• John
  • Felde
  • University of Maryland
  • USA

Latest Posts

  • James
  • Doherty
  • Open University
  • United Kingdom

Latest Posts

  • Andrea
  • Signori
  • Nikhef
  • Netherlands

Latest Posts

  • CERN
  • Geneva
  • Switzerland

Latest Posts

  • Aidan
  • Randle-Conde
  • Université Libre de Bruxelles
  • Belgium

Latest Posts

  • Vancouver, BC
  • Canada

Latest Posts

  • Laura
  • Gladstone
  • MIT
  • USA

Latest Posts

  • Steven
  • Goldfarb
  • University of Michigan

Latest Posts

  • Fermilab
  • Batavia, IL
  • USA

Latest Posts

  • Seth
  • Zenz
  • Imperial College London
  • UK

Latest Posts

  • Nhan
  • Tran
  • Fermilab
  • USA

Latest Posts

  • Alex
  • Millar
  • University of Melbourne
  • Australia

Latest Posts

  • Ken
  • Bloom
  • USA

Latest Posts

Christine Nattrass | USLHC | USA

View Blog | Read Bio

The Quark Gluon Plasma

Much of the press coverage of the LHC has discussed the search for the Higgs, but ALICE was designed for something completely different.  We’re studying a hot, dense phase of nuclear matter called the Quark Gluon Plasma (QGP).

What we’re doing is mapping out the phase diagram of nuclear matter.  This is roughly what we know about the phase diagram of nuclear matter:

The x-axis, baryochemical potential, is the amount of energy needed to add a baryon (such as a proton or a neutron) to the system.  You can think of it as analogous to pressure in the phase diagram of water.   The y-axis is the temperature in MeV – one MeV is roughly 1160 megakelvin (MK) and 175 MeV is about 2,000,000MK.  That’s about 250,000 times hotter than the core of the sun.  At lower temperatures and lower baryochemical potentials (lower densities), we have a hadron gas.  (Neutrons and protons are hadrons.  The hadron gas is actually mostly made up of pions, a hadron with about 10% of the mass of the proton.)  At higher temperatures, we get the Quark Gluon Plasma.  Neutron stars are somewhere to the right, way off the chart.

We create a QGP in the lab by colliding two nuclei together at high energies.  The LHC will collide lead nuclei at a center of mass energy of 5.5 TeV per nucleon, which is over 27 times greater than the highest high energy nuclear collisions so far.  (Nucleons are the protons and neutrons in the nucleus.)  We expect our first data from lead-lead collisions this fall, if everything goes according to plan.  From studies of gold-gold collisions at the Relativistic Heavy Ion Collider at Brookhaven National Laboratory on Long Island, we learned that a QGP is a liquid of quarks and gluons – and set record for the highest temperature reached in the laboratory in the process.  The fact that it acts like a fluid actually indicates the attraction between quarks and gluons isn’t negligible, as had been predicted – this is why it’s sometimes called a strongly-interacting QGP, or sQGP.  (At one point someone proposed a variation on this, the bound state QGP, but this name was thankfully abandoned.)  We’re actually not really sure what we’ll see at the LHC – will it still behave like a liquid?  Or will it behave like a gas?

What we’re doing is kind of like colliding two ice cubes together to study the phase diagram of water.  If you had two ice cubes in outer space (where it’s really cold) and you got them moving really fast and slammed them into each other, the ice would melt for a very short time and then as the system expanded, it’d freeze again and you’d end up with little flakes of ice scattering all over the place.  On the surface of the Earth, nuclei are essentially frozen nuclear matter.  We have to figure out the properties of melted nuclear matter by looking at the frozen shrapnel.  We have the luxury of having better ways to study water than by smashing ice cubes together, but we don’t have any other way of studying nuclear matter.

Everyone learns that water has three phases, solid, liquid and gas.  This is a pretty good approximation, but it is a simplification.  There are 15 different phases of ice (that we know about).  Analogously, the diagram above is definitely an approximation.  There’s probably a lot of structure in there – a lot of extra phases – we don’t know about.  In the era of the LHC, we’ll be able to move over about two orders of magnitude in temperature along this phase diagram because both the LHC and RHIC are actively taking data.  There have also been a lot of improvements in detector technology and data analysis techniques in the last couple decades, so we’ll be able to do better measurements than ever before.  We’re in the golden age of the field.

ps –  I haven’t forgotten about the questions from my first post on Facebook but I just haven’t gotten to them all.  I did post replies to the comments posted on the blog in the comments section.

