MatIntel PhysicsCheck — PC-2026-0331-00001-LI

MATINTEL · PHYSICS VALIDATION LABORATORY

NO LLM · DETERMINISTIC

5.4M

Materials records

697K

Peer-reviewed papers

67K

Patents

81K

Synthesis routes

3

CC BY 4.0 databases

Physics Validation Report

PhysicsCheck™

PC-2026-0331-00001-LI ·March 2026 ·PhysicsCheck v0.2.0 ·22 pages

Domain

Semiconductors

Solar

Coatings

Batteries

Hydrogen

Catalysts

Ceramics

Biomaterials

Additive Mfg

Claim Under Review

"

Silicon anode achieves 1,800 mAh/g capacity with 80% retention after 500 cycles at 0.5C — specifically Crystalline Si

Claim Assessment

UNDEMONSTRATED

Physics allows it — no published study has demonstrated this compound performance simultaneously

Investment Risk Tier

MODERATE

Path to LOW: independent validation closes the evidence gap — §2.1, p.6

Dr. Luc Durand, PhD

Pierre and Marie Curie University · CEA Saclay · JRC Ispra · LETI · ICMAB

Scientific Director · MatIntel Physics Validation Laboratory

MATINTEL · PHYSICS VALIDATION LABORATORY

NO LLM · DETERMINISTIC

PC-2026-0331-00001-LI · Summary Card p. 2 of 22

Summary Card

Physics validation · Evidence analysis · Risk assessment

PC-2026-0331-00001-LI

Claim Under Review

Semiconductors

Solar

Coatings

Batteries

Hydrogen

Catalysts

Ceramics

Biomaterials

Additive Mfg

"

Silicon anode achieves 1,800 mAh/g capacity with 80% retention after 500 cycles at 0.5C — specifically Crystalline Si

Physics Feasibility — does physics make it possible?

✓

POSSIBLE

No physics law prohibits this claim.

12 constraints · all pass

!

UNCERTAIN

One or more constraints raised concern.

✕

IMPOSSIBLE

A physics law is violated.

Evidence Strength — has it been proven?

How the claimed values compare to published science

Storage Capacity

MatIntel extracted 663 measurements in 505 papers

Minimum	Median	Maximum	This claim
280mAh/g	950mAh/g	3,200mAh/g	1,800mAh/g

Durability at 500 cycles

MatIntel extracted 312 measurements in 218 papers

Minimum	Median	Maximum	This claim
21%retention	66%retention	96%retention	80%retention

✕

Both values together have no published precedent

No published study reports 1,800 mAh/g AND 80% retention AND 500 cycles together on Crystalline Si. Each exists in the literature separately — the combination does not.

∴

Physics allows it · Both values well above the average and median across published science · Compound performance never demonstrated together → physically possible, not yet proven

Claim Assessment

SUPPORTED

Physics possible. Evidence confirms.

UNDEMONSTRATED

Physics possible. Compound claim not yet proven.

CONTESTED

Evidence contradicts the claim.

INVALIDATED

Physics law violated.

Investment Risk Tier

LOW

Claim supported. Standard diligence.

MODERATE

Validate before committing.

HIGH

Resolve before investing.

DO NOT INVEST

Physics violation confirmed.

Path to LOW — independent electrochemical validation closes the evidence gap. See §2.1, p.6.

Expert Assessment

Expert Review Pending. MatIntel Expert Board covers 9 materials domains. Domain expert review is included with every full PhysicsCheck report.

MATINTEL · PHYSICS VALIDATION LABORATORY

NO LLM · DETERMINISTIC

PC-2026-0331-00001-LI · Table of Contentsp. 3 of 22

§1Table of Contents

Cover

Verdict, risk tier, corpus scale, claim

1

Summary Card

6 decision blocks: claim, feasibility, evidence, assessment, risk, expert

2

§1 Table of Contents

Navigation index

3

§2 Executive Summary

BOTTOM LINE, conviction score, key findings, next steps

4–6

§3 Commercial Readiness

TRL, scale qualifier, FTO signal, key risk

7–8

§4 Parameter Landscape

Benchmark bars, percentile context, compound claim

9

§5 Constraint Scorecard

12-rule verdict table (R01–R12)

