EDITORS' NOTE: This article originally was published on Jan. 19, 2014.
In 2004, two colleagues at N.C. State asked professor Stefan Franzen for his help. They sold him on research they said could create world-changing inventions through a process that mimicked natural biological evolution.
Scientists, they claimed, could devise tools to split water into oxygen and hydrogen, producing an endlessly renewable and clean fuel. Or they could create superconductors to transmit electricity at room temperature, revolutionizing the business of distributing power.
It was a heady and intellectually stimulating time, a chance to work on a project that combined two of the most potent scientific buzzwords of the period: nanoparticles and RNA, the body's messenger service for genetic information.
Never miss a local story.
"I was very jazzed about it, " Franzen said. "I was in a very fanciful and creative mood, taken up with the possibilities."
A decade later, Franzen is still consumed. Not by the project, but by his attempts to force his former colleagues to admit they built their research on a false foundation.
After years of going through all the prescribed channels in the science community and at N.C. State, he's taking his story public.
Franzen is convinced that his colleagues knew early on that their research was flawed, and he was outraged when they refused to correct their misrepresentations. Legal threats and investigations ensued. An acrimonious battle raged in the arcane journals of research chemistry. He says university lawyers and administrators were more worried about controlling damage to N.C. State's reputation than about maintaining scientific standards and ethics.
Franzen is 56, a tenured professor of chemistry who also works two months a year at Zhejiang University, one of China's most prestigious schools. An author of 175 scientific papers, his research currently focuses on proteins in marine organisms that can clean up toxins in the environment.
His quest to correct the record raises larger issues. How much false science is published? Who polices the misconduct, and how well?
The questions burn at the core of one of the region's economic engines. The Triangle is home to growing alliances among businesses, academics and venture capitalists. Three major research universities bring in hundreds of millions of dollars in grants annually, and last week President Barack Obama came to town to announce more would be coming to N.C. State. Dozens of research-based companies in and around Research Triangle Park thirst for the discoveries and will pay handsomely for them.
But scientific misconduct has been on the rise. A 2012 study from the National Academy of Sciences found a tenfold increase since 1975 in scientific articles retracted because of fraud.
Researcher misconduct was the cause of two-thirds of the retractions, which were more likely to occur in prestigious journals. The competition for grant money and the pursuit of academic prestige were among the reasons, authors of the study have said.
Bruce Eaton, 59, has had a long career in industry and academics. He is named on at least 70 patents, helped start two companies and consulted on product launches. Dan Feldheim, 46, described by colleagues as a big thinker, has a wide research background, from DNA research to the development of tools to diagnose lung diseases such as cancer, pneumonia and drug-resistant tuberculosis.
In May 2004, both were professors at N.C. State. Eaton and Feldheim - with the help of Ph.D. candidate Lina Gugliotti - published new research in Science, one of the world's most prestigious journals.
The scientists wrote that they had used RNA to create tiny crystals of palladium, a metal used in catalytic converters and thousands of other industrial products. And while palladium is valuable, the greater promise was that scientists could potentially create any number of tiny structures - a microfactory inside a test tube, with RNA as the machinery that could churn out valuable superconductors and high-strength materials.
The key ingredient was a mix containing an enormous number of RNA sequences - the number 1 followed by 14 zeros. NCSU applied for a patent for the process of selecting useful RNA sequences as well as some of the resulting compounds. Eaton and Feldheim were listed as inventors, entitled to a portion of any royalties. Venture capitalists and investors refer to such patented compounds as the "secret sauce."
Feldheim and Eaton said they mixed the RNA solution in water with a compound containing mostly carbon but also 20 percent palladium by weight.
After two hours at room temperature, they filtered and processed the solution.
In the dry, precise language of scientific journals, Eaton and Feldheim trumpeted their findings in Science. Their secret sauce had formed tiny hexagonal palladium crystals, they said. All life on Earth is carbon based, and DNA and RNA regulate the creation of carbon-based cells. Feldheim and Eaton claimed they were now able to step out of the carbon-based world and deploy RNA to bring about the formation of metals.
The paper was good news for the professors. Eaton, who had previously trademarked the phrase “Evolutionary Chemistry,” moved in 2005 to the University of Colorado and joined a biotech firm, SomaLogic, as its director of research.
After the Science paper was published, Feldheim and Eaton asked Franzen, an expert in attaching bits of RNA and DNA to various surfaces, to join them. They wanted to apply for a $1 million grant from the Keck Foundation.
Franzen signed on. Both he and Feldheim had started teaching at N.C. State in 1997 and had collaborated on seven previous scientific papers.
