Robert Lourie, “Probing the Nucleus with Electrons” - LNS46 Symposium: On the Matter of Particles

Search transcript...

LOURIE: OK, it's certainly a great pleasure to be back here at MIT. I was here for exactly one decade. Four years as an undergraduate, working with Bill, four years as a graduate student working with Bill, and two years as a post-doc working with Bill. Somebody said to me, yesterday, I think it was Bob Jaffee, that was an awfully long sentence. But I didn't consider it that way. It has always been a pleasure.

And I think the remark that Fred made about Bill, I think, probably characterizes the experience better than I could. He said that, talking about the spectrometer, Bill wanted this particular design. People told him he couldn't have it. Bill got it anyway. And it worked out just fine. And I think that was a good lesson that I learned from how to practice in the world of physics from Bill.

I want to talk about some things that have been done, some things that just are being done, and some focus on where we're going with electrons both around the world but mainly with the emphasis on Bates and CEBAF. And I'll break it up into two sections, one dealing with nuclei proper and the other dealing with the nucleon itself, both of which are receiving extensive attention.

The first part, one might subtitle as, particle physics meets the many body problem. And there are certain issues that I'll go into where the nucleon itself, which is sort of a particle physics object, meets directly aspects of the many body problem. And in particular this is the question of nucleon structure and nuclei. If those nucleons are modified it is certainly because of the presence of the many body environment. I'll say something about the phenomenon of color transparency and correlations.

And then I want to talk about three aspects of low energy QCD that are receiving a lot of attention. These are the electric form factor of the neutron, the excitation of the nucleon to the delta, and the strange quark content of the nucleon. I'll just very briefly flash this as Bill laid out a lot of this groundwork. What makes the whole new generation of studies possible, of course, is the technical advancement of the new facilities. The single most important being the continuous beams, but there's many others. High intensity, access to internal targets with tens of milliamps of circulating current, polarization observables, and both large acceptance and out of plane detection of reaction products.

So let's start with nucleons and nuclei. First question one might ask is, are they in there? And the answer is, To some extent, sure. Because one sees the incoherent scattering from nearly three constituents, which tells us that the nucleus is composed of nucleons. The next question one might ask then is, are they the same as the nucleons you get when you buy a bottle of hydrogen? And the answer is no at some level. And I want to show four results that relate to these two questions.

The first, are they in there, we saw some of this Heppell data. But one does indeed see quite nicely in the unseparated cross-section, throughout the periodic table from carbon to nickel to lead, this structure. That is the Doppler broadened quasi free peak early Fermi gas calculations by, at the time young, [INAUDIBLE]. And everything appears here to be under control in the sense that at that time, this is data from the early 70s. You see the quasi elastic structure. It has the right quantitative strength. The nucleus is composed of quasi free nucleons.

However once these data were separated into their longitudinal and transverse components, one sees that there's many ways to represent this problem. One way is to represent it in terms of the so-called y scaling variable, which, in a first approximation, is the nucleons longitudinal momentum in the initial state. And when you divide out the appropriate nucleonic response, one would expect a universal response. That is, the data should scale as a function of this single variable. And there's several momentum transfers plotted here. One sees that they indeed do scale in terms of that variable. But the deviation from our most simple idea, which would say that these two response functions scale equally, and one sees that they do not.

MODERATOR: The television people would like you to stand in the limelight more.

LOURIE: Oh. Ah, TV. So there's one problem in our understanding that may be related to the question of nucleons and nuclei. I will say a little more about it. Let me just quickly remind you of the first EMC effect, the ratio of the response function in a heavy system to deuterium. And in particular one sees throughout this intermediate x region, where x is now the longitudinal momentum carried by the struck quark, this degradation. And remembering what the structure functions are, the quark momentum distributions, this degradation at intermediate x says that there are fewer quarks with those higher momentum values, meaning longer range correlations.

So in some sense there are many mechanisms proposed, of course, to account for the EMC effect. But they all share the feature that they have looked at the quarks, have longer correlation lengths inside the nucleus. And in some sense that certainly is a many body modification of the structure of the nucleon and the nucleus. Another place where that might be relevant-- return again to the Coulomb sum rule, which as Dirk indicated, measures essentially the charge of the nucleus and a two proton correlation term which one expects to at least get smaller if not go to zero as the momentum transfer increases.