10

§6 Detailed Analysis

Per-rule evidence, threshold vs actual, commercial implication

11–12

§7 Patent Landscape

Corpus + live search, signal badges, FTO advisory

13–14

§8 Literature Evidence

Corpus + live literature, semantic ranking

15

§9 Commercial Landscape

Funding signals, competitor activity

16

§10 Expert Commentary

Domain expert review

17

§11 Cross-DB Verification

Three-source agreement table

18

§12 Methodology

Pipeline, score formula, audit trail

19–21

§13 Disclaimer

Legal boundaries

22

MATINTEL · PHYSICS VALIDATION LABORATORY

NO LLM · DETERMINISTIC

PC-2026-0331-00001-LI · Executive Summaryp. 4 of 22

62%

Conviction Score

9

Pass

3

Warn

0

Fail

0

N/A

UNDEMONSTRATED

Claim Assessment

§2Executive Summary

62% (MODERATE) — The physics checks out, but the compound performance has no published precedent. One independent validation study moves this to LOW risk.

BOTTOM LINE

The physics allows this performance — no fundamental law is violated. However, no published study has demonstrated 1,800 mAh/g AND 80% retention simultaneously on Crystalline Si; independent lab validation is required before MODERATE risk can move to LOW.

The claim was evaluated against 12 deterministic constraint rules using three CC BY 4.0 databases (Alexandria, OQMD, Materials Project). The corpus contains 663 measurements of specific capacity and 312 measurements of cycle retention for silicon-based anode materials, drawn from 505 and 218 papers respectively. Specific capacity of 1,800 mAh/g sits at the 80th percentile of the published distribution (typical range 585–1,503 mAh/g). Cycle retention of 80% at 500 cycles sits at the 94th percentile (typical 55–76%). Both values individually are within the published range — but no single study has demonstrated both simultaneously on Crystalline Si.

Conviction score: Feasibility 100% × 40% + Precedent 46% × 35% + Coverage 80% × 25% = 62%.

Key Findings

PASSAll 12 physics constraints passed — no fundamental law prohibits this performance for Crystalline Si.

WARNThe claimed combination of high capacity and high retention is an extraordinary pairing — no published study has achieved both simultaneously at these levels.

WARNNo validated synthesis route achieves this performance at commercial electrode loading (>3 mg/cm²). Lab demonstrations use coin cell format at 1–3 mg/cm².

WARNLiterature consensus shows that silicon anodes at >1,500 mAh/g typically suffer rapid capacity fade due to volume expansion — sustaining 80% retention at 500 cycles is above the consensus.

MATINTEL · PHYSICS VALIDATION LABORATORY

NO LLM · DETERMINISTIC

PC-2026-0331-00001-LI · Conviction Score Detailp. 5 of 22

§2Conviction Score — Component Breakdown

Feasibility × 40% + Precedent × 35% + Coverage × 25% = Conviction

100% × 0.40 + 46% × 0.35 + 80% × 0.25 = 62%

100%

Physics Feasibility

Weight: 40% · Contribution: 40 points

Binary component. 100% if all physics constraint rules (R01–R08) pass, 0% if any fail. For this claim, all 8 physics rules passed — no fundamental law of physics prohibits Crystalline Si achieving 1,800 mAh/g capacity. The theoretical maximum for Si is 3,579 mAh/g (Li₁₅Si₄), placing the claim at 50% of the physical ceiling.

Result: All R01–R08 PASS. No physics violation detected. Full 40 points awarded.

46%

Precedent

Weight: 35% · Contribution: 16.1 points

Percentile rank of claimed values in the published distribution, with compound claim penalty. Individual percentiles: specific capacity at 80th (above median of 1,000 mAh/g), cycle retention at 94th (above median of 63%). Average individual score would be ~87%. However, no published study demonstrates both values simultaneously on Crystalline Si — compound claim penalty reduces to ~46%.

Result: Compound claim penalty applied. Individual properties within range — simultaneous demonstration absent. 16.1 points awarded (of 35 possible).

80%

Evidence Coverage

Weight: 25% · Contribution: 20 points

Corpus completeness for this material-property combination. Silicon anode is one of the most studied anode materials in the corpus: 663 measurements of specific capacity across 505 papers, 312 measurements of cycle retention across 218 papers. Coverage is high (80%) — the corpus has sufficient data to make the constraint evaluation robust.

Result: High coverage. 663 + 312 measurements across 505 + 218 papers. 20 points awarded.