  • TimG

    Interesting post… I’m glad someone is giving some blogging attention to the other goals of the LHC besides finding the Higgs.

    I understand that the temperature achieved is a function of the speed of the colliding nuclei. But what determines the baryochemical potential? I take it it’s not just E=mc^2. You mention density, but if it’s a function of density then what determines the volume? I mean, the colliding nuclei aren’t bound within a finite volume like neutrons trapped in the gravity well of a neutron star, are they?

    Thanks in advance for any further explanation you can provide.

  • Cedric

    ” The y-axis is the temperature in MeV – one MeV is roughly 1160 megakelvin (MK) and 175 MeV is about 200MK.”

    175 MeV is 2e12 K …

    “The LHC will collide lead nuclei at a center of mass energy of 5.5 TeV”

    I really think it is not 5.5 TeV … But [3.5|5|7] Z TeV with Z = 82 for fully stripped lead ions …

  • TimG

    While I’m asking questions… is there a straightforward explanation of why lead ions will be used as opposed to some other element? Thanks again.

  • Cedric

    Ok … 2.76*2 ~ 5.5 TeV per nucleon and in center of mass at maximum rigidity.

  • Christine Nattrass

    Cedric – you’re right on the temperature, that was some last minute edits gone wrong. One should not do math before coffee. And yes I am using the energy per nucleon, the standard way for reporting collision energies in heavy ion physics. I added “per nucleon,” lest anyone get confused. Thanks for reading closely.
    Tim – this is a really good question.

    Let me first clarify I’m not saying that the baryochemical potential is the density – they’re related but in a non-trivial way. You could plot the phase diagram of water with temperature on one axis and pressure on the other. Or you could plot temperature on one axis and volume on the other. Or entropy on one axis and pressure on the other. We can choose to plot a phase diagram with all sorts of different variables. We could choose to plot the phase diagram of nuclear matter with the density on the x-axis.

    The mathematical definition of the baryochemical potential is the partial derivative of the internal energy with respect to the number of baryons, holding everything else – including the volume – constant. It’s roughly the amount of energy you need to add a baryon. But that doesn’t mean you can do an experiment where you hold the volume constant, and our system does indeed expand. Everyone has heard of density so that’s somewhere to start, and generally the baryochemical potential (at a constant volume and temperature) increases with increasing density. The baryochemical potential depends on the density and the temperature and many other variables.

  • Christine Nattrass

    Tim – another good question. The reason for using lead is more historical and technical, to take advantage of local expertise.

    We want heavy ions – something large enough that we can talk about collective effects of particles. Temperature, for instance, is a measure of the average kinetic energy of particles in a system. If you have three particles, it doesn’t make sense to talk about a temperature. It’s not clear exactly what is large enough – copper ions were collided at RHIC and there were some differences between copper and gold, but those differences were fairly small. But when we’ve compared results for gold ions and lead ions, we haven’t seen any significant differences. So beyond being “heavy,” it doesn’t matter for us – we can get the same physics out.

    But for making and accelerating the beam, it does matter. The whole process starts with a source that sputters out lead ions. You then have to accelerate the ions slightly, strip some more of their electrons off them, accelerate them more, strip off the rest, and then bring them up to full speed. You strip electrons by shooting the beam through a thin foil, and you actually get a distribution of electrons – so maybe you want the lead ions with 30 electrons left at a particular stage, but you’ll get electrons with 20-40 electrons left. You then have to select the ions with the charge you want, and in some stages of acceleration you’ll use a magnet to bend the beam. The beam will bend by a certain amount, depending on the charge to mass ratio. The number of electrons stripped by a particular foil is different for different ions, and the optimal charge to mass ratio is usually different for different ions. To switch the type of ions, you may have to get a new magnet or use a different foil and this may even require some substantial construction because all of this can change the radius of curvature of the beam.

    So switching beams is not trivial. Given that there’s no physics motivation, it’s not worth the work. What we’re left with is lead, because CERN has long used lead beams and is has a lot of expertise and the facilities to make lead beams well. Brookhaven uses gold beams, because Brookhaven has long used gold beams…

    Of course CERN has used other species, such as carbon and silicon, for heavy ion collisions. For light ions, there actually are a lot of differences between different beam species. In principle, the LHC could probably collide carbon ions fairly easily, but there’s just no reason we’d want to do that at this point in time.

  • Christine: nice discussion of the LHC heavy ion programme.
    Tim: Another fact about lead 208, the isotope used in the LHC, is that it is the heaviest stable isotope of any element. It’s preferable not to have unstable stuff around. In fact, I believe the isotopically pure lead used in small quantities in the CERN ion source is more expensive than gold so it wasn’t an economy measure ;).

