February7,2008 8:15 ProceedingsTrimSize: 9inx6in ringberg BEYOND THE STANDARD MODEL AT HERA: STATUS 6 AND PROSPECTS 0 0 2 E. PEREZ ∗ n a CE-Saclay, DSM/DAPNIA/Spp, F-91191 Gif-sur-Yvette J and DESY, Notkestrasse 85, D-22607 Hamburg. 7 E-mail: [email protected] 1 1 An overview of experimental results on searches for new phenomena at HERA is v presented. The complementarity with searches performed at other experiments 8 is discussed and the prospects for a discovery, using the full HERA data to be 2 delivereduntilmid-2007,arepresented. 0 1 0 6 1. Introduction 0 / Althoughremarkablyconfirmedsofarbylowandhighenergyexperiments, x the StandardModel(SM)ofstrong,weakandelectromagneticinteractions e - remainsunsatisfactory. Manymodelsof“newphysics”havebeenproposed p to address the questions which are unexplained by the SM. Experimen- e h tally, the observation of deviations with respect to the SM predictions is a : v key part of the existing and future high energy programmes. At HERA, i searchesfornewphenomenahavebeencarriedoutusingaluminosityofup X −1 to 280pb . Besides anexcessofatypicalW-likeeventsobservedatH1all r a measurements are so far in good agreement with the SM expectation and constraints on models for new physics have been obtained. 2. Searches for new phenomena in inclusive DIS NeutralCurrentDeepInelasticScattering(NC DIS)is measuredatHERA for values of the photon virtuality Q2 up to about 40000GeV2. Although the precision is still statistically limited at the highest Q2, the good agree- ment of the measurements with the SM expectation allows stringent con- straintsonnewphysicstobeset1. Forexample,afinitequarkradiuswould reduce the high Q2 DIS cross section with respect to the SM predictions, such that the current data rule out quark radii larger than 0.85 10−18 m, · assuming that the electron is point-like. Similarly, the effective scale of ∗Talk given at the Workshop “New Trends in HERA Physics”, Ringberg (Germany), October 2-72005. 1 February7,2008 8:15 ProceedingsTrimSize: 9inx6in ringberg 2 eeqq contact interactions is constrained to be larger than typically 5TeV - a similar sensitivity being achieved from the preliminary Run II Drell-Yan data of the Tevatron experiments 2. The effective mass scale associated to theexchangeofKaluza-Kleingravitonsinmodelsassumingadditionallarge space dimensions is constrained to be larger than about 0.8TeV, slightly below the LEP and Tevatron bounds. The longitudinal polarisation of the lepton beam in the HERA II data has been exploited to measure the polarisationdependence of the Charged Current (CC) DIS cross section 3. Although these data constrain in prin- ciple the existence of a right-handed W boson, the sensitivity is below the W mass bounds obtained at the Tevatron. However these measurements R confirm the left-handed nature of the weak interaction in the t-channel. 3. Model-dependent searches AsHERAisnotanannihilationmachine,thepair-productionofnewheavy particles, which could occur in e+e− or pp collisions via their coupling to gauge bosons, has a very low cross-section at HERA. Instead, searches for the single production of new particles are performed at the H1 and ZEUS experiments. The cross-section for such processes depends on the unkown coupling of the new particle to the SM fields. Hence these searches do not provide absolute constraints on the mass of new particles. Conversely, the observationofasignalforthesingleproductionofanewparticlewouldpro- vide information not only on its mass, but also on this unknown coupling. In the following example constraints obtained on leptoquarks, squarks in R-parity violating supersymmetry and excited fermions are presented. 