Research
Published work
Each drop ships as a paired narrative note and public-record technical artifact (a defensive disclosure under 35 U.S.C. §102 or an open dataset under CC-BY-4.0). Every drop carries a Zenodo DOI for third-party timestamp.
Research Note 08 · Published May 2026
Open IND blueprint for a humanized hetero-octameric chaperonin built on the human TRiC/CCT scaffold (CCT1 through CCT8) with 54 engineered amino acid substitutions across the eight subunits and an N-terminal TAT cell-penetrating peptide (RKKRRQRRR) fused to CCT1, CCT3, CCT6A. Predicted 19 to 28 percent slowing of cognitive decline on the regulatory-favored Clinical Dementia Rating Sum of Boxes (CDR-SB) at 18 months in early symptomatic Alzheimer's disease (central estimate 23 percent; upside potential 30 to 43 percent if Pathway B or C operates intracellularly with high TAT-CPP delivery efficiency). Three plausible mechanism pathways with different ATP requirements: anti-aggregation by substrate sequestration (ATP-independent; well-precedented per Sontag 2013, Tam 2006); entropic-pulling disaggregation analogous to HSP70-mediated extraction of α-synuclein from fibrils per Wentink 2020 (ATP-required, intracellular only); chaperonin-specific apical-domain disaggregation (ATP-required, intracellular only; without direct precedent). Pathways B and C are restricted to the intracellular ATP-saturated compartment because extracellular CSF [ATP] (~10 nM) is severely undersaturated relative to chaperonin K_M(ATP) (~25 to 100 μM); under all three scenarios the extracellular fraction operates Pathway A. The compartment-based efficacy model is calibrated to lecanemab CLARITY-AD CDR-SB (27.1 percent) and bracket-checked against five other anti-amyloid antibody clinical-trial readouts on combined-population CDR-SB; 5 of 5 non-calibration anchors land within or borderline-near bracket. The construct provides intracellular access via TAT-CPP fusion to the approximately 80 percent intracellular tau cytotoxic load that no antibody-class therapy can reach; does not engage Fc-receptor-mediated complement or vascular amyloid clearance; predicted to carry no excess ARIA risk over background incidence. Annual cost at competitive generic equilibrium 7 to 15 times lower than the anti-amyloid antibody class. Three falsification-class bench experiments at $65,000 total distinguish among the mechanism pathways; ATP dose-response on synthetic Aβ42 oligomer disaggregation is the primary determination. The construct artifact (FASTA sequences, structured per-subunit mutation table, manufacturing route, formulation buffer, release criteria, dose regimen) is dedicated to the public domain under CC0 1.0 Universal; paper text and analysis are licensed under CC-BY 4.0. Anonymous research collaboration; humanitarian distribution; no patent applications filed or intended.
- Reader impact
- An open IND blueprint for a humanized cell-penetrating chaperonin construct that addresses the structural gap left by the anti-amyloid antibody class: intracellular tau access via TAT-CPP fusion (covering the approximately 80 percent intracellular tau cytotoxic load that antibodies cannot reach), broad-spectrum apical-surface engagement of soluble Aβ and tau oligomers (not plaque-specific, not Aβ-restricted), no Fc-receptor-mediated ARIA mechanism, and competitive-generic cost (7 to 15 times lower than the antibody class). Predicted clinical magnitude on the regulatory-favored CDR-SB endpoint sits within the antibody-class span: above aducanumab EMERGE high (22.0 percent), below lecanemab CLARITY-AD (27.1 percent) and donanemab TRAILBLAZER-ALZ2 combined population (29 percent). The construct artifact is dedicated to the public domain under CC0 and explicitly invites manufacture by contract manufacturing organizations, generic biopharmaceutical companies, academic medical centers, non-profit drug-development consortia, and regulators without permission, royalty, or attribution. The deposit serves as defensive prior art under 35 U.S.C. § 102 and analogous foreign-jurisdiction novelty provisions including European Patent Convention Article 54.
