recent results on hard and rare probes from atlas · motivation and recent atlas results i qgp...

Recent results on hard and rare probesfrom ATLAS

Mariusz PrzybycieńAGH University of Science and Technology, Krakow, Poland

(on behalf of the ATLAS Collaboration)

Motivation and recent ATLAS resultsI QGP exists for a short time (t ∼ 10 fm/c ∼ 10−23 s) – need a probe from inside the

collision region.

I Hard probes are produced early in the HI collision, in a process which cross section isnot changed by presence of strongly interacting medium, i.e. can by calculated in pQCD.

I Passing through the medium hard probes interact weakly or strongly with it providinginformation on its properties.

The following measurements will be reviewed in this talk:Measurement of angular and momentum distributions of charged particles within and aroundjets in Pb+Pb and pp collisions at

√sNN = 5.02 TeV with the ATLAS detector. Phys. Rev. C

100, 064901 (2019)Measurement of suppression of large-radius jets and its dependence on substructure in Pb+Pbat 5.02 TeV with the ATLAS detector. ATLAS-CONF-2019-056Measurement of W± boson production in Pb+Pb collisions at

√sNN = 5.02 TeV with the

ATLAS detector. Eur. Phys. J. C 79, 935 (2019)Z boson production in Pb+Pb collisions at

√sNN = 5.02 TeV measured by the ATLAS

experiment. Phys. Lett. B 802, 135262 (2020)Measurement of light-by-light scattering and search for axion-like particles with 2.2 nb−1 ofPb+Pb data with the ATLAS detector. ATLAS-CONF-2020-010

M. Przybycień (AGH UST) Hard and rare probes of QGP with ATLAS LHCP 2020 25 - 30 May 2020 1 / 14

https://doi.org/10.1103/PhysRevC.100.064901

https://doi.org/10.1103/PhysRevC.100.064901

https://cds.cern.ch/record/2701506

https://doi.org/10.1140/epjc/s10052-019-7439-3

https://doi.org/10.1016/j.physletb.2020.135262

https://atlas.web.cern.ch/Atlas/GROUPS/PHYSICS/CONFNOTES/ATLAS-CONF-2020-010/

Jets as probes of the Quark-Gluon PlasmaI High transverse momentum partons, produced in hard scattering process, propagating

through the medium of strongly interacting nuclear matter, lose energy, resulting inthe phenomenon of ‘jet quenching’.

I Magnitude of the suppresion is expected to depend on both the pT dependence ofenergy loss as well as the shape of initial jet pT spectrum.

I Nuclear modification factor: RAA =1

Nevt

1〈TAA〉

(d2Nproc

dpTdy

∣∣∣∣cent

)/d2σppproc

dpTdy

I A continuous suppression of jet production from peripheral to central Pb+Pb collisionswith respect to pp reference is observed.

I RAA is observed to grow slowly (quenching decreases) with increasing jet pT.

[GeV]T

p

AA

R

0.5

1

40 60 100 200 300 500 90040 60 100 200 300 500 90040 60 100 200 300 500 90040 60 100 200 300 500 90040 60 100 200 300 500 90040 60 100 200 300 500 90040 60 100 200 300 500 90040 60 100 200 300 500 90040 60 100 200 300 500 90040 60 100 200 300 500 900

0 - 10%20 - 30%40 - 50%60 - 70%

ATLAS = 5.02 TeVNNs = 0.4 jets, R tkanti-

| < 2.8y|-1 25 pbpp, -12015 data: Pb+Pb 0.49 nb

and luminosity uncer.⟩AA

T⟨ Phy

s.L

ett.

B79

0(2

019)

108


Charged-particle distributions within and around jetI Yields of charged-particles as a function of particle’s pT

and distance r from the jet axis:

D(pT, r) =1Njet

12πr dr

dnch(pT, r)dpT

I D(pT, r) decrease as a function of distance from the jet axis.I The rate of fall-off increases sharply for higher pT particles, with most of these being

concentrated near the jet axis.I The distributions exhibit a difference in shape between Pb+Pb and pp collisions, with

the Pb+Pb distributions being broader at low pT and narrower at high pT.

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

r

4−10

3−10

2−10

1−10

1

10

210

]-1

) [G

eVr,

Tp

(D

Pb+Pb pp < 1.6 GeV

Tp1.0 <

< 2.5 GeVT

p1.6 < < 4.0 GeV

Tp2.5 <

< 6.3 GeVT

p4.0 < < 10.0 GeV

Tp6.3 <

< 25.1 GeVT

p10.0 < < 63.1 GeV

Tp25.1 <

< 158 GeVjet

Tp126 <

10%−0ATLAS

-1 = 5.02 TeV, 0.49 nbNNsPb+Pb -1 = 5.02 TeV, 25 pbs pp

R=0.4tkanti-

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

r

4−10

3−10

2−10

1−10

1

10

210

]-1

) [G

eVr,

Tp

(D

Pb+Pb pp < 1.6 GeV

Tp1.0 <

< 2.5 GeVT

p1.6 < < 4.0 GeV

Tp2.5 <

< 6.3 GeVT

p4.0 < < 10.0 GeV

Tp6.3 <

< 25.1 GeVT

p10.0 < < 63.1 GeV

Tp25.1 <

< 158 GeVjet

Tp126 <

80%−60ATLAS


R=0.4tkanti-

PRC 100 (2019) 064901


Ratio of charged-particle yields in Pb+Pb and pp collisionsI Use the observable RD(pT,r) = D(pT, r)Pb+Pb

/D(pT, r)pp to quantify

modifications of the yields due to the QGP.I RD(pT,r) is shown as a function of r for different pT and centrality ranges for jets with

126 < pjetT < 158 GeV.

I Central and midcentral collisions: RD(pT,r) > 1 for particles with pT < 4 GeV,and (for r > 0.05) shows a depletion for particles with pT > 4 GeV.

I Peripheral collisions: no significant r-dependence. (not shown)I Closer look at centrality dependence: Magnitude of the excess or suppresion increase

for more central events in both pT intervals.

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

r

0

2

4

)r, T

p (

DR

< 1.6 GeVT

p1.0 < < 2.5 GeVT

p1.6 < < 4.0 GeV

Tp2.5 < < 6.3 GeV

Tp4.0 <

< 10.0 GeVT

p6.3 < < 25.1 GeVT

p10.0 < < 63.1 GeV

Tp25.1 <

< 158 GeVjet

Tp126 <

10%−0ATLAS

-1 = 5.02 TeV, 0.49 nbNNsPb+Pb -1 = 5.02 TeV, 25 pbs pp R=0.4tkanti-

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

r

0

2

4)r,

Tp

(D

R

T,2p

T,1p

10%− 0 20%− 10 30%− 20 40%− 30 60%− 40 80%− 60

ATLAS-1 = 5.02 TeV, 0.49 nbNNsPb+Pb

-1 = 5.02 TeV, 25 pbs pp < 158 GeVjet

Tp126 < R=0.4tkanti-

< 2.5 GeVT,1

p1.6 < < 10.0 GeV

T,2p6.3 <

PRC 100 (2019) 064901


Difference vs. ratio of charged-particle yieldsI The difference of the charged-particle yields measured in Pb+Pb and pp collisions:

∆D(pT, r) = D(pT, r)Pb+Pb −D(pT, r)pp

quantifies modifications of the yields due to the QGP in terms of particle density (perunit area per GeV).

