MC-TH-95/04
May 1995
On Z and Z + iei Production in Heavy Ion Collisions
V.Kartvelishvili
Department of Physics and Astronomy,
Schuster Laboratory, University of Manchester,
Manchester M13 9PL, U.K.
R.Kvatadze and R.Shanidze
Joint Institute of Nuclear Research,
Duhna, Moscow Region, RU-141980, Russia
Abstract
Z+jet production in heavy ion collisions at the LHC is proposed as a possible probe of the
properties of the dense hadronic matter. It is shown that the accuracy of this measurement
with general purpose LHC detectors and under reahstic experimental conditions can be
high enough. It is argued also that Z-boson production and subsequent leptonic decay is
a good reference process for this measurement, as well as for the QCD process of high Pt
jet pair production.
On leave from High Energy Physics Institute, Tbilisi State University, Tbilisi, GE-
380086, Republic of Georgia.
1 Introduction
The main physics motivation of heavy ion experiments at LHC is possible observation
and study of the phase transition from the confined hadronic matter to the plasma of
deconfined quarks and gluons (quark-gluon plasma, QGP). At the center of mass energies
of few TeV per nucleon pair and for heavy nuclei A ~ 200, the energy density in central
nucleus-nucleus collisions must be well above the expected phase transition value (see e.g.
reviews [0, 3)-
Unfortunately, at present we know no single measurement that could give an unam-
biguous indication that the phase transition to QGP has taken place, so several observable
effects should be measured in order to establish the existence of QGP and study its proper-
ties. A number of potentially measurable effects have been suggested and widely discussed
in the literature (see e.g. refs. [@, H). A list of those taken from ref. [|], obviously not
exhaustive and in no particular order, is shown below :
1. transverse energy distributions;
2. charged multiplicities;
3. transverse momentum distributions of identified particles;
4. strangeness production;
5. identical particle interferometry;
6. production of vector mesons with low and high mass;
7. continuum of lepton pairs;
8. continuum of direct photons;
9. high transverse momentum jets.
Detailed analysis of these topics goes far beyond the scope of this paper, but we would
like to make some comments. Items 1 to 4 will certainly be sensitive to the final state
effects, but it seems that reliable and unambiguous theoretical predictions about QGP
manifestations are difficult to obtain p|. Same is true for low invariant mass lepton pairs
and light vector mesons, as well as for particle interferometry, with additional difficulties
for the latter caused by very high precision of momentum measurements required. It also
seems that the continuums of high invariant mass lepton pairs and direct photons will
be very hard to study, mainly because of large irreducible backgrounds and low rates.
Anyway, these two processes can be only considered as reference processes (see Sect. 3)
as final state interaction should not significantly affect leptons and photons.
In fact, some recent investigations imply that direct probes of QGP should be hard
enough in order to resolve sub-hadronic scales and distinguish confined and deconfined
media p . So, one is essentially left with two topics most widely discussed in the context of
QGP manifestations: heavy quarkonium production and high transverse momentum jets.
1
Yet again, the use of charmonium states {J/ip-, ip' ■, . . .) at LHC heavy ion experiments may
be questionable, as their production rates even at average transverse momenta should be
mostly dominated by weak decays of bottom hadrons, which take place far beyond the
area where QGP can be formed. T family production at LHC looks more promising
especially when the peculiarities of the recent Tevatron data are explained theoretically.
As for high transverse momentum jets,
p + p, A + ]et + X, (1)
their production mechanisms in hadron-hadron collisions are well understood in all acces-
sible ranges of the initial energy and jet transverse momentum. It is believed that their
properties should be affected by the dense hadronic matter 0, ^, resulting in measurable
effects like additional energy loss and acoplanarity. Detailed investigation of high Pt jet
production in heavy ion collisions at LHC and its comparison to that in pp interactions
are well under way |10|, 11, 12]. Process (|l]) is especially interesting as jet quenching
patterns are expected to be different for hadronic gas and QGP g,
We propose another hard process which also allows to investigate the effects of nu-
clear interaction on the propagation of a fast moving color charge, and can potentially
distinguish the two states of the dense hadronic matter. This is the production of a
leptonically decaying Z boson with high transverse momentum, in association with an
energetic hadronic jet:
p + p, A + A^ fx+fx-) + jet + X. (2)
The detailed studies of the influence of the dense matter on jet development can be found
in the literature (see e.g. refs. 0, ||, We do not consider them here; the purpose
of this paper is just to show that such an investigation is theoretically interesting and
experimentally feasible at the LHC heavy ion facility. As shown in Sect. 2, the cross
section of this process is high enough and the background can be easily controlled using
physically motivated cuts. For the purposes of QGP study, process (0) incorporates
most of the useful features of (|l|), and offers some additional possibilities, which are also
discussed in Section 2.
