Nuclear PhysicsB403 (1993) 3—24 ~ ~North-Holland

Searchfor Z°decaysto two leptonsandachargedparticle—antiparticlepair

DELPHI Collaboration

P. Abreut, W. A~am~,T. Adye~.E. Agasi~’,R. Aleksanarn G.D. A1ekseev~,A. Algeritm, P. AllenaY, S. Almehedw. S.J.Alvsvaagd,U. Ama1di~,

A. Andreazza~,P. Anti1ogus’~,W-D. Ape!°,R.J.Apsimon~,Y. ArnoudamB. Asrnan~,J-E.Augustinr, A. Augustinus~~,P. B~j11~n~,P. Bambader,

F. Baraot, R. Barate~,G. Barbie1lini~’,D.Y. Bardin~,G.J. Barkerah,

A. Baroncelliao,0. Barring~,J.A. Barrio~,W. Bartlaz, M.J. Batesak,M. Battagliatm,M. Baubillierv, K-H. Becks~’,C.J. Beeston~’,M. Begalliai,

P. Beilliere’, Yu. Be1okopytov~l~,P. Beltran, D. Bcnedich,A.C. Benvenuti~,M. Berggrenr,D. Bertrandb,F. Bianchiat, M.S. Bi1enky~,P. Billoirv,

J. Bjarncw, D. Bloch”, S. B1yth~’,V. Boccia~,P.N. Bogolubov’~,T. Bologneseam M. Bonesiniaa,W. Boniventoaa,P.S.L. Boothu,G. ~

H. Borner~C. Bosioao, B. Bostjancica~,S. Bosworth~’,0. Botneraw,E. ~ B. Bouquet’, C. Bourdariosr,T.J.V. Bowcock~’,M. Bozzok,

S. Braibantb,P. Branchiniao,K.D. Brand~°,R.A. Brenner~,H. Briand”,C. Bricmanb,R.C.A. Browns,N. Brummer~~d,J-M. Brunet~,L. Bugge~’~,

T. Burana~,H. Burmeister~,J.A.M.A. Buytacrt~,M. Caccia~,M. Ca1vi~,

A.J. CarnachoRozasap,R. Campion~’,T. Carnporesi~,V. Canale~,F. Caob,F. Car~na~,L. Carroll U C. Casok,M.V. Castillo Gimenez~~,A. Cat~ai~,F.R. CavaIlo~,L. Cerritoaf, V. ~ A. Chana,Ph. Charpentjer~,

L. Chaussardr,J. Chauveauv,P. Checchia’°,GA. Chelkov~,L. Chevalier~’m,P. Chliapnikovao,V. Chorowiczv,J.T.M. Chrinay, P. CollinSat,

J.L. Contreras~,R. Contrik, E. Cortinaay,G. Cosrne’,F. Couchotr,H.B. Crawleya,D. Crenne1la~(,G. Crosettik, M. Crozon~,J. Cuevas~

S. Czellarm, E. Dah1-Jensen~,B. Dalmagner,M. Dam~~,G. Damgaard~,G. Darbok, E. Daubieb, A. Daurn°,P.D. Daunceyat, M. Davenport~,

P. Davidv, J. Davies~,W. Da Silva”. C. Defoix~,P. Delpierre~,N. Demariaat,A. De Angelis~’,H. De Boeckb,W. De Boer°,C. De Clercqb,

M.D.M. De FezLaso”~,N. De Groota~~,C. Dc La Vaissierev,B. De Lotto~’,A. De Minaa, H. Dijkstra~,L. Di Ciaccio~’,J. Dolbeau~,M. DonszeI~~nn~,K. Doroba’~,M. Dra~~s~,J. Drees~’,M. Drjsae,Y. D~four~,F. Dupont~,

D. Edsalla,L-0. Eekaw,P.A.-M. Eeroia~~R. Ehret°,T. Ekelofaw,G. Ekspongas,A. Elliot Peisertal,J-P. Engelh,N. ErshaidaY,D. Fassouliotisae,

0550-3213/93/S06.00 © 1 993—ElsevierSciencePublishersB.V. All rights reserved

RAPID COMMUNICATION

4 DELI’I!I (ol/aborailon / Z0 decat,sea/c/I

M. F~indt~,M. FernandezAlonsoap.A. Ferreray.T.A. Filippasae,

A. Firestonea H. Foethg E. Fokitis ae F. Fontanellik K.A.J. ForbesUJ-L. Foussctz,S. Franconx,B. Franekak,P. Frenkiel~,D.C. Fries0,A.G. Frodesend,R. Fruhwirth~,F. Fulda-Quenzcr’,K. Furnival”’,

H. Furstenau°,J. Fustcr~D. Garnbaat,C. Garciaay,J. Garciaap,C. Gaspar~,U. Gasparini”, Ph. ~avi1l~~~,E.N. Gazisae,J-P.Gerberh,P. GiacomeIli~,

R. Gokieli~,B. Golobar, V.M. Golovatyuk~,J.J.GomezY Cadenas~,G. Gopal~,L. Gorna,M. Gorski~,V. Graccok,A. Gr~nt~,F. Grardb,

E. Grazianiao G. Grosdidierr E. Grossg P. Grosse-Wiesrnanng

B. Grossetetev,S. ~urneny~ka~, J. Guy’~,U. Hacdinger°,F. HahnuJM. Hahn°,S. Haider~’,A. Hakansson’~,A. Hallgren~,K. Hamacher~,

G. HamelDc Monchenaultam,W. Haoad F.J. Harris’~,V. Hedbcrgw,T. ~ J.J.Hernandez”~,P. Herquetb,H. H~rr~,T.L. Hessingu,

I. Hietanentm,C.0. Higginsu,E. HigonaY, H.J. Hjlk~~,S.D. Hodgson~’,

T. Hofmokl’~,S-0. Holrngren~,D. Holthuizen’~,P.F. Honoree,J.E. Hoopera~),M. Houldenu,J. Hrubecaz,K. Huetb,P.O. Huithas,

K. Hultqvistas,P. loannouc,P-S. Iversend,J.N. Jacksonu P. Jalocha~,G. Jar1skog~v,P. Jarry”tm, B. Jean-Marie’,E.K. Johansson~,D. Johnsonu,

M. Jonker~L. Jonsson’~,P. Juilloth, G. Kalkanis’, G. Kalrnus~,F. Kapustav,M. Kar1~~~n~,E. Karvelas’, S. Katsanevase,E.C.Katsoufishie,R. Keranen~J. Kesternanb,B.A. Khomenko’’, N.N. Khovanski0,B. King”,

N.J. Kjaer~,H. Klein~,A. Klovning~’,P. Kluit~1,A. Koch-Mehrin~,J.H. Koehne°,B. Koene~’,P. Kokkinias’, M. Koratzinos~,A.V. Korytov”,

V. ~ C. Kourkournelis’, 0. Kouznetsov”,P.H. Krarner~,C. Kreuter°,J. Krolikowskiu~,I. Kronkvist~,U. Kruener-Marquis~,

W. Krupinski~,K. Kulkaaw, K. Kurvinentm, C. Lacastaay,C. Lambropoulos’,J.W. Lamsaa,L. Lanceri”, V. Lapina~,i-P. Laugicr”tm, R. Lauhakangastm,

G. Lederaz,F. Ledroit~,R. Leitner~,Y. Lemoigne”tm, J. Lernonneb,G. Lenzenu~,V. Lepeltier’, T. Lesiak~,J.M. Levy”, E. Lieb~,D. Likoaz,

