Investigation Report: "Breaking the Fixed Output: Harnessing Photonic Reabsorption and Photothermal Effects for Tunable NIR Waveguiding in a Flexible Organic Crystal" (Angew. Chem. Int. Ed., 2026; DOI: 10.1002/anie.5856537)

Academic fraud report · Geng Detector

Summary

This report evaluates eight concerns in a 2026 Angewandte Chemie paper on flexible organic crystal NIR waveguiding. Verdict: doubtful (yellow). The most significant concern is Figure 5g, where serial infrared thermograms at varying 660 nm irradiance show identical crystal outlines, noise patterns, and hotspot positions, differing only in pseudocolor—a pattern consistent with recoloring of a single captured frame rather than independent sequential imaging. Additional concerns include implausibly overlapping three-point bending curves (Figure 1g), overly smooth wavelength-distance scatter (Figure 3g,h), an ideal six-point linear temperature-irradiance relationship (Figure 4f), zero-degradation photothermal and optical switching cycles (Figure 4h, 5j), and suspiciously close optical loss coefficients for straight vs. bent crystals (0.0812 vs. 0.0824 dB/mm). Reference list contains duplicate entries, misparsed author names ("A. Broadband"), and dense self-citation (>15 papers by Hongyu Zhang). None of these findings constitutes definitive proof of fabrication; the issues are consistent with selective smoothing, data idealization, or editorial carelessness. Confidence is moderate. Cautious: request raw thermal video, force-displacement curves, and raw spectra before escalation.

Verdict

Doubtful (Yellow). Multiple figures show implausibly ideal data textures and one figure (Figure 5g) shows strong prima facie evidence of single-frame pseudocolor reuse. No finding is independently conclusive; raw data are required for confirmation. Not recommended for formal editorial complaint until the authors respond.

Key findings

Figure 5g (thermal imagery, severe concern): Serial thermograms at increasing 660 nm laser irradiance share identical crystal shape, background noise, and hotspot location; only pseudocolor differs. No camera jitter, airflow artifacts, or baseline drift is visible between frames. The text acknowledges repeated laser-power switching, which should yield independent captures.
Figure 1g (mechanical curves): Three three-point bending stress–strain curves for DPTD crystals overlap almost perfectly in the 0–0.002% elastic region with no specimen-to-specimen noise, despite acknowledged pre-measurement dimensional variation.
Figure 3g,h (waveguide peak wavelength vs. propagation distance): Data follow a near-textbook staircase/linear trend with negligible scatter; spectrometer measurements typically exhibit ±1–2 nm random fluctuations.
Figure 4f (max surface temperature vs. irradiance): Six data points fall on a near-perfect line with no visible outliers despite acknowledged IR-camera calibration disruption near 150 °C.
Figure 4h (100 photothermal cycles) and Figure 5j (200 on/off cycles): "Highly consistent" and "no signal decay" performance is unusual for an organic crystal cycled near its 189 °C melting point (peak ~147 °C).
OLC values (text p. 5): Straight crystal 0.0812 dB/mm vs. bent 0.0824 dB/mm—only 1.5% difference, atypical for a visibly bent waveguide.
Reference list metadata errors: Duplicate entries for A. V. Kumar et al. 2022 ([50] and [63]) with identical title and volume mismatch; "A. Broadband" appears as an author (clearly parsed from a title). Dense self-citation: ≥15 of 69 references list Hongyu Zhang, mostly 2024–2026.

Evidence highlights

Figure 5g: pixel-level alignment of crystal contour and noise across all irradiance panels; only the colorbar LUT changes.
Figure 1g: zero-noise superposition of three nominally distinct specimens in the elastic regime.
Figure 3g,h: absence of measurement noise around the fitted trend.
Figure 4f: R²-approaching-unity distribution of six points in a multi-coupled physical process.
Text p. 5: OLC values differing by 0.0012 dB/mm to four decimal places.
References [50] and [63]: duplicate author, year, journal; volume differs; "A. Broadband" parsed from a title.

Notes

Severity scale used by the source: 🟠 (Figure 5g) and 🟡 (all others). Concerns do not yet meet the threshold for institutional or editorial referral.
Request original data: Figure 5g raw thermal video/timestamps, Figure 1g raw force–displacement curves, Figure 3g,h raw spectra, Figure 4f raw IR temperatures per cycle, full OLC attenuation curves for straight vs. bent crystals.
Limitations: visual analysis of a PDF cannot confirm pixel-level identity; smoothing, idealization, or selective display remain plausible non-fraudulent explanations. Numeric evidence retained as reported.