It’s 1986, and there’s a puzzle on Dave Alburger’s desk. Not Ernö Rubik’s latest toy, but the data from a four-year experiment to measure the half-life of the rare radioactive isotope silicon-32. On one level, the numbers fit together just fine, adding up to a half-life of 172 years, in keeping with previous estimates.
There’s a devil in the detail, however. The sample’s radioactivity has not been dropping steadily over time, as the textbooks demand. It has fallen, to be sure, but superimposed on that decline is an odd, periodic wobble that seems to follow the seasons. Each year, the decay rate is at its greatest around February and reaches a minimum in August.
If we know anything about radioactivity, it’s that this kind of thing just doesn’t happen. Radioactivity decreases predictably over time. That’s why we can tell the age of rocks, fossils and prehistoric artefacts by the activity of radioactive atoms within them, and why nuclear waste becomes less toxic over time.
Advertisement
The fault was surely in some detail of the experimental set-up. Yet try as they might, Alburger and his colleagues at the Brookhaven National Laboratory on Long Island, New York – all nuclear physicists highly versed in this kind of painstaking measurement – couldn’t find it. Eventually they published the result anyway, noting that although the variations were a puzzle, they had no bearing on their value for silicon-32’s half-life (Earth and Planetary Science Letters, vol 78, p 168).
And there the result languished, a scientific skeleton in the closet. Until last year, that is, when it was rediscovered and dusted down by and Jere Jenkins of Purdue University in West Lafayette, Indiana. They think the data fits into an emerging pattern indicating that radioactivity is not quite the immutable process we assume it to be. Instead, it is susceptible to unseen interference from an unexpected quarter – the sun.
This controversial view goes against the grain established by , the New Zealand-born physicist who discovered the structure of the atom. In 1930, he and colleagues measured the decay rates of various isotopes, concluding that “the rate of transformation of an element has been found to be a constant under all conditions”.
We have since learned that certain decays can be influenced by electromagnetic fields, but Rutherford’s core conclusion stands firm. Atoms in a chunk of radioactive material decay with an equal probability within a given time. It’s a random process at the atomic level: you can’t tell when any one atom will pop, but the fewer there are left, the less frequently it occurs. The result is a characteristic curve of activity that falls exponentially over time.
When Fischbach and a student, Shu-Ju Tu, stumbled upon Alburger’s old results, they were not looking to overturn that picture. Rather the reverse: they had developed a new test of randomness and were using nuclear decay data to see if it worked. The Brookhaven results stopped them in their tracks. “We could see just by looking at it that the data was not random,” says Fischbach. Intrigued, he and Jenkins began combing the results from other groups to see if anyone else had reported a similar seasonal effect.
Sure enough, someone had. It was not as clear-cut as the Brookhaven case, but in 1998 a team at Germany’s national metrology lab, the in Braunschweig, had seen an annual variation in the decay rate of radium-226, an isotope with a half-life of about 1600 years. The experiment had run for 15 years in the 1980s and 90s (Applied Radiation and Isotopes, vol 49, p 1397).
Do two swallows a summer make? Countless measurements of the radioactivity of many different elements have been made over the years. If just two had thrown up an anomaly – even the same anomaly – surely the error must lie in the experiments?
Yes and no. Tests of relatively few isotopes would throw up a subtle annual oscillation, even if it were a general feature. For a start, catching such variations requires decay rates to be counted over several years, impossible for the great majority of radioactive isotopes which have half-lives shorter than a few dozen years. Equally, counting experiments are not performed on stabler isotopes that decay over hundreds or thousands or millions of years at all: the change in count rates over the course of an experiment lasting even years would be too small to be measurable. That leaves relatively few elements, like silicon-32 or radium-226, with half lives of a few dozen to 1000 or so years, that would show the effect.
For Fischbach, the significant thing was that the results were both from world-class laboratories. It would not be surprising for odd variations to occur in a long-term experiment. Buildings and the equipment they contain heat up and cool down over the course of a year. Environmental parameters such as atmospheric pressure and humidity also change over time.
Alburger and his colleagues, though, had meticulously designed their experiment to avoid such problems. They measured the decay rate not only of silicon-32, but also of chlorine-36, a much longer-lived isotope, under the same conditions. By measuring the ratio of the decay rates, any systematic errors resulting from the way the experiment was set up or changes in its environment should have cancelled out. But they didn’t.
Fischbach and Jenkins considered various possible explanations. Eventually, they . The seasonal variation seemed to track precisely the 3 per cent change in the distance between the Earth and the sun as the planet completes its slightly elliptical orbit. The closer Earth was to the sun, the higher the decay rate was. It was a convincing fit, but only half an answer. What on Earth – or off it – could be behind such a correlation?
Nuclei such as silicon-32 undergo beta decay, during which a neutron in the atomic nucleus decays into the slightly less massive proton. As it does so, it emits an electron and a near-massless particle, an antineutrino. As antineutrinos are notoriously difficult to detect, beta decay is signalled simply by a nucleus spontaneously emitting an electron.
Fischbach and Jenkins suggest that another reaction would, in theory, have the same signature. If a neutrino – a sister particle to the antineutrino – knocked into a neutron in an atomic nucleus, it would produce a proton and an electron. The nuclear fusion reactions that power the sun’s core are spewing neutrinos equally in all directions. The further away from that source you go, the more spread out those neutrinos are. The higher flux of neutrinos through the Earth when it is close to the sun would therefore bump up nuclear decay rates (see diagram).