And I'll remind you of two results both at Bates and Saclay. Very detailed studies were done on the few body systems. In tritium one sees that it goes to the value z equals 1 quite nicely. And on helium 3 and 4 it's either converged or nearly converged to the value z equals 2. When one goes to heavy systems, essentially all heavy systems with the possible exception of uranium indicate this quite sizable suppression that Dirk alluded to of the longitudinal response. And it's much larger than can be realistically expected from the effects of correlations.

So one suggestion that was made, and that sort of gets back to the EMC, is that maybe the longer correlation length manifests itself as truly an increased radius. This is the swollen nucleons hypothesis. These things get bigger. You construct this sum rule by dividing out the free proton form factor. So if the nucleon was larger in the medium, the form factor would be dropping faster, and you would make therefore an error in dividing that out. That's a simple minded picture.

One question, then, is are these two effects-- are EMC and this deviation from our simple ideas in the quasi elastic region related? That's certainly not a question that I'm prepared to answer or I think has been answered. Let me just show you kind of a list of the things that people have thought about in response to these questions. That the nucleus exhibits an increased radius and magnetic moment in the medium by various mechanisms.

Renormalized vector meson masses and vacuum polarization involving vector meson in a sort of vector meson dominance picture has been popular. The idea of pion excess and nuclei, that there are more pions per nucleon in a nucleus because in addition to the loops that connect the nucleon to itself, one now has the possibility of meson exchange. Quark exchange between two nucleons as they pass by-- it's certainly an allowed process for them to exchange quarks, which gives them on the average a larger length scale for those quarks to wander around.

And that's just sort of the one example of the general class of topics, partial deconfinement, where the quarks are free to wander from nucleon to nucleon, or at least amongst neighboring nucleons, giving you the so-called 6 quart bags, to the extreme limit where they would be free to wander through the entire nucleus. That is, it would percolate.

How might you attempt to look for this? Well there's been two-- there's been several. I will talk about two experimental attempts. One is to measure the momentum transfer dependence of the proton knockout reaction, which if we believe that we are interacting with a nearly free proton inside the nucleus, then the momentum transfer dependence should be that of the free nucleon cross-section. And the other approach is to measure the longitudinal transverse character of that same cross-section where you know you're only knocking out a single proton and going to a bound state of the residual system.

Both these types of experiments have been done at Metcalf and here at Bates. It's easier to cover a wider range of momentum transfer just measuring the cross-section. And one sees here, for carbon knockout, that, within the arrow bars, both the knock out for the P shell of carbon and the S shell of carbon over a pretty wide range of momentum transfer, almost up to a GeV squared, where the nucleon form factor has changed by more than an order of magnitude over this range. To within about the 15% level these data are flat. We've divided out-- plotted as a spectroscopic factor. But what that means is we've divided out the free nucleon response.

So what these data are telling you are, at the level of their errors, the response versus momentum transfer of a proton in carbon looks like the free one. Can then put limits on how much the radius could change on the basis of these data, and it's about 15%. And certainly one hopes-- one expects that with the 100% duty factor and more intensity, these arrow bars will over the next several years be brought down considerably.

The other way to look at this is to look at this so-called this ratio, RG, which for a free nucleon turns out to just be the ratio of GmP over GeP, which is the magnetic moment of the proton, 2.79. And that's indicated by the horizontal line here. This has been measured at Metcalf and at one point at Bates. And one sees that, over a more modest momentum transfer range, a slight deviation for these several nuclei in shells from the free space value. But I indicate that once one does real DWIA calculations of this, and puts in the proton distortion and some differential effects that are totally conventional, those effects are consistent with these data.

There is-- I think I end that section with just echoing Dirk's remark. This question of, at intermediate energies of a longitudinal transverse response, the suppression of the Coulomb sum rule remains an important open question on the one hand. And we certainly haven't seen in the same energy region any smoking gun signature for nucleon modification in the medium, the although we've looked.

Nucleons, when we knock them out of the nucleus. Well, the kind of things I was just talking about were related to initial state properties. The final state may also be quite interesting. In particular, there's been a prediction now that's been around for quite a while, the phenomenon of color transparency. It's a prediction of perturbative QCD, that there's a lack of absorption in Hadron nucleus interactions at high q squared.