J. Lindgrenm,R. Lindnerb~),A. Lipniackau~,I. L~r~ia,B. Loerstadw,M. Lokajicek~,J.G. Loken~’,A. Lopez~Fernandez~,M.A. LopezAguerah~P,M. LOSad D. Loukas’, J.J.Lozanoay,P. Lutz~,L. Lyonsah,G. Maehlumaf,

J. Maillard~,A. Maiot, A. Maltezos’, F. Mandl~,J. Marco”~,M. Margoni~,J-C.Marjn~A. Markou’, T. Maronb~~,S. Martiay, F. Matorras”,

C. Matteuzzi~a,G. Matthiae”~,M. Mazzucatoal,M. Mc Cubbinu,R. Mc Kaya, R. Mc Nulty”, G. Meola’~,C. Meroni~,W.T. Meyer”,

M. Michelottoai ~ Mikulec az L. Mirabito X WA. Mitaroffaz,G.V. Mitselrnakher0,U. Mjoernrnark’~’,T. Moaas,R. Moellera~),K. Moenig~,

M.R. Mongek, P. Morettinik, H. Mueller0, W.J. Murray ,Ik, B. Muryn~,G. Myatt~’,F.L. Navarriae,P. Negri~a,R. Nicolaidouc, B.S. Nielsen~,

B. Nijjhar’~,V. ~ P.E.S.Nilsend, p~Nissas,A. Nornerotski”’.v~Obraztsov~,A.G. Olshevski0.R. Oravatm,A. ~ K. Osterbergm,

RAPID COMMUNICATION

DELPHI Collaboration / L° decaysearch 5

A. Ouraou”m,M. Paganoniaa,R. Painv H. Palka”, Th.D. Papadopoulou~~e,L. Papc~,F. Parodik,A. Passeriao,M. Pegoraro”’, J. Pennanentm,L. Peraltat.

V. ~ H. Pernegger’~,M. Pernicka~~z.A. Perrottac,C. PetridouhiU 4~Petrolinik L. Petrovichaq F. Pierream M. Pimenta

0. Pingoi)~,S. Plaszczynski’.0. Podobrin°,M.E. Pol~,G. Polok”,P. Poropata~~,V. Pozdniakov”, P. Privitera°.A. Pullia”’. D. Radojicic~,

S. Ragazzi””, H. Rahmani~e,P.N. Ratoffs,A.L. Read”, p. ~N.G. Redaelliaa,M. Regler~~z,D. ~ P.B. Rentonall L.K. Resvanisc,

F. Richard’, M. Richardson”.J. Ridky~,G. Rinaudo’°,I Roditi~A. Romeroat,1. Roncagliolok,P. Ronchese”’,C. Ronnqvistm,El. Rosenberg”,

S. Rossi~,E. R0550~P. Roudeau’,T. Rovelli’. W. Ruckstuhl~d,

V. Ruhlrnann~Kleiderarn,A. Ruiz~,K. ~ H. Saarikkom,

Y. Sacquin”tm,G. Sajot’, J. Salt~’~,J. Sanchez~,M. Sannino~”,S. Schael~,H. Schneider°,M.A.E. Schynsb~~,G. Sciollaal, F. Scuri””, A.M. Segar”,

A. Seitz°,R. Sekulin~,M. Sessaau.G. Sette’~,R. Seufert°,R.C. Shellardaj,I. Siccama&~,P. Siegrist”tm, S. Simonettik, F. Simonetto”’, AN. Sisakian’’,

G. Skjevling”, G. Smadjaam~,0. Smirnova’’. G.R. Srnith~k,R. Sosnowski~,D. Souza~SantosaJ,T.S. Spassoff’,E. Spiritiao,S. Squarcia’~,H. Staeck~,

C. Stanescu”°,S. Stapnes”’,G. Stavropoulos’,F. Stichelbautb,A. Stocchi’,J. Straussaz,J. Straver~,R. Strub’’, B. Stugu’’, M. ~M. Szeptyckaba,i. szyrnanskiba,T. Tabarelliaa,0. Tchjkjlevaq

G.E. Theodosiou’.A. Tilquinz. J. Tirnmerrnans~,V.G. Timofeev”,L.G. Tkatchev”, T. Todorov1’. D.Z. Toet”~.0. Tokertm, B. Tome’.

E. Torassa”’, L. Tortora”°,D. Treil1e~,U. Trevisank,W. Trischuk~,G. Tristran~,C. Troncon”, A. Tsjrou~E.N. Tsyganov”, M. Turala”,

M-L. Turluer”tm, T. Tuuvam, l.A. Tyapkinv. M. Tyndel””, S. Tzamarias”,S. Ueberschaerb~),0. ~ v~ ~-~act G. Valenti’, E. VaIlazza~~t,

J.A. Valls Ferrer”~,C. VanderVeldeb.G.W. Van Apeldoorn~’.P. Van Dam~’,M. Van Der Heijden~’,W.K. Van Doninck”, P. Vaz~,G. Vegnia~,

L. Ventura”’, W. Venusa~(,F. Verbeureb,M. Verlato”’, L.S. Vertogradov’’,D. Vilanovaam,P. Vincent~,L. Vitaletm, E. Vlasov”~,A.S. Vodopyanov”,

M. VolImerb~),M. Voutilainentm. V. Vrba”°,H. Wahlenu~,C. Walck”5,F. Waldner”, A. Wehr’~,M. Weierstallb~~,P. Wejlharnrner~,J. Wernerb~),A.M. We~herell~J.H. Wickensb,G.R.WilkiflSon”~’,W.S.C. Williams’”’,

M. Winter”, M. Witek~,G. Wormser’, K. Woschnagg~~v,N. Yamdagni~~s,P. Yepes~,A. Zaitsevaq,A. Zalewska~,P. Zalewskir, D. Zavrtanik”’,E. Zevgolatakos’,0. Zhangb~~.N.I. Zimin~,M. Zilo R. Zuberi’’~’,

R. ZukanovichFunchal~,G. Zumerle”’, J. Zuniga”~

“ Ames Laboratory and Departntentof J’/itsics, 10/ta ,Stcilr’ LJ,i,e/sitY,Ames Il 50011, US.1

b P/tistcsDepartment, (nit. Imistelhng -imtttrerpen, Umtirersiteitsp/ein I, B-26 /0 14//ni/k, Belgium

and huE, LLB- 1 LB Plein/aan 2, B- /050 Brussels, Belgium, and l’actilte desSciences,(jim.dc If/tat Moos,Ar. Maist,iau 19, 8-70001lons, Belgium

RAPID COMMUNICATION

6 DELPhI Collaboration / Z0 decaysea/c/I

Physics Laboratory, Lm,iversit~’ofAtliens,SolonosSt,. 104, GR-10680 Athens,Greeced DepartmentofJ’hysics, University of Bergen,Allégaten55, N-5007 Bergen, Norway

Dipartimento di Fisica, LTnirersith di Bolognaand INFN, Via Irnerio 46, 1-40126 Bologna, Italy

CollegedeFrance, Lab. dePhysiqueCorpusculaire,1N2P3-CNRS,F-75231Paris Cedex05, France

g C’L’Rtvi ChI-1211 Geneva23, Switzerlandh CentredeRechercheNuclcaire, 1N2P3 - CNRS/ULP- BP2O, F-67037Strasboto’gCedex,France

/ Institute ofNuclear Physics, N.C.S.R.De,nokritos, P.O. Box 60228, GR-15310At/tens, GreeceJ FZU, Inst. ofPhysicsof the (‘AS. Il/gb EnergyPhysicsDivision, Na Slovance2, CS-18040

