Probing Nuclear PDF and Gluon Saturation At The LHC with Forward Direct Photons

2 MB·

by  Mauro R. Cosentino

Probing Nuclear PDF and Gluon Saturation At The
LHC with Forward Direct Photons
7
1
0
2 Mauro R. Cosentino∗, for the ALICE FoCal Collaboration
n CentrodeCienciasNaturaiseHumanas,UniversidadeFederaldoABC
a CEP09210-580,SantoAndre,SP,Brazil
J
E-mail: [email protected]
0
3

Inrelativisticnuclearcollisionssomeoftheimportantaspectstobeaddressedaretheeffectsof
]
x
thenuclearPDFandthegluonsaturation. IntheLHCthebestwaytoaddressthesequestionsis
e
- bymeansofpAcollisionsandinparticularthroughthemeasurementofdirectphotonproduction
l
c intheforwarddirection. Inordertoachievethismeasurementanewforwardcalorimeter(FoCal)
u
n isproposedasanupgradetotheALICEexperiment. Theproposeddetectorwillcovertherange
[ 3.5 < η < 5, probing the gluon distributions at x∼10−5 and Q∼ p > 4 GeV. We will discuss
T

1 performancestudiesanddemonstratethatextremelyhigh-granularitycalorimetryisrequiredfor
v
suchmeasurement. WewillalsopresentafewresultsfromR&Dforthisproject.
7
7
6
8
0
.
1
0
7
1
:
v
i
X
r
a
The26thInternationalNuclearPhysicsConference
11-16September,2016
Adelaide,Australia
∗Speaker.
(cid:13)c Copyrightownedbytheauthor(s)underthetermsoftheCreativeCommons
Attribution-NonCommercial-NoDerivatives4.0InternationalLicense(CCBY-NC-ND4.0). http://pos.sissa.it/
ProbingnPDFandGluonSaturationwithForwardDirectPhotons MauroR.Cosentino
1. Motivation
The inner structure of hadrons, known as Parton Distribution Functions (PDFs) for protons
and neutrons and as nPDFs for nuclei, is of fundamental importance in the understanding of the
physics results coming from high energy hadron colliders, such as the LHC. Since the PDFs are
initial state configurations, they can impact many of final observables like cross sections [1] or
azimuthalangularcorrelations[2].
Inadditiontotheabove,nPDFsareofgreatinterestthemselves,giventhefactthattheDGLAP
equations, that describe their evolution, predict a divergence of the gluon densities for decreasing
valuesoflongitudinalmomentumfractionx,violating,forexample,unitarity. Fig. 1(left),presents
the PDFs for the proton as obtained from Deep Inelastic Scattering (DIS) from ZEUS data [3],
showingthegluondensitytendencytodiverge. Thegluondensity-schematicallypresentedinthe
rightpanelofFig. 1-increasesfordecreasingxandincreasingmomentumtransfer,Q2,uptothe
point where non-linearity effects are expected to become important and a saturation of gluons is
expectedtolimitthegluondistribution. GluonsaturationisdescribedbymodelssuchastheColor
Glass Condensate (CGC) [4], which describe quantitatively the initial states of highly energetic
protonsandnuclei.
Figure 1: Evolution of PDFs. Left: proton PDF for various fits. (Gluon and sea-quark distributions - xg
andxS-arescaleddownbyafactorof20[3]). Right: thegluonsaturationdiagramaccordingtoevolution
equations.
Asaturationscaleforthepartondistributionofgluonscanbeestimatedby
α (cid:18)A(cid:19)1/3
Q2(x)≈ S xG(x,Q2)∝ (1.1)
s πR2 x
showing that for a given value of x, the heavier the nucleus1, the larger is Q2, making this a more
s
accessible measurement. In addition to this, the minimum accessible value of x can be estimated
asafunctionoftherapidityyandthecollisionenergyby2
2p
x ≈ √ T e−y (1.2)
min
s
NN
1TheAdependencecomesfromG∝AandR∝A1/3
2FromanapproximationofaLOcalculationofa2→2process.
1
ProbingnPDFandGluonSaturationwithForwardDirectPhotons MauroR.Cosentino
Hence, equations 1.1 and 1.2 show that saturation can be more easily probed in collisions
of protons on heavy nuclei targets with measurements taken in the forward region. The higher
collisionenergyoftheLHCprovidesaccesstoxaslowasx∼10−5 at(cid:104)y(cid:105)≈4. Theuseofprotons
asprojectileshasthepurposeofavoidingthefinalstateeffectspresentinA−Acollisions.