MATINTEL · PHYSICS VALIDATION LABORATORY

NO LLM · DETERMINISTIC

PC-2026-0331-00001-LI · Recommended Next Stepsp. 6 of 22

§2.1Recommended Next Steps

Request from the startup before proceeding:

1

Independent 500-cycle retention test on pouch cell format BEFORE TERM SHEET

Third-party electrochemical validation on pouch cell (not coin cell) at ≥3 mg/cm² Si loading. Must demonstrate 1,800 mAh/g AND 80% retention AND 500 cycles simultaneously under controlled conditions.

Fraunhofer FFB Münster · ZSW Ulm · MEET Battery Research Centre · €2,500–5,000 · 2–4 weeks

2

Full electrochemical test protocol disclosure BEFORE TERM SHEET

C-rate profile, voltage window (V_min–V_max), electrode loading (mg/cm²), electrolyte formulation, and temperature conditions. Without this, no external party can replicate the claim.

3

Synthesis protocol and cost model BEFORE SERIES A CLOSE

Full synthesis route, precursor sources, batch size, yield at pilot scale. Target cost <$50/kg for automotive applications. No validated process achieves this performance at commercial electrode loading.

4

FTO clearance from IP counsel IMMEDIATE

5 patents identified in adjacent space (1 RELEVANT — SK ON CO LTD). FTO opinion required from qualified IP counsel before any commercial deployment or licensing discussion. See §7, p.13.

MATINTEL · PHYSICS VALIDATION LABORATORY

NO LLM · DETERMINISTIC

PC-2026-0331-00001-LI · Commercial Readinessp. 7 of 22

§3Commercial Readiness

§3.1 Technology Readiness Level

TRL 3 Proof of Concept

Performance demonstrated in coin cell format at laboratory scale. No evidence of pouch cell or prismatic cell validation. Crystalline Si anode at this capacity has been demonstrated only under controlled lab conditions with low electrode loading (1–3 mg/cm²).

§3.2 Scale Qualifier

Current demonstrated scale: Lab-scale, coin cell format, approximately 1–3 mg/cm² Si electrode loading, controlled electrolyte environment.

Commercial scale requirement: Pouch cell or prismatic cell format, >3 mg/cm² electrode loading, standard electrolyte, ambient temperature range.

Gap: No validated process achieves this performance at commercial electrode loading. Silicon volume expansion (~300% during lithiation) remains the primary scaling obstacle — strategies that work at <3 mg/cm² loading frequently fail at commercial loading due to mechanical degradation of the electrode structure.

MATINTEL · PHYSICS VALIDATION LABORATORY

NO LLM · DETERMINISTIC

PC-2026-0331-00001-LI · Commercial Readiness (continued)p. 8 of 22

§3.3 IP Status — FTO Signal

RELEVANT 5 patents identified in adjacent space

5 patents identified. FTO opinion from qualified IP counsel required before any commercial deployment or licensing discussion. 1 patent (SK ON CO LTD — lithium-doped silicon-based oxide negative electrode) is directly relevant to silicon anode materials. See §7, p.13 for full patent landscape.

§3.4 Key Risk

No published study has demonstrated the claimed compound performance (1,800 mAh/g + 80% retention + 500 cycles) simultaneously on Crystalline Si at any electrode loading — the compound claim has no precedent, and the scaling gap from coin cell to commercial format introduces additional risk.

MATINTEL · PHYSICS VALIDATION LABORATORY

NO LLM · DETERMINISTIC

PC-2026-0331-00001-LI · Parameter Landscapep. 9 of 22

§4Parameter Landscape & Benchmark

Specific Capacity · Crystalline Si

663

Measurements

505

Papers

80th

Percentile

1,000

Median mAh/g

4,200

Record mAh/g

Cycle Retention · 500 cycles at 0.5C

312

Measurements

218

Papers

94th

Percentile

63%

Median retention

96%

Record retention

Si Specific Capacity — Claimed vs. 663 Published Values

Specific Capacity · Crystalline Si

663 measurements · 505 papers · C-rate / electrode loading vary

1,800 mAh/g

0 585 median 1,000 1,503 4,200 mAh/g

80th

percentile
typical 585–1,503 mAh/g

The claimed value sits above the IQR (585–1,503 mAh/g) but below the published record of 4,200 mAh/g. Physically achievable for Crystalline Si (theoretical max: 3,579 mAh/g).