- Who can implement
- Any contract manufacturing organization with biologics expression capability (CHO or HEK293 mammalian-cell expression of eight separate subunit constructs followed by in-vitro reconstitution), any generic biopharmaceutical company, any academic medical center running disease-modifying Alzheimer's trials, any non-profit drug-development consortium, or any regulator. The deposit publishes the eight engineered subunit FASTA sequences, the structured per-subunit mutation table with positions and rationales, the manufacturing route, the intrathecal-grade formulation buffer (50 mM phosphate at pH 7.4 with 150 mM NaCl, 1 mM glutathione, 0.01% polysorbate-80), the release criteria (post-assembly cryo-EM resolution ≤ 5.0 Å, ≥ 95 percent canonical-positioning rate at the eight-subunit assembly checkpoint), the dose regimen (15 to 20 mg every four weeks intrathecal), and three falsification-class bench experiments at approximately $65,000 total that distinguish among the construct's three plausible mechanism pathways. See Engagements for commissioned scoping work or custom-class disclosures across additional disease-modification targets.
- Disclosure landing HTML, headline prediction, mechanism pathways, cross-validation against the antibody class, falsification path, manufacturer invitation
- Zenodo deposit DOI 10.5281/zenodo.20100749 · 5 files · ~1.5 MB · CC0 on construct artifact (FASTA + JSON) · CC-BY 4.0 on paper text + cover
Research Note 07 · Published May 2026
Defensive publication of 1,908,879 sequence-distinct allosteric peptide modulator candidates targeting the MT1 (gene MTNR1A, UniProt P48039) and MT2 (gene MTNR1B, UniProt P49286) melatonin receptors. The disclosure decomposes to 35,198 unique linear backbones × 90 distinct form-variant chemistries spanning seven chemistry families: lipidation (988,540 compositions, 51.8%), PEGylation (397,104, 20.8%), glycosylation (303,985, 15.9%), Fc fusion (105,606, 5.5%), prodrug variants (78,314, 4.1%), glucuronidation (35,202, 1.8%), and macrocyclization (36 compositions; small subset by count, with predicted-affinity quality competitive with the form-variant families). Three-tier disclosure architecture: 100 Tier 1 deep-dossier candidates with 15-section per-candidate HTML covering sequence, predicted structure, mutation map, predicted pharmacology with form-variant Kd caveat, family-level synthesis route, MT1/MT2 radioligand-displacement binding-affinity assay, cAMP HTRF functional confirmation, predicted PK profile, cost-and-timeline estimates, and provenance with bundle SHA-256 and OpenTimestamps Bitcoin block-header attestation; 1,000 Tier 2 table-form stubs; 1,907,779 Tier 3 manifest-only entries. The deposit additionally publishes a WIPO Standard ST.26 Sequence Listing at the unique-backbone level (35,198 entries with feature annotations for form-variant modifications) for examiner-side BLAST searchability. The universal-protocol package comprises eight family-level synthesis protocols (base Fmoc SPPS, lipidation, PEGylation, glycosylation, Fc fusion, prodrug, glucuronidation, cyclization) and three form-variant-independent characterization protocols (radioligand-displacement binding assay with [125I]-2-iodomelatonin tracer and Cheng-Prusoff IC50→Kd correction, cAMP HTRF functional readout with allosteric-cooperativity discrimination per the Christopoulos-Kenakin allosteric ternary complex model, RP-HPLC + ESI-MS + AAA quality control), satisfying the §102 enablement element under the In re Wands factors for any specific disclosed composition without per-candidate inventive choice. Each disclosed composition carries a readiness-quality letter (A / B / C / D) summarizing its predicted-affinity tier; the letter is a pure-pharmacology readout and the deposit makes no clinical or therapeutic-efficacy claim. Predicted Kd values rest on a single research-grade anchor (an 18-residue endozepine-derived hypothesis scaffold sequence, not a contiguous fragment of human DBI; closest 18-residue DBI window is 13/18 identity) and produce predictions at the receptor-class-template level only. The deposit is dedicated to the public domain under CC0; multi-jurisdictional §102 declaration covers US (35 U.S.C. §102 AIA), EU (EPC Article 54), UK (Patents Act §6), JP (Patent Act §29), and CN (Patent Law Article 22). Bundle SHA-256 anchored to Bitcoin via OpenTimestamps.
