The Journal of the American Chemical Society has retracted a 2009 paper on ethylene polymerization after the authors said they were unable to replicate their findings.
The article, “Bimetallic Effects for Enhanced Polar Comonomer Enchainment Selectivity in Catalytic Ethylene Polymerization,”came from the lab of Tobin Marks, a highly decorated — and grant-and-royalty-generating — chemist at Northwestern University.
ChemBark this week reported that the retraction seems to have been prompted by a reader who contacted the blog and the journal in August, raising questions about the data in the article. Check out the post, because it highlights particular trouble spots in the figures.
According to the retraction notice:
The authors have been unable to reproduce the synthesis and spectroscopic characterization of the ethylene/acrylate copolymer described in this article. Accordingly, the authors are retracting this publication due to concerns over the validity of the aforementioned data. The authors regret any confusion that may have been created by publication of this work.
In contrast, the results reported in related publications regarding the synthesis of a binuclear catalyst and its ethylene homopolymerization properties,1 and on ligand steric and fluoroalkyl substituent effects on enchainment cooperativity and stability in bimetallic nickel(II) polymerization catalysts,2 have been reproduced.
The two references are to 1) Rodgriguez, B.A.; Delferro, M.; Marks, T.J. Organometallics 2008, 27, 2166-2168; and 2) Weberski, M.P.; Chen, C.; Delferro, M.; Marks, T.J. Chem. Eur. J. 2012, 18, 10715-10732.
The study has been cited 53 times yet to be cited, according to Thomson Scientific’s Web of Knowledge.
Marks was the subject of a profile in Crain’s Chicago Business last April, which reported that he has brought Northwestern “more than $100 million in grants and royalties” since coming to the school in 1970.
He ranks fifth among Chicago-area inventors, with 126 U.S. patents since 1980, according to the database compiled by Ms. [Deborah] Strumsky, a researcher at the University of North Carolina at Charlotte. Mr. Marks estimates his total patent count tops 200.
Marks didn’t respond to our request for comment. But ChemBark said its attempt to reach him by email was met with this reply from Northwestern:
Your inquiry to a Northwestern University professor, Tobin Marks, was referred to me, as I am the chief communications officer for Northwestern.
As you note in your email, two Northwestern University faculty members and a former Ph.D. student have retracted a publication that appeared several years ago in the Journal of the American Chemical Society. The authors are Brandon A. Rodriguez, who received a Ph.D. from Northwestern in 2009, Tobin Marks, a professor of chemistry, and Massimiliano Delferro, a research assistant professor of chemistry. The article was retracted because the authors were unable to reproduce a portion of the data described in the article.
Northwestern University has established processes and procedures for reviewing issues relating to research integrity. If concerns were to be raised regarding the data that was the subject of the retraction, the University would use those procedures in its review. Part of those procedures, which follow the steps mandated by the federal government, is that any review remains confidential.
Thank you for your interest in Northwestern.
Best wishes,
Alan Cubbage
ChemBark promises a follow-up post that might shed more light on this case.
Update, 12:30 p.m. Eastern: Corrected number of citations with strike-through. For some reason, Thomson Scientific has two entries for this paper, one of which said zero citations. The other, which we’ve now found, says 53.
Hat tip: Rolf Degen
If a JACS paper doesn’t get cited even once in four years, it could possible imply that many of the readers had concerns about it.
I am no polymer specialist, but that last sentence by Cubbage is extremely worrisome to me: “steps mandated by the federal government, is that any review remains confidential”. How can one promote transparency, change and reform in science and science publishing when a review is forced, according to Cubbage, by the US Government, to stay mum? Does the same apply to ORI “reviews” or “reports”? Maybe a US scientist – or ORI – would care to explain in more detail this law and/or provide a web-link because this is extremely important.
That official reply from Northwestern is a big flashing warning sign that the university is aware of problems and actively covering up. They seem to be saying that research integrity laws and rules are a gag order from the government. But no other university interprets it that way.
Since anyone who looks at the paper can see that it appears to have been faked, the university could have simply stated that investigations are underway etc etc. Instead they seem intent on not investigating, not commenting and not taking responsibility.
