Here in this section we talk about how one goes about 'seeing' different particles using the ATLAS detector.  By this I don't really mean we actually see little tiny particles zipping about – cool as that might be – but more along the lines of 'seeing' the way a digital camera sees.  If you think about it, it's the camera's CCD (charge-coupled device) which actually interacts with the collected photons from the dog on the other end of the lens (or whatever the subject of your photo is).  Of course your eyes too, in a sense, behave much like a digital camera, only they don't store the image (in the same sense).  There are tons of things a digital camera can do but our eyes can't.  And vice versa.

Also in the case of a digital camera one can then look at the image after the fact, even multiple times.  One could even come up with algorithms encoded in computer programs which can tell us a whole ton about the image even with no one looking at it, simply by looking for correlations between the millions of individual pixels.  

So remember, it's the camera that captured the photons and turned their energy into those little pixelated bits of information in such an arrangement that allows to recognize the familiar form of a dog on the other side.


In the case of ATLAS, sure, it can also 'see' photons, but it can see a whole lot more than that!  At the end of the day, the main way it does its seeing is still by capturing photons (well, technically electrons, since it measures electric currents).  But many types of fundamental particles (the guys you've learned about way back in Part I!) can 'speak' with the ATLAS detector, the same way photons 'speak' to your digital camera.  Often it's done indirectly, but they still ultimately transmit information in some way – information that physicists can later use to put pieces of the puzzle back together.

Before we get going with particles, let's talk about what parts makes up the detector:

Overview of the Components of ATLAS

If you break the detector apart into different key layers, or sub-detectors, then there are really three main categories to consider.  These are summarized below.  Note that the innermost and outermost regions of the detector make use of magnetic fields in the context of what we just talked about in the last section.  And remember that this is really in 3D of course: try to picture particles being created at the very centre of the detector and flying radially outwards.

Maybe doesn't sound like much, but get this: the entire detector is close to 45 m long by 25 m in diameter.  Its mass is close to 7000 metric tonnes (!).  It was built underground in the huge cavern where it sits today.  Its scale really is mind-blowing if you see it in person, which is actually possible during shutdown periods.  (Side note: if you're in the Geneva area anyway, why not get a guided tour of CERN?  You can get more info here if you're interested).

Also, if you've got an extra six or seven minutes to spare right now (+ particularly if you're a Ravel fan) check out a time-lapse video made from sequential photos ( that redundant?) taken over the course of the assembly of ATLAS here.

Interactions of different types of particles in the various sub-Detectors:

The title pretty much says it all!  Here's a colourful table that captures the gist of it:

Here the coloured portions in the table mean the indicated particle interacts with that particular component, whereas black means it doesn't interact.  There are a few special cases: muons only leave a little bit of energy in the calorimeters (they don't interact with the material as much), so the corresponding entry is lightened to highlight that fact.  Also see the triangular box associated with the photons?  That's there because photons can often decay into an electron-positron pair, and when that happens those particles do leave tracks in the inner detector.  It's not the actual photons themselves per se, but their decay products, so it still (kind of) counts.  Hence the triangle.

We'll now go through each of these individually, starting with the electron.

Here we're talking about either electrons or their anti-matter counterparts (which you'll remember are called positrons).  The latter is implied as well whenever I talk about electrons.  They interact very slightly differently from one another (particularly once they slow down), but we group them together here.  Since electrons/positrons are electrically charged, they leave tracks in the inner region of the detector.  If they're electrons they curve one way; if they're positrons they curve the other way. 

Electrons rarely make it past the first layer of the calorimeter system.  Ultimately they leave a shower of energy in the calorimeters, which is read out as a signal.  That shower of energy is initiated by the electrons bumping into particles in the calorimeter material (e.g. lead, copper, tungsten, and a splash of liquid argon), thereby giving off photons, the photons subsequently decaying to electron-positron pairs, and so on and so forth.  But that's as far as they make it.

Next up: photons.

We already mentioned these guys in the summary table above, but in the calorimeter photons behave very similarly to electrons.  The biggest difference is that they don't leave a track in the inner region of the detector (since they're neutral).  Again, with the proviso that they can of course first decay to electron-positron pairs, and these would then leave tracks. 

So even though it's not depicted above you could envision a set of tracks that begin partway through the inner detector region from when that conversion happens.

What about muons?

Muons are somewhat strange.  In a sense they're kind of like the heavier cousin of the electron.  But in another sense they behave entirely differently, at least from our perspective.

