Darleane Hoffman sat in her office on the hilly rise above the Berkeley campus, admiring the view across San Francisco Bay. In the distance she could look out and see the old federal prison on Alcatraz, the city’s main tourist attraction, and past that, the rolling fog slipping like white sheets under the Golden Gate Bridge. It was Monday, 19 April 1999. It would have been Glenn Seaborg’s eighty-seventh birthday.
Almost 50 years earlier, she had missed the discoveries of einsteinium and fermium while she sat outside Los Alamos, waiting for HR to realise that women could be scientists too. Now, her team leaders wanted to tell her something important. She had feared the worst but had been reassured on the phone that it was good news.
In her early seventies, Hoffman gave the impression of being someone’s sweet grandmother. Foolishly, some graduates tried to get into her group assuming she’d be a pushover. It was an opinion that was quickly dispelled – particularly if you dared to call element 105, officially dubnium, anything other than ‘hahnium’. Tough and supportive in equal measure, universally respected across the chemical world, Hoffman had dedicated her life to making sure nuclear chemistry wouldn’t die, even if she had to resuscitate it one student at a time.
Berkeley was still playing catch-up with the other labs, but finally seemed to have turned a corner. Super-HILAC had been turned off in 1993, and the experiments had moved to the 88-inch cyclotron; SASSY had also been replaced by a new gas-filled separator (this time with a more fitting emergency valve). With it, the team hoped to hit even lower detection limits.
Matti Nurmia had returned to Finland. In his place came three additions. The first was Darleane Hoffman’s former postdoc Ken Gregorich. Tall and wiry, with a neatly trimmed goatee and a crop of hair skirting a bald crown, Gregorich was the epitome of the ‘work hard, play hard’ attitude of nearby Silicon Valley. In the lab he was a relentless and meticulous researcher; at home he ran ultra-marathons for fun. He had worked with Hoffman since the mid-1980s and was at the vanguard of a new generation not poisoned by the Cold War. Sharp, measured and precise in his approach, to Gregorich the transfermium storms were water under the bridge. He just wanted to do good science.
Next, on a Fulbright scholarship from the Soltan Institute for Nuclear Studies in Warsaw, Poland, was Robert Smolańczuk (Berkeley didn’t have the budget for another permanent member of the team, so the secondment was the best solution). A theoretical physicist, while Smolańczuk had been at GSI he had published some eyebrow-raising calculations. According to his theory, the cross sections for the elements beyond 114 wouldn’t vanish into the realms of statistical improbability, but rather be large enough to detect: a massive 670 picobarns. In the 1940s such a low limit would have felt impossible; by the turn of the millennium, the element hunter’s toolbox had improved so dramatically it felt like it was easy pickings – the kind of cross section that could produce hundreds of atoms a week. If Berkeley was willing to give Smolańczuk a shot, he was confident they could leap beyond anything Dubna were trying to do and find element 118. It was controversial, flying against all known wisdom. But then, wasn’t that what element hunting was all about?
The final new arrival was considered a coup: Berkeley had tempted element superstar Victor Ninov to leave GSI and join their team. Ghiorso, technically retired but still doggedly riding his recumbent bike into work every morning, saw the newcomer as the future of element discovery. ‘Victor Ninov,’ Ghiorso would tell anyone who would listen, ‘reminds me of a young … well, a young Al Ghiorso!’ On the back of her colleague’s glowing recommendation – and letters praising his talent from his colleagues at GSI – Hoffman had placed complete faith in him; while Gregorich led the team in running the machines, Ninov had brought over his unique computer program from GSI to analyse the results. He was the only one who knew how it worked, but Berkeley didn’t need anyone else: he was the best in the world at what he did.
Hoffman’s hand had been twisted toward Smolańczuk’s madcap idea. When Dubna claimed element 114 had been found, the Berkeley team had been a mere eight months behind them, ready to do the same experiment. However, they hadn’t been able to get hold of the large quantities of plutonium-244 or calcium-48 needed (or the permission required to use plutonium in the hills above one of the most populous metro areas in the US). Options narrowed; all that was left was ‘Robert’s reaction’: firing krypton into lead.
