The Eight-Second Gap: How Long Does a Neutron Live

Bölüm metni

A neutron that leaves the nucleus and is left on its own decays within about fifteen minutes. But measure that span by two different methods and you get two different answers — and the eight seconds between them cannot be blamed on experimental error. For twenty years, either someone misunderstands their own apparatus, or some of the neutrons are quietly going somewhere else.

The neutrons sitting right now in the nuclei of the atoms of your body have stood there, undecayed, since the first three minutes of the universe. Thirteen point eight billion years. Take one of those same neutrons out of the nucleus, leave it alone in an empty room, and it dies within a quarter of an hour. It breaks apart, it scatters, it turns into a proton, an electron and an antineutrino. Immortal inside the nucleus, mortal outside it. The whole difference lies in the presence of its neighbours: the binding energy inside the nucleus does not leave it the tiny budget it needs in order to decay.

That budget really is tiny. The mass of a free neutron, put in units of energy, is nine hundred thirty-nine point five six five megaelectronvolts. The combined mass of the proton and the electron it leaves behind when it decays is nine hundred thirty-eight point seven eight three. The entire affair takes place in the zero point seven eight two of space between those two numbers. The neutron can decay because it fits into that thin energy gap between itself and its own products; close the gap, and there would be no such temporary thing in the universe as a free neutron.

So what is the answer to the question of how long it lives? For a single neutron there is no answer. Decay is random; a neutron does not age, does not tire, and a neutron that has already been waiting for an hour is exactly as likely to decay in the next second as one that has just been born. What can be measured is the average of a crowd. That average, the mean lifetime, is about fourteen and a half minutes. For half of a crowd of neutrons to be used up, about ten minutes is enough. These two numbers are two different expressions of the same thing, and physicists speak of the mean lifetime.

Why is such a simple question still being asked? Because the decay of the neutron is the cleanest, plainest example of the weak nuclear force we have in our hands. No accelerator is needed, no exotic particle is needed; there is only the neutron, and it falls apart of its own accord. That is why its lifetime behaves like a constant of nature: a calibration number that goes in and out of other calculations and decides whether they are right. And when you are measuring a calibration number, an accuracy of one percent is not enough for you; you want one part in a thousand, even one part in ten thousand.

Humanity has found two ways of measuring this, and the two of them work on exactly opposite logic.

The first way is the bottle. You fill a vessel with neutrons, you wait, and then you count the ones left over. It sounds easy, but holding a neutron in a container is not easy: having no charge, it cannot be caught with electric fields, and it passes straight through walls. The solution is to slow the neutrons down to an extreme. Moving at a few metres a second, the pace of a walking human being, these ultracold neutrons are so feeble that gravity can hold them and a strong magnet can bounce them back. The trap built at Los Alamos National Laboratory, in the American state of New Mexico, does exactly that: a bathtub-shaped vessel whose floor and sides are woven out of permanent magnets and whose lid is nothing but gravity. The neutrons never touch any material surface; they sway on a magnetic cushion. Fill it, wait, pour out what remains and count. And you do not even need an absolute number: if you compare what is left after waiting thirty minutes with what is left after waiting ninety, the slope of the exponential decline gives you the lifetime directly. Between two thousand seventeen and two thousand nineteen the Los Alamos team counted more than forty million neutrons this way, and in two thousand twenty-one they published this result: eight hundred seventy-seven point seven five seconds, with an uncertainty below half a second. Zero point zero three nine percent. The most precise measurement made to date.

And the weak point of the bottle lies in exactly the same place. The vessel cannot tell the difference between a neutron decaying and a neutron escaping. Every neutron that leaks out through a gap at the edge of a wall, that warms up as it strikes and so gets free of the trap, that has its magnetic orientation flip and so escapes the magnet's push, shows up in the ledger as having decayed. Which means every hidden loss in the bottle method makes the lifetime look shorter than it is.

The second way is the beam, and it counts the dead. You pass a beam of cold neutrons coming out of a reactor through a volume whose length is known down to the millimetre. Every neutron that decays inside that volume leaves a proton behind; the proton has charge, so it can be held with electric and magnetic fields and counted one by one. Then you measure how many neutrons there are inside that same beam: you place in the beam's path a thin layer of lithium-six, which swallows a known small fraction of the neutrons going through, and from the number of absorptions you work out the flux. How many protons were born per second, how many neutrons were in that volume; the ratio of those two gives you the lifetime. The American national measurement body, the National Institute of Standards and Technology, the Institute of Standards, has been working with this method for years, and the result it published in two thousand thirteen was eight hundred eighty-seven point seven seconds, with an uncertainty of about two seconds.