For reasons beyond the scope of this introduction, they travel right through the calorimeter modules.  On their way there, they do of course leave tracks in the inner region of the detector (after all, they're electrically charged).  In the calorimeters they also do actually ionize some atoms and thereby deposit some of their energy there for us to measure, but it's typically only a tiny fraction of their total energy.  Then they just sail on through.  Luckily (again, by design) we have something put in place to try to identify them on the other side – the muon spectrometer system.  Notice I didn't say 'to catch them on the other side'.  That was on purpose: we don't actually catch them!  They fully escape the confines of the detector!  They're actually the only guys who make it out that far (well, other than neutrinos which you'll hear about shortly).  But crucially in the muon spectrometer we have a different configuration of huge toroid magnets (here toroid just refers to the shape) and these steer the muons along a curved path.  When we reconstruct a track out there, the amount of curve indirectly tells us the momentum of the muon, kind of like in the centre of the detector. But out here we know we're only dealing with muons.

What happens with a given muon next?  It actually keeps going!  But that's ok, it's given us the info we need and it won't actually get too much further anyway.

Another interesting fact: there are things called cosmic ray muons.  These are bombarding the Earth's atmosphere all the time.  These actually represent a (small) source of what's called background in our experiment, because they're real muons, but they're not coming from the proton-proton collisions.  They fly through the detector, but from all sorts of angles (as long as they're coming from above).    We can actually require tracks in both the inner detector and muon spectrometer and make sure the tracks point more or less to the centre of the detector – that helps us disregard the other cases so they don't lead us down the wrong path.  Remember that ATLAS isn't that deep underground (only ~ 100 m or so).  There are other types of experiments that are really affected by these cosmic ray muons.  So much so that they have to go deeper – much much deeper – underground, typically in abandoned mine shafts (~ a few km's underground!), so that they're less affected by the background.  More on this later.

Quarks and gluons are next, and we group them together here.

Ok here it gets messy...  Basically nature doesn't let quarks fly around on their own.  They always have to be in what are called colour-neutral states.  What does that mean?  Well, they have a type of 'charge' called colour charge.  It's kind of like electric charges (+) and (-), except there are three types, and we call them – somewhat misleadingly – red, green, and blue.  All of the composite particles we know about which are made up of quarks (called hadrons) have no net colour.  There are also anti-colours too, just to confuse you.  But basically saying 'no net colour' translates to red+green+blue (that gives neutral overall) for example, or even red + anti-red (that also gives neutral overall).

Here are a few examples:

But notice that we don't have red+red+green (just as an example). That wouldn't balance out colour-wise.

This is real science by the way (in case you were starting to question it).

So you're wondering: does a red up quark behave differently from a blue up quark?  Well, yes and no.  On its own (if it could be on its own) there would be no inherent difference.  But when you start having interactions between many different particles, the so-called colour flow can actually impact the final results.  Again, here you just let nature do its thing and take care of the accounting.  But remember how we talked about the fundamental forces of nature early on?   We talked about gravity and electromagnetism (which you knew about already) and also mentioned the less-familiar strong and weak nuclear forces.  Well, colour charge is to the strong force what electric charge is to electromagnetism.  The rules are very different for those forces, but the idea is the same.

You probably guessed it, but the colour we're talking about here is not the colour you're used to.  It's just a name, plus we can think of it adding together in the same way that coloured light does.  But that's where the analogy stops – the quarks aren't actually red, blue and green!  ...unless it helps you to picture them that way, in which case, go right ahead!

Why do we say the strong force is so strong, in fact?  Or why do we give it that name?  Well, if you have an electron and a proton close to each other and they're not moving too quickly, they'll naturally form a simple hydrogen atom.  It will take a certain amount of force or energy to pull them apart from one another.  Similarly, rather than an electron and a proton, two quarks in a bound state will be held together by the strong force.  But try to pull those apart and the force grows exponentially as the distance increases.  Eventually you're putting so much energy into pulling the two quarks apart that it's energetically favourable for nature to let them split apart and create another quark-antiquark pair out of thin air such that you get two bound states where you initially had only one. 

Crazy, right? 

There's so much force binding those quarks together as you try to pull them apart, that you're better off appeasing nature by giving them each a new partner.  Making matter (the new particles) out of energy (from your pulling them apart).  That way everybody's happy: colour 'neutrality' is preserved and the quarks stay in bound states – sure you had to roll up your sleeves and make two new particles out of energy, but nature prefers things that way.