    The LHC could well accelerate argon ions in a few years’ time since they will likely be set up for another experiment. Before that, there is also interest in hybrid collisions of protons with lead nuclei, which pose some interesting questions for the accelerator. RHIC already collided deuterons with gold but the problems were quite different since they have independent rings of magnets for the two beams. The LHC has its famous two-in-one magnet design.

  • Jonathan Clift

    Thanks for the interesting post.

    How do you know that the plasma is/was there? (Sorry if this is a really stupid question, but won’t it then just expand into whatever it is you get smashing protons together {the hadron gas?} before it then forms particles that you can detect?) And, is this a quantum Thing (sorry for the imprecise language, but ‘thing’ was the best I could manage – maybe Boojum would suit better?), like a very very complicated Feynman diagram, or is it ‘real’? Talking about it appearing to have liquid-like properties certainly makes it sound like a real thing (though I’ve read enough of these blogs to know it’s dangerous to push an analogy too far).

  • TimG

    Thanks for the answers, Christine (and you too John)! I’m looking forward to hearing more about the ALICE experiment in your future blog posts.

  • Christine Nattrass

    Jonathan – these are excellent questions.

    The question of how we know the QGP is there is really complicated and actually would be a good subject for an entire blog post. Or perhaps several blog posts. For now I’m not going to answer it now because I don’t think I could do it justice in a few sentences, but I will come back the issue and devote at least one and possibly more posts specifically to the question.

    Yes, it is a quantum system. Everything obeys quantum mechanics. Everything also obeys quantum electrodynamics and quantum chromodynamics (which is where Feynman diagrams come from). We actually do see many effects which are directly quantum effects. For instance, you can see correlations between particles from a common source due to quantum effects (see the wiki on Hanbury Brown and Twiss effect) and we have observed these correlations in heavy ion collisions. We use quantum electrodynamics and quantum chromodynamics to calculate many things see. It’s actually not so simple to say there’s a really complicated Feynman diagram to describe the system because even for a simple process you have to consider multiple Feynman diagrams. In practice we can’t always do calculations to describe our measurements in quantum chromodynamics and quantum electrodynamics – not because they’re wrong but because the calculations are too hard. So then we use other models for how things interact.

    We do have considerable evidence that a QGP behaves like a liquid. This also not an easy question to answer quickly, but let me point you to a couple other resources:
    1. http://www.sciencedaily.com/releases/2010/02/100215101014.htm
    2. The BrookhavenLab channel on youtube has quite a few nice videos and “RHIC: The Perfect Liquid” is particularly relevant
    If you search “The perfect liquid” you’ll find more about this. Briefly, (relativistic) hydrodynamical models work to describe heavy ion collisions. The calculations are not trivial. There’s enough material to write a blog post on this, but until there are data from the LHC, it’d be more RHIC centered, so I think I won’t do a post on that because this is an LHC blog.

    Excellent questions – keep them coming!

  • Jasper

    Actually neutron stars are actually quite close to the deconfinement phase transition. You can also see this because at moderate densities in the stars there are neutrons. While maybe for the highest densities (in maximum mass neutron stars) near the core of the star the quarks in the neutrons become deconfined. Thus neutron stars might actually contain the deconfinement transition. A more accurate guess for the phase diagram can be found at: http://www.gsi.de/fair/experiments/CBM/Phasendiagram.jpg .

  • Christine Nattrass

    Jasper – Thanks for your valuable insight on neutron stars. I was thinking of the core of massive neutron stars, where several exotic forms of nuclear matter have been proposed, rather than typical neutron stars. However, this was not the main emphasis of my post.

    The GSI picture is definitely much prettier. However, I think the differences exemplify the quantitative uncertainty in the field. Our best estimates for key features of the phase diagram have been evolving rapidly. For instance, Figure 4 of http://arxiv.org/pdf/hep-lat/0701002v1 shows different theoretical calculations for the location of the critical point (if one exists), and just in the last 10 years calculations for this point have covered most of the range of my sketch above. Not all calculations are equal, but this is far from settled. The hypothetical critical point on the GSI sketch is far from any of the calculations. The location of a critical point (if one exists) would pin down the location of the phase boundary, and without knowing where this is, we really don’t know where the phase boundary is. So I would strongly disagree that the GSI figure is more accurate. The structure of the the phase diagram is still far from certain – which is exactly why this is such an exciting and interesting field!