Robert Hofstadter

The Crystal Ball Experiment

Category: Lectures

Date: 29 June 1982

Duration: 38 min

Quality: HD MD SD

Subtitles: EN

Robert Hofstadter (1982) - The Crystal Ball Experiment

Instead of an abstract, the text below is a short introduction by an editor.This is the last of four reports on Robert Hofstadter’s post-Nobel project: construction and operation of total absorption detectors for gamma ray spectroscopy. The earlier reports were given in 1968, 1971 and 1973

Ladies and gentlemen, I don't understand German very well but I heard something in that introduction about prognostication. And the crystal ball is not a device that reads the future. It's actually an experimental device that looks into the wonderful and marvellous laws of nature and tries to find out as precisely as possible what some of those things are. I'm going to be talking about the crystal ball experiment today and this experiment involves a very large number of physicists. When it began, it involved a collaboration of about 32 physicists. And there's been a discontinuity recently which I will tell you about a little later and it's now an experiment that is being conducted by about 80 physicists. So obviously what is being studied is a complicated thing. And obviously my talk is going to be complicated in some way as well. For that reason, I decided that I would give some kind of an introduction to the students here in case they don't know what scintillations are or what sodium iodide crystals are. And anyway it gives me a great excuse to talk about sodium iodide which I invented in Princeton in 1948. And I've been working on that material in one way or another ever since. So I'm kind of wedded to it. I'm also wedded to the study of gamma rays. I've always liked gamma rays. And when I first went to Princeton I started to study means of detecting gamma rays and I worked on a device that is called a conduction counter, it was an insulator actually. And at low temperatures it became a counting device. It was an insulator at low temperatures and became a counting device there. After a little while, it turned out that there were some difficulties with using it as a counting device because it did require low temperature. Although nowadays it operates very much like a germanium lithium drifted type of detector. So in some sense, perhaps I may go back to it someday. But in the meantime, perhaps some of you know that in 1947, right after the war, a German physicist by the name of Kallman invented the scintillation counter, which used large lumps of naphthalene crystal. And that crystal enabled gamma rays to be measured in quantity and also slightly in energy. At that time if you would go around to any laboratory in the United States, it would smell of mothballs because naphthalene is the thing that makes mothballs. And all the nuclear physicists were using that material. And I tried it too. But since I had had some experience with conduction type counters I knew the literature and in particular the German literature of Hirsch and Paul who had worked on luminescence of crystals for many years and particularly on the alkali halides. And so I looked up a few of the articles that the Paul group had written and I recognised that one could make a scintillation counter of some of the alkali halides. And within a day or two I had made sodium iodide scintillate. And also caesium iodide and rubidium iodide and several other things like that. And so I will show you a little bit about that early history. And I feel that I was very lucky in finding that material because it has turned out to be enormously useful in medicine and geology and all branches of science. And I am particularly pleased that it has had so many humanitarian applications. Now, in 1950, I went to Stanford and started to work on electron scattering because there's a very fine accelerator there. But I carried along this information on sodium iodide. And I started some graduate students working on the problem of detecting high energy gamma rays because the accelerator was going to produce high energy gamma rays. And so with a graduate student by the name of Asha Kants who has disappeared from view with Asha Kants we developed a large, we made measurements that could establish how large detectors could work in totally absorbing a high energy gamma ray or a high energy electron. And this work lead eventually to theories which predict now how showers, electromagnetic showers, which is what gamma rays and electrons produce in large crystals, can be estimated and how you can estimate things like resolution and so on. So I want to tell you a few of those things before I tell you about the crystal ball because the crystal ball is made of sodium iodide. This crystal ball is a beautiful device and it merits some discussion on its own. So I would like to do that now by starting with some slides. If I can have the first slide. Now, this first slide shows the energy level diagram of positronium and I've shown this rather than the energy level diagram of hydrogen because this bears a resemblance to the mesons that I'm going to discuss, like charmonium and bottomonium and so on, obsolonium. The point is that Bohr in 1913 gave us the idea that photons could jump between energy levels. In other words we could have transitions which would give us photons. Now, the same kind of thing happens in high energy physics, there are energy levels and there are transitions between those levels or among those levels. And one can thereby find the structure of those mesons or particles or whatever they are by studying those transitions which lead back to the energy level diagram. And positronium consists of a positron and electron circulating around each other with this energy level diagram. And there's a great resemblance as I've said to much higher energy systems. In particular, I'd like to call your attention to a set of seven levels: this one, this one, this one, these three and that one up at the top. And later you will see that this level corresponds to a state in a meson