Single W and Anomalous Single Top Production at HERA
aa r X i v : . [ h e p - e x ] M a y Single W and Anomalous Single Top Production at HERA D. M. South a ∗ a Technische Universit¨at Dortmund, Experimentelle Physik V, 44221 Dortmund, Germany
The search for events containing isolated leptons (electrons or muons) and missing transverse momentumproduced in e ± p collisions is performed individually and in a common phase space with the H1 and ZEUSdetectors at HERA in the period 1994–2007. The presented H1+ZEUS data sample corresponds to an integratedluminosity of 0.97 fb − , and comprises the complete high energy data from the HERA programme. A total of87 events are observed in the data, compared to a Standard Model prediction of 92.7 ± P XT >
25 GeV in the e + p data, luminosity 0.58 fb − , 23 data events are observed comparedto a SM prediction of 14.6 ± W production are performed by H1, where the measured crosssections are found to be in agreement with SM predictions. A complementary search by H1 for events containingan isolated tau lepton and missing P T is also presented. A measurement of the W polarisation fractions isperformed by H1, where the presented results are found to be in agreement with the SM. Finally, the H1 isolatedlepton events are examined in the context of a search for anomalous single top production. In the absence of aclear signal, an upper limit on the anomalous top production cross section σ ep → etX < .
16 pb is established at the95% confidence level, corresponding to a limit an upper bound on the anomalous magnetic coupling κ tuγ < .
1. The HERA Electron–Proton Collider
ZEUSH1
Figure 1. The HERA e ± p collider at DESY. Thelocation of the H1 and ZEUS experiments is in-dicated, as well as the pre-accelerator PETRA.The HERA e ± p collider, located in Hamburg,Germany, and shown in figure 1, was in opera-tion in the years 1992–2007. Protons with an en- ∗ on behalf of the H1 and ZEUS Collaborations. ergy up to 920 GeV were brought into collisionwith electrons or positrons of energy 27.6 GeV attwo experiments, H1 and ZEUS, each of whichcollected about 0.5 fb − of data. Together withmeasuring the structure of the proton, the deepinelastic collisions (DIS) produced at HERA, ata centre of mass energy up to 318 GeV, providedan ideal environment to study rare processes,set constraints on the Standard Model (SM) andsearch for new particles and physics beyond theStandard Model (BSM).
2. Events with Isolated Leptons and P miss T Events containing a high transverse momentum( P T ) isolated electron or muon and missing P T have been observed at HERA [1,2,3,4]. An ex-cess of HERA I data events (1994–2000 which ismostly in e + p collisions) compared to the SM pre-diction at large hadronic transverse momentum P XT was reported by the H1 Collaboration [2].This was not confirmed by the ZEUS Collabo-ration, although the analysis was performed in amore restricted phase space [4].1 D. M. South p q W + l + n l q’ q’e + e + Figure 2. Feynman diagram of the process ep → eW ± ( → lν ) X , which is the main SM signal con-tribution to the search for events with isolatedleptons and missing transverse momentum.The main SM contribution to the signal topol-ogy is the production of real W bosons via photo-production with subsequent leptonic decay ep → eW ± ( → lν ) X , as illustrated in figure 2, wherethe hadronic system X has typically low P T . Theequivalent charged current (CC) process, ep → νW ± ( → lν ) X , also contributes to the total sig-nal rate, although only at a level of about 7%.The production of Z bosons with subsequentdecay to neutrinos ep → eZ ( → ν ¯ ν ) X resultsin a further minor contribution to the total sig-nal rate in the electron channel at a level of 3%.SM signal processes are modelled using the eventgenerator EPVEC [5]. The main W productionvia photoproduction component is reweighted toa NLO calculation [6], which has a theoretical un-certainty of 15%.SM background enters the electron channel dueto mismeasured neutral current (NC) events andthe muon channel due to lepton pair (LP) eventsin which one muon escapes detection, both casesresulting in apparent (fake) missing transversemomentum. CC background, which contains in-trinsic missing transverse momentum, enters thefinal sample in both lepton channels, where a finalstate particle is interpreted as the isolated elec-tron or muon. This process is not included in the present ZEUS analysis.