I An excess in the charged-particle yield density for Pb+Pb collisions compared to ppcollisions is observed for charged particles with pT < 4 GeV, mainly within the jet cone.

I This behaviour is different to the ratios RD(pT,r) where the biggest excess is observedat large r due to the fact that the pp signal is small there.

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

r

0

5

10

15

]-1

) [G

eVr,

Tp

(D ∆

< 1.6 GeVT

p1.0 < < 2.5 GeV

Tp1.6 <

< 4.0 GeVT

p2.5 < < 6.3 GeV

Tp4.0 <

< 10.0 GeVT

p6.3 < < 25.1 GeV

Tp10.0 <

< 63.1 GeVT

p25.1 <

< 251 GeVjet

Tp200 <

10%−0ATLAS


R=0.4tkanti-

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

r

0

2

4

)r, T

p (

DR

< 1.6 GeVT

p1.0 < < 2.5 GeVT

p1.6 < < 4.0 GeV

Tp2.5 < < 6.3 GeV

Tp4.0 <

< 10.0 GeVT

p6.3 < < 25.1 GeVT

p10.0 < < 63.1 GeV

Tp25.1 <

< 251 GeVjet

Tp200 <

10%−0ATLAS

-1 = 5.02 TeV, 0.49 nbNNsPb+Pb -1 = 5.02 TeV, 25 pbs pp R=0.4tkanti-

PRC 100 (2019) 064901


Suppression of large-radius jets

200 300 400 500 [GeV]

Tp

0.5

1.0

AA

R PreliminaryATLAS, 5.02 TeV-1 257 pbpp, -1Pb+Pb 1.72 nb

0-10%10-20%20-40%40-60%60-80%

|<2.0 y| 1.0 jetsR =Reclustered

200 300 400 500[GeV]

Tp

0.4

0.6

0.8

1.0

AAR PreliminaryATLAS

, 5.02 TeV-1257 pbpp, -1Pb+Pb 1.72 nb=1.0 reclustered jets (this analysis)R= 0.4 (PLB 790 (2019) 108)R

|<2.0 y|

ATLAS-CONF-2019-056

I Large-radius jets (R = 1) are obtained by reclusterringsmall-radius (R = 0.2) jets with pjetT > 35 GeV.

I For re-clustered jets the energy radiated between theR = 0.2 sub-jets is removed and lost energy is notrecovered, so (re-clustered R = 1 jets) 6= (R = 1 jets)

I Distance definition:

dij = min(p2T,i, p

2T,j) ·∆R2

ij , ∆Rij =√

∆φ2ij + ∆y2

ij

I Nuclear modification factor:

RAA =1

Nevt

1〈TAA〉

(d3Njet

dpTd√d12dy

∣∣∣∣cent

)/d3σppjet

dpTd√d12dy

d12 referes to the two jets before the final clusteringstep in the anti-kt jet finding algorithm.

I Clear suppression of large-radius jet production incentral collisions.

I The magnitude of the RAA is lower for large-radius jetsthan for small-radius jets.


Suppression of large-radius jetsI d12 referes to the two jets before the final clustering step in the anti-kt jet finding

algorithm, corresponding to the hardest splitting in the jet, so it characterizes the jetsubstructure scale. For jets with only a single sub-jet (SSJ) d12 = 0.

I Significantly different values of RAA for SSJ compared to jets with multiple sub-jets.I The SSJ jets are less suppressed with respect to those with higher sub-jet multiplicity.

I Continous increase of the suppression with increasing centrality.

I For SSJ jets, RAA grows slowly with increasing jet pT.

I No significant pT dependence of RAA is seen for jets with multiple sub-jets. (not shown)

50 100 150[GeV]12d

0.5

1.0

1.5


, 5.02 TeV-1257 pbpp, -1Pb+Pb 1.72 nb0-10%10-20%20-40%40-60%60-80%

<251 GeVT

|<2.0, 200< py|1.0 jetsR =Reclustered

SSJ 0 100 200 300 400⟩

partN ⟨

0.5

1.0

1.5

AA


Single sub-jet

< 20 GeV12d14<

< 41 GeV12d29<

< 84 GeV12d59<

<251 GeVT

p|<2.0 200< y| 1.0 jetsR =Reclustered

200 300 400 500 [GeV]

Tp

0.5

1.0

1.5

AA


0-10%10-20%20-40%40-60%60-80%

|<2.0 Single-sub jety| 1.0 jetsR =Reclustered A

TL

AS

-CO

NF

-201

9-05

6


Probing QGP with electroweak bosonsI Electroweak (EW) bosons and their leptonic decay products do not interact strongly.

Therefore if formed in a hard parton-parton interaction at a very early stage of thePb+Pb collision they carry out unmodified information about the geometry of thenuclei at the collision time.

I In LO, W+(W−) bosons are produced by interactions of u(d) valence quarks and d(u)sea quarks.

I Z and W boson production yields per binary collision do not depend on centrality.This is consistent with a view of heavy ion collisions as a superposition of binarynucleon-nucleon collisions.

⟩ part

N⟨0 100 200 300 400

ev

ents

N |y|<

2.5

ll→

ZN ⟩

coll

N⟨9

10

2

4

6

ATLAS = 2.76 TeVNNsPb+Pb

-1 = 0.15 nbintData 2011 L

ee→Z ll →Z

µµ →Z

ZT

All p

<10GeVZT

p

<30GeVZT

10<p

>30GeVZT

p

PR

L11

0,02

2301

(201

3)

⟩part

N⟨0 50 100 150 200 250 300 350 400

even

tsN

fiduc

ial

N ⟩co

ll N⟨

910

0

5

10

15

20

25

30

35

40

Data±W+W-

W

POWHEG CT10±W+W-

W

-1 0.14-0.15 nb≈ Ldt ∫ = 2.76 TeVNNsPb+Pb

ATLAS

EP

JC75

(201

5)23


W/Z boson yields in Pb+Pb collisions at√sNN = 5.02 TeV

|l

η|

[pb]

lηdNd ev

tN

1 ⟩

AA

T⟨ 1

150

200

250

300

350

400

450

500ATLAS

-1=5.02 TeV, 0.49 nbNN

sPb+Pb, ν+l →+W

> 25 GeVT

νl,p > 40 GeVTm| < 2.5

lη| 0-80%

DataCT14EPPS16nCTEQ15

|l

η|0 0.5 1 1.5 2 2.5

Dat

aP

red.

0.81

|l

η|

[pb]

lηdNd ev

tN

1 ⟩

AA

T⟨ 1

150

200

250

300

350

400

450ATLAS

-1=5.02 TeV, 0.49 nbNN

sPb+Pb, ν−l →−W

> 25 GeVT

νl,p > 40 GeVTm| < 2.5

lη| 0-80%

DataCT14EPPS16nCTEQ15

|l

η|0 0.5 1 1.5 2 2.5

Dat

aP

red.