3.1. Lepton-quark resonances: Leptoquarks An intriguing characteristic of the Standard Model is the observed sym- metry between the lepton and the quark sectors, which is manifest in the representation of the fermion fields under the SM gauge groups, and in their replication over three family generations. This could be a possible indication of a new symmetry between the lepton and quark sectors, lead- ing to “lepto-quark” interactions. Leptoquarks (LQs) are new scalar or vector colour-triplet bosons carrying both a baryon and a lepton number. Severaltypes ofLQscanbe predicted,differingintheir quantumnumbers. The interaction of the LQ with a lepton-quark pair is parameterized by a coupling λ. Depending on its quantum numbers a LQ couples to eq, νq or both. Thebranchingratioβ foraLQtodecayintoanelectronandaquark February7,2008 8:15 ProceedingsTrimSize: 9inx6in ringberg 3 SCALAR LEPTOQUARKS WITH F=0 l 1 D E H1 CI D U L C X E ~ S 1 / 2, L H1 single prod. -1 10 OPAL indir. limit D0 pair prod. 200 250 300 350 400 M (GeV) Figure1. Examplemass-dependent upperbounds onthe Yukawa couplingλofafirst generation scalar leptoquark to the electron-quark pair. These are shown for a LQ of F =0couplingtoane+ andadquark,whichdecays exclusivelyintoed. can be fixed by model assumptions, or can be treated as a free parameter. Decay modes other than eq and νq are usually neglected. In electron-proton collisions, first generation LQs might be singly pro- duced via the fusion of the incoming lepton with a quark coming from the proton. The productioncross-sectionroughlyscaleswithλ2 times the par- ton density function of the relevantpartonevaluatedat x =M2 /s, √s Bj LQ being the centre-of-mass energy and M the LQ mass. Hence, e+p (e−p) LQ collisions provide a larger sensitivity to LQs with fermion number F = 0 (F =2) since the u and d parton density is larger than that of antiquarks at large x . LQs might be observed as a resonant peak in the lepton-jet Bj mass spectrum ofNC or CC DIS events. No suchsignalhas been observed by the H1 and ZEUS experiments 4. Existing constraints on a scalar leptoquark which decays solely into an electron and a quark (β = 1) are summarised in Fig. 1. For an electro- magnetic strength of the coupling λ (λ2/4π = α , i.e. λ 0.3), the em ≃ HERA experiments rule out LQ masses below 290GeV. Constraints ∼ derived from the search for pair produced LQs at the Tevatron do not de- pend on the coupling λ and a lower mass bound of 256GeV is set by the D0 experiment 5. Example constraints on more general LQ models where β = 1 are shown in Fig. 2 assuming that the LQ decays exclusively into 6 February7,2008 8:15 ProceedingsTrimSize: 9inx6in ringberg 4 β eeeejjjj 11 II --bb + ff CCDDFF RRuunn IIII I 22 0.8 n .. u pp R xx ee d, 66 00 0.6 e 00 DD n 00.. eenn jjjj bi == m ll o 11 0.4 c HH 0 D 0.2 HH11 ll == nn jj 00..33 M LQ 100 200 300 400 (GeV) Figure 2. Example constraints on first generation scalar leptoquarks decaying exclu- sively into eq and νq. For λ = 0.3, the H1 constraints obtained from the ej and νj analyses are indicated by the dotted curves, and the combined bound is shown fortwo λvalues. ThefuturesensitivityoftheTevatronforaluminosityof2fb−1 isalsoshown. eq and νq. Leptoquarks decaying with a large branching ratio into νq are not easily probed at the Tevatron due to the large background for final states containing only jets andmissing transversemomentum. Incontrast, with a similarly good signal to background ratio for the eq and νq final states, HERA experiments can be sensitive on LQs which decay with a largebranchingratiointoνq,providedthatthecouplingλisnottoosmall. ThecurrentHERAboundsaremorestringentforF =0LQsduetothe much larger HERA I luminosity in e+p collisions (about 110pb−1 in e+p −1 − and 15pb in e p). Assuming the existence of a F =0 LQ in the HERA kinematicrange,andthatitscouplingtotheeqpairisequaltotheHERAI upperlimit,asignificanceof 4σcouldbereachedineachexperimentwith ane+pluminosityof350pb−∼1 andcombiningtheH1andZEUSdatacould allow the 5σ threshold to be reached. Assuming a LQ production cross- sectionequaltohalfthe existinglimitthe discoverypotentialislimited. In contrast a much larger discovery potential remains for F = 2 leptoquarks, − which will be probed with the e p data to be delivered until mid-2007. It should be noted that an ep collider would be the ideal machine to study a leptoquark signal. The fermion number of the leptoquark could February7,2008 8:15 ProceedingsTrimSize: 9inx6in ringberg 5 easily be determined by comparing the signal rates in e+p and e−p colli- sions; the polarisation of the lepton beam provides a handle to determine the chiral couplings of the LQ; and the good signal to backgroundratio in the νq channel allows the LQ coupling to the neutrino to be studied. 