Investigation Report: Phase change doped sand storage for isothermal thermoelectric heat delivery (DOI: 10.20517/energymater.2026.73)

Academic fraud report · Geng Detector

Summary

Verdict: Questionable (insufficient evidence for confirmed misconduct). The original Chinese report flagged four potential issues: duplicate Figure 9/10 numbering, use of Gemini 2.5 Pro for concept figures, a truncated word in Table 2, and absence of pixel-level image evidence. A careful re-review found that the supplied manuscript text was truncated at line 973 and did not contain the captions, Declarations, or Table 2 content cited in the original findings. Consequently, the figure-numbering, AI-generation, and table-truncation claims could not be substantiated from the available material. The fourth finding (no experimental images, pure COMSOL simulation) was confirmed: the paper is a finite-element simulation study, and the derived calculation keff = 0.70 × 1.4 + 0.30 × 0.55 = 1.145 W m⁻¹ K⁻¹ is arithmetically consistent. Confidence is limited by missing manuscript sections; no fabricated data is apparent in the provided text.

Verdict

🟡 Questionable — based on the supplied truncated manuscript (lines 1–973), none of the originally flagged concerns can be independently verified, and no evidence of data fabrication or image manipulation is present in the available text. The findings remain unconfirmed rather than refuted.

Key findings

⚠️ Figure 9/10 numbering inconsistency — Not verifiable in supplied text. The manuscript fragment does not contain the phrase "Figure 10 presents six complementary analyses" nor the caption "Figure 9. Cyclic stress response and material-pair selection." Highest figure reference in provided text is Figure 9G.
⚠️ AI-generated concept figures (Gemini 2.5 Pro) — Not verifiable. The Declarations section is absent from the supplied fragment; no mention of Gemini 2.5 Pro appears in the available text.
⚠️ Table 2 truncation ("Sensible storag") — Not verifiable. Table 2 content is not included in the supplied fragment, though "Iron filings" is referenced at line 379.
✅ Pure simulation study confirmed — Line 432 identifies the work as a COMSOL Multiphysics simulation. No wet-lab images (Western blots, gels) or error-bar-bearing biological data are present.
✅ Mathematical derivation consistent — Resistance calculation at lines 797–798 is correct: keff = φs·ks + φd·kd = 0.70 × 1.4 + 0.30 × 0.55 = 1.145 W m⁻¹ K⁻¹; subsequent R = ln(Ro/Ri)/(2πk) formulation is mathematically valid.

Evidence highlights

Manuscript truncation point: provided text ends at line 973, excluding final figure captions, Declarations, and likely Table 2.
Simulation platform declaration: line 432 of the supplied text.
Numerical check: lines 797–798 — keff = ϕsks + ϕdkd = 0.70 × 1.4 + 0.30 × 0.55 = 1.145 W m⁻¹ K⁻¹.
Doi: 10.20517/energymater.2026.73 (https://dx.doi.org/10.20517/energymater.2026.73).
Authors (per report): Qirui Ding, Jiacheng Liu, Changhui Song, Lili Zeng, Weicheng Cui.
Journal: Energy Materials, 2026.

Notes

All four original findings carry "insufficient evidence" status due to manuscript truncation; they are flagged for verification rather than confirmed.
Request the complete PDF, including the figures, figure captions, Table 2, and Declarations, to revisit the figure-numbering, AI-generation, and table-truncation claims.
If misconduct is suspected at the simulation-data level, the authors' raw COMSOL .mph files and any ANSYS .wbpj project files should be requested for parameter audit.
AI-generated conceptual figures (Figures 1–2) should be reviewed for internal scientific consistency, even though their generative origin is not, by itself, a violation of typical journal policy when disclosed.
Authors are encouraged to correct any figure-numbering errors and Table 2 typos if confirmed in the full manuscript.