Search for the chiral magnetic effect in U+U & Isobar collisions

Prithwish Tribedy

1

CME workshop, RHIC & AGS Annual Users' Meeting, June 7-10, 2016Brookhaven National Laboratory, Upton, NY,USA

(for the STAR Collaboration)

• Results for U+U & Au+Au collisions from STAR

• What else can we try with the existing data

• Outlook for collisions of Isobars @ RHIC

2

STAR Experiment at RHIC

Large Coverage: 0 < φ < 2π, |η| < 1.0 Uniform acceptance: transverse momentum (pT) and rapidity (y) Excellent particle identification capabilities (TPC and TOF)

5

Year √sNN (GeV)

Minimum Bias

Events(106)

2010 62.4 67

2010 39 130

2011 27 70

2011 19.6 36

2014 14.5 20

2010 11.5 12

2010 7.7 4

BES-I Dataset TPC MTD Magnet BEMC BBC EEMC TOF

HFT @ Maria & Alex Schmah

•  M. Anderson et al., Nucl. Instrum. Meth. A 499 (2003) 659 •  W. J. Llope., Nucl. Instrum. Meth. A 661 (2012) S110–S113

12/02/16

Flow is the dominant source of background for signals of CME

How can U+U collisions be used to disentangle the two effects ?

What else can we try : Isobars

Outline

5

B

J

n5

3

Next Step: Can we use U+U collisions to learn about CME ?

What have we learned from the U+U collisions at RHIC ?

5

0.85 0.9 0.95 1 1.05 1.1 1.15

2v

0.02

0.022

0.024

0.026

0.028

0.03

1% Most Central ZDC Selection

U+U Au+Au

⟩Mult⟨Mult/0.85 0.9 0.95 1 1.05 1.1 1.15

2v

0.02

0.022

0.024

0.026

0.028

0.03

0.125% Most Central ZDC Selection

=0.13hardGlauber xGlauber Const. QuarksCGC-IPGlasma

ZDC Centrality Cut1−10 1

310×

slop

e

20−

15−

10−

FIG. 3. (Color online) Top panels: charged particle v2{2} vs.normalized multiplicity within |η| < 1.0. The upper panel isfor the top 1% most central events based on the smallness ofthe ZDC signal, while the middle panel is for the top 0.125%.Small boxes indicate the possible range of variation of v2 fromuncertainties in the efficiency corrections on the x-axis. Modelcomparisons are described in the text. Bottom panel: Theslopes as a function of increasingly tighter ZDC centralityselections. The systematic uncertainties are shown as bands.

slope, which indicates the effect of the impact parame-ter is still prominent (otherwise we expect the Au+Auslope to be nearly flat or even positive). The middlepanel of Fig. 3 shows the 0.125% most central events.The negative slope for Au+Au collisions is smaller inmagnitude, indicating the effects from non-central col-lisions are reduced and the variation in multiplicity inAu+Au collisions is mainly driven by fluctuations. Thebottom panel of Fig. 3 shows how the slopes extractedfrom v2 vs normalized multiplicity evolve with succes-sively tighter ZDC sections. While the slope for Au+Aucollisions becomes less negative, the slope for U+U colli-sions becomes steeper as the centrality selection is tight-ened. This demonstrates that the variation of multiplic-ity in the 0.125% U+U collisions is dominated by thedifferent geometries made possible by the prolate shapeof the uranium nucleus and that tip-tip collisions producemore multiplicity than body-body collisions. Systematicuncertainties shown as bands on the slope were estimated

by varying the fit range and efficiency corrections. Othersources of systematic error are smaller and sub-dominantcompared to the variation due to the range of efficienciesused in the error analysis. Due to large statistical errors,no conclusions could be drawn from studies of v2{4} ver-sus multiplicity in these events. We also measured v3{2}in central collisions and found that v3{2} in the 0.125%most central collisions are (1.410±0.006)×10−2 for U+Uand (1.380± 0.008)× 10−2 in Au+Au collisions (statisti-cal errors only). The slope of v3 vs multiplicity was smalland negative in both systems at about −0.005± 0.002.The U+U data in the top panels of Fig. 3 are com-