Prague 8, Czechoslovakiak Dipartimnento di Fisica, Universith di Genovaand JNFN, Via Dodecaneso33, 1-16146

Genova,ItalyInstitut desSciencesNuclCaires, 1N2P3-CNRS,UniversitC de Grenoble 1, F-38026

Grenoble, FianceReseam’chInstitute for High Energy Physics,SEFT, Siltavuorenpenger20 C, SF00170

Helsinki, Finland“ Joint Institute for Nuclear Research,Dubna,HeadPost Office, P.O. Box 79, 10/ 000

Moscow, RussianFederationInstitut für ExperimmientelleKernphysik, Universität Karlsruhe, Postfach6980, D-7500

Karlsruhe 1, Germany~ High EnergyPhysics Laboratory, Institute ofNuclear Physics, UI. Kaivior,’ 26 a, PL-30055

Krakow 30, Poland“ Centro Brasileiro dc PesquisasFisicas, ma Xarier Sigaud /50, RJ-22290Rio deJaneiro, Brazil

(~,u,’e,’sitCde Panis-Sud,Lab. de l’Acchlhrateur LinCaim’e, 1N2P3-CNRS,Bat 200, F-91405Orsa,’, France

S Schoolof Physicsand Materials, Universityof Lancaster, LancasterLA] 4YB, UKLIP, 1ST, FCUL - .4v. Elias Game/a, 14 - 10, P-1000Lisboa (‘odex, Portugal

Departmentof Physics,University of Liverpool, P.O. Box /47, Liverpool L69 3BX, UKV LPNHE, 1N2P3-CNRS,UniversitésParis 1’I et VII, Tour 33 (RdC), 4 placeJussieut,F- 75252

Paris Cedex05, FranceDepartmentof Physics,University of Lund, So”l~’egatamt14, S-22363Lund, Sit’eden

UniversitC ClaudeBerna,’d deLyon, IPNL, 1N2P3-CNRS,F-69622 I ‘illeurbanneC’edex, FranceY UniversidadComplutense,Avda. Coinplutenses/n, E-28040 Madrid, Spain

Unit’. dl4ix - Marseille II - CPP, 1N2P3-CNRS,F-13288 Marseille Cedex09, Franceaa Dipartimmiento di Fisica, Università di Milano and INFN, I ‘ia Celoria /6, 1-20/33Milan, Italy

ab Niels Bohr Institute, Blegdamnsrej17, DK-2100 C’openhagen0, Den,miarkac NC, Nuclear Centreof ‘tIFF, Charles University, Am’eal MFF, I’ Holesorickach2, CS-180 00,

Prague 8, Czechoslovakiaad NIKHEF-H, Postbus41882, NL-1009 DB Amsterdam,The Nethem’lands

ae National TechnicalUniversity, Physics Department,Zogm’afou Campus, GR-15773

Athens,Greecead PhysicsDepartment,University of Oslo, Blinder,,, N- /000 Oslo 3, Aoru’ai’ag Dpto. Fisica, Univ. Oviedo, C’/P.JimenezCasas,S/N-33006Ot’iedo, Spain

ah Departmentof Physics,University of Oxford, Keble Road, OxfordOX/ 3R11, UKai Dipartimnento di Fisica, Universith di Padova and INFN, ‘ia Mar:olo 8, 1-35131Padua, Italy

“~Depto. deFisica, Pont/f/cia Univ. Cathlica, C.P. 38071 RJ-22453Rio de Janeiro, Brazilak RutherfordAppletonLaboratory, Chilton, Didcot OXJ I OQX, UK

af Dipartimento di Fisica, Uni,’ersith di RomnaII and INFN, Tor I ‘ergata, 1-00173Rome,Italyam Centred’Etude de Saclay,DSM/D.IPNIA, F-9119] G/fLsum’~YvetteC’ede.v, France

~ Dipa,’timnentodi Fisica-Universith di Salem’no, 1-84100Salerno, ItalyIstituto Superioredi Sanith, 1st. Naz. di Fisica NucI. (INFN,), Via/c ReginaElena 299, 1-00161

Rome, Italy“P C.E.A.F.M., CS/C. - Unit’. (‘antabria, Arda. los C’astros, S/N-39006Santander,Spain

aq Inst. for Ihigh Enem’gy Physics, Serpukov

RAPID COMMUNICATION

DELPHI Collaboration / Z°decayscare/i 7

P.O. Box 35, Protvino (Mosco/vRegion). RussianI”ederation

J. StefanInstitute and Departimientof Physics, University of Ljumb/jana,Jamova39, SI-6 1000

Ljubljana, Slo,’enia~ Fysikum, StockholimiUniversity, Box 6730, 5’- 113 85 Stockholmmi,Siveden

a~Dipartimnento di I”isica Sperimmientale,Universith di Torino and INFN, I ‘ia P. Giuria 1, 1-10125Turin, Italy

Dipartimmiento di Fisica, Università di Triesteand IIVFV, I ia A. I ‘a/eGo 2, 1-34127Trieste, Italy

and Istituto di Fisica, Universith di (‘dine, 1-33100(dine, Italy~W Departimientof RadiationSciences,University qf L~’psalaP.O. Box 535, S-751 21

Uppsala, Sweden“~ IFIC’, I’alencia-CSIC, amid D.F..4.M.N., U. dc Valencia,Arda. Dr. .lloliner 50, E-46100

Burjassot(Valencia,), Spainaz Institut fir Hochenergiephysik,Osterm’. ,4kad. d. U ‘issensch.,Nikolsdorfem’gasse18. .4-1050

Vienna,Austriaba Inst. Nuclear Studiesand Uni,’e,’siti’ of U ‘arsair, (‘1. Ilo:a 69, PL-00681 Warsaw, Poland

bb Fac/ibereichPhysik, Universityof 14’ttppei’tal, Post/a~’li/00 127, D-5600 U’uppertal 1, Germany

Received26 May 1993(Revised10 June 1993)

Acceptedfor publication 15 June 1993

Basedon a sampleequivalentto 365000 hadronicZ0 decays,the searchin DELPHIdatafor pairs of leptonsaccompaniedby a pair of chargedparticles is described.A totalof 11 eventswere found in the electronchannel,9 in the muon channeland 7 in the tauchannel.Resultson leptonpairs with a radiatedphotonare also presented.The datafromall channelsarecompatible with the expectationsfrom standardprocesses.However, oneeventwas found in the tau channelwith an unusually high massof the chargedparticlepair.

1. Introduction

This article describesa study of eventsconsistingof a lepton pair and onepair of chargedparticles,recordedin the DELPHI detectorat LEP. Theleptonpair, f+Fy, may be e+e, ~ or i~r. The otherpair of particles,hereafter

called V, canbe eithera lepton pair or a hadronpair. Eventsof this type canbeproducedthrough second-orderQEDprocesses,whenan off-mass-shellphotonfrom initial or final-statebremsstrahlungmaterializesinto a pair of fermions.Calculationsof ratesanddifferential distributions havebeenmadein ref. [1].Thesesimple final statesare of interest for several reasons.They allow one

to extendthe precisiontestsof QED from the final stateswith a real photonradiatedto the case where a virtual photon is produced.Also, although theyare rather rare, they can be a significant backgroundin searchesfor the Higgsbosons[2j. Finally, an anomalyin the productionrateor distributionof eventscould be a signalof new physicsbeyond the standardmodel. A possiblesmallexcessin the ~ channelwas suggestedin 1990 by the AMY collabo-ration [3j, working in the TRISTAN/KEK accelerator,and later the ALEPH

collaboration [4] reportedan excessof t~rV events.In July 1991 the Markil

RAPID COMMUNICATION

8 DELPhI Col/ahom’ation/ L° deca,’search

[5] collaborationusing data taken at PEP/SLACand the OPAL [61 and theDELPHI [71collaborationsusingdatafrom LEP/CERN all quotedno discrep-ancieswith the standardmodel. More recently,the AMY collaborationreportedthat with the increasein statistics,their dataare not significantly different fromexpectations[8].

