NEW PARTICLE SEARCHES

J.-F. Grivaz

Laboratoire de FAcce'le'rateur Lin6aire,IN2P3-CNRS et University de Paris-Sud, F-91405 Orsay

Rapporteur talk at the International Europhysics Conferenceon High Energy Physics, Brussels (Belgium), July 27 - August 2, 1995

U.E.Rde

I Universite Paris-Sud

Institut Nationalde Physique Nucleaire

etde Physique des Particules

Batiment 200 - 91405 ORSAY Cedex

LAL 95-83October 1995

NEW PARTICLE SEARCHES

J.-F. Grivaz

Laboratoire de 1' Acce*le"rateur Line'aire,IN2P3-CNRS et University de Paris-Sud, F-91405 Orsay

Rapporteur talk at the International Europhysics Conferenceon High Energy Physics, Brussels (Belgium), July 27 - August 2, 1995

NEW PARTICLE SEARCHESLAL 95-83

October 1995

J.-F. GRIVAZ

Laboratoire de VAccMrateur Liniaxre, IN2P3-CNRS el University de Paris-Sud, 91405 Orsay, France

Results on new particle searches from LEP, HERA and the Tevatron are presented. Limits on standardand non-standard Higgs bosons, on supersymmetric particles with and without R-parity conservation, onleptoquarks, on additional leptons, quarks and gauge bosons are reported. Fn-pnrtifntnr, thf? mnrm Inwfr

2 2gd mudel HiggH bo«oii in 05.2<3cV/c2, it is 173-GcV/c2 for the gluino and 650CcV/e2

fei a Z ' w Hit-standard mcTdel couplings. Events with unexpected topologies were observed at LEP andat the Tevatron.

1 Introduction

The results on new particle searches presented at thisconference have been obtained by the experiments in-stalled at the three highest energy machines presentlyin operation: LEP at CERN, an e+e~ collider with acentre-of-mass energy of 91.2 GeV, close to the Z mass;HERA at DESY, an asymmetric machine where 27 GeVpositrons collide on 820 GeV protons, corresponding toa 300 GeV centre-of-mass energy; and the Tevatron atFermilab, a pp collider with a centre-of-mass energy of1.8 TeV. Although no energy increase took place at any ofthese machines recently, substantial luminosity improve-ments have been achieved, resulting in an enhanced sen-sitivity for a number of searches: the number of hadronicZ decays collected was almost doubled at LEP in 1994,with up to 3.7 million such events analyzed by a singleexperiment; at HERA, the luminosity integrated in 1994was 3.2 pb"1, to be compared to 0.6 pb"1 in 1993; andat the Tevatron, while less than 20 pb"1 had been ac-cumulated up to 1993, results are now coming, based onthe analysis of 70 pb"1.

In the following, all the limits reported are given atthe 95% confidence level.

2 The standard model Higgs boson

The search for the standard model Higgs boson ispresently performed only at LEP, where the sole relevantproduction mechanism is the so-called bremsstrahlungprocess, e+e~ ->• HZ* -»• Hff. For Higgs boson massesof current interest, i.e. well beyond the bb thresholdbut also well below that for H —> WW, the main Higgsboson decay mode is H -4 hadrons, of which ~ 95%into bb, with the remaining ~ 9% into T+T~ . The finalstate topology further depends on the Z* decay products.Unfortunately, the Z* -f qq mode, which accounts for~ 70% of the Z* decays, cannot be used because of thehuge contamination from normal hadronic Z decays. Itis also the case for the final states involving Z* -f T+T~decays, which contribute only ~ 3% and suffer in ad-

dition from the r+r~qq four-fermion final state back-ground. In the end, the usable contributions come fromthe Z* -¥ e+e~ and Z* -»• fi+ft~ modes, ~ 7% together,and mostly from the Z* -> vv mode which accounts for20% of the final states.

In 1994, Higgs boson mass lower limits were re-ported at the 60 GeV/c2 level1. With a statistics almostdoubled, a substantial improvement could be hoped for.However, for a Higgs boson mass of 65 GeV/c2, the pro-duction cross-section is only 40% of what it is for a massof 60 GeV/c2. Therefore, only a few GeV/c2 increase inthe sensitivity range is to be expected.

2.1 Search in the (H-* hadrons)(Z* -+ vv) channel

The main background to the (H —> hadrons) (Z* -+ vv)channel comes from hadronic Z decays with missing en-ergy either real, due to neutrinos from heavy quark de-cays, or fake, due to jet energy measurement fluctua-tions. The tools to cope with these backgrounds havebeen elaborated over the years and comprise typically: aselection of events well contained in the detector; theseevents should be acollinear, which rejects qq final states,and acoplanar, which rejects qq"7 final states where theradiated photon escapes down the beam pipe; and toreject qqg final states, requirements on the isolation ofthe missing momentum and on the event aplanarity areimposed. In addition, detector dependent cuts may beapplied to avoid "weak regions" such as the boundariesbetween calorimeter sub-systems.

To establish their selection methods, the four LEPexperiments produce unbiased background Monte Carlosamples of a size similar to or somewhat larger than thatof their data; they isolate potentially dangerous config-urations and generate samples enriched accordingly andcorresponding to an integrated luminosity typically anorder of magnitude larger than that of the data. Thestatistical treatments applied at this point become unex-pectedly different:

• OPAL2 choose a set of classical cuts intended to betight enough for the analysis to remain unchanged

for the whole lifetime of LEP1, and indeed the mod-ifications to the analysis they reported last year arerather minor.