pared to the Glauber-xhard model (asssuming v2 =ε2⟨v2⟩/⟨ε2⟩). The ZDC response was modeled by calcu-lating the number of spectator neutrons from the Glaubermodel (accounting for the charge to mass ratio of thenucleus) and folding each neutron with the known ZDCresolution for a single neutron. The Glauber-xhard modelsignificantly over-predicts the observed slope for U+U.This indicates that the variation in multiplicity betweentip-tip collisions and body-body collisions is smaller thananticipated if multiplicity has a significant contributionproportional to Nbin. Given this failure, we investigatetwo alternatives with no explicit Nbin dependence: aconstituent-quark Glauber model (Glauber-CQ) [18, 19]and the IP-Glasma model [17] based on gluon satura-tion [16]. The Glauber-CQ model neglects Nbin andcounts the number of participating constitutent quarksNCQ with each nucleon being treated as three constituentquarks distributed according to ρ = ρ0 exp(−ar) witha = 4.27 fm−1 [19]. This model with σqq = 9.36 mb pro-vides a good description of transverse energy and multi-plicity distributions at RHIC [19] and a better descrip-tion of v2 fluctuations than a nucleon based Glaubermodel [24]. In our simulation, for each NCQ, we samplean NBD with parameters tuned to match the distribu-tions from p+p [25] and Au+Au at 200 GeV (n = 0.76,and k = 0.34 for |η| < 0.5 and n = 2.9 and k = 0.86 for|η| < 1). For both Glauber models we use two sets of pa-rameters for the nuclear geometry, one corresponding tothe more commonly used values [29] (dashed lines) andthe new parameters proposed in Ref. [30] (solid lines).The effect of the different parameter sets is small. TheIP-Glasma and Glauber-CQ model are also compared tothe Au+Au data (Glauber-xhard is left off for clarity) butbecause of significant uncertainty in the actual shape ofa Au nucleus, it is difficult to draw conclusions from thiscomparison.In U+U collisions, both the IP-Glasma model and the

Glauber-CQ model predict slopes closer to the data. Inthe Glauber-CQ model, even though there is no depen-dence on Nbin, the average number of quarks struck ina nucleon (NCQ/Npart) is larger in tip-tip than in body-body collisions so that tip-tip collisions create more mul-tiplicity. This leads to a strong anti-correlation betweenNCQ/Npart and ε2 which in turn translates into a nega-

L. Adamczyk et al. (STAR Collaboration) Phys. Rev. Lett. 115, 222301 (2015)

U+U data contradicts strong binary-collision dependence of multiplicity

• Limitations of two-component model in MC-Glauber : Modifications : Quark-Glauber (nucl-th/0302071, 1509.06727), TRENTO (1412.4708), Shadowed Glauber (1510.01311)

• Evidence of color coherence & CGC like initial state : CGC —> Weak dependence of multiplicity on shape (Schenke, PT, Venugopalan 1403.2232)

• Dominance of fluctuations, small control in triggering shape : 35% variation in dN/dη —> 12%

variation in v2 in (<1% ZDC)

4

4

FIG. 4: (Color online) Elliptic flow and the magnetic field(arbitrary units) in Au+Au and U + U collisions as functionof multiplicity. The arrows indicate the multiplicities corre-sponding to the top 2% of the collision cross section.

nity than Au+Au collisions. First, the relative variationin v2 is almost a factor of 2 larger that that in Au+Aucollisions. Also, the variation in elliptic flow in Au+Aucollision is mostly determined by fluctuations in the ini-tial eccentricity, which are still not very well known. InU+U collisions the elliptic flow variation is mostly due tovariation in orientation of the nuclei at the moment of col-lision. The corresponding estimates have much smalleruncertainty.

While selection of the events based on the number ofspectators is very useful, it seems to be also possible todisentangle CME and background correlations based onlyon the dependence of the signal on charged multiplicity.Figure 4 presents the dependence of the elliptic flow andmagnetic field on charged multiplicity. The elliptic flowdependence is di↵erent for two systems, with U + U col-

lisions exhibiting a characteristic kink (cusp) at multi-plicity ⇠ 1000 [25], reflecting the fact that high(er) mul-tiplicity events have predominantly tip-tip orientation;the latter also leads to a decrease in elliptic flow. Beingmostly determined by correlation of the multiplicity withthe number if participants, the magnetic field has simi-lar dependence on multiplicity for both collision systems.The di↵erence in the dependencies of the magnetic fieldand elliptic flow on charged multiplicity can be used a asa test for the nature of correlations contributing to thesignal.The charge separation dependence on the strength of

the magnetic field can be further studied with collisionof isobaric nuclei, such as 96

44Ru and 9640Zr. These nuclei

have the same mass number, but di↵er by the charge.The multiparticle production in the midrapidity regionwould be a↵ected very little in collision of such nuclei,and one would expect very similar elliptic flow. At thesame time the magnetic field would be proportional tothe nuclei charge and can vary by more than 10%, whichcan results in 20% variation in the signal. Such variationsshould be readily measurable. The collisions of 96

44Ru and9640Zr isotopes have been successfully used at GSI [26] in astudy of baryon stopping. Collisions of isobaric nuclei atRHIC will be also extremely valuable for understandingthe initial conditions, and in particular the initial velocityfields, the origin of directed flow, etc.In summary, the estimates presented in this Letter

show that a detailed analysis of central Au+Au and U+U

collisions should be able to disentangle CME and back-ground correlations contributing to the signal observedby STAR.Discussions with J. Dunlop and P. Filip are grate-

fully acknowledged. This work was supported in partby the US Department of Energy, Grant No. DE-FG02-92ER40713.

[1] E. V. Shuryak and T. Schafer, Ann. Rev. Nucl. Part. Sci.47, 359 (1997).

[2] D. Diakonov, Prog. Part. Nucl. Phys. 51, 173 (2003).[3] D. Kharzeev, R. D. Pisarski and M. H. G. Tytgat, Phys.

