A Notable Discrepancy at the LHCb Experiment

As many of you have already read on the BBC or on the physics blogs, such as Resonaances, Quantum Diaries and Cosmic Variance, a new discrepancy — this one quite substantial, and appearing in a measurement that is harder to question than some others we’ve seen this year — has appeared at a Large Hadron Collider (LHC) experiment, LHCb. (The two general purpose experiments, ATLAS and CMS, are the ones mainly involved in the search for the Higgs particle and for other new phenomena at high energy. The LHCb experiment is mainly focused on something else: measuring the properties of the bottom and charm quarks in even greater detail than has already been achieved.  LHCb’s efforts are thus largely, though not entirely, complementary to those of CMS and ATLAS. A fourth big experiment, ALICE, is mainly focused on the collision of heavy atomic nuclei, though ATLAS and CMS participate in this effort too.)

For a quick and understandable explanation (if a little thin on details), I suggest you read what Sean Carroll had to say on Cosmic Variance. There are a few small points I don’t precisely agree with, but on the whole he says something rather similar to what I could say in a short post. Resonaances’s report is much more detailed and very suitable for experts or budding near-experts.

On my end, things are a little busy with recommendation letters to write for young people applying for postdoctoral positions, and with a paper approaching final stages. But let me give a short summary for now, with more to come.

We need two definitions:

• Parity (P): an imaginary transformation of the world in which you imagine how it would look in a mirror.
• Charge conjugation (C): a similar imaginary transformation where you imagine replacing every electrically charged particle with a corresponding particle of opposite charge. You need to do a few other things too. I’m being vague here because it’s a bit subtle what you do, and the details matter.

Suffice it to say that if you made a  list of the known types particles of nature, including their antiparticles, you would find this list would not be the same if you flipped our world using parity, or in a world that was flipped by charge conjugation. C and P are not symmetries of the particles of nature; in the language of physicists, P is violated, and C is violated, at the 100% level by the very structure of the Standard Model. This was one of the great discoveries of the 1950s, and arguably one of the greatest discoveries in physics of the last century.  [You can see a hint of the violation of parity in the second figure, marked Figure 3 (sorry), of this article, where you can see that particles labeled as left- and right- are organized differently; you can get some of the point from the text of that article, but I realize that further explanation will be needed.]  Violation of partiy was suggested as theoretically possible in 1956 by Lee and Yang (who won a Nobel prize) and discovered in their suggested experiment by Wu later that year (and she arguably ought to have won a Nobel prize herself.) 

CP, however — which involves imagining doing C and then P, or P and then C (the order doesn’t matter), leaves the list of particles and antiparticles completely unchanged. It just flips particles to antiparticles, but since there’s an antiparticle for every particle, nothing has changed on that list. Again, I’ll have to explain this better in a later article.

In 1964, however, it was discovered that despite this, CP is not a symmetry of nature.  Even though it is a symmetry of the list of particles and antiparticles, it is not a symmetry of the way the particles interact with one another — one might say that it is not a symmetry of the forces of nature. (This is a loose way of saying it, but you’ll have to forgive me for now.) This was discovered by Cronin and Fitch, who … yep … won a Nobel Prize for it. It turns out that this effect can easily be understood within the Standard Model’s particles and forces as long as there are three generations of quarks (the 3 generations being [1] up and down, [2] charm and strange, [3] top and bottom.) This was pointed out in 1973 by Kobayashi and Maskawa — before the third generation or quarks was discovered or suspected — and so, since they essentially and correctly predicted that a third generation would eventually be found, they too won a Nobel Prize.

CP violation was originally seen as a very small effect in the decays of Kaons.  Kaons, like all hadrons, are full of quarks, antiquarks and gluons, but — unlike a proton, which has two extra up quarks (i.e. two more up quarks than up anti-quarks) and one extra down quark — a kaon has one extra strange quark and one extra up or down anti-quark (or vice versa, an extra strange anti-quark and one extra up or down quark.) [Hadrons that have one extra quark and one extra antiquark are called “mesons” nowadays — names often shift over the years, and decades ago, before the existence of quarks was known, the name meant something a bit different.] But once you know that there is CP violation that comes into the Standard Model through Kobayashi and Maskawa’s mechanism, you can easily calculate that it is going to show up in some other places. The best place to see CP violation is in B mesons (hadrons with an extra bottom quark and an extra up or down antiquark, or vice versa). B mesons have been studied in great detail in the past two decades, and quite large effects of CP violation in B mesons have been observed. But so far, none of those measurements has been wildly different from what the Standard Model predicts; in other words, there is no sign yet that any new unknown phenomena are affecting the properties of B mesons.