• ALEPH3 also use a set of cuts, but the cut valuesfor the most critical variables are optimized everyyear in an automatic way in order to cope with theincrease of the absolute background level broughtwith the additional data. To this end, they use theexpectation value 1̂ 95(0;) of the 95% CL upper limiton the number of signal events produced, calculatedunder the assumption of no signal contribution as

N95(x) = Zn>oVb(z){n)C9s(n)/e(x).

Here, Vb(x){n) - e-b^b(x)n/n\ is the probabilityto observe n events, given a background expectationb(x) for a cut location x; £95(5) is the value of the95% CL limit set when n events are detected, i.e.3.00 for n = 0, 4.74 for n = 1, e t c . ; and e(x)is the signal efficiency for the cut location x. Asthe cut becomes tighter, the background level andthe signal efficiency decrease simultaneously, and thevalue chosen is the one which minimizes N95.

• L34 now use a similar method. All cuts are simulta-neously tightened in steps, and the optimization isperformed on the step number.

• DELPHI5 aim at casting all the information into asingle variable, for instance the output of a neuralnet or, more recently, of a discriminant analysis, andthey cut on this single variable to reduce the back-ground to a given level; they need however to applyadditional specific cuts both beforehand and after-wards.

Altogether, 13 million hadronic Z decays were analyzedby the four experiments, with efficiencies ranging from 16to 31% and mass resolutions from 5 to 10 GeV/c2, bothfor a 65 GeV/c2 Higgs boson. Three candidate eventswere selected, while five were expected from backgroundprocesses. The largest of the three candidate masses is5

(38±8)GeV/c2.

2.2 Search in the (H~* t*~t~)(Z" -> vv) channel

Here too, a large potential background comes fromhadronic Z decays containing leptons from semileptonicheavy quark decays. This background can however belargely reduced by tight requirements on the energies andon the isolations of the two leptons.

Such a reduction does not occur, however, for thebackground from the process e+e~ —>• l+l~qq, in theconfiguration where the four-fermion final atate resultsfrom a Z decay into a high mass qq pair which radi-ates a virtual photon converting into an 1+1~ pair. In-deed, the topology of this background is indistinguishable

from that of the signal, but the b-tagging technique for-tunately provides an efficient discriminating tool: whilemost of the hadronic Higgs boson decays are expectedto be into bb, the four-fermion final state is reduced byan order of magnitude not only because only 22% of thehadronic Z decays are into bb but also because the emis-sion of the virtual photon is reduced by the square of theelectric charge, 1/3, of the b-quark.

In addition, the mass of a Higgs boson candidatecan be reconstructed with high accuracy, better than1 GeV/c2 at 65 GeV/c2, as the mass of the system recoil-ing against the lepton pair. As a result, any candidateevent will contribute only within a restricted mass do-main.

For masses in excess of 40 GeV/c2, nine events areselected in 11 million hadronic Z decays3'45 (OPAL didnot analyse their 1994 data for this channel). This issomewhat on the low side compared to the expectationof fifteen events from background processes, but the un-certainty on this prediction is rather large, especially be-cause of the poorly known QCD corrections to the four-fermion final state production. There is no particularmass accumulation among the nine candidate events.

2.3 Higgs boson mass limits

The limits reported by the four LEP experiments2'3'4'5

are given in Table 1.

Table 1: 95% CL lower limits reported by the four LEP collabora-tions (or the mass of the standard model Higgs boson, in GeV/c2.

Experiment

Mass limit

ALEPH

63.1

DELPHI

55.4

L3

60.2

OPAL

59.1

The most straightforward way to determine whethera given Higgs boson mass is excluded by the four ex-periments considered simultaneously would be to simplyadd the numbers of signal events expected to be found inall three channels (HJ/P, H(i+fi~ and He+e~) by all fourexperiments and to compare it with the number of candi-date events reported contributing at that mass value (forinstance, one could consider only those candidates ly-ing within two standard deviations, and reduce the num-ber of events expected by 5% accordingly). With such amethod, a limit of about 65 GeV/c2 would be set, withone candidate event contributing in that mass region.

This method can be criticized however because itdoes not result from the application of a well defined apriori prescription. For instance, while all collaborationsunanimously decided long ago not to use the Z* —• qq

decays, they chose only progressively and not simultane-ously to stop using the H ->• T+T~ decays. Implicitly, thebasis for such choices is that channels which bring toomuch background for too little added efficiency shouldnot be used. This can be turned into a quantitative state-ment using the variable N95 defined previously6: givena set of analyses, each with its efficiency, its expectedbackground level and its mass resolution, the subset tobe chosen is the one which minimizes N95. With thisprescription, it turns out that, out of the twelve analysesreported (three channels times four experiments), fourshould be removed from the combination. The result ofthis procedure is shown in Fig.l, where the mass infor-mation for the candidate events in the eight remaininganalyses has been treated as advocated in Ref. 7. Theresulting 95% CL lower limit for the mass of the standardmodel Higgs boson is 65.2 GeV/c2.

50 52 54 56 58 60 62 64 66 68 70

Figure 1: Number of events expected and 95% CL upper limit onthe number of observed events in the eight selected analyses.

3 Non standard Higgs bosons

3.1 Search for Z -> Hf

In the standard model, the process Z ~¥ H7, which pro-ceeds mainly via top quark and W boson loops, occurswith a rate far too low to be worth consideration, es-pecially in view of the huge background from final stateradiation in hadronic Z decays. This rate may howeverbe enhanced in some extensions of the standard model,usually involving anomalous couplings, hence the analy-ses reported by DELPHI8 and by OPAL9. Isolated ener-getic photons are selected in hadronic events, and a peak

is looked for in the photon energy spectrum above thefinal state radiation continuum. As in the case of theHl+1~ final state, the search sensitivity is enhanced byan order of magnitude when b-tagging is applied. De-cay rates thirty times larger than the standard modelpredictions are thus excluded for Higgs masses around50 GeV/c2.