Another problem in heavy ion physics is that the cross sections here are not easy
to normalize properly. This means that instead of comparing a distribution in nucleus-
nucleus collisions with a similar one obtained in nucleon-nucleon interactions, one has to
specify an appropriate reference process, and compare the ratios of two processes. In Sect.
3 we argue that Z boson inclusive production with a subsequent leptonic decay,
p + p, A + A^ Z{^ fi+fi-) + X, (3)
is a good choice for the reference process, which satisfies all necessary criteria.
We have performed a Monte-Carlo simulation of the processes (|) and (|]), assuming
a realistic detector model, which can detect Z — > ^^fi^ decays and high Pt jets in the
rapidity range \Y\ < 2.5, for lead-lead collisions at LHC energies. Some results of our
simulation and the discussion of various experimental aspects are presented in Sect. 4,
and conclusions are drawn out in Sect. 5.
2
2 Z + jet production
At multi-TeV accelerators, high transverse energy jet production is expected to be com-
pletely dominated by the parton subprocess of (quasi)elastic gluon-gluon large angle scat-
tering. Hence, in heavy ion collisions at the LHC one could investigate the propagation
of 5^/iton-initiated jets in the dense hadronic matter. The cross section of this process
is relatively high and large statistics can be accumulated. QCD-motivated theoretical
expectations are well tested at available energies and no serious discrepancies have been
found. So, the process (|I]) is expected to be the major source of information about the
effects of the dense matter on the fast-moving gluon fragmentation process in heavy ion
collisions at the LHC. Possible changes in the energy loss mechanism in QGP as opposed
to dense hadronic gas p, Q make this process even more interesting.
Nevertheless, some criticism should be expressed:
• As all participating partons in the initial hard scattering subprocess are gluons,
it may be rather difficult to distinguish the effects of the initial state nuclear in-
teractions like gluon shadowing from genuine final state dense matter effects. In
order to do this, one has to know the gluon distribution functions in heavy nuclei
at small fractional momenta, which would require tedious analyses of, say, charm
and beauty production in lepton-nucleus collisions, and/or high Pt jet production
in proton-nucleus interactions.
• No information can be extracted about the influence of the dense hadronic matter
on the fragmentation of gnar/c-initiated jets.
• Both jets in the event are equally affected by the dense hadronic matter. Most prob-
ably, the two jets are affected independently, which means that certain measurable
effects like acoplanarity will be slightly reduced. But some possible coherent effects
may escape detection.
Consider now the process (0) of a high Z-boson production in association with a
jet. In this case, relevant parton subprocesses are q{q) + g ^ Z + q{q) and q + q ^ Z + g.
Explicit calculations show that in the rapidity region |y| < 2.5 and for both jet and Z
transverse momenta exceeding 50 GeV/c, the relative contribution of these two parton
subprocesses is ~ 70 % and ~ 30 %, correspondingly. At least one of the initial partons is
(almost certainly) a quark, and as far as quark distribution functions are readily measured
in high energy lepton-nucleus collisions, initial state nuclear effects can be taken into
account much easier and far more reliably than for the process (|I|). Moreover, in about
70% of events the detected jet takes its origin from a quark or an antiquark, and by
choosing large \Y\ values of the Z + jet system this proportion can be made even higher.
This means that the process (0) gives a unique possibility to study the effects of the
nuclear matter on a high purity sample of light- quark-indnced jets Q.
^One could also speculate about tagging one of the 6(6)-quark induced jets in the processes p + p, A +
b + b + X and measuring changes in the 6(&)-quark fragmentation process due to nuclear effects, but
experimental feasibility of this study requires further investigation |l3|| .
3
The process has yet another advantage: the transverse momentum of the jet is
highly correlated to that of the Z (in fact, neglecting the primordial Pt distribution of
initial partons, in the absence of the initial-state radiation both P^'s are exactly equal
and opposite), so the jet in this case can be considered as tagged, and the effects of its
further evolution should be easier to determine, as opposed to the two jet case, where
both jets are affected by the final state effects.