Rev. Lett. 81, 512 (1998); D. Kharzeev and R.D. Pis-arski, Phys. Rev. D 61, 111901 (2000).

[4] S. A. Voloshin, Phys. Rev. C 62, 044901 (2000).[5] L. E. Finch, A. Chikanian, R. S. Longacre, J. Sandweiss

and J. H. Thomas, Phys. Rev. C 65, 014908 (2001).[6] D. Kharzeev, Phys. Lett. B 633, 260 (2006).[7] D. Kharzeev and A. Zhitnitsky, Nucl. Phys. A 797, 67

(2007).[8] D. E. Kharzeev, L. D. McLerran and H. J. Warringa,

Nucl. Phys. A 803, 227 (2008).[9] K. Fukushima, D. E. Kharzeev and H. J. Warringa, Phys.

Rev. D 78, 074033 (2008).[10] P. V. Buividovich, M. N. Chernodub, E. V. Luschevskaya

and M. I. Polikarpov, Phys. Rev. D 80, 054503 (2009).[11] M. Abramczyk, T. Blum, G. Petropoulos and R. Zhou,

arXiv:0911.1348 [hep-lat].[12] P. V. Buividovich, E. V. Lushchevskaya, M. I. Polikarpov

and M. N. Chernodub, JETP Lett. 90, 412 (2009) [PismaZh. Eksp. Teor. Fiz. 90, 456 (2009)].

[13] Pand CPodd e↵ects in hot and dense matter, RIKENBNL workshop, Brookhaven National Laboratory, April2010, http://www.bnl.gov/riken/hdm/

[14] S. A. Voloshin, Phys. Rev. C 70, 057901 (2004).[15] B. I. Abelev et al. [STAR Collaboration], Phys. Rev. Lett.

103, 251601 (2009).[16] B. I. Abelev et al. [STAR Collaboration], Phys. Rev. C

81, 054908 (2010).[17] V. Skokov, A. Y. Illarionov and V. Toneev, Int. J. Mod.

Phys. A 24, 5925 (2009).[18] S. Voloshin and Y. Zhang, Z. Phys. C 70, 665 (1996).[19] A. M. Poskanzer and S. A. Voloshin, Phys. Rev. C 58,

1671 (1998).[20] S. A. Voloshin, A. M. Poskanzer and R. Snellings, in

Landolt-Boernstein, Relativistic Heavy Ion Physics, Vol.

Au+Au (ultracentral) ε~0, B~0

U+U (ultracentral) ε≠0, B~0

Reaction plane & B-field direction is strongly correlated in Au+Au—> Not true for U+U

y EP

SP

EPSP

z S P S P

y E P E P

Correlation between B-field & eccentricity

Search for non-zero v2 & zero CME

Can U+U collisions disentangle flow & signals of CME ?

Voloshin 1006.1020

Qualitative picture

5

• General (3-particle) correlator :

• Lowest order (3-particle) charge sensitive correlator :

• The CME correlator :

C112 = �cos((�±1 + ��

2 � 2�3))�

Cm,n,m+n = �cos((m�1 + n�2 � (m + n)�3))�

�a,b � �cos(�a1 + �b

2 � 2�3)�v2{2} � �cos(�a + �b � 2�RP )�

(3P-cumulant method) (event-plane method)

Observables for CME

v2{2}2 = �cos(2(�1 � �2))�

• U+U 193 GeV : Year 2012 (Min-bias/ultra-central)

• Au+Au 200 GeV : Year 2004, 2007 (Min-bias), 2011 (ultra-central)

• Centrality selection :

– TPC uncorrected multiplicity |η|<0.5

– ZDC East & West ADC

• Common QA cuts :

– |Vr|< 2, |Vz| < 20, |Vz - vpdVz|< 2 cm

• Acceptance cuts: |η|<1, 0.2 GeV/c<|pT|

6

Weight estimation : bin in sagitta, η-φ

TPC acceptance (used in this analysis)

Details of the data set

7

0

0.002

0.004

0.006

0.008

0.01

0.012

0.5 1 1.5

Au+Au 200 GeV∆ η>0.025

v 2 {

2}2

∆ η

2.5 x 70-80%1.7 x 60-70%1.2 x 50-60%40-50%30-40%20-30%10-20%5-10%0-5%

-1.5

-1

-0.5

0

0.5

1

1.5

0 10 20 30 40 50 60 70 80

Au+Au 200 GeV Run4,Run11

γa,b x

103

Centrality %

STAR OS (0909.1717)STAR SS (0909.1717)This analysis OSThis analysis SS

Au+Au results —> baseline for measurements in U+U

STAR Preliminary

-1.5

-1

-0.5

0

0.5

1

1.5

0 10 20 30 40 50 60 70 80

Au+Au 200 GeV

γa,b x

103

Centrality %

Run4 OSRun4 SS Run11 OSRun11 SS

Au+Au 200 GeV Run 11 Run 4 (PRC 81 (2010) 054908)