3.2 Higgs bosons in the MSSM

In the minimal supersymmetric extension of the standardmodel, the MSSM, the neutral Higgs sector contains, asin any two-doublet model, two CP-even bosons, h andH with mn < mjj, and one CP-odd boson, A. Theircouplings are controlled by a, the mixing angle in theCP-even sector, and by tan/9, the ratio of the vacuumexpectation values of the two Higgs fields. At the treelevel, the Higgs boson masses and the angles a and 0can all be calculated from only two inputs, e.g. m^ andtan /?. Additional parameters are however needed in thecalculation of the radiative corrections which, because ofthe large top quark mass, turn out to be quite substan-tial for the CP-even Higgs bosons. Besides mj, theseparameters are the average scalar top mass, mi, and theamount of mixing in the scalar top sector, controlled bythe parameters At and ft.

The cross-section for the process e+e~ —• hZ* isequal to its standard model analogue, up to a reductionfactor sin2(/? — a). In addition, a new production mech-anism takes place if kinematically allowed, e+e~ —• hA,the cross-section of which is proportional to cos2(/? — a),which renders the two processes complementary. Fortan/? > 1, as theoretically preferred, the main decaymodes of h and A are into r + r~ and bb, with branch-ing ratios not much different from those of the standardmodel Higgs boson. Exceptions arise when the chan-nel h —>• AA is kinematically allowed, and then usuallydominant, or when decay channels into supersymmetricparticles are open, a possibility which will be discussedfurther down.

Searches for the MSSM Higgs bosons have been re-ported at this conference by ALEPH10 and by L311.Both use their most recent results on the standard modelHiggs boson3'4, and in addition updated their searches forZ -4 hA decays. For large h and A masses, the kinematiclimit is reached using the r+r~qq final state which can bemade essentially background free by suitable topologicalcuts. A similar sensitivity was also achieved by ALEPHin the bbbb final state, using b-tagging in four-jet events.The L3 result is shown in Fig. 2 in the {m^yin^) plane fortypical top and scalar top mass values, and the ALEPHresult in Fig. 3 in the (mn,tan/3) plane, with the depen-dence on the mixing in the scalar top sector also indi-cated.

!

70

(0

10

40

30

20

10

L3-

not allowed

bythMfy

II

//

Excluded by L3«tS5 %CL

i

I allowwl /

'] /

f/ not allowad

/ bytlMonr

\ 1 1 « 1 »

20 40 M M

Mh(GeV)

Figure 1: Domain excludedtan/? > 1, for m t = 176

ed by L3, in the ("»h'mA) plane and forGeV/c2, m-t = 1 TeV/c", and no scalar

top mixing.

140

mh (G«Vfe2)

Figure 3: Domain excluded by ALEPH, in the (mn,tan/3) plane, form t = 175 GeV/c2 and mj = 1 TeV/c2. The full lines correspondto the case of no scalar top mixing, and the dotted and dashed

lines to cases of typical and maximal mixings, respectively.

For large tan 0, only Z —¥ hA contributes; the searchis then kinematically limited, except when tan/? is solarge that the h and A widths can no longer be ne-glected, allowing maases beyond rn^/2 to be probed.The sensitivity at low tan j3 may still improve somewhatvia the search for a standard model like Higgs boson ine+e- ->hZ*.

3.3 Invisible Higgs bosons and mono jets

The searches for the standard model Higgs boson are notsensitive to invisible Higgs boson decays. Such decaysmay occur, and even be dominant, for instance in super-symmetric models if the channel h —> \X i s kinemati-cally accessible (x is the lightest neutralino, commonlyassumed to be the lightest supersymmetric particle), orin majoron models. However, the bremsstrahlung pro-cess e+e~ -> hZ*, with Z* -> ff, can still be used in thatcase if if is an 1+1~ or a qq pair, but of course not ifit is a UV pair. The topology is similar to the one ex-pected from the production of a standard model Higgsboson with Z* -4 i/u, and the same analysis can thereforebe used, extended to the monojet topology to cope withhigh mass invisible Higgs bosons and thus with low massvisible systems arising from the Z* decay.

The result of such an analysis performed by OPAL2

is shown in Fig. 4. It can be seen that, for a standardmodel like production cross-section, the excluded massdomain extends to 67.5 GeV/c2, i.e. beyond the limitfor the standard model Higgs boson; this is because theZ* —• qq mode, with its large branching ratio, can beused here, in contrast to the case of the standard modelHiggs boson search. For very light invisible Higgs bosons,the reduction in the production cross-section is of at leastthree orders of magnitude with respect to the standardmodel one. Similar results were also reported by L34.

Following their search for monojet events, performedinitially in the context of a search for invisibly decayingHiggs bosons, ALEPH reported last year the observationof two unusual events12: a 3 GeV/c2 mass e+e~ pairand a 5 GeV/c2 hadronic system, with momenta trans-verse to the beam axis of 20 and 18 GeV/c, respectively.They proposed an interpretation of these events as dueto the process e+e~ —¥ i/Pff which occurs predominantlyvia the so-called conversion diagram e+e~ —> Z*7* withZ* —> i/i> and 7* —> ff. The probability of observing suchlarge transverse momenta was estimated to be at the fewpercent level. This year, DELPHI13, L314 and OPAL1516

presented results from new dedicated searches for mono-jet events. They all found candidates, but with moder-ate transverse momenta and in agreement with standardmodel expectations.

SJJOtVft*

95% CL excluded hadroateoajklale

OPALpreliminary

M » 30 N (0 70

Figure 4: For an invisibly decaying Higgs boson, 95% CL upperlimit reported by OPAL for the ratio of the production cross-sectionto its standard model analogue, as a function of the Higgs bosonmass. The cusps correspond to the occurrence of candidate events.

