Specific Physics Question Addressed by Heavy Flavor Studies

(Compiled by Ramona Vogt)

The charge to the working groups are:
a. Identify which of the physics questions can be addressed by the class of measurements to be studied by the working group, and outline the structure of a scientific program that can successfully attack these questions and lead to convincing answers.
b. Identify the compelling questions that can be addressed with RHIC II that are unique to RHIC and cannot be addressed at the LHC. Which provide crucial complementary information to the planned LHC measurements?
c. Specify the required information, kinematic range, statistical precision, and required beam combinations to address the physics questions. What technical advances (experimental and theoretical) are required beyond the capabilities currently in place at RHIC?
Below is a list of physics topics for 1) charm, 2) bottom, 3) charmonium and 4) bottomonium. Not all the physics topics dovetail neatly with the list of physics questions that the working groups are supposed to address. In addition, there are obvious overlaps with other working groups such as forward physics and pA as well as electromagnetic probes (through quarkonium --> l+l- and semileptonic open heavy flavor decays), high p_T and even equation of state.

Given these overlaps however, it seems to me that the first physics question for the working groups regarding the nature of the phase transition, hadronization and evidence for deconfinement is the one that heavy flavors most naturally addresses, particularly in the case of quarkonium. Other questions addressed would have to do with the initial state, medium properties and thermalization.

Preliminary list of questions/comments on heavy flavor topics:

1) Charm
Hadronization is not all that well understood. The Peterson function is falling out of favor relative to other ways of modeling the e+e- data (such as those in my paper on charm and bottom in pp at RHIC with Matteo Cacciari and Paolo Nason that they fit to the e+e- data in Mellin moment space) which end up considerably reducing the effect of fragmentation on the charm p_T distribution. This smaller hadronization effect would consequently reduce the need for p_T broadening to describe the charm hadron spectra at lower energies and perhaps also its effect at higher energies although that last is less clear.

There are obviously some medium effects on charm such as the observed finite flow. How strong this is and what kind of medium it indicates certainly needs more data for clarification. Does charm flow tell us anything about the equation of state and do we need a thermalized system for charm to flow?

Initial state (before the collision -- the nuclear gluon density) and initial state (immediately after the collision -- the density of produced gluons) both
affect the charm spectrum.

For the first, it would be better to measure charm production as a function of rapidity rather than p_T, like PHENIX has done for J/psi in d+Au. Something similar would, given small enough error bars, perhaps give a better handle on the absorption cross section for J/psi -- if it's an initial state effect alone, then the J/psi and charm dA/pp ratio should be similar as a function of rapidity. To do that, it is important to measure charm away from midrapidity, the further forward the better, like extracting the charm yield through muon spectra in the PHENIX muon arms. What is the status on this? Is anyone working on it?

For the second, the calculations of medium induced energy loss depend on the density of produced gluons to predict the expected energy loss for charm. Right now they get the gluon dN/dy by fitting the R_AA for light hadrons and it would be interesting to see how the R_AA for charm compares to these predictions. The data so far look surprisingly similar to the light hadron R_AA but it's still rather preliminary.

2) Bottom
The issues are the same except the effects are all expected to be weaker because of the larger bottom mass and the higher scale.

3) J/Psi and Charmonium
In dA and pp, good baseline distributions as a function of p_T and y are needed. Here more luminosity would certainly help. Would improved efficiency do a similar job without a luminosity upgrade and, if so, is a substantial improvement in efficiency possible in any of the planned detector upgrades? We are already learning something about the initial state (before the collision) from J/psi in d+Au but more data would make a more compelling case.

It would be nice to get a handle on the absorption mechanism as well as the cross section, as I already mentioned above. One way to do that would be to see if the J/psi and the chi_c are affected the same way in dA (or pA) collisions or not. In NRQCD, the chi_c is supposed to be produced mostly in singlet states so the absorption pattern would be different for the J/psi and the chi_c. While it is probably not doable in AA, it would be nice to see if it is possible to do in dA and/or pp. We learn a lot about the production/absorption mechanism with such a measurement and that, in turn, helps us understand deconfinement better. What is the status of the chi_c at RHIC? Is anyone looking into trying to measure it in pp and/or dA?

Medium effects/evidence for deconfinement -- J/psi suppression was supposed to be one of the gold-plated pieces of evidence for deconfinement. We know it's a lot more complicated than that but we might still be able to tell the difference between suppression alone and suppression with regeneration. The charm cross section feeds into this too: if the charm cross section is really more than 1 mb at RHIC, then it would considerably increase the apparent J/psi rate in central collisions and that does not seem to be the case at the moment. The charm rate is needed also for comparison since the regeneration models tend to normalize the J/psi rate to open charm. For charm to be a normalization in this measurement, it seems necessary to get a more accurate measurement of the total cross section.