0.81

EP

JC79

(201

9)93

5

I Combined measurements in electron andmuon decay channels.

I Normalized W±/Z boson yields as functionof |η`|/y, integrated over centrality, arecompared with free proton PDF (+isospin)and nPDFs.

I All predictions describe the shapes well, butthe normalisation is underestimated bynPDFs (10-20%).

0 0.5 1 1.5 2 2.5

20

40

60

80

100

120

[pb]

y/d

Z N

d-1 ⟩

AA

T⟨ -

1ev

tN

ATLAS -1Pb+Pb, 0.49 nb

=5.02 TeVNNs0-100% centrality

<116 GeVll m66<

>20 GeV lT

p

|<2.5 lη|-

l+l → Z

CT14 NLO (Pb+Pb isospin)

EPPS16 NLO

nCTEQ15 NLO

0 0.5 1 1.5 2 2.5|y|

0.8

1

D

ata

The

ory

PL

B80

2(2

020)

1352

62


W/Z boson yields in Pb+Pb collisions at√sNN = 5.02 TeV

⟩part

N⟨

[nb]

evt

N1

⟩A

AT⟨

νl →

WN

1.2

1.4

1.6

1.8

2

2.2ATLAS

-1=5.02 TeV, 0.49 nbNN

sPb+Pb,

> 25 GeVT

νl,p > 40 GeVTm| < 2.5

lη|

: ν+l →+W Data CT14: ν−l →−W Data CT14

unc.⟩AA

T⟨

⟩part

N⟨0 50 100 150 200 250 300 350 400

Dat

aP

red.

0.8

1

⟩part

N⟨0 50 100 150 200 250 300 350 400

AA

R

0.4

0.6

0.8

1

1.2

1.4

1.6 ATLAS-1=5.02 TeV, 0.49 nbNNsPb+Pb,

-1=5.02 TeV, 25 pbs, pp

> 25 GeVT

νl,p > 40 GeVTm| < 2.5

lη|

: ν−l →−W Data CT14

: ν+l →+W Data CT14

norm. unc.

EPJC 79 (2019) 935

I Normalized W± boson yields as function of〈Npart〉 are in good agreement with theorypredictions for 〈Npart〉 > 200, but a slightexcess of the data over predictions in moreperipheral collisions is observed.

I Normalized Z boson yields consistent withthe pp cross-section at all centralities andshow only weak dependence on 〈Npart〉.

I These are also reflected in RAA behaviour.

0 100 200 300 400⟩

partN⟨

300

400

500

[pb]

Z N

-1 ⟩A

A

T⟨ -

1ev

t N

Glauber v2.4

Zppσ


-1, 25 pbpp=5.02 TeVs, NNs

0 100 200 300 400⟩

partN⟨

1

1.2

A

AR

MC isospin

error scaled up by 3⟩part

N⟨

PL

B80

2(2

020)

1352

62


Possible shadowing in inelastic nucleon-nucleon cross section2

data we leave it out from the analysis. The idea is tofirst nail down the EW-boson cross sections by using anext-to-next-to-leading order (NNLO) perturbative QCD(pQCD) with state-of-the-art PDFs for protons and nu-clei. Using the theory prediction on the left-hand-side ofEq. (1), we can then determine σinel

nn within the same MCGlauber implementation as in the experimental analyses.We find that the data favor a significant suppression inσinelnn . We show that this is compatible with predictions

from an eikonal minijet model with nuclear shadowing.We also demonstrate that the unexpected enhancement

seen by ATLAS in the ratios RW±,ZPbPb towards peripheral

collisions disappears with the found smaller value of σinelnn .

II. NUCLEAR SUPPRESSION IN σinelnn

The observables we exploit in this work to extract σinelnn

are the rapidity-dependent nuclear modification ratiosfor W± and Z boson production in different centralityclasses. Experimentally these are defined as

RexpPbPb(y) =

1

〈TAA〉1

NevtdNW±,Z

PbPb /dy

dσW±,Zpp /dy

, (2)

where the per-event yield is normalized into nucleon-nucleon cross section by diving with the mean nuclearoverlap 〈TAA〉 = 〈Nbin〉c/σinel

nn obtained from a MCGlauber model calculation. For minimum-bias collisionsthe same quantity can be calculated directly as a ratiobetween the cross sections in Pb+Pb and p+p collisions,

RtheorPbPb(y) =

1

(208)2dσW±,Z

PbPb /dy

dσW±,Zpp /dy

. (3)

We have calculated the cross sections in Eq. (3) at NNLOwith the mcfm code (version 8.3) [40]. For the protonswe use the recent NNPDF3.1 PDFs [41] which provide anexcellent agreement to ATLAS data for W± and Z bosonproduction in p+p collisions at

√s = 5.02 TeV [42]. The

nuclear modifications for the PDFs are obtained from theEPPS16 NLO analysis [43] which includes Run-I data forW± and Z production in p+Pb collisions at the LHC [44–46] and provide an excellent description of the more re-cent Run-II data [47]. The available NNLO nuclear PDFs[48, 49] do not include any constraints beyond deeply in-elastic scattering, so the applied PDFs provide currentlythe most accurately constrained setup for the consid-ered observables. The factorization and renormalizationscales are fixed to the respective EW boson masses.

The ratios RtheorPbPb and Rexp

PbPb are compared in the up-per panel of Fig. 1. For W±, Rexp

PbPb is formed by divingthe normalized yield in Pb+Pb from Ref. [37] with thecorresponding cross section in p+p from Ref. [42] addingthe uncertainties in quadrature. The plotted experimen-tal uncertainties do not include the uncertainty in 〈TAA〉.The theoretical uncertainties derive from the EPPS16 er-ror sets and correspond to the 68% confidence level. Note

0.4

0.6

0.8

1.0

1.2

1.4

0.4

0.6

0.8

1.0

1.2

1.4

0.0 0.5 1.0 1.5 2.0 2.5

RPbPbwithσinel

nn

=70

mb

EPPS16 (68% uncert.)

ATLAS W− data, 0-80%

ATLAS W+ data, 0-80%

ATLAS Z data, 0-100%

√s = 5.02TeV

RPbPbwithσinel

nn

=41.5mb

y`±, yZ

EPPS16 (68% uncert.)

ATLAS W− data, 0-80%

ATLAS W+ data, 0-80%

ATLAS Z data, 0-100%

√s = 5.02TeV

FIG. 1. Nuclear modification ratios of W± and Z, computedfrom pQCD (solid lines with error bands) and from ATLASdata [37, 38] with σinel

nn = 70 mb (upper panel) and 41.5 mb(lower panel).

that the W± measurement is for 0–80% centrality insteadof full 0–100%. However, for rare processes like the EWbosons the contribution from the 80–100% region is neg-ligible so the comparison with the minimum-bias calcu-lations is justified. It is evident that with σinel

nn = 70 mbboth the W± and the Z data tend to lie above the cal-culated result, which we will interpret as an evidence ofnuclear suppression in σinel

nn as explained below.By equating Eqs. (2) and (3) we can convert each

data point to 〈TAA〉. The outcome is shown in theupper panel of Fig. 2. The obtained values tend tobe higher than the nominal 〈TAA〉 = 5.605 mb−1 (0–100%) and 〈TAA〉 = 6.993 mb−1 (0–80%) which assumeσinelnn = 70 mb, see Table I. The fact that the preferred

values of 〈TAA〉 are independent of the rapidity stronglysuggests that the original mismatch in RPbPb is a nor-malization issue – the nuclear PDFs predict the rapiditydependence correctly.