The dataanalyzedhere wererecordedby DELPHI during 1990 and 1991 in

a samplecorrespondingto 365000 hadronicZ°decays.The eventsweretakenat centreof massenergiesbetween88.2and 94.2GeV with 74% at the peakofthe Z°crosssection~the integratedluminosity was 15.7pb’.

The organizationof the paperis the following: in sect. 2 a descriptionof the

essentialfeaturesof the DELPHI detectoris given. In sect. 3 the Monte-Carlosimulationis presented.The eventselectionis explainedin detail in sect.4 andthe eventidentification is describedin sect. 5. In sect. 6 theresultsaregiven, andin sect. 7 a checkwith H’y eventsis presented.Finally conclusionsare drawnin

sect. 8.

2. Apparatus

An exhaustivedescription of the DELPHI detector can be found elsewhere[9]. The specific propertiesrelevant to this analysisarenow described.

Chargedparticle tracksare measuredin the 1.2 T magneticfield by threecylindrical trackingchambers:the Inner Detector(ID) coverspolaranglesfrom30°to 150°at radii 12 to 28 cm, theTime ProjectionChamber(TPC), the maintrackingdevice,coverspolar anglesfrom 20°to 1600 and radii between35 and111 cm andthe Outer Detector (OD) coverspolar anglesfrom 43°to 137°atradii between198 and 206 cm.

The Microvertex Detector(VD) consistsof threecylindersmadeof two inde-pendenthalf-shellsinsertedbetweenthe beampipe andthe ID. Eachhalf-shell

containscoaxiallayersof microstrip detectorslocateda radii 6.3 (only for 1991data), 9 and 11 cm. They measurethe transversecoordinatesandcover polaranglesfrom 43°to 137°.The VD wasusedin the analysisafter the eventselec-tion to checkon possiblephotonconversionsoccurringbetweenthe VD and theTPC sensitiveregions.

The energy of the photons and electrons is measured by the High DensityProjection Chamber (HPC) and by the ForwardElectromagneticCalorimeter(FEMC). The HPC has nine layersof lead andgas coveringpolar anglesfrom43°to 137°andradii between208 and260 cm.The FEMC hasleadglassblockscoveringpolar anglesfrom 10°to 36°andfrom 144°to 170°.

Hadronshowerenergiesaremeasuredby combiningthe measurementsfromthe Hadron Calorimeter (the instrumentediron return yoke for the magnet),

coveringpolar anglesfrom 10°to 170°,andthe electromagneticcalorimeter.

RAPID COMMUNICATION

DELPHI (‘ollabom’atiomi / /0 decai’search 9

Muonsareidentifiedby their penetrationthroughtheyoke to theMuon Barrel

(MUB) and Forward (MUF) chambers,which havelayersembeddedin andoutsidethe iron yoke,andcoverpolaranglesfrom 9°to43°,from 52°to 128°andfrom 137°to 17 1°.The calorimetersalso distinguishhadronor electromagneticshowersfrom minimum ionizing particles.

TheSmallAngleTagger(SAT) wasusedto measuretheluminosity. It consistsof an electromagneticcalorimetercoveringpolar anglesfrom 43 to 1 35 mrad.

3. The simulationof known physicsprocesses

At the Z°pole the main processescontributingto thef~fV signalarethe Z°decaysinto leptonsandthei-channel photonexchange,bothwith virtual photonproductionproviding the other pair of chargedparticles.There is a small con-tribution to the signal from multiperipheraland conversion-typediagrams[1].

The generationof the l+8V eventsusesa Monte-Carloprogramthat is de-

rived from the onedescribedin ref. [1], but with the Z°diagramsaddedto thephotonones.The electromagneticcoupling constantis assumedequalto 1/128

for the Z°propagatorand 1 / 137 for thephotonpropagator.Thisgeneratorpro-gramdoesnot include initial radiation, hencethe crosssectionsobtainedweremultiplied by a correctionfactor dependingon the energy in thecentreof mass

(0.74 at the Z°peak).The programwasmodified to accountfor final statescontaininga pion pair.

This wasdone by generatingC +[/L+/1 events,changingthe muonpair in apion pair, andthenweighting eachof them accordingto thetimelike pion formfactorasmeasuredin the ratio of~~rvm_productionto /1+u~ production [10] inlow-energye+e annihilations.This factor,which wasassumedto be zeroabovea V massof 1.6 GeV, takes into account both the resonantmassdependencearoundthe p regionandthe different angulardecaydistributionsfor pion andmuonpairs.Final statescontainingotherhadronpairshaveacontribution one

orderof magnitudesmaller [11] andwere not simulated.A total of 500 events(about 250 pb~)weregeneratedfor eachfinal state,

requestingpair invariant massgreaterthan 50 MeV/c2 andpolar anglegreaterthan 15°(except for the i~tV channel).They weresubmittedto the full de-

tectorsimulation [12] allowing for all decaysand interactions,andthe resultswere processedby the samereconstructionand analysis programsas the rawdatafrom LEP.

In addition to the simulatedC +1_V events,other sets of generatedeventswere also passedthrough the detectorsimulation, reconstructionand analysisprogramsin orderto estimatethebackground.Thedatasamplesusedcomprised20 pb’ of simulated Bhabhaevents (etc -~ e~e(;’)) from the programBABAMC [13], 35 pb’ of simulatedMuon events (eke jr~u(y)) from

RAPID COMMUNICATION

10 DELPHI Collaboration / Z°decaysearch

DYMU3 [14], 50 pb~of simulatedTau events (eke r~t~)’)) fromKORALZ [15] and 25 pb~of simulatedhadronicevents (e~e~ q~j(~’))from JETSET7.3[16].

4. Event selection

Thechargedparticlemultiplicity distinguishesC ~CV eventsfrom mostdilep-tonandhadronicfinal states.Candidateeventsarerequiredto havefourcharged

particlesor, to accepty+z’V eventsin which one of the r decaysinto threechargedpions,six chargedparticles.Backgroundeventscomefrom t~t eventswith 3-prongdecay of oneor both of the t’s, C + C ~‘ eventswherethe photon

converts,and hadronicjets at small anglesto the beam,where the efficiencyfor reconstructingparticlesis low. In order to reducethesebackgrounds,thecharacteristictopology of C’CV eventsis exploited, favoring a quasi 3-bodyconfiguration, andwith a wide range in the invariant mass of the CV system(constrainedin the i~r events)or of the V pair (peakedat very low massesin the contributionfrom realphotonconversions).

In the following subsectionsthe successiveselectioncriteria will bedescribedin detail.

4.1. SELECTION OF CHARGED PARTICLES

Charged particles must come from the interaction region within 6 cm along thebeamdirectionand within 3 cm in the transverseplane (4 cm for polar angleslower than 40°or greaterthan 140°).They shouldhavea polar anglebetween

25°and 155°,andmomentumgreaterthan0.2 GeV/c. Eventsmust havefouror six chargedparticleswith zero total charge.If therewere only four charged

particles,up to two additionalchargedparticlesnotcomingfrom theinteractionregionwereallowed;thesemay be theresult of an associatedphotonconversion.To removephoton-photoninteractionsthe sumof the absolutemomentaof the

chargedparticlesmust beabove 15 GeV (seefig. 1) andtheremustbe at leasttwo jets. Thejet topologywasdeterminedusingthejet clusteralgorithm JADE[16] with default parameterYcut = 0.05.