I tai#4& KHtlMttt M. Mill* flM »ITn | t"H* Ih4> IMHtttit-ifa* M I) lulfh 1 tatI VM M.tll ltd M.I kin If.ltltfl .IU 41, .ll| BM|*> ll^nikift I) Httfh I t«l<K-i-w >.... m n. w. »..- m i W •"««;• ;•:«

Figure 5: The OPAL monojet event, with unexpectedly large massand transverse momentum

One highly atypical event was however reported byOPAL16. This event, an 18.4 GeV/c2 mass hadronic sys-tem with a 28.3 GeV/c transverse momentum, is dis-played in Fig. 5. There is essentially no longitudinal mo-mentum, and the mass of the invisible recoiling systemis 50 GeV/c2. Although they expect about one eventwithin their selection cuts, the probability for such largemass and transverse momentum values is only 2%, andthis seemingly large level is obtained only once the W-photon scattering process e+e~ -> Ped~ue~ is included,with the spectator electron remaining undetected in thebeam pipe.

4 Supersymmetric particles

In supersymmetric theories, any ordinary particle, orrather any of its degrees of freedom, has a supersymmet-ric counterpart of which it differs by half a unit of spin.Each charged lepton or quark thus has two scalar part-ners called sleptons, In and li, or squarks, qfi and qL;and to each neutrino is associated a sneutrino, v. Thegauge bosons have fermionic partners called gluino, g,photino, 7, wino, W^ and zino, Z, and the Higgs bosons(two doublets at least are needed) also have fermionicpartners called higgsinos. The winos and charged hig-gsinos mix to form two charginos, the lighter of whichis denoted x*. while the photino, the zino and the neu-tral higgsinos mix to form at least four neutralinos. Inthe MSSM, there are only two Higgs doublets and there-fore only four neutralinos, denoted x> X1 > x" and x'" m

increasing mass order.The lightest neutralino, \, is commonly assumed to

be the lightest supersymmetric particle, the LSP. It in-teracts only weakly with ordinary matter and, if R-parityis conserved, it is stable. R-parity, a new multiplicativequantum number which takes the value +1 for ordinaryparticles and —1 for their superpartners, is assumed to beconserved in the following, except when explicitly stated.By R-parity conservation, supersymmetric particles areproduced in pairs and (cascade) decay until the LSP isreached. These stable, neutral, weakly interacting LSPsescape detection, which is at the origin of the celebratedsignature of supersymmetry: missing energy.

4-1 Gluinos, squarka and stops

If light enough, squarks and gluinos, which are stronglyinteracting particles, should be abundantly produced atthe Tevatron in reactions such as pp -+ qqX, pp -> qgX,PP -* 88^. Depending on the mass hierarchy betweensquarks and gluinos, the subsequent decays will be: i)q -+ qg and g -* qq(x or \± or *')> or it) g -+ qqand q -> q(x or x* or x')- The branching ratios intothe various charginos and neutralinos are highly modeldependent and are usually calculated within the MSSM

for specific parameter choices (in particular for /*, thehiggsino mass term, and tan/3). The same parametersgovern the subsequent \' o r X* decay chains.

By inspection of the squark and gluino decay pat-terns, it can be seen that the final state should exhibit asubstantial number of jets, and also missing energy car-ried away by the LSPs. Both CDF17 and DO18 have up-dated their limits on squark and gluino production. Asan example, the DO selection requires a missing trans-verse energy of at least 75 GeV, three jets with a trans-verse energy in excess of 25 GeV, and no azimuthal cor-relation between the directions of the missing energy andof the jets. The background from W+jets, with W —> lv,is reduced by a veto on isolated leptons. The numberof events selected in 13.5 pb"1 is fourteen, in agreementwith expectation from standard model backgrounds. Thedomain thus excluded in the plane of the squark andgluino masses is shown in Fig. 6, where it is assumedthat all but the top squarks are mass degenerate. Thesensitivity to gluino pair production in the case of verylarge squark masses was increased by a complementarysearch where the requirement on the amount of missingenergy is reduced to 65 GeV and that on the jet ener-gies to 20 GeV, but in which at least four jets ratherthan three are required. Only five events remain, andthe gluino mass limit thus obtained is 173 GeV/c2.

600

600

I "3 300

800

100

i '

I

I I

• DZERO experiment 96% CX, —

. DZKRO UPDATEh (PREUMINARrf

masses are assumed to be equal at unification scale, thestop masses get renormalized differently because of thelarge top Yukawa coupling, and the stops tend to becomelighter in this process than the other squarks. Secondly,the off-diagonal terms in the squark mass matrix are usu-ally neglected with respect to the diagonal terms; this isbecause they are proportional to the quark mass whilethe diagonal terms are proportional to the direct renor-malized squark masses. This assumption is likely not tobe valid in the case of stops because of the large value ofthe top quark mass, in which case one of the stop masseigenstates will be significantly lighter than any of the di-agonal mass terms. In addition, if the mass of this lightstop is smaller than mj,+mxd: and than m^ + myj +mx,its main decay channel will be into ex via a loop. Thefinal state arising from stop pair production in pp colli-sions will therefore contain two c-jets and missing energy.Since the above described analyses require at least threejets, they are not sensitive to such a topology. A dedi-cated analysis has therefore been developed by DO19 tocope with this particular signature, the results of whichare shown in Fig. 7. The LEP limit has been somewhatimproved to 45 GeV/c2 by ALEPH2Q recently (not shownin Fig. 7).

