aa r X i v : . [ h e p - e x ] M a y SEARCHES FOR NEW PHYSICS AT THE TEVATRON
M. Jaffr´eon behalf of the CDF and D0 collaborations
LAL, Universit´e Paris-Sud, CNRS/IN2P3,91898 Orsay Cedex, France
The Tevatron collider has provided the CDF and D0 experiences with large datasets as inputto a rich program of searches for physics beyond the standard model. The results presentedhere are a partial survey of recent searches conducted by the two collaborations using up to6 fb − of data. The standard model (SM) of particles, despite its remarkable description of experimental data atthe elementary particle level, has some deficiencies to explain what is observed in the universe :lack of anti-matter, existence of dark matter,. . . Working at the energy frontier, as was the caseat the Tevatron for so many years, gives experimentalists the hope to discover new non-SMparticles which would indicate some direction to follow at explaining these SM deficiencies.Over the years, the CDF and D0 experiments have gained experience in the detector re-sponses to all particle types. It allows to look at a large number of different final states search-ing for deviations from the SM expectations. As the knowlegde of detector particle responsesbecomes more accurate, the complexity of final states can increase.For a given final state signature, the non-observation of deviations from the SM predictionallows to constrain several models at once.
Experimentally, in a hadronic environment, dielectron, dimuon or diphoton final states are easyto identify. A lot of extensions to the SM predicts the existence of new particles which couldbe observed as narrow resonances decaying into a pair of leptons or photons. Among those arenew spin-1 gauge boson 1 or spin-2 Randall-Sundrum (RS) graviton 2.The observed dilepton mass distributions observed in large datasets (around 5 fb − or more)by CDF3,4 and D05 are in agreement with the SM expectations. The largest invariant dielectronmass observed by CDF is 960 GeV(Fig. 1). No new resonance is observed above the Z gaugeboson mass. This allows both experiments to set 95% C.L. cross section limits which can becompared to model predictions; as an example, the benchmark SM-like Z ′ is now excluded formasses below 1 TeV.The same spectra can be used to extract limits for the graviton G mass; higher limits canbe achieved with the diphoton spectra 6 as the G branching fraction into two photons is twicethat into a lepton pair. But, the best limits are obtained by combining dielectron and diphotonnalyses7 ,
3. Fig. 2 shows the excluded domain, obtained by CDF, in the two parameter spaceof the RS model, k/M
P l , the ratio of curvature of the extra dimension to reduced Planck scale,and the G mass. ) Dielectron Mass (GeV/c
100 200 300 400 500 600 700 800 900 1000 N o f eve n t s -2 -1
10 100 200 300 400 500 600 700 800 900 1000 -2 -1 DATADrell-YanQCDDibosontt ) -1 CDF Run II Preliminary (L=5.7 fb
Figure 1: Invariant dielectron mass distribution com-pared to the expected backgrounds in CDF data. ) Graviton Mass (GeV/c400 500 600 700 800 900 1000 1100 P l M k / Graviton Mass (GeV/c400 500 600 700 800 900 1000 1100 P l M k / ee analysis analysis gg analysis gg ee+ ) -1 CDF Run II Preliminary (L=5.7 fb
RS-graviton 95% CL exclusion
Figure 2: Excluded domains in the RS parameterspace ( k/M Pl , graviton mass) obtained in the dielec-tron analysis, the diphoton analysis and their combi-nation Production of pairs of gauge vector bosons has been observed by CDF and D0 at the levelpredicted by the SM. This sector is however not well constrained due to the smallness of thecross sections. There is still room for an extra source of dibosons from the decay of massivecharged or neutral particle as new gauge boson W ′ or RS graviton G . This has been looked atby CDF 8, and recently by D0 9 using 5.4 fb − of data. One of the bosons is allowed to decayleptonically and the other hadronically leading to two event signatures ( ll + jets, l + jets + E / T ) where the lepton l is