L. Gregor Herten, “The Future of L3 at LHC” - LNS46 Symposium: On the Matter of Particles

HERTEN: So far, we heard the results we had with L3 at LEP. And now I will continue and explain our plans for the future, namely our plans at LHC. And first, what is LHC? So here at CERN-- this is the CERN complex showing the LEP ring and also these pre-injector systems here. And the plan is to use the lap tunnel, insert some new magnets and new beam and use this for PP collisions at 16 TeV. And this is called the Large Hadron Collider, or LHC.

Now, why are we actually interested to go to LHC with the L3 detector? And of course, as you know, so far the Standard Model is actually doing very well. And in fact, the tests we are doing at the moment with L3 show, for instance, that there are three families. And also the tests show that these zero couplings are in agreement with the Standard Model to a very high precision.

Clearly, there are two missing points. And the first is the top quark. And the second is the Higgs boson. The top quark, presently, we know should be larger than something like 90 GeV. And I think everybody agrees, or most people agree, that the top should exist. And if it's below 150 GeV, it will be found very soon, hopefully, at Fermilab. If it's heavier than 150, probably it will first be found at LHC or SSC.

Now, the next part is actually somewhat more interesting. Namely, it's the Higgs boson. And this is, the Higgs mechanism is responsible to generate masses for W's and Z's. And we hope with L3, at the time when LHC comes up, that we will be able to explore Higgs masses again, up to something like 90 GeV. So therefore, the interest is to explore masses beyond this 90 GeV range.

And of course, you can think now, for instance, the Higgs mechanism in the Standard Model is not correct. For instance, there could be some more complicated Higgs mechanism. So in other words, you have more than just one Higgs doublet, which is, for instance, predicted by supersymmetry. And then you should also find many more new supersymmetric particles. Or there is no Higgs mechanism at all. And then current theories predict there should be other types of particles around, in the mass range above 100 GeV.

Or maybe the Higgs mechanism is correct. And clearly, then, we also will look for some new types of interaction, so new types of vector bosons, Z prime and W prime. So in any case, I think the solution to the major problems and questions we have in the Standard Model at the moment lie in this range around 1 TeV. And therefore, they are accessible at LHC, which has a center of mass energy of 16 TeV.

Now for us, clearly, the most important question is the experimental task, the experimental challenge. And this can best be seen here in this diagram where we see the cross-section in proton-proton collision as function of the center of mass energy. And for instance, see, we have the CERN collider, Fermilab, LHC, and SSC. So the first observation is that, indeed, the interesting processes like Higgs productions, Z prime production, have cross-sections which are extremely small, only about 1 picobar.

So that means we have to go to colliders which have an extremely large luminosity. But on the other hand, you see also that the real background, namely the total cross section, is something like 10 orders of magnitude larger. So if you have a high luminosity machine, you have a huge amount of background produced all the time. So therefore, we need to build detectors which are able to cope with this kind of background to extract signals, let's say 1 event out of 10 to the 10 background events. And this is really the experimental challenge.

And so what we are planning, then, is the following, to perform a specialized experiment at LHC. And namely, this is done after the completion of the LEP program, and also the LEP 200 program. And we intend to use much of the existing L3 detector, and also the existing infrastructure we have. And clearly, we have to modify the detectors. And this is based on the results we are having then, and we are doing at the moment, in this R&D efforts.

Now, when I say we are planning to do this, then, this is the existing L3 collaboration. And here, just to give you the names of all the institutes involved-- and these are 38 institutes of 14 countries. And also, in addition recently, a lot of other groups have expressed interest to continue this work with L3 at LHC. And so recently, we have an addition for this project, 22 institutes from seven countries, which are listed here, which want to explore also the pp collision with the modified L3 detector.

OK. So how can we modify L3 for proton collisions? And this is, in fact, how L3 looks like, or will look like at the time when the LEP 200 program is started. So the changes, then, that we will have, also the forward muon system, which is indicated here. So our plan then, is starting from this detector to, at the moment, we are having two designs, two possible designs.

Namely, after the completion of the LEP program, there will be one year to modify the detectors and to install new detectors. And then there's one possibility to go to a specialize muon electron detector this way, which then, in the first phase, which means a minimum modification of L3 will allow us to explore luminosities up to 10 to 33. And then in the next step, in a second modification, we will be able to explore very high luminosities, the highest luminosities at LHC, of 2 times 10 to 34.