Since each 〈TAA〉 maps to σinelnn through MC Glauber,

we can also directly convert RexpPbPb to σinel

nn . Here, wehave used TGlauberMC (version 2.4) [50] which is thesame MC Glauber implementation as in the considered

arX

iv:2

003.

1185

6[h

ep-p

h]

I Measurements of EW bosons production inHI collisions depend on normalisations fromthe Glauber model calculations, where theinelastic nucleon-nucleon cross section, σinelnn ,is taken from pp measuremnets, σinelpp = 70 mb.

I Gluon shadowing and saturation phenomenaat small x, might be more pronounced in HIcollisions, especially at high energies, and asa result the σinelnn might be reduced.

I Recently K. Eskola et al. (arXiv:2003.11856[hep-ph]) suggested to treat the EW bosonsas ”standard candles” and used the ATLASdata to obtain a new value of σinelnn .

I Using EPPS16 and NNPDF3.1 (+MCFM)they obtained theory predictions for RAA.By comparing it with the measured RAA vs.η` or yZ it is possible to convert each datapoint to a new value for 〈TAA〉, which viathe Glauber MC are directly related to σinelnn .

I New value from the fit: σinelnn = 41.5+16.2−12.0 mb


Possible shadowing in inelastic nucleon-nucleon cross sectionI Significant change of σinelnn leads to a rather modest change of 〈TAA〉 in central events.

The impact grows towards more peripheral centrality classes, where also other effectssuch as possible centrality dependence of σinelnn or the neutron-skin effect may becomeimportant to explain the data behaviour. 4

0.6

0.8

1.0

1.2

1.4

0 10 20 30 40 50 60 70 80 90 100

0.6

0.8

1.0

1.2

1.4

RPbPbwithσinel

nn

=70mb

Centrality [%]

ATLAS W− ATLAS W+ ATLAS Z data

Original TAA uncertainty

EPPS16min.bias

EPPS16min.bias

√s = 5.02TeV

√s = 5.02TeV

RPbPbwithσinel

nn

=70mb

EPPS16min.bias

0.6

0.8

1.0

1.2

1.4

0 10 20 30 40 50 60 70 80 90 100

0.6

0.8

1.0

1.2

1.4

RPbPbwithσinel

nn

=41.5mb

Centrality [%]


Fitted σinelnn uncertainty

EPPS16min.bias

EPPS16min.bias

√s = 5.02TeV

√s = 5.02TeV

RPbPbwithσinel

nn

=41.5mb

EPPS16min.bias

FIG. 3. The centrality-dependent nuclear modification ratios for W± and Z boson production in Pb+Pb collisions from ATLAS[37, 38] compared to NNLO pQCD calculation with EPPS16 nuclear modification with the nominal value of σinel

nn = 70.0 mb(left) and with the nuclear-suppressed value σinel

nn = 41.5 mb (right).

(close-to) minimum-bias collisions. The impact, however,grows towards more peripheral centrality classes, see Ta-ble I. To illustrate this, Fig. 3 compares the centralitydependent Rexp

PbPb before and after rescaling the data by

〈TAA(σinelpp )〉/〈TAA(σinel

nn )〉 using the fitted σinelnn . The left-

hand panels show the original ATLAS data including thequoted 〈TAA〉 uncertainties, and in the right-hand panelsthe data have been rescaled and the uncertainties followfrom the σinel

nn fit. The striking effect is that the mysteri-ous rise towards more peripheral collisions in the originaldata becomes compatible with a negligible centrality de-pendence, the central values indicating perhaps a mildlydecreasing trend towards peripheral bins. As discussede.g. in the ATLAS publications [37, 38], such a suppres-sion could be expected from selection and geometricalbiases associated with the MC Glauber modeling [54].Also other effects such as possible centrality dependenceof σinel

nn and the neutron-skin effect [55, 56] may becomerelevant to explain the data behaviour in the far periph-ery.

IV. MINIJETS WITH SHADOWING

To study the plausibility of the obtained suppression inσinelnn , we calculate its value in an eikonal model for minijet

production with nuclear shadowing. The model is basedon a similar setup as in Ref. [57] but in the eikonal func-tion we include only the contribution from the hard mini-jet cross section σjet(

√snn, p0, [Q]), calculated at lead-

ing order in pQCD. The transverse-momentum cutoff p0(which depends on

√snn, scale choice Q and the pro-

ton thickness) and the width of the assumed Gaussianproton thickness function we fix so that the model repro-duces σinel

pp = 70 mb matching the COMPETE analysis

[58] at√s = 5.02 GeV. The free proton PDFs are here

CT14lo [59], and we take the nuclear PDF modificationsfrom the EPPS16 [43] and nCTEQ15 [60] analyses. Theresults for σinel

nn , obtained with p0 and proton thicknessfunction width fixed to the the p+p case, are shown inFig. 4. The error bars are again from the nuclear PDFsscaled to the 68% confidence level. As expected at thefew-GeV scales, the predicted σinel

nn depends strongly onthe factorization/renormalization scale Q, but within the

4

0.6

0.8

1.0

1.2

1.4

0 10 20 30 40 50 60 70 80 90 100

0.6

0.8

1.0

1.2

1.4

RPbPbwithσinel

nn

=70mb

Centrality [%]


Original TAA uncertainty

EPPS16min.bias

EPPS16min.bias

√s = 5.02TeV

√s = 5.02TeV

RPbPbwithσinel

nn

=70mb

EPPS16min.bias

0.6

0.8

1.0

1.2

1.4

0 10 20 30 40 50 60 70 80 90 100

0.6

0.8

1.0

1.2

1.4

RPbPbwithσinel

nn

=41.5mb

Centrality [%]


Fitted σinelnn uncertainty

EPPS16min.bias

EPPS16min.bias

√s = 5.02TeV

√s = 5.02TeV

RPbPbwithσinel

nn

=41.5mb

EPPS16min.bias

FIG. 3. The centrality-dependent nuclear modification ratios for W± and Z boson production in Pb+Pb collisions from ATLAS[37, 38] compared to NNLO pQCD calculation with EPPS16 nuclear modification with the nominal value of σinel

nn = 70.0 mb(left) and with the nuclear-suppressed value σinel

nn = 41.5 mb (right).

(close-to) minimum-bias collisions. The impact, however,grows towards more peripheral centrality classes, see Ta-ble I. To illustrate this, Fig. 3 compares the centralitydependent Rexp

PbPb before and after rescaling the data by

〈TAA(σinelpp )〉/〈TAA(σinel

nn )〉 using the fitted σinelnn . The left-

hand panels show the original ATLAS data including thequoted 〈TAA〉 uncertainties, and in the right-hand panelsthe data have been rescaled and the uncertainties followfrom the σinel

nn fit. The striking effect is that the mysteri-ous rise towards more peripheral collisions in the originaldata becomes compatible with a negligible centrality de-pendence, the central values indicating perhaps a mildlydecreasing trend towards peripheral bins. As discussede.g. in the ATLAS publications [37, 38], such a suppres-sion could be expected from selection and geometricalbiases associated with the MC Glauber modeling [54].Also other effects such as possible centrality dependenceof σinel

nn and the neutron-skin effect [55, 56] may becomerelevant to explain the data behaviour in the far periph-ery.