either an electron or a muon and E / T is the missing transverse energycarried away by the neutrino. The two leptons or the lepton and E / T are first combined to forma Z or W candidate. Since for a very high massive resonance, the two bosons would be highlyboosted, the two hadronic showers from the decay of the second boson may be reconstructedas a single massive jet. So, D0 has increased its signal sensitivity by first trying to assign the W ( Z ) hypothesis to a jet with a mass larger than 60(70) GeV before any two jet combination.Figure 3 shows for the data and the expected backgrounds the distributions of the reconstructedresonance and transverse masses in the dilepton and single lepton channels, respectively. Thepredicted distributions of a 600 GeV W ′ or G are also shown. Cross section limits are extractedfrom the absence of any significant excess of events in data from which one deduces lower boundsfor W ′ and G masses of 690 and 754 GeV respectively. Vectorlike quarks (VQ) share many characteristics of the SM quarks with the distinctive excep-tion that their left and right components transform in the same way under SU (3) × SU (2) × U (1).They can be singly produced via the electroweak interaction and may decay into a W or Z bo-son and a SM quark. D0 10 has separated the analysis in two independant channels accordingto the leptonic decay of the vector boson ( ll + jet, l +jet + E / T ). The event’s leading jet intransverse momentum is assumed to come from the VQ decay. Figure 4 shows, for the dataand the expected backgrounds, the distributions of the reconstructed resonance and transversemasses in the dilepton and single lepton channels, respectively. The absence of any statistically W or WZ Transverse Mass [GeV]
300 400 500 600 700 800 E ve n t s / G e V
10 DataW+jetsttWW+WZSSM W’(600)RS G(600)
WW or WZ Transverse Mass [GeV]
300 400 500 600 700 800 E ve n t s / G e V -1 , 5.4 fb ˘ D WZ Mass [GeV]
300 400 500 600 700 800 E ve n t s / G e V WZ Mass [GeV]
300 400 500 600 700 800 E ve n t s / G e V -1 , 5.4 fb ˘ D (a) (b) Figure 3: Distributions of the reconstructed
W W or W Z transverse mass (left) and
W Z mass (right) for data(dots), estimated backgrounds, and the estimated contributions of a 600 GeV SSM W ′ and a 600 GeV G . significant excess in data allows to derive cross section limits which are compared to VQ pro-duction in two scenarios. VQ not coupled to down-quarks and decaying exclusively to W q areexcluded for masses below 693 GeV; if decaying to Zq the lower bound of the mass is 551 GeV.In the alternate scenario, i.e. no coupling to up-quarks, the mass limits are 403 GeV ( W q ) and430 GeV ( Zq ). (GeV) QT M0 200 400 600 800 1000 1200 E ve n t s / G e V
10 (GeV) QT M0 200 400 600 800 1000 1200 E ve n t s / G e V (a) -1 DØ, L = 5.4 fbDataV+jetsDibosonTopMultijet Wq fi D Q = 500 GeV Q m (GeV) llj M0 200 400 600 800 1000 1200 E ve n t s / G e V
10 (GeV) llj M0 200 400 600 800 1000 1200 E ve n t s / G e V (b) -1 DØ, L = 5.4 fb Zq fi U Q = 500 GeV Q m Figure 4: (a) Vector-like quark transverse mass and (b) vector-like quark mass for the single lepton and dileptonchannels, respectively. Expected distributions for 500 GeV signals decaying as Q D → W q and Q U → Zq . A minimal extension of the standard model is obtained by the addition of a new unbroken SU(N)gauge group. Such a group is characterized by the mass of the new fermions (quirks),Q, and thestrength of the gauge coupling, Λ. D0 11 has considered the case where the quirks are charged,and Λ << M Q , and M Q = 0 . − Q ¯ Q pair produced in p ¯ p collisions will not hadronize. The quirks in the pair will stay connected,as with a rubber band, the two tracks will not be resolved by the tracking system and theywill be reconstructed as a single straight highly ionizing track. This is the first search of thiskind. Event selection requires an isolated track and a very high transverse momentum jet backto back. Such events are triggered by requiring jets and substantial E / T . Analysing 2.4 fb − of data, no excess of events at large ionization loss is observed over the expected backgrounddetermined from isolated tracks in an orthogonal data sample. From the cross section limits onthe quirk production, D0 extract limits on the quirk mass depending on N, the number of colorsin the new gauge group, of 107, 119 and 133 GeV for N=2, 3 and 5 respectively. Search for leptonic jets
