LongRange NearSide Angular Correlations in ProtonProton Interactions in

Long-Range, Near-Side Angular Correlations in Proton-Proton Interactions in CMS First observation in p-p or p-pbar collisions! Kajari Mazumdar LHC Physics Seminar, DHEP, TIFR October 23, 2010.

Ar. Xiv: 1009. 4122 (hep-ex), CERN-PH-EP-2010 -031 Submitted to Journ. of High Energy Physics

Prelude • Trying to understand the analysis and its implications. • We have not participated in this analysis and would like to know more about the analysis. • Several slides are taken from presentations by G. Tonelli (Spokesperson, CMS collaboration) and G. Roland (leader of the analysis) at CERN on September 21, 2010. http: //indico. cern. ch/conference. Display. py? conf. Id=10744 Additionally materials collected from various presentations by Wei Li and private communications. • CMS press release can be obtained from http: //cms. cern. ch • Finally, we are proud to be in CMS collaboration, to have various colleagues with expertise in various aspects.

Summary of the seminars on 21. 9. 2010 Short-range and long-range angular collisions studied in pp collision with CMS at LHC • Observed long-range, near-side correlations in high-multiplicity events. -- signal grows with event multiplicity -- effect is maximal in transverse momentum range of 1 - 3 Ge. V/c • Long-range, near-side correlation is not seen in low multiplicity events and generators, but resembles effects seen in heavy-ion collisions at high energies. • The feature has survived all possible checks for various effects. • Collaboration decided to submit the paper to expose our findings to the scrutiny of the scientific community at large. Since there a number of potential explanations, today’s presentation is focussed on the experimental evidence in the interest of fostering a broader discussion on the subject. The incoming Heavy Ion run of LHC during end of the year will be an additional important test bench.

Preliminaries Starting November 2009, LHC delivered proton-on-proton collisions at 900 Ge. V, 2, 360 Ge. V before moving on to 7000 Ge. V from March 30, 2010. We have few more weeks with pp collision at 7 Te. V, before heavy ion collision. The sample of hard scattered events, till now is too small to really see anything new at high energy scale. Integrated luminosity at present is about 3. 5 /pb. However, even with low luminosity, lower energy operation, LHC is already turning out to be a no-loose machine, albeit, in a different sense! • Physics commissioning studies in CMS experiment have shown that, in general, the detector is behaving extremely well, and the results match Monte Carlo predictions till now. The level of preparedness of the experiment is very high. • No significant discrepancy with known theory observed. The first physics with hadron collision: minimum bias events mainly soft particles (pions, kaons of average momentum few hundred Me. V/c). measurements have to be explained by mostly phenomenological models rather than by perturbative QCD, and the description can be improved with new data. study basic properties at the new kinematic regime offered by LHC. Need high resolution, highly efficient detector to “track” soft particles.

CMS All Silicon Tracker : detection of charged particles (Pixel Detector, Tracker Inner and Tracker Outer, Tracker Endcap, . . ) r- θ view, beam parallel to horizontal axis r-f view, beam out of/into the plane TOB TIB PD TOB TIB TEC TID PD • Coverage up to |h|<2. 5; extremely high granularity, due to the small cell size and high longitudinal segmentation, to keep low occupancy (~ a few%) also at LHC nominal luminosity. • It is the largest Silicon Tracker ever built: Strips: 9. 3 M channels; Pixels: 66 M channels. Operational fractions: strips 98. 1%; pixel 98. 3%

Pixel detector: 3 barrel layers at radii between 4. 4 and 10. 2 cm Single point resolution: 10 mm in r-f, 25 mm in z Silicon detector: 10 barrel layers upto 1. 1 m Resolution of transverse momentum for particles with 1 Ge. V/c: 0. 7% at h = 0, 2% at |h|=2. 5 Primary vertex (requires atleast 3 tracks) to lie within 4. 5 cm of nominal collision point along Z and 0. 15 cm in direction transverse to beam. Track counting done with dz, dxy error ~ 100 mm

D angle between two tracks in the longitudinal plane Df angle between two tracks in the transverse plane

Probing new energy frontier starts with the study of minimum bias events in the new energy regime: charge multiplicity, average momentum distribution, etc. Charged hadron multiplicity in mimimum bias events at different energies. Strategy: make the best possible use of the large eta coverage of CMS detector, the redundancy of the apparatus and the flexibility of trigger system. High multiplicity events are rare. Special trigger needed to collect large statistics sample of high multiplicity events. For primary/online data collection implemented event trigger condition in 2 steps: • total transverse energy in the event > 60 Ge. V • the multiplicity of charged tracks > 70/85 (above pt>400 Me. V/c, |Dh|<2, within dz < 0. 12 cm of a single vertex, with z < 10 cm) Statistics for high multiplicity events enhanced by O(103) Compared to min. bias events.