So, just remember that try as you might (as far as we've been able to measure!) you can't ever get a quark in isolation!

You can think about the strong force another way:

Atomic nuclei are incredibly tiny.  They're also made up of composite particles which are either electrically neutral (neutrons) or positive (protons).  But like charges repel each other, right?   So what's holding all of those protons in the nucleus together?  Why doesn't the nucleus just explode from all of the positive charge squeezed into such a small region with no negative charge to balance it out?  Sure, they neutrons are neutral so they don't repel one another, but there can be a huge number of protons in a very small space for the heavier types of atoms on the periodic table and something, somehow, is keeping them from flying apart.  What is it?

The answer: the strong nuclear force.

For an atomic nucleus, the strong nuclear force and the electromagnetic force are competing with each other in a way.  The strong force is (as its name implies) much stronger, but can only work on small distances scales.  If you try to pack more and more neutrons and protons onto a nucleus, eventually the electromagnetic force will triumph and the nucleus will split apart.

Trade-off between the electromagnetic and strong nuclear forces: an atomic nucleus can only accommodate certain numbers of neutrons and protons – adding any more will eventually cause it to become unstable and decay.


What this means in the context of ATLAS is that if we manage to have a process that occurs in a proton-proton collision resulting in a quark being produced, it won't travel any distance really at all before additional quarks and gluons start popping up out of empty space to make sure no quarks are ever left on their own (i.e. in some state with a net overall charge).  The gluons can split apart into quark-antiquark pairs too.

So like I said, it's messy.

What starts as a single charm quark (just for example, as depicted in the cartoon below) ultimately ends with what's called a parton shower.  The word ‘parton’ is just a fancy, collective term to denote either a quark or a gluon.  Now the cartoon is just to convey the general idea.  I leave it to physicists to scrutinize the details (are there any of you reading this?).   Ok, here's the kicker: we actually don't have a detailed mechanism for exactly how parton showers actually work on a step-by-step basis.  If you're used to the ideas of quantum physics, it might not seem that strange to you anyway that we can't really talk about the actual meanderings of a single particle itself – we can talk about probabilities and bulk properties, sure, and we have several models which describe the data extremely well, but the theory is not worked out exactly.  But that's ok!

At the end of the day, you have a spray of hadrons (composite particles made up of quarks) and gluons, as well as their subsequent decay products and those are what fly through the various components of the ATLAS detector, not the quarks or gluons themselves. So you typically will see many tracks (not just one!), some calorimeter energy deposits, and (in the case that there are muons in any decays) muons in the final muon spectrometer.  

I neglected to include muons in the cartoon below, but just to highlight that they're not nearly as energetic as the muons we mentioned above as a separate category.  Might be some neutrinos in there too.  Again, omitted from the cartoon below.

A complete mess, right?  But again, that's ok in the end.  We can reconstruct objects we call jets, which are an attempt to put all of this mess together, and which are then treated as a single, reconstructed object.  They are the closest thing we can piece together and which act as a probe to the initial gluon or quark we're really interested in.

I said that the objects we reconstruct are called jets and that those provide a good probe for the original quark or gluon of interest.  We can think of jets as being formed from energy deposits in the calorimeter, or from a bunch of tracks in the inner detector.  Or maybe a mix of both just to confuse you!

But if we focus on making jets out of energy deposits in the calorimeter, let me just walk you through the basic idea of how we actually go about reconstructing a jet – from beginning to end.  On a superficial level of course.

Let's use a top quark for our example.  A top quark can be produced in the centre of the ATLAS detector from time to time.  We showed in Part II how a pair of top quarks can be produced.  Regardless of how that happens, let's say we've managed to produce a top quark somehow and it's sitting right in front of us.  What happens to it next?  Well, it never sits around for long.  It actually decays faster than we could measure with any clock!  So it always decays incrediblly quickly, and when it decays a certain fraction of the time it does so into three other quarks, and those guys each produce a shower or spray of energy ultimately leading to what we'll reconstruct as a jet.  So where do we start?  What is it we're actually putting together?  How do we reassemble the top quark we're really interested in?  It started as one single particle, decayed into three 'lighter' quarks, and these in turn each resulted in huge sprays of energy and particles.  What a mess!

Well, ultimately those final, more stable, shower particles travel through the material in the calorimeters where they ionize the material (liquid argon, for example).  Ionize here means they liberate electrons by giving them a kick of energy and releasing them from their nucleus. 