Hoffman and Ghiorso had both urged doing it as soon as possible – if Smolańczuk was right, there was nothing to stop GSI or Dubna doing it first. ‘[It was] a strange reaction that no one thought would go,’ Ghiorso would later recall for the New York Times, ‘but because it was relatively easy, we thought, “What the heck, we have nothing to lose.”’ Gregorich had agreed, arguing that the efficiency of the 88-inch cyclotron was so high that even if they didn’t find 118, it would give him a chance to improve its systems. Finally, Ninov relented and threw himself into the analysis with his usual verve.
The experiment had started on 8 April 1999 and ran for four days. At first, nothing happened. The team departed for the Easter break, leaving Ninov to check their results. Now, almost two weeks later, Hoffman watched as a trio of researchers – Gregorich, Ninov and Walter Loveland, on sabbatical from Oregon State University and there to soak up how Berkeley conducted their experiments – entered her office. They brought with them a sheet of data.
While the experiment ran, Ninov’s analysis tool, Goosy, had found three distinct alpha decay chains resulting from fusion. Two of them matched Smolańczuk’s predictions perfectly. The numbers coming from Ninov’s analysis were too good to put down to random chance.
Ninov laughed. ‘Does Robert talk to God, or what?’
Berkeley had discovered element 118.
The reaction from the team had been a mix of excitement and disbelief, even before their results had made their way to Hoffman’s desk. Loveland’s first response had been ‘What the hell is going on?’; Gregorich was surprised too; Ninov had been so taken aback he had urged his collaborators to keep the results quiet and not tell Hoffman (Loveland and Gregorich overruled him). For her part, Hoffman felt a pinch of excitement, but kept her cool. Science, she knew, thrives on verification: the experiment had to be repeated or the results were meaningless. She was too seasoned to get her hopes up on what could be a phantom.
‘OK,’ she said. ‘Let’s do it again.’
The Berkeley team started a second experiment running ‘Robert’s reaction’. By the first week of May, they had another perfect chain that matched Smolańczuk’s calculations (the previous chain that didn’t follow the pattern was discarded). Dubna’s single atom of 114 wasn’t enough to convince IUPAC of a discovery; Berkeley’s three atoms were solid, unchallengeable proof. Hoffman and Ghiorso began to dream of snatching another element out from under Oganessian’s nose. Already, thoughts turned to its name. Berkeley had seaborgium; why not ‘ghiorsium’ too?
Wary of the false reports of discoveries throughout the Cold War, Berkeley Lab decided to proceed with caution, conducting an internal review to rule out any embarrassing mistakes. The staff double-checked everything: the ion source, the accelerator and the detectors. Nothing was wrong. Finally convinced, in June 1999 Hoffman and Ghiorso called a press conference and published their claim in full. Anyone who had been near to the experiment was added to the paper, with Ninov as first author. ‘Needless to say,’ Hoffman and Ghiorso wrote in The Transuranium People, published that year, ‘this news is an enormous surprise to the scientific world. Now there is no question, the Superheavy Island [of Stability] actually exists! […] We have convincing evidence of 114, 116 and 118! This opens up a whole new region for study.’
Finally, on Glenn Seaborg’s birthday, Darleane Hoffman had her element. It seemed too good to be true.
It was.
Hoffman, Ghiorso and their team had just fallen victim to the most audacious fraud in science history.
* * *
Every scientist makes mistakes. Science functions by ‘failing forward’, constantly tinkering ideas, experiments and approaches to get things a little closer to right each time. Research fraud – faking your results, lying to yourself and the world – is the opposite of everything science stands for. When you’re caught (and you always are), it destroys your reputation, your colleagues’ reputation and your lab’s reputation. In physics, the best-known scandal is probably the work of Jan Hendrik Schön, who seemed to have made miracle breakthroughs in semiconductors that were nothing of the sort. When news of his scientific misconduct broke, it saw 28 papers in leading journals retracted.
