For whom the Bell tolls?

July 12, 2002

Allegations of research fraud have scandalised physics. Steve Farrar reports.

The machine sitting in Jan Hendrik Schön's old laboratory at the University of Konstanz in Germany did not appear out of the ordinary. Plenty of laboratories boasted far more sophisticated equipment.

But since 2000, scientists have been awed by the revolutionary electronic devices that Dr Schön produced with the help of that bog-standard sputtering machine.

At 31, Dr Schön became the rising star at the world-famous Bell Labs in the US. His work was at the cutting edge of nanotechnology and some considered a Nobel prize to be a formality.

Now the sky has fallen in on the young physicist. Leading scientists have raised serious doubts about his work, and since May a team of inquiry has been picking through his data for signs of artifice.

Dr Schön told The THES : "I will try to help them as much as possible."

The allegations of research fraud have shaken the physics community. If proved, they threaten not only Dr Schön's career but also the good name of Bell Labs. Their stain may even stretch to the respected academic journals that carried Dr Schön's research.

The quiet, hard-working German postdoc left Konstanz for Bell Labs in 1998. He joined respected physicist Bertram Batlogg in research into electronic devices made from organic materials and was put to work a floor below where the world's first transistor was tested in 1947.

In 1999, Dr Schön enjoyed his first success with the fabrication of a field effect transistor made from a single organic crystal. Its performance was superior to anything similar created before and it gave him his debut paper in Science, in February 2000.

More advances followed. An organic transistor just a single molecule in size, the first plastic superconductor and an organic laser - the marvels pioneered by this previously unheard of scientist came thick and fast.

Dr Schön's research appeared in a flurry of academic papers - 74 in two and a half years, including 15 in the most prestigious journals, Nature and Science .

It was an astonishing body of work, albeit produced with more than 20 co-authors. The results stoked hopes of a revolution in electronics and fuelled interest in the electrical properties of organic materials.

At Oxford University, John Singleton tried in vain to get hold of one of the Bell Labs' devices to test. At Sheffield University, Richard Jones's department of physics debated whether to attempt to replicate the results. At Nottingham University, Peter Beton was inspired to take an alternative approach to the work.

Elsewhere, particularly in Japan and the US, many scientists have tried to reproduce Dr Schön's research but so far have always failed to produce the same effects.

Laurence Eaves, professor of physics at Nottingham, remarked: "The equipment required to make and measure the transistors is not very expensive and is available in many labs around the world, yet they seemed to have a magic touch that put their research years ahead of the competition."

Others were baffled at the mechanisms responsible for the observed effects. Dr Schön could not give a concrete explanation but noted that whatever it was, it clearly worked.

An examination of Dr Schön's work in last week's issue of Science focused on one aspect in the fabrication of many of the devices that particularly baffled scientists. This involved the laying down of an insulating layer of aluminium oxide, a process that brought success only when Dr Schön carried out the work on the old sputtering machine back in Germany.

One scientist closely associated with the research said doubts were raised within Bell Labs many months ago as to how the results were being achieved, but there was little more than circumstantial evidence of a problem.

Then it was noticed that two graph plots from different experiments in different published papers appeared identical, right down to the supposedly random pattern of noise.

Staff members contacted colleagues outside the laboratory who determined that five of Dr Schön's papers involved graphs that seemed partially duplicated. They reported the matter to the journals in question - Nature, Science and Applied Physics Letters - and to the laboratory authorities.

Bell Labs promptly announced that an inquiry would be undertaken by a committee of independent researchers, headed by Malcolm Beasley, professor of physics at Stanford University.

The committee is not expected to report its findings until late summer, but it has already widened its remit to include more than a dozen papers.

Saswato Das, a spokesman for Bell Labs, said the company had taken action immediately after questions as to the scientific validity of the data had first been raised two months ago.

"There had been scientific concerns right from the start but that's not uncommon in science when experiments are done and theory doesn't explain the results," he said.

The physics community has a track record of probity and honesty. The fact that no one has been able to reproduce the results has coloured opinions, although such concerns had long gone unreported.