- Reader impact
- Defensive publication of 1,908,879 distinct (backbone × form-variant × cyclization) compositions forecloses composition-of-matter patent claims on every disclosed composition under 35 U.S.C. §102 in the United States and the equivalent novelty provisions in major jurisdictions (EPC Article 54, UK Patents Act §6, JP Patent Act §29, CN Patent Law Article 22). Eleven stated utility surfaces (allosteric modulation of MT1, allosteric modulation of MT2, modulation of circadian rhythm, sleep-wake cycle, mood and affective state, pain pathways, immune signaling, cellular proliferation and oncology research, metabolic pathways, cardiovascular function, reproductive function, neuroprotection) maximize method-of-use foreclosure surface for any specific composition any party would otherwise attempt to claim post-publication. The deposit does not block §103 obvious-variants claims; method-of-use claims for indications outside the eleven stated utilities, formulation claims, combination-therapy claims, and compositions including a disclosed peptide as a non-disclosed component are unaffected and explicitly preserved.
- Who can implement
- Any peptide medchem team, peptide-bioconjugation lab, recombinant-Fc-fusion expression facility, generics manufacturer, or pharma BD function operating in the melatonin-receptor allosteric peptide space. Each Tier 1 dossier carries the family-level synthesis protocol pointer, the universal MT1 and MT2 radioligand-displacement binding-affinity assay protocol, the universal cAMP HTRF functional readout protocol, RP-HPLC + ESI-MS + AAA quality-control standards, predicted PK profile, and cost-and-timeline estimates. Implementation requires per-candidate Fmoc SPPS synthesis (or family-specific synthesis route per the form-variant chemistry), RP-HPLC purification with ESI-MS and AAA quality control, radioligand-displacement binding assay validation against MT1 and MT2 stable HEK293 membrane preparations, and functional cAMP confirmation. See Engagements for commissioned scoping work or custom-class disclosures across additional receptor families.
- Disclosure landing HTML, deposit summary, three-tier architecture, universal-protocol package, predictor properties
- Zenodo deposit DOI 10.5281/zenodo.20012096 · 2 files · 65.2 MB on Zenodo (1,135 files extracted) · CC0 / Public Domain Dedication · OpenTimestamps Bitcoin attestation
Research Note 06 · Published May 2026
A public reference catalog of 84 aging-dysregulated human miRNAs with integrated dossiers covering López-Otín hallmark mapping, centenarian-signature classification across six independent cohort studies, predicted target footprint, GTEx v8 tissue exposure across ten aging-relevant tissues, mouse and rat ortholog mapping, SenMayo senescence-associated secretory phenotype overlay, top fifteen HIGH-tier predicted targets, and freedom-to-operate clearance across six antisense modulator chemistries (LNA-2'-MOE-PS gapmer 16 nt, 2'-MOE-PS gapmer 22 nt, mimic 22 nt, GalNAc-2'-MOE-PS gapmer 16 nt, PMO 22 nt, tough-decoy 60 nt) against a 5,596-patent corpus across 252 assignees. Twenty-five candidates pass the canonical antisense chemistry without HIGH or MEDIUM patent encumbrance flag and are publishable at canonical specificity without licensing. Eight candidates classify as DIRECT_SASP under hypergeometric over-representation of senescence-associated secretory phenotype genes (Saul 2022 SenMayo) among their top-500 strongest predicted binders; six of the eight are also IP-clean at canonical chemistry (miR-130a-3p, miR-130b-3p, miR-148a-3p, miR-148b-3p, miR-152-3p, miR-365a-3p), forming a SASP-axis senolytic discovery panel. The hero candidate miR-130a-3p hits all three independent evidence axes simultaneously: SASP-axis (top-500 predicted binders contain twelve SenMayo signature genes including EREG, IGF1, IL15, TNFRSF1B, MMP13, MMP10, TNF, IGFBP5; log2 fold-enrichment 1.917, hypergeometric p = 8.3 × 10⁻⁵), centenarian-preserved (preserved-LLI in the German-Kiel PBMC cohort of ElSharawy 2012 Aging Cell with confidence MEDIUM), and IP-clean canonical at LNA-2'-MOE-PS gapmer 16 nt with HIGH = 0 and MEDIUM = 0. The literature-evidence corpus is 559 audited rows with 83.0% A-grade primary-source verification across all twelve López-Otín hallmarks; the bundle additionally ships the 21-row centenarian-preserved-LLI signature, a 24-row intervention-responsive table, the cross-species and tissue-exposure tables, the SenMayo per-candidate and per-target tables, and the full 504-pair clearance grid as per-pair JSON. The catalog is released CC-BY-4.0; the predictor implementation and underlying methodology are held proprietary.