3.2. R-parity violating supersymmetry Insupersymmetric(SUSY)modelswheretheso-calledR-parity(R )isnot p conserved,squarkscouldberesonantlyproducedatHERAsimilarlytolep- toquarks. In addition to the “LQ-like” decays into eq and possibly νq the squarks also undergo decays into gauginos (the supersymmetric partners of gauge bosons) and an exhaustive searchrequires a large number of final statestobeanalysed. ThishasbeenpioneeredbytheH1collaborationin6 where the full HERA I dataset has been used to set constraints on super- symmetric models. A similar preliminary analysis looking for the SUSY partnerofthetopquarkhasbeenperformedbytheZEUSexperiment. Ex- ample constraints are shown in Fig. 3. For a squark R -violating coupling p ZEUS '131 l ZEUS (prel.) 99-00 e+p 100 GeV < M < 300 GeV 2 -300 GeV < m < 300 GeV 1 tanb =6 L C % E xclu d e d at 9 5 -1 10 l ‘ 1 3 1 (APV) -2 10 Excluded in part of SUSY parameters 100 120 140 160 180 200 220 240 260 280 300 M (GeV) stop Figure 3. Example mass-dependent constraints on the coupling of the stop to an e+d pair. AscanoftheSUSYparameterspacehasbeenperformed. Thelightshadeddomain is ruled out at 95% confidence level for any value of the parameters, which determine thedominantfinalstatesinwhichthesignalcouldbeobserved. Ascanbeseenfromthe narrownessofthedarkband,thesensitivityoftheanalysisisnearlymodel-independent. February7,2008 8:15 ProceedingsTrimSize: 9inx6in ringberg 6 of the electromagnetic strength, lower mass bounds of 270 280GeV can − be set. Within constrained SUSY models where a few parameters deter- mine the full Higgs and supersymmetric spectrum, stringent bounds were derivedonthesquarkmassfromsearchesforHiggs,sfermionsandgauginos at LEP. However, part of the SUSY parameter space remains open for a discovery at HERA II, for a reasonably largeR coupling 6. The case of p 6 a light stop or sbottom is of high interest for HERA II since the bounds coming from Tevatron are less stringent than those obtained assuming five degeneratesquarks. Inparticular,thesensitivityonthesbottom,whichhas a larger production cross-section in e−p than in e+p collisions (e−u ˜b), − → will considerably increase with the HERA II e p data. In case the squarks are too heavy to be produced at HERA, the t- channel exchange of a selectron or sneutrino could allow for single gaugino production. This processhas beenconsideredintwoclassesofSUSY mod- els, differing in the dominant decay mode of the produced gaugino 7. In both cases the analyses slightly improve the previous bounds if the rele- vant R-parity violating coupling is quite large. These are the first SUSY constraints set at HERA which are independent of the squark sector. 3.3. Fermion-boson resonances The observedreplicationofthree fermionfamilies motivates the possibility of a yet unobserved new scale of matter. An unambiguous signature for a new scale of matter would be the direct observation of excited states ∗ of fermions (f ), via their decay into a fermion and a gauge boson. In the most commonly used model 8, the interaction of an f∗ with a fermion and a gauge boson is described by a magnetic coupling proportional to ′ 1/Λ where Λ is a new scale. Proportionality constants f, f and f result s in different couplings to U(1), SU(2) and SU(3) gauge bosons. Existing constraints on excited electrons are shown in Fig. 4, under the assumption ′ ∗ that f = f . Searches for pair produced e at LEP allowed to rule out masses below about 103GeV, independently of the value of the coupling f/Λ. In contrast, searches for single e∗ production at LEP, HERA 9 and Tevatron10 setmassboundswhichdependonf/Λ. ThefutureHERAand Tevatron sensitivities, also depicted in Fig. 4, show the discoverypotential ofHERAII forexcitedelectrons. The caseofexcitedneutrinosis alsovery interesting for HERA II, since their production cross-section is larger in e−p than in e+p collisions by typically one order of magnitude. February7,2008 8:15 ProceedingsTrimSize: 9inx6in ringberg 7 1) -V e ZEUS T f = + f’ Λ ( H1 f/ DELPHI + OPAL, Direct (Prelim) 10 LEP Combined, Indirect (Prelim) CDF Run II 1 Tevatron 2 fb-1 -1 HERA II 1 fb-1 10 100 150 200 250 300 M (GeV) e* Figure4. Existingconstraintsonexcitedelectronmassesandcouplings,assumingthat f = f′. The decreasing curve shows the hyperbola f/Λ = 1/Me∗. The future HERA andTevatronsensitivitiesarealsoshownasdottedcurves. 