Nearly 12 years after Science published a nanoparticle synthesis paper, the journal has decided to retract it. The withdrawal is in response to a National Science Foundation report that found the paper’s authors used poor research practices and misrepresented data, although the report also concluded that they did not engage in misconduct.
The paper, which is on RNA-mediated formation of palladium nanoparticles, was written by Bruce E. Eaton, Daniel L. Feldheim, and grad student Lina Gugliotti—all of whom were at North Carolina State University (NCSU) when it was published (Science 2004, DOI: 10.1126/science.1095678). Eaton and Feldheim are now at the University of Colorado.
The team reported in 2004 that putting select RNA sequences into an aqueous solution of tris(dibenzylideneacetone)palladium(0) caused hexagonal-shaped inorganic Pd nanoparticles to form. The researchers concluded that the RNA catalyzed the process. C&EN wrote a story about the work at the time (April 19, 2004, page 9), asking Gerald F. Joyce, of Scripps Research Institute California, to provide expert comment. Joyce was enthusiastic about the results: “We now realize that RNA is adept at synthesizing inorganic materials, and this discovery takes the field of RNA-directed evolution in a whole new direction.”
The development instead veered off in an unexpected direction. A group led by NCSU physical chemist and biophysicist Stefan Franzen, a one-time collaborator of Feldheim’s and Eaton’s, tried to better understand the RNA-mediated process but couldn’t replicate the work. Franzen’s team found that the reaction caused hexagonal particles to form but that the shapes contained carbon and thus were not strictly inorganic Pd nanoparticles. The researchers also determined that RNA didn’t seem to play any role in the particles’ formation. Franzen says he discussed his observations with Feldheim and Gugliotti and also tried to communicate with Eaton about them. Eaton says Franzen never contacted him at this time or later on. In any case, the researchers weren’t able to resolve the matter.
In 2006, an NCSU administrator suggested that the seeming irreproducibility of the research be evaluated as a possible case of scientific misconduct. The resulting investigation concluded in a 2008 report that electron microscopy experiments to study particle formation, shape, and composition had not been carried out correctly. It also concluded that solution conditions needed to solubilize the organopalladium starting material had not been described properly in the Science paper. The report charged Feldheim, Eaton, and Gugliotti with negligence but not with research misconduct. The University of Colorado also investigated the matter, concluding that there was nothing wrong with the study’s methodology or with the way the Science paper was written, a far different conclusion than NCSU’s. Meanwhile, NSF launched its own investigation.
The dispute over a Science paper has been raging for 12 years.
May 2004: Eaton, Feldheim, and Gugliotti publish their Science paper.
Dec. 2006: NCSU launches its investigation into the research.
June 2008: NCSU concludes its investigation.
Sept. 2008: NSF launches its investigation.
Sept. 2008: Franzen and coworkers submit letter to Science, surfacing problems with the original work.
Aug. 2010: Feldheim, Eaton, and coworkers publish a Journal of Materials Chemistry paper that repeated the 2004 work and confirmed its findings.
Feb. 2013: Franzen, De Yoreo, and coworkers publish Particle & Particle Systems Characterization paper concluding that the original work was flawed and that RNA does not catalyze the formation of inorganic Pd particles.
Sept. 2013: NSF OIG posts preliminary findings online.
Oct. 2015: NSF posts final decision memo online.
Feb. 2016: Science retracts the 2004 paper.
At one point during this period, Franzen asked Eaton and Feldheim to cooperate on a technical note to Science about the problems, but they declined, Eaton says, because “the note was full of technical false statements” and the request “was posed as an ultimatum.” So Franzen and coworkers instead ended up airing the problems in a September 2008 letter to Science. In a reply, Feldheim, Eaton, and Gugliotti claimed that a redo of the electron microscopy work showed that the original findings had been valid and that they had made the organopalladium reagent soluble by adding a cosolvent, although that information hadn’t been included explicitly in the original paper.
Later on, the Colorado group repeated the complete 2004 study and reported in a scientific paper having confirmed all their original findings (J. Mater. Chem. 2010, DOI: 10.1039/c0jm02050h). But Franzen subsequently published a Technical Comment in the same journal (J. Mater. Chem. 2013, DOI: 10.1039/c3tb20820f) that critiqued the methodology and findings of the 2010 paper. Franzen; materials scientist James J. De Yoreo, then at Lawrence Berkeley National Laboratory; and coworkers also reported in another journal (Part. Part. Syst. Charact. 2013, DOI: 10.1002/ppsc.201200114) that the particles made by the process described by Eaton and Feldheim are nothing more than aggregates of the starting material. This is similar to how snowflakes form, a process that doesn’t require RNA to proceed, Franzen says.
Preliminary findings of NSF’s investigation, posted online in 2013 by the agency’s Office of Inspector General (OIG), were that the 2004 Science study had been conducted without due care, that the findings were not fully supported by the paper’s data, and that the authors had negligently failed to correct errors. Eaton tells C&EN that the OIG report contained 11 incorrect statements and that he and his coworkers responded to it by pointing out “documented facts proving the report was false.”
NSF’s recently issued final decision memo suggests the agency’s OIG did not find those arguments persuasive. In it, NSF’s OIG stated that Eaton, Feldheim, and Gugliotti had “recklessly falsified research data” and had departed significantly from accepted practices. The OIG recommended a finding of research misconduct, that a letter of reprimand be issued to the researchers, and that they publish a correction in Science.
NSF eventually declined to accept its OIG’s recommendation for a finding of research misconduct. But the agency did conclude “that the subjects’ actions were a significant departure from standard research practices.” It issued the letter of reprimand and declared the subjects ineligible for NSF funding until they submitted a correction to Science.
Editor-in-Chief Marcia McNutt says Feldheim did submit the required correction but that the editors at Science “do not think a correction is appropriate given the concerns” raised by the OIG report. The journal retracted the paper instead (Science 2016, DOI: 10.1126/science.351.6273.569-a). Eaton says he corresponded extensively with McNutt in an effort to get her to rethink the retraction decision, but McNutt insisted retraction was the appropriate move. Eaton and Feldheim did not agree to the retraction. “I truly know that RNA-mediated formation of nanoparticles does reproduce, as reported in Science and the other publications,” Eaton says.
Franzen says the process hasn’t been easy or pleasant. “It has plagued me for the last nine years. I’ve been investigated, I’ve been threatened by lawyers, and I’ve been charged with stealing samples. But I didn’t do anything wrong, and I have been exonerated of everything.”
He insists Gugliotti is actually blameless because her lab work was flawless, her lab notes were complete, and she didn’t write the 2004 Science paper. C&EN was unable to locate Gugliotti by press time.
Asked to comment on whether issues with the 2004 paper have adversely affected related scientific research, Joyce said: “Science is a self-correcting process—even 12 years after the fact.”