Cycle Retention · 500 cycles at 0.5C

312 measurements · 218 papers · C-rate and cycle count vary

80%

0% 55% median 63% 76% 100%

94th

percentile
typical 55–76%

80% retention at 500 cycles places this claim at the 94th percentile. At 1,800 mAh/g capacity, published retention is typically <60% at 100 cycles due to volume expansion.

Compound claim: no simultaneous precedent. No published study reports 1,800 mAh/g AND 80% retention AND 500 cycles simultaneously on Crystalline Si. Each property individually falls within the published range — the combination has no precedent in the literature.

Values measured under differing conditions — not all directly comparable to this claim.

MATINTEL · PHYSICS VALIDATION LABORATORY

NO LLM · DETERMINISTIC

PC-2026-0331-00001-LI · Constraint Scorecardp. 10 of 22

§5Constraint Scorecard

12 deterministic constraint rules evaluated. All inputs reproducible. See §6 for detailed analysis of WARN results.

Rule	Rule Name	Verdict	Key Finding
R01	Ashby Bound	WARN	Extraordinary combination — capacity + retention pairing exceeds published envelope
R02	Formation Energy	PASS	Crystalline Si is thermodynamically stable under operating conditions
R03	Composition	PASS	Elemental composition validated — Si is a known anode material
R04	Crystal Structure	PASS	Diamond cubic structure confirmed for Crystalline Si
R05	Phase Stability	PASS	Si stable under lithiation/delithiation within claimed voltage range
R06	Thermal Limit	PASS	Operating temperature within safe bounds for Si anode chemistry
R07	Dimensional Analysis	PASS	All units dimensionally consistent — mAh/g, %, cycles
R08	Theoretical Maximum	PASS	1,800 mAh/g < theoretical 3,579 mAh/g for Li₁₅Si₄
R09	Synthesis Feasibility	WARN	No validated synthesis route at commercial electrode loading (>3 mg/cm²)
R10	Literature Consensus	WARN	Retention claim above consensus — Si at >1,500 mAh/g typically <60% at 100 cycles
R11	Cross-DB Consistency	PASS	Alexandria, OQMD, Materials Project agree within ±15% tolerance
R12	Novelty Check	PASS	Si is a known material — not a novel composition requiring additional validation

MATINTEL · PHYSICS VALIDATION LABORATORY

NO LLM · DETERMINISTIC

PC-2026-0331-00001-LI · Detailed Constraint Analysisp. 11 of 22

§6Detailed Constraint Analysis

Three rules returned WARN. Nine rules passed. No rules failed. Detailed evidence and commercial implications for each WARN finding below, followed by PASS confirmations.

§6.1 WARN Rules

R01 Ashby Bound — Extraordinary Combination WARN

The property map shows 1,800 mAh/g capacity plotted against 80% cycle retention. While each value individually falls within the published distribution, the combination lies outside the demonstrated envelope for Crystalline Si. No published measurement in the corpus of 663 capacity values and 312 retention values achieves this pairing simultaneously.

Threshold vs actual:
Capacity: claimed 1,800 mAh/g vs published envelope max at ≥80% retention: no data point
Retention: claimed 80% at 500 cycles vs published envelope max at ≥1,500 mAh/g: <60%

Commercial implication: The claimed performance occupies an uncharted region of the capacity–retention property space. Until an independent lab confirms this combination, the compound claim remains undemonstrated and represents the primary investment risk.

R09 Synthesis Feasibility WARN

81K synthesis routes in the MatIntel corpus. No route produces the claimed compound performance at electrode loading >3 mg/cm². Published Si anode synthesis methods achieve high capacity OR high retention — not both at commercial scale. Volume expansion of ~300% during lithiation causes electrode mechanical degradation at higher loading.

Threshold vs actual:
Commercial loading threshold: ≥3 mg/cm² vs demonstrated loading at claimed performance: 1–3 mg/cm² (coin cell only)
Synthesis routes achieving compound claim: 0 of 81K in corpus · 0 in live search

CORPUS · 81K ROUTES LIVE · SEMANTIC SCHOLAR LIVE · arXiv PREPRINTS

Live search across Semantic Scholar and arXiv preprints (retrieved 2026-03-28 22:34 UTC) confirmed no additional synthesis routes demonstrating the compound claim at commercial loading. Recent preprints describe Si/C composite approaches and prelithiation strategies — none report simultaneous 1,800 mAh/g + 80% retention at 500 cycles on crystalline Si.

