+ All documents
Home > Documents > Search for charged Higgs bosons at LEP 2

Search for charged Higgs bosons at LEP 2

Date post: 24-Nov-2023
Category:
Upload: independent
View: 1 times
Download: 0 times
Share this document with a friend
16
Transcript

EUROPEAN ORGANIZATION FOR NUCLEAR RESEARCH

CERN�EP������

� June ����

Search for Charged Higgs Bosons

at LEP �

DELPHI Collaboration

Abstract

A search for pair produced charged Higgs bosons has been performed in the highenergy data collected by DELPHI at LEP with

ps � ��� �� and ��� GeV

The analysis uses the ���� c�s�� and c�s�cs �nal states and a combination ofevent shape variables di�jet masses and jet �avour tagging for the separationof a possible signal from the dominant W�W� and QCD backgrounds Thenumber of selected events has been found to be compatible with the expectedbackground The lower excluded value of the H� mass obtained by varying theH� � hadrons decay branching ratio has been found to be �� � GeV�c�

�Submitted to Phys� Lett� B�

ii

P�Abreu��� W�Adam��� T�Adye��� P�Adzic��� Z�Albrecht�� � T�Alderweireld� � G�D�Alekseev�� � R�Alemany� �

T�Allmendinger�� � P�P�Allport��� S�Almehed�� � U�Amaldi � N�Amapane��� S�Amato��� E�G�Anassontzis� �

P�Andersson��� A�Andreazza� S�Andringa�� � P�Antilogus�� � W�D�Apel�� � Y�Arnoud� B��Asman���

J�E�Augustin�� � A�Augustinus � P�Baillon � P�Bambade�� F�Barao��� G�Barbiellini�� � R�Barbier�� �

D�Y�Bardin�� � G�Barker��� A�Baroncelli� � M�Battaglia�� � M�Baubillier�� � K�H�Becks��� M�Begalli� �

A�Behrmann��� P�Beilliere � Yu�Belokopytov��� � K�Belous�� � N�C�Benekos��� A�C�Benvenuti�� C�Berat���

M�Berggren��� D�Bertini�� � D�Bertrand�� M�Besancon� � M�Bigi�� � M�S�Bilenky�� � M�A�Bizouard� � D�Bloch�� �

H�M�Blom��� M�Bonesini�� � W�Bonivento�� � M�Boonekamp� � P�S�L�Booth��� A�W�Borgland� � G�Borisov� �

C�Bosio��� O�Botner�� E�Boudinov�� � B�Bouquet�� C�Bourdarios� � T�J�V�Bowcock��� I�Boyko��� I�Bozovic�� �

M�Bozzo��� P�Branchini� � T�Brenke��� R�A�Brenner�� P�Bruckman� � J�M�Brunet� L�Bugge�� � T�Buran���

T�Burgsmueller�� � B�Buschbeck�� � P�Buschmann�� � S�Cabrera�� M�Caccia��� M�Calvi��� T�Camporesi�

V�Canale��� F�Carena� L�Carroll�� � C�Caso��� M�V�Castillo Gimenez� � A�Cattai� F�R�Cavallo� � V�Chabaud�

M�Chapkin�� � Ph�Charpentier � L�Chaussard��� P�Checchia�� � G�A�Chelkov��� R�Chierici�� � P�Chliapnikov�� �

P�Chochula� � V�Chorowicz��� J�Chudoba�� K�Cieslik� � P�Collins � R�Contri�� � E�Cortina� � G�Cosme��

F�Cossutti� J�H�Cowell��� H�B�Crawley�� D�Crennell�� � S�Crepe��� G�Crosetti��� J�Cuevas Maestro���

S�Czellar�� � M�Davenport � W�Da Silva�� � A�Deghorain� � G�Della Ricca�� � P�Delpierre�� � N�Demaria�

A�De Angelis � W�De Boer��� C�De Clercq�� B�De Lotto��� A�De Min�� � L�De Paula�� � H�Dijkstra�

L�Di Ciaccio��� � J�Dolbeau � K�Doroba��� M�Dracos��� J�Drees��� M�Dris�� � A�Duperrin�� � J�D�Durand�

G�Eigen�� T�Ekelof�� G�Ekspong�� � M�Ellert� � M�Elsing � J�P�Engel�� � B�Erzen��� M�Espirito Santo���

G�Fanourakis�� � D�Fassouliotis�� � J�Fayot��� M�Feindt�� � A�Fenyuk��� P�Ferrari��� A�Ferrer�� E�Ferrer�Ribas� �

F�Ferro��� S�Fichet�� � A�Firestone� � U�Flagmeyer�� � H�Foeth� E�Fokitis�� � F�Fontanelli�� � B�Franek���

A�G�Frodesen�� R�Fruhwirth�� � F�Fulda�Quenzer� � J�Fuster�� A�Galloni�� � D�Gamba��� S�Gamblin� �

M�Gandelman�� � C�Garcia�� C�Gaspar� M�Gaspar�� � U�Gasparini�� � Ph�Gavillet � E�N�Gazis��� D�Gele���

N�Ghodbane�� � I�Gil�� F�Glege��� R�Gokieli��� � B�Golob�� � G�Gomez�Ceballos�� � P�Goncalves�� �

I�Gonzalez Caballero�� � G�Gopal�� � L�Gorn���� � Yu�Gouz��� V�Gracco��� J�Grahl�� E�Graziani� � C�Green���

H�J�Grimm��� P�Gris�� G�Grosdidier� � K�Grzelak�� � M�Gunther� � J�Guy��� F�Hahn� S�Hahn��� S�Haider�

A�Hallgren� � K�Hamacher��� J�Hansen��� F�J�Harris��� V�Hedberg��� S�Heising�� � J�J�Hernandez�� P�Herquet��

H�Herr� T�L�Hessing��� J��M�Heuser��� E�Higon� � S�O�Holmgren��� P�J�Holt��� S�Hoorelbeke� � M�Houlden�� �

J�Hrubec��� K�Huet�� G�J�Hughes��� K�Hultqvist�� � J�N�Jackson��� R�Jacobsson � P�Jalocha� R�Janik��

Ch�Jarlskog�� � G�Jarlskog�� � P�Jarry�� B�Jean�Marie� � E�K�Johansson��� P�Jonsson��� C�Joram� P�Juillot�� �

F�Kapusta��� K�Karafasoulis�� � S�Katsanevas�� � E�C�Katsou�s��� R�Keranen�� � B�P�Kersevan���

B�A�Khomenko��� N�N�Khovanski�� � A�Kiiskinen�� � B�King��� A�Kinvig�� � N�J�Kjaer��� O�Klapp�� � H�Klein�

P�Kluit�� � P�Kokkinias�� � M�Koratzinos� V�Kostioukhine�� � C�Kourkoumelis� � O�Kouznetsov�� M�Krammer���

E�Kriznic�� � P�Krstic��� Z�Krumstein�� � P�Kubinec� � J�Kurowska��� K�Kurvinen��� J�W�Lamsa�� D�W�Lane��

P�Langefeld�� � V�Lapin��� J�P�Laugier�� R�Lauhakangas�� � G�Leder��� F�Ledroit��� V�Lefebure�� L�Leinonen�� �

A�Leisos��� R�Leitner� � G�Lenzen��� V�Lepeltier� � T�Lesiak�� M�Lethuillier� � J�Libby�� � D�Liko�

A�Lipniacka�� � I�Lippi�� � B�Loerstad��� J�G�Loken��� J�H�Lopes��� J�M�Lopez��� R�Lopez�Fernandez�� �

D�Loukas�� � P�Lutz�� L�Lyons��� J�MacNaughton��� J�R�Mahon�� A�Maio��� A�Malek��� T�G�M�Malmgren���

S�Maltezos�� � V�Malychev�� � J�Marco��� R�Marco�� � B�Marechal�� � M�Margoni�� � J�C�Marin� C�Mariotti�

A�Markou��� C�Martinez�Rivero� � F�Martinez�Vidal� � S�Marti i Garcia� J�Masik��� N�Mastroyiannopoulos�� �

F�Matorras��� C�Matteuzzi�� � G�Matthiae�� � F�Mazzucato�� � M�Mazzucato�� � M�Mc Cubbin�� � R�Mc Kay��

R�Mc Nulty��� G�Mc Pherson��� C�Meroni��� W�T�Meyer�� E�Migliore�� � L�Mirabito�� � W�A�Mitaro�� �

U�Mjoernmark��� T�Moa��� M�Moch��� R�Moeller� � K�Moenig � M�R�Monge��� X�Moreau��� P�Morettini�� �

G�Morton��� U�Mueller�� � K�Muenich�� � M�Mulders�� � C�Mulet�Marquis�� � R�Muresan�� � W�J�Murray���

B�Muryn���� � G�Myatt��� T�Myklebust�� � F�Naraghi�� � M�Nassiakou�� � F�L�Navarria�� S�Navas�� K�Nawrocki���

P�Negri��� S�Nemecek��� N�Neufeld� R�Nicolaidou� � B�S�Nielsen� � P�Niezurawski�� � M�Nikolenko����� �

V�Nomokonov��� A�Normand��� A�Nygren��� V�Obraztsov��� A�G�Olshevski�� � A�Onofre��� R�Orava���

G�Orazi��� K�Osterberg��� A�Ouraou�� M�Paganoni�� � S�Paiano� � R�Pain�� � R�Paiva��� J�Palacios�� � H�Palka��

Th�D�Papadopoulou��� � K�Papageorgiou�� � L�Pape� C�Parkes� F�Parodi�� � U�Parzefall�� � A�Passeri��

O�Passon��� M�Pegoraro��� L�Peralta�� � M�Pernicka�� � A�Perrotta�� C�Petridou�� � A�Petrolini�� � H�T�Phillips�� �

F�Pierre�� M�Pimenta�� � E�Piotto��� T�Podobnik�� � M�E�Pol�� G�Polok� � P�Poropat��� V�Pozdniakov�� �

P�Privitera�� � N�Pukhaeva��� A�Pullia�� � D�Radojicic�� � S�Ragazzi�� � H�Rahmani�� � P�N�Rato��� A�L�Read���

P�Rebecchi � N�G�Redaelli�� � M�Regler�� � D�Reid�� � R�Reinhardt�� � P�B�Renton��� L�K�Resvanis� � F�Richard� �

J�Ridky�� � G�Rinaudo�� � O�Rohne��� A�Romero��� P�Ronchese�� � E�I�Rosenberg� � P�Rosinsky� � P�Roudeau� �

T�Rovelli� � Ch�Royon�� V�Ruhlmann�Kleider� � A�Ruiz��� H�Saarikko�� � Y�Sacquin� � A�Sadovsky�� � G�Sajot���

J�Salt�� D�Sampsonidis�� � M�Sannino�� � H�Schneider�� � Ph�Schwemling�� � B�Schwering�� � U�Schwickerath�� �

M�A�E�Schyns��� F�Scuri�� � P�Seager��� Y�Sedykh��� A�M�Segar��� R�Sekulin�� � R�C�Shellard� � A�Sheridan�� �

M�Siebel�� � L�Simard� � F�Simonetto�� � A�N�Sisakian�� � G�Smadja�� � O�Smirnova�� � G�R�Smith��� A�Sokolov�� �

iii

A�Sopczak��� R�Sosnowski�� � T�Spassov��� E�Spiriti� � P�Sponholz�� � S�Squarcia�� � C�Stanescu� � S�Stanic�� �

K�Stevenson�� � A�Stocchi� � J�Strauss��� R�Strub�� � B�Stugu�� M�Szczekowski�� � M�Szeptycka�� � T�Tabarelli�� �

F�Tegenfeldt�� F�Terranova��� J�Thomas��� J�Timmermans��� N�Tinti�� L�G�Tkatchev��� S�Todorova���

A�Tomaradze�� B�Tome��� A�Tonazzo� L�Tortora�� G�Transtromer��� D�Treille � G�Tristram�

M�Trochimczuk�� � C�Troncon��� A�Tsirou� M�L�Turluer� � I�A�Tyapkin��� S�Tzamarias�� � O�Ullaland �

V�Uvarov��� G�Valenti� � E�Vallazza�� � G�W�Van Apeldoorn�� � P�Van Dam��� W�K�Van Doninck� � J�Van Eldik�� �

A�Van Lysebetten� � N�Van Remortel� � I�Van Vulpen��� N�Vassilopoulos�� � G�Vegni��� L�Ventura��� W�Venus��� �

F�Verbeure�� M�Verlato��� L�S�Vertogradov��� V�Verzi��� D�Vilanova� � L�Vitale��� E�Vlasov���

A�S�Vodopyanov�� � C�Vollmer��� G�Voulgaris� � V�Vrba��� H�Wahlen��� C�Walck��� C�Weiser��� D�Wicke�� �

J�H�Wickens�� G�R�Wilkinson � M�Winter�� � M�Witek�� G�Wolf� J�Yi�� O�Yushchenko�� � A�Zaitsev�� �

A�Zalewska� � P�Zalewski�� � D�Zavrtanik�� � E�Zevgolatakos�� � N�I�Zimin����� � G�C�Zucchelli�� � G�Zumerle��

�Department of Physics and Astronomy� Iowa State University� Ames IA ���������� USA�Physics Department� Univ Instelling Antwerpen� Universiteitsplein �� BE���� Wilrijk� Belgiumand IIHE� ULB�VUB� Pleinlaan �� BE����� Brussels� Belgiumand Facult�e des Sciences� Univ de l�Etat Mons� Av Maistriau � � BE����� Mons� Belgium�Physics Laboratory� University of Athens� Solonos Str ���� GR����� Athens� Greece�Department of Physics� University of Bergen� All�egaten ��� NO����� Bergen� Norway�Dipartimento di Fisica� Universit�a di Bologna and INFN� Via Irnerio �� IT����� Bologna� Italy�Centro Brasileiro de Pesquisas F��sicas� rua Xavier Sigaud ���� BR���� � Rio de Janeiro� Braziland Depto de F��sica� Pont Univ Cat�olica� CP ����� BR������ Rio de Janeiro� Braziland Inst de F��sica� Univ Estadual do Rio de Janeiro� rua S�ao Francisco Xavier ���� Rio de Janeiro� Brazil�Comenius University� Faculty of Mathematics and Physics� Mlynska Dolina� SK������ Bratislava� SlovakiaColl�ege de France� Lab de Physique Corpusculaire� IN�P��CNRS� FR������ Paris Cedex ��� FranceCERN� CH����� Geneva ��� Switzerland��Institut de Recherches Subatomiques� IN�P� � CNRS�ULP � BP��� FR����� Strasbourg Cedex� France��Institute of Nuclear Physics� NCSR Demokritos� PO Box ����� GR������ Athens� Greece��FZU� Inst of Phys of the CAS High Energy Physics Division� Na Slovance �� CZ���� ��� Praha �� Czech Republic��Dipartimento di Fisica� Universit�a di Genova and INFN� Via Dodecaneso ��� IT���� Genova� Italy��Institut des Sciences Nucl�eaires� IN�P��CNRS� Universit�e de Grenoble �� FR����� Grenoble Cedex� France��Helsinki Institute of Physics� HIP� PO Box � FI������ Helsinki� Finland��Joint Institute for Nuclear Research� Dubna� Head Post O�ce� PO Box � � RU���� ��� Moscow� Russian Federation��Institut f�ur Experimentelle Kernphysik� Universit�at Karlsruhe� Postfach ��� DE����� Karlsruhe� Germany�Institute of Nuclear Physics and University of Mining and Metalurgy� Ul Kawiory �a� PL������ Krakow� Poland�Universit�e de Paris�Sud� Lab de l�Acc�el�erateur Lin�eaire� IN�P��CNRS� B�at ���� FR� ���� Orsay Cedex� France��School of Physics and Chemistry� University of Lancaster� Lancaster LA� �YB� UK��LIP� IST� FCUL � Av Elias Garcia� ����o� PT����� Lisboa Codex� Portugal��Department of Physics� University of Liverpool� PO Box ���� Liverpool L �BX� UK��LPNHE� IN�P��CNRS� Univ Paris VI et VII� Tour �� �RdC�� � place Jussieu� FR������ Paris Cedex ��� France��Department of Physics� University of Lund� S�olvegatan ��� SE���� � Lund� Sweden��Universit�e Claude Bernard de Lyon� IPNL� IN�P��CNRS� FR� �� Villeurbanne Cedex� France��Univ d�Aix � Marseille II � CPP� IN�P��CNRS� FR������ Marseille Cedex � � France��Dipartimento di Fisica� Universit�a di Milano and INFN� Via Celoria �� IT������ Milan� Italy�Niels Bohr Institute� Blegdamsvej ��� DK����� Copenhagen �� Denmark�NC� Nuclear Centre of MFF� Charles University� Areal MFF� V Holesovickach �� CZ���� ��� Praha �� Czech Republic��NIKHEF� Postbus ������ NL���� DB Amsterdam� The Netherlands��National Technical University� Physics Department� Zografou Campus� GR������ Athens� Greece��Physics Department� University of Oslo� Blindern� NO����� Oslo �� Norway��Dpto Fisica� Univ Oviedo� Avda Calvo Sotelo s�n� ES������ Oviedo� Spain��Department of Physics� University of Oxford� Keble Road� Oxford OX� �RH� UK��Dipartimento di Fisica� Universit�a di Padova and INFN� Via Marzolo �� IT������ Padua� Italy��Rutherford Appleton Laboratory� Chilton� Didcot OX�� OQX� UK��Dipartimento di Fisica� Universit�a di Roma II and INFN� Tor Vergata� IT������ Rome� Italy�Dipartimento di Fisica� Universit�a di Roma III and INFN� Via della Vasca Navale ��� IT����� Rome� Italy�DAPNIA�Service de Physique des Particules� CEA�Saclay� FR� �� � Gif�sur�Yvette Cedex� France��Instituto de Fisica de Cantabria �CSIC�UC�� Avda los Castros s�n� ES�� �� Santander� Spain��Dipartimento di Fisica� Universit�a degli Studi di Roma La Sapienza� Piazzale Aldo Moro �� IT������ Rome� Italy��Inst for High Energy Physics� Serpukov PO Box ��� Protvino� �Moscow Region�� Russian Federation��J Stefan Institute� Jamova � � SI����� Ljubljana� Slovenia and Laboratory for Astroparticle Physics�Nova Gorica Polytechnic� Kostanjeviska �a� SI����� Nova Gorica� Slovenia�and Department of Physics� University of Ljubljana� SI����� Ljubljana� Slovenia