>«0

6

100 200 300 400 600 600Gluino mass (G«V/o2)

DO PRELIMINARY

Figure 6: Squark and gluino mass limits obtained by DO. The fullline results from the analysis requiring at least three jets, while thelong dashed line indicates the improvement achieved when incor-

porating the analysis requiring at least four jets.

As stated above, these exclusion domains were de-termined for ten mass degenerate squarks. There arehowever reasons to believe that a supersymmetric part-ner of the top quark, dubbed stop, could be significantlylighter than the other squarks. Firstly, even if squark

Figure 7: Domain excluded by DO in the (mj, mx) plane.

First generation squarks can also be searched atHERA in ep collisions via the process ep -+• eq wherea neutralino is exchanged in the ^-channel. The topol-ogy of the final state is an acoplanar electron-jet pairand is easy to disentangle from the deep inelastic scat-tering background. The HI analysis21 excludes selec-trons and squarks such that the sum of their masses is

smaller than 100 GeV/c2, provided that the exchangedneutralino mass is smaller than 30 GeV/c2 and that thisneutralino is dominantly gaugino-like.

4-2 Charginos and nevtralinos

The Z coupling to charginos is large, so that the Z to-tal width measurement at LEP is sufficient to excludecharginos with masses up to m^/2, irrespective of themodel parameters. On the contrary, the couplings of theZ to the various neutralinos strongly vary with these pa-rameters, and may become very small for some choices;for instance, the Z\x coupling vanishes when x i8 a

pure photino. It is therefore worthwhile updating theneutralino searches as the number of Z decays collectedincreases. Results were reported at this conference byOPAL2 and recently published by L322. The kinemat-ically most favourable decay channel Z -»• XX can beaddressed only via the invisible Z width measurement.The channels investigated are therefore Z -+ \x! tua^Z -> X'X'J with subsequent x' -*• X& o r X1 -*• Xf decays.The L3 results are shown in Fig. 8 as excluded domainsin the (M2, (i) plane of the MSSM for two values of tan/3(M2 is the gaugino mass term associated with the SU(2)z,gauge group).

Assuming unification of the gaugino masses, theseresults can be combined with the DO gluino mass limitquoted above which basically excludes Afj values smallerthan 49 GeV/c2. It then results a x mass lower limit of20 GeV/c2.

Charginos and neutralinos have also been searchedat the Tevatron where they could be produced via theprocess qq" -f W* -+ x+x'- A clear signature ariseswhen both gauginos decay into leptons: x+ —• x'" a ndX1 -*• x'+l • The final state is characterized by threeleptons, some missing energy and no jet activity. Theresults reported last year by CDF, based on 19 pb"1,have not been updated17. The chargino mass limit theyobtain, 45 GeV/ca, is similar to that achieved at LEP.However, this limit is calculated within the MSSM andis strongly dependent on some of the model parameters,in particular those which determine the leptonic decaybranching ratios of the charginos and neutralinos; thechoice they make, m* = 1.2m», turns out to be quitefavourable. Besides, when gaugino mass unification isassumed, the DO gluino mass limit quoted above trans-lates into a more constraining chargino mass limit, e.g.60 GeV/c2 for tan/? = 2, with ft = -400 GeV/c2 as cho-sen by CDF. The conclusion is that, at the Tevatron, thedirect gluino search is presently more sensitive than thegaugino search via the trilepton signature.

250

200

150 -

100 -

50 -

-200 -100 100 200

250

200 -

150 -

100 -

50

•200 •100 100 200H(QeV)

Figure 8: Domains excluded by L3 in the (Mj, it) plane of theMSSM, for two value* of tan 0, using the full and invisible Z width

measurements and direct searches for neutralinos.

4-3 R-parity violation

In the MSSM, the superpotential contains the terms

hLHjS, h'QHj), h"QH2U

which are responsible for the Yukawa couplings of theHiggs fields to the ordinary fermions, and thus for thefermion masses. In the above expression, the generationindices have been dropped for simplicity. Other terms arehowever allowed by the general constraints of supersym-metry, renormalizability and gauge invariance, namely

XLLE, X'LCfD, \"TWD.

In contrast to the MSSM terms, they violate the lep-ton or baryon numbers and would lead, if simultaneously

present, to proton decay at an unacceptable rate. R-parity conservation, with R = (_i)3fl-^+2*> w as intro-duced to forbid all such terms, but this may be viewedas a somewhat ad hoc prescription to which there are al-ternatives, for instance B-parity conservation which for-bids only the X"UDD terms. When R-parity conserva-tion is not enforced, the phenomenological analysis be-comes extremely intricate unless the further assumptionis made that one of the R-parity violating couplings dom-inates over all the others, in a fashion similar to the topYukawa coupling in the R-parity conserving sector. Themain consequence of R-parity violation is that the LSPis no longer stable. For instance, and introducing nowthe generation indices, a dominant A[23 coupling will in-duce decays such as \ -> r~/i+j/e, and a dominant A'131

coupling decays such as \ ~* dbPe-A comprehensive search for signals of supersymme-

try in Z decays was performed recently by ALEPH23,under the assumption of a dominant A-type term. TheLSP, assumed to be the lightest neutralino, decays totwo charged leptons and a neutrino. The final statesignatures therefore become multileptons with moderatemissing energy, instead of the usual large missing energy.Slepton, sneutrino and squark pair productions lead tosix-lepton, to four-lepton and to four-lepton+two-jet fi-nal states, respectively. The associate \x' productionleads to six-lepton or four-lepton+two-jet final states.But the most dramatic change occurs with the LSP pairproduction, Z —>• \X> which, instead of being addressedonly via the invisible Z width measurement, can now bedetected with four leptons in the final state. The resultof these analyses is that all limits on supersymrnetric par-ticles are at least as constraining as in the MSSM case.An example is shown in Fig. 9 where the effect of thedetectability of the \X final state is demonstrated.

