When Science Goes Wrong (40 page)

That was the basic idea, but up until then it hadn’t worked. Either the energy of the incoming nucleus was too low to overcome the reticence posed by the target nucleus’s Coulomb barrier, or it was too high and the target nucleus simply disintegrated, like a toad struck by a flying princess. There didn’t seem to be any ‘just-right’ level of energy – the gentle kiss that would allow the long-hidden Prince Charming to step forth.

Things seemed to change in 1998. A young Polish theoretical physicist by the name of Robert Smolanczuk, who was then at the Berkeley Lab’s German rival, GSI, did new calculations of the expected probabilities for the creation of superheavy elements by various fusion reactions. He reported that if atoms of
280
Pb (the commonest natural isotope of lead) were bombarded with a beam of
86
Kr nuclei (an isotope of the inert gas krypton) that had been accelerated to exactly the right energy, there was a surprisingly good probability of scoring hits that would generate nuclei of element 118 – a superheavy element far beyond what had been created up until then. ‘What Smolanczuk picked up on,’ Gregorich told me, ‘was that you need to bombard at an energy level that’s just over the Coulomb barrier: this happens to be the correct energy needed to make the product.’

Nuclear chemists calculate the probability of such successful hits in units called barns, whose name derives not from some famous scientist named Barn or Barnovsky, but from the phrase ‘can’t hit the broad side of a barn’. Most theorists thought that the probability for such reactions fell off exponentially with increasing atomic number, so that by the time one got to element 118, it might be down in the femtobarn range (one quadrillionth of a barn) or less, making the reaction essentially unachievable even in a lifetime of trying. Smolanczuk, on the other hand, pegged the probability as being many orders of magnitude higher, at 670 picobarns – nearly one billionth of a barn. This was still a challenging proposition but within the realm of possibility.

According to Smolanczuk’s calculations, the compound nucleus formed by the fusion reaction would evaporate just one neutron. The reaction would be a ‘cold fusion’, contrasted with the ‘hot fusion’ reactions in which the compound nucleus was more highly excited and evaporated several neutrons, like the
48
Ca/
244
Pu example described earlier. Thus the reaction would leave an atom of
293
118 – a nucleus with 118 protons and 175 neutrons. On the basis of shell theory, this atom seemed to lie on the western shore of a magic island: it could probably exist for a very brief period – a fraction of a millisecond – but it would have too few neutrons to last for longer.

Darleane Hoffman, still a leader of the Berkeley group in spite of having been officially retired for seven years, invited Smolanczuk to join the lab, which he did in October of 1998. His ideas got a very enthusiastic response. Hoffman, along with Albert Ghiorso, urged the junior members of the team to put Smolanczuk’s reaction to the test, and quickly. Smolanczuk had told the German and Russian groups about his ideas, and so the quest for element 118 had suddenly become an international horse race – in Hoffman and Ghiorso’s minds, at least.

Ken Gregorich, the designer of the gas-filled detector, and Walter Loveland, the visitor from Oregon State, took the lead in setting up the experimental apparatus. For the data analysis, they turned to Victor Ninov.

Ninov was born in Bulgaria, but had obtained his doctorate at GSI before coming to Berkeley. While at the German lab he took part in the research that led to the discovery of elements 107 to 112. Darleane Hoffman and Ken Gregorich hired Ninov away from GSI in 1996. Because of his achievements at the German lab, the move was thought to be quite a coup for the Berkeley group – an acquisition that would greatly increase their chances of finding a superheavy element of their own.

At Berkeley, Ninov worked closely with Gregorich on the construction of the gas-filled separator and its associated instruments. His most valuable sphere of expertise, however, was in developing and using the software that analysed the output of the detectors. This software had to search the instruments’ output files, which were binary files recorded on magnetic tape. The software looked for events whose time, location and magnitude corresponded to what would be expected for the breakdown of the nuclei that were being sought.