- Reader impact
- An integrated reference catalog covering 84 aging-dysregulated human miRNAs with the upstream curation an academic group, a biotech R&D team, or a generics manufacturer would otherwise have to rebuild from scratch: literature evidence with primary-source citations, López-Otín hallmark mapping, centenarian-signature direction and confidence, predicted target footprint, tissue exposure, SenMayo SASP overlay, cross-species ortholog mapping, and freedom-to-operate clearance across six antisense modulator chemistries against a 5,596-patent corpus. Every (candidate, chemistry) clearance verdict is reproducible from the bundled per-pair JSON against the named patent corpus on the snapshot date. The catalog is field infrastructure: a downstream developer can pick a candidate and start work without further consultation.
- Who can implement
- Any aging-therapeutics, senescence, RNA-therapeutics, or oligonucleotide therapeutics team. The 25 canonical-clean publishable miRNAs and the 6 IP-clean SASP-axis candidates pass the canonical LNA-2'-MOE-PS gapmer 16 nt chemistry without any HIGH or MEDIUM patent encumbrance flag and can be synthesised at canonical specificity without licensing concern against the surveyed corpus. Implementation requires per-candidate chemistry selection (the dossier names the cleanest available chemistry for each encumbered candidate), oligonucleotide synthesis at the recommended chemistry, and standard target-engagement and tissue-exposure validation. See Engagements for commissioned scoping work or custom-catalog commissions covering additional candidate axes (cell-type resolution, additional senescence panels, indication-stratified prioritisation, combination protocols).
- Research note landing HTML, hero callout, headline subsets, tier structure, clearance methodology
- Zenodo deposit DOI 10.5281/zenodo.19952344 · CC-BY-4.0 · 10-page PDF + 48 MB zip · 84 dossiers + 9 audited TSVs + 504 per-pair clearance JSONs + README + LICENSE
Research Note 05 · Published April 2026
A closed-form multi-region human miRNA target binding affinity atlas across 3'UTR, 5'UTR, and CDS, with computational m6A modulation flag. 2,656 mature human miRNAs (miRBase v22) by 19,542 canonical human protein-coding transcripts (GENCODE v47); 153,160,752 (miRNA, transcript, region) predictions in a single SQLite database. Each region applies a closed-form biophysical weakening factor: 3'UTR is the canonical post-transcriptional regulatory region; CDS uses a three-state Markov chain steady-state ribosome-displacement model; 5'UTR uses a scanning-ribosome displacement model with a 40-nt cap-proximity exclusion mask. The CDS pipeline validates against an n=25 literature panel with all three validation gates passing (median weakening 2.60x); the 5'UTR pipeline validates against an n=18 scanning + n=2 cap-proximity literature panel with all four validation gates passing (median weakening 2.00x). The 3'UTR pipeline validates at Pearson R = 0.847 against the 224-record Wee 2012 + Salomon 2015 single-molecule kinetics panel and AUC = 0.9339 multi-site against TarBase v9. Per-site DRACH consensus motif overlap (9.33M sites annotated) flags potential m6A modulation. Two additional artifacts ship with the bundle: a CLIP-seq concordance bench against ENCORI / starBase peaks reports median top-100 enrichment 2.83× and median AUC 0.619 across n=602 miRNAs (apples-to-apples vs TargetScan v8 on the n=254 TS-conserved subset is also reported), and a drug-design demonstration notebook recovering 37/37 published on-target genes across five clinical-stage miRNA therapeutics. The atlas database, README, CHANGELOG, and citation metadata are released CC-BY-4.0 in a single SQLite deposit; the predictor implementation and underlying methodology are held proprietary.
