The most obvious design challenge to detect the Higgs particle is that the colliding particles needed to have energies that are sufficient to produce the Higgs. Not knowing the mass of the particle complicated things but having a good idea that the upper limit of mass should be around 1 TeV helped. (For previous posts in this series, click on the Higgs folder just below the blog post title.)
In the previous post in the series, we saw that when you do not know what the mass is of the particle you are seeking to create, proton-proton collisions have some advantages. To see why the LHC was designed to have the energies it has, we need to realize that when two protons collide, you can visualize it as essentially two bags containing quarks and gluons smashing into each other. The reactions that produce new particles are those caused by an individual quark or gluon in one proton bag colliding with an individual quark or gluon in the other proton bag. The energy of the individual constituents of each proton will of course be less than that of the proton itself.
The average value of energy of an individual quark or gluon inside a proton is about one-sixth the energy of the proton as a whole, so if you want to be able to reach 1 TeV of collision energy of the elementary particles within the protons, you need each colliding quark or gluon to have an energy of at least 0.5 TeV which requires each proton to have an energy of at least 3.0 TeV. This means that you need the proton accelerator to reach a combined energy of the two beams of at least 6.0 TeV. This set the lower bound of the design energy though you need to go higher in order to build in a cushion and also to later probe excited states of the Higgs, if it has any. The LHC was designed to eventually reach energies of 14 TeV. It found the Higgs after having reached energies of 7 TeV, and it reached energies of 8 TeV before it was shut down for two years for the final upgrade to 14 TeV.
But even with the required energies being reached, there are other hurdles to overcome. Protons contain only up and down quarks and gluons. These two quarks are extremely light and the gluons are massless. But we saw that the strength of the Higgs particle’s interactions with other particles (and hence the probability that those colliding particles will result in a Higgs being produced) increases with the mass of that particle. Since the up and down quarks are very light and the gluons are massless, the situation does not look good for producing Higgs particles. The solution to this is the same as that for the detection of the Higgs in which the final particles detected were two photons or four charged leptons, all of which either do not interact directly with the Higgs or do so weakly. In other words, we have to look at multi-step processes in which the intermediate particles are heavy ones that do interact strongly with the Higgs.
One of the mechanisms that give rise to the most Higgs particles is via a two-step process with two gluons colliding. There are various possible production modes. A gluon in each of the colliding protons may first produce a massive particle (like the top quark) and its antiquark. These are very short-lived virtual particles and one pair of quark-antiquark can annihilate each other while producing a Higgs particle. In another mode, a quark in each of the protons may produce a W particle and the two W particles can then annihilate each other to produce a Higgs.
The catch is that the introduction of each intermediate step in a process results in the probability of that process decreasing and so such Higgs-producing reactions are of extremely low probability. Since both the production and decay channels involve these multi-step processes, the chances of producing and detecting a Higgs is extremely low. Hence in order to get a decent number, we need to create a vast number of proton-proton collisions, and that was another major challenge for the accelerator’s designers. They had to create a huge number of proton-proton reactions in order to have a hope of seeing the Higgs particle with sufficient statistical robustness.
The LHC consists of two proton beams circulating in opposite directions along a circular tube of circumference 26 kilometers. The protons in each beam are not spread uniformly but are grouped in small bunches about an inch long, each bunch containing about 100 billion (1011) protons. Currently there are 1380 bunches in each beam, with the eventual target being 2808. The two beams do not cross each other except at the locations of the two detectors known as ATLAS and CMS. At those locations the beams are narrowed to a width of just one-thousandth of an inch and every second 20 million of these bunches cross each other.
But despite the vast numbers of protons in each bunch, most of them simply pass by each other without colliding and only 20 proton-proton collisions occur when one bunch going in one direction passes through a bunch going in the opposite direction. Of course, since there are 20 million such crossings each second, that means each second results in 400 million proton-proton collisions. But even with this vast number, only about 15 of them were expected to result in a Higgs particle being produced. So you can see the immensity of the background problem and why it is compared to finding a piece of hay in a haystack.
There is another price to be paid in trying to increase the number of collisions in order to increase the chances of producing Higgs particles. Inferring the existence of the Higgs requires highly detailed information on the detected particles. A single proton-proton collision generates about one megabyte of such high-precision data. 400 million collisions per second result in large amounts of data that cannot possibly be stored for later leisurely analysis. You have to have some combination of software and hardware (called ‘triggers’) that can look at the data produced in real time and decide what looks ‘interesting’ and keep and dump the rest. Of the 400 million collisions per second, only a few hundred make the cut. Clearly there is a danger that with such quick decision-making you may throw out some good data but there is no choice.
The LHC eventually hopes to double the number of bunches in each beam, thus doubling the rate of production of Higgs particles and obtaining better statistics.
Next: Other design challenges of the LHC