IV. MINIJETS WITH SHADOWING

To study the plausibility of the obtained suppression inσinelnn , we calculate its value in an eikonal model for minijet

production with nuclear shadowing. The model is basedon a similar setup as in Ref. [57] but in the eikonal func-tion we include only the contribution from the hard mini-jet cross section σjet(

√snn, p0, [Q]), calculated at lead-

ing order in pQCD. The transverse-momentum cutoff p0(which depends on

√snn, scale choice Q and the pro-

ton thickness) and the width of the assumed Gaussianproton thickness function we fix so that the model repro-duces σinel

pp = 70 mb matching the COMPETE analysis

[58] at√s = 5.02 GeV. The free proton PDFs are here

CT14lo [59], and we take the nuclear PDF modificationsfrom the EPPS16 [43] and nCTEQ15 [60] analyses. Theresults for σinel

nn , obtained with p0 and proton thicknessfunction width fixed to the the p+p case, are shown inFig. 4. The error bars are again from the nuclear PDFsscaled to the 68% confidence level. As expected at thefew-GeV scales, the predicted σinel

nn depends strongly onthe factorization/renormalization scale Q, but within the

arX

iv:2

003.

1185

6[h

ep-p

h]


Axion-like particles in light-by-light scattering

rapidity of the outgoing X1X2 system. The energy of thephotons is expressed through ω1=2 ¼ Wγγ=2 expð�YX1X2

Þ,in the nucleus-nucleus center-of-mass frame. Experimentalcuts on transverse momenta and rapidities are in the sameframe, which for colliding beams is equivalent to thelaboratory frame (the rest frame of the detectors). Thevariables b1 and b2 are impact parameters of the photon-photon collision point with respect to parent nuclei 1 and 2,respectively, and b ¼ b1 − b2 is the standard impactparameter for the A1A2 collision. The absorption factorS2absðbÞ assures UPC implying that the nuclei do notundergo nuclear breakup. The photon fluxes (Nðωi;biÞ)are expressed through a nuclear charge form factor of thenucleus. In our calculations we use a realistic form factorwhich is a Fourier transform of the charge distribution inthe nucleus. More details can be found e.g., in Ref. [20].We expect that the equivalent photon approximation shouldbe reliable for Wγγ > 0.5 GeV but its accuracy deservesfuture study.The elementary cross section σγγ→X1X2

in Eq. (2.1) for theγγ → γγ scattering is calculated within LO QED withfermion loops (see left panel of Fig. 1 in Ref. [8]). Theone-loop box diagrams were calculated with the help ofthe Mathematica package FORMCALC [21] and theLOOPTOOLS library based on [22] to evaluate one-loopintegrals. In the numerical calculations we include boxdiagrams with leptons and quarks only. At low energies,say Mγγ < 5 GeV, there is no common agreement on thereliability of calculations with quarks/antiquarks. It is notclear which values of their masses should be used. Due tofractional charge of quarks their contribution is smaller thanthat of electrons or muons. However, all contributions addcoherently and due to interference effect they may modify

(enhance) the cross section by up to a factor of 2. At lowenergies one could also include pionic loops. The relevantcalculations are not trivial and require further study. Onewould need to include hadronic form factors since thecharged pions in the loop may be off mass shell but in thiscase it is not clear how to ensure gauge invariance. Pionicloops were discussed recently in the context of hadroniclight-by-light corrections to the anomalous magneticmoment of muon [23–26]. Inclusion of the W-boson loopsis not necessary because this contribution is important onlyfor energies larger than twice the W boson mass [27].Our results have been compared to and agree withRefs. [28–30]. Other production mechanisms were con-sidered in Refs. [8,17], but their contributions are muchsmaller in the low diphoton mass region Mγγ < 5 GeVconsidered here.

A. Pion pair production

The elementary cross section for γγ → ππ was studied indetail in Ref. [9]. Both γγ → πþπ− and γγ → π0π0 reactionswere considered within the physical framework describingexisting experimental data. Two of us calculated, for thefirst time, both the total cross section and angular distri-butions and demonstrated significance of resonances,continuum and pQCD mechanisms in these processes.Following [9], here we include nine resonances, γγ →πþπ− → ρ� → π0π0 continuum, Brodsky-Lepage andhandbag mechanisms. A detailed formalism and descrip-tion of these subprocesses can be found in Ref. [9]. Theangular distribution for the γγ → π0π0 process can bewritten in standard form with the help of the λ1, λ2 photonhelicity-dependent amplitudes, as a function of z ¼ cos θ,where θ is the pion scattering angle

(a) (b)

(c)

FIG. 1. The continuum γγ → γγ scattering (a), γγ → resonances → γγ (b), and the background mechanism (c).

LIGHT-BY-LIGHT SCATTERING IN ULTRAPERIPHERAL … PHYS. REV. D 99, 093013 (2019)

093013-3

Pb

Pb Pb

Pb *( )

*( )

rapidity of the outgoing X1X2 system. The energy of thephotons is expressed through ω1=2 ¼ Wγγ=2 expð�YX1X2

Þ,in the nucleus-nucleus center-of-mass frame. Experimentalcuts on transverse momenta and rapidities are in the sameframe, which for colliding beams is equivalent to thelaboratory frame (the rest frame of the detectors). Thevariables b1 and b2 are impact parameters of the photon-photon collision point with respect to parent nuclei 1 and 2,respectively, and b ¼ b1 − b2 is the standard impactparameter for the A1A2 collision. The absorption factorS2absðbÞ assures UPC implying that the nuclei do notundergo nuclear breakup. The photon fluxes (Nðωi;biÞ)are expressed through a nuclear charge form factor of thenucleus. In our calculations we use a realistic form factorwhich is a Fourier transform of the charge distribution inthe nucleus. More details can be found e.g., in Ref. [20].We expect that the equivalent photon approximation shouldbe reliable for Wγγ > 0.5 GeV but its accuracy deservesfuture study.The elementary cross section σγγ→X1X2

in Eq. (2.1) for theγγ → γγ scattering is calculated within LO QED withfermion loops (see left panel of Fig. 1 in Ref. [8]). Theone-loop box diagrams were calculated with the help ofthe Mathematica package FORMCALC [21] and theLOOPTOOLS library based on [22] to evaluate one-loopintegrals. In the numerical calculations we include boxdiagrams with leptons and quarks only. At low energies,say Mγγ < 5 GeV, there is no common agreement on thereliability of calculations with quarks/antiquarks. It is notclear which values of their masses should be used. Due tofractional charge of quarks their contribution is smaller thanthat of electrons or muons. However, all contributions addcoherently and due to interference effect they may modify

(enhance) the cross section by up to a factor of 2. At lowenergies one could also include pionic loops. The relevantcalculations are not trivial and require further study. Onewould need to include hadronic form factors since thecharged pions in the loop may be off mass shell but in thiscase it is not clear how to ensure gauge invariance. Pionicloops were discussed recently in the context of hadroniclight-by-light corrections to the anomalous magneticmoment of muon [23–26]. Inclusion of the W-boson loopsis not necessary because this contribution is important onlyfor energies larger than twice the W boson mass [27].Our results have been compared to and agree withRefs. [28–30]. Other production mechanisms were con-sidered in Refs. [8,17], but their contributions are muchsmaller in the low diphoton mass region Mγγ < 5 GeVconsidered here.