- Reader impact
- An open SQLite database of multi-region miRNA target binding predictions across 2,656 mature human miRNAs by 19,542 canonical human protein-coding transcripts × 3 regions (3'UTR + 5'UTR + CDS) with per-site DRACH m6A overlap counts. All three region pipelines are derived in closed form from cellular context (translation rate, scanning ribosome occupancy, cap-binding exclusion) with no parameter tuning against the validation panels. Three primary query modes: by miRNA × region, multi-region for one (miRNA, gene), and m6A-modulated targets.
- Who can implement
- Any miRNA biology, RNA therapeutics, or oligonucleotide therapeutics team. The atlas covers full miRBase v22 hsa (2,656 mature miRNAs) across three transcript regions per pair (3'UTR + 5'UTR + CDS) with per-site DRACH consensus m6A flag. The deposit is human-only; mouse, rat, per-paralog AGO (AGO1/3/4), per-tissue accessibility, experimentally-validated m6A peak overlay, ASO/siRNA design overlays, and ceRNA network propagation are tracked as the v3.1 capability-pack development queue or available on commission. See Engagements for commissioned scoping work or custom-atlas commissions.
- Research note landing HTML, multi-region overview, validation tables, query patterns
- Zenodo deposit DOI 10.5281/zenodo.19932285 · CC-BY-4.0 · SQLite 69.78 GB on Cloudflare R2 · README + LICENSE + CHANGELOG + CITATION + bench bundle
Research Note 04 · Published April 2026
The GLP-1 family is the most commercially valuable peptide drug class in modern pharmaceuticals. Approved members (semaglutide, liraglutide, tirzepatide, dulaglutide, exenatide) and clinical-stage molecules (retatrutide, cotadutide, survodutide, mazdutide, cagrilintide) collectively underpin a market exceeding USD 100 billion in projected annual sales. The originator IP estate is concentrated and weakly differentiated at the sequence level: most claims rest on small substitution patterns around a conserved central scaffold, and the extant patent corpus does not adequately disclose the full sequence space surrounding each approved peptide. This drop publishes 3,800 novel sequence-distinct analogs across five structural classes (GLP-1R mono-agonists, glucagon mono-agonists, GLP-1R / GCGR dual agonists, GLP-1R / GCGR / GIPR triple agonists, oxyntomodulin-like extended scaffolds) as public-domain prior art under 35 U.S.C. §102. Each candidate ships with a complete bench-execution package: predicted pharmacology, full Fmoc SPPS synthesis route, RP-HPLC purification protocol, radioligand-displacement binding affinity assay, Cisbio HTRF cAMP functional confirmation protocol, sema-class developed-form pharmacokinetic profile, projected receptor occupancy and required weekly dose at the PIONEER-1 anchor, and predicted clinical efficacy at week 26 and week 68 calibrated to STEP-1, SURPASS-2, SURPASS-3, SUSTAIN-7, and PIONEER-1 Phase 3 endpoints. 82 candidates are immediately bench-ready (Kd ≤ 1 nM at primary target, required dose ≤ 100 mg/week at PIONEER-1 anchor); 76 more are formulation-near. A single Tier A candidate is bench-validatable for approximately USD 7-12k and 2-3 weeks of parallel synthesis-and-assay work. Once published, no party may claim composition-of-matter novelty on any of the 3,800 specific disclosed sequences in applications filed after the disclosure date.