The event selection employed by the H1 [7] andZEUS [8] analyses is very similar and may be sum-marised as follows: The identified lepton shouldhave high transverse momentum P lT >
10 GeV,be observed in the central region of the detec-tor and be isolated with respect to jets and othergood quality tracks in the event. The eventshould also contain a large transverse momen-tum imbalance, P miss T >
12 GeV. Further cutsare then applied, which are designed to reduceSM background, whilst preserving a high level ofsignal purity. Event quantities which are sensitiveto the presence of undetected energetic particlesin the event are employed such as the azimuthalbalance of the event, the difference in azimuthalangle between the lepton and the hadronic sys-tem and the longitudinal momentum imbalance.To ensure that the two lepton channels are ex-clusive, and may therefore be combined, electronevents must contain no isolated muons.
Both H1 and ZEUS have recently performedthe analysis of the electron and muon channelson their respective complete HERA I+II e ± p datasets, which correspond to approximately 0.5 fb − per experiment [7,8].A total of 59 events are observed in the H1data, compared to a SM prediction of 58.9 ± P XT >
25 GeV, a total of 24 events are ob-served compared to a SM prediction of 15.8 ± e + p data com-pared to a SM prediction of 8.9 ± e + p data [2]thus remains at the 3.0 σ level for the completeH1 e + p data [7].In the ZEUS analysis of the completeHERA I+II data, 41 data events are ob-served in agreement with the SM prediction of48.3 ± P XT region, where 11 events are seenin the e ± p data compared to a SM prediction of13.1 ± A study of the selection efficiency for signal pro-cesses found the H1 and ZEUS analyses to be ingle W and Anomalous Single Top Production at HERA P XT >
25 GeVare also found to be in the region of overlap of thetwo analyses. Nevertheless, in order to coherentlycombine the results from the two experiments, acommon phase space has been established.This common selection is based on the H1event selection [2,7], but covers a more restrictedlepton polar angle range of 15 ◦ < θ l < ◦ ,as that employed in the ZEUS analysis [8]. Thesignal expectation rates of the H1 and ZEUS anal-yses using the common selection are found to becomparable, taking into account the respectiveluminosities of the data sets and the signal pro-cesses included. More details on the combinationof the H1 and ZEUS analyses can be found in [11].The results of the combined H1+ZEUS analy-sis are summarised in table 1. The signal contri-bution, mainly from real W production, is seento dominate the total SM expectation in all datasamples. At large hadronic transverse momentum P XT >
25 GeV a total of 29 events are observedin the H1+ZEUS e ± p data compared to a SMprediction of 25.3 ± e + p data alone,23 events are observed with P XT >
25 GeV com-pared to a SM prediction of 14.6 ± σ . Seventeen of the 23 data eventsare observed in the H1 data compared to a SMexpectation of 7.1 ± σ .Figure 3 (top) shows the transverse mass, M lνT and P XT distributions of the H1+ZEUS e ± p HERA I+II data for the combined electron andmuon channels. The distribution of events in M lνT is compatible with the Jacobian peak expectedfrom W production. Similarly, the observed P XT spectrum displayed in figure 3 (bottom) is com-patible with that expected from W production,peaking at low values of hadronic transverse mo-mentum. An additional change with respect to the H1 event selec-tion is that the cut on longitudinal momentum imbalancein the electron channel is simplified in that it is alwaysapplied. This is found to make only a negligible differenceto the SM expectation. (GeV) n lT M E ve n t s -1 = 87 Data