A. Pion pair production

The elementary cross section for γγ → ππ was studied indetail in Ref. [9]. Both γγ → πþπ− and γγ → π0π0 reactionswere considered within the physical framework describingexisting experimental data. Two of us calculated, for thefirst time, both the total cross section and angular distri-butions and demonstrated significance of resonances,continuum and pQCD mechanisms in these processes.Following [9], here we include nine resonances, γγ →πþπ− → ρ� → π0π0 continuum, Brodsky-Lepage andhandbag mechanisms. A detailed formalism and descrip-tion of these subprocesses can be found in Ref. [9]. Theangular distribution for the γγ → π0π0 process can bewritten in standard form with the help of the λ1, λ2 photonhelicity-dependent amplitudes, as a function of z ¼ cos θ,where θ is the pion scattering angle

(a) (b)

(c)

FIG. 1. The continuum γγ → γγ scattering (a), γγ → resonances → γγ (b), and the background mechanism (c).

LIGHT-BY-LIGHT SCATTERING IN ULTRAPERIPHERAL … PHYS. REV. D 99, 093013 (2019)

093013-3

Pb

Pb Pb

Pb*( )

*)(

a

5 10 15 20 25 30 [GeV]γγm

0

5

10

15

20

25

30

Eve

nts

/ GeV PreliminaryATLAS

= 5.02 TeVNNsPb+Pb Signal region

-1Data, 2.2 nb)γγ → γγSignal (

γγ →CEP gg ee→γγ

Syst. uncertainty

I Measured fiducial cross section (pγT > 2.5 GeV,|ηγ | < 2.4, mγγ>5 GeV, pγγT < 1 GeV, Aco < 0.01):

σfid = 120± 17 (stat)± 13 (sys)± 4 (lumi) nb

I The measured diphoton invariant mass is used to setupper limits at the 95% CL on the ALP productioncross section in the process γγ → a→ γγ and alsoon the ALP coupling to photons (1/Λa).

I Best exclusion limits so far over the mass range of6 < ma < 100 GeV.

20 40 60 80 100 120ma [GeV]

10 1

100

1011/

a [T

eV1 ]

LEP

LEP

CMS [PLB 797 (2019) 134826]

ATLAS (this work)

CDF

LHC

ATLAS Preliminary

Existing constraints from JHEP 1712 (2017) 044

AT

LA

S-C

ON

F-2

020-

010


Summary

Detailed study of the yields of charged-particles inside and around jets in Pb+Pb andpp collisions provided information about the modification of the jet at large distancesfrom the jet axis that can be used to constrain models of how the jet is modified bythe presence of the QGP and how the QGP responds to the jet.

A significant evolution of RAA with√d12 is observed, at small

√d12 values, indicating

a significant difference in the quenching of large-radius jets having single sub-jet andthose with more complex substructure. No significant dependence of RAA on

√d12 is

observed at larger√d12.

Shadowing in nucleon-nucleon cross section in HI collisons might be a possibleexplanation of the underestimate of the measured Z/W boson yields in Pb+Pbby the nPDF predictions as well as the excess of W yields in peripheral collisions.

The measured diphoton invariant mass distribution is used to set new exclusion limitson the production of axion-like particles.

More details and results from the heavy ion physics program realized by ATLAS areavailable on https://twiki.cern.ch/twiki/bin/view/AtlasPublic/HeavyIonsPublicResults

Thank you for your attention!


https://twiki.cern.ch/twiki/bin/view/AtlasPublic/HeavyIonsPublicResults

Backup slides


The ATLAS detectorDetector coverage:Inner Detector (ID):

|η| < 2.5Calorimeter (CAL):

|η| < 3.2 (EM)

|η| < 4.9 (HAD)

3.2 < |η| < 4.9 (FCal)

Muon Spectrometer (MS):

|η| < 2.7Zero Degree Cal. (ZDC):

|η| > 8.3 @z = ±140 m

MB Trig. Scint. (MBTS):

2.1 < |η| < 3.9

Magnetic fields:• 2T solenoid field in ID

• Toroidal field in MS

Identification of minimum-bias p+Pb and Pb+Pb collisions:measurement of spectator neutrons in ZDC and chargedparticle tracks (pulse hight and arrival times) in MBTS.


Nuclear modification factor RAA for the inclusive jetsI The RAA value is independent of rapidity

at low jet pT. For jets with pT & 300 GeVa sign of a decrease with rapidity is observed.

I The magnitude of the RAA as well as itsevolution with jet pT and rapidity areconsistent with those reported in a similarmeasurement performed with Pb+Pbcollisions at

√sNN = 2.76 TeV in the

overlapping kinematic region.

[GeV]T

p

AA

R

0.5

1

40 60 100 200 300 500 90040 60 100 200 300 500 90040 60 100 200 300 500 90040 60 100 200 300 500 90040 60 100 200 300 500 90040 60 100 200 300 500 90040 60 100 200 300 500 90040 60 100 200 300 500 90040 60 100 200 300 500 90040 60 100 200 300 500 900

and luminosity uncer.⟩AA

T⟨

= 2.76 TeV [PRL 114 (2015) 072302]NNs0 - 10%, = 5.02 TeVNNs0 - 10%, = 2.76 TeV [PRL 114 (2015) 072302]NNs30 - 40%, = 5.02 TeVNNs30 - 40%,

ATLAS | < 2.1y| = 0.4 jetsR tkanti-

Phy

s.L

ett.

B79

0(2

019)

108

|y|

|<0.

3)y(|

AA

R|)

/y(|

AA

R

0.6

0.8

1

1.2

< 200 GeVT

p158 < 0 - 10 %,

|y|

0.6

0.8

1

1.2

< 251 GeVT

p200 < 0 - 10 %,

|y|

0.6

0.8

1

1.2

< 316 GeVT

p251 < 0 - 10 %,

|y|0 0.5 1 1.5 2 2.5

0.6

0.8

1

1.2

< 562 GeVT

p316 < 0 - 10 %,

= 5.02 TeVNNs = 0.4 jets, R tkanti-

ATLAS

-12015 Pb+Pb data, 0.49 nb

-1 data, 25 pbpp2015


Charged-particle distributions within and around jetI The observed behavior inside the jet cone, r < 0.4, agrees with the measurement of

the inclusive jet fragmentation functions, where yields of fragments with pT < 4 GeVare observed to be enhanced and yields of charged particles with intermediate pT aresuppressed in Pb+Pb collisions compared to those in pp collisions.