High multiplicity events The particle densities in the high multiplicity events of proton-proton collisions at 7 Te. V begin to approach those in high-energy collisions of nuclei such as copper. It is natural to study the two particle angular correlations in LHC and compare the results with the ones obtained in relativistic heavy ion colliders like RHIC.

The schematization of the collision is cut into pieces and modeled in different ways, though, actually, the pieces are correlated. 2 -particle correlation probes the connection between various pieces. clusters

The Past and The Future Independent cluster model depicts: 1. independently produced clusters from the initial interaction. 2. Isotropic decay of these clusters in their CM system into hadrons. Clusters: jets, resonances, strings, …, having short range correlation. 2 -particle correlation probes essentially the mechanism of multiparticle production in high energy collision of hadrons. At high energy collisions, the mechanism of hadronisation and possible collective effects due to high particle densities can be studied, Various correlations have been studied extensively in previous experiments at ISR, SPS, RHIC. STAR, PHOBOS experiments at RHIC has reported observation of longrange (high |Dh|) angular correlation at near side (|Df| ~ 0). Heavy Ion collisions at LHC by end of 2010, will provide more opportunity to study the effect in detail.

Types of correlation Angular correlation can be both short- and long-range which characterizes QCD in the energy ranges encountered in the experiment. Short-range correlations in minimum bias events has a typical width of Dh ~ 1. The correlation strength and extent can be parametrized in terms of a simple cluster model and the strength quantified in terms of cluster size (average no. of particles in a cluster) and width (separation of particles in pseudorapidity) However, this does not really lead to basic understanding of the process. The long range correlation, for high |Dh| values, may be significantly affected by the presence of hot and dense matter formed in high energy collisions of hadronic matter. Possible to study till now only in heavy-ion facilities, but now, may be already in proton-proton collisions at LHC. It could as well be manifestation of some jet properties at high energies. The physical origin of long range correlation is not yet well-understood.

Steps to make 2 -particle correlation for a given momentum bin Signal: pairs from same event • Take each event and make all possible 2 -particle pair combination. • For each pair calculate Dh and Dj and fill a histogram. • Then average over all events • Normalization of the histo not too important. However, the distribution is normalised to unit integral in the present analysis. Background: essentially product of 2 single particle distributions. • take random combinatorial pairs, (using event mixing procedure to kill any correlation) • Select 2 events randomly and make pairs using one particle from each event having similar vertex and multiplicity. Take Dh, Dj always positive and fill other quadrants by reflection symmetry around Dh = 0, Dj = 0

Correlation function

Correlation function Calculate ratio for each multiplicity bin first and then weight by average multiplicity of each bin <N>. R is a measure of correlated pairs/pairticle/event. Multiplicity weighting factor: N-1 = total number of pairs/particle/event

Features of correlation plot: 1

Features of correlation plot: 2

Features of correlation plot: 3

2 -particle correlation in minimum bias data

Result from real data: 1 Cut the region of |Dh|<0. 06, |Dj|<0. 06, to reduce secondary effects (tracks from photon conversions, weak decays, event not rejected by impact parameter cut.

Result from real data: 2

Result: 3, the novel feature!

Result from Monte Carlo generated events

Multiplicity and Pt dependence: Turn on of the Ridge

High Multiplicity and Pt dependence: 2 MC has minimum at Dj = 0 Data has local maximum at Dj = 0

Systematic uncertainties • Negligibly small statistical uncertainty. • However the signal is subtle and unexpected. Estimate systematic uncertainties • Is there a way to fake the signal qualitatively? Check for effects due to: • Pile-up + beam background, • Detector noise, acceptance, efficiency • Trigger efficiency, bias • Reconstruction efficiency, fakes • Bugs in analysis code. • Apply data-driven methods No indication of effect that would fake ridge signal.