Hidden Valley (HV) scenarios 12 contain a hidden sector which is weakly coupled to SM parti-cles. They become popular as they provide convincing interpretation of observed astrophysicalanomalies and discrepancies in dark matter search. New low mass particles are introduced inthe hidden sector, and the dark photon, which is the force carrier, would have a mass aroundone GeV or less and would decay into a fermion or pion pair. The case of decays to lepton pair( electron or muon ) is particularly attractive. SUSY is often included in HV models, one couldhave a situation where the lightest neutralino will decay to a dark photon and ˜ X , the lightestSUSY particle of the hidden sector, which will escape detection, leading to large E / T . As thedark photon is light, it will be highly boosted in the neutralino decay, and the two leptons willbe close to each other. Experimentally, one has to change the isolation criteria usually appliedto identify leptons. The presence of a track of opposite charge close to the lepton candidatewill sign the so-called leptonic jet ( l -jet). Using 5.8 fb − of data, D0 13 has searched for pairproduction of l -jets in three configurations : ee , µµ and eµ . No evidence of l -jets is observed inthe distributions of the electron and muon pair masses (Fig. 5). Limits on the production crosssection, around 100 fb for a 1 GeV dark photon, are obtained. Di-electron mass (GeV) E ve n t s / . G e V D g M( )=0.9 GeV D g M( )=1.3 GeV D g M( -1 ˘ Da)
Di-electron mass (GeV) E ve n t s / . G e V Di-muon mass (GeV) E ve n t s / . G e V -1 ˘ D Muon l-jetsBackground)=0.3 GeV D g M( )=0.9 GeV D g M( )=1.3 GeV D g M( -1 ˘ Db)
Di-muon mass (GeV) E ve n t s / . G e V Figure 5: Invariant mass of dark photon candidates with two isolated l -jets and E / T >
30 GeV, for (a) electron l -jets and (b) muon l -jets ( two entries per event, the eµ events contribute to both plots). The red band showsthe shape of the mass distribution for events with E / T <
20 GeV. The shaded blue histograms show the shapesof 8 MC signal events added to backgrounds, for three masses of the dark photon.
CDF 14 has searched for a fourth generation bottom-like quark ( b ′ ). Current limits push b ′ tobe heavier than the sum of the t quark and the charged gauge boson W masses. The analysis isrealised using 4.8 fb − of data assuming b ′ to decay exclusively to t and W . The b ′ pair producedin p ¯ p interactions will decay into two b quarks and four W ’s. One of the W decays leptonicallyand the others hadronically. The final state is then characterised by an isolated lepton (electronor muon), E / T and many jets; one of the jets is required to be tagged as a b -jet. All the quarkjets will not be reconstructed either because they fall below the transverse momentum cut orbecause their hadronic showers are overlapping and they are merged in a single jet. The analysisis performed in three jet multiplicity bins : 5, 6 and 7 jets or more. The other variable whichhelps fighting the SM backgrounds, mainly t ¯ t and W + jets production, is H T , the scalar sumof the transverse momentum of the lepton, jets and E / T . Signal is expected to appear in the lastmultiplicity bin and for high H T values. The H T distributions for the three multiplicity binsare shown in a single plot (Fig.6) using the variable J et − H T equal to H T , H T + 1000 GeV or H T + 2000 GeV for events with 5, 6, and 7 jets or more respectively. In the absence ofny significant excess of events, cross section limits are set as a function of the b ′ mass. Whencompared to the NLO b ′ cross section 15, one gets a lower bound for its mass of 372 GeV (Fig. 7). E v e n t s DataBkg. + b’Total Bkg.ttW,Z,tQCDDi-boson T Jet-H0 500 1000 1500 2000 2500 3000 O b s . - E xp . -10010 DataBkg Unc.b’