1.1 SaturationI:IndicationsfromRHIC
AttheRelativisticHeavyIonCollider(RHIC),measurementsaimedtoprobegluonsaturation
areperformedind-Aucollisions. AtRHIC,twoofthemoststrikingresultsthatsupporttheCGC
model are the ones presented in Fig. 2. On the left the inclusive hadron p spectra scaled by the
T
equivalentnumberofproton-protoncollisionsgivenbythevariable
d2N/(dηdp )|
T dAu
R = (1.3)
dAu (cid:104)N (dAu)(cid:105)d2N/(dηdp )|
coll T pp
shows a much stronger suppression in the forward region ((cid:104)η(cid:105)=4) than in more central rapidity
values,withtheoverallpictureshowingapatternofincreasingsuppressionwithrapidity.
Figure2: RHICd+Auresults. Left: R η dependency,showinganincreasingsuppressionpattern. Right:
dAu
Comparisonoftwohadronazimuthalcorrelationsinp+pandd+Au,whereasignificantsuppressionofthe
awaysideisobserved.[5]
The other important result is the one on the right of Fig. 2, where the azimuthal angular
distribution, with respect to a trigger particle, of neutral pions measured in the forward region is
compared between the two colliding systems - pp and d-Au. There is a strong suppression of the
awayside(∆φ ≈π)inthed-Aucollisionswithsmallimpactparameter,whenthedeuteronhitsthe
Au nucleus where the nucleon - and hence parton - density is higher. Both results are consistent
with CGC predictions - that predicts that the recoil parton is in a coherent state with many others
- but other hadronic final state interactions, such as hadronic absorption, inital state energy loss
and/ormultiplescatteringseffectcannotberuledout.
1.2 SaturationII:IndicationsfromLHC
At the LHC the scenario is similar to the one at RHIC in the sense that there are several
results presenting modification in p–Pb collisions with respect to the same observables in pp that
maybethemanifestationofinitialstateeffects,butthathavemanyotherpossiblehadroniceffects
contributing to them. Fig. 3 presents some of the ALICE experiment p–Pb results where the
2
ProbingnPDFandGluonSaturationwithForwardDirectPhotons MauroR.Cosentino
comparison to pp is avoided by the creation of an observable based on the Forward-to-Backward
RatioorR ,whichissimplytheyieldmeasuredinforward(protongoing)dividedbytheyieldin
FB
thebackward(leadgoing)direction,withinthecommonrapiditycoverageofboth.
+ ) µµ→ 1.5 ALIC2E.9 p6 <P|by |s<N3N.=553.02 TeV RFB 2.22 p Pb sNN = 5.02 TeV,2 µ.9±6←< |cy,b d|e<c3a.5ys4
(RφFB 1 111...468 ALICE Preliminary cms
1.2
1
0.5 0.8
data 0.6
HIJING 0.4 NLO (MNR) with EPS09 shadowing
DPMJET 0.2 systematic uncertainty on normalization
00 1 2 3 4 5 6 7 00 2 4 6 8 10 12 14 16
p (GeV/c) p (GeV/c)
T ALI−PREL−80458 T
Figure3: ALICER resultsinp–PbintheLHC.Left: R forφ mesonsshowingastrongsuppression.
FB FB
Right: R forµ ←c,b,presentingaslightsuppressionatlow p .[6]
FB T
On the left side of Fig. 3, the R of φ mesons shows a strong suppression for which the
FB
physical mechanism is not yet clear. On the right side of the figure, the R of muon from charm
FB
and bottom decays is consistent with shadowing. However the initial state kinematics (x, Q2) of
heavyflavourarenotwelldefined,makingtheseresultsnoyetveryconclusive.
In summary we may say that in p(d)−A collisions, there are striking results at RHIC, and
severalresultsattheLHCshowingsuppressionofhadronproductioninprocessesthataresensitive
tolow-xgluondensity,butallofthemarebasedinhadronicprobesthataresubjecttomanyother
sources of influence other than the initial parton distribution. The natural alternative to this is to
make use of electromagnetic probes, that can be sensitive to the gluon parton distribution and be
freeofundesiredfinalstatedynamics.
1.3 ProbingSaturationwithDirectPhotons
Dileptons from Drell-Yan (DY) processes can be used to probe the nPDFs in the very same
way as in DIS, since they are equivalent processes, as shown by the Feynman diagrams on the
left of Fig. 4, where the scattering electron in DIS corresponds to the outgoing (anti)lepton pair.
Fromthoseonecanseethat,infact,DYissensitivetothegluondistributionsonlyatNLO,which
decreases significantly the cross section of the desired processes in one hand and on the other the
LOprocessguaranteesthatmostofthedileptonswillbebackground.
Thecasefordirectphotonsismorepromising. TherightsideofFig. 4showsthe"Compton-
like" diagram which is a LO process sensitive to the gluon parton distribution function. The ex-
pectedluminosityofthefutureLHCp–PbrunsmaynotallowaDYmeasurement,whilethedirect
photon,withitshighercrosssection,canbeapromisingprobe.