STAR Preliminary

Measurement of v2{2} & γab in Au+Au collisions

8

0

0.002

0.004

0.006

0.008

0.01

0.012

0.5 1 1.5

U+U 193 GeV∆ η>0.025

v 2 {

2}2

∆ η

2.5 x 70-80%1.7 x 60-70%1.2 x 50-60%40-50%30-40%20-30%10-20%5-10%0-5%

pT >0.2 GeV

0

0.002

0.004

0.006

0.008

0 0.5 1 1.5 2

← track merging

40-50 % U+U 193 GeV

v 2 {

2}2

∆ η

DataHBTv2v2 + HBT

STAR Preliminary

�a,b =�cos(2(�a

1 + �b2 � 2�3))�

v2{2}

Removing two major artifacts: • Track merging (apply Δη > 0.025) • Short range -correlations (Gaussian fit)

Measurement of v2{2} in U+U collisions

9

-0.1

0

0.1

U+U 0-5 %

⟨cos

(φa 1

+ φb 2

- 2 φ

3 )⟩ ×

104

same-signopps-sign

U+U 5-10 %

same-signopps-sign

U+U 10-20 %

same-signopps-sign

-0.4

0

0.4 U+U 20-30 %

⟨cos

(φa 1

+ φb 2

- 2 φ

3 )⟩ ×

104

same-signopps-sign

U+U 30-40 %

same-signopps-sign

U+U 40-50 %

same-signopps-sign

-2

0

2

0.5 1 1.5

U+U 50-60 %

⟨cos

(φa 1

+ φb 2

- 2 φ

3 )⟩ ×

104

∆ η

same-signopps-sign

0.5 1 1.5

U+U 60-70 %

∆ η

same-signopps-sign

0.5 1 1.5

U+U 70-80 %

∆ η

same-signopps-sign

STAR Preliminary

Need to remove two major artifacts :

• Track merging • apply Δη > 0.025

• Short-range -correlations • do : (OS - SS)

�cos(2(�a1 + �b

2 � 2�3))�C112 =

�� = ��1,2

Differential measurement of the C112 correlator

10

-12

-8

-4

0

4

8

0 10 20 30 40 50 60 70 80

U+U 193 GeV

γa,b x

104

Centrality %

Gang(1210.5498) OSGang(1210.5498) SSThis analysis OSThis analysis SS

Centrality bins finer than 0-10% is needed to probe the shape of

Uranium

STAR Preliminary

Event plane OSEvent plane SS3P Cumulant OS3P Cumulant SS

Results using Cumulant and Event-plane methods

�a,b � �cos(�a1 + �b

2 � 2�3)�v2{2}

� �cos(�a + �b � 2�RP )�

11

0.1

1

10

100

1000

10000

0 100 200 300 400 500 600 700 800

U+U 193 GeV

Entri

es

Refmult

Min-bias0-2%2-4%4-6%6-8%8-10%

-Method 1 (using RefmultCorr)

1

10

100

1000

10000

0 1000 2000 3000 4000 5000 6000 7000 8000 9000

U+U 193 GeV

Entri

esZDCUnAttenuated (East+West)

Min-bias8-10

6-84-62-40-2

-Method 2 (using ZDCs)

Binning on multiplicity Binning on spectators

Centrality Selection in 0-10% events

After removing track merging and HBT peak

12

Refmult bins ZDC bins

Stronger variation of v2 with multiplicity compared to spectators

0

0.0005

0.001

0.0015

0.002

0.5 1 1.5

U+U 193 GeV, 0-10%∆ η>0.025

v 2 {

2}2

∆ η

8-10%6-8%4-6%2-4%0-2% 0

0.0005

0.001

0.0015

0.002

0.5 1 1.5

U+U 193 GeV, 0-10%∆ η>0.025

0-1%v 2

{2}

2

∆ η

8-10%6-8%4-6%2-4%1-2%STAR Preliminary

STAR Preliminary

Estimation of v2{2} (varying multiplicity & spectators)

13

Observations in 0-10%:

• Strong correlation : nearly linear dependence between γab & v2

• γab~0 for v2≠0 0

0.05 0.1

0.15 0.2

0.25 0.3

0.35 0.4

0.025 0.035 0.045

U+U 193 GeV 0-10%∆η>0.025

(γO

S -γSS

) × 1

04

v2 {2}

RefMult binningZDC binningGang (1210.5498)

STAR Preliminary

Event plane (Refmult)

γab-v2 correlations (varying multiplicity & spectators)

γab-v2 correlations (varying multiplicity & spectators)

14

0 0.05

0.1 0.15

0.2 0.25

0.3 0.35

0.4

0.025 0.035 0.045

U+U 193 GeV 0-10%∆η>0.025

(γO

S -γSS

) × 1

04

v2 {2}

RefMult binningZDC binningGang (1210.5498)

U+U collisions Au+Au collisions

• Observation-I : linear dependence (Δγ-v2)• Observation-II : Δγ =(γOS - γSS)~0 for non-zero v2

STAR Preliminary

Event plane (Refmult)

0

0.5

1

0.01 0.02 0.03 0.04 0.05

Au+Au 200 GeV

U+U 193 GeV(γO

S -γSS

) × 1

04v2 {2}

Run 4 (0909.1717)Run 7 (1302.3802)Run 11 (0-1%)

STAR Preliminary

PRC 81 (2010) 054908PRC 88 (2013) 6, 064911

15

Can model calculations provide some insights ?