The other possibility we are thinking of is to go to a muon electron and photon detector. This requires some more modification in the beginning. But then we will also be able, right from the start, to really explore the highest luminosities. And then in the second phase, we will install a very precise crystal calorimeter as a photon detector. And this will allow us, especially, to explore this Higgs decay into two photons. And of course, we will be able to run at the highest luminosities.

OK. So let me first now go this line here, namely, the specialized muon electron detector, and explain the plans we have at the moment. And this is now how this detector would look like, where we see here the existing L3 detector in blue, the existing muon chambers, and these are the muon chambers for LEP 200. Then we will add, in the very forward direction, the magnetized [INAUDIBLE] system with proportional chambers here to extend the coverage for muons in the forward direction. Then we will insert a superconducting coil at this point, which provides a magnetic field of six Tesla. And we will replace the calorimeter inside, and also the inner tracker inside, to be able to run at very high luminosities.

Now in the other view, this detector looks as follows, where you see here this typical L3 structure with the return yoke, coil, muon system. And then here, the superconducting coil, calorimeter, and inner tracker. So let me explain a little bit the idea behind this. And first, the magnetics, the field system. Where now this superconducting coil provides this field in the inner region, which is 6 Tesla, the field return, in the field return, we also use this forward [INAUDIBLE] system. And then the field returns through this muon chamber volume. And here we have a field of something like 0.5 Tesla.

Inside of this high magnetic field region of 6 Tesla, we have then a new calorimeter system, which is shown here, where you can distinguish three parts. The central part indicated here, which has an absorber of tungsten with silicon readout in the front, and the gas readout in the back, with some copper plates. Then the forward region here, we have a copper absorber with gas chamber readout. And in the very forward region, in the end cap region, again, because the particle density is very high, we have tungsten with some gas readout.

And also you can see here the new inner tracker system, indicated in red, where we have several layers of inner trackers, which are used in this high magnetic field to measure the momentum of particle with very high precision. So the total thickness of this new calorimeter, the instrumented thickness, is 6 to 7 lambda. And together with the superconducting coil, the total filtering thickness is larger than 9 lambda, everywhere here in this region.

Now let me explain a little more in detail the inner tracker system. As I already mentioned, there are several layers of these new trackers. And each layer consists of straws, namely five layers of straws. And each straw has the diameter of 4 millimeters, with a precision of 120 microns. Outside of the calorimeter, we have a somewhat larger tubes, which have 5 millimeters times 10 millimeters cross-section. And here, the precision is 160 micron. And as you see here, from the expectation that the occupancy per tube and per bunch is of the order of a few percent in all of these layers.

Here, just indicate again how many channels we need for this system, where for each tracker here, it's indicated the amount of electronic readout channels you need to find tracks in the system. And in total, we need something like 100,000, a little bit more than 100,000 readout channels. And also what's indicated here in the forward region, we have these multi-wire proportional chambers to find tracks in a very forward region where the particle density is very high.

Now, what is the basic idea with this detector? Why do we have this large magnetic field? And this can best be shown in this picture where we see a high momentum muon, 500 GeV, overlaid with 20 minimum bias events. And that is, in fact, what you would see every 50 nanoseconds. At LHC, you would have such a picture.

And you see already with all these events here, it's very difficult to isolate the real interesting track. But clearly, a 6 Tesla field helps a lot because all these backgrounds have very low pT. And therefore, they curl up in this slot magnetic field such that, at this point where you have your tracker-- for instance, here's a tracker layer, here's another one-- the particle density actually from these background tracks is very small. And therefore, this allows you to identify this track of the muon track very cleanly.

And this can be seen a little bit more quantitatively in these diagrams, where we have the occupancy in different trackers versus the magnetic field. For instance, here is a tracker at 5.5 meter radius, 0.7, 0.8, and 1.3 meter radius. And you see if you, for instance, have 0 magnetic field, you have a very high occupancy, 20%. But then going down here with increasing magnetic field, you see at 6 Tesla, you reduce your occupancy quite a lot. Here, it's 70%-- 70 centimeters. And you see the further you are, you will have a very small occupancy. And you clearly can see the effect of the large magnetic field.