N 11.2 – = 92.7 SM N H1+ZEUS Data (prelim.)All SMSignal ) -1 p, 0.97 fb – events at HERA I+II (e missT + P m e, H E RA E x o t i cs W o r k i ng G r oup o ve r f l o w (GeV) n lT M E ve n t s -1 (GeV) XT P E ve n t s -1 = 87 Data
N 11.2 – = 92.7 SM N H1+ZEUS Data (prelim.)All SMSignal ) -1 p, 0.97 fb – events at HERA I+II (e missT + P m e, H E RA E x o t i cs W o r k i ng G r oup (GeV) XT P E ve n t s -1 Figure 3. The transverse mass M lνT (top) andhadronic transverse momentum P XT (bottom) dis-tributions of the combined H1+ZEUS e ± p HERAI+II data. The data (points) are compared tothe SM expectation (open histogram). The sig-nal component of the SM expectation is given bythe striped histogram. N
Data is the total numberof data events observed and N SM is the total SMexpectation. The total uncertainty on the SMexpectation is given by the shaded band. D. M. South
Table 1Summary of the combined H1+ZEUS results in the search for events with isolated electrons or muonsand missing transverse momentum, shown for the electron and muon channels separately and combinedfor the full HERA I+II e + p , e − p and e ± p data. The number of observed data events is compared to theSM prediction. The results are shown for the full sample and for events with P XT >
25 GeV. The signalcomponent of the SM expectation, dominated by real W production, is given as a percentage of the totalSM prediction in parentheses. The uncertainties contain statistical and systematic uncertainties addedin quadrature.H1+ZEUS Preliminary Electron Muon Combined l + P miss T events at obs./exp. obs./exp. obs./exp.HERA I+II (Signal contribution) (Signal contribution) (Signal contribution)1994-2007 e + p Full Sample 39 / 41.3 ± ± ± − P XT >
25 GeV 12 / 7.4 ± ± ± e − p Full Sample 25 / 31.6 ± ± ± − P XT >
25 GeV 4 / 6.0 ± ± ± e ± p Full Sample 64 / 72.9 ± ± ± − P XT >
25 GeV 16 / 13.3 ± ± ±
3. Cross Sections
The H1 results described in section 2.1 are usedto calculate production cross sections for eventswith an energetic isolated lepton and missingtransverse momentum ( σ ℓ + P miss T ) and for single W boson production ( σ W ), for the latter of whichthe branching ratio (Γ=0.24) for leptonic W de-cay is taken into account [12]. For the isolatedlepton cross section, Γ = 1. The cross sections aremeasured in the phase space 5 ◦ < θ ℓ < ◦ , P ℓT >
10 GeV, P miss T >
12 GeV, and where thelepton is isolated from any jet by at least one unitin η − φ .Cross sections are calculated using the formula: σ = N data − N MCbkd L Γ A with A = N MCrec N MCgen , (1)where N data is the number of data events, N MCbkd is the Monte Carlo (MC) estimate of number ofbackground events and L is the total data lu-minosity. A is the acceptance, where N MCrec and N MCgen are the number of reconstructed and gen-erated events in the signal MC, respectively. SMsignal processes are described in section 2. Itshould be noted that the small SM contributionto N MCrec from Z production is signal for σ ℓ + P miss T ,whereas it is considered background for σ W . The results are shown in table 2 with statisti-cal (stat) and systematic (sys) uncertainties andare found to be in good agreement with the SMpredictions.