0

1

2

3

4

5

6

0.05 0.075 0.1 0.125 0.15

U+U 193 GeV0-10 %

-⟨|eB

2 |cos

(2(Ψ

B-Ψ

2)) ⟩

(fm

)-4

Ellipticity ε2

RefMult binningZDC binningPr

ojec

ted

B-fie

ld

0

1

2

3

4

5

6

-⟨|eB

2 |cos

(2(Ψ

B-Ψ

2)) ⟩

(fm

)-4

Linear dependence (Δγ-v2) + offset also seen in model (B-ε2)

Proxy for γab & v2 —> MC-Glauber model simulation

16

0

1

2

3

4

5

6

0.05 0.075 0.1 0.125 0.15

U+U 193 GeV0-10 %

-⟨|eB

2 |cos

(2(Ψ

B-Ψ

2)) ⟩

(fm

)-4

Ellipticity ε2

RefMult binningZDC binningPr

ojec

ted

B-fie

ld

0

1

2

3

4

5

6

-⟨|eB

2 |cos

(2(Ψ

B-Ψ

2)) ⟩

(fm

)-4

0

0.5

1

1.5

2

0.05 0.075 0.1 0.125 0.15

Y/⟨ Y

⟩ 0-1

0%

ε2

Y = |eB2|Y = cos(2(ΨB-Ψ2))

Linear dependence (Δγ-v2) + offset also seen in model (B-ε2)Central events —> direction of B de-correlates with event-plane

Proxy for γab & v2 —> MC-Glauber model simulation

Can model calculations provide some insights ?

17

0

1

2

3

4

5

6

0.05 0.075 0.1 0.125 0.15

U+U 193 GeV0-10 %

-⟨|eB

2 |cos

(2(Ψ

B-Ψ

2)) ⟩

(fm

)-4

Ellipticity ε2

RefMult binningZDC binningPr

ojec

ted

B-fie

ld

0

1

2

3

4

5

6

-⟨|eB

2 |cos

(2(Ψ

B-Ψ

2)) ⟩

(fm

)-4

Background driven scenario : Δγ 0 when v2 0

B-field driven scenario : Δγ 0 when cos(ΨB-Ψ2) 0 even if v2 ≠0

New challenge for 100% background driven models

Proxy for γab & v2 MC-Glauber model simulation

Can model calculations provide some insights ?

0

0.5

1

0.01 0.02 0.03 0.04 0.05

Au+Au 200 GeV

U+U 193 GeV(γO

S -γSS

) × 1

04

v2 {2}

Run 4 (0909.1717)Run 7 (1302.3802)Run 11 (0-1%)

STAR Preliminary

PRC 81 (2010) 054908PRC 88 (2013) 6, 064911

18

S.Chatterjee & PT (1412.5103)

What else can we try with the existing U+U data ?

10-5

10-4

10-3

10-2

10-1

0 10 20 30 40 50 60 70

P(|

L-R

|)

|L-R| (Spectator asymmetry)

U+U

Au+Au

RandomTip-TipBody-Tip

y

z

21

y

z

y

z

|L−R| ~ |L−R|min |L−R| ~ |L−R|max|L−R|min<|L−R|< |L−R|max

L: left going spectator neutronsR: right going spectator neutrons

A new tuning parameter to disentangle ε2 and B-field

Binning in |L-R| it is possible to trigger body-tip events : B ε2

Interesting geometry of U+U collisions

y

z

21

y

z

y

z

|L−R| ~ |L−R|min |L−R| ~ |L−R|max|L−R|min<|L−R|< |L−R|max

Prolate shape of Uranium can be used to trigger events so that:

Large spectator asymmetry is generated when one nucleus engulf the other : Body-Tip collisions

S. Chatterjee and P. TribedyPhys. Rev. C 92, 011902(R)

• v2 decreases

• B-field increases or remains unchanged

10-5

10-4

10-3

10-2

10-1

0 10 20 30 40 50 60 70

P(|L

-R|)

|L-R| (Spectator asymmetry)

U+U

Au+Au

RandomTip-TipBody-Tip

MC-GlauberL: left going spectator neutrons R: left going spectator neutrons

arXiv:1412.5103

3

Interesting geometry of U+U collisions

y

z

21

y

z

y

z

|L−R| ~ |L−R|min |L−R| ~ |L−R|max|L−R|min<|L−R|< |L−R|max

Prolate shape of Uranium can be used to trigger events so that:

Large spectator asymmetry is generated when one nucleus engulf the other : Body-Tip collisions

