Journal of Physics: Conference Series First Results of Searches for New Physics at TeV with the detector To cite this article: Paolo Azzurri J. Phys.: Conf. Ser. 5 75 View the article online for updates and enhancements. Related content - New Physics with the ATLAS detector: experimental prospects Giovanni Sirausa and the Atlas collaboration - Searches for supersymmetry with the detector at the LHC Edmund Widl and the Cms collaboration - A search for new physics in dijet mass and anular distributions in pp collisions at s=7 TeV measured with the ATLAS detector G Aad, B Abbott, J Abdallah et al. This content was downloaded from IP address 7.44.. on //8 at 9:6
Journal of Physics: Conference Series 5 () 75 doi:.88/74-6596/5//75 First Results of Searches for New Physics at s=7 TeV with the detector Paolo Azzurri Scuola Normale Superiore, Piazza dei Cavalieri 7, 566 Pisa, Italy E-mail: paolo.azzurri@cern.ch Abstract. First searches for new physics phenomena usin the LHC 7 TeV proton-proton collision data collected by the detector in are reviewed. Results are presented of searches for new physics in events with hadronic jet pairs, and for heavy stable chared particles, includin a dedicated search for lon-lived particles that stop in the detector and decay in periods between beam crossins.. Introduction Durin the LHC runnin, the experiment [] recorded over 4 pb of pp collisions at 7 TeV center-of-mass enery. The lare momentum transfers accessible at these collision eneries allow to test for new short rane physics scenarios beyond the predictions of Quantum Chromodynamics (QCD) in dijet events, and to search for the production of new heavy quasistable particles []. In the followin, results for the search of new physics in dijet events are presented for resonances in the dijet mass spectrum in Section, and for new interactions affectin dijet anular distributions in Sections and 4. The outcome of searches for eneric heavy stable chared particles are iven in Section 5, while the pursuit of lon-lived luinos that stop and decay in the detector is detailed in Section 6.. Narrow dijet resonances In proton-proton collisions QCD predicts that the invariant mass spectrum of the two jets with larest p T falls steeply and smoothly, while new massive objects that couple to quarks (q) and luons () would show up as resonant structures in the dijet mass. A data sample of.9±. pb is employed for this analysis [], usin a sinle-jet trier with a transverse enery threshold of 5 GeV, that is measured to be over 99.5% efficient for dijet masses above GeV. Jets are reconstructed with the anti-k T alorithm [4] and R =.7. The reconstructed jet enery E is corrected as a function of p T and η for the non-linearity and inhomoeneity of the calorimeter response [5]. The two jets with the leadin p T define the dijet system. The inclusive dijet mass is shown in Fiure, where the data points are compared with a prediction from PYTHIA [6], that is in areement within the % jet enery scale uncertainty. The data is also compared with a smooth fit to a functional form with four free parameters, and there is no indication of narrow resonances. The dijet mass data points, the fitted backround parameterization, and the expected dijet resonance shapes are used to set specific Bayesian limits on new particles decayin to parton Published under licence by Ltd