In the case of element 118, the original
293
118 nucleus was expected to decay by giving off a sequence of alpha particles, each of which consisted of two protons and two neutrons. These alpha emissions were what the detector actually detected. The first alpha emission would mark the breakdown from
293
118 to
289
116 – itself an undiscovered element. The next would mark the breakdown of that nucleus to
285
114, and so on all the way down to
269
106, which is seaborgium. The entire cascade was expected to take about two seconds. The software was designed to pick this characteristic chain of alpha emissions out of millions of irrelevant events.

Ninov was clearly under a great deal of pressure in the last few months of 1996. ‘We… convinced Victor Ninov that the reaction should be run as soon as possible,’ wrote Ghiorso and Hoffman later, ‘as we greatly feared GSI or Dubna might do it first.’ Whether this pressure came only from Ghiorso and Hoffman, or also from Gregorich and Loveland is not clear. When I talked to the latter two men in 2006, both of them suggested that the experiment was planned more as a shakedown cruise for the new equipment than as a confident attempt to find a new chemical element.

For several months, technical difficulties with the apparatus delayed them. The general level of motivation jumped considerably in January of 1999, however, when a report came in that the Dubna lab had created a single atom of element 114, using the calcium-plutonium reaction described earlier.

The Russians had run through more than a million dollars’ worth of
48
Ca to achieve that one seemingly successful strike. If superheavy nuclei are defined as those with an atomic number greater than 112 – the usage favoured by Hoffman and Ghiorso – then this was the first superheavy nucleus ever created. What’s more, the nucleus stayed intact for all of half a minute before breaking down. If this finding was genuine, the Russians had already landed on or near an island of stability. ‘[W]e felt happy that at last the Magic Island had been found,’ wrote Ghiorso and Hoffman in their 2000 book, ‘and we redoubled our efforts to get our own experiment under way.’

On April 8, 1999, the Berkeley experiment finally began. Over a period of four days, the lead foil target was bombarded with 700,000,000,000,000,000 krypton nuclei, each of them boosted to an energy of nearly half a million electron volts. When Ninov applied his software to the resulting data, he found two alpha-decay chains. The energies of the alpha emissions and the time intervals between them were remarkably close to the values predicted by Robert Smolanczuk. It seemed clear that the run had produced two atoms of element 118, which had decayed into another never-before-seen element element – 116 – and then into element 114 and even lighter elements.

To be sure of the result, the group ran another experiment a couple of weeks later, in which they hurled more than twice as many krypton atoms at the target as they had done during the first run. The researchers were therefore expecting that they might see four or more alpha chains. In fact they only saw one, but it was a beauty, again confirming Smolanczuk’s predictions. So the research team assumed that the lower yield on the second run was just a statistical fluctuation, and they added the one sighting to the previous two, meaning that three atoms of element 118 had now been created.

‘Does Robert talk to God or what?’ exclaimed Ninov, according to Ghiorso and Hoffman’s memoir. There was general amazement at the close fit between the observations and Smolanczuk’s theory. Clearly there was some initial worry about this among the researchers, but the worry eventually gave way to jubilation. ‘It was such a startling discovery that strenuous efforts were made to find out if anything had gone wrong,’ wrote Ghiorso and Hoffman, ‘but nothing obvious was uncovered… Now there is no question, the Super-Heavy Island actually exists!’

The findings were quickly written up and submitted to
Physical Review Letters
, a journal that specialises in rapid publication of newsworthy findings. The paper appeared in print on August 9, 1999, only three months after the experiments were completed. The paper had 15 authors. First came Ninov, Gregorich, and Loveland – the central players – followed by a group of other faculty members who had played at least a peripheral role, including Ghiorso, Hoffman and Swiatecki. The list was rounded off with a gaggle of graduate students who, as Loveland put it, ‘took shifts minding the separator in the middle of the night and stuff like that.’