Resonant production of single supersymmetric par-ticles is also a feature of models with R-parity violation.This has been addressed at HERA where squarks couldbe produced by positron-quark fusion, e.g. e+d —>• 0 if aA'nl coupling is dominant. If the produced squark subse-quently decays via the same R-parity coupling, u —>• e+d,the final state will exhibit a positron-jet coplanar pair;the deep inelastic scattering background is reduced bythe requirement of a high y value, and a peak in thepositron-jet mass is then searched over the remainingcontinuum. If the squark decays to a quark and to theLSP via a gauge coupling, u —>• \i\, the LSP further de-cays with R-parity violation, \ —• e+du or x ~• e~du;in the latter case, the decay electron has the wrong signwith respect to the beam, which leads to a backgroundfree topology. Both signatures have been analysed byHI24. Since the R-parity violating coupling is involvedin the production mechanism, the squark mass limit de-pends on the value of that coupling. Results obtainedwith 3.15 pb"1 are shown in Fig. 10 where it can be seen

that squark masses up to 240 GeV/c2 are excluded for

LJL,t.I f.K.Lt.X.4 H.+ I.f) . l.V.t.t.1.4.

- 2 -1.5 - 1 -0.5 0 0.5 1 1.5

Figure 9: Domain excluded by ALEPH in the ("»«,, 41) plane inthe case of R-parity violation with a A-type coupling: from the Zwidth measurement (light grey), from the search for neutralinogin the Z —¥ xx' a n c ' Z —¥ x'x' modes (heavy grey), and from the

search in the Z -+ XX mode (black).

10

160 GeV

80GeV40GeV

H1 preliminary. . . . i . . . . i . . . . i . . . . i

SO 73 100 125 150 173 200 223 250

M,~(G.V)

Figure 10: Domain excluded by HI in the (m^, \[u) plane.

5 Pot pourri of other non-standard particles

5.1 Heavy gauge bosons

Extra Z-like bosons, denoted Z', are predicted in most ex-tensions of the standard model based on extended gaugegroups (but not in the simplest SU(5) GUT). Extra W-like bosons may also occur in such extensions, eitherof the sequential type, denoted W , or mediating right-handed currents, denoted W/j. In the latter case, right-handed, and probably heavy, neutrinos are also expected.

Heavy Z' bosons can be easily detected at the Teva-tron through the reactions qq" -* Z' -+ e+e~ or fi+p~.They should show up as a peak in the lepton pair massover the Drell-Yan continuum. A recent CDF17 analysis,based on 70 pb~ l , excludes a Z' with standard-model likecouplings up to 650 GeV/c2.

If heavy enough, extra W bosons of the sequentialtype will decay to a W-Z pair. CDF17 have investigatedin 67 pb"1 the topology arising from subsequent W -+ ei/and Z -»• qq* decays, namely a high energy electron, sub-stantial missing transverse energy and a pair of energeticjets. The mass reconstructed from these four ingredientsshould show a jacobian peak at the mass of the W'. Thedata are compatible with the background from W+jetsproduction, which excludes such an extra W in the rangefrom 205 to 400 GeV/c2.

Heavy right-handed W bosons have been searchedby DO25 in 13.5 pb^of data. They look for the decaychain Wn —• eun -> eeqq in two ways. The inclusiveelectron spectrum should show an excess at high energy,uncorrelated with any significant missing energy, in con-trast to the W+jets background. For large VR masses,the direct electron becomes too soft, but the complete fi-nal state, two electrons and two jets, can be resolved andeasily disentangled from the background, mostly due toZ+two-jets production. As can be seen in Fig. 11, a WJIwith mass smaller than 500 GeV/c2 is excluded as longas the associated right handed neutrino mass is smallerthan 240 GeV/c2.

5.2 Isosinglet heavy neutrinos

In many extensions of the standard model, ordinary neu-trinos are massive and mix, e.g via a see-saw mecha-nism, with heavy right handed neutrinos. Such isosin-glet neutrinos can be produced in Z decays only throughtheir small mixing with ordinary neutrinos. Therefore,associate production of a light and a heavy neutrino isfavoured over pair production of heavy neutrinos not onlyfor phase space reasons but also because only one powerof the mixing angle is involved in the amplitude. Onceproduced, the heavy neutrino decays into a neutrino orcharged lepton and a fermion pair, with a lifetime alsocontrolled by the mixing angle. Many topologies aretherefore to be considered, depending on the decay mode

of the heavy neutrino, its mass and its lifetime. For in-stance, monojets emerging from a vertex detached fromthe interaction point could occur. DELPHI13 have per-formed a comprehensive analysis of all these topologies,the result of which is shown in Fig. 12.

.700

100 200 M O W O 100 600 700

Figure 11: Domain excluded by DO in the plane of the W/t mavi. the I/R mass.

MassofNHL(GeV)f

Figure 12: Domains excluded by DELPHI and by other experi-ments in the plane of the isosinglet neutrino mass vt. the square

of the mixing angle.

5.S Leptoquarks

Extended gauge group and compositeness models oftenpredict particles carrying both lepton and baryon num-bers, and called leptoquarks. Such leptoquarks decayinto a charged lepton or neutrino and a quark, all withina given generation, with model dependent branching ra-tios.

Being coloured objects, leptoquarks can be abun-dantly pair produced in pp collisions. CDF17 updatedtheir limits on second generation leptoquarks, using67 pb~ of data. The signature they considered is theone which arises when both produced leptoquarks de-cay into a muon and a quark: two muons and two jets,all with transverse energy larger than 20 GeV, and withthe muon pair mass not falling near the Z mass. For abranching ratio into fiq of 100%, they exclude leptoquarkmasses smaller than 180 GeV/c2. The limits obtained forother values of the branching ratio are shown in Fig. 13.