discovered by Sam Ting and Burt Richter which decays into three gluons. This one decays into three gamma rays. And I'll come back to that in a little while. So this is kind of an example of what we're going to see later on. The physics is not too different in some respects. It's the same kind of physics; all you have to do is change the names. And I hope to show you something of that sort. Now let me make a little diversion to sodium iodide. May I have the next slide please? Incidentally, Willis Lamb reminded me that I talked about sodium iodide here probably in 1965 but no one else will remember. So I hope you'll excuse me if I do it again. I did it at that time in connection with tests of quantum electro dynamics which Sam Ting has improved on very greatly. But this is the original sodium iodide detector with the sodium iodide at this end of the tube, an evacuated quartz tube fastened on to a small photomultiplier tube, a very small tube. Next slide please. Here is the active material and you see the dimensions of this are very small, this is only half an inch in size. The material had to be kept inside a quartz tube because it's terrible hygroscopic. And that's the way you have to work with this material. You have to preserve it from contact with the air. Next slide please. A little bit later I grew some crystals of sodium iodide in this form. And you see they have good geometry. They're beautiful crystals and I knew that something beautiful had to come out of this. So we actually set up an apparatus which is shown at the top, not important. Jack McIntyre, a graduate student and I and we made some measurements on the Compton effect. And we verified the Klein-Nishina formula at energies in the cobalt 60 range. Next. When we prepared this nice crystal instead of that amorphous kind of crystal you saw before and used radioactive elements, in particular this one is gold 198, with a 411 kilovolt line, we immediately got this remarkable plot on the oscilloscope. This is a photo peak and this is the Compton distribution that corresponds to the absorption and scattering of 411 kilovolt gamma rays in sodium iodide. Next slide please. We also did cobalt 60 right then, just about at the same time, and observed the two photo peaks corresponding to the two peaks in cobalt 60 and the double Compton edge corresponding to those two different gamma rays. Now, we also realised that there was some escape of Compton gamma rays, the scattered gamma rays in the crystal. And if you would make a big crystal you could probably capture the second gamma ray. So some time later we made bigger crystals. This one is about a foot across, 13 inches, something like that. And in that crystal you could see the two gamma rays with a Compton edge greatly suppressed and here you even see the addition of those two. This lead to the idea that by going to large crystals, you could capture all the energy in a gamma ray or in electron or positron showering particles. Next slide please. This shows the showering phenomenon in a very crude way. Where an electron comes in and makes bremsstrahlung and you get pairs and pairs you get gamma rays and the gamma rays make pairs and so on. And if the material is big enough you can capture a very large fraction of the whole shower and thereby measure the energy of the gamma ray. And this is the work that Asha Kants and I were engaged in when I went to Stanford. Next slide please. Sometime later we had the Harshaw Chemical Company grow larger crystals for us. This one weighs 1,000 pounds and is 30 inches in diameter and almost a foot thick. This is Barry Hughes who was very glad to receive that crystal when we first got it. And that crystal has been used in subsequent experiments to test electrodynamics to a high order of accuracy. Next slide please. The experiments I am going to talk about today have been carried out at the two mile accelerator at Stanford. And this is a representation, a picture of the accelerator along here. And this is the end station area where the experiments are done. Next slide please. Here is the Spear storage ring facility and our experiments were done in this little laboratory here. This is the SSRL, synchrotron radiation project, this and this, where I and my collaborators are also carrying out some medical experiments. Next slide please. Now, after the J/Psi was discovered, we tried with some of those large crystals to detect gamma rays from the decay of the Psi and we were thwarted to some extent because the multiplicity involved in the decay was too great and too many particles would enter the large crystal at the same time. And so the first gamma ray measurement was actually made at DESY in the decay of the Psi. But we soon followed up with a system, with an experiment in which we used a large number of sodium iodide crystals around the region where positrons and electrons collide. And at the Psi prime which you'll see in just a moment, we detected some interesting things. May I have the next slide? This is an experiment done in collaboration with Princeton and John's Hopkins. And this is the gamma ray spectrum of the Psi that we found and I would say that there is no structure in that but the Psi prime which at that time was called Psi 3684. You see that there is some evidence of some kind of line structure. So we pursued that but realised that it would be good to have better modularity. Next slide please. And so we devised the idea of the crystal ball. And the crystal ball comes from originally a platonic figure called the icosahedron which has 20 equilateral triangular faces. And if you divide up those faces into four parts, you get a system like this. And then if you further divide each one of these small parts into nine parts as shown here, you end up with 720 individual modules or crystals. Note that in order to do an experiment you have to have some holes in this device. And here is such a hole and there is one on the other side and there is a tunnel in between. This is the original design of the crystal ball. Next slide please. This is a