Commercial implication: The startup must disclose their synthesis route and demonstrate reproducibility at pilot scale. Without a validated process at commercial loading, manufacturing risk remains unresolved. TRL 3 — see §3 for full assessment.

R10 Literature Consensus WARN

The published consensus for Crystalline Si anodes at >1,500 mAh/g capacity shows retention typically dropping below 60% within 100 cycles. Claiming 80% retention at 500 cycles at 1,800 mAh/g is above the literature consensus by a significant margin. This does not mean it is impossible — but it exceeds the state of published science.

Threshold vs actual:
Literature consensus retention at >1,500 mAh/g: <60% at 100 cycles vs claimed: 80% at 500 cycles
Claimed percentile: 94th (median: 63%, IQR: 55–76%)

Commercial implication: The startup is claiming performance above the published consensus. The VC should request raw electrochemical data and independent replication before treating this as demonstrated.

MATINTEL · PHYSICS VALIDATION LABORATORY

NO LLM · DETERMINISTIC

PC-2026-0331-00001-LI · Detailed Analysis (continued)p. 12 of 22

§6.2 Passing Rules — 9 of 12 Confirmed

The following 9 constraint rules passed without concern. 3 rules (R01 Ashby Bound, R09 Synthesis Feasibility, R10 Literature Consensus) returned WARN — see §6.1.

PASS

R02 · Formation Energy: Crystalline Si has a formation energy of 0.00 eV/atom — thermodynamically stable under standard operating conditions. All three databases agree (§11).

PASS

R03 · Composition: Si is a validated anode material. Elemental composition is stoichiometrically consistent with lithium-ion battery chemistry.

PASS

R04 · Crystal Structure: Diamond cubic structure (Fd3m) confirmed for Crystalline Si. Structure is well-characterised across all three databases.

PASS

R05 · Phase Stability: Si remains phase-stable under lithiation/delithiation cycling within the voltage window implied by the claim (0.01–1.5V vs Li/Li⁺).

PASS

R06 · Thermal Limit: Operating temperature for Si anode cycling is well within the thermal decomposition threshold. No thermal risk at 0.5C rate.

PASS

R07 · Dimensional Analysis: All units are dimensionally consistent — mAh/g (specific capacity), % (retention), cycles (integer), C-rate (dimensionless).

PASS

R08 · Theoretical Maximum: Claimed 1,800 mAh/g is 50% of the theoretical maximum for Si (3,579 mAh/g for Li₁₅Si₄). No thermodynamic ceiling violation.

PASS

R11 · Cross-DB Consistency: Alexandria, OQMD, and Materials Project agree on formation energy, band gap, and density within ±15% tolerance. See §11 for full verification table.

PASS

R12 · Novelty Check: Silicon is a well-known anode material with extensive published literature (505+ papers in corpus). No novelty flag — additional validation burden for novel compositions does not apply.

MATINTEL · PHYSICS VALIDATION LABORATORY

NO LLM · DETERMINISTIC

PC-2026-0331-00001-LI · Patent Landscapep. 13 of 22

§7Patent Landscape

5 patent(s) flagged RELEVANT — FTO opinion required before scaling. Corpus + live search results below.

5

Total patents

1

Relevant

0

Blocking

Patent ID	Title · Applicant	Filed	Signal
Corpus Search — EPO OPS + GCP BigQuery · Snapshot 2026-03-01
US20260074194.A1	Lithium-Doped Silicon-Based Oxide Negative Electrode Active Material SK ON CO LTD [KR]	2025-11-19	RELEVANT
WO2026038938.A1	Silicon Carbon Composite, Anode Active Material, Secondary Battery LG ENERGY SOLUTION LTD [KR]	2025-08-14	NO CONFLICT
WO2026039222.A1	Electrolytes for Lithium-Ion Batteries with Micro-Sized Silicon Anode UNIV MARYLAND [US]	2025-08-04	NO CONFLICT
Live Search — EPO OPS + Google Patents · Retrieved 2026-03-28 22:34 UTC
US20260074194.A1	Lithium-Doped Silicon-Based Oxide Negative Electrode Active Material SK ON CO LTD · Live confirmation	2025-11-19	RELEVANT
CN-120681761-A	Preparation method of silicon-carbon anode material Live search only — not in corpus snapshot	2025-06-06	NO CONFLICT

MATINTEL · PHYSICS VALIDATION LABORATORY

NO LLM · DETERMINISTIC

PC-2026-0331-00001-LI · Patent Landscape — FTO Advisoryp. 14 of 22

§7Patent Landscape — FTO Advisory

FTO Advisory — action required: 5 patents reviewed. 1 relevant patent identified (SK ON CO LTD). FTO opinion required from qualified IP counsel before any commercial deployment or licensing discussion. This analysis identifies signals only — it does not constitute a legal FTO opinion.

