This is the third in a series of posts covering the basics of particle physics, originally posted back in 2003. In this installment, I talk about some of the hardware involved, specifically the CLAS detector at Jefferson Lab, because I’ve heard a good number of talks about that.
It should be noted that the inspiration for this whole thing was the announcement of the discovery of a “pentaquark” particle at a couple of accelerators. That discovery is by no means certain, but I’m still fairly happy with the explanatory aspects of these posts. I’m certainly not bothered enough to re-write them.
So feel free to ignore pentaquark-specific comments in these reposts…
At the end of my last post about the creation of pentaquarks, I left things hanging a bit. Specifically, I didn’t explain how the experimenters go about figuring out that these things were created in the first place. In essence, this comes down to explaining what the graph at the bottom of this page is, and how they get it.
The basic process is the same for all particle physics type experiments, and traces its lineage back to the immortal Ernest Rutherford, who was the first physicist to earn fame by shooting small particles at the nucleus of an atom and finding surprising results. Granted, things have gotten a lot more complicated since Rutherford’s day– his shocking discovery was simply the fact that there is a nucleus in the first place (this was deemed one of the most beautiful experiments in physics)– but the method is unchanged. Essentially, you get a bunch of small particles (protons, electrons, photons), fire them at some sort of target (atoms, protons, anti-protons), and look at what comes out after the collision.
Just by itself, that sounds like a pretty horrible way to do science– it’s akin to trying to figure out how to make a wristwatch by dropping one off a tall building and looking at the pieces that fly off. But the real story is actually even worse than that, for two reasons: 1) As noted in the previous post, what emerges from a collision isn’t necessarily what you started with, because some of the energy of the colliding particles can be used to create new particles out of nothing, and 2) With very few exceptions, none of these particles last very long once they are created, with the more exotic particles falling apart into other things within a tiny fraction of a nanosecond. Reconstructing what happened in a particle collision is a very difficult process indeed, requiring whopping huge, intimidating detectors, and a horrible alphabet soup of acronyms (CEBAF, CLAS, SPring-8, LEPS).
I’m far from an expert on how this stuff is done, but I’ll try to lay out the basic bit of apparatus that were used in one of the pentaquark experiments (specifically, the ones in Hall B of the Jefferson Lab in Virginia). The details vary a bit from accelerator to accelerator, and experiment to experiment, but the basic ideas are generally the same.
The central goal of the whole process is to identify the particles that come screaming out of the target area after a collision occurs- and they are screaming, moving at a fair fraction of the speed of light. Particle identification comes down to measuring a handful of properties– charge, mass, total energy, and momentum– and deducing what the particle was on that basis. To pick a simple example, if a positively charged particle with a mass of just under 1 GeV (1 giga-electron-volt, or 1,000,000,000 times the energy an electron acquires when accelerated across a 1 volt potential difference) passes through your detector, you can confidently say that that was a proton. A neutral particle with a similar mass would be a neutron, while a negatively charged particle of the same mass would be an anti-proton.
The clever trick here is that all these properties are worked out indirectly, by tracking the motion of the particles after the collision. You have to do it this way, for a number of reasons– these are subatomic particles, after all, some of which don’t last very long, so you can’t just pin them down and interrogate them at your leisure. Instead, particle-physics experiments set up situations where they can record the tracks of large numbers of particles, and identify them later on from the way they move.
There are generally a few layers of detectors in any such experiment, each of which serves a slightly different purpose. In the specific case of the CLAS (“CEBAF Large Acceptance Spectrometer,” where CEBAF stands for “Continuous Electron Beam Accelerator Facility.” This is a government lab, so they need lots of acronyms…) detector used for the “pentaquark” results, these fall into three general categories: scintillators, calorimeter, and drift chambers. These are arranged in concentric layers around the central region where the collision occurs. (A spiffy exploded diagram of the CLAS detector exists, because I’ve seen it in talks, but I can’t find it on the J-Lab site, the bastards, so you’ll just have to use your imagination– picture a big glass and metal onion, several stories high…)
Scintillators are basically blocks of glass, which are prepared in such a way that a particle slamming into the glass will produce a small flash of light, which gets picked up by a detector. These are the outermost layer of detectors, and are mostly used for timing purposes: when you see the flash, you know that a particle hit the detector. If you know when the collision occurred at the center of the detector, you can use this to work out the speed of the particle.
Calorimeters are used to measure the energy of some types of particles. Basically, when the particle enters the calorimeter, it loses some or all of its energy to the calorimeter, which tallies up the energy gain, and calculates the total energy of the particle. These are the penultimate layer in the CLAS detector, just inside the scintillators.
The most important components of the detection system are the drift chambers. These consist of large volumes of space divided into small cells with an array of fine wires. When a charged particle passes through one of these cells, it causes a little “blip” of current in one of the wires. By keeping track of successive “blips,” you can stitch together a map of the particle’s track through the chamber– it passed this wire, then that one, then that one over there, and so on.
“Big deal,” you say. After all, knowing exactly what path the particle took on its way out doesn’t necessarily get you any information that you don’t already have. That’s why you put the whole thing in a large magnetic field.
A charged particle (an electron, say) moving in a magnetic field will feel a force that depends on three things: 1) the velocity of the particle, 2) the magnetic field, and 3) the charge of the particle. An electron moving horizontally through a vertical magnetic field will feel a force to the left, while a positron would feel a force to the right. The faster the particle, the bigger the force, and the stronger the field, the bigger the force.
This means that particles passing through a drift chamber in a magnetic field will follow curved tracks, and by following the tracks, you can get a lot of information about the particle, and putting this together with the information from other detectors gets you all the information you need. You can get the sign of the charge from noting which direction the particle curves (positive charges go one way, negative charges the other). If you know the strength of the field and the speed of the particle (from the scintillators), you can figure out the mass (heavier particles curve more slowly than light ones) from Newton’s Laws (at least, the relativistic analogue thereof). (Technically speaking, what you get is the ratio of the charge to the mass, which gives you the mass for known charges– if you don’t know what the charge is, you need the mass to complete your knowledge, and you can get that from other detectors.)
The CLAS detector actually has three layers of drift chambers. Particles leaving the collision region pass through one chamber with no magnetic field, then one with a large field applied, then a third with no field again. I’m not entirely sure why they use three layers– possibly to provide a couple of bits of straight track that can be traced back to the source, but I really don’t know.
Putting all of this together gives you a complete description of the collision– you know what two things hit each other (because you set it up, choosing the particles that make up the beam and the target), and by stitching together all the information from the detectors described above, you can work out what came out of the collision region (what you can detect, anyway– some of the products may be things like neutrinos that pass right through the detector without leaving a track). That’s where the difficult part starts…
But a crude sketch of how you use this to identify new types of particles will have to wait for another post…