The beam's weak point, too, is the mirror of its method. Here you are condemned to two absolute counts: of the protons and of the neutrons. A proton that goes unnoticed reduces the number of the dead and makes the lifetime look long. A neutron flux that has been overcounted does the same thing. So the errors of the beam method tend to make the lifetime look longer than it is.

And here is where the collision happens. The bottle says eight hundred seventy-seven point seven five; the beam says eight hundred eighty-seven point seven. One answer is fourteen minutes and thirty-eight seconds, the other fourteen minutes and forty-eight seconds. The puzzle got its name while the difference between them stood at around eight seconds; as the measurements sharpened, the difference did not close, it widened towards ten seconds. Put in statistical language, the probability that the two results are merely scatter around the same true value is a few parts in a thousand. So either there is an error in one of the two experiments that nobody has found yet, or the neutron is doing something that is not written down in our ledger.

You might ask what difference eight or ten seconds make in a lifetime of a quarter of an hour. The difference is that this number is not a duration but a setting. In the first minutes of the universe the neutron's lifetime ran like an hourglass, and the rate at which that hourglass emptied determined the composition of the matter we see today.

It went like this. About one second after the Big Bang, the universe grew too cold to go on turning protons into neutrons and neutrons into protons. At that moment a count was taken, and there was roughly one neutron for every six protons. But the story did not stop there. For those neutrons to get inside a helium nucleus and reach safety, the universe had to cool enough to hold deuterium together, and that took minutes. Throughout those minutes free neutrons went on decaying. When the door to helium production finally opened, about three minutes after the explosion, every surviving neutron was locked away into helium almost instantly. The only reason the ratio slipped from one in six to one in seven is this clock of death, ticking all through the wait. The fact that a quarter of the ordinary matter in the universe today is helium by mass has nothing to do with stars; that helium was laid down in those three minutes, and its quantity was written by the neutron's lifetime.

So the calculation can be set up like this: if the neutron lives longer, more neutrons are still alive when the door opens, and there is more helium. Here you would want to call the universe itself as a referee. You look at distant primordial gas clouds that stars have not contaminated, you measure the helium fraction, and you say which lifetime it fits. Unfortunately the referee cannot yet see a difference this fine. A change of one percent in the lifetime shifts helium's mass fraction by only about zero point two percent. The ten-second dispute moves a helium fraction of around zero point two four seven by some six parts in ten thousand. Whereas the uncertainty on the helium fraction read off the sky is still of the order of a few parts in a thousand. So in this case the universe is a knowledgeable witness whose eyesight is not quite sharp enough. If observational sensitivity comes down to one part in a thousand, its testimony will begin to speak.

The place where the number really hurts at the moment is elsewhere: the internal consistency of the weak force itself. The neutron's rate of decay is directly tied to the largest entry in the table that lists the probabilities of quarks turning into one another, the transition amplitude between the up quark and the down quark. If you measure the neutron's lifetime and the direction in which the electron prefers to fly off as it decays, you can calculate that entry. Then you add up the squares of the numbers in the table's first row; theory says the total should come out to exactly one. That sum is one of the sharpest examinations the Standard Model faces, and in recent years an uncomfortable gap at the level of one part in a thousand has been under discussion; a tension that does not yet count as a discovery, but that will not go away either. The neutron could have entered this debate as an independent witness. It does not, because the one point one percent uncertainty in its lifetime carries a blurring of half a percent into that critical number, whereas the precision at which the debate is taking place is of the order of three parts in ten thousand. The marks on the ruler in your hand are twenty times thicker than the notch you need to measure.

And now the most provocative side of the puzzle. What these two methods measure is not, in fact, one and the same thing. The bottle counts the survivors; which is to say, it is sensitive all at once to every route by which a neutron can disappear. The beam counts protons; which is to say, it sees only a decay that produces a proton. Think about it: what would happen if about one percent of neutrons were leaving by an invisible route that produces no proton? The bottle would count those losses as decays too, and find the lifetime short. The beam would never see that loss at all, and find the lifetime long. The sign of the difference, its size, its direction; all of it sits exactly on the observed picture. In two thousand eighteen, Bartosz Fornal and Benjamín Grinstein put this possibility on the table as a serious proposal: perhaps the neutron, with a probability of one percent, is turning into dark matter.

The idea was exciting, but it had to pass through a very narrow door, and nature had long since fixed the door's measurements.