- Reader impact
- Defensive publication of 3,800 sequence-distinct analogs forecloses composition-of-matter patent claims on the disclosed sequences under 35 U.S.C. §102 in the United States and the equivalent novelty provisions in major jurisdictions (EPC Art. 54, UK Patents Act §6, JP Patent Act §29, CN Patent Law Art. 22). 158 viable candidates (Tier A bench-ready and Tier B formulation-near) carry predicted Kd, dose, and clinical-efficacy projections sufficient for direct bench validation. The deposit does not block §103 obvious-variants claims; method-of-use, formulation, and combination-therapy claims are unaffected and explicitly preserved.
- Who can implement
- Any peptide medchem team, generics manufacturer, or pharma BD function operating in the GLP-1 family space: Novo Nordisk, Eli Lilly, Boehringer Ingelheim, Zealand Pharma, AstraZeneca, Hanmi, Pfizer, plus generics manufacturers preparing post-CoM-expiration entry on the disclosed scaffolds (Teva, Sandoz, Hikma, Apotex, Aurobindo, Cipla, Dr Reddy's, Sun Pharma, Lupin, Mylan/Viatris, Amneal). Implementation requires per-candidate Fmoc SPPS synthesis, RP-HPLC purification with ESI-MS and AAA quality control, radioligand-displacement binding assay validation, and functional cAMP confirmation. Each per-candidate dossier carries every catalog reference, vendor SKU, buffer recipe, and protocol step required to execute validation without further consultation. See Engagements for commissioned scoping work.
- Disclosure landing HTML, deposit summary and tier system
- Zenodo deposit DOI 10.5281/zenodo.19801439 · 18 files · 46 MB on Zenodo (4,816 files · 123 MB extracted) · CC-BY-4.0
Research Note 03 · Published April 2026
Rybelsus is oral semaglutide at about one-percent bioavailability. The same absorption ceiling applies to the full metabolic-peptide class at three-to-five-kilodalton molecular weight: liraglutide, tirzepatide, exenatide, cagrilintide, retatrutide, survodutide. Ninety-nine percent of every swallowed dose is wasted, and that ratio sets the retail-price floor across the category. The absorption barrier is four problems in series: gastric protease, an eighty-micrometer mucus gel, a tight-junction-sealed columnar epithelium near 1000 ohm-square-centimeter TEER, and first-pass hepatic retention. Rybelsus addresses the first and partially the third. The other three are unaddressed. This drop publishes an integrated multi-mechanism oral tablet at class level: SNAC plus CAGE ionic liquid plus a selective tight-junction modulator plus a PLGA-PEG mucus-penetrating nanoparticle plus a thiolated chitosan mucoadhesive, with the transcellular and paracellular routes physically orthogonal and combining additively on F_epithelium. Predicted human bioavailability is four to fourteen percent theoretical, two to seven percent effective pre-bench, two to fourteen times Rybelsus across all seven peptides simultaneously.
- Reader impact
- The architecture covers all seven clinically relevant 3-to-5 kDa metabolic peptides simultaneously: semaglutide, liraglutide, tirzepatide, exenatide, cagrilintide, retatrutide, survodutide. Liraglutide and exenatide composition-of-matter have already expired in the United States (August 2025 and 2017 respectively); the architecture clears a generic-oral path for both via dedicated Mitragotri-independent claim variants (Claim 22-bis and Claim 43-bis). Publication places the integrated multi-mechanism architecture and 108 numbered claims in the public domain as §102 prior art under CC-BY-4.0, foreclosing downstream patent filings on the integrated route-additive architecture across the full peptide class.