From the experimental point of view, two high transverse momentum muons in (H)
can be triggered with high efficiency and low background, using general purpose LHC
detectors CMS and ATLAS. The muon trigger of these detectors provides /i detection
with the Pt threshold of few GeV/c. Though it is optimized for nominal pp interactions,
a similar muon trigger can be successfully used in heavy ion interactions as well. As for jet
studies, jets with transverse energies above 50 GeV can be recognized with high efficiency
and low background even in central Ph — Ph collisions [O
3 Z production as a reference process
As mentioned above, serious normalization problems of the cross sections measured in
heavy ion collisions can be avoided, if an appropriate reference process is found, and ratios
of the two processes are compared in nucleus-nucleus and proton-proton interactions.
It is most important that the reference process is not sensitive to final state nuclear
interaction effects, so production of the large invariant mass lepton pair continuum seems
to be an obvious candidate. Explicit calculation in realistic experimental conditions show,
however, that a significant background coming from heavy quark semileptonic decays and
low production rates of lepton pairs with high invariant mass M (/i"*"/!") > 100 GeV make
this process almost useless for the above purpose.
We argue that the production of the Z boson and its subsequent leptonic decay Z —>■
H~^IJ,~ should be a good choice for a reference process for a number of reasons:
• Z decays weakly into muons, and final state interactions are clearly negligible in
this case;
• the cross section is high enough, and the background can be shown to be well under
control, so that the expected error can be made sufficiently small;
• the invariant mass of the produced system = Mz is large enough, not too far
from the average invariant mass of relevant high Pt jets, Z + jet system and other
possible hard processes of interest, which makes their production kinematics quite
similar;
• momenta of initial partons involved are large enough, xi^2 ~ y^s/s > 0.02, so that
the very small x region, most strongly affected by the initial state nuclear interaction
like gluon shadowing, is avoided.
4
• As the Z-boson is produced mainly by quark- ant iquark fusion mechanism, a very
good self-consistency check can be performed, by using quark and antiquark distribu-
tion functions extracted from deep inelastic lepton-nucleus scattering for describing
Z inclusive distributions in heavy ion collisions.
• General purpose LHC detectors are capable of measuring Z production with high
efficiency and resolution.
So, we propose to use the LHC heavy ion facility to study: i) the ratio of the number of
high Pt jets to the number of detected Z — > /^'''Ai" decays, and ii) the ratio of the number
oi Z + jet events to the same number of detected Z-boson decays, in various heavy ion
collisions as well as in pp interaction. Note that for the latter process we expect significant
cancellation of initial-state nuclear effects in the ratio, as initial quarks are involved in
both the numerator and the denominator. Expected statistics will allow to measure the
variation of these ratios with the transverse energy fiow and other quantities of interest.
4 Experimental aspects and results of simulation
At LHC heavy ions will be accelerated up to the energies E = Ep ■ {2Z/A) per nucleon
pair, where Ep = 7 TeV is the proton beam energy for LHC. In the case of Pb nuclei the
energy per nucleon pair will be 5.5 TeV and the design luminosity for a single experiment
is about L ^ 1.0 X 10^^ cm~^s~^. The event rate for Pb — Pb interactions is expected to
be in the range of 6 — 7 kHz, but only 2 — 3 % of it will correspond to central collisions.
The rates of jet, Z+jet and Z production (processes (0-0)) in nucleus-nucleus collisions
were obtained from pp interactions at the same energy using the linear A-dependence:
aAA = A^"" X (jpp, with a = 1.0. The cross-sections in pp collisions were evaluated
using the PYTHIA 5.7 Monte-Carlo program |T£|, with the default structure function
set (CTEQ2L) and the /T-factor equal to 2. The effects of deflection for quark structure
functions with small fractional momenta in a nucleus relative to a free nucleon, and energy
losses of high P^ partons in the dense matter were not taken into account.
The charged particle density in mid-rapidity range for the central Pb — Pb interactions
at LHC is expected to be around 2000 — 4000. Several Monte-Carlo event generators have
been created for simulating such collisions. HIJING FRITIOF |TB] and VENUS
17| are often used for nucleus-nucleus collision studies at SPS energies. We have used
another approach to the simulation of this process, implemented in the SHAKER Monte-
Carlo program, where gross features of soft particle production in heavy ion collisions are
reproduced using some simple parametrizations. The main parameter for this generator
is the charged particle density in central rapidity region, according to which the rates
of various types of particles are computed. Particle production was generated according
to the flat rapidity distribution in the region |y| < 2.5 and experimental distributions
measured at the Tevatron were used as transverse momentum spectra. After that, the
5
simulated central heavy-ion collision event was added to a hard pp interaction event with
a high Pt jet and Z production. The following kinematical cuts have been applied:
c.m.s. rapidity Y of the Z and the jet should be within the limits \Y\ < 2.5;
transverse momentum of each muon is larger than 20 GeV/c. Such muons will be
measured with very high efficiency and low background |[TT[| ;
• transverse energy of the jet is larger than 50 GeV . Reconstruction efficiency for jets
with smaller transverse energy is fairly low, and large contamination from false jets
is anticipated;
• transverse momentum of the muon pair is larger than 50 GeV/ c. This cut eliminates
the background from uncorrelated lepton pairs and maintains the balance of the
transverse energy in the event.