Patent Signal Summary

RELEVANT

US20260074194.A1 — SK ON CO LTD: Lithium-doped silicon-based oxide negative electrode active material. Filed November 2025. Directly relevant to silicon anode compositions. Claims cover lithium-doped Si-oxide materials that may overlap with crystalline Si anode formulations. IP counsel review required.

NO CONFLICT

WO2026038938.A1 — LG Energy Solution: Silicon carbon composite anode material. Claims focus on Si/C composite processing method — distinct from pure crystalline Si anode approach. No conflict identified.

NO CONFLICT

WO2026039222.A1 — Univ Maryland: Electrolyte formulation for micro-sized silicon anodes. Claims are electrolyte-specific — does not restrict silicon anode material composition. No conflict identified.

NO CONFLICT

CN-120681761-A: Preparation method of silicon-carbon anode material. Si/C synthesis route — distinct from crystalline Si approach. Identified via live search only (not in corpus snapshot).

MATINTEL · PHYSICS VALIDATION LABORATORY

NO LLM · DETERMINISTIC

PC-2026-0331-00001-LI · Literature Evidencep. 15 of 22

§8Literature Evidence

12 papers ranked by semantic similarity to claim — corpus (697K indexed papers) + live Semantic Scholar results.

12

Papers found

697K

Corpus papers

4

Live papers added

Authors	Title · Journal	Year	DOI
Corpus Search — 697K indexed publications · Hybrid semantic + keyword
Chan C.K. et al.	High-performance lithium battery anodes using silicon nanowires Nature Nanotechnology	2008	10.1038/nnano.2007.411
Wu H. et al.	Stable Li-ion battery anodes by in-situ polymerization of conducting hydrogel Nature Communications	2013	10.1038/ncomms2941
Liu X.H. et al.	Size-dependent fracture of silicon nanoparticles during lithiation ACS Nano	2012	10.1021/nn302879x
Magasinski A. et al.	High-performance lithium-ion anodes using a hierarchical bottom-up approach Nature Materials	2010	10.1038/nmat2725
Ko M. et al.	Scalable synthesis of silicon-nanolayer-embedded graphite for high-energy lithium-ion batteries Nature Energy	2016	10.1038/nenergy.2016.113
Jin Y. et al.	Self-healing SEI enables full-cell cycling of a silicon-majority anode Science	2017	10.1126/science.aam6014
Li J.-T. et al.	Artificial SEI for silicon anode with long cycle life Advanced Energy Materials	2018	10.1002/aenm.201801234
Zuo X. et al.	Silicon based lithium-ion battery anodes: A chronicle perspective review Nano Energy	2017	10.1016/j.nanoen.2017.01.035
Live Search — Semantic Scholar · Retrieved 2026-03-28 22:34 UTC
Szczech J.R. & Jin S.	Nanostructured silicon for high capacity lithium battery anodes Energy & Environmental Science · 105 citations	2011	10.1039/C0EE00281J
Obrovac M.N. & Christensen L.	Structural changes in silicon anodes during lithium insertion/extraction Electrochemical Solid-State Letters · 45 citations	2004	10.1149/1.1652421
Ryu J. et al.	Practical silicon anodes for high-energy lithium-ion batteries Joule · 89 citations	2021	10.1016/j.joule.2021.06.003
Zhang Z. et al.	New horizons for inorganic solid state ion conductors Energy & Environmental Science · 62 citations	2023	10.1039/D2EE03828A

Note: None of the above papers report the compound claim (1,800 mAh/g AND 80% retention at 500 cycles) simultaneously. 4 new papers added to MatIntel corpus from live search.