Franzen was curious: What was the step-by-step formation of the crystals? How did the RNA fold the palladium into a hexagon? Why wasn't it a rod or a cube, the usual shapes for palladium crystals?
The bold grant proposal owned up to being perhaps too radical for more timid foundations "that do not possess the global long-range vision of The Keck Foundation, " it said. The foundation's mandate is to support pioneering discoveries in science, engineering and medical research. One of the Keck board members was Larry Gold, a University of Colorado professor and founder of SomaLogic.
The grant proposal asked the foundation to imagine a fantasy vision of the world where any magnetic, optical, chemical or physical characteristic could be formed at will.
"Everything from new ultra high strength materials to room temperature superconductors might be possible, " the trio wrote. "These materials would in turn enable new technologies for medicine, transportation, information storage and alternative energy; technologies that will have a profound impact on the world in which we live."
Keck approved the grant. Feldheim and Eaton went on to receive more than $700,000 in related grants from the U.S. Department of Energy and the National Science Foundation.
In November 2005, Feldheim and Eaton announced more progress in the Journal of the American Chemical Society. They had used RNA to form hexagonal crystals of platinum, another useful metal.
In a note at the end of the paper, they disclosed a minor-sounding detail: Their aqueous solution, mixed with the secret RNA sauce and the carbon-metal compound, was not 100 percent water. It contained 5 percent of a solvent called THF. The importance of this would become apparent years later.
But at the time, Feldheim's laboratory continued to research RNA and the creation of what it claimed to be metal crystals. As is normal, Ph.D. students and post-doctorates performed most of the work, under the supervision and direction of Feldheim and Franzen. Eaton had left for Colorado.
The students engaged in a constant attempt to confirm or disprove the work and expand the research. This brought Lina Gugliotti and Donovan Leonard together in fall 2005.
The two graduate students met in one of the electron microscopy rooms on Centennial Campus to run tests that would check whether the formed crystals were indeed palladium.
Gugliotti was the Ph.D. candidate who had, under the supervision of Feldheim and Eaton, carried out the research published in Science, which carried her name as a contributing author. She brought samples of the palladium crystals.
Leonard, a Ph.D. candidate in materials science, would operate the electron microscopes. Leonard set up the microscope to examine the diffraction pattern. Every crystal has a unique diffraction pattern - an image of electron beams bouncing off atoms. In preparation for the test, Leonard had created models of palladium diffraction images.
They were nothing like the images he saw in the microscope.
"It was clear we weren't looking at palladium, " Leonard said. "There was a lot of head scratching."
Leonard said he had calibrated the machine correctly. Maybe it was a bad sample. The grad students ran more tests and got the same results.
Leonard ran a chemical analysis that showed the particles contained a lot of carbon. It was not pure palladium, which is an element and a metal. He took his concerns to Franzen, who suggested it might be a bad sample. Franzen said the significance of the results didn't dawn on him at the time.
Leonard re-read the Science article and tried to talk with Feldheim about the data.
On Dec. 6, 2005, to a group that included Feldheim, Leonard presented his findings: The hexagons weren't metal and contained very little palladium.
Franzen and Leonard recall that Feldheim was angry and defensive, accusing Leonard of not knowing what he was doing.
Leonard said Gugliotti pulled out diffraction images from the original research. He asked whether she had indexed the images, in effect calibrating the microscope and producing a benchmark image of a known substance. Indexing provides a unique fingerprint of a material. Unindexed images are useless random patterns.
Gugliotti admitted to Leonard, Franzen and all present that she had not indexed the images.
Leonard said that subsequent exchanges with Feldheim were unpleasant. When Leonard asked to meet, the response was, " 'We're busy.' 'Don't worry.' 'Sorry we brought it up to you.' 'We're going to do it elsewhere.' "
Or it was combative.
"He was yelling at me, 'Leave this palladium stuff alone!' " Leonard said. "As a student, I was confused."
Lots of solvent
After receiving her Ph.D., Marta Cerruti joined N.C. State for a two-year post-graduate stint. In spring 2006, Cerruti started a series of experiments funded by the Keck grant. She uncovered a string of problems with the RNA research.
In the course of her research, Cerutti routinely ran basic control tests, including one without RNA, the building block of Feldheim and Eaton's evolutionary chemistry.
Even without the RNA, the experiment produced the hexagonal particles.
She ran it again. Same result, leading her to ask questions about the core of the research: Did RNA have any role at all? Did the notion of evolutionary chemistry have any value?
"It was really scary, " Cerruti recalled. "There was no need for RNA to make the particles."