So the muon system in this design, the muon detector, the present L3 muon detector provides very good momentum resolution, up to something like 200 GeV. And also, this muon detector serves to identify muons and to find the track such that we can trace back the track to the inner system, to the inner trackers. And then the inner tracker will provide a very good momentum resolution, especially for tracks larger than 200 GeV.

And here now to summarize the resolution for muons, I have the resolution, momentum resolution, S function of the polar angle theta. And you see that for tracks below something like 100 GeV, the resolution is typically in the range of 2%. And even at very high momentum of 500 GeV, the resolution stays around 5%.

Now, if you take all the parameters of this proposed detector, including the efficiency to identify electrons, muons, and so on, I have here a simulation for Higgs events. And we're talking here about 150 GeV Higgs boson and the decay of, into Z0, Z0, go into electron and muon, together with all possible backgrounds. And also what's included in this simulation are 80 minimum bias events overlaid with each Higgs event.

And clearly you see this very clean Higgs peak in all these modes of e-e-, mu-mu- for muon, and also for electron. And in each case, the precision-- let's say the mass resolution-- is of the order of 1% in all these cases. So therefore, clearly, we'll be able to see the Higgs if it happens to be in that mass range. But also, for instance, if the Higgs is very heavy-- 600 GeV, for instance-- again, this is a simulation, the same decay mode with all background included. And in one year, we would expect to see a signal of a 600 GeV Higgs of this type.

And just to show the momentum resolution of this detector for muons, for instance, here we have a Z prime, extremely heavy of 4 TeV, decaying into mu plus, mu minus. And that would be what we would observe, where you have here the natural width. And with a muon detector, we can reconstruct this e-prime with a very high precision. And we obtain, then, a mass resolution of 7% for such a 4 TeV object.

So now let me come to the next line, the next design we have at the moment. We are thinking about, namely, a specialized muon, electron, and photon detector. And the design considerations now are somewhat different. Because again, we have to build a detector which is able to cope with this huge background we have at LHC. And now we keep a low field, a low magnetic field. But then we go to a very large radius to reduce the occupancy at the detector. And also, we have a crystal calorimeter, especially for high precision photon detection.

Now this detector, then, looks as follows, where you see, again, the L3 components here-- the magnet, the coil. And also we keep the outer muon chamber layer, here indicated in blue. But then the modifications we apply are as follows, that we move some muon chambers to the outside, to the outside of the magnet system, indicated here.

And also, we insert two layers of resistive plate chambers, which I use now to identify muons outside of the iron. And then they allow us to trace back the muon to the inner system, where then with the inner tracker system we are able to measure the momentum of these muons. And also we, then we'll have a hadron calorimeter at very large radius, this crystal calorimeter, and a new inner tracker system.

Now let me start from the inside to explain this detector. Here is the new tracker system we will have in this detector, where you see layers of trackers at 140 centimeter radius and 280 centimeter radius. And these trackers are built out of proportional chambers and drift tubes. And also in the forward region, where we have proportional chambers between 22 and 40 degrees, and gas microstrip chambers in the very forward region. And the total thickness of this tracker is smaller than 2% of x0.

So why do we need these different technologies? And this can be seen here where in the barrel region-- so this is theta-- so the barrel region, we see here the particle density per unit area per second. So in the central tracker, we have particle densities up to 10 to the 5. And then this density increases sharply in the forward region, such that here we can use drift tubes and proportional chambers. Here we use proportional chambers. And in the very forward region, we use microstrip gas chambers, which have a very high, can have a very high rate.

Now the system, then, has the following resolution, where we have here as function of theta, again, the momentum resolution of this tracker system. And we see that for 100 GeV tracks, the resolution is typically around 3%, from 10 to 90 degrees. And even at 500 GeV, will still see that the resolution is of the order of 12% or so, which then is used to, for instance, measure the momentum of higher energy muons.

Now let me come to this crystal calorimeter, which is shown here, is like a barrel, which consists of crystals. And we will have 100,000 crystals arranged in such a barrel structure with a radius of 3 meters. Now, which crystal can we use? For instance, one candidate which is under discussion at the moment is Cerium fluoride.

And to compare now Cerium fluoride with the BGO crystals we have presently, you'll see that the density is slightly smaller, 6. Radiation length is a little bit larger, not much larger. But what is important now for LHC is the fact that the decay time for this crystal is much smaller. It's only 20 nanosecond compared to 300 for BGO. And also what is interesting that the temperature dependence is very much smaller than BGO, which makes it much simpler to perform precise calibrations.