New interaction
It’s a neat idea, with just one catch. For it to work, neutrinos must interact with neutrons much more readily than has ever been measured. “There would have to be some kind of additional interaction that for some reason had never been observed before,” says , a nuclear physicist at the University of California, Berkeley. “That seems unlikely.”
agrees. A physicist at the Fermilab particle accelerator facility in Batavia, Illinois, his work had already questioned a claim that the energy of particles varied with the seasons when hitting an underground detector at the DAMA experiment at in central Italy. This variation, with a maximum in June and a minimum in December, had been proposed as the signature of the solar system’s passage through a sea of dark matter thought to perfuse our galaxy (żěè¶ĚĘÓƵ, 26 April 2008, p 14). But Cooper’s analysis suggested that subtle seasonal effects affecting DAMA’s detectors could not be discounted as the cause.
Following a suggestion made by Fischbach and Jenkins, Cooper tested the new claim by looking at the trajectories of space missions powered by radioisotope thermoelectric generators. These RTGs harness the heat created by plutonium as it undergoes beta decay to produce electricity. If the new idea were correct, the further out in the solar system the spacecraft travels, the smaller the flux of solar neutrinos would be and therefore the slower the rate of plutonium decay.
One spacecraft seemed an ideal candidate for investigation. NASA’s Cassini mission to Saturn, launched in 1997, followed a trajectory first towards the sun, gaining energy in a gravitational slingshot around Venus, and then outwards past Earth and Jupiter. What Cooper found was: . The power from Cassini’s generators fell exponentially as it flew both towards and away from the sun in exactly the way it would have done on Earth.
A quiet end to a heretical theory? Fischbach and Jenkins don’t think so. They counter that the power developed by Cassini’s generators is proportional to the difference in temperature between the plutonium it contains and the outside of the spacecraft. This temperature difference changes in accordance with the square of the probe’s distance from the sun in exactly the opposite way to the neutrino flux. It would therefore almost perfectly cancel out any variations in the decay rate as measured by the RTG’s power output.
The duo is now busy combing the scientific literature for other evidence which might bear out their solar neutrino theory. They have had some success. There is the case, for example, of Ken Ellis, a medical physicist at Baylor College of Medicine in Houston, Texas, who over nine years found seasonal variations of about 0.5 per cent in the decay rate of plutonium-238 used for radiation studies of the chemical composition of the human body .
The evidence is bitty, however, and the consensus is that much more is needed before the theory can be properly assessed. Alvin Sanders, a physicist at the University of Tennessee in Knoxville, thinks there could be something in it. He reckons it might also hold the key to another curiosity – the fact that when the age of trees judged using carbon-14 dating is compared with their age gauged by counting their rings, the discrepancy between the two gets larger and smaller over a cycle of about 200 years.
“Wiggles in carbon-14 dates are well known as a nuisance,” says Sanders. Fluctuations in incoming cosmic rays and in Earth’s magnetic field have been proposed as explanations, but Sanders thinks that . The well-documented 200-year period of sunspot activity, known as the de Vries/Suess cycle, would cause variations in the number of neutrinos being emitted by the sun, which would in turn influence carbon-14 decay rates.
“What we are seeing may be the heartbeat of the sun,” says Sanders. It means that carbon-14 data from Earth, as a proxy for the sun’s neutrino activity, could allow us to determine the history of the sun’s internal reactor stretching back thousands of years. Quite generally, if Fischbach and Jenkins should ultimately be proved right, nuclear decay would represent a powerful way to detect neutrinos – something that currently requires experiments on a huge scale – and a new type of telescope with which to peer inside the sun.
At the moment that is speculation. Twenty years on, the mystery of Alburger’s result remains, and until it is explained nothing should be dismissed out of hand. Researchers are still searching for a mundane explanation. Last month, Tom Semkow, a physicist with the New York State Department of Health in Albany, and his colleagues proposed that despite all the precautions taken, some of the variation in the Brookhaven data might be explained by seasonal temperature changes. Their idea is that hot air is less dense and absorbs fewer beta particles, increasing the count rate registered at a detector (Physics Letters B, vol 675, p 415). Even if that is right, though, it doesn’t look to be enough to explain the whole effect.
Alburger himself, long since retired, is almost apologetic that the issue remains unresolved. “I am sorry that I am unable to throw any further light on these curious and as-yet-unexplained results,” he told żěè¶ĚĘÓƵ. Fischbach and Jenkins might have made a worthy stab at explaining them, but it looks likely that this skeleton will be hanging in the closet for a while yet.
Wibbly-wobbly neutrinos
If neutrinos cause wobbles in nuclear decay, it seems nuclear decay may return the favour by producing wobbles in neutrinos.
The background is research from the in Darmstadt, Germany, which found a strange periodicity in the decay of two heavy radioactive ions, praesodymium-140 and promethium-142.
These ions undergo a process similar to beta decay known as electron capture, in which a proton in the nucleus absorbs an electron and changes into a neutron, emitting a neutrino. This decay changes the mass of the ions, and can be identified as a change in the speed at which they race around a magnetic storage ring.
Yuri Litvinov and colleagues, who carried out the experiments, found a standard exponential decay with half-lives of 3 minutes, 23 seconds for praesodymium-140 and 40.5 seconds for promethium-142. But superimposed on each was an oscillation, with the measured decay rate increasing and decreasing every 7 seconds – just like in the Brookhaven case, but on a much shorter timescale .
Unlike in the Brookhaven case, there is a growing consensus as to the cause. Neutrinos come in three different “flavours” with different tiny masses, and they can oscillate between these forms. The suggestion is that this quantum-mechanical oscillation changes the momentum associated with an ion’s decay, and so affects the time it takes the isotope to move around the ring. The regular oscillation between two states of the neutrinos emitted by the isotopes produces the observed 7-second cycle.
If so, that is an exciting development, because it raises the possibility of studying the behaviour of neutrinos in an entirely new way – and in a much smaller-scale experiment than has ever been possible before.