In December last year, Paul Solomon, a research staff member at IBM's laboratory at Yorktown Heights, together with some colleagues, wrote to Nature questioning the scientific rationale behind one of Dr Schön's papers: "It seemed to go against some of the basic things I've assumed to be correct and I thought some explanation was in order."

He was told that the journal did not disagree with the points he made and that a reviewer who checked the paper for Nature had raised similar concerns, resulting in Dr Schön's interpretations being toned down.

Nevertheless, the original article had appeared on the strength of the data as presented. Dr Solomon was informed his letter would not be published.

Dr Solomon was disappointed. "I'm not really objecting to Nature publishing [Dr Schön's article]," he said. "But if scientists express concern I think such concerns should be published. If they had been and it is found that the data were fabricated at least they would have maintained their integrity."

Karl Ziemelis, Nature 's physical sciences editor, said Dr Solomon's contribution was considered by a referee of the original paper who felt it did not add substantially to the discussion over the data's interpretation.

Others felt both Nature and Science had been caught out.

Philip Anderson, Nobel laureate, emeritus professor of physics at Princeton University and consulting director of Bell Labs' physical research laboratory until 1984, said that papers containing exciting results were not subjected to the same level of scepticism as more mundane papers.

"If, in the end, it turns out there has been fraud, part of the blame goes with the management of the labs but also part with the journals," he said. "Everyone I've talked to implicated to a great extent the journals and, in particular, Science and Nature . They were eager to compete with each other to get these exciting papers."

Robert Laughlin, Nobel laureate, professor of physics and applied physics at Stanford and a former researcher at Bell Labs, was equally scathing of both management of the laboratory and the journals.

He said he did not understand why the mysterious mechanism behind Dr Schön's results had not been probed further by his colleagues.

"The Bell Labs of old would have descended on this great science problem like a swarm of bees," he said.

Furthermore, Professor Laughlin said that in his opinion Science and Nature were "afraid of being scooped by the other."

"The real culprit is the practice of selling journals on the basis of what is sexy," he said. "They have no one to blame but themselves."

One researcher closely associated with Dr Schön's work said he believed that the two journals publishing the research made it harder for Bell Labs'

management to initiate action.

Dr Ziemelis insisted that all Dr Schön's papers had been subjected to rigorous peer review no different from any other paper. At no point did any referee tell the journal there was a problem that would prevent publication or question the integrity of the data.

"We're an obvious scapegoat," he said. "These were such significant, high-profile and well-received results and emotions are running high within the community."

In Dr Schön's case there was not only the importance of the research to be weighed against any misgivings. There was also the good reputation of his co-authors and that of the prestigious institution he worked at - Bell Labs boasts that six of its physicists have received Nobel prizes.

Donald Kennedy, editor of Science, wrote last week that peer review was no protection against clever fraud. "There is little journals can do about detecting research misconduct," he said.

Attempting to reproduce research was the scientific community's best weapon against fraud, according to Richard Friend, Cavendish professor of physics at Cambridge University and one of the UK's leading researchers into organics.

"Almost by definition, major breakthroughs in science are not instantly explicable," he said. "Two years seems to be an entirely reasonable length of time to suspend disbelief."

When the Beasley committee reports its findings that grace period will be up.

If it concludes that there was fraud, the price could be high for Bell Labs. Its parent company, Lucent Technologies, has seen its share price collapse since 2000 and suffered a further drop in the wake of the WorldCom corporate fraud.

Nature and Science will also have to consider whether any useful lessons can be learnt.

Dr Schön, meanwhile, is still at Bell Labs awaiting the outcome of the investigation.

Back in Germany, even the sputtering machine is in limbo. Dr Schön told a conference in April that for some months he had been unable to get it to produce the kinds of remarkable device it had once supplied.

You've reached your article limit.

Register to continue

Registration is free and only takes a moment. Once registered you can read a total of 3 articles each month, plus:

  • Sign up for the editor's highlights
  • Receive World University Rankings news first
  • Get job alerts, shortlist jobs and save job searches
  • Participate in reader discussions and post comments
Register

Have your say

Log in or register to post comments