sodium iodide crystal which is placed at some distance from the intersection region where electrons and positrons collide. There is a photomultiplier which is put on the end so as to receive the light from that individual module. All modules are optically isolated from each other. Next slide please. This is the way the assembly was made at the Harshaw Chemical Company by Harshaw personnel and by our own people. And you see the individual crystals being assembled. This was not an easy job because you can think of what might happen if you put all of these in and then the last one won't go in or the last several won't go in. So these all had to be surveyed and the surveying equipment is shown there but it doesn't show up very well on the slide. May I have the next slide please? This is the completed hemisphere for the crystal ball. The next slide please. And this shows the hemisphere with a cap that will go down on it and all the holes in there for the photomultipliers. And you see the surveying equipment here that had to be used in putting that together. That was quite an elaborate job. Next slide please. And this is the way the crystal ball was put together at SLAC, at the Spear storage ring. Of course these 2 hemispheres join each other, they join together. And in here one has the beam pipe. Here are some end cap photomultipliers of sodium iodide. And in here are wire chambers and proportional chambers that allow you to identify and track charged particles and so distinguish those from gamma rays which do not leave a trace as they go through. Next slide please. Here is the actual device itself. You can see it's very complicated. This is Ian Kirkbride who has a very personal relationship with the crystal ball, he goes wherever it goes; he doesn't let it get out of his sight. And he makes sure that it's in good condition. This represents the one half and this the other half and they are joined together. Next slide please. Here they are joined together. Next slide please. This is the tunnel inside the crystal ball before the end caps are put on and this show the beam pipe going into the interior of the crystal ball. Next slide please. And this shows the end cap crystals here as they are located around, just outside the tunnel. May I have the next slide please? Now I want to show you how some events are recorded in this device. This is the so called BABA event in which E plus and E minus of high energy are scattered into E plus and E minus and each leaves a shower in the crystal ball. Of course, we have made a flat representation of all the crystals. Now, this is one electromagnetic shower and this is another electromagnetic shower. I'm sorry that the focus is not very good but I think you can get the idea. And we can identify where the centre of the shower occurred and we can also add up all the individual energies in the crystals and get totals. And for example here are two total energies. Next slide please. This one shows three Gammas recorded. Here is one Gamma, here is another Gamma and here is another Gamma. Next slide please. Here is another one, where there is Pi plus and Pi minus and three Gammas. Here, here, here, there and there. Those are the two Pis, they do not make showers the way electrons or gamma rays do. Next slide please. The previous slide had an orientation or representation of this kind; the three gamma rays here and the two pions which left tracks in the wire chambers and proportional chambers. Next slide please. Now, the resolution in energy of this crystal ball is given on this table. For caesium 137 which is at 600 and something kilovolts, 6/10 of an MeV the resolution is 20%. If you know what you can do with this material, you can get a resolution of about 5% with a good crystal, a good individual crystal. But because one has to put all these modules together and add their outputs in various ways, the resolution deteriorates so that for .667 MeV you get only 20%. For an energy of 1,842 MeV you get 4.7% for full width half maximum. Now I will make a prediction here that this kind of resolution will be increased by, will be improved by factors of 100 or 1,000 someday and in the not too distant future. And in that case we'll see incredible detail which we can only guess at now. It certainly won't be done with sodium iodide, it will be done with something, I don't know what. Next slide please. Now, this is one of the results we obtained and this curve has become rather famous I would say. If you remember the energy level diagram in positronium, you will see that in charmonium, in which this is the lowest level, I wonder if that wouldn't come out more clearly on some of the transparencies. Let me try a transparency of that. I think I have one. Yes, could I have this thing set up again here? You see the energy level diagram for charmonium, that is the J/Psi system. And you see the same set of energy levels: this one, this one, this, these three and that one. And the transitions between them are indicated and are numbered. And these spectral lines here refer to that energy level diagram. And you see that there is the same kind of splitting that occurred in positronium, here it is in charmonium where the energies are multiplied by orders of magnitude. Also shown are two little peaks here which have been investigated in great detail. Because these levels were not known, that is the theorists who do QCD theory tell us that there should be states like these, just as one sees in positronium. And there should be transitions from here to here, from here to there and from there to there. And if you identify with the numbers there, this one is number eight which corresponds to that transition, about 640 MeV, not kilovolts, MeV and this one that corresponds to this little transition here at about 90 MeV. Now, this transition right here is not shown in the Psi prime spectrum because it would occur from the Psi to that level. And so I think that the next slide shows that. I don't have a transparency of that. Next slide please. Well, before I get to that, this shows the branching