First, mass. The hypothetical dark particle the neutron would turn into has to be lighter than nine hundred thirty-nine point five six five megaelectronvolts; otherwise there is not enough energy for the transformation. But it cannot be too light either, because then the same transformation becomes possible inside nuclei as well. Here a quiet but decisive witness steps in: beryllium-nine. Among the stable nuclei, it is the one that holds on to its neutron most loosely; prising that neutron off takes only one point six six four megaelectronvolts. If the dark particle had a mass below that threshold, beryllium-nine nuclei would fall apart of their own accord; they would end up as two helium nuclei and a dark particle. Beryllium-nine is still lying around, not falling apart. So the dark particle's mass is squeezed into a window between nine hundred thirty-seven point nine and nine hundred thirty-nine point six, a window not even two megaelectronvolts wide. Before the hypothesis was even born, the mere existence of a stable nucleus had closed off most of the possibilities.

Then came the turn of evidence. If the neutron emits a photon alongside as it turns into a dark particle, that photon's energy has to lie between seven hundred eighty-two and sixteen hundred sixty-four kiloelectronvolts; the lower bound is set by the energy budget of ordinary decay, the upper bound by beryllium's loose grip. So the place to search was marked out to the centimetre. The Los Alamos team arranged sensitive detectors around the neutron trap and looked at precisely that range. Not a single line. A photon channel at the one percent level was ruled out with ninety-nine percent confidence. And what if an electron-positron pair comes out alongside instead? Another experiment in the same laboratory, measuring the electrons from neutron decay, excluded such a channel with a total energy above one hundred kiloelectronvolts at a certainty well beyond five sigma. A team working at the Grenoble research reactor in France also ruled out, independently, a one percent contribution across ninety-five percent of the permitted mass window.

So the visible versions of dark decay are dead. What is left standing is a channel that leaves no trace at all, wholly dark: a dark particle plus a dark carrier, zero signal in the laboratory. Searching for this directly is hardly possible, but it is not entirely unconstrained either. If neutrons can turn into a dark particle, that transformation works best where neutrons are most abundant, inside neutron stars; it changes the star's internal pressure, its structure, the greatest mass it can bear. The neutron stars we observe, reaching twice the mass of the Sun, therefore place an indirect but real limit on the hypothesis. It has to be said plainly here: dark decay is not a proven explanation today, only a possibility that has not yet been completely eliminated.

Meanwhile the boring explanation, the possibility that there is a systematic error in one of the experiments, has gained strength from an unexpected quarter. At Japan's accelerator facility J-PARC, the Japan Proton Accelerator Research Complex, a third approach has been set up. This apparatus also uses a cold neutron beam, but instead of the protons it counts the electrons that come out of the decay inside a gas chamber, and it measures the number of neutrons in the beam from the reactions given by helium-three nuclei mixed into that same gas. The same question, an entirely different geometry, an entirely different family of errors. The result that came out of the data the Japanese centre collected between two thousand fourteen and two thousand twenty-three was announced at the end of two thousand twenty-four: eight hundred seventy-seven point two seconds. Its statistical uncertainty is one point seven seconds, its systematic uncertainty still around four seconds, so it is not as sharp as the bottle. But what it points to is clear: it agrees with the bottle, and it stands in tension with the average of the beams that count protons. This changes the geography of the puzzle. The fault line may not be beam against bottle at all, but only in the proton trap's own accounting.

That is why the next step was not to argue but to build. The beam experiment at the Institute of Standards is being rebuilt from the ground up; the goal is to test every known systematic effect at the level of one part in ten thousand and to push the final uncertainty far below one second. The answer from this experiment could close the puzzle from either direction. If the new beam result comes down to around eight hundred seventy-seven, the eight-second difference will quietly go on the record as a measurement error, and nobody will go on hunting for a dark particle. If instead it brings its uncertainty down to half a second and still insists on eight hundred eighty-eight, the idea that the neutron has an invisible escape route turns into a serious candidate for a discovery.

What this puzzle really teaches is perhaps this: when two careful measurements fail to agree, physics does not vote on which of them is the more likeable. It publishes the difference openly, it writes down every single time which method was used to measure the lifetime, and it builds a third instrument to close the gap. The neutron is the simplest unstable particle we can hold in our hands; there is nothing mysterious about its quark structure, nothing complicated about its decay scheme. And even so, we cannot say how long it lives to one part in a thousand. Countless neutrons are passing through the room you are in right now, every second, and every one of them knows better than we do how many seconds it has before it falls apart.

Nöron Science'de bugünkü bölümün sonuna geldik. Yeni bir hikâyede yeniden buluşmak üzere.