- Who can implement
- Any oral peptide formulator working at 3-to-5 kDa peptide payload: Novo Nordisk, Eli Lilly, Boehringer Ingelheim, Zealand Pharma, AstraZeneca, Hanmi, Pfizer, plus generic manufacturers preparing to enter the post-CoM-expiration generic-oral market for liraglutide and exenatide (Teva, Sandoz, Hikma, Apotex, Aurobindo, Cipla, Dr Reddy's, Sun Pharma, Lupin, Mylan/Viatris, Amneal). Implementation requires per-peptide envelope tuning, CAGE ionic-liquid form selection, PLGA-PEG nanoparticle process calibration, and gastric-corpus versus duodenal claudin-profile bench validation. See Engagements for commissioned scoping work.
Research Note 02 · Published April 2026
A Starship-class Mars departure burn requires the receiver vehicle to aggregate close to its full 1500-ton methalox load in low Earth orbit across a multi-flight tanker campaign. Ship-to-ship cryogenic transfer at this scale has not flown. The aggregation campaign raises six coupled cryogenic fluid-management problems, each absent from SpaceX's public patent record: boil-off suppression at the thousand-ton scale across a multi-week orbital hold; zero-g liquid acquisition at the tank outlet; slosh mitigation at transfer transients; zero-g mass gauging to sub-one-percent aggregate accuracy; cryogenic androgynous docking with fluid transfer at fleet-scale cycle life; and attitude control of a 104-meter mated-tandem stack across live mass transfer. This drop works each pain point from first principles, quantifies bounds at mission-relevant scale, and publishes the integrated six-subsystem architecture as a §102 defensive publication. A companion technical disclosure gives the full integrated architecture at patent-application-grade specificity with a 15-flight worked example, alternative embodiments, and prior-art distinction.
- Reader impact
- The architecture is directly applicable to Starship-class Mars, lunar, and HLS operations, and to every other cryogenic-in-orbit program at scale: NTR, electric-propulsion tugs, commercial depots, Lunar Gateway resupply. Publication places the integrated six-subsystem architecture in the public domain as §102 prior art, foreclosing downstream patent filings on the aggregate claim. Individual component interventions remain as cited prior art under their respective filings (Boeing, ULA, MDA, MERL, Altius/Voyager, NASA KSC).
- Who can implement
- Any aerospace team building a cryogenic orbital aggregation architecture: SpaceX, Blue Origin, ULA, Lockheed Martin Space, Boeing, Sierra Space, NASA MSFC, and third-party propellant-depot operators. Implementation requires vehicle-specific parameter-fitting (tank geometry, MLI installation topology, cryocooler sizing against specific solar array capacity, seal material selection against docking frequency, MPC tuning against specific vehicle inertia). See Engagements for commissioned scoping work.
Research Note 01 · Published April 2026
Bloom's first deployed cohort of solid-oxide stacks fell well short of the ten-year service-life expectation. Three cathode-side mechanisms degrade the cell at once, and commercial mitigation addresses only the most visible. This drop derives each mechanism from first principles and specifies a graded-microstructure cathode architecture that suppresses all three on the existing co-fire process flow. Composition windows and process parameters at patent-application specificity, with a 32-page technical disclosure published alongside the narrative note.
- Market impact
- The architecture applies across most commercial SOFC platforms: Mitsubishi Power, Ceres Power, SolydEra, Weichai, Bosch, Doosan, as well as Bloom Energy itself. A competently-filed and licensed patent family on graded-microstructure cathodes clears an estimated $300M to $700M NPV over the term, with a midpoint near $400M for the specific bundle disclosed. Publication forecloses competitor filings on the integrated architecture as §102 prior art.
- Who can implement
- Any SOFC or SOEC maker running LSCF-family cathodes on MCO-spinel interconnect coatings. Implementation requires thermal-expansion matching to the maker's specific LSCF formulation, powder-size and binder-burnout calibration for the existing tape-cast line, and QA integration. See Engagements for the scope of commissioned follow-on work.