9> 08+It

0.6

0.4

0.2

100 160 200

Second Generation Leptoquark Mass (GeV/c )

Figure 13: Domain excluded by CDF in the plane of the secondgeneration leptoquark mans vs. the branching ratio into a muon

and a quark.

Limits on first generation leptoquarks were set byDO26 who consider not only the case where both lepto-quarks decay into an electron and a quark, but also thatwhere one of the two leptoquarks decays into a neutrinoand a quark. The latter topology is selected requiring anelectron and two jets, all with transverse energy above20 GeV, a missing transverse energy larger than 40 GeVand a transverse mass of the electron-neutrino systemlarger than 105 GeV/c2 in order to remove the W+jetsbackground. With 14 pb""1 of data analysed, a mass

lower limit of 131 GeV is set if the branching ratio intoan electron and a quark is 100%.

First generation leptoquarks can also be produced byelectron or positron-quark fusion at HERA. The produc-tion mechanism and the decay signatures are very simi-lar to those involved in the search, discussed above, forsquarks produced with R-parity violation. The resultsreported by HI27 extend the DO limit on first generationleptoquarks as soon as the coupling at the productionvertex is larger than about a tenth of the electromag-netic coupling.

5-4 Excited leptons and quarks, augluons

Excited leptons and quarks are a common feature of com-positeness models. At HERA, excited electrons can beresonantly produced in electron-photon collisions, wherethe photon is radiated by the proton beam. The excitedelectron e* can decay into e^ or, if its mass is sufficient,into eZ or i/W. The final state topology is then a copla-nar ef pair, an electron-jet-jet system or an acoplanarpair of jets with missing transverse energy, where thetwo jets result from a Z or W decay. These characteris-tic signatures have been investigated by ZEUS28 who setimproved limits on the compositeness scale as a functionof the excited electron mass. HI29 achieve similar limitsin the case of e* —> e7.

Excited quarks could be produced at the Tevatron.Their decay into a quark and a gluon would lead to apeak in the jet-jet mass distribution, at the mass of theexcited quark, over the QCD continuum. The level of thiscontinuum can be inferred from a fit to the side bands inorder to minimize the theoretical uncertainties. CDF17

analysed this way 70 pb"1 of data, from which theyexclude excited quarks with masses up to 600 GeV/c2.This same analysis allows limits to be set on various ob-jects decaying into two jets, as shown in Fig. 14. Forinstance, axigluons, a feature of "chiral colour" modelswhere colour-SU(3) is viewed as the diagonal group ofSU(3)t xSU(3)fl, are excluded up to masses of 1 TeV/c2.This is a rather unexpectedly high value for a limit froma direct search in the pre-LHC era...

6 For the record

A number of results were submitted, which are not dis-cussed in detail here due to lack of time. These includenew limits on:

• Lepton flavour violations in Z decays30;

• Z decays31 to 777, f/7 and 777;

• High mass 77 resonances32 in e+e~ —• 1+1~77,

Invisible resonances33 in e+e~ —>• 7X;

10

1 "e

» •

w»

w

1

*

N \ CDF Preliminary 1A + 1B (70 ptr')\ \ \ • 95* Cl Upper Um«

(3aO<M<47O<Mtr) ̂ . \

Z—>aS """"" xS."(4io<H<«o(»0 \

E>OlquarkD->qq(37(KM<4aaC«V)

h)^<2.0. le<w«1<2/3

» MO 400 5(9 MO

_ AxigluonA-^qa. i(200<kV< 1000 GfV EsehidMl)

_ _ _ Excited Q u o r k q * ^ q g(2t»<lf<SO0 OW Excluded) "

i ) T«chnirtio^,-fg-fq)),gg\ . (270<H<SI0C«VE«c(udod)

*\̂ -

.X^k Nno MO MO 1000 <ioo

New Paitide Man (G»V/c*)

Et=59GeV

Figure 14: Upper limits obtained by CDF on the cross-sections forvarious new particles.

• Fourth generation Dirac neutrinos34, mixing domi-nantly with electron-type neutrinos, and producedin ud -+ W -»• ei>4;

• Excited neutrinos35 in Z -> vv* -> vv"i\

• Particles with anomalous charge in Z decays36.

Some relevant results were reported in other sessions:

• A contact interaction analysis of deep inelastic scat-tering in ep collisions by HI37;

• A search for new particles decaying into tt byCDF38;

• A search for strangelets in Pb-Pb collisions at158 A GeV/c at the SPS39.

Finally, two "zoo events" were presented by CDF17,for which no clear standard model interpretation is avail-able. The first event, shown in Fig. 15, contains two elec-trons and two photons, all well isolated and with trans-verse energies in excess of 30 GeV; in addition, there isno jet activity and there are 53 GeV of missing transverseenergy. The other event contains three isolated leptons,one muon and two electrons, with transverse energies of27, 23 and 182 GeV, a jet with an 83 GeV transverseenergy, and 106 GeV of missing transverse energy.

7 Outlook

Next year, LEP data collected in Fall '95 at a centre-of-mass energy of 140 GeV should allow chargino masses

lt=53GeV

Figure IS: One of the CDF zoo events.

up to almost 70 GeV to be probed, even with only afew pb"1 of integrated luminosity. By the end of RunIb, both CDF and DO will have accumulated a wealthof new data, with integrated luminosities well in excessof 100 pb"1, of which all new particle searches shouldlargely benefit. It will be interesting to watch whetherthe population of the zoo will stabilize or keep increasing.