Cerruti discovered another problem: Other researchers in Feldheim's lab were running related RNA experiments using 70 to 100 percent THF solvent, she said, a far cry from the "aqueous solution" described in the Science paper. The solvent, not the RNA, was the key to creating the particles.
Cerruti regularly shared her results with her mentors. She couldn't have received more different reactions.
Franzen was shocked. He had wanted to believe in the power of RNA. But the data didn't add up. The hexagons were not sharply defined crystals. The hexagons were apparently 90 percent carbon, not 100 percent palladium.
In a June 2006 letter to Feldheim and Eaton, Franzen proposed new experiments. He pushed his arguments with Feldheim at a meeting in Raleigh and at a scientific conference in Poland. Franzen said Feldheim privately admitted there were serious issues, that the results could be a product of experimental error or a faulty test arrangement. They would run more experiments to find out what was happening, Feldheim said.
Cerruti got a very different response from Feldheim. In September 2006, she performed several of the tests in front of Feldheim, using his lab and his equipment.
"I said, 'What do we do? We should say something; there is something wrong in what has been published, ' " said Cerruti, who had a few months left in her two-year contract.
"He said, 'You don't have that much time. Go finish something else.' "
Cerruti said it was awkward to be caught between two strong-willed professors.
"I felt terrible, " she said. "It was a very, very bad time in my career."
But in the lab and in the chemistry department, things were about to get worse.
Cast of characters
Stefan Franzen, 56
Tenured professor of chemistry at N.C. State University who has spent eight years trying to correct the scientific record of former colleagues Eaton and Feldheim.
Bruce Eaton, 59
Former NCSU professor, now at the University of Colorado, a founder of two biotech companies and named on 70 patents.
Dan Feldheim, 46
Former NCSU professor and collaborator with Franzen on seven journal articles before the two fell out; currently professor at the University of Colorado.
Lina Gugliotti, 35
Graduate student with Eaton and Feldheim who received her Ph.D. in 2006. She performed the experiments for the questioned 2004 Science article. Now lives in Connecticut.
Donovan Leonard, 38
Ph.D. candidate in material sciences in 2005. Discovered through an electron microscope that the purported palladium crystals were not palladium. Now works at Oak Ridge National Laboratory.
Marta Cerruti, 35
Postdoctorate fellow who discovered in 2006 that RNA was not a factor in the formation of the purported palladium crystals. Now a professor at McGill University in Montreal.
Understanding the science
DNA: molecules that encode genetic instruction for all living organisms
RNA: molecules that act as DNA's messenger service within cells
Palladium: a rare and valuable element and catalyst with many industrial uses
Nanoparticles: particles smaller than the tiniest bacteria. Nanoparticle research has been an area of intense interest with potential uses in the fields of medicine, electronics and optics.
THF: a common industrial solvent, colorless with a smell similar to nail polish remover
Diffraction: a process of shooting beams of electrons through crystals to produce a unique fingerprint of a substance
Indexing: calibration of an electronic microscope that involves taking an image of a known substance, usually aluminum, to ensure accurate diffraction images
May 2004: Science article published on RNA forming hexagonal palladium crystals.
January 2005: Research begins with help of $1 million Keck Foundation grant.
December 2005: Donovan Leonard presents evidence that hexagonal particles are not palladium.
Spring 2006: Marta Cerruti uncovers more flaws in Science paper.
About the reporter
Joseph Neff, 53, is a veteran investigative reporter who has written extensively about criminal justice and health care. He and Mandy Locke revealed broad misconduct at the State Bureau of Investigation in 2010, and he exposed the prosecutorial misconduct of former Durham District Attorney Mike Nifong in the Duke lacrosse case.
In 2012, Neff teamed with David Raynor and reporters at The Charlotte Observer to produce "Prognosis: Profits, " a five-part series that exposed huge profits at some nonprofit hospitals and large salaries paid to top executives. That series was honored with several national awards, including grand prize in the Robert F. Kennedy Journalism Awards and the American Society of News Editors' award for Local Accountability Reporting.
About the reporting
This series is based on interviews with Stefan Franzen, Donovan Leonard, Marta Cerruti and other experts in the field. Reporting also included reviews of hundreds of pages of documents: journal articles, grant proposals, emails, legal correspondence and an internal investigation at N.C. State.
We requested interviews with professors Bruce Eaton and Dan Feldheim through voicemail, email, letters and spokesmen for the University of Colorado and SomaLogic, the Colorado company that licensed Eaton's patents. They did not respond.
Lina Gugliotti, the Ph.D. candidate who performed much of the research, did not respond to phone messages.