Actually a lot of universities interpret CFR 93 that way. They often refuse to release reports of investigations, citing that rule’s confidentiality provision. But we don’t see things that way. See the second half of this post: http://www.retractionwatch.com/2012/02/24/climate-science-critic-wegman-reprimanded-by-one-university-committee-while-another-finds-no-misconduct/
It makes the difference between incompetence and malfeasance. Note that CFR 93 (according to the quoted section in your previous blog) suggests confidentiality for misconduct proceedings, not for misconduct. Regardless of CFR 93 the university is free to tell us whether they think misconduct has occurred at their institution. They can withhold information as to what disciplinary or other actions have been taken.
Note the official statement says “If concerns were to be raised regarding the data…” They are clearly pretending that no concerns have been raised, and that was the response to a letter specifically raising such concerns!
I think you are incorrect here, Ivan. The Federal Regulations on research misconduct require that federal agencies and the research institutions under their jurisdiction maintain the confidentiality of the individuals and the processes involved in assessment, inquiry and investigation. The regulatory clause exception for informing “those with a need to know” was intended to include institutional officials (like chairs, deans, counsels, integrity officers, financial officers, IT investigators, etc.) who would be involved in the handling of the case at the institution. Complainants are generally kept informed in confidence of the status of the case as it proceeds. Witnesses, especially coauthors, are also contacted in confidence. Lawyers for respondents and court officials in the case of legal filings and suits, and members of Congress with oversight authority over the agencies, may also be informed if they have a need to know. Editors, whether as complainants or as persons who may have evidence about submitted manuscripts or status of papers, may also be informed.
But reporters, bloggers, critics, and curious persons, as well as other individuals and officials do not have such a “need to know” and would not fall under this exemption from the federal (and the institutional policies) confidentiality mandates.
Nonetheless, that obligation to maintain confidentiality does not exist when the institutional or federal case is closed with a finding of research misconduct. At that misconduct finding stage, the institutions and federal agencies may make public such findings and administrative actions (sanctions) imposed. The Office of Research Integrity (ORI) has done this since 1992, making public statements that name the person who committed the specific acts of falsification, fabrication, and/or plagiarism, their position and institution, the specific data or text containing the misconduct, any publications that require or have been retracted or correction, and the term of action (often debarment from federal funding, or requirement for supervision and certification of any HHS/NIH-supported research for a given period (1, 2, 3, 5, or 10 years, or lifetime). The reports on which ORI based its findings are subject to federal Freedom of Information Act requests by reporters, lawyers, and the public. Many institutions also release such reports finding research misconduct.
However, if the ORI (or the institution) does not find that significant research misconduct was committed, then the requirement for confidentiality continues — unless the respondent, who was found not to have committed research misconduct, makes a formal request for a public statement or a letter that they can use.
According to the JACS website, the paper has been cited 60 times. Which means that there is a problem with Retraction Watch’s method of finding citations. Go to the abstract site of the paper and you’ll see the citing papers.
Right. And Google scholar gives it away:
http://scholar.google.de/scholar?hl=de&q=%E2%80%9CBimetallic+Effects+for+Enhanced+Polar+Comonomer+Enchainment+Selectivity+in+Catalytic+Ethylene+Polymerization%2C%E2%80%9D&btnG=&lr=
Thanks for flagging this. We’ve corrected the citations, and added an update at the end about how this error happened.
It seems that the first record in WoS refers to a correction published in 2010 to the original 2009 article.
the paper was published the year of the student’s graduation. it seems to me that he was in a hurry to graduate in four years and took some degree of creative license in prepping some data that he thought the work required to be published. i sincerely doubt professors marks or delferro would knowingly publish fabricated data, so it looks like this was just a maverick student who wanted to immediately start at Dow rather than hang around another year or two until he had real results
the organometallics paper from the same authors is also troubling, in the SI, the scheme S4 and scheme S5 appear to be the same spectra with the peak at 34.2 deleted in scheme S5. http://pubs.acs.org/doi/suppl/10.1021/om800208f
The spectra certainly look similar, but I don’t think they look identical. The peak at ~21 ppm looks different (better shimmed in Scheme S5). Further, in Scheme S4 there’s a tiny peak at ~24 ppm that’s absent in Scheme S5. There are other differences.
Iron Chemist, I think you should look at the noise of the spectra here. The noise shouldn’t be the same for different NMR experiments.
The spectra look like they have identical baselines to me too.