4. Searches for deviations from the SM in rare processes 4.1. The “isolated lepton events” Within the Standard Model, W production at HERA has a cross-section of about 1pb. When the W decays leptonically, the final state contains an isolatedlepton, missing transversemomentum, and a usually softhadronic system. This process has been measured using the HERA I data 11 in the “electron” (W eν ) and “muon” (W µν ) channels, and a general e µ → → agreementwith the SM predictionwas observed. However,for largevalues ofthetransversemomentumofthehadronicsystem,PX,anexcessofevents T wasreportedbytheH1Collaboration11. Thisexcesswasnotconfirmedby a ZEUS analysis 12, differing from the H1 analysis in terms of background rejectiona. An abnormally large rate of high PX events is also observedby the H1 T experiment 13 in the HERA II data. Combining the e and µ channels and the HERA I and HERA II datasets, which amount to a total luminosity aThe non W contribution to the expected background was about 50% at large PTX in theZEUSanalysis,whileitamounts to15%onlyintheH1analysis. February7,2008 8:15 ProceedingsTrimSize: 9inx6in ringberg 8 l+Pmiss events at HERA 1994-2004 (e+p, 158 pb-1) l+Pmiss events at HERA 1998-2005 (e-p, 121 pb-1) T T entsents110022 HA1ll DSMata (prelim.) NNDSMat a = = 2188.5 ± 2.6 entsents110022 HA1ll DSMata (prelim.) NNSDMat a = = 1125.8 ± 2.2 vv Signal vv Signal EE EE 1100 1100 11 11 1100--11 1100--11 00 1100 2200 3300 4400 5500 6600 7700 8800 00 1100 2200 3300 4400 5500 6600 7700 8800 PPXX ((GGeeVV)) ee aanndd µµ cchhaannnneellss PPXX ((GGeeVV)) ee aanndd µµ cchhaannnneellss TT TT Figure5. Distributionof the transversemomentum of the hadronic system PTX inse- lectedeventsrecordedin(left)thee+pdatasampleand(right)thee−pdatasample. The hatchedhistogramshowstheexpectationfromW productionwhilethetotalexpectation isgivenbytheopenhistogram. of 279pb−1, 17 events are observed at PX > 25GeV for a SM expecta- T tion of 9.0 1.5. Amongst the 6 new events observed in HERA II, 5 were recordeddu±ringthee+prunning(53pb−1)andoneduringthee−prunning (107pb−1). Fig. 5 shows the observed PX distributions separately for the T e+p and e−p datasets where HERA I and HERA II data are combined, together with the corresponding SM expectations. The observed and ex- pected numbers of events are given in table 1. While the observation in − the e pdata isconsistentwiththe SMexpectation,15eventsareobserved at PX > 25GeV in the e+p data for an expectation of 4.6 0.8 events. T ± The probability that the expected yield fluctuates to 15 events or more corresponds to a 3.4σ fluctuation. The ZEUS experiment has recently carried out a re-analysis of the e channel using the 99-00e+p data, resulting in a largerpurity in W events. The positron data taken at HERA II have been analysed in the same way, −1 such that the total luminosity amounts to 106pb . The results are also shown in table 1. At PX > 25GeV one event is observed in the data, in T agreement with the SM expectation of 1.5 0.18. ± Although the rate of events observed in the e channel in the positron data is larger in H1 than in ZEUS, both experiments are compatible with −1 each other within 2.5σ for an average rate of about 4 events per 100pb in this channel. Assuming that such events are observed in the future H1 −1 and ZEUS data at an average rate of about 7 8 events per 100pb − February7,2008 8:15 ProceedingsTrimSize: 9inx6in ringberg 9 combining the e and µ channels, a significance of 4σ could be reachedb fromthe combinedH1 andZEUSdatasets bydoubling the e+p luminosity. It should be noted that new physics scenarios can be found which could explain that