What then?  How could we collect those electrons?  Well, we can apply a strong electric field (which you read about earlier!) and this pulls them towards a postively charged plate (the orange +V bar in the cartoon below).  The more energy deposited, the more electrons are released.  So if we 'count' electrons (or have a machine do it for us), we indirectly get a sense of how much energy was deposited in the detector.  It might help to remember that a steady stream of electrons is really what we call electric current! (again, it might help, but you might in fact just be wondering when this bloody section will end...we're getting there).  So we measure electric currents and that tells us how much energy we had.

The tiniest or most fundamental unit in a calorimeter is something we call a cell.  Each cell can have some energy associated with it (if there were indeed sufficient electrons collected there).  We can then group together these cells in a clever way to form clusters.  The clusters we then group into individual jets (almost done...), and, finally, we put three jets together (right-most picture below) to make the now long-gone top quark we set out to reconstruct.  We put the three jets together because each of those jets is associated with one of the light quarks which the top quark decayed into.

Here's a picture of what I mean, from start to finish:


Then what?

Ok here's what's cool: we can ask for how much energy and momentum that collection of jets has.  Tons and tons of individual cells with small amounts of energy.  And guess what?  When we add it all up it should be related to the original energy and momentum of the original top quark!  How do we know this?  From theoretical predictions for one thing, but also from from years and years of measurements from actual data!

In some senses this might be intuitive.  If that's the case, great!  I mean, after all, you started with a beautiful vase; you dropped it on the ground; it shattered into pieces; and you then tried (in vain) to glue the pieces all together again before your parents got home.  It's kind of along those lines, but here it's not just 'pieces' as it was with the vase.  It's a mix – some 'pieces' (particles), sure, but also some energy. But that amounts to the same thing, remember?  Matter? Energy?  As Einstein told us, we don't really care what form it happens to be in.  That said, in terms of putting it together, it seems easier for our minds to visualize putting 'stuff' back together.

But the bottom line is we can ultimately get at the properties of quarks or gluons we're interested in measuring by putting all the Lego pieces back together to make jets and using those jets as a probe to the original quarks or gluons.

Again, quarks and gluons are the most complicated types of particles to explain here, but tons of the rarer, more massive particles we try to produce at the LHC do end up decaying to quarks, so we really do need to understand them.  What's more, some of the more exotic, so-called 'Beyond the Standard Model' theories predict the existence of as-of-yet unobserved particles which can also decay to quarks, so again: unavoidable.  They play a crucial role in what we're trying to do.

Neutrinos are next (and last!) on the list.

In some ways neutrinos (whose name means 'little neutral one') are the easiest to talk about here. 

And in some ways they're the most complicated.

Easy, because they essentially don't interact at all with the detector material, so we don't measure them.  That makes the cartoon depicting their passage through the various components super simple:

They're also complicated for the very same reason: they don't interact!  Neutrinos are, like all fundamental particles as far as we know, truly point particles.  They have no physical size, but they have a tiny but non-zero mass – much smaller than that of an electron.  They travel at nearly the speed of light.  But they can carry away huge amounts of energy, in true ghost-like fashion, from centre of the detector where they were originally produced, and that causes problems in terms of the energy accounting.

To see just how little they interact, have a look at this example below.  We use a photon – a particle of light – to illustrate the first concept:

That's a ridiculously large distance!  But ok, a solid light-year of lead is hard to wrap our mind around...

There's another way of looking at how little neutrinos interact with the matter we see around us.  From current models (and confirmed by experiment!) every second there are hundreds of trillions of neutrinos passing through your body.  Yes, your body.  Right now and every second.  Hide underground, stick a tinfoil hat on – absolutely no difference.  They also fly right through the Earth more or less.  After all, it's only ~ 12,000 km in diameter (way less than a light-year) and moreover far less dense than lead!  Of all those neutrinos passing through your body, if you waited a full century (!), you could expect one single bloody neutrino (on average) to interact with the matter in your body.  No, I didn't just make that number up.

Instead of using just our bodies as detectors, we could go about things in a different way.  If you had two people, you'd only have to wait 50 years; 10 years if you had ten people.  You get the idea.  Actually let's forget using people as the targets (probably a good idea) and just use a huge volume of something else for our detector.  Ice, say.  Think that sounds crazy?