So here, just to list a few properties of Cerium fluoride, it has very good mechanical properties. It has an emission in the visible region. I said there is nearly no temperature dependence. And also, it has a very good radiation hardness. So this is certainly a very good candidate. And the question, of course, in crystal calorimeters is-- always it was exactly the same with BGO-- how do you get all these crystals you need? And clearly, at the moment, if you would buy it, you had to pay a huge amount of money.

But we have these very good connections to the Shanghai Institute of Ceramics. Because also the BGO detector has been built there. And they are willing to produce such a crystal of 50 cubic meter for 70 million Swiss francs. And already they have a lot of the infrastructure available because of the construction of the BGO calorimeter. And also, one should note that China is, in fact, the first producer of Cerium oxide and other rare earth materials. So therefore, it's clearly a very good possibility for us to get all the crystals we need.

And now just let me go a little bit into detail where I just point out the resolution we can obtain with this detector. And clearly, the resolution is the crucial point here. That's the reason why we go to crystal calorimeters. And here, this formula gives you the resolution sigma over E. I don't want to go into detail to explain all this. So this is just the intrinsic part for Cerium fluoride. Then there are other important points which we have to control, namely, for instance, noise, pile up, and of course, the calibration.

Now noise, for instance, you need to keep the electronic noise low. You need to go to a low capacitance readout. And that's the reason why we prefer a low magnetic field. And therefore, we can use [INAUDIBLE], which give fairly low noise. And to control this pile up problem, we go to large radius. Because the pile up scales like 1 over R-squared. And that's the reason why we place these crystals far away from the vertex.

And there's actually another very good reason to go to a large radius, and this is seen here. The vertex uncertainty at LHC is typically 5.7 centimeters. And such, if you are looking for the Higgs goes to 2 photon, you have a natural uncertainty in the Higgs mass because you don't know really the vertex of the event. And therefore, if you see this here, this uncertainty coming from the vertex uncertainty in the Higgs mass as function of the detector radius, you see clearly this big improvement if you go to a 3 meter radius compared to, let's say, to a 1 meter radius. And clearly what we need to see the Higgs in this mode, we need something which is of the order of 0.5%. And that's the reason why we place this detector at 3 meter radius.

Now as I mentioned already that we are really interested in this Higgs goes to gamma gamma mode, and with this detector then, with this calorimeter, we would observe in one year the following signal where you see here this really enormous background and this signal here of the Higgs. And clearly from this, you already see what you really need is an extremely high performance calorimeter. If you have a calorimeter with poor resolution, you will really see nothing here.

And now performing a background subtraction, you see these kind of peaks for various Higgs masses between 80 and 160 GeV. And in each case, you have typically something like 12, even 20, 18-- at very low masses, something like 7-- sigmas. And this corresponds to this kind of integrated luminosity. And this channel is actually very important because if the Higgs happens to be in that mass range of 80 to 130, let's say, this will be the only way to find it.

All right. So if the Higgs happens to be heavier, clearly this detector is also able to explore these other decay modes. And again here, we have the decay mode in Z0, Z0, into electron or muon. And again, we have electron muon for electron and for muon mode. And also with this detector, the mass resolution is of the order of 1%, even for electrons, in this case, slightly better. But again, if the Higgs is heavier than this, for instance, again, the 600 GeV, also with this detector, then, we will clearly see in one year a signal for the Higgs. And of course, here now, this Higgs is very broad. And this is really dominated by the natural width of this 600 GeV Higgs boson.

So in fact, you see that with this detector, we will be able to cover the complete mass range of the Higgs, from 90 GeV up to 600 and higher. And therefore, I think even when I'm 46, probably there's a very good chance that we actually know where the Higgs mechanism is there or whether it's just some mathematical trick. But of course, also, we will find something else if it's there, for instance, the Z prime of 4 TeV again. And now with the electron mode, we clearly can reconstruct the Z prime like with the natural width there. You see here this very good agreement with the natural width curve.

So therefore to conclude, what we are planning is to build a detector for LHC. And the spirit is really that we want to upgrade the existing detector to use most of the facilities we have. And this upgrade should be simple enough such that we can implement it in a very short time-- it's in one year or so-- so we are ready when LHC is ready. And also, so we can do this with moderate additional resources. And clearly the designs emphasize running at the highest LHC luminosities. OK. Thank you.