χ / ndf. / Prob.4 p.69e-6 ± 5.98e-8 p 5.77 ±.77 p 6.994 ±.648 p.658 ±.658 Symposium on Prospects in the Physics of Discrete Symmetries Journal of Physics: Conference Series 5 () 75 doi:.88/74-6596/5//75 dσ/dm (pb/gev) 4 - - -4 q* (.5 TeV) (.9 pb ) Fit % JES Uncertainty QCD Pythia + Simulation Excited Quark Strin S ( TeV) q* (.5 TeV) η <.5 & η <. S ( TeV) 5 5 Dijet Mass (GeV) Fiure. Dijet mass spectrum. The data points are superimposed with a smooth fit (solid line) and to QCD predictions [6] (shortdashed line). The shaded band shows the jet enery scale uncertainty. Possible excited quark sinals and strin resonance sinals are also shown. Graph σ BR A (pb) 4 5 (.9 pb ) - η <.5 & η <. 95% CL Upper Limit Gluon-Gluon Quark-Gluon Quark-Quark Resonance Models Strin Excited Quark Axiluon/Coloron E 6 Diquark W Z RS Graviton 5 5 5 Resonance Mass (GeV) Fiure. Production upper limits for resonances decayin to luon-luon, quarkluon, and quark-quark dijets, compared to theoretical predictions from different models. pairs. The obtained 95% confidence level (CL) upper limits on the resonances production are shown in Fiure and compared to the expected productions in different models. The resultin excluded mass ranes for strin resonances [7], excited quarks [8], axiluons [9], colorons [] and E 6 diquarks [] all extend previous exclusions [], while for new aue bosons W and Z [], and Randall-Sundrum ravitons [] no mass interval is excluded with the present data.. Dijet centrality ratio In proton-proton collisions the QCD dijet production is dominated by t-channel processes and peaks in the forward direction (alon the beam axis) while new physics scenarios, includin models of quark compositeness, predict more isotropic anular distribution, leadin to enhanced jet production at smaller pseudorapidity values. The dijet centrality ratio R η between the number of events with the two leadin jets in the reion η <.7 and the number of events with the two jets in the reion.7 < η <., is defined to use less anular information than a full dijet anular distribution analysis, but allows for fine binnin in the dijet mass and cancels many sources of systematic uncertainty, providin an accurate test of QCD and probe to new physics. The analysis [4] is based on the same data sample, events selection and jet reconstruction used for the previously described dijet resonances search []. Fiure shows the predicted and observed values of R η a a function of the dijet mass m jj. The measured R η is nearly flat with a value of about.5, as predicted by QCD, while the presence of quark contact interactions would lead to an increase of R η above a value of m jj that depends on the compositeness enery scale Λ. A lo-likelihood-ratio R LL is used to test for the presence of quark compositeness, and, iven the consistency of the data with the QCD hypothesis, a modified frequentist method with the CL s approach [5], is employed to derive the 95% CL limit Λ > 4. TeV, as displayed in Fiure 4
Journal of Physics: Conference Series 5 () 75 doi:.88/74-6596/5//75 R η.6.4..8.6.4..9 pb Null Hypothesis Syst. Uncertainty Λ = TeV Λ = 4 TeV 5 5 5 Dijet Mass (GeV) R LL -5-5 - -5 -.9 pb Limit:! > 4. TeV 95% CL s SM SM ± " SM ± " 4 5! (GeV) Fiure. The observed dijet centrality ratio as a function of m jj compared with QCD predictions (solid line), with the systematic uncertainty band, and to two hypotheses of quark contact interactions with Λ =, 4 TeV. Fiure 4. Exclusions for the contact interaction scale Λ. We show R LL versus Λ for the data (solid line), the 95% CL s (dashed line), and the SM expectation (dotted line) with σ (dark) and σ (liht) bands. 4. Dijet anular distributions The complete dijet anular distribution is measured thouh the differential cross section (/σ dijet )(dσ dijet /dχ dijet ) where χ dijet = (+ cos θ )/( cos θ ), and θ is the polar scatterin anle in the dijet center-of-mass (CM) frame [6]. This choice is motivated by the fact that dσ dijet /dχ dijet is rather flat for QCD (Rutherford) parton scatterin. A data sample of up 6±4 pb is employed for this analysis, and events are selected with five inclusive sinle-jet triers, with different jet transverse momentum p T thresholds. Offline jets are reconstructed with the anti-k T alorithm [4] with R =.5, and