4. Events with an Isolated Tau and P miss T The H1 Collaboration has also performed asearch for events with an isolated tau lepton andlarge missing transverse momentum, using thefull HERA I+II e ± p data and the hadronic one–prong tau decay mode [13]. This search is com-plementary to the electron and muon searches de-scribed above, and provides a test of lepton uni-versality. In addition, some BSM scenarios favourthe third lepton generation, which could lead toan enhancement from the subsequent leptonic de-cay of the tau lepton.The event selection is based on that usedin [14], with improvements to the tau identifi-cation algorithm, in particular to the track iso-lation [13]. The event signature is a narrow, lowtrack multiplicity jet (tau–jet) in coincidence withmissing transverse momentum.The hadronic transverse momentum distribu-tion of the final sample is shown in figure 4,where 20 events are observed in the data com-pared to a SM prediction of 19.5 ± ingle W and Anomalous Single Top Production at HERA σ ℓ + P miss T ) and for single W production ( σ W ) with statistical (stat.) and systematic (sys.) uncertainties.The results are compared to the SM prediction with the associated theory uncertainty (th.).H1 HERA I+II Data SM σ ℓ + P miss T ± ± ± σ W ± ± ± / GeV XT P E ve n t s -2 -1 H1 Data (prelim.)All SMSignal = 20
Data
N 3.2 – = 19.5 SM N ) -1 p, 471 pb – events at HERA I + II (e missT + P t / GeV XT P E ve n t s -2 -1 Figure 4. The hadronic transverse momentumdistribution of τ + P miss T events in the H1 e ± p HERA I+II data. The data (points) are com-pared to the SM expectation (open histogram).The signal component of the SM expectation isgiven by the striped histogram. N
Data is the totalnumber of data events observed and N SM is thetotal SM expectation. The total uncertainty onthe SM expectation is given by the shaded band.is dominated by charged current events and thesignal purity is observed to be much lower thanthe electron muon channels, at around 14%. For P XT >
25 GeV one event is selected in the data,compared to a SM prediction of 0 . ± .
5. Measurement of the W Boson Polarisa-tion Fractions
The H1 measurement of the W boson polarisa-tion fractions at HERA makes use of the cos θ ∗ distributions in the decay W → e/µ + ν , where θ ∗ is defined as the angle between the W bosonmomentum in the lab frame and the charged de-cay lepton in the W boson rest frame. The lefthanded F − , longitudinal F and right handed F + polarisation fractions are constrained by the rela-tion F + ≡ − F − − F . The cos θ ∗ distributionsfor W + bosons are given [15] by: dNd cos θ ∗ ∝ (1 − F − − F ) ·
38 (1 + cos θ ∗ ) + F · (cid:0) − cos θ ∗ (cid:1) + F − ·
38 (1 − cos θ ∗ ) . (2)For W − bosons the cos θ ∗ distributions have op-posite sign. To allow the combination of thedifferent W boson charges, the value of cos θ ∗ is multiplied by the sign of the lepton charge q ℓ = ± W boson can be successfully reconstructed, basedon the method described in [16]. Additional se-lection criteria are applied to ensure a reliablecharge measurement, so that the resulting chargemisidentification is well below 1% [12]. Electronand muon events originating from tau decays of W bosons are considered background in this anal-ysis, since for these events the cos θ ∗ distribu-tions are not expected to be described by equa-tion 2. The final event sample consists of 21 elec-tron events and 10 muon events with an overallsignal purity of 83%.The measured q ℓ cos θ ∗ distribution, correctedfor acceptance and detector effects, is shown infigure 5 (top), compared to the SM prediction.The cross section fit to the model defined in equa- D. M. South *) q cos( l q-1 -0.5 0 0.5 1 * ) / pb q c o s l / d ( q s d HERA I+II DataFitSM Preliminary H1 W Polarisations at HERA F - F HERA I+II DataSMSingle Top N o t a ll o w ed H1 Preliminary s s F - F W Polarisations at HERA
Figure 5. Top: A simultaneous fit (solid his-togram) of F − and F to the measured differen-tial cross section (points), where the error barsdenote the statistical uncertainty only. The SMprediction (dashed histogram) is shown with a15% theoretical systematic uncertainty (hatchedarea). Bottom: The fit result for the simulta-neously extracted W boson polarisation fractions F − and F (point) with 1 and 2 σ CL contours.The predictions for the SM prediction (triangle)and anomalous single top production via FCNC(square) are also shown. tion 2 is also shown. In the fit, the optimal val-ues for F − and F are simultaneously extracted,the result of which is shown in figure 5 (bottom).The measured W boson polarisation fractions arefound to be in good agreement with the SM pre-diction and compatible with anomalous single topproduction via FCNC within 1 σ confidence level(CL).Values of F − and F are also extracted in fitswhere one parameter is fixed to its SM value. Theresults are presented in table 3 and show goodagreement with the SM. The quoted systematicuncertainties are propagated from the differentialcross section calculations.
