Purdue University has opened an investigation into "extremely serious" concerns regarding the research of a professor who said he had produced nuclear fusion in a tabletop experiment, the university announced yesterday.
Fusion is the process the sun uses to produce heat and light, and scientists led by Rusi P. Taleyarkhan, a professor of nuclear engineering at Purdue, said they were able to achieve the same feat by blasting a container of liquid solvent with strong ultrasonic vibrations.
The vibrations, they said, collapsed tiny gas bubbles in the liquid, heating them to millions of degrees, hot enough to initiate fusion. If true, the phenomenon, often called sonofusion or bubble fusion, could have far-reaching applications, including the generation of energy.
The research first appeared in 2002 in the journal Science, but controversy had erupted even before publication. Dr. Taleyarkhan, then a senior scientist at Oak Ridge National Laboratory in Tennessee, reported the detection of neutrons, which are the telltale signs of fusion, but two other scientists at Oak Ridge, using their own detectors, said they saw no signs of neutrons.
Dr. Taleyarkhan, who joined the Purdue faculty in 2003, and his colleagues have published two additional papers in major physics journals, amid the continuing skepticism of other scientists. No other scientists have been able to reproduce the findings.
The university began a review of the research and the accusations last week, Sally Mason, the university provost, said in a statement. "The research claims involved are very significant," Dr. Mason said, "and the concerns expressed are extremely serious."
Dr. Mason said that the review was being conducted by Purdue's Office of the Vice President of Research and that the results would be announced publicly.
Dr. Taleyarkhan did not return phone calls or respond to an e-mail message seeking comment.
Meanwhile, Brian Naranjo, a graduate student at the University of California, Los Angeles, said his analysis of data from the last scientific paper that was published by Dr. Taleyarkhan's group showed a chance of less than one in 10 million that the emission pattern could have been generated by fusion.
Instead, Mr. Naranjo said that the pattern of particles seen in the experiment much more closely matched that given off by californium, a radioactive element that is used in Dr. Taleyarkhan's laboratory. With $350,000 from the Defense Department, Seth J. Putterman, a professor of physics at U.C.L.A. and the thesis adviser to Mr. Naranjo, has tried to build a replica of Dr. Taleyarkhan's apparatus and has not seen any signs of fusion.
Dr. Putterman said he told Dr. Taleyarkhan of the calculations last week on a visit to Purdue. "He didn't have any clear answers," Dr. Putterman said. "From my perspective, his answers were not satisfactory."
Californium is present in Dr. Taleyarkhan's laboratory, stored in a closet about 15 feet from the experiment — close enough to generate the results reported in Dr. Taleyarkhan's paper if it had been stored improperly.