The 118 scandal, which occurred at virtually the same time, was more devastating for its community. The first cracks in the Berkeley discovery had emerged within months of the paper’s publication. Back in Germany the GSI team were struggling to repeat the Berkeley results. The institute’s superheavy programme had not had a success since 1996 and the new director, Hans Specht, had quarrelled with Sigurd Hofmann. ‘He gave us a hard time,’ Hofmann recalls. ‘We immediately got the beam time to repeat the experiment for 118; krypton is an easy beam [to produce], and lead targets are easy too. After one week, when he hadn’t seen anything, our director shouted at us: “You’re not able to make such experiments, you’re too stupid!” But after another two weeks, we still didn’t see anything either.’ (Specht told me he does not remember this, although it was clear when we corresponded there was no love lost between him and Hofmann.)
Teams in France and Japan then tried the experiment. Neither found any of the miraculous decay chains the Americans had reported. Stranger still, when Ninov attended conferences, he was reluctant to speak about the amazing feat Berkeley had accomplished. Questions from the audience were deflected, sidetracked or, as had happened at GSI, casually forgotten about. The man of the hour seemed to want to avoid all mention of his greatest achievement.
In the spring of 2000 the Berkeley team decided to silence their critics and run the experiment again. What had once been so easy to produce became invisible: the 118 decay chains weren’t there. For the next year, the laboratory went over the data, even calling in an independent group to analyse the experiments and make recommendations.
In April 2001 the team gave their amazing krypton-into-lead reaction another shot. Two-thirds of the way through their beam time, Ninov gave everyone the news they had been waiting for: his analysis showed clear evidence of a decay chain for element 118.
The news should have brought relief; instead, it brought the opposite. ‘In the time that passed,’ Loveland recalls, ‘there were other people who had become expert in Goosy, including my postdoc, Don Peterson.’ Keen to make sure everything was correct, Peterson decided to go back into the raw data, before it had passed through Ninov for analysis, to check the chain was accurate. ‘Don looked at this stuff and said, “I can’t find the event!” At that point, I thought “Oh God, what’s happening?”’
Loveland checked over his colleague’s work, pulling up the original data from the machine itself. Peterson was right: the alpha decay chain Ninov claimed to have found wasn’t there. ‘Depending on whose software you used, if Ninov used it or Don Peterson used it, you got different answers,’ Loveland explains. ‘And that is just not right. At that point I started yelling to everyone that something was terribly, terribly wrong.’
In June 2001 the team went back to review the original 1999 data tapes – the raw, unfiltered information spewed out before it had been processed by Ninov. None of the chains he had taken up to Hoffman’s office existed. A third Berkeley committee agreed: there was no sign of element 118. There never had been.
For decades, Berkeley had openly mocked and questioned the Russian data. Now their own research was flat-out wrong. What had been the highlight of Hoffman’s career and the crowning glory of Ghiorso’s had become the darkest moment of their lives. Meeting with the group, everyone except Ninov agreed to retract the claim. While Ninov conceded the records didn’t show his element, he remained adamant about what he had seen during his analysis.
Nobody was listening to him. Goosy had a habit of spawning quirky, corrupted data, but a review showed it hadn’t made a mistake; and, even if it had, the probability of spawning three chains that fit the predictions exactly wasn’t worth contemplating. The different committees also ruled out the possibility that someone had wiped the event data from the original tapes. That left only one possibility, the final committee decided:
There is clear evidence that at least one of the 118 element decay chains published in 1999, and also the candidate in the 2001 data, were fabricated. This fabrication was performed by capturing the output of the data analysis program in a text editor and then systematically altering some events and inventing others in order to present data that would appear to be an element 118 decay chain.
Someone had, in 1999 and 2001, copy-pasted evidence of element 118 into the raw data. Someone analysing data only one man saw, and who knew how to use a computer program only one man could interpret.
