Integrity Review Report: Binary anion and cation co-doping enhance sulfide solid electrolyte performance for all-solid-state lithium batteries

Academic fraud report · Geng Detector

Summary

This report flags a paper published in Energy Materials (2024) as highly suspicious (DOI: 10.20517/energymater.2023.78). Two serious internal contradictions were confirmed. First, the experimental section states that cyclic voltammetry (CV) was performed at a scan rate of 0.1 mV s⁻¹, while the results section describes the same CV analysis at 1 mV s⁻¹ — a tenfold discrepancy that fundamentally changes peak current and position per the Randles–Sevcik equation. Second, the anode material is inconsistently reported: the cell assembly section specifies a Li/In alloy (Li/In = 3:7), while the electrochemical measurement section describes lithium foil as the anode for CV. These two parameters are not interchangeable in solid-state electrochemistry. Additionally, the reported XPS binding energies for Sn 3d₅/₂ appear ~1 eV below standard reference values for both Sn⁴⁺ (485.4 eV vs. expected ~486.5–487 eV) and Sn⁰ (483.7 eV vs. expected ~484.6–485 eV), suggesting potential calibration or peak-fitting issues. Limitations: figures could not be inspected at pixel level, and XPS shifts may partly reflect instrumental calibration rather than fabrication.

Verdict

🟠 Highly suspicious. Two severe internal contradictions (CV scan rate and anode identity) were verified against the manuscript text and are difficult to reconcile with carelessness alone.

Key findings

CV scan rate contradiction (Severity: 🔴, Confirmed): Experimental section states 0.1 mV s⁻¹; Results section describes the CV at 1 mV s⁻¹. A 10× scan-rate error is not trivial — peak currents scale with the square root of scan rate per the Randles–Sevcik equation, and peak separations also change.
Anode material contradiction (Severity: 🔴, Confirmed): Cell assembly section specifies Li/In alloy (Li/In = 3:7) as anode; CV electrochemical-measurement section specifies lithium foil. In all-solid-state lithium batteries, Li/In and Li metal have different interfacial chemistry, reference potentials, and stability windows against sulfide electrolytes.
XPS binding-energy deviation (Severity: 🟡, Confirmed within text): Sn 3d₅/₂ reported at 485.4 eV for Sn⁴⁺ and 483.7 eV for Sn⁰ — both ~1 eV below typical reference values (Sn⁴⁺ ~486.5–487 eV; Sn⁰ ~484.6–485 eV). May reflect missing C 1s charge correction or subjective peak fitting.
Raw-data access (Severity: 🟡, Insufficient evidence): Core conclusions rest on XRD, Raman, EIS Nyquist plots, and long-cycle data presented as line figures with no error bars or replicate counts reported in the text (200 cycles at 0.1 C mentioned in Figure 7 caption). Pixel-level image inspection was not possible from the supplied PDF.

Evidence highlights

Page 5 (Experimental): "employing a scan rate of 0.1 mV s⁻¹" for CV.
Page 7 (Results and Discussion): "at a scan rate of 1 mVs⁻¹" when discussing the same CV (Figure 4).
Page 4 (Assembly of ASSLBs): "Li/In alloy was used as the anode"; "Li/In alloy (Li/In = 3:7) was positioned on the other side."
Page 5 (Electrochemical measurements): "lithium foil as anodes" for the CV setup.
Page 9 (XPS analysis): "Sn is in the +4 valence state, with binding energies of the Sn 3d5/2 peak at 485.4 eV"; "elemental Sn peak at binding energies of 483.7 eV."
Figure 7 caption mentions cycling stability and Coulombic efficiency; main text references "after 200 cycles at 0.1 C".
DOI: 10.20517/energymater.2023.78

Notes

The two confirmed contradictions (scan rate, anode material) are mutually independent and both appear in the authors' own wording, making copy-paste errors or template mismatches a plausible partial explanation — but the magnitude and electrochemistry-relevance of the errors warrant author response.
XPS shifts could in principle reflect legitimate instrument calibration differences, but the simultaneous ~1 eV offset on both Sn⁴⁺ and Sn⁰ peaks is suspicious and should be checked against the authors' C 1s reference (commonly 284.8 eV).
Recommended follow-up: request raw .vms or equivalent XPS files, full CV parameter logs, the original cell-assembly notebook entry, and any replicate datasets behind Figures 1–7 and the supplementary materials.
This AI-assisted report is an indicator only; formal misconduct findings require institutional investigation.