You basically need three things to happen for this to be true. You need to produce small objects at high q squared. Basically the spatial extent of the object being roughly 1 over q. Then those small objects have to have small cross-sections. Basically because they will have a small colored dipole moment. And then those objects have to get out of the nucleus before they expand. Because if you, say, knock out a proton where initially the three quarks were quite close together, it's going to blow up and by the time it's reached its asymptotic state-- your detector. So if those conditions are met, one might expect to see a dramatic effect in nuclei.

Let me just show you what the effect might be. Here's the picture of what the q square dependence is based on this cartoon. At high q squared it moves further out before it expands. The red circle is of course the nucleus. So you would expect increased transparency as you go to higher q squared. While at fixed q squared you would expect more transparency for a small nucleus than a big one. And that general feature of course is borne out by all the more detailed calculations, of which I show one here. You see the decreasing transparency as a gets bigger and the increasing transparency as the momentum gets higher.

Now this would be-- unambiguous detection of this would certainly be of great interest. There are several experimental initiatives. There is data, somewhat controversial to explain at the moment, on P2P at Brookhaven, which I won't talk about at all beyond that. There is ee prime p data from SLAC experiment Any 18 that is currently under analysis. The data has all been taken and we should know an answer on that before too awful long. And there are proposals for more of this at SLAC going to higher momentum transfer. Any 18 went to about 7 GeV squared. They want to double that. And both ed prime p and also ed prime p with polarization at somewhat lower momentum transfers at CEBAF.

Let me just quickly show you some preliminary results, although not the q square dependence of the SLAC experiment. But to someone who has done coincidence experiments at a 1% duty, the fact that they get these data at 10 to the minus four duty, I think, is really quite impressive. You see missing energy spectra on carbonate at [INAUDIBLE] squared, the P and S shell structures are cleanly seen, no background beneath the nucleon threshold.

And probably the most impressive picture is gold at seven [INAUDIBLE] squared. Even under those extreme conditions, low cross-section, exceeding low duty factor, they obviously still get a very nice, clean signal out of there. And we certainly look forward to the results of that experiment. At CEBAF-- let me show you one projected result. There's the standard. One wants to look at carbon and oxygen and so forth. And the standard picture showing no Glauber treatment of the final state interactions, which are essentially energy independent, and then the kind of size of effect you'd expect from color transparency.

The interesting thing about this one proposal is that, in addition to exploiting high resolution-- even in the [INAUDIBLE] set up, you can even separate the P3 halves and P one half shell in oxygen, it will also attempt to measure the component of proton polarization in the final state normal to the reaction plane. In plane wave impulse approximation, that is if the proton had no further interactions with the nucleus, this component is identically zero. It's like the fifth response function that Bill Turchinetz alluded to.

An observable that is rigorously zero in the absence of final state interactions is therefore a quite sensitive probe of those final state interactions. And it will be quite interesting to measure both the cross-section and the momentum transfer dependence of this induced polarization simultaneously.

There was a crucial assumption that goes into essentially all of the previous slide, in particular the picture and also the color transparency prediction. That is that the scattering results from an interaction with a one body current. That is that all the energy and momentum of the photon is absorbed on a single nucleon. There's been a series of experiments at Bates, at Saclay, and NIKHEF that reveal the limitations of this concept.

One sees for example interaction with two nucleon pairs, that is a deuteron like object, in the ground states of helium 3 and 4. One sees an additional component to the transverse response function that's associated with the two nucleon threshold even when you sit right on top of the quasi elastic peak. That is, at x equals 1. And, to my mind anyway, these are the most dramatic data that indicate this effect. It was the thesis work of Paul [INAUDIBLE].

This is the missing energy spectrum on carbon. The longitudinal response, the 1p, and the 1s structure, and then zero out here. And just a quick glance shows you how qualitatively different the transverse response function is. Excess strength out to the deepest missing energies that were measured. Obviously, still the s structure present. But there's clearly something present, that's at the moment not understood quantitatively, in the transverse response. It's definitely not there in the longitudinal. If you look at the difference between the two it rises up from two nucleon threshold. That's where that excess strength starts increasing and that's subsequently been confirmed by NIKHEF measurements on a couple of nuclei.

