Physicists with retraction for a “pattern that was unphysical” lose another for manipulation

journal of applied physicsIn September, we wrote about the retraction of a physics paper for “a pattern that was unphysical.”

The team, whose first author, R.K. Singhal refused to sign the notice, has had another paper retracted, this one in the Journal of Applied Physics. Here’s the notice for “Study of electronic structure and magnetization correlations in hydrogenated and vacuum annealed Ni doped ZnO:”

Subsequent to publication of the above article, was discovered that the data involving the Ni 2 XPS spectra shown in the inset in Figure 4 appeared to have been manipulated, and as a consequence, the results presented should not be relied upon and may be scientifically unsound.

AIP Publishing LLC therefore retracts this article.

The paper has been cited three times, according to Thomson Scientific’s Web of Knowledge.

Update, 11 a.m. Eastern, 6/26/14: A commenter points out that this is actually the third retraction for this group. Here was the second.

9 thoughts on “Physicists with retraction for a “pattern that was unphysical” lose another for manipulation”

  1. In the Figure 4 insert, the three curves appear to be duplicates that have been offset to give three different peaks. The data also goes noisy-smooth-noisy, which doesn’t look realistic.
    I notice that the journal is retracting the paper without reference to the authors.

  2. As mentioned by Dan Zabetakis, the inset in Fig. 4 is a ridiculous collage. Three time the same signal (red, black, blue), glued with hand-done noise at low and high energy. The result is a noisy signal which becomes suddenly clean between 851 and 858 eV, followed by noise… The most hilarious part is the noise just below 858 eV, which seems to be taken from a 2D Brownian motion trace.
    I can’t believe that referees didn’t catch such a mockery. The insert was probably not present in the first submitted manuscript.

    1. My interpretation is that the author used a spline-smoothing function on the lines in a graphics program. That would give the lines an excessively curvy character. It might have been Excel or Illustrator, but I don’t usually use that function.

  3. They also published a corrigendum for a 2010 article:
    Again, the problem deals with XPS data. Some details are given, though:
    “It appears that the […] spectra […] had been mistreated due to a series of wrong processes during data analysis i.e., the normalization, background subtraction and smoothing processes”.

    For sure, a “pattern” is emerging.

    1. Yes, you’re right. Using my hunting instinct during lunch break I ended up with the following list (I think I mainly cite articles themselves here).

      Covered by RW: (J. Phys. D 44, 165002) (Appl. Surf. Sci. 257, 585) (J. Appl. Phys 115, 239902)

      Found by Sylvain – a corrigendum, no retraction: (J. Mag. Magn. Mater 366, 64)

      additionally: (Mater. Lett. 65, 1485)

      there’s an erratum: (J. Appl. Phys. 107, 113916)

      some corrigenda: (Mater. Chem Phys. 126, 998) (J. Phys. Condens. Matter. 19 326201)

      seemingly not (yet?) influenced: (Mater. Chem. Phys. 132, 534) – Fig. 2a is the prima candidate for giving a reason for retraction, spectra for CeO2:Htd and CeO2 seem to be very similar. (Mater. Lett.65, 825) – this one is not obvious (Appl. Phys. Lett. 98, 092510) – Fig. 2a presents spectra labeled as TiO2 and TiO2:H:Ht, they should therefore be different? This one is heavily cited! (Appl. Surf. Sci. 257, 1808) – this is a comment on a competitor’s work who found different spectra! Wow! (Appl. Surf. Sci. 257, 1053) – I have strange feelings when looking at Figs. 1b and the inset in 1c (at least!), funny missing data points at different energy also in 1e. (Appl. Phys. Lett. 97, 172503) – again the spectra from heated and untreated samples are VERY similar (Fig. 2) (Phys. Stat. Sol. 207, 2373) – Compare eventually Figs. 11 and 12. For example. (Solid State Commun. 150, 1154) – Figs. 2a and c (J. Alloys Comp. 496, 324) – it now starts to become ridiculous. I have seen the spectra shown in Figs. 6 and 7 in SEVERAL of the other listed works as well! Same noise and kinks but different samples, even materials. Most of these papers are seemingly from 2010 or early 2011. No one had any idea? Jan-Hendrik Schön is calling! Shame on the coauthors

      1. There is definitely a similarity to Jan-Hendrik-Schoen frauds. I quote a different excerpt of the corrigendum which Sylvain Bernès found (R.K. Singhal, A. Samariya, S. Kumar, Y.T. Xing, U.P. Deshpande, T. Shripathi, E. Saitovitch, “Corrigendum to “Defect-induced reversible ferromagnetism in hydrogenated ZnO:Co” [J. Magn. Magn. Mater. 322 (2010) 2187]”, “Journal of Magnetism and Magnetic Materials”,
        “[. . .] The concern is that the background features look identical in these Co 2p spectra despite a shift observed in the spectrum of hydrogenated sample. This should not have been so. [. . .]”

Leave a Reply

Your email address will not be published. Required fields are marked *