How are particles measured?
September 11, 2008 7:35 PM   Subscribe

How are the attributes of atomic and subatomic particles measured?

I guess you can't measure a particle without slamming something into it, right? How is this actually done?

Bonus: Is it this that forces us to use uncertainty in our current model? Does measuring a particle's location, say, change its momentum?
posted by Citizen Premier to Science & Nature (5 answers total) 4 users marked this as a favorite

Which particles? Electrons are measured in, say, Millikan's oil drop experiment. Most of the LHC type physics is measurement of more permanent properties, like say, the Higgs boson's mass [or finding other particles' charge, or dipole moment. Things that determine how they act in a magnetic or electric field, or in the presence of gravity]. The principle you are referring to is called backaction, although you may have heard it attached to the Heisenberg uncertainty principle.

It's a consequence of the way things work that you can never have two "good" numbers for position and momentum, energy and time, and some other "observables". On the other hand, there are other properties which can be known at the same time--say x position and y momentum.

The only reference I can find for this is this, whose full text says unhelpfully,

"Thanks to quantum-mechanical effects, however, important information about Higgs bosons can be obtained from experiments that are unable to produce them directly. Because of the Heisenberg uncertainty principle, which allows elementary particles to appear and disappear spontaneously, physicists are able to constrain the masses of as-yet-unobserved particles [HN9]. Known as virtual effects, these fleeting apparitions may nevertheless have profound physical consequences. A Higgs boson should induce such effects, with magnitudes that depend on its mass. But because the effects are small, only high-precision experiments can elucidate them."
posted by gensubuser at 9:19 PM on September 11, 2008

this is a very big question...

the old school way is to point a particle with known properties at something you want to measure, and the way the unknown thing scatters the known thing tells you information about the unknown. this is, for instance, how rutherford and marsden were able to figure out that gold atoms have a nucleus and aren't a uniform blob of matter. in those days the scattering direction was measured by the glowing spots on a phosphor screen, which emits light when charged particles hit it.

this is the way it was done for a long time, with a beam of particles pointed at a target of something you wanted to study, and some screen or something to measure the scattering angle. turns out there's a lot you can infer from just the direction of scattering.

and it's still the way a huge number of things are studied, be they by neutron scattering, or x-ray scattering, electron diffraction (which is scattering + interference), and so on with many many variations. the instrumentation has improved (though phosphor screens are still used for lots of things, including seeing the beam in the LHC the other day). photomultiplier tubes, calorimeters, scintillators, cerenkov radiation monitors are all ways of detecting the arrival of various types of particles, and there are many others.

so. a feature of quantum mechanics and the wave-like nature of things is something called the de broglie wavelength, which in simplest terms says that the wavelength of a particle decreases as the particle increases in energy. waves won't be scattered by something much smaller than their wavelength. so, if you want to be able to resolve the structure of smaller things, like the structure of the nucleus (which was just a blob to rutherford, but now we know has protons and stuff) you have to use a higher and higher energy probe beam. so with nuclear physics the particle accelerators were born.

but, because E=mc2, when your probe beam has a lot of energy, it can hit something, lose energy, and create out of thin air some new particles. this process is called pair production. so now, you have your beam whacking into a target and creating particles, and hey, maybe these particles are interesting things. so if you put some instrumentation on your target, you can study the particles. which is interesting, because they are often things like pions and kaons that are unstable and not something you can make a target out of.

one of the earliest ways of doing this was to put the target inside a strong magnet. this is because charged particles that are in motion have their path bent by a magnetic field, and they move in circular (or, strictly, helical) paths. then, right behind your target you put some sort of thing that tracks the motion of the particles.

bubble chambers and cloud chambers were the earliest ways of doing this. take a photo of the bubble chamber at the right time and you can see the tracks that the charged particles leave, and if you know the strength of the field, you can reconstruct the collision events and deduce the particle charge, and mass, and measure the lifetime, and things like that.

at some point, someone realized that making heavier and heavier particles, which takes a higher and higher energy beam on the target, was difficult. so the collider, which runs two beams into each other, was born. at the place where the two beams collide (which is called the IP or interaction point), you put some instrumentation.

the way it's done today is to wrap the IP in successive layers of instrumentation, like a cylindrical onion, along with the big magnet that makes the charged particles go in circular paths. there's a picture of the CMS layers here. usually the innermost few layers is something called a vertex tracker, which is essentially a CCD like in a digital camera, but with a cylindrical geometry. these cylinders are nested around each other like the tubes in a telescope when it's not extended. the two beams enter through the ends, collide and make some new particles, which travel through each layer and a pixel lights up. if you know the geometry of the pixels, you can reconstruct the particle paths close to the collision point. these things are replacing drift chambers, and are both basically like an electronic version of the cloud chamber.

so from the curved (or straight) tracks in the tracker you can get the charge and the momentum. all of this will be wrapped in a huge superconducting magnet coil, and outside that will usually be a calorimeter, where the particle is absorbed and the total energy is measured. there will also be lots of other instrumentation depending on exactly what the experiment is looking for and the details of the particular study, like the muon detectors at the CMS. also, on top of that, the events you're interested in studying may be rare, so there has to be a huge and sophisticated triggering system to analyze all the data on the fly, separate out the background and only record the stuff that you're interested in.

anyway with all this data, and a metric ass-ton of CPU cycles, you can deduce how many particles came out, what their charges are, their directions, lifetimes, masses and energies and so on. often these particles are the secondary decay products of the particles you are really interested in, and so by conservation laws (charge, momentum, energy) you can say something about those original particles created in the collision.

anyway, sorry for writing a novel, but like i said, it's a big question. the reason is that it's a whole field of research. anwyay if there are any terms i used that you don't know, you can surely google them. apologies if i left out something important.
posted by sergeant sandwich at 9:26 PM on September 11, 2008 [4 favorites]

Short answer: by measuring how they interact with other particles/fields.
posted by msittig at 10:18 PM on September 11, 2008

Bonus: Is it this that forces us to use uncertainty in our current model? Does measuring a particle's location, say, change its momentum?

One of the misunderstandings people have about Uncertainty is that it is some sort of experimental limitation we discovered.

Just the opposite, it arises directly, mathematically from the formulation of quantum mechanics. Things like position/momentum or Energy/Time are non-commuting variables. These are variables which, in the definition of the wave function, are Fourier duals.

In Classical mechanics you can "extract" physical properties in any order (the coordinates in property-space commute) but the Math just doesnt work out that way in quantum mechanics. Now, what does that mean for physical reality? Thats something we've been trying to get our heads around ever since.
posted by vacapinta at 3:00 AM on September 12, 2008 [1 favorite]

« Older Why can't I disable root login on this Mac?   |   It's been on my mind: what legal say do the dead... Newer »
This thread is closed to new comments.