6. Search for Anomalous Single Top QuarkProduction
The production of single top quarks is kine-matically possible in ep collisions at HERA dueto the large centre of mass energy of up to √ s =320 GeV. The dominant SM process for singletop production at HERA is the charged currentreaction e + p → ¯ νt ¯ bX ( e − p → ν ¯ tbX ), which has anegligible cross section of less than 1 fb. However,in several extensions of the SM the top quark ispredicted to undergo Flavour Changing NeutralCurrent (FCNC) interactions, which could lead toa sizeable anomalous single top production crosssection at HERA [17].Such events are of interest as the signature ofthe top decay to b and W with subsequent lep-tonic decay matches that of the isolated leptonevents discussed in section 2. In particular, thehadronic final state from the fragmentation of the b quark would exhibit high P T , and thus this pro-cess could provide an explanation of the data ex-cess observed at high P XT by the H1 experiment. Anomalous single top production is describedby an effective Lagrangian where the interactionof a top with u -type quarks via a photon is de-scribed by a magnetic coupling κ tUγ . The con-tribution from the charm quark is expected to besmall at the large proton longitudinal momentumfraction x needed to produce a top quark, and is Although it should be noted that single top productioncannot explain the observed difference between the H1 e + p and e − p data. ingle W and Anomalous Single Top Production at HERA W polarisation fractions F − and F with statistical (stat.) andsytematic (sys.) uncertainties. The central values are obtained by fixing one parameter to the SMprediction and fitting the other. The SM prediction is obtained from a two parameter fit to the SM q ℓ cos θ ∗ distribution, where the quoted uncertainty is statistical only.H1 HERA I+II Data SM F − ± ± ± F ± ± ± κ tcγ ≡ Z boson v tUZ are supressed at HERAdue to the large Z mass, and are also neglectedin the present H1 analysis ( v tUZ ≡ W decay channels to electrons and muons.The first step in the analysis forms a top prese-lection of this event sample, by demanding goodtop quark reconstruction and lepton charge com-patible with single top production [19]. Threeobservables are investigated in this preselection,namely: the transverse momentum of the recon-structed b quark candidate, the reconstructed topmass, and the W decay angle calculated as theangle between the lepton momentum in the W rest frame and the W direction in the top quarkrest frame. The observed data distributions ofthese quantities agree well with the SM expecta-tion within the uncertainties and no evidence forsingle top production is observed.The observables are then combined into amulti-variate discriminator, which is trained us-ing a single top MC as the signal model and a W boson MC as the background model. The re-sulting discriminator output for the electron andmuon channels is found to provide good separa-tion between W and top MC events.Limits on the signal cross section are extractedfrom the discriminator spectra using a maximumlikelihood method [19]. Likelihood functions arecalculated for the electron and muon channel sep-arately. An upper bound on the cross section of σ ep → etX < κ tuγ < κ tuγ and v tuZ . The top mass isassumed to be m t = 175 GeV in order to com-pare with previous results. The new preliminaryH1 result presented here extends the bound on κ tuγ into a region so far uncovered by currentcolliders. Also shown in figure 6 are results fromthe L3 experiment at LEP [20], the CDF exper-iment at the Tevatron [21] and results from theZEUS experiment using HERA I data [4]. A newresult from CDF [22], not shown in figure 6, de-rives a limit on the branching ratio B ( t → Zq ) of3.7%, which translates as an upper limit on theanomalous vector coupling of v tuZ .