‘In short,’ the committee found, ‘it is very difficult to reconcile all these circumstances on any basis other than with Ninov being the fabricator of the claimed 118 decay chains.’ In 2002 Berkeley Lab found Victor Ninov guilty of scientific misconduct and fired him.
The recriminations were swift. The rest of the team received harsh criticism that only one person had been left checking the results, creating a weak link where someone could interfere with the scientific method. ‘I had hired a world-recognised expert and we were trusting him to do a job,’ Gregorich told the New York Times. There were no safeguards because what happened was so unthinkable.
Ghiorso’s appraisal was blunt. ‘It’s good that Seaborg died before this,’ he was quoted as saying in the same article. ‘He would have been one of the co-authors. This would have just about killed him.’
* * *
At GSI, elements 110, 111 and 112 had all been confirmed after Ninov had departed, removing any possibility that the German claims were also faked. Even so, Hofmann wanted to make sure everything was correct. He asked a colleague to go back through the old files from the ‘miracle years’ of 1994 to 1996. Had anyone faked the data in them too?
‘It took about three or four hours,’ Hofmann told me. ‘We looked, and what existed was an alpha decay of polonium into lead.’ This was chicken feed for the element team: just the random radioactive noise you’d expect in a particle accelerator. ‘Then we looked at the subsequent printouts. On the old computers, they had version numbers, and we found them with the help of our computer centre. There were different versions in Ninov’s old computer … he had changed the information from the background event into 112. He had added, step by step, additional numbers to make it look like an alpha decay chain from 112. It wasn’t completely accurate, and I could see something was wrong.’
Hofmann remembered the events of 1996, when Ninov had burst into his office before lunch with the ‘discovery’ of element 112, only to hesitate for hours before giving him a simple printout. The strange delay suddenly made sense: had Ninov rushed off to fabricate the data?
‘I told the new director [Walter Henning, who had replaced Specht],’ Hoffmann explains. ‘He gave me some good advice: look through all the data Ninov was involved with. And we found a second chain, from the discovery of 110, that had been manipulated too.’ Of 34 chains, the two fakes stood out like sore thumbs. GSI quickly published a retraction of the altered chains, euphemistically referring to ‘inconsistencies’ in the data. Everyone knew exactly what they meant.
Fortunately for Hofmann and the Germans, they had always treated Ninov’s spurious chains with caution, and the rest of their work was beyond reproach: the two bad apples were removed before the whole barrel was tainted. Within a decade, the GSI elements were all confirmed as real.
The GSI elements had names too. On 10 December 1997 Hofmann declared a ‘names-finding day’ at GSI, gathering up his team, as well as Matti Leino and guests from Dubna and Bratislava. Hofmann went through the names sent in by colleagues as suggestions, with Andrey Popeko writing them down on the blackboard. The team then went around the table, finishing with a list of 30 possibilities. ‘The arguments raged fast and furiously,’ Hofmann wrote. ‘Painfully, one at a time, names were ringed with red chalk until, in the end, we had found the three we needed.’
Element 110 became ‘darmstadtium’ (one suggestion that was rejected, which came from an American school class, was ‘policium’ – in Germany, 110 is the emergency dial code for the police). For 111 and 112, GSI chose to honour famous scientists throughout the ages. Element 111 would be ‘roentgenium’, after Wilhelm Conrad Röntgen, the discoverer of the X-ray; and element 112 would be ‘copernicium’, after Nicolaus Copernicus, the Renaissance scientist who showed that the Earth orbited around the Sun.
They would be – to date – the last elements discovered at GSI. In 1999 the Germans had dreamed of creating a ‘superheavy element factory’, possibly exploring as far as element 126. Hofmann asked for permission to start using calcium-48 and actinide targets too, confident that GSI could easily overhaul the Dubna–Livermore team. His request was turned down. ‘We could have started experiments in hot fusion,’ Hofmann told me. ‘But the proposition was rejected, and not with friendly words. It was a dead thing.’1
* * *
Victor Ninov has always insisted on his innocence. His response to the Berkeley investigation in February 2002 was frank: ‘At no time did I knowingly engage in any form of misinterpretation of data or scientific misconduct […] I have never intentionally altered, invented, fabricated, corrupted, deleted or concealed data […] I stand by the integrity of my research.’