PRC 98 (2018) 024908

)T

p(D

R

0.5

1

1.5

2

0-10%

ATLAS

< 158 GeVjet

Tp126 <

) 2

0.5

1

1.5

2

| < 2.1jet

y|

=0.4 jetsRtkanti-

< 200 GeVjet

Tp158 <

2

0.5

1

1.5

2

< 251 GeVjet

Tp200 <

= 5.02 TeVNN

sPb+Pb,

-10.49 nb

2

0.5

1

1.5

2

< 316 GeVjet

Tp251 <

= 5.02 TeVs , pp

-125 pb

)T

p(D

R

0.5

1

1.5

2

30-40%

) 2

0.5

1

1.5

2

2

0.5

1

1.5

2

2

0.5

1

1.5

2

[GeV]T

p

1 10210

)T

p(D

R

0.5

1

1.5

2

60-80%

[GeV]T

p

1 10210

0.5

1

1.5

2

[GeV]T

p

1 10210

0.5

1

1.5

2

[GeV]T

p

1 10210

0.5

1

1.5

2

I RD(pT) =D(pT)Pb+Pb

D(pT)pp

I D(pT) =1Njet

dnch

dpTI Excess of soft particles

observed in central Pb+Pbcollisions exhibits only asmall pjet

T dependence.

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

r

0

2

4

6)r, T

p (

DR

T,2p

T,1p

< 158 GeVjet

Tp126 <

< 200 GeVjet

Tp158 <

< 251 GeVjet

Tp200 <

< 316 GeVjet

Tp251 <

10%−0ATLAS-1 = 5.02 TeV, 0.49 nbNNsPb+Pb

-1 = 5.02 TeV, 25 pbs pp R=0.4tkanti-

< 2.5 GeVT,1

p1.6 < < 10.0 GeV

T,2p6.3 <

PRC 100 (2019) 064901


Ratios of yields integrated over pT and rI The yields integrated over the pT range of 1−4 GeV and distance r from the jet axis:

Θ(r) =

∫ 4 GeV

1 GeV

D(pT, r) dpT P (r) =

∫ r

0

∫ 4 GeV

1 GeV

D(pT, r′) dpT dr

′

provide a concise look at the enhancement region, using ratios:

RΘ(r) = Θ(r)Pb+Pb/

Θ(r)pp RP (r) = P (r)Pb+Pb/P (r)pp

I Most of the excess of charged-particles with pT in the range 1−4 GeV in Pb+Pbcollisions is observed in the peripherals of the jets (r ≈ 0.4).

I A significant pjetT dependence is seen in the RP (r) distributions for the most-centralPb+Pb collisions.

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

r

0

1

2

3

)r(Θ

R

< 158 GeVjet

Tp126 <

< 200 GeVjet

Tp158 <

< 251 GeVjet

Tp200 <

< 316 GeVjet

Tp251 <



10%−0

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

r

0.5

1

1.5

2

)rP

(R

< 158 GeVjet

Tp126 <

< 200 GeVjet

Tp158 <

< 251 GeVjet

Tp200 <

< 316 GeVjet

Tp251 <



10%−0

PRC 100 (2019) 064901



50 100 150[GeV]12d

0.5

1.0

1.5


, 5.02 TeV-1257 pbpp, -1Pb+Pb 1.72 nb0-10%10-20%20-40%40-60%60-80%

<251 GeVT


SSJ

50 100 150[GeV]12d

0.5

1.0

1.5


, 5.02 TeV-1257 pbpp, -1Pb+Pb 1.72 nb0-10%10-20%20-40%40-60%60-80%

<316 GeVT


SSJ

I Significantly different values of the RAA for SSJcompared to jets with more complex sub-structure.

I RAA sharply decreases with increasing√d12 for small

values of the splitting scale followed by flattening forlarger

√d12.

I This is qualitatively consistent with the effects wherethe medium can not resolve partonic fragments below acertain transverse scale.

50 100 150[GeV]12d

0.5

1.0

1.5


, 5.02 TeV-1257 pbpp, -1Pb+Pb 1.72 nb0-10%10-20%20-40%40-60%60-80%

<398 GeVT


SSJ 50 100 150[GeV]12d

0.5

1.0

1.5


, 5.02 TeV-1257 pbpp, -1Pb+Pb 1.72 nb0-10%10-20%20-40%40-60%60-80%

<500 GeVT


SSJ

AT

LA

S-C

ON

F-2

019-

056



0 100 200 300 400⟩

partN ⟨

0.5

1.0

1.5

AA


Single sub-jet

< 20 GeV12d14<

< 41 GeV12d29<

< 84 GeV12d59<

<251 GeVT


0 100 200 300 400⟩

partN ⟨

0.5

1.0

1.5

AA


Single sub-jet

< 20 GeV12d14<

< 41 GeV12d29<

< 84 GeV12d59<

<316 GeVT


I Continous increase of the suppression with increasingcentrality.

I The SSJ jets are less suppressed with respect to thosewith higher sub-jet multiplicity.

0 100 200 300 400⟩

partN ⟨

0.5

1.0

1.5

AA


Single sub-jet

< 20 GeV12d14<

< 41 GeV12d29<

< 84 GeV12d59<

<398 GeVT


0 100 200 300 400⟩

partN ⟨

0.5

1.0

1.5

AA


Single sub-jet

< 20 GeV12d14<

< 41 GeV12d29<

< 84 GeV12d59<

<500 GeVT

p|<2.0 398< y| 1.0 jetsR =Reclustered A

TL

AS

-CO

NF

-201

9-05

6



200 300 400 500 [GeV]

Tp

0.5

1.0

1.5

AA


0-10%10-20%20-40%40-60%60-80%

|<2.0 Single-sub jety| 1.0 jetsR =Reclustered

200 300 400 500 [GeV]

Tp

0.5

1.0

1.5

AA


0-10%10-20%20-40%40-60%60-80%

<14 GeV12d|<2.0 6< y| 1.0 jetsR =Reclustered

I For SSJ jets, RAA grows slowly with increasing jet pT.

I No significant pT dependence of RAA is observed forjets with multiple sub-jets.

200 300 400 500 [GeV]

Tp

0.5

1.0

1.5

AA


0-10%10-20%20-40%40-60%60-80%

<29 GeV12d|<2.0 20< y| 1.0 jetsR =Reclustered

200 300 400 500[GeV]

Tp

0.5

1.0

1.5


, 5.02 TeV-1257 pbpp, -1Pb+Pb 1.72 nb0-10%10-20%20-40%40-60%60-80%

<121 GeV12d|<2.0 84< y|1.0 jetsR =Reclustered A

TL

AS

-CO

NF

-201

9-05

6


Reconstruction of Z/W bosons with the ATLAS detector

[GeV]eem

2−10

1−10

1

10

210

310

410

Eve

nts/

GeV ATLAS

-1=5.02 TeV 0.49 nbNNsPb+Pb, -e+e → Z

0-100% centrality

Data-e+e → Z

EM backgroundMulti-jet

-τ+τ → ZTop quarks

60 70 80 90 100 110 120 [GeV]eem

0.51

1.52

Pre

d.D

ata

µη2− 1.5− 1− 0.5− 0 0.5 1 1.5 2

µη/d

Nd

1000

2000

3000

4000

5000

6000

Dataν−µ→−W

Multi-jetντ→W

−τ+τ→Z−µ+µ→Z

ttATLAS

=5.02 TeVNNsPb+Pb, , 0-80%-10.49 nb

[GeV]T

µp30 40 50 60 70 80 90 100

[1/G

eV]

Tµp

/dNd

200

400

600

800

1000

1200

1400

Dataν+µ→+W

Multi-jetντ→W

−τ+τ→Z−µ+µ→Z

tt

ATLAS=5.02 TeVNNsPb+Pb,

, 0-80%-10.49 nb

Z/W bosons are measured in their leptonic (` = e, µ)decay channels: W± → `±

( — )

ν and Z → `+`−

Events collected with single-lepton triggers:

peT > 15 GeV and pµT > 14 GeV

High quality reconstruction and isolation (minimizehadronic activity around) requirements on the leptons.