MATINTEL · PHYSICS VALIDATION LABORATORY

NO LLM · DETERMINISTIC

PC-2026-0331-00001-LI · Commercial Landscapep. 16 of 22

§9Commercial Landscape

Silicon anode technology is a highly active commercial space with multiple well-funded players pursuing similar performance targets. Major battery manufacturers (Samsung SDI, LG Energy Solution, SK ON, Panasonic) hold significant patent portfolios in the adjacent space. Several startups (Sila Nanotechnologies, Enovix, Group14 Technologies) have raised Series B+ funding for Si anode approaches, with Sila reaching commercial deployment in consumer electronics.

Commercial intelligence is non-peer-reviewed market data and does not affect the physics verdict or conviction score.

Company	Signal	Type	Amount / Date
Sila Nanotechnologies	Series F funding for silicon anode commercialisation; deployed in Whoop 4.0	FUNDING	$590M · 2024-01
Enovix	3D silicon lithium-ion battery — Fab-2 production line operational	PRESS	2024-06
Group14 Technologies	Silicon-carbon composite anode material — partnership with Porsche	FUNDING	$400M · 2023-05
SK ON CO LTD	Patent filed: lithium-doped silicon-based oxide negative electrode	PATENT	2025-11
LG Energy Solution	Patent filed: silicon carbon composite anode active material	PATENT	2025-08

MATINTEL · PHYSICS VALIDATION LABORATORY

NO LLM · DETERMINISTIC

PC-2026-0331-00001-LI · Expert Commentaryp. 17 of 22

§10Expert Commentary

Expert Review Pending. MatIntel Expert Board covers 9 materials domains. Domain expert review is included with every full PhysicsCheck report.

Domain expert review is included with every full PhysicsCheck report. The MatIntel Expert Board covers 9 materials science domains: Batteries, Semiconductors, Solar, Coatings, Hydrogen, Catalysts, Ceramics, Biomaterials, and Additive Manufacturing. Expert commentary does not modify the physics verdict — it provides domain context that a deterministic constraint engine cannot capture, such as emerging unpublished results, known laboratory artefacts, and manufacturing readiness nuances.

MATINTEL · PHYSICS VALIDATION LABORATORY

NO LLM · DETERMINISTIC

PC-2026-0331-00001-LI · Cross-DB Verificationp. 18 of 22

§11Cross-Database Verification

Three CC BY 4.0 databases checked for internal consistency. Agreement tolerance: ±15% between sources.

Database	Property	Value	Data Tier	Agreement	Retrieved
Alexandria	Formation Energy (Si)	0.00 eV/atom	DFT Computed	PASS	2026-03-01
OQMD	Formation Energy (Si)	0.00 eV/atom	DFT Computed	PASS	2026-03-01
Materials Project	Formation Energy (Si)	0.00 eV/atom	DFT Computed	PASS	2026-03-01
Alexandria	Band Gap (Si)	1.12 eV	Experimental	PASS	2026-03-01
Materials Project	Band Gap (Si)	1.11 eV	DFT Computed	PASS	2026-03-01
Alexandria	Density (Si)	2.33 g/cm³	Experimental	PASS	2026-03-01
OQMD	Density (Si)	2.33 g/cm³	DFT Computed	PASS	2026-03-01

3 databases agree within ±15% tolerance. Cross-database consistency confirmed for formation energy, band gap, and density properties of Crystalline Si. Internal corpus data is reliable for constraint evaluation.

MATINTEL · PHYSICS VALIDATION LABORATORY

NO LLM · DETERMINISTIC

PC-2026-0331-00001-LI · Methodologyp. 19 of 22

§12Methodology — How PhysicsCheck Works

Every verdict is produced by the same deterministic pipeline. The same claim, the same corpus snapshot, always produces the same result.

1

Claim submitted

Text · Domain · Material

2

12 Physics Constraint Rules

DETERMINISTIC · NO LLM

R01 · Ashby Bound

WARN

R02 · Formation Energy

PASS

R03 · Composition

PASS

R04 · Crystal Structure

PASS

R05 · Phase Stability

PASS

R06 · Thermal Limits

PASS

R07 · Dimensional Analysis

PASS

R08 · Theoretical Maximum

PASS

R09 · Synthesis Feasibility

WARN

R10 · Literature Consensus

WARN

R11 · Cross-DB Agreement

PASS

R12 · Novelty Flag

PASS

Corpus

OFFLINE

A

Alexandria

3.2M records

O

OQMD

1.3M entries

M

Materials Project

0.9M structures

S

Synthesis corpus

81K routes

Live Search

REAL-TIME

OA

OpenAlex

697K papers

SS

Semantic Scholar

papers

EP

EPO OPS

27K patents

GC

Google Patents

40K patents

Score Calculation

Three independent dimensions · weighted combination

Physics Feasibility

40%

Most critical — physics either allows it or not

Precedent

35%

Where claim sits in published distribution

Evidence Coverage

25%

Completeness of corpus for this material

Conviction Score

(Feasibility × 0.40) + (Precedent × 0.35) + (Coverage × 0.25)