At this first levelabout 4100 eventswere kept.

4.2. SUPPRESSIONOF r~v BACKGROUND

The backgroundfrom t~i eventscan be much reducedby applying a cuton the lowest invariant massof all possibletriplets of particles:in a perfectlyreconstructedtau decay, this mass should not exceed 1.7 GeV, while for thesignalno suchconstraintis present.For the samplewith four chargedparticlesa

RAPID COMMUNICATION

DELPhI C’ollabo,’atiom, / ~0 decat’seam’c/i Ii

0

DELPHI500

LI) IH-z[0

4QQ

300-

200-

10: ilO 20 310 40 510 60 70 810 910 100

Sum of Momenta (GeV/c)

Fig. 1. Sum of the momentaof the chargedparticles.

minimal massof 2.5 GeV/c2wasimposedon all combinationsof threechargedparticles.For the samplewith six chargedparticles,all possiblecombinationsof threechargedparticleswith total charge+ 1 and invariant masslower than1.7 GeV/c2werecounted.Eventswith oneandonly onesuchcombination(atdecaycandidate)andan invariantmassof the otherthreeparticlesgreaterthan2.5 0eV/c2 wereretained.The acceptedparticlecombinationwasreplacedbyits total momentum,reducingthis caseto four chargedparticles.

Fig. 2 shows that the triplet massspectrumis dominatedup to 1.6 GeV/c2by the x into 3m contribution.The reliability of the Monte Carlo simulationofthe 3m massdistribution wasverified usingselecteddecaysof the Z°into t~r[17]. It canbe seenthat the 2.5 GeV/c2cut removesmostof this background;

the contaminationremainingin the final samplewasestimatedto be 1.6 ±0.5events.This contributioncomesfrom eventsin whichoneof the chargedpionshassuffereda nuclearinteraction in the materialbetweenthe decayvertexandtheTPC, resultingin abadlyreconstructedr mass.Above 1.6 GeV/c2,theexcess

RAPID COMMUNICATION

12 DELPIII Collahoratiomi / /0 decayseai’cli

~000 DELPHI

U

>Solid — DATA

~ 800 -d Dashed — r Background MC

- Dotted —I1V Monte—Carlo*1 00if)H-z[0

600 -

400 -

200-- CL.T

0-0 2 4 6 8 10

MASS (GeV/c2)

Fig. 2. Invariant massof the threenearestchargedparticles, for eventswith only four tracksandtotal chargezero.The solid line representsthe data, the dashedline representsthe Monte-Carlofor the simulation of ~ background,and the dotted line representsthe Monte-Carlo for the

signal for a luminosity correspondingto hundredtimes the luminosity of the datasample.

abovetheexpectedC ~CV signalis mainlydueto the backgroundfrom hadronic

eventsand leptonic eventswith convertedphotons.After the massselection3 16 eventsremained.

4.3. REJECTIONOF CONVERTED PHOTONS

Photonconversionsare recognizedsincethey give two chargedparticletrackswhich havealmostthe samepolar angleandvery small invariant mass. In thetransverseplane the circular trajectoriesof the two tracksare almosttangent.Therefore,for every pair of particles,the transversetrajectorieswereextrapo-

latedto thepointwherethetangentswereparallelandthedistancebetweenthemwassmallest.The pair wasconsidereda convertedphoton if at that point the

RAPID COMMUNICATION

DELI’IlI C’o/laboratiomi / /0 decayscare/i 13

140- DELPHI

U

~ 120(2

0-

DATA

80 - i/il/I//i IIV Monte—Carlo *10

60 -

40 -

20- CUT

0 - r~—~r~-1‘9—n-i r r-r~0 0.~ 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

MASS (GeV/c2)

Fig. 3. Smallestinvariantmass (assumingthe sameazimuthal angle)of all possiblecombinationsof two chargedparticles.Thcsolid line representsthe data,and the shadedregion representstheMonte-Carlo for the signal for a luminosity correspondingto ten timesthe luminosity of the data

sample.

invariant masswasbelow 85 MeV/c’2 and the distancebetweenthe trajectorieswasless than4.5 cm. For particleswith momentumlower than0.6 6eV/c, the

cut wasincreasedto takeinto accountmultiple scatteringeffects.Forthe samplewith six chargedparticles,the threeparticlesmaking the t candidatewere not

used.In fig. 3 the invariant massof thepair is shown,comparingdataandsignalMonte-Carlosimulation (multiplied by 10). No eventsfrom the BhabhaandMuon Monte-Carlosimulatedsamplessurvive this selection.

Thesecuts removed246 eventsandthus 70 candidateeventsremained.

4.4. REMOVAL OF HADRONIC BACKGROUND

ThepreferredtopologiesofP’~1Veventsare 1-3 (i.e. two jetswith theV andonelepton in the samejet) and 1-1-2 (i.e. threejetswith the V in a separatejet),

RAPID COMMUNICATION

14 DL’LPlII Collaboration / ~0 decaysearch

while hadronicbackgrounddueto lost tracksoften have2-2 and 1-3 configu-

rations.Hadronic eventsin 1-3 topologyarehighly suppressedby the previous

triplet masscut. Theeventswith 2-2 topologywere rejected.With this cut 39 ofthe 2-2 type eventswereremovedfrom the 70 candidateevents.This criterionis stablefor variationsof up to +0.03 in the .T’cut parameter.

The 39 eventsremovedhadmoreextra-tracks,that is tracksnot extrapolating

to the interactionregion, and very few identified leptons;indeed, only 6 outof the 31 remainingcandidateshad extra tracks, while 23 out of the 39 2-2eventshadat leastone extratrack, andonly 4 out of the 39 2-2 eventshadanidentifiedlepton (3 candidatemuonswith momentumgreaterthan 5 6eV/c,and1 candidateelectronwith energygreaterthan 5 GeV/c). Theseresultsconfirm

that the rejectedsamplemostly containedq~events.The quasi 3-body configuration of the 1+1—V eventswas usedas a further

selection,to eliminate hadroniceventswith undetectedparticles.After mergingthetwo particlesthat maketheV into a singleparticle,theresultingconfigurationshould be planar,evenfor t+rV events,sincein this casethe chargeddecayproductsfollow closely the direction of the tau. Thereforethe sumof its three

internal anglesshould be approximately360°.Rejectingthe eventsfor whichthereis no combination of two oppositelychargedparticlesfor the V giving asumof anglesgreaterthan 357°,2 out of the 31 candidateswereremoved.Theefficiencyof this cut is high for thesignal: 99%for thee~e V andjt ~jiV eventsand 85%for the t4tV events.

Finally, in hadronicevents,largeamountsof hadronicenergynotassociatedto

chargedparticlescanbe expected,dueto neutralparticlesandchargedparticleslost in the forward region, while for the ~ eventsthis is not the case:99%of the simulatede~eVandp~j~Veventsand95% of the simulatedz~yVeventsdeposit less than 10 6eV in the hadronic calorimeterunassociatedtochargedparticles.Thereforetwo candidateswhichhad depositsof unassociatedhadronicenergygreaterthan 10 GeV wereremoved.