7. Summary
Searches for events containing isolated leptonsand missing transverse momentum produced in e ± p collisions at HERA are presented, performedindividually and in a common phase space withthe H1 and ZEUS detectors at HERA in the pe-riod 1994–2007. In the complete H1+ZEUS highenergy data sample, luminosity 0.97 fb − , a to-tal of 87 events are observed in the data, com-pared to a Standard Model (SM) prediction of92.7 ± P XT >
25 GeV in the e + p data, luminosity0.58 fb − , 23 data events are observed comparedto a SM prediction of 14.6 ± W productionare performed by H1, as well as a measurement D. M. South | g tu k | -1
10 1 | t u Z v | -1 | g tu k | -1
10 1 | t u Z v | -1 = 0 tcZ = v g tc k = 175 GeV t m H1 Preliminary (HERA I+II)Excluded
Excl. by ZEUSExcl. by CDFExcl. by L3
Figure 6. Exclusion limits at 95% CL on theanomalous κ tuγ and v tuZ couplings obtained atHERA (H1 [19] and ZEUS [4] experiments), LEP(L3 experiment [20]) and at the TeVatron (CDFexperiment [21]). Anomalous couplings of thecharm quark are neglected κ tcγ . Limits are shownassuming a top mass m t = 175 GeV.of the W polarisation fractions. The H1 isolatedlepton events are also examined in the context ofa search for single top production. No clear signalis observed, and an upper limit on the anomaloustop production cross section of σ ep → etX < .
16 pbis established at the 95% confidence level.
REFERENCES
1. C. Adloff et al. [H1 Collaboration],
Eur. Phys.J. C (1998) 575 [hep-ex/9806009].2. V. Andreev et al. [H1 Collaboration], Phys.Lett. B (2003) 241 [hep-ex/0301030].3. J. Breitweg et al. [ZEUS Collaboration], Phys.Lett. B (2000) 411 [hep-ex/9907023].4. S. Chekanov et al. [ZEUS Collabora-tion], Phys. Lett. B (2003) 153[hep-ex/0302010].5. U. Baur, J. A. Vermaseren and D. Zeppenfeld, Nucl. Phys. B (1992) 3.6. K. P. Diener, C. Schwanenberger and M. Spira, Eur. Phys. J. C (2002) 405[hep-ph/0203269];P. Nason, R. R¨uckl and M. Spira, J. Phys. G (1999) 1434 [hep-ph/9902296];M. Spira, [hep-ph/9905469].7. H1 Collaboration, contributed paper to HEP-EPS 2007 , Manchester, abstract ,H1prelim-07-063.8. ZEUS Collaboration, contributed paper to
HEP-EPS 2007 , Manchester, abstract ,ZEUS-prel-07-001.9. H1 Collaboration, contributed paper to ICHEP 2006 , Moscow, H1prelim-06-162.10. ZEUS Collaboration, contributed paper to
ICHEP 2006 , Moscow, ZEUS-prel-06-012.11. H1 and ZEUS Collaborations, contributedpaper to
HEP-EPS 2007 , Manchester, ab-stract , H1prelim-07-162, ZEUS-prel-07-029.12. H1 Collaboration, contributed paper to
HEP-EPS 2007 , Manchester, abstract ,H1prelim-07-161.13. H1 Collaboration, contributed paper to
HEP-EPS 2007 , Manchester, abstract ,H1prelim-07-064.14. A. Aktas et al. [H1 Collaboration],
Eur. Phys.J. C (2006) 699 [hep-ex/0604022].15. K. Hagiwara, R. D. Peccei, D. Zeppenfeld andK. Hikasa, Nucl. Phys. B (1987) 253.16. T. Stelzer, Z. Sullivan and S. Willen-brock, Phys. Rev. D (1997) 5919[hep-ph/9705398].17. H. Fritzsch and D. Holtmannspotter, Phys.Lett. B (1999) 186 [hep-ph/9901411].18. A. Aktas et al. [H1 Collaboration], Eur. Phys.J. C (2004) 9 [hep-ex/0310032].19. H1 Collaboration, contributed paper to HEP-EPS 2007 , Manchester, abstract ,H1prelim-07-163.20. P. Achard et al. [L3 Collaboration],
Phys.Lett. B (2002) 290 [hep-ex/0210041].21. F. Abe et al. [CDF Collaboration], Phys. Rev.Lett.80