This strength out in the continuum seems to grow and grow, as others experiment's at Bates have revealed. In particular, the thesis work of Larry Weinstein and John Morrison. As one looks at the strength in the missing energy continuum versus q squared, it starts to rise up. It may become flat. It's hard to say on the basis of three points, but it comes up from numbers around 30% to numbers of order 40%. Now that's 40% of the cross-section in the deep missing energy continuum in processes that are presently not understood. And so one has to be quite careful about interpreting missing energy spectra in terms of just one or a body of responses.

And the questions that the initiatives will address here. Bill also had those up. I phrase them just a little bit differently. How does the hadronic many body system absorb the energy and momentum of a virtual photon? What is the role of multi-nucleon currents? And what are the role of subnucleonic degrees of freedom with the ultimate goal of finding the limits to a conventional meson baryon description?

Now I want to switch from nuclei to a nucleus, the nucleus of hydrogen, and talk about things that go beyond the constituent quark model. If the proton were really just U and D quarks in a spherically symmetric SU(6) wave function, there are three things that would be rigorously 0. One is the electric form factor of the neutron. The second are the quadrupole components of the N to delta transition. And the third, essentially by definition, is the strange quark content of the nucleon. So those are three things, if they're nonzero, require us to go beyond this.

Let's talk about the neutron electric form factor first. Its derivative at low q squared is the neutron charge radius squared. That charge radius is measured quite accurately, scattering the thermal neutrons from atomic electrons, to have this number. It's negative, meaning there's more negative charge at big R in the neutron. It's positive in the center. And for comparison I show the proton radius here and you take the square root. The neutron radius is not small. 16% of that of the proton. And the neutron form factor is not being non-zero, telling you how the deviations from symmetry in those quark wave functions, it's a sensitive probe quark configuration.

It's also difficult to measure because it is small. In the cross-section it occurs quadratically, and some with the large magnetic form factor squared. So what one would want as observable is that a linear ln GeN, and that means going to polarization studies. There are several prospects and techniques for attacking GeN, a couple of which have already been developed and used at Bates in initial what I call proof of principle studies.

These are the scattering of polarized electrons from polarized helium 3 in two different groups with two different targets and the neutron spin transfer that Dirk mentioned. Working at relatively low Q squared in these initial experiments that developed the technology and showed yes, indeed, these things can be done this way. Under development are polarized solid state targets, ND3 at Virginia and Deuteronatomic beam source for blast at Wisconsin.

The next generation of experiments will try to cover Q squareds up to about 0.6 GeV squared at Bates and maybe 2 GeV squared at CEBAF, with uncertainties of order 10% to 15%. Let me show you what the current situation is. All data. Now here are the two measurements from Bates on the polarized helium 3 and the neutron recoil polarization, which is preliminary. Only point I want to make here is that both these measurements were very statistics limited. The techniques work. They've been demonstrated. And one hopes to achieve the kind of error bars-- one really wants to see 10% to 15%. It basically-- a little more current, a little more beam time. But there's no question that these methods are feasible.

And what's on the table for the next round of experiments would give error bars like that, based on a couple of different models. But here are two more Bates points using the south hall ring, CEBAF points using the recoil polarization, and CEBAF points using the polarized deuterium target. And once this program is complete, it may be the end of-- late this century, will have made a tremendous difference in our knowledge of the neutron form factor.

They also will be attacked with polarized internal targets and the blast detector that Bill showed, both polarized helium 3 and polarized deuterium. And I show the projected error bars and the results of several either parametrizations of the neutron form factor or the case when it's zero. And in all cases one sees large sensitivity in the measured asymmetries to GeN. And it's certainly important, because one has to go to a nucleus to get the neutron target to do this recoil neutron detection. Polarized helium 3, polarized deuterium will all give us confidence that what we're seeing is GeN and not maybe nuclear physics.