After all the selections,27 candidateeventswereretained,which constituted

the final sample.Applying this analysisto thesimulatedq~backgroundsample,1.8+ 1.0events

survive the cuts.However, as the q~Monte-Carlo programsare not well tested in the low

chargedmultiplicity domain,an estimateofthe hadronicbackgroundwasmadefrom the data.In this estimatethe eventswith non zero total chargewere used,sincethe incompletehadroniceventsdo not necessarilyhavezero total charge.In fact,before thetopologycut, wherethe othermainbackgroundswerealreadyhighly suppressed,39 zerototalchargeeventsand32 nonzerototal chargeeventswerefound with topology2-2. Removingthe chargeconservationcut, two non-zero totalchargeeventssurviving all the othercuts werefound.However,afterinspection,oneof thesetwo eventswasidentified as aC+1V eventwhere there

RAPID COMMUNICATION

DELPHI Collaboration / ~0 decayscald/i 15

wasa wrong chargeidentification in oneof the tracks(0.7 eventsareexpectedfrom the CCV Monte-Carlo).Therefore,accordingto the ratio of the hadronicsamplesin the zero and non zero total chargeevents,the remaininghadronicbackgroundin the candidatesampleis estimatedto be 1 .2 + 1.2 events.

This is in good agreementwith the aboveMonte Carlo estimate; taking theweightedaverageonegets 1.6 ±0.8 events.

5. Identification of events

The 27 candidateeventswereclassifiedas t+TV, e+eV orp~uV, accord-

ing to the multiplicity, the total measuredenergy,andthe lepton identification.To definethe V, eachpair of oppositelychargedparticleswasmergedinto a

singleone.If theresulting3-particleconfigurationhadthe sumof its threeinter-nal angleshigher than357°,the pair waschosenas the V candidate.Whenmorethan one such pairexisted,and if the two particlescomposingthe V werenotidentifiedas leptonsof different flavour, theonewith thelowest masswastaken.

5.1. IDENTIFICATION OF r~rV EVENTS

Events with 6 chargedparticles, or eventswith four chargedparticlesand

sumof their absolutemomentalowerthan 65 GeV/candtotalelectromagneticenergybelow 65 GeV, wereconsideredT+rV events.This cut retains 86% ofthe simulatedr+1V events,while the contaminationfrom simulatede+eVor p4 jrV eventsis 1%. Therewere7 eventsclassifiedas

All eventsbut one containedat leastoneidentified electronor muon, a con-

dition which was not imposedon the selectionand which confirms that thecandidateeventsarenot significantly dominatedby hadronicbackground.Thespreadin mass of the lowest masstriplet of chargedparticle tracksand thefact that at least one particle of this triplet was identified as a lepton in 5out of the 7 r+z~Vevents, also excludesany significant contaminationdueto poorly reconstructedt~r decayinginto 3m. The total backgroundto thischannel is estimated,as describedin the previoussections, to be 1.6 ±0.5eventsfrom t~t backgroundand 1.6+ 0.8 eventsfrom hadronicbackground.

5.2. IDENTIFICATION OF e~eVAND1G1iV EVENTS

The remaining20 eventswere classifiedase+e_V or ~14jrV events.

Out of these20 events, 11 events were identified as e+e—V by demandingthat one of the leptons (not making the V) had Eem/pbetween0.6 and 1.7,

RAPID COMMUNICATION

16

U

V ‘5’....LD0005.,) .-INOC —‘,

- ‘~ . E ~ ~ 0. c~0 0. 00 — 00 — ~‘ ‘ ~ 00 00 ~ (‘I ‘.0 ‘0 ‘0 00 0’. ‘r’. ~ ‘ri 00~ C’. r— ~ — 00 ‘.z — r— r’, i-i r’.i ~n f-i ‘.0 r— 0’. 0 00 — r’ 0 — 00 00

U_~ U,,U ~~ 0°

~ T.. ‘~“~r—~° 0.’0’0~0.C~0~’.~

11~!~0.~0.~0

~ ~

> ~ ~U L’.~ ~ ~

0.00’..

— 0 —‘ C’. — r— ‘.0 r’ 00 r— 0- C’. © r— — ‘.0 ,. 0- — ‘.f ‘c r”~‘.0 r”l 0. — 0- ‘.0>0’.,, ... l’—---~0-r~---

-. 0.00.0. ~

~‘ 0 .. 0..-. > . 0 C

~UU00 0 .VOOCCCCCUUCVUVU000VC.t.~ .“.‘.~ GU00~0~ C. V00CC~~~UU0UVUUC.0~V0

C 000 0

-C‘0r—r’-- ~~t©©0—-o0-~r0—0

— ~ ‘f~ ~‘ ‘f~’.0.0.—0’ —00—h-’fc0r-000.0.~cI’0C’~’f’.r-r’-—,A0.o~C’.CO 0 00 C’. 00 ‘f’ r— ‘./0 r— 00 ‘.— 00 0- ‘.0 00 00 C’. 0- 00 ‘.0 ‘f’ 0- C’— C’— C”- C’) 00’0 rC’

.5 u’.” 0U .~ 0 ~ U

H ..~EOCo.0 ‘.‘.~ C. VVC.~’-U~C.C.VUC.VV00V~’V’.-U~’-U‘.~ ~- oo~~o UU&’.~V~.C.C.VVC.VVC.C.V~..V~UUU00) V.0 S’.-

.2~- 0~•~~U ‘.-‘ U 0

!~ > 0- 0- 0- ‘.0 0- 0- o- ‘a- ‘a- ‘-a- ‘a- ‘a- ‘a- ‘a- o ‘a- ‘a- ‘.o ‘a- ‘a- ~r0 U a 5 PC 00 0. 0- — 0. C’~0. f’. 0- ‘.0 ‘.0 00 00 ‘0 0. 0. ‘0 C— (“I 0. ‘PC t~—0. ‘0 C) ‘a-

000.’a-— ‘5 ‘~ .u 0. 00 0. — — ‘PC C ‘.OC-i 00 0. — 0. 00 C’. — C-i C’. iN 0- C-- —, r— ~n 00 (N “1’CO Uv U CO

0.,,, oo©~r—h ~a-©iN0.00C-0.0-’PC’.”.iN’0-~’PC0’.

0)) ~ (N000-(N’.~ ~N0.0-00r’I’fr’i’f~iC- 0.’00-”1~ ~

‘.,, > U0’. 0) 00000.r—r 0-0-0-~0-0-0-~’PC’0~C--0.©0.’PC©’.Cr’-’PC

E iNC$r”ir~r”i (NINC’ir0(Nc-)NINC-IC’.0.0.C!NC’--C’f—’a-C’..0 L’.~’ ——0.0-CN

~ -~ 0.0.000.0. 0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.000.00 C’.