S. Chatterjee and P. TribedyPhys. Rev. C 92, 011902(R)

• v2 decreases

• B-field increases or remains unchanged

10-5

10-4

10-3

10-2

10-1

0 10 20 30 40 50 60 70

P(|L

-R|)

|L-R| (Spectator asymmetry)

U+U

Au+Au

RandomTip-TipBody-Tip

MC-GlauberL: left going spectator neutrons R: left going spectator neutrons

arXiv:1412.5103

3

MC-Glauber 0.6

0.8

1

1.2

1.4

1.6

0 10 20 30 40

Y/

⟨Y⟩

|L-R|

0-20%(a)

0 10 20 30 40

|L-R|

20-40%(b)

Y = |eB|2 Cos(2 (ΨB - Ψ2SP))

Y = |eB|2 Cos(2 (ΨB - Ψ2PP))

Y = ε2

0 10 20 30 40

|L-R|

20-40% (L or R <60)

U+U Random

(c)

0.6

0.8

1

1.2

1.4

1.6

0 10 20 30 40

Y/

⟨Y⟩

|L-R|

0-20%(a)

0 10 20 30 40

|L-R|

20-40%(b)

Y = |eB|2 Cos(2 (ΨB - Ψ2SP))

Y = |eB|2 Cos(2 (ΨB - Ψ2PP))

Y = ε2

0 10 20 30 40

|L-R|

20-40% (L or R <60)

U+U Random

(c) 0.6

0.8

1

1.2

1.4

1.6

0 10 20 30 40

Y/

⟨Y⟩

|L-R|

0-20%(a)

0 10 20 30 40

|L-R|

20-40%(b)

Y = |eB|2 Cos(2 (ΨB - Ψ2SP))

Y = |eB|2 Cos(2 (ΨB - Ψ2PP))

Y = ε2

0 10 20 30 40

|L-R|

20-40% (L or R <60)

U+U Random

(c)

0.6

0.8

1

1.2

1.4

1.6

0 10 20 30 40

Y/

⟨Y⟩

|L-R|

0-20%(a)

0 10 20 30 40

|L-R|

20-40%(b)

Y = |eB|2 Cos(2 (ΨB - Ψ2SP))

Y = |eB|2 Cos(2 (ΨB - Ψ2PP))

Y = ε2

0 10 20 30 40

|L-R|

20-40% (L or R <60)

U+U Random

(c)

19

Spectator asymmetry in U+U to disentangle Δγ & v2

Spectator asymmetry —> Triggers event with different B but

same ε2 & vice-versa

0.0

6.0

12.0 ΨRP = Ψ2SP

U+U Random

|L-R|=0

(a)

0-10%10-20%20-30%30-40%

0.0

4.0

8.0

0.1 0.15 0.2 0.25 0.3 0.35

−

⟨ |e

B|2 C

os(2

(ΨB

- ΨR

P )) ⟩ (

fm)−

4

ε2

ΨRP = Ψ2PP

U+U Random

|L-R|=0

(b)

0-10%10-20%20-30%30-40%

Look for similar trend between Δγ & v2 with ZDC asymmetry

0 0.05

0.1 0.15

0.2 0.25

0.3 0.35

0.4

0.025 0.035 0.045

U+U 193 GeV 0-10%∆η>0.025

(γO

S -γSS

) × 1

04

v2 {2}

RefMult binningZDC binningGang (1210.5498)Event plane (Refmult)

STAR Preliminary

S.Chatterjee & PT (1412.5103)

Analysis under progress (challenges : ZDC response to neutrons)

Zr +Zr Ru +Ru Au +Au

20

Outlook for Isobar collisions at RHIC

e-A scattering experimentZr Ru

R0 5.07 5.14a(d) [fm] 0.48 0.46β2 0.06 0.13β4 0 0

Zr RuR0 5.05 5.13a(d) [fm] 0.45 0.45β2 0.18 0.03β4 0.01 0.009

comprehensive model deduction

Ref: Gang Wang, QCD Chirality workshop ‘2016

(Woods-Saxon parameters)

Idea is to change B-field without changing background

Single collision in IP-Glasma model (b=0)

Ru4496

Ru4496+ Zr40

96Zr40

96+

Different B-field with same flow background is expected √s =200 GeV

http://starmeetings.physics.ucla.edu/sites/default/files/gang_wang.pdf

Q. Y. Shou et al arXiv:1409.8375

100

101

102

103

104

105

106

0 200 400 600 800 1000

√sNN=200 GeV

Entri

es

Multiplicity |η|<0.5

U+UAu+AuRu+Ru

Zr+Zr

- Two component MC-Glauber corrected dN/dη - Using parameters compatible to e-A scattering data

Comparisons of multiplicities for centrality estimation

100

101

102

103

104

105

106

0 200 400

√sNN=200 GeV

Entri

esMultiplicity |η|<0.5

Ru+RuZr+Zr

Comparison between different systems —> Zr+Zr & Ru+Ru similar

21

0.8

1

1.2

1.4

1.6

0 10 20 30 40 50 60 70 80

√sNN=200 GeV

Rat

io o

f ⟨ B

2 ⟩Centrality %

Au/RuRu/Zr

20

25

30

35

40

45

50

0 10 20 30 40 50 60 70 80

√sNN=200 GeV

⟨ B2 ⟩

[ fm

-4 ]