Let's turn to the end of delta transition. In the quark, SU(6) quark model, this is the spin flip of a single S 1/2 quark. So you have-- given you a spin 3/2 state. Just a single quark spin flip, which would mean the transition's pure magnetic dipole. But for example, the color hyperfine interaction that results from a 1 gluon exchange potential, among other things, can give D state components in the nucleon and delta wave function, very much in analogy to the way the tensor force gives D state components in the deuteron wave function. And that would allow nonzero Coulomb at electric quadrupole amplitudes.

Having D state in there, that there might be a significant D state in there, was suggested a long time ago by Glashow to resolve some problems in the simple quark model had, we'll say, calculating the axial vector coupling constant. The F to D ratio. And more recently it's been suggested that these D state components are relevant for the role of quark angular momentum contributions to the spin structure functions.

There are the exact same remarks about quadrupole components apply as I made for the neutron. They're sensitive. They're difficult. And you need some observables linear in them to get at them. There are-- Dirk showed one picture of the e2. The e2 is actually known a little less, a little worse, than the c2. Here you see that the existing data don't even really agree on the sign. Whereas the Coulomb quadrupole, although with large error bars, at least seems to favor a value around maybe minus 5% to minus 7%. These are applied as ratio to the dipole.

There are several initiatives to look at N to delta, both at Bates, using polarized electrons and recoil proton polarization and polarized electrons and out of plane detection of the protons. These will be quite complimentary studies. And at CEBAF, of course, to extend to higher momentum transfers. The same technique with polarized protons and also polarized proton targets in class and unpolarized measurements covering a very wide range of momentum transfers.

And when all the CEBAF experiments are done, one might hope to go from this to cases like that, where the larger points results from the recoil polarization. And those are the same class points that Dirk showed. How am I doing time?

MODERATOR: You're on time.

LOURIE: Oh. Well let me-- I got two more slides. I'm going to say something about parity. It's the lab-- the last-- I won't maybe motivate it so much. The strange quark content of the nucleon. We expect that there are some strange quarks in there on the basis of pion nucleon sigma, the polarized EMC, and simple cartoons like so, where you can have Ss bar components mixed in in a quark gluon picture. Or even in a more conventional picture, you can have proton fluctuations into lambda K. All of these things can give you a strange quark component to the nucleon.

One way to access it is by parity violating electron scattering on the proton. Just use the Z0 as a probe of the hadronic structure. And the important point is that there's a contribution to the electric part that's seen by the Z that comes from strange quarks. And similarly for the magnetic. By combining the photon and the Z0, you make the strange quark contribution explicit. It shows up explicitly in the asymmetry. So you measure that asymmetry. There's many initiatives to measure that asymmetry.

Let me just say what the initiatives are and then go to my last slide. The initiative's based on two facts. One is that the strange quark content has unknown low Q squared properties, that means radius and contribution to the magnetic moment, and it also has an unknown Q square dependence. The sample experiment at Bates is looking at very backward angles at low Q squared, the goal being to go after the strangeness contribution to the magnetic moment, mu strange.

Hall A at CEBAF will use forward angle scattering, trying to make very precise measurements using the existing high resolution spectrometers, over this momentum transfer range. There is a proposed new detector for Hall C at CEBAF that could go to low Q squared at forward angles and also then be useful at backward angles to try to separate the electric and magnetic strangeness contributions. And finally, there is a proposal being prepared for N station A at SLAC that would use high energy electrons and forward angle to look at this.

My last slide is just-- if I can actually push every parity violating, electron scattering proposal or experiment that's currently underway. They actually all fit on one slide. The green curve is the standard model calculation with no strange quarks. And the black curve in each case is the fit of Bob Jaffee, probably extended to Q squareds where it has no business being extended to. We see sample over here. We see CF of the Hall C forward angle measurements here, the Hall A measurements here, CEBAF backward angle measurements here, and finally at the highest momentum transfers, the SLAC measurements, and in each case the deviation from the appropriate curve.

The final remark I want to make about this picture, in addition. This is a decade of very nice physics if it gets done. We learn a tremendous amount. New window on the nucleon with the weak neutral current. And all these experiments aim for reasonable precision. And I know that at least one of the co-spokesmen on every one of these experiments has either got his thesis at MIT LNS or was a post doc at LNS. And I think that carries on the spirit of the laboratory in a fine way to take it to this new probe of the nucleon, and to have every experiment coming from someone who had an association with the laboratory at one time. Thank you.