But in D mesons (hadrons with an extra charm quark and an extra up or down antiquark, or vice versa) the Standard Model predicts that most CP violation effects should be quite small — and for various technical reasons, quite difficult to observe. That sounds sad, until you realize that it still offers a huge opportunity. If there are any unknown particles or forces in nature that also violate CP, they could create small effects on D mesons that could still be much larger than predicted by the Standard Model… and perhaps large enough to be observable. Moreover, if you observed them, you would be sure they did not come from the Standard Model, because of our reasonable confidence that the Standard Model’s CP-violating effects are small for D mesons.

So what happened on Monday, at the HCP conference in Paris where it first became widely known, is that the LHCb experimental collaboration announced that they had looked for CP violation in D mesons, by comparing certain types of decays of D⁰ mesons, which have an extra charm quark and up anti-quark, to those of D⁰ anti-mesons, which have an extra charm anti-quark and an up quark.  They found these occur at slightly different rates, an effect which violates CP. [Details of how this was done will come in a later article.]  If the Standard Model were correct, any difference should have been too small to observe… probably much less than 0.1%. But instead, they saw a difference of about 1 percent. More precisely, they saw an effect that was 0.82%, with a statistical error of 0.21% and a systematic error of 0.11%. Using the formalities of statistics, this is equivalent to 3.5 standard deviations away from zero — to be compared with the 5.0 that is the rough standard used in the field for judging a case to be completely convincing. [Caveat: statistical uncertainties can only be interpreted if you assume nobody goofed. If experimenters, or their theorist colleagues, make an error, a measurement can arrive at a very convincing discovery … of the mistake! But I have no particular reason to think any mistakes were likely made here, so let me ignore this possibility for now.]

If this result holds up, it means that there are particles and forces not included in the Standard Model that violate CP in a new way. That would be huge news, of course — the Standard Model has kept us imprisoned in a single theoretical framework for more than 30 years, and any breakdown, for any reason, would be enormously important. But to figure out why it is breaking down would require some more information than this one measurement, of a single number, can provide.

Well, now what? First, we have the usual list of caveats. Most hints of new physical phenomena go away over time. That said, 3.5 standard deviations is rather large — large enough to deserve some serious attention. On the other hand, the various experiments at the LHC are making lots of measurements, and the probability that, by chance, one of those measurements will deviate from experiments by a very large number of standard deviations is pretty high. (The usual story — if you flip a coin ten times, the chance that you will get it to land on the same side ten times in a row isn’t that high, but if 1000 people do it, the chance that this will happen to one of their coins is very high indeed.) And if you read the conclusions from the LHCb talk, you’ll see the experimenters are suitably cautious (any professional experimentalist has seen things like this  go away over time with more data.) But for certain we should pay very close attention. By March, or even before, the LHCb experiment should have analyzed the rest of their 2011 data — the measurement was made with only half of the data collected this year — and next year they should get about double the data again. What that means is that somewhere between next spring and next fall, LHCb should have enough data to exceed the 5 standard deviation threshold, if this effect is real. The only problem is that no other experiment will be able to check the result anytime soon, as far as I know. We’ll have to hope that whatever is causing the effect — whatever new particles and forces might be responsible — shows up directly, in some other way, at the ATLAS and CMS experiments.

By the way, whenever you read about CP violation and LHCb in the press — this week’s BBC article is no exception — you will always see the mantra repeated that “The result may help explain why we see so much more matter than antimatter” in the universe.   This is not entirely propaganda, but there’s quite a bit of that going on. It’s quite a long and subtle argument (and the famous physicist Andrei Sakharov, of nuclear weapons and Soviet imprisonment fame, is credited for it) that takes you from CP violation to the matter-antimatter asymmetry of the universe. It certainly is true that an understanding of the matter-antimatter asymmetry may require a better understanding of CP violation; but it is certainly not true that new observations of CP violation will somehow, just by themselves, explain anything about the matter-antimatter asymmetry. You also have to understand details of the early history of the universe, which you can’t get from LHC experiments alone. You needn’t just take my word for this not-actually-heretical statement; you’ll see that my colleague Sean Carroll too is always careful to add this caveat when he writes about this subject, as he did this week.