Like-sign vs. unlike sign Correlations calculated separately for pairs of same sign and opposite signs No dependence on relative charge sign

Analysis code

Track reconstruction code

Trigger

Events background: 1

Events background: 2

Event pileup Pileup effects are suppressed due to excellent resolution.

Analysis with tracks paired with photons

Conclusion CMS experiment has measured 2 -dimensional angular correlations between particles with high |Dh| over full range of Dj for proton-proton collision at cm energies of 0. 9, 2. 36 and 7 Te. V. A variety of features are observed due to short- and long-range correlations. The most interesting feature is the first observation of near-side, long-range correlation in high-multiplicity events at 7 Te. V which resembles the observation at RHIC experiments. The physical origin of the correlation is not yet understood. Further detailed studies are needed. CMS decided to report about the finding to the HEP community.

Backup

Event and track selection

Minimum Bias Data sample at different energies Nevent integrated luminosity energy 168, 854 10, 902 150086 3. 3 mb-1 0. 2 mb-1 3. 0 mb-1 900 Ge. V 2360 Ge. V 7000 Ge. V For high multiplicity analysis data corr. To int. lumi = 980 /nb Data at 7 Te. V Event multiplicity corrected for all detector and reconstruction algorthm

Short range correlation vs. sqrt(s)

Quantifying cluster properties and their energy dependence On average every 2 -3 charged particles, typically pions are produced in a correlated Fashion cluster mass ~ 1 Ge. V Clusters are also getting narrower with increasing energy.

2 -particle correlation

Tracking performance in high multiplicity events p. T>1. 0 Ge. V/c | |<1. 0 o 20<N<35 • 90<N<110 p. T>1. 0 Ge. V/c The CMS Tracker feels at home in high charged particles multiplicity environment. It has been designed to tackle thousand of tracks per event as it will happen with LHC running in pp at nominal luminosity. | |<1. 0 It is foreseen to provide good tracking performance also for the imminent Heavy Ion running of LHC where we expect to have order of 104 tracks per event.

Corrections Event selection efficiency: low for low multiplicity events Triggering vertexing

Tracking/acceptance efficiency correction Overall efficiency 76%, for p. T ~ 100 Me. V/c efficiency = 55% No significant change after correction, since event multiplicity is high

Clusters are partially lost at the edge of Acceptance, correlation affected by 20 -25

Track Selection: Track impact parameter significance for track p. T between 100 to 200 Me. V/c Very good agreement between data and monte carlo down to soft tracks Use MC to determine efficiency of track selection

Energy dependence of cluster properties

Data Collection in p-p collision Dedicated high multiplicity trigger in the two steps. Level 1 (L 1): Sum of the total transverse energy ET (ECAL, HCAL, and HF) > 60 Ge. V. High-level trigger (HLT): number of online tracks built from the three layers of pixel detecto >70 (85). Statistics for high multiplicity events enhanced by O(103). Total datasets corresponding to 980 nb-1

CMS PHOBOS p. T>0. 1 Ge. V/c high multiplicity pp 7 Te. V comparable to ~18 nucleon pairs, each colliding at 62. 4 Ge. V in Cu. Cu -2 +2 The particle densities in the high multiplicity events of proton-proton collisions at 7 Te. V begin to approach those in high-energy collisions of nuclei such as copper. It was considered natural to study the two particle angular correlations in LHC and compare the results with the ones obtained in relativistic heavy ion colliders like RHIC.

Zero Yield At Minimum (ZYAM) Strength of the near side ridge and its dependence on pt and multiplicity Can be quantified by calculating the associated yield: number of other Particles correlated with a specific particle. ZYAM uses R(Dj) integrated between |Dh|=2. 4 and 4. 8. First fit a polynomial in the range of 0. 1 to 2. 0 and find the minimum, FZYAM. Integrated R(Dj) between o to FZYAM and multiply with background integrated over Dh between 2. 0 to 4. 8 Assume that away side jet contribution, the background, negligible. The uncertainty in fitted minimum gives the uncertainty of the associated yield

Detector