1 CERN-TH-97-195 hep-ph/9708437 HERA DATA AND LEPTOQUARKS IN SUPERSYMMETRY G. Altarelli a aTheoretical Physics Division,CERN, CH-1211 Geneva 23, and Terza Universit`a di Roma, Rome, Italy I present a concise review of the possible evidence for new physics at HERA and of the recent work towards a theoretical interpretation of the signal. It is not clear yet if the excess observed at large Q2 is a resonance or a continuum (this tells much about the quality of the signal). I discuss both possibilities. For the continuum case one considers either modifications of the quark structure functions or contact terms. In the case of a resonance, a leptoquark, the most attractive possibility that is being studied is in terms of s-quarks with R-parity violation. In writing this script I updated the available information to include the new data and the literature presented up to August 1, 1997. 1. Introduction + 2 = 5 events with about 1.2 + 0.8 = 2 expected. The distribution of the first H1 data suggested The HERA experiments H1 [1] and ZEUS [2], a resonance in the NC channel. In the interval recently updated in ref. [3], have reported an 187.5 <M<212.5 GeV , which corresponds to excess of deep-inelastic e+p scattering events at x ' 0.4, and y>0.4, H1 in total finds 7 + 1 = 2 > 4 2 8 events with about 1 + 0.5 = 1.5 expected. But large values of Q ∼ 1.5 × 10 GeV ,inado- main not previously explored by other experi- in correspondence of the H1 peak ZEUS observes ments. The total e+p integrated luminosity was a total of 3 events, just about the expected num- of 14.2 +9.5 = 23.7 pb−1, at H1 and of 20.1+13.4 ber. In the domain x>0.55 and y>0.25 ZEUS = 33.5 pb−1 at ZEUS. The first figure refers to observes 3 + 2 events with about 1.2 + 0.8 = 2 the data before the ’97 run [1] [2], while the sec- expected. But in the same domain H1 observes ond one refers to part of the continuing ’97 run, only 1 event in total, more or less as expected. whose results were presented at the LP’97 Sym- We see that with new statistics the evidence posium in Hamburg at the end of July [3]. In for the signal remain meager. The perplexing fea- −1 the past, both experiments collected about 1 pb tures of the original data did not improve. First, − each with an e beam. A very schematic descrip- there is a problem of rates. With more inte- 2 > 4 tion of the situation is as follows. At Q ∼ 1.510 2 grated luminosity than for H1, ZEUS sees about GeV in the neutral current channel (NC), H1 the same number of events in both the NC and observes 12+6 = 18 events while about 5+3 = 8 CC channels. Second, H1 is suggestive of a reso- were expected and ZEUS observes 12 + 6 = 18 nance (although the evidence is now less than it events with about 9 + 6 =15 expected. In the was) while ZEUS indicates a large x continuum charged current channel (CC), in the same range 2 (here also the new data are not more encourag- of Q , H1 observes 4+2 = 6 events while about ing). The difference could in part, but appar- 1.8+1.2 = 3 were expected and ZEUS observes 3 2 > ently not completely [4], be due to the different ceivable at very large x, x ∼ 0.75, which is too methods of mass reconstruction used by the two large even for ZEUS. The compatibility with the experiments, or to fluctuations in the event char- Tevatron is also an important constraint. This is acteristics. Of course, at this stage, due to the because ep scattering is linear in the quark den- limited statistics, one cannot exclude the possi- sities, while pp¯ is quadratic, so that a factor of bility that the whole effect is a statistical fluctu- 1.5-2 at HERA implies a large effect also at the ation. All these issues will hopefully be clarified Tevatron. In addition, many possibilities includ- by the continuation of data taking. Meanwhile, ing intrinsic charm (unlessc ¯ 6= c at the relevant it is important to explore possible interpretations x values [11]) are excluded from the HERA data of the signal, in particular with the aim of identi- in the CC channel [12]. In conclusion, it is a fying additional signatures that might eventually fact that nobody sofar was able to even roughly be able to discriminate between different expla- fit the data. This possibility is to be kept in mind nations of the reported excess. if eventually the data will drift towards the SM and only a small excess at particularly large x