2. TheALICEexperimentandtheFoCalasanUpgrade
Themeasurementofdirectphotonsatforwardrapidityinppandp–PbcollisionsatLHCisthe
mainphysicsmotivationfortheproposedFoCalupgradeoftheALICEexperiment. Theproposal
3
ProbingnPDFandGluonSaturationwithForwardDirectPhotons MauroR.Cosentino
Figure4: Feynmandiagramsfortheproductionofdifferentelectromagneticprobes. Left: Drell-Yanpro-
cesses and its correspondence to DIS, showing their gluon sensitivity only at NLO. Right: Direct photon
productionina"Comptonlike"diagramwhichissensitivetogluonsalreadyatLO.
ispresentlyunderdiscussionwithintheCollaborationandwouldbetargetedforinstallationduring
thethirdlongshutdownoftheLHCin2024.
The current setup of ALICE [7] is shown schematically in Fig. 5. On the left is presented
the overview of the experiment with labels showing the places of the approved upgrades[8], and
the proposed position for the FoCal (in red). The right side of Fig.5 presents the side-view of
the experiment with the possible location for the two parts of the FoCal: the electromagnetic and
hadroniccalorimeters.
Figure5: SchematicviewoftheALICEexperiment. Left: currentALICEconfigurationwiththeindication
oftheupgrades. Right: longitudinalviewoftheALICEexperimentandtheplacementfortheFoCal.
The FoCal project aims for a calorimeter for γ and π0 measurements with enough resolution
to reconstruct π0s either by the invariant mass of the two γ showers reconstructed separately or
through the shape of the reconstruct overlapped showers. To cover the largest rapidity possible,
theFoCalisplacedat∼8mfromtheinteractionpoint,covering3.3<η<5.3,whereπ0senergyare
high and their decays opening angle is small. Hence, the Molière radius (R ) should be as small
M
as possible, to enable very high spatial resolution. For that it also needs a very high granularity
read-out.
4
ProbingnPDFandGluonSaturationwithForwardDirectPhotons MauroR.Cosentino
Figure6: TheFoCalstrawmandesign. Left: thelongitudinalsegmentation, dividedinlayersofabsorber,
low(LG)andhigh(HG)granularityreadout. Right: theperpendicularprofilesoftheLGandHGlayersand
thedetailontherightendschematicallyshowsthegranularityofasingleHGcell.
The current design to meet these requirements is of a sampling calorimeter with 24 layers of
3.5 mm thick tungsten (∼1 X ) as absorber and Si sensitive layers of two kinds: low granularity
0
(LG)SiPADs(∼1cm2)andhighgranularity(HG)pixel(∼1mm2). Theelectromagneticsection
of the FoCal - depicted as the strawman design of Fig. 6 - is placed at ∼7 m of the interaction
point,outsidethemagnet.
2.1 Performanceofdirectphotonmeasurement
ThecurrentdesignhasbeensimulatedusingtheALICEanalysisframework-basedonROOT,
PYTHIAandGEANT3[9]-toevaluateitsperformanceinthemeasurementofdirectphotons. The
main challenge of this measurement is the extremely high background of decay photons, the vast
majorityofthemcomingfromπ0 decays. Consequentlythesuccessofthismeasurementdepends
ontheefficientπ0 identificationtoenabletherejectionoftheirdecayproducts.
The π0 identification can be achieved either by their invariant mass reconstruction or by the
shapeoftheirreconstructedelectromagneticshowers. Inthefirstcase,thetwophotonshaveenough
spatialseparationtoallowtheirelectromagneticshowerstobereconstructedintwoseparatedclus-
ters, from which an invariant mass can be calculated, and in the case of this being consistent with
the π0 mass both clusters (photons) are rejected. The design with HG layers can provide a two-
showerseparationdistanceofonlyafewmm.
Thecaseofshowershapemethodisnecessarywhentheπ0energyishigherthanacertainlimit
whichbooststhedecayphotonstobesoclosespatiallythattheirshowerscannotbereconstructed
separately, even with the resolution of a few mm. In this case, an invariant mass calculation is
not possible, however the overlapped shower shape is very different from that of a single electro-
magneticshower. Thiseffectcanbeevaluatedgiventhatthecalorimeterhasenoughgranularityto
revealthedetailsofthereconstructedcluster. ThedesignwithHGlayershasarejectionpowerofa
factor∼5higherthantheoptionwithoutit. Therejectionpowerofthosemethodscanbeobtained
from the left side of Fig. 7, where the ratio of photons to all clusters is presented after each cut.
In some cases one of the decay photons cannot be reconstructed, and in such cases the other one
cannot be reject using the previous methods. At these high energies, π0s are generally associated
withjets,andthiskindofbackgroundcanonlyberemovedbymeansofisolationtechniques,since
direct photons should be alone (isolated), while π0s in jets have some hadronic activity around
them. For this same reason fragmentation photons can also be rejected with isolation techniques.