Final kinematic selection cuts on Z/W bosons:

|ηe| < 1.37, 1.52 < |ηe| < 2.47, |ηµ| < 2.4

Specific cuts for Z boson selection:

opposite charge leptons with p`T > 20 GeV,

invariant mass 66 < m`` < 116 GeV.

Specific cuts for W± boson selection:

p`T > 25 GeV, pmissT > 25 GeV,

mT =√

2p`TpmissT (1− cos ∆φ) > 40 GeV,

veto on Z events: m`` < 66 GeV, with p`T,2 > 20 GeV.

Multi-jet bkgr. (semileptonic decays of HF, decays of K-sand π-s in muon channel, photon conversions, misidentifiedhadrons, ...) - estimated from data using template fits.


Z/W boson production in pp collisions at√s = 5.02 TeV

Fiducial integrated and differential cross sections measured in pp collisions serveas a reference for Pb+Pb interactions.

0 0.5 1 1.5 2 2.5

| [pb

]ll

/d|y

σd

20

40

60

80

100

120ATLAS

-1=5.02 TeV 25 pbspp

-l+ l→ Z<116 GeVll66<m

>20 GeVlT

p|<2.5l

η|

Data

NNLO QCDCT14 NNLONNPDF3.1MMHT2014HERAPDF2.0ABMP16

|ll

|y0 0.5 1 1.5 2 2.5

D

ata

NN

LO Q

CD

0.91

1.1

EPJC 79 (2019) 128

Separate measurements in e and µ decay channels.

Combined integrated fiducial cross sections:

W+ : 2266± 9 (stat)± 29 (sys)± 43 (lumi) pb

W− : 1401± 7 (stat)± 18 (sys)± 27 (lumi) pb

Z : 374.5± 3.4 (stat)± 3.6 (sys)± 7.0 (lumi) pb

Overall good agreement with NNLO pQCD calculations.

A` =dNW+→`+ν`/dη` − dNW−→`−ν`/dη`dNW+→`+ν`/dη` + dNW−→`−ν`/dη`

0 0.5 1 1.5 2 2.5

| [pb

]lη

/d|

σd

350

400

450

500

ATLAS-1=5.02 TeV 25 pbspp

ν+ l→ +W>25 GeVνl,

Tp

>40 GeVTm|<2.5l

η|

Data


|l

η|0 0.5 1 1.5 2 2.5

D

ata

NN

LO Q

CD

0.91

1.1

0 0.5 1 1.5 2 2.5

| [pb

]lη

/d|

σd

200

250

300

350 ATLAS-1=5.02 TeV 25 pbspp

ν- l→ -W>25 GeVνl,

Tp

>40 GeVTm|<2.5l

η|

Data


|l

η|0 0.5 1 1.5 2 2.5

D

ata

NN

LO Q

CD

0.91

1.1 0 0.5 1 1.5 2 2.5

|η|

0.16

0.18

0.2

0.22

0.24

0.26

0.28

0.3

0.32lA

ATLAS-1=5.02 TeV 25 pbspp

ν l→ W>25 GeVνl,

Tp

>40 GeVTm|<2.5l

η| Data


0 0.5 1 1.5 2 2.5|l

η|

0.81

1.2 D

ata

NN

LO Q

CD


Binary scaling - Glauber MC comparison

⟩part

N⟨0 50 100 150 200 250 300 350 400

[nb]

evt

N1

⟩

AA

T⟨νl

→W

N

1.2

1.4

1.6

1.8

2


> 25 GeVT

νl,p > 40 GeVTm| < 2.5

lη|

ν+l →+W

MCGlauber v2.4MCGlauber v3.2CT14 + neutron skin

unc.⟩AA

T⟨

⟩part

N⟨0 50 100 150 200 250 300 350 400

[nb]

evt

N1

⟩

AA

T⟨νl

→W

N

1.2

1.4

1.6

1.8

2


> 25 GeVT

νl,p > 40 GeVTm| < 2.5

lη|

ν−l →−W

MCGlauber v2.4MCGlauber v3.2CT14 + neutron skin

unc.⟩AA

T⟨

EPJC 79 (2019) 935

Comparison of Glauber MC v2.4 and v3.2(includes different radial distributions of pand n in nuclei, which results in an evolutionof the effective p/n ratio with centrality).

No significant change of the measured yields.

Theory curves obtained using the CT14NLO PDF set include the neutron skin effectincluded in Glauber MC v3.2.

Slopes of Z boson RAA vs. 〈Npart〉:Glauber v2.4: (10± 7)%; v3.2: (5± 6)%

0 100 200 300 400⟩

partN⟨

300

400

500

[pb]

Z N

-1 ⟩A

A

T⟨ -

1ev

t N

Glauber v2.4Glauber v3.2

Zppσ


-1, 25 pbpp=5.02 TeVs, NNs

0 100 200 300 400⟩

partN⟨

1

1.2

A

AR

MC isospinMC isospin, Glauber v3.2

error scaled up by 3⟩part

N⟨

PL

B80

2(2

020)

1352

62


Comparison with HG-Pythia predictions

Peripheral decrease in production yield predicted by HG-PYTHIA (PLB773 (2017) 408)

In qualitative agreement with charged hadrons RAA suppression observed by ALICE.

Opposite trends observed by ATLAS in Z and W boson yields.

Theoretical predictions are calculated with the CT14 NLO PDF set multiplied by theRAA obtained from HG-PYTHIA.

Centrality (%)0 20 40 60 80 100

AA

R

0

0.2

0.4

0.6

0.8

1

1.2

| < 0.8 η = 5.02 TeV, charged particles, | NN

sPbPb,

c < 20 GeV/T

pALICE data, 8 <

HGPYTHIA, PLB 773 (2017) 408

0 20 40 60 80 100

Centrality [%]

0.4

0.6

0.8

1

1.2

1.4

1.6

AA

R

ATLAS-1=5.02 TeV, 0.49 nbNNsPb+Pb,

-1=5.02 TeV, 25 pbs, pp

Data HG-Pythia (isospin corrected)

-W Z

+W

-W Z

+W

PLB 793 (2019) 420 PLB 802 (2020) 135262


recent results on hard and rare probes from atlas · motivation and recent atlas results i qgp...

Documents