��Fysikum� Stockholm University� Box ���� SE���� �� Stockholm� Sweden��Dipartimento di Fisica Sperimentale� Universit�a di Torino and INFN� Via P Giuria �� IT������ Turin� Italy��Dipartimento di Fisica� Universit�a di Trieste and INFN� Via A Valerio �� IT������ Trieste� Italyand Istituto di Fisica� Universit�a di Udine� IT������ Udine� Italy

��Univ Federal do Rio de Janeiro� CP ���� Cidade Univ� Ilha do Fund�ao BR��� ��� �� Rio de Janeiro� Brazil�Department of Radiation Sciences� University of Uppsala� PO Box ���� SE���� �� Uppsala� Sweden�IFIC� Valencia�CSIC� and DFAMN� U de Valencia� Avda Dr Moliner ��� ES����� Burjassot �Valencia�� Spain��Institut f�ur Hochenergiephysik� �Osterr Akad d Wissensch� Nikolsdorfergasse ��� AT����� Vienna� Austria��Inst Nuclear Studies and University of Warsaw� Ul Hoza � PL����� Warsaw� Poland��Fachbereich Physik� University of Wuppertal� Postfach ��� ���� DE���� � Wuppertal� Germany��On leave of absence from IHEP Serpukhov��Now at University of Florida

� Introduction

The existence of a charged Higgs boson doublet is predicted by several extensions of theStandard Model and in particular by Supersymmetry At LEP charged Higgs bosonscan be produced in pairs and decay predominantly in either � �� or c�s �nal states � Asearch for pair�produced charged Higgs bosons has been performed based on the datacollected by DELPHI during the LEP runs at centre�of�mass energies

ps of ��� GeV

�� GeV and ��� GeV The results reported in this paper supersede those obtained inan earlier analysis of the DELPHI data limited to the ��� GeV and �� GeV runs ��� Similar searches have been performed by the other LEP experiments �� As the highercentre�of�mass energy gives sensitivity to Higgs masses closer to that of the W bosonthe background from e�e� � W�W� becomes more important The cross�section forW� production is two orders of magnitude higher than that predicted for H� of equalmass therefore searches are signi�cantly less sensitive to Higgs mass values close to theW� mass peak In order to retain sensitivity to the possible signal from charged Higgsboson decays e�ective rejection of this background is very important A new techniquehas been developed to improve the discrimination against the hadronic W decays in thesearch for H� candidates In section the event reconstruction and the de�nition of thediscriminating variables are discussed separately for the hadronic semileptonic and fullyleptonic �nal states with particular emphasis on the

ps � ��� GeV data Section �

describes the results from the combined DELPHI data at LEP

� Data Analysis

The analysis has been performed on the data collected by the DELPHI detector atLEP at centre�of�mass energies of ��� GeV �� GeV and ��� GeV The DELPHI de�tector and its performance have already been described in detail elsewhere ���� Thecorresponding integrated luminosities are � �� pb�� �� �� pb�� and �� � pb�� respec�tively

Charged particle tracks have been required to satisfy the following quality crite�ria Only particles with momentum larger than ��� MeV�c relative momentum error� p�p � � and track length larger than �� cm have been used In addition the impactparameter to the event primary vertex had to be smaller than � � cm in the projection onthe plane normal to the beam axis and smaller than �� � cm along the beam axis Kaonshave been identi�ed by the combined response of the RICH detectors and the speci�c ion�isation measured in the TPC ���� Neutral particles have been required to have energydeposit larger than �� MeV in the electromagnetic calorimeters or larger than ��� MeVin the hadron calorimeter