the reconstructed jet enery is corrected [5]. The differential dijet anular distributions for different m jj ranes, normalized to their respective interals, are shown in Fi. 5. The data are compared with perturbative QCD (pqcd) predictions at next-to-leadin order (NLO), for all m jj ranes. Given the ood areement with pqcd, the data are used to to set limits on quark compositeness represented by a four-fermion contact interaction term in addition to the QCD Laranian. As in the previous section, a statistical test based on the lo-likelihood-ratio R LL is used to derive limits on Λ, usin the modified frequentist CL s approach [5]. The results are shown in Fiure 6, and a 95% confidence level bound Λ >5.6 TeV is observed, the currently most strinent limit on the contact interaction scale of left-handed quarks. 5. Heavy Stable Chared Particles Many extensions of the standard model predict the existence of heavy (quasi-)stable chared particles (HSCPs), that would travel across the size of typical particle detector [], and could be observable as hih momentum (p) particles with exceptional ionization enery loss (de/dx). The search [7] is based on an interated luminosity of. pb, and has two approaches. A tracker-only selection looks for chared particles in the inner tracker with lare de/dx and p T. A tracker-plus-muon selection additionally requires associated hits in the outer muon
Journal of Physics: Conference Series 5 () 75 doi:.88/74-6596/5//75 dσ dijet /dχ /σ dijet dijet.7.6.5.4... QCD prediction Theory uncertainty Λ = 5 TeV M >. TeV (+.5) jj.8 < <. TeV (+.4) 4 6 8 4 6 χ.4 < <.8 TeV (+.). < <.4 TeV (+.5).85 < <. TeV (+.).65 < <.85 TeV (+.5).5 < <.65 TeV (+.).5 < <.5 TeV (+.5) L = 6 pb.5 < <.5 TeV dijet Fiure 5. Normalized dijet anular distributions in m jj ranes, shifted vertically (as iven in parentheses) for clarity. The data points include full uncertainty bars and are compared with the predictions of pqcd at NLO (solid historam) and with the predictions includin a contact interaction term of compositeness scale Λ = 5 TeV (dashed historam). The shaded band shows the effect on the NLO pqcd predictions due to factorization and renormalization scale variations and parton distribution functions uncertainties, as well as the uncertainties from the nonperturbative corrections added in quadrature. CL S..8.6.4 L = 6 pb observed expected expected ± σ expected ± σ 4 5 6 Λ [TeV] Fiure 6. The observed CL s (solid) and expected CL s (dashed) with one (two) standard deviation indicated by the dark (liht) band as a function of the contact interaction scale Λ, from the dijet anular distributions. The 95% confidence level limits on Λ are extracted from the intersections of the CL s lines with the horizontal line at CL s =.5. detectors. For both selections, the mass of the candidate is estimated from the measured p and de/dx. For candidate tracks with η <.5 and p T > 5 GeV/c, an estimator I as of the deree of compatibility of the observed chare measurements in the silicon detector with the minimum ionizin particle (MIP) hypothesis is used to point out non-relativistic HSCP candidates. The data and Monte Carlo (MC) normalized distributions of p T and I as for pre-selected trackeronly candidates are shown in Fiure 7. One MC sample contains events from standard QCD processes, while the second MC sample contains sinal events from pair-production of stable with a mass of GeV/c. Both samples are enerated with PYTHIA [6]. 4