ratios for the various transitions from those piece state levels. And I'd like to make a remark here that you see accuracies on the order of 9% or maybe 6% or something like that. And I've always had a strong interest in precision and 9% and 6% really bother me and I hope that someday somebody will get much better precision because you want to compare those numbers with what you can calculate from QCD. From quantum chromo-dynamics. I do not believe that quantum chromo dynamics at the present time can make a calculation that is better than about 20%. But on the other hand if we were to have these transition rates observed to within 1/2% or 1/10%, it would really make a test of QCD. And I hope that that will occur. Now the next slide please. Now, here are some other results. I'm just showing you a few results at random. The crystal ball has obtained many, many results and a lot of them are, as I said at the beginning, complicated. It would take a very long time to explain them. But the top figure is a Dalitz plot of decays of the, can we see that, yeah of the Psi. And from this you can see that the Eta meson is formed and the Eta prime meson is formed. We verified that also by working with decay processes in which we did not have a company in quantum electrodynamic transitions and you see the same kind of thing here. So these are three Gamma decays. The Eta goes into two Gammas. Next slide please. Here are the branching ratios that were observed for those transitions. The J/Psi going into three Gamma. Here is the transition into the Eta, a Gamma Eta state, Gamma Eta prime and the ratio. And these measurements early eliminated a couple of other states that had previously been found. We did not find these states and I think it's now recognised that any such states do not exist, at least within levels of that sort. Next slide please. Now, we also did make experiments of the kind Sam Ting talked about, the R measurements. And this is just a sample of our own measurements. These of course are to be collected with those of many other groups and from these one will try to tell how many quarks are produced and what total cross sections are. May I see the next slide please? Yes, I would like to just look at my transparencies for a moment. Oh, there was one small matter that I did leave out. That very small transition from the Psi to the Eta state was detected by us and that is shown here. This is the decay of the Psi and that's at about 120 MeV. That was detected and in an exclusive reaction it was detected also there. Now let me just say a few more words about some of our other findings. In both the discussion of Steve Weinberg and Sam Ting the word gluon appeared, or did it appear in your talk, Steve? Yes it did, ok. But nobody talked about combinations of gluons and theorists have predicted that there should be states in which gluons interact with each other. And they're not like photons in that respect, they're different in that respect. And so, inspired by the work of the theorists we have looked at transitions of the Psi into gluons. Possible gluons to see if there could be two gluons interacting with each other and a gamma ray emitted at the same time along with those. That is what has been predicted and in fact a three gluon type of decay is similar to the one that occurred in positronium. But you could also have two gluons and a gamma ray. And so we looked at the end of the Psi spectrum. And sure enough there was a lot of structure there. And I'm not going to go into the details of this because the arguments are rather intricate and complicated. However, as a result of making those measurements and following a very nice presentation by Don Coyne in our group, we believe that there are two possible candidates for combinations of gluons, a two gluon state in other words. And this indicates one of those possibilities, a resonance at about 1,640 MeV. In this case what we look for is a five gamma ray decay because each of these Eta mesons decays into two gamma rays. So, two and two made four and then the original gamma ray made five. And if you do that study carefully, we believe that this is an indication of a possible gluon or gluonium state. Some people call it a glue ball. And this has been called by our group the Theta state. This is another one and I won't go into details, this is called the Iota, the evidence for that is not quite as clear as for the Theta. It would be very advantageous if we had more events to look at. We ended up with about 2 million events for the Psi decay. If one could increase the luminosity of storage rings by a factor of five or ten, we would get, we would begin to get the kind of accuracy that spectroscopists would like to see. Now I'd like to go back to the slides very briefly. The one before this, please. This shows the crystal ball being moved out of SLAC. The idea was that we had done quite a bit of work at SLAC, the collaboration had done quite a bit of work at SLAC. And the Doris machine at DESY in Hamburg was going to be improved to give an increase of luminosity by a factor of ten which is just the kind of thing that I was talking about. And the energy was to be increased also so that one could make a study of the Upsilon meson which was a new meson involving the bottom quark. And so in April of this year, there's one half of the crystal ball being moved for the first time The next slide shows how we did that. This is the back end of a US air force plane, the C5 plane which is an enormous military plane. And it was our idea that we should use the military for a useful purpose so that... And the crystal ball is in there, including dehydrating equipment and also electronic equipment. Now it turned out that the C5 made a beautiful trip and made a beautiful landing at Rhein-Main base. And there was a little trouble in getting the truck off but they did get the truck off successfully. And then the truck made its way to Hamburg and broke down about half way. So the airplane trip was successful but the land trip wasn't. And so it took an extra day...