Figure 6: Distribution of the variable Jet- H T for dataand expected backgrounds. Contributions from a350 GeV b ′ signal on top of the expected backgroundsare also shown. The bottom pane shows the differencebetween the expected and observed number of events,as well as the total uncertainty on the expected num-ber of events. ] Quark mass [GeV/c
300 350 400 C r o ss - s ec ti on [f b ] Median68%95%TheoryObserved Figure 7: 95% C. L. observed upper limits on b ′ pro-duction cross section (red line) compared to the ex-pected median limit (black line) in simulated exper-iments without b ′ signal. Green and yellow bandsrepresent 68% and 95% of simulated experiments, re-spectively; the dashed line is the NLO b ′ productioncross section 15. CDF 16 has searched for dark matter through the production of an exotic fourth generation t ′ quark decaying to a t quark and a dark matter particle X . The decay of the t ′ to a b quarkand a charged gauge boson W is not allowed. The signal is searched for in the following eventtopology: an isolated lepton (electron or muon ), four jets or more, and very large E / T due tothe invisible particle X . The minimum value of E / T is optimised for each point in the ( t ’ mass, X mass) plane. The larger the mass difference, the higher the cut. The main SM backgroundsare t ¯ t pair and W + jets production. There are W bosons in the main backgrounds which areprominently visible in the transverse mass distribution of the lepton and E / T pair in the signaldepleted domains i.e. at low E / T or low jet multiplicity as seen in Fig. 8a and 8b. In the signalregion, no significant excess of events is observed (Fig. 8c). Cross section limits are obtainedwhich allow to exclude a t ′ mass below 360 GeV for a X mass below 100 GeV. The performance of the Tevatron has brought limits on BSM physics beyond one could haveexpected. LHC experiments will take over, but CDF and D0 have still assets with their largedatasets; their analysis will be oriented towards complex final states. Further details on physicsresults can be obtained from :
CDF D0 Figure 8: Transverse mass distributions of the lepton and E / T for data and the expected backgrounds in controlregions where signal is negligible (a and b) and where the W is visible and the region where it should appear (c).The bottom panes show the difference between expected background and observed events, as well as the totaluncertainty on the expected background events. Acknowledgments
The author would like to thank the CDF and D0 working groups for providing the material forthis talk, and the organizers of the
Rencontres for a very enjoyable conference and the excellentorganization.
References
1. P. Langacker,
Rev. Mod. Phys. , 1199 (2008).2. L. Randall and R. Sundrum, Phys. Rev. Lett. , 3370 (1999); , 4690 (1999).3. CDF Collaboration, Public note CDF 10405 (unpublished).4. T. Aaltonen et al (CDF Collaboration), Phys. Rev. Lett. , 121801 (2011).5. V.M. Abazov et al (D0 Collaboration),
Phys. Lett. B , 88 (2011).6. T. Aaltonen et al (CDF Collaboration), Phys. Rev. D , 011102(R) (2011).7. V.M. Abazov et al (D0 Collaboration), Phys. Rev. Lett. , 241802 (2010).8. T. Aaltonen et al (CDF Collaboration),
Phys. Rev. Lett. , 241801 (2010).9. W.M. Abazov et al (D0 Collaboration),
Phys. Rev. Lett. , 061801 (2010); arXiv:hep-ex/1011.6278, accepted by Phys. Rev. Lett.10. V.M. Abazov et al (D0 Collaboration),
Phys. Rev. Lett. , 081801 (2011).11. V.M. Abazov et al (D0 Collaboration),
Phys. Rev. Lett. , 211803 (2010).12. T. Han et al.,
J. High Energy Phys. , 008 (2008) 008; M. Strassler and K. Zurek, Phys.Lett. B , 374 (2007).13. V.M. Abazov et al (D0 Collaboration), Phys. Rev. Lett. , 211802 (2010).14. T. Aaltonen et al (CDF Collaboration),
Phys. Rev. Lett. , 141803 (2011).15. R. Bonciani et al , Nucl. Phys. B , 424 (1998); M. Cacciari et al , J. High Energy Phys.04