In principle, J/psi flow could be studied. We could look into how much luminosity would be required to do that. There should also be some level of energy loss for J/psi, similar to that for charm, but it hasn't really been studied. It's harder to calculate too, I would think, since calculations of the p_T dependence are rather painful and seem to require some kind of p_T broadening that makes it more difficult to sort out.

4) Upsilon
Same as charmonium for medium effects, just probably weaker. The main difference is the effect of a deconfining medium. There are a wide range of predictions on the temperature at which Upsilon production should be suppressed, all considerably higher than that for the J/psi. The Upsilon' and Upsilon'' states are another matter. It would be nice to get some handle on whether or not the Upsilon states are suppressed or not at RHIC and, if they are, by how much. This is a topic for further study, both for theory and experiment, as the discussion last week showed.


To do these various measurements for open heavy flavors, in my opinion, it would be most preferable to use reconstructed charm and bottom hadrons rather than relying on the electron and/or muon spectra alone. Is it possible to improve heavy flavor hadron reconstruction away from midrapidity? It would be a great help since at p_T > 4 GeV the electrons start to be mainly from b decays and the medium effects get blurred. The crossover from D to B dominance may also be at different p_T away from midrapidity. This is something that can be checked by calculation but is also sensitive to the respective total rates.

It would also be useful to compare charm (through the D channel) to J/psi over the same rapidity range as in PHENIX. This helps to get to more forward regions with lower partonic x to make the comparison. For b and Upsilon, I think it's hard to produce meaningful results without higher luminosity but I'd be glad to be convinced it's possible otherwise.

All of the above are interesting, reasonable measurements that would enhance our understanding of heavy flavor production in general as well as in the nuclear/QGP environment. As such, they should not be inherently political. The other two charges are more so.

As for point b, regarding the uniqueness of RHIC II, I would agree with Bill's message of a couple of weeks ago that the charge regarding RHIC II relative to LHC is not formulated very well. Especially in the case of heavy flavors, it is difficult to say that any of the measurements are unique to RHIC and cannot be done at LHC. One may argue, as Bill mentioned, that many of the J/psi's produced at the LHC will come from B decays and therefore the J/psi measurements at the LHC involve different physics. Is this really a problem though? There are plans at the LHC to distinguish between prompt and secondary J/psi's through displaced vertices, both in ALICE and CMS. (I don't know about ATLAS, sorry for any ATLAS afficianados out there.) These secondary J/psi's from B decays also have some interesting physics possibilities, perhaps helping reconstruct B decays through B --> J/psi K, but just in general contributing to the total understanding of B physics both in pp and in AA.

So, in general, it's hard to say that any of the heavy flavor measurements are unique to RHIC. The higher energy and greater luminosity at the LHC should make all this heavy flavor stuff easier thanks to the correspondingly higher rate. ALICE can reconstruct D's down to low p_T but perhaps not away from midrapidity, |y|<1. The other experiments can probably do high p_T D's but the high p_T tail of the charm distribution tells nothing about the total cross section. Some simple calculations show that the slope of the p_T distribution for heavy flavors is only sensitive to the renormalization/factorization scale, not the quark mass. CMS and ATLAS can probably do B mesons since that's part of their pp programs and since it's done to low p_T at the Tevatron, there's nothing preventing it from being done well at LHC. I'm not sure about tracking in AA for that but it should be doable in pp and p(d)A to get the baseline. (I don't think one can exclude the possibility of either p(d)A or lower energy runs at the LHC. There is considerable interest in doing pA at the LHC and the pp people will probably want a lower energy run to check back with Tevatron, like the Tevatron's 630 GeV run to compare with the CERN SppS data.)

CMS and ATLAS will have some trouble with low p_T J/psi. Without using a special trigger they probably can't go below p_T = 5 GeV at midrapidity but should have no problem at higher p_T, even at midrapidity eventually. I know CMS has looked into changing their trigger conditions and found they can go lower in p_T for the J/psi. In any case, ALICE can do J/psi to low p_T. The Upsilon is no problem for the LHC detectors, even at midrapidity, because of the larger mass and here I think the higher energy is going to favor LHC with respect to Upsilon break up scenarios. I think a complementary Upsilon measurement at RHIC would provide a crucial point of comparison relative to LHC measurements.

Thus, my feeling is that, at least for our group, RHIC II and LHC are very complementary rather than exclusive. This may not be a popular opinion, but for heavy flavors, I think it is a reasonable one. One can make arguments about relative rates due to run time etc. but it's not a question of who will win in this sort of comparison. The physics is sufficiently different at the LHC due to the different energy regime to make a compelling case for a strong heavy flavor program at both machines.

I won't comment now on the last point in the charge but I'd like to get some opinions on the issues raised here.

 Last Update:3/18/2005