Journal of Physics: Conference Series 5 () 75 doi:.88/74-6596/5//75.6 pb.6 pb fraction of tracks/ GeV/c - - Tracker - Only MC - MC - QCD fraction of tracks/. - - Tracker - Only MC - MC - QCD -4-4 -5-5 -6 5 p (GeV/c) T -6-7..4.6.8 I as Fiure 7. normalized distributions of p T (left) and I as (riht), compared to simulations from standard QCD processes and from pair-production of with a mass of GeV/c..6 pb σ (pb) Theoretical Prediction luino (NLO+NLL) stop (NLO+NLL) 95% C.L. Limits luino; % luino; 5% luino; % ; ch. suppr. stop stop; ch. suppr. 4 6 8 Mass (GeV/c ) Fiure 8. HSCP theoretical production cross section and observed 95% CL upper limits for different combinations of models and scenarios: pair production of supersymmetric stop and luinos; different fractions, f, of R-luonball hadronization states and chare suppression scenarios [7]. Only the results obtained with the most sensitive selection are reported: tracker-only for the chare suppression scenarios and tracker-plus-muon for all other cases. The bands represent the theoretical uncertainties on the cross section values. The mass of the candidates is extracted from the measured de/dx and p values, invertin a parametrized Bethe-Bloch fitted from data. The search is performed as a countin experiment with candidates required to have I as and p T reater than threshold values, and the candidates mass to be in the rane of 75 to GeV/c. No sinal candidates are observed in the data, with.74 ±. expected form standard processes for the tracker-only selection, and.5 ±.4 for the tracker-plus-muon selection. Therefore cross section upper limits at the 95% CL are set on the HSCP production for two benchmark scenarios: direct production of pairs and t pairs. The sinal efficiencies depend on the hadronization fractions and interactions of the resultin sinal R-hadrons. A summary of cross section upper limits, set with a Bayesian method, are shown in Fiure 8. The trackerplus-muon selection set a 95% CL lower mass limit of 98 GeV/c for lon-lived with a f =. 5
Journal of Physics: Conference Series 5 () 75 doi:.88/74-6596/5//75 fraction of R-luonball states, and of GeV/c for lon-lived t. In a pessimistic scenario of complete chare suppression the tracker-only selection yields a mass limit of GeV/c. 6. Search for stopped luinos A sinificant fraction of slow chared R-hadrons, resultin from the production of lon-lived luinos, can lose sufficient enery to be stopped inside the detector and may decay at later times (with χ ) resultin in jet-like sinals in the calorimeters, out-of-time with respect to LHC collisions. The presence of a stopped-particle sinal is searched for in 6 hours of trier live-time, followin collision data correspondin to pb recorded with cm s peak luminosity. A control sample of 95 hours of trier live-time collected durin fills with 7 7 cm s luminosity is used to estimate the backround rates [8]. A dedicated trier is used that requires a jet-like sinal when no beam is crossin, and outside a 75 ns window around beam crossins. Care is taken to reject beam-halo events, cosmic rays and calorimeter instrumental noise. Both a countin experiment and a time-profile analysis are performed for candidate events, buildin a probability density function (PDF) for the luino decay sinal. Fiure 9 shows an example of such a PDF for a µs lifetime, toether with the in-orbit positions of two candidate events..6.4. () Sinal PDF (τ = µs).8.6.4. 5 5 5 5 BX...8.8.6.6.4.4.. 5 5 4 45 5 55 6 5 5 4 45 BX BX Fiure 9. In-orbit positions of two observed events recorded durin an LHC fill with 4 collidin bunches. The decay profile for a µs lifetime hypothesis is overlaid. The bottom panels are zoomed views of the boxed reions around the two events. Finally no excess of sinal events over the expected backround rates is observed in the search sample, for any lifetime hypothesis, so that 95% CL limits are set on σ(pp ) BR( χ ) for a iven mass difference m m χ > GeV/c, and are summarized in Fiure. For m = GeV/c lifetimes from 75 ns to 5 s are excluded with the countin experiment. The excluded reions as a function of the luino mass are plotted in Fiure. For lifetimes between µs and s m < 7 