Having criss-crossed the world, I have never met a heavy element researcher who believes him. Most just want closure. Most, simply, want to know why.
It’s a hard question to answer. The glory of discovering an element doesn’t make a very good motive; the chance of a made-up alpha decay chain perfectly fitting the real results (which would have emerged eventually) and fooling a seasoned element hunter was virtually zero; Sigurd Hofmann’s immediate, at-a-glance dismissal of the fictional chain for element 112 was proof of that.
Nor was there pressure to succeed. No one, at any point, suggested Ninov’s career was in the balance if another element wasn’t found. He had already discovered three elements. He had nothing to prove.
Al Ghiorso, speculating, suggested that Ninov’s intent had been to buy time for Berkeley: by inserting the false chain, it won the team more beam time to find the actual element 118. But again, the argument doesn’t hold water. Robert Smolańczuk’s ideas were considered a Hail Mary – nobody had really believed they would succeed. Why buy time for an experiment that none of the team thought would work?
‘None of us fully understand what Victor did or why he did it,’ Loveland says. ‘He was extremely well thought of, a very talented man. If you ask me why, I have no real idea. Perhaps he became overconfident in his ability to predict what was going on. I was never able to come to an understanding with him. I spoke to him on a daily basis. At one time he alluded to other people interfering instead of him, but that was nonsense.’
The best guess comes from Sigurd Hofmann. ‘The astonishing thing is that [the first time data was fabricated, for element 110] it happened on 11 November. The chain Victor produced had a half-life of 11.19 minutes. If you know something about German carnivals, you’ll know they start on 11 November at 11.11 a.m. I think he meant it as a joke. But, with such a joke, he realised he could manipulate things and nobody would realise it.’ Perhaps it isn’t such a surprise element 118 was ‘discovered’ on Glenn Seaborg’s birthday after all.
For Loveland, the Ninov episode emphasises a positive: ‘Science works. You get an anomalous result, and if it’s right it gets stronger and stronger [as the experiment is repeated]. Sometimes it doesn’t. There have been other cases where outstanding events were announced and pulled back because they couldn’t be reproduced. In this case, it has the additional element that there appears to be fraud involved.’ In the long term, science self-corrects, finds answers and pushes forward. Ninov was caught and the discovery was retracted. Time to move on.
Even so, the Ninov scandal sent shockwaves through the superheavy community. Heinz Gäggeler, the man he had once called a friend and who had put Ninov up in his house while climbing the High Alps, feels betrayed. ‘Victor was so well received when he came to Berkeley,’ Gäggeler told me. ‘He had full support. And because of that, one didn’t look too carefully into the analysis he was doing. It was a total disaster. Did it destroy Berkeley? Of course it did. Berkeley was Berkeley. The outside world doesn’t want fake news. The show was over.’
The scandal brought an ignominious close to Al Ghiorso’s remarkable element-hunting career. For Ken Gregorich, the fallout was even more devastating. Today, 20 years down the line, it remains a sore subject on what has otherwise been a glittering career. When I asked him about it, he politely declined to go over it again. ‘It was a dark period, and it’s gone, and I’d rather leave it at that.’ I can’t blame him.
Yet greatest in the team’s sympathies was Darleane Hoffman. She had been denied part of the discovery of einsteinium and fermium because of sexism; now, so close, her dream of being an element discoverer was over. ‘When we thought we had [element 118],’ one former Berkeley researcher told me, ‘we had it for her. It was Darleane’s element. And that crushed us more than the retraction – that she didn’t have it.’
Notes
1 I’ve seen the internal report, and Hofmann isn’t kidding. However, it would be misleading and disrespectful to everyone involved to say this was the only reason GSI fell behind in the element race. The truth is more complicated.