In this kinematical region the contribution of parton subprocesses q + g Z + q and
q + q ^ Z + g are 71% and 29% respectively. The Monte-Carlo study has shown that
for Pt > 50 GeV/c the jet recognition efficiency of 95% can be achieved, while the
contribution of false jets can be reduced down to 10% [IT2 .
The dominant background of pairs in this case comes from two sources: Z
production with the misrecognition of a fluctuation in the transverse momentum flow of
soft particles as a jet, and QCD processes of heavy flavor pair cc,bb,tt production with
their subsequent semileptonic decays. However, the cut upon the transverse momentum
of the muon pair Prpf^~^^~ > 50 GeV/ c signiflcantly reduces both of them, so that the flnal
contribution of the background is less than 5% even without using the muon isolation
criteria. The background due to uncorrelated muon pairs from vr and K decays is negligible
for Pt'^ > 20 GeV/c. The expected number of jet + Z events is around 2000.
For the same kinematical region the expected number of high Pt jets in Pb — Pb
collisions (process (|l|)) is ~ 10^, while the expected number of detected Z — >• decays
(process (^) in the same reaction is around ~ 5 x 10^. The luminosity for lighter ions at
LHC should be much larger, so that numbers of events in case of Ca — Ca interactions are
expected to be about 100 times higher for the same period of running time, with similar
background levels.
5 Conclusion
We have shown that the high transverse momentum Z-boson production in association
with a hadronic jet in heavy ion interactions (reaction (|])) is a unique tool for investigating
fast colour charge (mainly light quark) propagation through dense hadronic matter and
searching for the manifestations of the quark-gluon plasma formation at high energy
densities. Despite its smaller cross section, which still happens to be large enough to be
measurable with sufficient accuracy using general purpose LHC detectors, the proposed
6
reaction can successfully complement similar studies of high transverse momentum jet
production (reaction (||)) in a number of aspects, and has several important advantages.
In order to compare the rates of various hard processes in different heavy ion and
proton-proton collisions, an appropriate reference process should be used. We argue that
for a number of theoretical and experimental reasons, the process of Z boson production
and its subsequent muonic decay (reaction (^)) is the best choice for such a process
satisfying all necessary conditions.
We would like to thank M.Bedjidian and D.Denegri for their interest and helpful
discussions.
References
[1] H.Satz, in ECFA LHC Workshop Proceedings, CERN 90-10, ECFA 90-133 (1990)
V.I, 188.
[2]
[3]
[4]
[5]
[6]
[7]
[9
[10
[11
[12
[13
[14
C.Singh, Phys. Rep. 236 (1993) 147.
H.Satz, Z. Phys. C 62 (1994) 683.
H.J.Specht, in ECFA LHC Workshop Proceedings, CERN 90-10, ECFA 90-133 (1990)
v.II, 1236.
H.Satz, Plenary talk at Quark Matter '95, Monterey, January 9 - 13, 1995; CERN-
TH/95-24.
R.Gavai et al., CERN-TH.7526/94.
CDF Collaboration, talk given by V.Papadimitriou at Rencontres de Moriond, Les
Arcs, March 19-26, 1995.
M.Gyulassy and M.Pliimer, Phys.Lett. B 243 (1990) 432.
J.Pan and C.Gale, Phys.Rev. D 50 (1994) 3235.
X.N.Wang and M.Gyulassy, Phys.Rev.Lett. 68 (1992) 1480.
CMS Technical Proposal, CERN/LHCC 94-38; ATLAS Technical Proposal,
CERN/LHCC/94-43.
V.Kartvelishvili, R.Kvatadze and R.Shanidze, CMS TN/95-052 (unpublished).
V.Kartvelishvili and R.Kvatadze, in preparation.
T.Sjostrand, PYTHIA 5.7 and JETSET 7.4 (Physics and Manual), CERN - TH.
7112/93.
[15] X.N.Wang and M.Gyulassy, Phys.Rev. D44 (1991) 3501.
[16] B.Anderson et al, Nucl.Phys. B281 (1992) 289.
[17] K.Werner and P.Koch, Z. Phys. C47 (1990) 215; K.Werner, CERN 90-10, ECFA
90-133 (1990) v.II, 1098.
8