The paper described the observation of the three alpha decay chains and the evidence that these had originated in three nuclei of the new element 118. From their data, the authors calculated that the probability of production of element 118 by the krypton-lead fusion reaction was about 2 picobarns – well shy of Smolanczuk’s estimate of 670 picobarns, but still an amazingly efficient reaction compared with what most nuclear chemists would have predicted.

The publication of the paper was accompanied by excited pronouncements from the scientists involved. ‘We jumped over a sea of instability onto an island of stability that theories have been predicting since the 1970s,’ said Ninov, according to the
Berkeley Lab Research Review
. In June, Hoffman and Ghiorso added a bubbly epilogue and some more exclamation marks to their already overloaded book. ‘We have convincing evidence for elements 114, 116, and 118!!’ they wrote. They mentioned their sadness that Seaborg had not lived to witness the discovery. (Seaborg was a posthumous co-author of the book, however.) All the major US newspapers carried stories about the discovery, often on their front pages. It was an American flag, after all, that the Berkeley group had planted on the Magic Island.

Very quickly, rival laboratories geared up to duplicate the feat of Ninov and his colleagues – but they couldn’t. Over a period of a year or so, GSI, GANIL (a French laboratory) and later RIKEN (in Japan) all announced their failure to create even a single 118 nucleus by the krypton-lead reaction, even with levels of bombardment that made the Berkeley Lab’s efforts seem like a mild peppering.

Meanwhile, back in Berkeley, the researchers were tearing down and rebuilding their equipment, including the gas-filled separator, in preparation for new studies. It was a time of great optimism. ‘It was clear that once this reaction worked, there were many other reactions you could open up – you could almost discover the other chemical elements one by one in a straightforward manner,’ Loveland told me. Loveland himself wrote a whole new suite of analytical software in preparation for the new work.

The Berkeley group’s reaction to the negative reports from overseas was fairly dismissive; they believed that the equipment in those other laboratories didn’t have the requisite sensitivity to replicate their own findings. In March of 2000, Darleane Hoffman was awarded the American Chemical Society’s Priestley Medal. In her acceptance lecture, she gave pride of place to the discovery of element 118.

Still, the Berkeley group eventually realised that it might be a good idea to repeat their own work before venturing further. So in early 2000 they conducted two more runs, totalling about the same length as the successful runs of the previous year. Not a single element 118 nucleus was spotted – a statistically very improbable result, assuming that the earlier runs were authentic. The group put together a committee of nuclear science specialists from other Berkeley labs to examine the problem, and in January of 2001 that committee wrote a report that focused on possible problems with the equipment during the 2000 runs, as well as on ways to resolve them. It seemed likely that Gregorich and his colleagues just needed to retune everything and they would soon get back to their winning ways.

Following the unsuccessful 2000 runs, the group did in fact spend about a year checking and improving their equipment. The next run began in April of 2001. This run, in Loveland’s phrase, was ‘massive’: the beam, the detectors and many other aspects of the experiment had been so thoroughly upgraded that the researchers expected to reap a rich harvest of element 118 atoms. Yet several days went by with no signal. Then, finally, Ninov announced success: he had found just one alpha decay chain with the unambiguous signature of an element 118 nucleus.

‘Victor came up with the chain,’ said Loveland, ‘and Don Peterson – who was the post-doc working with me – and myself were there, and we said, “Victor, we want to be able to see that chain,” because we now had our own software to analyse the data. We said, “Let’s go ahead, tell us where it is on the tapes, we’ll look at it.” We tried; it was an eventful weekend. We tried very, very hard and had absolutely no success. We couldn’t find this event at all… We talked to Victor and said, “There’s a terrible problem, we can’t find this.” He said, “Oh, I’ll show you,” and he pulled it up on his computer screen, and I said, “Oh my God, this is really tough, because now we’re caught in a situation where it depends whose software we’re using, and there’s something terribly wrong at that point.”’