GeV/c is excluded, and if only electromanetic (EM) interactions are considered the limit becomes m < GeV/c. References [] R. Adolphi et al. [ Collaboration], JINST (8) S84. [] M. Fairbairn et al., Stable massive particles at colliders, Phys. Rept. 48 (7), [] V. Khachatryan et al. [ Collaboration], Search for Dijet Resonances in 7 TeV pp Collisions at, Phys. Rev. Lett. 5 () 8 [arxiv:. [hep-ex]]. [4] M. Cacciari, G. P. Salam, and G. Soyez, The anti-k(t) jet clusterin alorithm, JHEP 4 (8) 6. [5] Collaboration, Jet Performance in pp Collisions at s=7 TeV, Physics Analysis Summary -PAS-JME-- (). [6] T. Sjostrand et al., Hih-enery-physics event eneration with PYTHIA 6., Computer Physics Commun. 5 () 8. 6
Journal of Physics: Conference Series 5 () 75 doi:.88/74-6596/5//75 ) [pb] χ σ(pp ) BR( 4 L dt = pb -7 max inst L = x cm - s m - m χ -6 = GeV/c -5-4 - 95% C.L. Limits: Expected: Countin Exp. Expected ±σ: Countin Exp. Expected ±σ: Countin Exp. Obs.: Countin Exp. Obs.: Countin Exp. (Neutral R-Baryon) Obs.: Countin Exp. (EM only) Obs.: Timin Profile - NLO+NLL (m = GeV/c ) 4 5 6 τ [s] Fiure. Expected and observed 95% CL limits on luino pair production cross section times branchin fraction as a function of the luino lifetime from the countin experiment and the time-profile analysis. Alternative R- hadron interaction scenarios are also presented. The NLO+NLL production expectation for m = GeV/c is also indicated. ) [pb] χ σ(pp ) BR( 4 L dt = pb max inst L = x cm - s m - m χ Not Sensitive = GeV/c 95% C.L. Limits Expected: µs - s Countin Exp. Expected ±σ: µs - s Countin Exp. Expected ±σ: µs - s Countin Exp. Obs.: µs - s Countin Exp. (EM only) Obs.: µs - s Countin Exp. Obs.: µs Timin Profile NLO+NLL 5 5 5 4 45 5 [GeV/c ] m Fiure. Upper 95% CL limits on luino pair production cross section times branchin fraction as a function of the luino mass with stron or EM only R-hadron interactions. The m m χ difference is set at GeV/c and m χ > 5 GeV/c is assumed. The NLO+NLL expected production are also shown. The indicated lifetimes maximize the countin experiment and time-profile analysis sensitivity. [7] L. A. Anchordoqui, H. Goldber, S. Nawata et al., Jet sinals for low mass strins at the Lare Hadron Collider, Phys. Rev. Lett. (8) 48. [8] U. Baur, I. Hinchliffe, and D. Zeppenfeld, Excited Quark Production at Hadron Colliders, Int. J. Mod. Phys. A (987) 85. [9] P. H. Frampton and S. L. Glashow, Chiral color: An alternative to the standard model, Phys. Lett. B9 (987) 57. [] E. H. Simmons, Coloron phenomenoloy, Phys. Rev. D55 (997) 678. [] J. L. Hewett and T. G. Rizzo, Low-enery phenomenoloy of superstrin-inspired E 6 models, Phys. Rept. 8 (989) 9. [] E. Eichten, I. Hinchliffe, K. Lane et al., Supercollider physics, Rev. Mod. Phys. 56 (984) 579. [] L. Randall and R. Sundrum, An alternative to compactification, Phys. Rev. Lett. 8 (999) 469. [4] V. Khachatryan et al. [ Collaboration], Search for Quark Compositeness with the Dijet Centrality 7
Journal of Physics: Conference Series 5 () 75 doi:.88/74-6596/5//75 Ratio in pp Collisions at s=7 TeV, Phys. Rev. Lett. 5 () 6 [arxiv:.449 [hep-ex]]. [5] A. L. Read, Modified frequentist analysis of search results (The CL(s) method), in First Workshop on Confidence Limits, CERN, Geneva, Switzerland, pp. 8. Jan,. [6] V. Khachatryan et al. [ Collaboration], Measurement of Dijet Anular Distributions and Search for Quark Compositeness in pp Collisions at 7 TeV, arxiv:. [hep-ex]. [7] V. Khachatryan et al. [ Collaboration], Search for Heavy Stable Chared Particles in pp collisions at s=7 TeV, arxiv:.645 [hep-ex]. [8] V. Khachatryan et al. [ Collaboration], Search for Stopped Gluinos in pp collisions at, Phys. Rev. Lett. 6 () 8 [arxiv:.586 [hep-ex]]. [9] W. Beenakker et al., Soft-luon resummation for squark and luino hadroproduction, JHEP (9) 4, 8