5
ProbingnPDFandGluonSaturationwithForwardDirectPhotons MauroR.Cosentino
usters 4p.p0 <s η = < 1 53. 0TeV ALFIoCCEa sl iumpuglraatidoen RpPb 1 sNN = 8.8 TeV ALICE simulation
all cl 1 7m position 0.8 4.0 < η < 5.0 FoCal upgrade
/γdir Statistical
Systematic (PYTHIA)
10−1 0.6 Systematic (JETPHOX)
0.4
10−2
dec rej + iso 0.2
dec rej (IM+SS)
isolation
R =0.4, pE+H < 3.0 GeV no selection
iso T,iso
10−3 0
4 6 8 10 12 14 0 5 10 15 20
p (GeV/c) ALI−SIMUL−69835 p (GeV/c)
ALI-SIMUL-69823 T T
Figure7: Directphotonmeasurementperformance. Left: rejectionofbackgroundphotonsaftertheappli-
cation of each analysis technique, decay rejection (Invariant Mass and Shower Shape), isolation and their
combination. Right: modificationfactorofdirectphotonmeasurement(designwithHGlayers).
These are the main motivations of the hadronic section of the FoCal, since the rejection power of
theisolationisgreatlyimprovedwithit,asshownintheleftsideofFig. 7.
Afterapplyingallrejectiontechniquesthefinalresultsoftheperformancesimulationsispre-
sentedintherightsideofFig. 7,asthe p dependencyofthenuclearmodificationfactor3 (R ).
T pPb
With such experimental precision the future results should be precise enough to refute or confirm
scenariossuchastheCGC.
2.2 R&Dandtestbeamresults
Figure 8: Test beam results. Left: two-shower separations. Right: lateral shower profiles for a 50 GeV
electronbeaminselecteddetectorlayers,eachcorrespondingto≈1X .
0
The FoCal collaboration is performing R&D for LG and HG technologies, the first is likely
to adopt Si-pads, while the HG layers have the Monolithic Active Pixel Sensors (MAPS) as its
3Calculationanalogoustoequation1.3.
6
ProbingnPDFandGluonSaturationwithForwardDirectPhotons MauroR.Cosentino
preferredoption,duetoitscapabilitytoallowveryhighpixeldensityaddinglittlematerialbudget.
A prototype of electromagnetic calorimeter using high granularity MIMOSA23 sensors [10] has
been built and tested. The prototype had a total number of 39 million pixels of 30 × 30 µm2 in a
4×4×11cm3 volumeandaMolièreradiusof∼11mm,althoughalargefractionoftheshower
energyiscontainedonlyinafewmm. Theresults,presentedinFig. 8,showthattheprototypehas
beensuccessfulinreconstructingelectromagneticshowerssmallenoughtoallowtheseparationof
onlyafewmmdistancebetweenthem.
3. Conclusions
In summary, the open question of the gluon saturation in the initial state hadrons and nuclei
can be addressed by means of the direct photon measurement at forward rapidity, and the ALICE
experiment, through the FoCal upgrade, with unprecedented spatial resolution, could provide the
measurementwiththenecessaryprecisiontodiscriminateamongdifferenttheoreticalmodels.
Acknowledgments
ThespeakerwouldliketothanktheFoCalcolleaguesfortheopportunityandfortheirsupport,
andalsotoacknowledgeCAPESandFAPESPforthefinancialsupport.
References
[1] A.D.Martin,inproceedingsofSchoolonQCD,lowxphysics,saturationanddiffraction,Acta
PhysicaPolonicaB392025(2008)[he h/0802.0161].
[2] H.-J.DrescherandY.Nara,Phys.Rev.C75034905(2007)
[3] ZEUSCollaboration,DESY-02-NNN,8thJuly2002,
[http://durpdg.dur.ac.uk/pdflib/zeuspdf/eps01.ps.gz]
[4] L.McLerranandR.Venugopalan,Phys.Rev.D49,3352(1994)
[5] STARCollaboration,Phys.Rev.Lett.97,152302(2006);Nucl.Phys.A854,168(2011)
[6] ALICECollaboration,[nucl-ex/1506.09206],Nucl.Phys.A931,546(2014)
[7] ALICECollaboration,Int.J.Mod.Phys.A29,1430044(2014)
[8] ALICECollaboration,J.Phys.G.41,087001(2014)
[9] R.BrunandF.Rademakers,Nucl.Inst.&Meth.inPhys.Res.A389,81(1997);R.Brunetal.,
CERN-DD-EE-84-1,(1987);T.Sjöstrand,S.MrennaandP.Skands,JHEP0605026,(2006)
[10] R.Turchettaetal.,Nucl.Instrum.Meth.A458677(2001)
7