and Q2 is left with comparable effects in NC and 2. Structure Functions CC, with e+ or e− beams. Since the observed excess is with respect to the Standard Model (SM) expectation based on the 3. Contact Terms QCD-improved parton model, the first question is whether the effect could be explained by some in- Still considering the possibility that the ob- adequacy of the conventional analysis without in- served excess is a non-resonant continuum, a voking new physics beyond the SM. In the some- rather general approach in terms of new physics what analogous case of the apparent excess of jet is to interpret the HERA excess as due to an ef- production at large transverse energy ET recently fective four-fermionee ¯ qq¯ contact interaction [13] observed by the CDF collaboration at the Teva- with a scale Λ of order a few TeV. It is interest- tron [5], it has been argued [6] that a substan- ing that a similar contact term of theqq ¯ qq¯ type, tial decrease in the discrepancy can be obtained with a scale of exactly the same order of magni- by modifying the gluon parton density at large tude, could also reproduce the CDF excess in jet values of x where it has not been measured di- production at large ET [5]. (Note, however, that rectly. New results [7] on large pT photons ap- this interpretation is not strengthened by more re- pear to cast doubts on this explanation because cent data on the dijet angular distribution [14]). these data support the old gluon density and not One has studied in detail [15] [16] vector contact the newly proposed one. In the HERA case, a terms of the general form similar explanation appears impossible, at least for the H1 data. Here quark densities are in- 4πηij µ ∆L = η 2 e¯iγ ei q¯j γµqj . (1) volved and they are well known at the same x (Λij ) but smaller Q2 [8], [9], and indeed the theory fits the data well there. Since the QCD evolution is with i, j = L, R and η a ± sign. Strong believed to be safe in the relevant region of x,the limits on these contact terms are provided by proposed strategy is to have, at small Q2,anew LEP2 [17] (LEP1 limits also have been consid- component in the quark densities at very large x, ered but are less constraining [18,19]), Teva- beyond the measured region, which is then driven tron [20] and atomic parity violation (APV) ex- at smaller x by the evolution and contributes to periments [21]. The constraints are even more HERA when Q2 is sufficiently large [8]. One stringent for scalar or tensor contact terms. APV possible candidate for a non perturbative effect limits essentially exclude all relevant AeVq com- at large x is intrinsic charm [10]. However it ponent. The CDF limits on Drell-Yan produc- turns out that a large enough effect is only con- tion are particularly constraining. Data exist 3 both for electron and muon pairs up to pair the production at HERA occurs from valence u or masses of about 500 GeV and show a remark- d quarks, since otherwise the coupling would need able e − µ universality and agreement with the to be quite larger, and more difficult to reconcile SM. New LEP limits (especially from LEP2) have with existing limits. However production from been presented [17]. In general it would be the sea is also considered. Assuming an S-wave possible to obtain a reasonably good fit of the state, one may have either a scalar or a vector HERA data, consistent with the APV and the leptoquark. I only consider here the first option, LEP limits, if one could skip the CDF limits because vector leptoquarks are more difficult to [22]. But, for example, a parity conserving com- reconcile with their apparent absence at the Teva- µ µ bination (¯eLγ eL)(¯uRγµuR)+(¯eRγ eR)(¯uLγµuL) tron. The coupling λ for a scalar φ is defined + + with ΛLR =ΛRL ∼ 4 TeV still leads to a marginal by λφe¯LqR or λφe¯RqL, The corresponding width 2 fit to the HERA data and is compatible with all is given by Γ = λ Mφ/16π, and the production existing limits [22,23]. Because we expect con- cross section on a free quark is given in lowest π 2 tact terms to satisfy SU(2) U(1), as they re- order by σ = 4s λ . flect physics at large energy scales, the above phenomenological form is toN be modified into Including also the new ’97 run results, the com- µ µ µ bined H1 and ZEUS data, interpreted in terms L¯LγµLL(¯uRγ uR +d¯Rγ dR)+¯eRγµeRQ¯Lγ QL), where L and Q are doublets [24].