Ok, we're getting side-tracked here (remember, the topic is neutrinos), but there's actually an experiment called the Ice Cube Experiment which is just that: a cube of ice.  But not just any cube of ice like the ones you put in your Glenfiddich (the connoisseurs gasp...I'm kidding, relax).  Seriously though, it's an actual cubic kilometer (!) of ice.  That means 1 km x 1 km x 1 km.  The experiment is situated underground at Earth's South Pole.  I mean literally at the South Pole.  For comparison, a cubic kilometer of water would be enough to fill a million swimming pools.  Still not easy to imagine a million swimming pools, but you get the point: it's an absolutely enormous volume of ice used as a detector.  Light-sensitive devices were lowered down vertical shafts drilled into the ice at evenly spaced intervals such that there's then an array of these devices (kind of like a 3D grid) to capture hints of interesting 'events' when a lucky neutrino actually interacts with the material in the ice.  

The Ice Cube experiment has a sister (but non-affiliated as far as I know) experiment called Antares on the floor of the Mediterranean Sea, and which operates on the same basic principle.  I won't say more about the experiments (the websites do a good job) but what you can take away from it is that you now know why neutrino experiments typically have to be so huge!  At the end of the day instead of waiting years for an interaction, Ice Cube physicists for example can observe roughly 300 interactions per day!

Another neutrino experiment deep in an abandoned mine shaft in Canada called SNO (Sudbury Neutrino Observatory) contained about 400,000 L of heavy water and ultimately confirmed the solar model as well as proved that neutrinos have a small but finite mass.

Yet other neutrino experiments make use of all of the high-energy neutrinos spewed out from nuclear fission reactors.  'Free neutrinos' in a sense, right?  Making use of these neutrinos is a fantastic way to test our models of neutrino interactions.  We can place detectors at various distances from the reactors and see how the measured data agree with the theory.

Finally a recently proposed experiment called Dune (which was fairly recently approved and will soon be under construction!) will use a high-energy neutrino beam which is blasted through the Earth (to a position about 1000 km away!) and into enormous targets consisting primarily of liquid argon (you can crudely think of liquid argon as a much denser, much much colder version of water).  The targets are buried in abandoned mine shafts somewhere in South Dakota.

I've mentioned abandoned mine shafts twice now, so let me explain.  Let's switch tracks for a second and talk about muons again.  The reason will be clear shortly.  Then we'll come back to neutrinos and be done.

Pretend you have some detector to measure the rates of cosmic muons coming from the upper atmosphere.  For some weird reason you decide to put the detector at a depth of 1 km below the surface of the ocean.  This would be a bad idea.  Why?  Many of the muons wouldn't make it to that depth!  But sure, whatever amount does make it that far, let's call it 100%.  It's a reference point.  Let's then go even deeper and see what happens.

If you were to then move your detector to 2 km below the surface, you'd get 10 times fewer muons compared to what you originally had.  Bump that up to 10 km, and you'd get a million times fewer muons.  It's an exponential drop off!

The numbers I'm quoting I took from the summary on Cosmic Rays (from the section on 'Astrophysics and Cosmology') available here.

Through the eyes of someone doing neutrino experiments on the other hand, this is an exciting concept!  They don't want to see muons!  In fact, they're an enormous source of what's called background to what they really care about.  So whereas muon rates drop at lower depths, neutrino interaction rates stay essentially the same.  What's a few measly kilometers of water – or solid earth, or lead for that matter – to a neutrino?  Nothing!

Of course neutrino experiments aren't typically underwater (except that one I talked about on the Mediterranean floor).  They're mostly underearth.  The advantage of placing a detector a few kilometers underground is that it shields you from a good fraction of unwanted muons, while leaving neutrino interaction rates mostly unaffected!

The detectors for many experiments involving neutrinos need to be deep below the surface of the Earth (typically at least a few kilometers). Abandoned mines offer the perfect setting.   Again, why?  Since neutrinos interact so rarely with a detector material, if you were to have your detector at the Earth's surface there would be tons of other reactions that would take place.  These would be too enormous a background; they would swamp or drown out your signal.  If, on the other hand, you were to put your detector underground, there would be no negligible difference in the rate of signal that you're looking for, but background events would be cut back substantially.

Getting back to ATLAS, just remember this: if there's a high-energy neutrino produced in a given collision, we don't see it – it doesn't interact with the detector – so it's effectively invisible to us.  But if there's only one, we can actually infer its presence.  I mean we 'see' it by not seeing it, so to speak.  That might sound strange, but wait until you've read the following section and you can re-read this if it still isn't clear.


So we're done!  Now we know more or less what to expect in terms of how each of our main particles interact in the detector.  Next up we can talk about looking for a few things at once – trying to identify a whole event.