0

~0 5 C’. — C-i C-i 00 0. ‘0 ,.., . 0. ‘a- ‘.0 (‘ 00 0. 0. 00 C- 0. 00 ‘~ 0.0 U — (C- 0. — C- C’. “1- 0’. ~C- ‘0 0. 0- ‘.00. ~ 0’. — ‘.0 — 0’. C’. C~- 0. 0.0) ~ ‘~ > ‘.0 ‘C’ ‘.0 CC- IN C’) CC- ‘.0 ‘f”. IN IN 00 ~“ 0. ‘.0 0. C’. (C- IN INU ~ > ~ C-i 0’. f-i — /0 — 00 (C- — — — Ifl-) ‘~ ‘_ CC- — 0- — CC- CC- 00 00V.0 0

O ~ — B C”) ‘a- C’. CC- ‘/‘, IN C” IN C’. 0”) ‘/0 f”. 00 ‘0 ‘/0 ‘/0 0. 0’. C’- CC- 0- — C”. .—. ‘.0•— o 0 © C’. (N 0- ‘.0 C’) 00 — 00 C”— ‘f-i CC- ‘a- 0 00 C’) 0. 0. — 0. 0’ C’) C- — 0’ C”.U — U 0. C’. 00 ‘f ‘PC — CC- ‘/0 C’. C’. ‘a- 00 C’ ‘PC ‘.0 C- C’. 0- ‘a- ‘1’ C’. 0. — C’) C-I CC- ‘0 r—

0~ 0.~IN~C-’a-’P’. rC(CINCC-0-’PC’fTh’.0’.0’.0t-00000000000000~ 0

0.C-ic”CC’)IN(NINC-CrNININI—ic’-Ir’ir’CC’)INr-Cr-ir”C(-Ir’Cr-l

CC”)

Lu u~~

RAPID COMMUNICATION

DELPHI Collaboration / /~ decam’scare/i 17

TABLE 2

Observednumbersof eventscomparedto the expectednumbersof events:/ + I — V is the expectedsignal, /‘).f~ representsthe added contribution from Bhabhaand Muon backgroundsimulated

samples.~r is the a-~r backgroundsampleand q0 is the hadronicbackgroundsample.

Channel Data Monte-CarloExpectations

Signal Backgroundl~I’ 1990 1991 Total I/V II:’ q~e~e 0 11 11 9.5±1.1 0 0 0

3 6 9 9.3±0.9 0 0 02 5 7 3.9 ±0.4 0 1.6 0.5 1.6 ±0.8

Total 5 22 27 22.7± 1.4 0 1.6±0.5 1.6 ±0.8

TABLE 3

Distribution of the candidatesaccordingto the V classification,as describedin the text.

l~l’ V = cc V = ~ V = mo V = ~u,m7r V =? Total

e~e 2 5 0 3 1 115 I 1 2 — 93 3 I — — 7

Total 10 9 2 5 1 27

where Eem was the energy measuredin the electromagneticcalorimeterandpthe measuredmomentum.

Of the remaining9 events,7 hadhits in the muonchambersassociatedto oneof the chargedparttclesnot making the V, andwere classifiedas JS1LV events.Of the other two events,onehad both leptonswith Eern/p below 0.1 andEhad/Pbelow 0.1 (Ehad is the energymeasuredin the hadroniccalorimeterassociatedto the particle)and sowasalso classifiedas ~i1iV event.The othereventhadoneleptonwith Ehad/Pbelow 0.1 andE~~/p= 5.9andtheotherlepton with Ehad/PandEem/p below 0.1. It wasassumedto be a ,L/C’V eventwhere the radiatedphotonwascollinearwith oneof the muons.

5.3. CLASSIFICATION OF THE V

If a particle making the V wasenergetic,it was classifiedusing the electro-magneticcalorimeterandmuonchambersinformation. Lower energyparticleswereidentified usingtheenergyperunit track-lengthdepositedin the gasof theTPC (dE/dx).

In all candidateeventsexceptone,at leastoneof theparticlesof theV wassuc-cessfullyidentified (with someambiguitybetweenmuonsandpionsfor low mo-mentumparticles)andthe V wascompatiblewith being a particle—antiparticlepair.

The information from the Microvertex Detectorwas used to confirm that

RAPID COMMUNICATION

15 DELPHI Collaboration / ~0 decaysearch

86 ~ ~:“ ~5’~ ‘96ECM (GeV)

3-

b)TTV DELPHI

if)I-zUs

1-

0 86 ‘ 87 ‘‘‘88’ ‘ ‘ 89 ‘ 909192 ‘ ‘ 93 ‘ ‘ 96ECu (CeV)

Fig. 4. Distribution of the centreof massenergyfor the (a) e’)eV and /i+/1V, and (b) trVcandidateevents,

photonconversionswereproperlyeliminatedby thecutsdescribedin section4.3.Of the 8 V’s with massbelow 0.4 6eV/c2 falling within the acceptanceof theVD, 7 hadassociatedhits in it, whereasmost of the conversionsshouldtakeplaceafterthis detector.

6. Results

The propertiesof the 27 candidatesare given in table 1. The i’esults of this

analysisare summarizedin tables2 and 3. Fig. 4 shows the distribution of thecentreof massenergyof the events,ECM. Figs. 5—7 show the comparisonsofinvariant massesandV energybetweenthe observeddataandthe expectationsfor the signal. The invariant mass of the pair C~C (the leptonsnot makingthe V), Inil, wascalculatedas the recoil massfrom the V, for all eventsexceptthe one with the V massof 17.5 6eV/c2. For this last case,a x2 4-constraint

RAPID COMMUNICATION

DELPHI Collaboration/ ~0 decayseai’c/t 19

9—

8- DELPHI

r a)eeV+~zV

5 ±DATA

d ~ ~_ t~IIV Monte—Carla

3 - /1111/, ‘r-r+qq Bacliground MC

0 24 68 ‘ 10 ‘ 12 14 16 ~ 20

V MASS (GeV/c2)

3,5-

DELPHI~2.5 -

b)TTV2-

2,1.5-

(2H- -

~0.5~ jI L ___

0 -f’~ , I , ‘r’r’fl,1,,,~ , ‘ T’’’P’fl’

0 2 4 6 8 10 12 14 16 18 20V MASS (GeV/c

2)

Fig. 5. Distribution of the V invariantmassfor the (a) c’) eV and ~i’~ ~i V, and (b) nV candidateevents,comparedto Monte-Carlosimulation for the signal (line) and background(shaded).

kinematicalfit of the four particlesmomenta(onebeingthe tau candidate,i.e.

the vector sum of threeparticles) wasperformedto improve the errors on thecalculatedmasses(thevaluesof the massesfor the V or 1+C havenot changedsignificantly from the previousmethods).

Within the statisticsthereis a reasonableagreementbetweendataandMonte-

Carlo predictions.Therearehoweversmalldiscrepancieswith the expectedbe-haviour.On fig. 4, thereis an indication of aspreadin centreof massenergyofthe 1+1_V candidates.There is also a small peakof 3 low V masseventsandan eventwith an unusuallyhigh V mass(17.47 + 0.17 6eV/c2) in the r~tVchannel (fig. 5-b)). The threeeventswith a low V masshad a centreof massenergyout of the Z°peak,andwerecharacterizedby a very energeticV.

The high V massevent is shown in fig. 8. In this eventthe V is composedoftwo identified muonsrecoilingagainsta 25.8+ 0.8 6eV/c2invariantmassT~Tsystem.The massresolutionfor this eventwascalculatedfrom a ,~2 4-constraintkinematicalfit of the four particlesmomenta.The numberof eventswith a V

RAPID COMMUNICATION

20 DELPhI Collaboration / Z0 decayscare/i

“U 6- DELPHI>

~ a)eeV~~V

~ + DATA

U IIV Monte—Carloliii. TT+qq Background MC

~ - ‘ ~3IQ ‘~~‘‘‘~1’’’’1’ 80

Ii MASS (CeV/c2)

2.5 - __________________________________________________________

DELPHI‘-‘.~. 2->

b)orV1.5

1- - - - — -

‘2~0’ 310 ~ 6’070 8’O9~

U MASS (GeV/c2)Fig. 6. Distribution of the invariant mass of the I ‘)/ — pair for the (a) e~eV and

1r~1i V.and (b) na-V candidateevents, comparedto Monte-Carlo simulation for the signal (line) and

background(shaded).

massgreaterthan 10 6eV/c2andC+1 masslower than40 6eV/c2 predicted

by the C~lV Monte Carlo is 0.013 + 0.013.