62%

Claim Assessment

UNDEMONSTRATED

Physics possible · compound claim not yet proven in literature

Investment Risk

MODERATE

Validate before committing · path to LOW exists

MATINTEL · PHYSICS VALIDATION LABORATORY

NO LLM · DETERMINISTIC

PC-2026-0331-00001-LI · Methodologyp. 20 of 22

§12Methodology — Score Formula & Audit Trail

Pipeline Flow

1

Claim Parsing

The submitted claim is parsed into material, properties, values, and conditions. No LLM interpretation — deterministic extraction ensures identical inputs always produce identical outputs.

2

RAG Evidence Retrieval

Corpus search across 5.4M materials records, 697K papers, 67K patents, and 81K synthesis routes. Hybrid semantic + keyword retrieval returns the most relevant evidence for the parsed claim.

3

12 Constraint Rules

Each rule (R01–R12) is evaluated deterministically against the retrieved evidence. PASS/WARN/FAIL/N/A verdicts are computed without LLM involvement. Identical claims always receive identical verdicts.

4

Scoring Algorithm

Three-component conviction score computed from rule results and corpus evidence. Compound claim penalty applied when no simultaneous precedent exists in the literature.

5

Report Generation

Structured report rendered from computed results. All content is data-driven. No editorial interpretation in decision-facing sections.

Score Formula

Feasibility × 40% + Precedent × 35% + Coverage × 25% = Conviction

100% × 40% + 46% × 35% + 80% × 25% = 62%

Feasibility (40%): Whether any physics law (R01–R08) is violated. Binary — 100% if all pass, 0% if any fail.

Precedent (35%): Percentile rank of claimed value in published distribution. Compound claim penalty applied: ~46% (no simultaneous precedent reduces from 87% average).

Coverage (25%): Corpus completeness for this material-property combination. 663 measurements, 505 papers = ~80%.

MATINTEL · PHYSICS VALIDATION LABORATORY

NO LLM · DETERMINISTIC

PC-2026-0331-00001-LI · Audit Trailp. 21 of 22

Audit Trail

Report Spec	PhysicsCheck Report Content Spec v1.0
Corpus Age	Last updated 2026-03-01 (30 days ago)
Engine Validation	200+ domain-specific test cases across 9 domains · Nightly canary: 10/10 at 05:30 UTC · matintel.ai/calibration
Constraint Rules	R01–R12 · Deterministic · No LLM in constraint evaluation loop
Databases	Alexandria 3.2M · OQMD 1.3M · Materials Project 0.9M · DiffSyn 81K · OpenAlex 697K
Generated	2026-03-31 10:00 UTC
Report ID	PC-2026-0331-00001-LI

MATINTEL · PHYSICS VALIDATION LABORATORY

NO LLM · DETERMINISTIC

PC-2026-0331-00001-LI · Disclaimerp. 22 of 22

§13Disclaimer

Scope and Limitations. This report validates one specific technical claim against published physics constraints, peer-reviewed literature, and patent evidence. It does not constitute investment advice, a recommendation to invest or not invest, a freedom-to-operate opinion, a materials science research paper, or a comprehensive startup evaluation. The investment decision belongs solely to the recipient. The report characterises the physics risk of the stated claim — no other dimension of the investment is assessed.

Data Sources. Evidence is drawn from three CC BY 4.0 databases: Alexandria (3.2M records), OQMD (1.3M records), and Materials Project (0.9M records), supplemented by 697K indexed papers via OpenAlex and 67K patents from EPO OPS and Google Patents. MatIntel makes no warranty on corpus completeness. Published data may contain errors, biases, or gaps. The constraint evaluation is deterministic and reproducible — but the quality of its output is bounded by the quality of the underlying corpus.

Confidentiality. Confidential. Prepared for the recipient named on the cover page. Distribution, reproduction, or disclosure to third parties requires prior written consent from MatIntel. Contact: [email protected].