such events are observed in e+p collisions only. For example, in supersymmetry with two R-parity violating couplings involving third generation fields, a top quark could be produced via t-channel sbottom exchange in e+d collisions. Due to the large value of Bjorken x needed to − produce a top quark in the final state, the corresponding process in e p collisions would have a much lower cross-section. Table1. SummaryoftheH1resultsofsearchesforeventswithisolatedelectronsormuons and missing transverse momentum for the e+p data (158 pb−1) and the e−p data (121 pb−1). Data from HERA I and HERA II are combined. The number of events observed by ZEUS in the electron channel, in 106pb−1 of e+p data, is also shown. The number of observed events at PTX > 25GeV is compared to the SM prediction. The quoted errors containstatisticalandsystematicuncertainties addedinquadrature. Electron Muon Combined PX > 25 GeV obs./exp. obs./exp. obs./exp. T H1 1998-2005 e−p Preliminary 121pb−1 2/2.4±0.5 0/2.0±0.3 2/4.4±0.7 H1 1994-2004e+p Preliminary 158pb−1 9/2.3±0.4 6/2.3±0.4 15/4.6±0.8 ZEUS 1999-2004e+p Preliminary 106pb−1 1/1.5±0.18 4.2. Multi-lepton events If the events reported above were to be explained by some anomalous W production mechanism, an anomalous rate for Z-like events could also be observed. Eventswithatleasttwoelectronsormuonsinthefinalstatehave been looked for by the H1 collaboration 14. A slight excess of high mass multi-electron events was observed in the HERA I dataset. The analysis has been repeated using the HERA II data and extended to include other multi-lepton topologies15. With a totalluminosity of 209pb−1 four events are observed with P P > 100GeV, three of which being HERA I leptons T ee events. This is slightly above the SM prediction of 0.81 0.14. ± bThis estimate is obtained by scaling the SM backgrounds of the H1 analysis in both theeandµchannels. February7,2008 8:15 ProceedingsTrimSize: 9inx6in ringberg 10 5. Conclusions Although some stringent bounds on new physics are set at LEP and the Tevatron, HERA appears to be very well suited to search for new phe- nomena in some specific cases. In particular, searches for new physics at HERA rarely suffer from huge SM backgrounds. Searches for leptoquarks, for the supersymmetric partners of the top or bottom quarks, and for ex- cited fermions might bring a discoverywith the HERA II data. This holds in particular for new physics processes for which the cross-sectionis larger in e−p than in e+p collisions, since the HERA I constraints are not too stringent in such cases. With the future e+p data to be delivered until mid-2007, the excess of atypical W-like events observed at H1, which cor- respondstoa3.4σ deviation,appearstobe the bestchanceforadiscovery at HERA II. References 1. H1 Collab., C. Adloff et al., Phys. Lett. B568 (2003) 35; ZEUS Collab., S.Chekanov et al., Phys.Lett. B591 (2004) 23. 2. D0Collab., D0 Note4552-CONF. 3. J. Meyer, these proceedings. 4. H1 Collab., Phys. Lett. B629 (2005) 9; ZEUS Collab., S. Chekanov et al., Phys.Rev.D68 (2003) 052004. 5. D0Collab., V.M. Abazov et al., Phys. Rev. D71 (2005) 071104. 6. H1Collab., A.Aktas et al., Eur. Phys. J. C36 (2004) 425. 7. H1Collab., A.Aktas et al., Phys. Lett. B616 (2005) 31; ZEUS Collab., con- tributedpapertoEPS’05,abstract#329;idem,contributedpapertoEPS’05, abstract #330. 8. K. Hagiwara, D. Zeppenfeld and S. Komamiya, Zeit. fur Phys. C29 (1985) 115; F. Boudjema, A. Djouadi and J.L. Kneur, Zeit. fur Phys. C57 (1993) 425. 9. H1 Collab., C. Adloff et al., Phys. Lett. B548 (2002) 35; ZEUS Collab., S.Chekanov et al., Phys.Lett. B549 (2002) 32. 10. CDF Collab., D.Acosta et al., Phys. Rev.Lett. 94 (2005) 101802. 11. H1 Collab., V. Andreevet al., Phys.Lett. B561 (2003) 241. 12. ZEUS Collab., S.Chekanov et al., Phys. Lett. B559 (2003) 153. 13. H1Collab.,documentpreparedfortheNov.2005DESY-PRCmeeting,avail- able at http://www-h1.desy.de/publications/H1preliminary.short list.html. 14. H1 Collab., A. Aktas et al., Eur. Phys. J. C31 (2003) 17; H1 Collab., A. Aktaset al., Phys. Lett. B583 (2004) 28. 15. H1 Collab., contributed to EPS’05, abstract #635.