7. Checks on £ty events

Anothertestof universalityin thethreeleptonicchannelswasdonebystudying

the production of Cl;’ events.This is directly related to the searchfor heavyexcited chargedleptonsdecayinginto two normal leptonsplus a photon, ananalysisalreadypublishedby DELPHI [181. This analysisalso correspondstoa total of 365000 hadronicevents.The eventselectionrequestedtwo chargedparticleswith polar anglesbetween25°and 155°plus a photonwith energyabove 2 GeV and with an isolation anglewith respectto the nearestchargedparticle larger than 30°.Table 4 gives the numberof events selectedfor each

RAPID COMMUNICATION

DELPIII C’ollalCoraiion / /0 dd’d’dii’ scald/i 21

—

~ 4~44- -f-DATA DELPHI11V Monte—Carlo a) eeV+~V

2 — — Ill/ill/li T’r+qq Background MC

V ENERGY (GeV)2.5-

DELPHI2- -- b)TTV

--

0 ~ ‘ ‘ ‘ 20 25’ 31Q ~ ~ ‘41Q’ 45 50

V ENERGY (CeV)

Fig. 7. Distribution of theenergyof the V for the (a)e’)eV and1’~jiV, and (b) na-V candidate

events,comparedto Monte-Carlo simulationfor the signal (line) and background(shaded).

channeltogetherwith the Monte Carlo expectationsfrom standardZ°decays;for this analysisthe selectionon the centreof massenergythat wasused in ref.

11181 wasremoved.Table 5 givesthe result of a modifiedstudy in which the minimum isolation

angle for the photonwasreducedIn 15°,thereforecoveringan angularregionsimilar to that of the/tV analysis.

As canbe seenin the tables,the numberof eventsfound is in agreementwiththe standardMonte Carlo predictions.

8. Summary and conclusion

A searchfor C~CVeventswith the DELPHI detector was performed on asamplecorrespondingto 365 000 hadronicZ°decays.The results indicate noanomalousbehaviourin any of the threechannels.e+eV, /1+/rV andy+y~~

RAPID COMMUNICATION

22 DELPHI Collaboration / Z0 decayscare/i

ID TB TI TX TV ST PA

DELPHI Interactive Analysis 66 35 0 0 0B,.,,: 04,8005 Rim’ 2827i lAS: 22-~,-C99l (66) (TO) 10) (8) 18) (0) (0)

En’ 3370 C3~:CO 0 3 0 0 0 0Pro.. 28-00C-C99C • ~ 24-Jo,-C992 (CC) (8) (0) (CC) (‘I) (0) 10

Fig. 8. Display of the a-’~a-Veventwith highermassfor the V (17.5±0.2GeV/c2). The particlescomposingthe V appearon the left-handside,giving hits in the outerlayersof the muonchambers,classifyingthe particlesasmuons.On the right-handside, a clusterof threechargedparticle trackswith invariant masslower than 1.7 GeV/c2, and a well-separatedtrack with associatedhadronicenergyare seen, identifying the event as a-+a-V. The energyin the centreof mass is 89.5 GeV.the massof the V is 17.47±0.17 GeV/d’2, and the massof the a-~a-— system is 25.8±0.8 GeV/c2(the massof the systemrecoiling from the V). Othercharacteristicscan be found in table 1 (1991

event 28271/3370).

The final sampleshowsa small excessin the low V massregionfor they+1Vchannel,althoughof no strongstatisticalsignificance.An eventwasfound withan unusuallyhigh V massof 17.47±0.17 6eV/c2 composedof two identifiedmuons,recoiling from a r systemwith invariant massof 25.8+ 0.8 6eV/c2.

An independentteston universality of the threelepton channelson l~C~

RAPID COMMUNICATION

DELPHI Collaboration/ Z°decayscare/i 23

TABLE 4

Number of II’;’ eventsfound for each channel,and numberof eventsexpectedfrom standardprocesses(photon isolation angle greaterthan 30°).

Channel Observed Expected

eey 166 169 ± 13123 126 ± 10

a-ny 75 72+6

TABLE 5

Numberof fly eventsfound when the minimum photon isolation angleO~is reducedto 15°,andnumberof eventsexpectedfrom standardprocesses.

15° 15°

Channel Observed Expected Observed Expectedeey 23 24+3 189 193+15

29 18±2 152 144+11a-ny 10 Il + 1.5 85 83+7

events,showsa good agreementbetweendataandexpectations.

We would like to thank R. Kleiss for helpful discussionsand J.Hilgart forproviding an improvedversionof the codefor the1+1—V model.Wearegreatlyindebtedto our technical collaboratorsand to the funding agenciesfor theirsupportin building andoperatingthe DELPHI detector,andto the membersofthe CERN-SLDivision for the excellentperformanceof the LEP collider.

References

[I] F.A. Berends,P. Daverveldt and R. Kleiss, NucI. Phys.B253 (1985) 421 and Comp. Phys.Comm.40 (1986)271., and P.Daverveldt,Monte Carlosimulationof two photon processes,Ph.D. Thesis,Rijksunjversiteit te Leiden, September,1985

[2] DELPHI Collaboration,P. Abreu etal., NucI. Phys. B373 (1992) 313] AMY Collaboration, Y.H. Ho et al., Phys. Lett. B244 (1990) 573[41ALEPH Collaboration, D. Decampet a]., Phys.Lett. B263 (1991) 112[51 MARK II Collaboration,T. Barklow et al., Phys.Rev. Lett. 68 (1992) 1316] OPAL Collaboration,PD. Acton Cl al., Phys. Lett. B 287 (1992) 38917] M.Pimenta, DELPHI Collaboration, Proc. Joint mt. lepton—photon symposium &

Europhysicsconferenceon high energyphysics, Geneva(July 1991) Vol. 1, p.379[8] AMY Collaboration, KEK preprint 92-36, AMY 92-1, May ‘92, Proc. IXth mt. Workshop

on photon—photoncollisions (March 1992) Univ. of California, San Diego, to be published[9] DELPHI Collaboration, P. Aarnio et al., Nuci. Instrum. MethodsA303 (1991) 233

[10] L.M. Barkovet al., NucI. Phys. B256 (1985) 365

1111 L. Bergstrom,R.W. Robinett,preprint USITP-90-05,Dep. Physics,Universityof Stockholm,and PSU/TH/64,Dep. Physics,The PennsylvaniaStateUniversity

[12] DELPHI Collaboration, DELPHI event generatorand detectorsimulation - User guide,DELPHI note 89-76 (1989), unpublished

RAPID COMMUNICATION

24 DL’LPIII Collaboi’aiion / ~0 decat’scare/i

[13] F.A. Berends.W, Hollik and R. Kleiss, NucI. Phvs.B304 (1988) 712[14] J.E. Campagneand R. Zitoun. Z. Phys.C 43 (1989) 469[15] S. Jadachand Z. Was. Comput.Phys. Comm. 36 (1985) 191[16] T. Sjostrand,Comput.Phys.Comm.28 (1983) 229[17] P. Privitera, Study of the decay a- —‘ 3m(ny)z’Twith the DELPHI detectorat LEP, Ph.D.

thesis,Institut für ExperimentelleKernphysik, UniversitDt Karlsruhe, January,1993[18] DELPHI Collaboration,P. Abreu eta].. Z. Phys.C 53 (1992) 41, updatedon DELPHI note

92-82 PHYS 193 (1992). submittedto the International High Energy PhysicsConference,Dallas, August 1992