Centrality %

U+UAu+AuRu+Ru

Zr+Zr

Estimation of B-field —> t=0, center of participant zone (in vacuum)

Comparisons of the magnetic fields

Approach based on arXiv:1111.1949 & arXiv:1412.5103

Estimation of B-field —> not effected by nuclear deformation Vladimir Skokov : QCD Chirality workshop ‘2016

22

http://starmeetings.physics.ucla.edu/sites/default/files/vladimir_skokov.pdf

CME signal depends on both |B| and direction ΨB

- Signal strength ~ |B2|cos(2(ΨB-Ψ2)) —> Projected field - About 20% difference in Ru+Ru vs Zr+Zr

0

2

4

6

8

10

0 10 20 30 40 50 60 70 80

√sNN=200 GeV

⟨ B2 c

os(2

(ΨB

- Ψ2)

) ⟩ [

fm-4

]

Centrality %

Ru+RuZr+Zr

0.8

1

1.2

1.4

0 10 20 30 40 50 60 70 80

√sNN=200 GeV

Rat

io o

f ⟨ B

2 cos

(2(Ψ

B - Ψ

2)) ⟩

Centrality %

(Ru+Ru)/(Zr+Zr)

Comparisons of the projected magnetic fields

23

Projection for CME correlator in Isobar collisions

• Step-II —> make projections for Ru+Ru & Zr+Zr

0

0.01

0.02

0.03

0 5 10 15 20

√sNN=200 GeV

Au+Au 0-50%

∆γ ×

Npa

rt

⟨ B2 cos(2(ΨB - Ψ2)) ⟩ [ fm-4 ]

Data [PRC88(2013)]A + B xα

Y = �� � Npart = � + � �B2 cos(2(�B � �2))��

• Step-I —> parameterize Au+Au data in terms of B

% Most central0 20 40 60 80

part

* N

γ∆

0

0.01

0.02

Ru+RuZr+Zr

(a)

= 200 GeVNNs

80% bgprojection with 1.2B events

% Most central0 20 40 60 80

(Ru+

Ru

- Zr+

Zr)

γ∆

Rel

. dif.

in

0.1−

0.05−

0

0.05

0.1 = 200 GeVNNs

(b)

)2∈ from σ rel. dif. (5γ∆

rel. dif. (case 1)2∈ rel. dif. (case 2)2∈

Contribution from the background has to be incorporated

STAR BUR 2017-18

24

Isobar ratio gives maximum 5σ (1.2 B) if CME is ~ 20% B-field driven

Significance of the expected signal in Isobar collisions

% Most central0 20 40 60 80

part

* N

γ∆

0

0.01

0.02

Ru+RuZr+Zr

(a)

= 200 GeVNNs

80% bgprojection with 1.2B events

% Most central0 20 40 60 80

(Ru+

Ru

- Zr+

Zr)

γ∆

Rel

. dif.

in

0.1−

0.05−

0

0.05

0.1 = 200 GeVNNs

(b)

)2∈ from σ rel. dif. (5γ∆

rel. dif. (case 1)2∈ rel. dif. (case 2)2∈

Background level (%)0 50 100

2∈ (R

uRu-

ZrZr

) w.r.

t γ

∆R

el. d

if. in

0

0.05

0.1

0.15

Sign

ifica

nce

0

5

10

15

20

25

case 1case 2

54

σ5 = 200 GeVNNs

20 - 60%

projection with 1.2B events

Equivalent running of RHIC : 3.5-weeks for each species (STAR proposal for BUR in Runs 18)

Ref: https://drupal.star.bnl.gov/STAR/system/files/STAR_BUR_Run1718_v19.pdf

STAR BUR 2017-18STAR BUR 2017-18

25

• Charge dependent azimuthal correlations have been studied in U+U & compared to Au+Au results.

• Strong correlation between γab & v2 observed in central events (<10%) with γab ~0 for v2>0 in both U+U and Au+Au collisions.

• New analyses under way to use U+U data to disentangle CME signal from backgrounds (specifically using spectator asymmetry).

• Collisions of Isobar looks very promising : (3.5 x 2) weeks of running with about (1.2 x 2) B events can provide about 5σ confidence of signal/bkg.

26

Summary / Outlook

backup

27

28

reflects the Glauber parameters from comprehensive model deductions (case 2). The difference in the initial eccentricityfrom the Monte Carlo Glauber simulation is also shown in the panel (b).