H�H� signal samples were produced with the PYTHIA generator ��� at four di�erentH� masses� �� GeV�c� �� GeV�c� �� GeV�c� and �� GeV�c� The q�q� QCD back�ground sample was also produced using PYTHIA and the four fermion �nal states in�cluding W�W� and Z�Z� were produced with the EXCALIBUR generator ��� The mainbackground contribution from four fermion processes is due to W�W� pairs which arereferred to as W�W� in the rest of the paper The VDM and QCD components of the twophoton interactions leading to hadronic �nal states were generated using TWOGAM ��� The generators of Berends Daverveldt and Kleiss ��� were used for the QPM componentand for leptonic �nal states In addition Z�Z� We�� and Ze�e� samples generated with

�Throughout the paper the charge�conjugate states are implicitly included

PYTHIA ������� and ������� events generated with the KORALZ ���� generator havebeen used for the estimate of the background in the fully leptonic channel

��� The hadronic channel

In the fully hadronic decay channel each charged Higgs is expected to decay into a c�spair producing a four jet �nal state

The hadronic four jet event selection followed in this analysis is the same as for theDELPHI neutral Higgs analysis ���� In addition events with two or more jets tagged ascontaining b quarks by the jet �avour tagging algorithm described below have also beenremoved In order to reduce the amount of QCD background and of wrongly reconstructedW�W� decays only events with four reconstructed jets using the Luclus algorithm ���with djoin � � � GeV�c� have been accepted

Energy�momentum conservation has been imposed by performing a ��C �t on theseevents and the di�erence between the two di�jet masses for each jet pairing has beencomputed A ��C �t assuming equal boson masses has been applied in order to improvethe resolution on the di�jet mass Mjj The di�jet combination giving the smallest �� hasbeen selected for the mass reconstruction Events for which the �� per degree of freedomof this combination exceeded � or the di�erence of the masses computed after the ��C �texceeded � GeV�c� have been rejected

Only events with �� GeV�c� � Mjj � �� GeV�c� have been considered for this analysis The two main sources of background in this channel are the q�qgg QCD background

and fully hadronic decays of W�W� pairs The charged Higgs boson is expected to cou�ple predominantly to c�s in its hadronic decay mode Therefore the QCD and W�W�

backgrounds can be partially suppressed with regard to the signal by selecting �nal statesconsistent with being c�s�cs A �avour tagging algorithm has been developed for the studyof multiparton �nal states ��� This tagging is based on the response of nine discrimi�nating variables� three of them are related to the identi�ed lepton and hadron contentof the jet two depend on kinematical variables and four on the reconstructed secondarydecay structure The �nite c lifetime is exploited to distinguish between c and light quarkjets while the c mass and decay multiplicity smaller than for beauty quarks was usedto discriminate against b jets Further s and c jets can be distinguished from u and djets by the presence of an identi�ed energetic kaon A likelihood variable was computedfrom the response of the individual jet variables as discussed below for the anti�WW andanti�QCD functions The responses for the individual jets were further combined into aglobal c�s�cs event probability

QCD background events di�er also kinematically from pair�produced bosons ���� Inorder to separate the signal from the QCD background the following variables have beenused together with the c�s�cs probability from the jet �avour tagging algorithm� the prod�uct of the minimum jet energy and the minimum di�jet angle min�Ejet� � min�jets� thedi�jet pair mass di�erence after ��C �t and the polar angle of the thrust axis Thesevariables have been combined to form an event anti�QCD likelihood function separatingQCD events from pair�produced bosons The likelihood function has been de�ned as fol�lows For each of the N discriminating variables the fractions FHH

i �xi� and FQCDi �xi� of

respectively H�H� and QCD events corresponding to a given value xi of the ith variable

have been extracted from a sample of simulated QCD and H�H� events with equal popu�lations The anti�QCD likelihood has been computed as the normalised product of theseindividual fractions

Qi���N FHH

i �xi���Qi���N FHH

i �xi� �Qi���N FQCD

i �xi�� The responseof this likelihood discriminates H�H� events from the QCD background but not from

W�W� events due to the similarity in the kinematics for these two processes Thereforeother signatures have been used to distinguish possible signal events from W�W� pairsand other four fermion background

Correctly reconstructed events from on�shell W� � q�q decays result in di�jet massesconsistent with the W mass and are therefore above the mass region of interest for thepresent H� search However wrong di�jet pairing decays of o��shell W bosons and eventswith incorrect jet reconstruction may result in an apparent boson mass below �� GeV�c� A simulation study has shown that the wrong di�jet pairing is the dominant source of thesebackgrounds Therefore all three possible jet pairings for a given event have been testedfor the hypothesis to be either a correctly paired W�W� or H�H� event Likelihoods havebeen computed for each pairing using the �� from the ��C �t the di�erence between thereconstructed di�jet mass and the nominal W mass and the boson production polar anglei e the polar angle of the di�jet momentum vector A jet�pairing W tag variable has beende�ned as LW

max��LHmax�LW

max� where LH�Wmax represents the largest of the three likelihoods

for the H and W hypotheses respectively This tag variable is peaked at zero in both H�H�

and QCD events while for the majority of the W�W� events it is uniformly distributedbetween � and �

Table �� Number of selected events and signal e�ciency in the hadronic �nal state at

di�erent stages of the event selection procedure forps � ��� GeV�

Selection Data Total W�W� QCD E�ciency H�H�

Bkg Bkg Bkg �MH��� GeV�c��

Event preselection ��� ��� � � � � ��� � ��� � � �� � � ��� Jets ��� �� � � � �� �� � � � � � ���� Cut �� � � � � � ��� � �� � � �� � � ���M Cut ��� � � � � � �� � � � � � � � ���� � MJJ � �� �� � � � � � �� �� � � � � � ��Probability Cuts �� �� � � � � � � � � � � � ��

Finally normalised fractions of H�H� signal events as a function of the polar angleof the event thrust axis the di�jet mass di�erence and the jet pairing W tag variablehave been combined with the probability from the jet �avour tagging algorithm into anevent anti�WW likelihood using the same procedure as for the anti�QCD likelihood TheZ�Z� production cross�section is small and the Z� mass is high enough to make the Z�Z�

background insigni�cant compared to QCD and W�W� backgrounds at masses below�� GeV�c�

Using the simulation the cuts on the two likelihood variables have been optimisedseparately for each centre�of�mass energy following a procedure discussed in section � �seeFigure �� The number of real data events expected backgrounds and the signal e�ciencyafter each of the set of cuts adopted in the analysis are given in Table � for the highestenergy sample

��� The semileptonic channel

In this channel one of the charged Higgs bosons decays into a c�s quark pair while theother decays into � �� Such an event is characterised by two hadronic jets a � candidateand missing energy carried by the neutrinos The dominating background processes areq�q� and q�qg events and semileptonic decays of W�W�

Charged particles with momentum greater than ��� MeV�c were used in this channeland at least eight were required for the preselected sample of events Further the energyin charged particles had to exceed ���

ps and the total detected energy had to be

greater than ���ps After a clustering into two jets using the Durham algorithm ����

the acollinearity had to be greater than ��� Events were also required to have no neutralparticles with energy above �� GeV and the energy detected in a cone of �� ����� halfaperture around the beam axis had to be less than � �

ps �� ��

ps� Further the angle

between the total momentum of the detected particles in the event and the beam axis hadto be greater than �� After clustering into three jets the jet with the smallest chargedparticle multiplicity was treated as the � candidate In case more than one jet had thesame number of charged particles the least energetic jet was chosen

� jet candidates were required to have less than eight particles of which less than fourwere charged At least one of the tracks in this jet was required to have track elementsin the Vertex Detector and the TPC and either energy deposit in the e m calorimeter orhits in the Outer Detector the Forward Chambers or the Muon Chambers In additionthe visible energy of the � candidate had to be greater than ��

ps

The mass of the decaying bosons Mjj was reconstructed using a constrained �t re�quiring energy and momentum conservation and equal mass of the decaying bosons Thethree components of the momentum vector of the neutrino and the magnitude of the� momentum have been treated as free parameters reducing the number of degrees offreedom in the �t from � to � The � lepton direction was approximated by that of thereconstructed jet Only events with mass in the range �� GeV�c� � Mjj � �� GeV�c�

have been further considered in the analysis Likelihood functions for a given event to be a QCD or a W�W� background event have

then been de�ned similarly to the case of the hadronic channel described above The polarangle of the total momentum the logarithm of the clustering distance de�ned as the ycutvalue in the Durham algorithm for which the number of jets changes from two to threeand the maximum angle between two particles from the � candidate have been used asdiscriminating variables to de�ne the event anti�QCD likelihood For the event anti�WWlikelihood the variables used were the reconstructed polar angle of the negatively chargedboson �where the charge has been taken to be that of the leading charged particle fromthe � decay� the visible � energy the angle between the two hadronic jets and the c�sprobability of the hadronic di�jet

The cuts on the two likelihood variables have been optimised separately for the threeps energies as discussed in section � �see Figure � The number of real data events

expected backgrounds and the signal e�ciency after each of the set of cuts adopted in theanalysis are given in Table for the highest energy sample

Table � Number of selected events and signal e�ciency in the semileptonic �nal state

at di�erent stages of the event selection procedure forps � ��� GeV�

Selection Data Total W�W� QCD E�ciency H�H�

Bkg Bkg Bkg �MH��� GeV�c��

Event Preselection ��� �� � � � �� � �� � � �� � � �� selection ��� �� � � � �� � �� � � � � � ��� � Mjj � �� �� � � � � � � � �� � � �� � � �Probability Cuts � � � � � � � � � � �� � � �

��� The leptonic channel

In this channel both charged Higgs bosons decay into a � �� pair These events arecharacterised by two acollinear slim jets and missing energy carried away by the unde�tected neutrinos The backgrounds are due to leptonic decays of W�W� pairs and otherfour fermion processes such as e�e� � W�e��� two fermion processes �f �f � such ase�e� � q�q��� and e�e� � ������� and two photon interactions ����

Events containing to � charged particles of more than ��� MeV�c with a totaldetected energy in the event not exceeding ���

ps and a measured energy in charged

particles greater than ���ps have been selected in the fully leptonic channel In order

to reject cosmic ray events at least one of the charged particles has been required to havea distance in space within � cm from the position of the LEP colliding beam envelope The energy detected within a cone of ��� half aperture around the beam axis had tobe lower than ���

ps The component of the total energy transverse to the beam axis

had to be larger than � ��ps After clustering the event into two jets by the Durham

algorithm the maximum angle between two particles within a jet had to be smaller than��� the angle between the beam axis and either jet and the angle between the two jetshad to be larger than ��

In order to reject events ful�lling the above preselection cuts where the jets are back�to�back and radiative production of Z� events with a photon along the beam pipe thefollowing angular cuts have been applied The angle between the projections of the jetson the plane perpendicular to the beam axis as well as the angle between the two jetsmust not exceed ����

The total energy of each of the two jets E� and E� had to be greater than ��ps

In addition the energy of the least energetic jet E� was also required to be smaller than���

ps �see Figure ��

Table �� Number of selected events and signal e�ciency in the leptonic channel at di�erent

stages of the event selection forps � ��� GeV�

Selection Data Total W�W� f �f �� E�ciency H�H�

Bkg Bkg Bkg Bkg �MH��� GeV�c��

Event Preselection �� �� � � � � � � ��� � � � �� � � �Angular Cuts � � � � � � � � � � � � �� � � �Energy Cuts �� �� � � � � �� � � � � � � � � � �

Due to the presence of the missing neutrinos in the decay of each of the two bosons itis not possible to reconstruct the boson mass in the leptonic channel on an event�by�eventbasis and the signal corresponds to an inclusive excess of events compared to the expectedbackgrounds The numbers of real data events those from expected backgrounds and thesignal e�ciencies after each set of cuts adopted in the analysis are given in Table �

� Results

��� Optimisation of the selection

The analysis applied atps � ��� GeV has been repeated for the data collected atp

s � ��� GeV and �� GeV After the event preselection procedure described above the

separation of the candidate signal events from the backgrounds in the semi�leptonic andfully hadronic �nal states has been obtained by cuts on the two likelihoods characterisingthe probability for a given event to be due to either a W�W� decay or a QCD process Thevalues of these cuts have been optimized by minimizing on a sample of simulated eventsthe value of the H�H� production cross�section that could be excluded with statisticsequivalent to those of the real data This minimization has been carried out simultaneouslyfor the two probabilities The correlations of the two probabilities for the backgroundW�W� and QCD samples the H�H� signal events and the real data are shown in Figure �and Figure for the hadronic and semileptonic channels respectively

Due to the di�erences in the integrated luminosity H�H� production cross�section andbackgrounds the �nal cuts on the likelihood variables for the hadronic and semileptonicchannels were optimised independently at the di�erent

ps energies

These criteria have been applied to simulated signal H�H� decays The mass reso�lutions have been estimated to be � � GeV�c� for the hadronic and � � GeV�c� for thesemileptonic channel The number of real data and background events and the estimatede�ciencies for these selections for di�erent H� masses are summarised in Table � for thethree �nal states

Table �� Number of events expected background and signal e�ciencies for di�erent

charged Higgs masses in the hadronic semileptonic and leptonic channels�

ps Channel Data Expected E� MH � E� MH � E� MH � E� MH �

Bkg �� GeV�c� �� GeV�c� �� GeV�c� �� GeV�c�

��� c�s�cs � � � � � � � �� � �� � � � ��� c�s�cs � � � � � � � �� � �� � � � ���� c�s�cs �� �� � � � � � � � � � � ����� c�s� �� � � � � � � �� � �� � �� � ���� c�s� �� � � � � �� � �� � �� � ����� c�s� �� � � � � � � � �� � �� � �� � ���� � ���� � � � � � � � �� � � � �� � ���� � ���� � � � � � � � � � � �� � ����� � ���� �� �� � � � � � � � �� � � � ��

��� Determination of the mass limit

No excess of events compared to the expected backgrounds has been observed in anyof the three di�erent �nal states investigated A lower limit for a charged Higgs bosonmass has been derived at the ��� con�dence level as a function of the hadronic Higgsdecay branching ratio BR�H � hadrons� The con�dence in the signal hypothesis CLshas been calculated using the likelihood ratio technique ���� That is �rst we �nd theratio of the likelihood of the observed candidates assuming signal plus background tothat found using the background only hypothesis The con�dence levels CLsb and CLb

are computed as the fractions of Gedanken experiments with respectively signal plusbackground or background only which gave likelihood ratio values smaller than thoseobserved for the data Finally we make the conservative step of de�ning CLs as the ratioof these probabilities� CLs �

CLsb

CLb

The required Gedanken experiments have been made by Monte Carlo simulation usingthe expected background and signal rates as well as the background and signal probabil�

ity density functions of one discriminating variable in each channel In the hadronic andsemi�leptonic channels the discriminating variable is the reconstructed mass while in theleptonic channel the energy of the most energetic jet has been used The distributions forthe discriminating variable of signal events obtained by the simulation at di�erent H�

mass values for eachps have been interpolated for intermediate mass values The sig�

nal e�ciencies have been �tted with polynomial functions to obtain the expected signalrate at any given mass Uncertainties in the expected background and in the signal e��ciency have been accounted for These uncertainties are due both to the �nite simulationstatistics available and to possible di�erences in the response of the selection variablesin data and simulation The background events in simulation have been reweighted suchthat each variable agreed in shape with that of the real data and the background hasbeen recomputed The overall background uncertainty has been obtained by summing inquadrature the di�erences in the background estimates after changing each of the vari�ables and the statistical error A Gaussian smearing of the central values of the numberof expected background events by their estimated uncertainties has been introduced inthe limit derivation program The uncertainties on the signal e�ciencies are dominatedby their statistical errors and have also been accounted for The results are summarisedin Figure � Independent of the hadronic decay branching ratio a lower H� mass limit ofMH� � �� � GeV�c� can be set at the ��� con�dence level For fully leptonic decays ofthe charged Higgs boson the limit becomes MH� � �� � GeV�c�

� Conclusion

A search for pair�produced charged Higgs bosons has been performed using the fullstatistics collected by DELPHI at LEP at

ps of ��� GeV �� GeV and ��� GeV Candi�

date events have been searched for in the � ���� c�s� �� and c�s�cs �nal states No signi�cantexcess of candidates has been observed and a lower limit on the charged Higgs mass of�� � GeV�c� has been set at the ��� con�dence level

