performance metric (say, efficiency or degradation rate) is not necessarily met by a reference that reports a different device or scale, such as a small research cell compared with a full-size module.
- Not public, or wrong date. The accessibility and date challenges in Section 3.
- Antedating (pre-AIA § 102(a) references only).
- Unexpected results or teaching away (relevant to obviousness).
6. A hypothetical walk-through (illustrative only)
Suppose a patent, filed after March 16, 2013, claims a solar cell comprising a layered stack with a passivation layer in a specified thickness range and a dopant concentration range.
- Identify the critical date. Find the earliest effective filing date the patent can claim, including any valid priority.
- Search the archives. Look through OSTI, NREL’s publication listings and conference proceedings from before that date.
- Find a candidate reference. Suppose an NREL conference paper published two years before the critical date describes a stack with the same layers in the same order, with a passivation layer thickness inside the claimed range and a doping level inside the claimed range.
- Chart the claim. Map each limitation to specific passages, figures and tables in the paper.
- Establish public accessibility. Gather the proceedings’ publication date, the catalog record and evidence that the paper was indexed and downloadable.
- Anticipate the counterarguments. The patentee argues the paper doesn’t disclose the specific dopant type. If it doesn’t, anticipation fails, but the paper may still support obviousness in combination with another reference.
- Evaluate enablement. The paper gives fabrication conditions sufficient for a skilled person, so the presumption of enablement stands unless rebutted.
If every limitation is found and the document is properly dated and accessible, the claim is anticipated. If one element is missing, the same paper may still anchor an obviousness case.
7. Building the evidence
- Secure authoritative copies of the reference and its metadata as early as possible. Web content can change.
- Establish the date through publisher or repository records, proceedings metadata, DOIs, archived captures and, where needed, a declaration from a records custodian or librarian.
- Prove accessibility with indexing evidence, distribution information and search results showing how a researcher would have found it.
- Authenticate the document. In court that means meeting the evidence rules on authentication and in Patent Office proceedings it means submitting proper exhibits and declarations.
- Use an expert to explain what the reference teaches to a skilled person, why it is enabling and how each element maps to the claim.
- Consider the authors. A declaration from the author can help, but authors may be difficult to reach or reluctant and their testimony can cut both ways.
- Check government sources. Requests to DOE for records may be possible, though timing is uncertain.
8. Choosing the forum
| Forum | Burden | What prior art counts | Notes |
| District court or ITC | Clear and convincing evidence (Microsoft v. i4i, 2011) | Any prior art, including public use and on sale | Full range of defenses. Juries may find technical prior art harder to follow |
| Inter partes review (IPR) | Preponderance of the evidence | Patents and printed publications only, on novelty and obviousness grounds | Institution is now heavily discretionary (below) |
| Post-grant review (PGR) | Preponderance | Broader grounds, but only within nine months of grant | Narrow window |
| Ex parte reexamination | Substantial new question of patentability | Patents and printed publications | Not subject to the discretionary-denial factors that now shape IPR |
The IPR landscape has tightened considerably. Reports from 2025 and 2026 describe a shift toward director-controlled institution decisions and expanded discretionary denial. In a precedential decision in May 2026, Director Squires said the Office should use discretion so that PTAB challenges act as a quick and cost-effective alternative to district court litigation rather than an expansion of it. Factors that have been applied include parallel litigation, prior challenges, examiner error and “settled expectations” based on how long the patent has been in force. An October 2025 notice of proposed rulemaking would codify many of these changes, so check whether it has been finalized.
For solar disputes specifically, a March 11, 2026 memorandum added new discretionary considerations that are notable in an industry with heavy overseas manufacturing: whether accused products are made in the United States, whether the patent owner’s competing products are made in the United States and whether the petitioner is a small business sued for infringement.
By contrast, commentators note that ex parte reexamination is governed by the “substantial new question” standard and is not subject to Fintiv-type timing considerations or settled expectations. That has made it a more attractive route for prior-art challenges based on publications, though it offers no adversarial participation for the requester once filed. Because this area is moving fast, confirm the current rules with counsel before choosing a forum.
9. The patentee’s perspective
- Search before you file. Reasonable prior art searches should include OSTI and NREL’s publications, not only patent databases, especially for technology areas where national laboratories are active.
- Duty of disclosure. Under 37 CFR § 1.56, known material information must be disclosed. Failing to cite a material DOE or NREL reference can support an inequitable conduct claim, although the standard after Therasense v. Becton Dickinson (Fed. Cir. 2011, en banc) requires but-for materiality and specific intent to deceive.
- Watch your own publications. If inventors are authors of NREL or DOE-funded papers, publication starts the clock on the grace period (one year under the AIA) and can become prior art against your own later filings.
- Government rights. If the invention arose under a DOE funding agreement, the Bayh-Dole framework imposes disclosure, election and reporting obligations and grants the government a license. This isn’t an invalidity issue, but noncompliance can jeopardize title.
- Draft claims with a range of fallbacks. Dependent claims that add specific structure or parameters preserve validity if the broad claim falls.
Bottom line
DOE and NREL research can be powerful prior art against solar patent claims, but only when the reference is demonstrably public, reliably dated, enabling and complete enough to meet every claim limitation as arranged. More often, government-lab literature works as the foundation of an obviousness case. The forum matters as much as the reference: with IPR institution increasingly discretionary, district court, ex parte reexamination and PGR (where timely) deserve serious consideration.
I can also put this into a Word document or formatted page if you’d like.