If our observed charge separation really contains the CME contribution, (Δγ × Npart) should be a function of (eB/mπ2)2cos[2(ΨB − ΨRP)]. Based on the STAR measurements for Au+Au collisions at 200 GeV, we have the projections of(Δγ×Npart) for 400M Ru+Ru and Zr+Zr collisions at 200 GeV as functions of centrality in Fig. 9 (a), assuming that thebackground contributes two thirds of the observed signal. Since previous STAR results have revealed the chargeseparation in both Au+Au and Cu+Cu at both 200 GeV and 62.4 GeV, and 96Ru (96Zr) bridges 197Au and 63Cu, thereshould be no worry about the isobaric nuclei being too small to create the pertinent physics. The error bars in Fig. 9 (a)reflect the statistics of 4 × 108 minimum bias events. The systematic uncertainties in the projections are largely canceledout with the relative difference in γ between Ru+Ru and Zr+Zr, shown in Fig. 9 (b). When we combine the events of20−60% collisions, the difference is 5σ above 0 for both cases of Glauber inputs, which are obtained from e-A scatteringexperiments (case 1) and comprehensive model deductions (case 2). On the other hand, if our charge-separationobservable solely is driven by collective flow, the relative difference in γ should closely follow the relative v2 (oreccentricity) difference, depicted by the blue curves in the panel (b). The relative difference in the initial eccentricitycomes from the Monte Carlo Glauber simulation, and it is highly consistent with 0 for peripheral events, and goesabove/below 0 in central collisions owing to the deformation of the Ru/Zr nucleus. For the centrality range of interest,20−60%, the v2 difference should be very close to zero, so that the related backgrounds stay almost the same for Ru+Ruand Zr+Zr. Therefore, the isobaric collisions provide a unique test to pin down the underlying physics mechanism forthe observed charge separation. As a by-product, v2 measurements in central collisions will discern which informationsource is more reliable regarding the deformity of the Zr and Ru nuclei.

FIG. 10: Magnitude (left axis) and significance (rightaxis) of the relative difference in the projected γ×Npartbetween 9644Ru+9644Ru and 9640Zr+9640Zr at 200 GeV.

FIG. 11: Same as Fig. 10, except with the event planefrom the EPD instead of the TPC. The EP{EPD}resolution is assumed to be 80%.

When a different background level is assumed, the magnitude and significance of the projected difference in γbetween Ru+Ru and Zr+Zr change accordingly, as shown in Fig. 10. The future measurements of the isobaric collisiondata will determine whether there is a finite CME signal observed in the γ correlator, and if the answer is “yes”, willascertain the background contribution, when compared with this figure. Up to now, the projections have been made withthe event plane reconstructed with the STAR TPC. By 2018 when the isobaric collisions are supposed to occur, STARwill have installed the Event Plane Detector (EPD), which will provide an independent event plane fromforward/backward rapidity regions. The γ correlator with the EPD event plane will better control the systematics due tothe three-particle correlation from clusters. Fig. 11 shows the magnitude and significance of the projected difference in γwith the event plane reconstructed from the EPD with 400M events for each collision system at 200 GeV. Assuming theEPD event plane resolution is 80% over the centrality range of interest, we will be able to obtain the CME signal with asignificance better than 5.5σ if the background contribution is less than two thirds. If the EP{EPD} resolution turns outto be only 60% instead of 80%, we will need 700M events for each collision system to achieve the same significancelevel.Table I lists the expected relationship between Ru+Ru and Zr+Zr in terms of experimental observables for elliptic

flow, CME, CMW and CVE, assuming that the chiral effects are major physics mechanisms for the correspondingobservables. With this assumption for the CMW observable, we have carried out a 700M-event projection for the slopeparameter r, and found the r ratio of Ru+Ru over Zr+Zr to be 1.08 ± 0.08 for 20−60% collisions, which is only a 1σ

Significance of the expected signal in Isobar collisions

Study done by Gang Wang

Projection for 400M events

29

Spectator asymmetry in U+U collisions

Experimental challenges :

• Response of ZDC to neutrons

• Clustering of nucleons

that introduces artificial de-correlation

Body-Tip events are experimentally triggered by asymmetry of ZDCs

Analysis in this direction (separating signals of flow & CME) and systematic studies are under progress

A few QA plots

30

0

0.002

0.004

0.006

0.008

0.01

0.012

0.014

0.016

-0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5En

tries

Sagitta

0

0.005

0.01

0.015

0.02

0.025

0.03

0 0.5 1 1.5 2 2.5 3 3.5

Entri

es

DCA

10-7

10-6

10-5

10-4

10-3

10-2

0 100 200 300 400 500 600 700 800

Entri

es

Refmult

0.0025

0.003

0.0035

0.004

0.0045

0.005

0.0055

0.006

0.0065

-40 -30 -20 -10 0 10 20 30 40

Entri

es

Vz [cm]

� =C

(1. + exp(�(pT + 0.1)/0.15))

�

0

10

20

30

40

50

60

70

80

90

100

-0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5

Entri

es x

106

Sagitta

• Acceptance binning for weight calculation :

Sagitta = charge*((20.*pT/3.) - √((20.*pT/3.)2 - 0.752))

Weight = 1/(entries in η-φ) * 1/

The tracking efficiency :

31

Weight estimation for cumulant calculations

32

B-field simulations : Dominance of fluctuations

t=0, x=<x>, y=<y>, z=0, U+U collisions