Acknowledgements

We are greatly indebted to our technical collaborators to the members of the CERN�SL Division for the excellent performance of the LEP collider and to the funding agenciesfor their support in building and operating the DELPHI detector We acknowledge in particular the support ofAustrian Federal Ministry of Science and Tra�cs GZ ��� �����III�a���FNRS�FWO BelgiumFINEP CNPq CAPES FUJB and FAPERJ BrazilCzech Ministry of Industry and Trade GA CR ��������� and GA AVCR A������Danish Natural Research CouncilCommission of the European Communities �DG XII�Direction des Sciences de la Mati�ere CEA FranceBundesministerium f ur Bildung Wissenschaft Forschung und Technologie GermanyGeneral Secretariat for Research and Technology GreeceNational Science Foundation �NWO� and Foundation for Research on Matter �FOM�The NetherlandsNorwegian Research CouncilState Committee for Scienti�c Research Poland P��B����� P��B���� andSPUB�P���������JNICT�Junta Nacional de Investiga!c"ao Cient#$�ca e Tecnol#ogica PortugalVedecka grantova agentura MS SR Slovakia Nr �����������Ministry of Science and Technology of the Republic of SloveniaCICYT Spain AEN������� and AEN�������The Swedish Natural Science Research CouncilParticle Physics and Astronomy Research Council UKDepartment of Energy USA DE�FG����ER�����

References

��� DELPHI Collaboration P Abreu et al� E� Phys� J� C � ������ � and Phys� Lett�

B ��� ������ ��� �� ALEPH Collaboration R Barate et al� CERN�EP������� submitted to Phys� Lett�

B%L� Collaboration M Acciarri et al� Phys� Lett� B ��� ������ ���%OPAL Collaboration G Alexander et al� Phys� Lett� B ��� ������ ���

��� DELPHI Collaboration P Aarnio et al� Nucl� Instr� and Meth� A ��� ������ � ��� DELPHI Collaboration P Abreu et al� Nucl� Instr� and Meth� A ��� ������ �� ��� M Battaglia and P M Kluit Particle Identi�cation using the DELPHI RICH De�

tectors to appear in the Proc� of the �rd Int� Workshop on Ring Imaging Cherenkov

Detectors� RICH �� Nucl� Instr� and Meth� A ������ ��� T Sj ostrand Comp� Phys� Commun� �� ������ �� ��� F A Berends R Pittau R Kleiss Comp� Phys� Commun� � ������ ��� ��� S Nova S Olshevski and T Todorov in Cern Report ����� Vol II � ��� F A Berends P H Daverveldt and R Kleiss Comp� Phys� Commun� �� ������ ��

�� and ��� ���� S Jadach B F L Ward and Z Was Comp� Phys� Commun � ������ ��� ���� DELPHI Collaboration P Abreu et al� CERN�EP������ to be published in E�

Phys� J� C ������ ��� A similar jet �avour tagging technique has been used in DELPHI Collaboration P

Abreu et al� Phys� Lett� B �� ������ �� ���� DELPHI Collaboration P Abreu et al� E� Phys� J� C � ������ ��� ���� S Catani et al� Phys� Lett� B �� ������ �� ���� A L Read Optimal statistical analysis of search results based on the likelihood ratio

and its application to the search for the MSM Higgs boson at and �� GeVDELPHI ������ PHYS ���

���

���

���

���

� ���� ��� ���� ��

���

���

���

���

� ���� ��� ���� �

���

���

���

���

� ���� ��� ���� �

����� �������� ����������

����

� ��

����

� ��

����

����

����� �������� ����������

����

� ��

����

� ��

����

����

����� �������� ����������

����

� ��

����

� ��

����

����

����� �������� ����������

����

� ��

����

� ��

����

����

���

���

���

���

� ���� ��� ���� �

Figure �� Correlation plots of the event anti�QCD and anti�WW likelihoods used todiscriminate between signal H�H� events and backgrounds for simulated QCD events�upper left� W�W� events �upper right� signal H�H� for MH � �� GeV�c� �lower left�and the

ps � ��� GeV real data �lower right� in the hadronic channel The box indicates

the signal region used in the analysis

��

���

���

���

���

� ���� ��� ���� ��

���

���

���

���

� ���� ��� ���� �

���

���

���

���

� ���� ��� ���� �

����� �������� ����������

����

� ��

����

� ��

����

����

����� �������� ����������

����

� ��

����

� ��

����

����

����� �������� ����������

����

� ��

����

� ��

����

����

����� �������� ����������

����

� ��

����

� ��

����

����

���

���

���

���

� ���� ��� ���� �

Figure � Correlation plots of the event anti�QCD and anti�WW likelihoods used todiscriminate between signal H�H� events and backgrounds for simulated QCD events�upper left� W�W� events �upper right� signal H�H� for MH � �� GeV�c� �lower left�and the

ps � ��� GeV real data �lower right� in the semileptonic channel The box

indicates the signal region used in the analysis

��

����

���

����

���

����

� ��� ��� ��� ��� ���

��

���√�

� ��√�

����

���

����

���

����

� ��� ��� ��� ��� ���

���√�

� ��√�

����

���

����

���

����

� ��� ��� ��� ��� ���

��

���√�

� ��√�

����

���

����

���

����

� ��� ��� ��� ��� ���

����

���√�

� ��√�

Figure �� Correlation plots of the total energy fraction of the less energetic jet E��ps vs

that for the more energetic jet E��ps used to discriminate between the signal and the

backgrounds in the fully leptonic channel for simulated W�W� decays �upper left� ��events �upper right� signal H�H� forMH � �� GeV�c� �lower left� and the

ps � ��� GeV

real data �lower right� The lines show the cuts de�ning the signal region used in theanalysis The distribution of E��

ps for the selected events has been used as discriminating

variable in the limit derivation program

��

�� ��������

�� �

� √

���

����

���

���

���

���

���

���

���

���

���

�� �� �� �� �� �� �� ��

Figure �� The ��� con�dence level observed and expected exclusion regions for H� in theplane BR�H� hadrons� vs MH� obtained from a combination of the search results in thehadronic semileptonic and fully leptonic decay channels at

ps � ��� GeV �� GeV and

��� GeV The sharp decrease at �� GeV�c� in the leptonic channel re�ects the maximumsignal mass considered in the lower energy samples


Recommended