The forthcoming years will be even more exciting fornew particle searches, and in particular for Higgs hunt-ing, with the advent of LEP2 in 1996 and its foreseenenergy upgrades up to 192 GeV, and with the main in-jector coming into operation at Fermilab toward the endof the millenium.

Acknowledgments

I wish to thank the ALEPH, CDF, DELPHI, DO, HI, L3,OPAL and ZEUS Collaborations for making their resultsavailable to me, even in a preliminary form, before thebeginning of this conference. I am grateful to PatrickJanot for his help in the combination of the LEP resultson the standard model Higgs boson. I also thank Ste-fan Simion for his assistance in the preparation of thismanuscript.

References

1. M. Pohl, "Search for New Particles and New Interac-tions", in Proceedings of the XXVIIth InternationalConference on High Energy Physics, Glasgow, Scot-land, U.K., 20-27 July 1994, Eds. P.J. Bussey andI.G. Knowles, pp. 107-118.

11

2. OPAL Collaboration, Search for Neutralinos, Stan-dard Model and Invisibly Decaying Higgs Bosons inAcoplanar Jets or Monojet Topologies, contributedpaper No. 339.

3. ALEPH Collaboration, Improved Mass Limit for theStandard Model Higgs Boson, contributed paper No.414.

4. L3 Collaboration, Search for Neutral Higgs bosonsproduction through the process Z -¥ Z*H°, con-tributed paper No. 107.

5. DELPHI Collaboration, Update of the StandardModel Higgs boson search, contributed paper No.529.

6. P. Janot, Nucl. Phys. B (proc. suppl.) 38 (1995)264.

7. J.-F. Grivaz and F. Le Diberder, Nucl. Instr. Meth.A 333 (1993) 320.

8. DELPHI Collaboration, Update of the search forthe Z decay into a Higgs boson and a photon, con-tributed paper No. 533.

9. OPAL Collaboration, Search for a Narrow Reso-nance in the Z Decays into Hadrons and IsolatedPhotons, contributed paper No. 331.

10. ALEPH Collaboration, Search for the hA ->• bibbfinal state in two Higgs doublet models, contributedpaper No. 415.

11. L3 Collaboration, Search for Pair Production ofNeutral Higgs Bosons in Two-Doublet Models, con-tributed paper No. 98.

12. D. Buskulic et al., ALEPH Collaboration, Phys.Lett. B 334 (1994) 244.

13. DELPHI Collaboration, Search for Monojets andI so single t Neutral Heavy Leptons, contributed paperNo. 538.

14. L3 Collaboration, Search for Monojets in Z decayswith the L3 Detector, contributed paper No. 115.

15. OPAL Collaboration, Search for Low MultiplicityEvents with Anomalous Topologies Using the OPALDetector at LEP, contributed paper No. 336.

16. OPAL Collaboration, Search for Monojet Events Us-ing the OPAL Detector at LEP, contributed paperNo. 337.

17. CDF Collaboration, Exotic physics at CDF, con-tributed paper No. 769;CDF/PUB/EXOTIC/PUBLIC/3192, 3199, 3210.

18. DO Collaboration, Search for Squarks and Gluinos inpp collisions at the DO Detector, contributed paperNo. 434.

19. DO Collaboration, Search for Light Top Squarks withthe DO Detector, contributed paper No. 435.

20. ALEPH Collaboration, Search for scalar top quarksin e+e~ collisions at LEP1 energies, contributed pa-per No. 416.

21. HI Collaboration, First Search for the Minimal Su-persymmetric Model at HERA, contributed paperNo. 463.

22. M. Acciarri et al., L3 Collaboration, Phys. Lett. B350 (1995) 109.

23. D. Buskulic et al., ALEPH Collaboration, Phys.Lett. B 349 (1995) 238.

24. HI Collaboration, A Search for Rp-Violating SUSYSquarks in HI at HERA, contributed paper No. 784.

25. DO Collaboration, Searches for new gauge bosons us-ing the DO Detector, contributed paper No. 437.

26. DO Collaboration, Search for First and Second Gen-eration Leptoquarks at DO, contributed paper No.432.

27. HI Collaboration, A Search for Leptoquarks in HIat HERA, contributed paper No. 462.

28. ZEUS Collaboration, Duvet Searches for ExcitedFermions in ep Collisions, contributed paper No.373.

29. HI Collaboration, A Search for Excited Electrons inep scattering at HERA, contributed paper No. 464.

30. DELPHI Collaboration, Search for Lepton FlavourNumber violating Z Decays, contributed paper No.532;L3 Collaboration, Search for Lepton Flavour Viola-tion in Z Decays, contributed paper No. 110;OPAL Collaboration, A Search for Lepton FlavourViolating Z Decays, contributed paper No. 335.

31. DELPHI Collaboration, An updated analysis of thee+e~ —• 77(7) reaction at LEP energies, con-tributed paper No. 531;OPAL Collaboration, see Ref. 15.

32. DELPHI Collaboration, Search for a high mass res-onance in /+r"77, J/P77, and qTjyj events, con-tributed paper No. 530.

33. OPAL Collaboration, Measurement of Single Pho-ton Production in e+e~ Collisions near the Z Reso-nance, contributed paper No. 342.

34. DO Collaboration, Search for Fourth GenerationNeutral Heavy Leptons, contributed paper No. 433.

35. DELPHI Collaboration, Search for New PhenomenaUsing Single Photon Events in the DELPHI Detec-tor at LEP, contributed paper No. 534.

36. OPAL Collaboration, Search for Heavy ChargedParticles and for Particles with Anomalous Chargein e+e~ Collisions at LEP, contributed paper No.334.

37. F. Eisele, Deep inelastic scattering, these proceed-ings.

38. A. Menzione, Top physics, these proceedings.39. I. Tserruya, Heavy ion physics, these proceedings.

12