Instead of an abstract, the text below is a short introduction by an editor.

This is the last of four reports on Robert Hofstadter’s post-Nobel project: construction and operation of total absorption detectors for gamma ray spectroscopy. The earlier reports were given in 1968, 1971 and 1973. The title of this talk, “The Crystal Ball Experiment” refers to experiments performed with a detector in the shape of a ball of crystals. The “ball” is really an icosahedron consisting of more than 700 individual crystal modules enclosing a collision chamber connected to two beam pipes through which the colliding particles enter. The crystals are looked at from the outside by photomultipliers, connected to each other so that gamma rays emanating from the collisions can be detected in coincidence. During the decade since the 1973 lecture, one can say that his project had been a total success. As Hofstadter remarks, the number of physicists working on the Crystal Ball project at the very beginning was 32. It had increased to as many as 80 at the time of the lecture. The Crystal Ball was first set up at the SPEAR colliding beam facility at SLAC, where it in particular looked at transitions between energy levels of charmonium, the quark-antiquark system discovered in 1974 by the 1976 Nobel Laureates B. Richter and S.C.C. Ting. In 1968, Hofstadter used the blackboard for pedagogical explanations, but this time he has so many interesting results that the lecture becomes a slide show and his “next slide please” is repeated N times, where N is a large number. At the end of his lecture he shows pictures from the transport of the Crystal Ball detector from SLAC to the DESY laboratory in Hamburg, Germany, where it was installed at the DORIS colliding beam facility. There both intensity and energy of the beams were larger than at SPEAR and the plan was to study gamma rays from transitions in systems containing the bottom quark. As far as I am aware, the Crystal Ball is still used, but now (2012) it has moved to Mainz.

Anders Bárány