Lifetime and production rate of beauty baryons from Z decays HAL Id: in2p3-00002239 http://hal.in2p3.fr/in2p3-00002239 Submitted on 14 Jun 1999 HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés. Lifetime and production rate of beauty baryons from Z decays P. Abreu, W. Adam, T. Adye, E. Agasi, I. Ajinenko, R. Aleksan, G D. Alekseev, P P. Allport, S. Almehed, S J. Alvsvaag, et al. To cite this version: P. Abreu, W. Adam, T. Adye, E. Agasi, I. Ajinenko, et al.. Lifetime and production rate of beauty baryons from Z decays. Zeitschrift für Physik C Particles and Fields, Springer Verlag, 1995, 68, pp.375-390. �in2p3-00002239� http://hal.in2p3.fr/in2p3-00002239 https://hal.archives-ouvertes.fr EUROPEAN ORGANIZATION FOR NUCLEAR RESEARCH CERN{PPE/95{54 21 April 1995 Lifetime and production rate of beauty baryons from Z decays DELPHI Collaboration Abstract The production and decay of beauty baryons (b-baryons) have been studied using 1:7 � 106 Z hadronic decays collected by the DELPHI detector at LEP. Three di�erent techniques were used to identify the b-baryons. The �rst method used pairs of a � and a lepton to tag the b-baryon decay. The second method associated fully reconstructed �c baryons with leptons. The third analysis re- constructed the b-baryon decay points by forming secondary vertices from iden- ti�ed protons and muons of opposite sign. Using these methods the following production rates were measured: f(b ! b-baryon) � BR(b-baryon ! �s`��`X) = (0:30 � 0:06 � 0:04)%; f(b ! b-baryon) � BR(b-baryon ! �c`��`X) = (1:18 � 0:26+0:31�0:21)%; f(b ! b-baryon) � BR(b-baryon ! p����X) = (0:49 � 0:11�+0:15�0:11)%: The average b-baryon lifetime was determined to be: � = 1:21+:21 �:18(stat:) � 0:04(exp:syst:)+:02�:07(th:syst:) ps. (To be submitted to Zeit. f. Physik C) ii P.Abreu 21 , W.Adam 50 , T.Adye 37 , E.Agasi 31 , I.Ajinenko 42 , R.Aleksan 39 , G.D.Alekseev 16 , P.P.Allport 22 , S.Almehed 24 , S.J.Alvsvaag 4 , U.Amaldi 9 , S.Amato 47 , A.Andreazza 28 , M.L.Andrieux 14 , P.Antilogus 25 , W-D.Apel 17 , Y.Arnoud 39 , B.�Asman 44 , J-E.Augustin 19 , A.Augustinus 31 , P.Baillon 9 , P.Bambade 19 , F.Barao 21 , R.Barate 14 , G.Barbiellini 46 , D.Y.Bardin 16 , G.J.Barker 35 , A.Baroncelli 40 , O.Barring 24 , J.A.Barrio 26 , W.Bartl 50 , M.J.Bates 37 , M.Battaglia 15 , M.Baubillier 23 , J.Baudot 39 , K-H.Becks 52 , M.Begalli 6 , P.Beilliere 8 , Yu.Belokopytov9, A.C.Benvenuti5, M.Berggren41, D.Bertrand2, F.Bianchi45, M.Bigi45, M.S.Bilenky16, P.Billoir 23 , D.Bloch 10 , M.Blume 52 , S.Blyth 35 , V.Bocci 38 , T.Bolognese 39 , M.Bonesini 28 , W.Bonivento 28 , P.S.L.Booth 22 , G.Borisov 42 , C.Bosio 40 , S.Bosworth 35 , O.Botner 48 , B.Bouquet 19 , C.Bourdarios 9 , T.J.V.Bowcock 22 , M.Bozzo 13 , P.Branchini 40 , K.D.Brand 36 , R.A.Brenner 15 , C.Bricman 2 , L.Brillault 23 , R.C.A.Brown 9 , P.Bruckman 18 , J-M.Brunet 8 , L.Bugge 33 , T.Buran 33 , A.Buys 9 , M.Caccia 28 , M.Calvi 28 , A.J.Camacho Rozas41, T.Camporesi9, V.Canale38, M.Canepa13, K.Cankocak44, F.Cao2, F.Carena9, P.Carrilho 47 , L.Carroll 22 , C.Caso 13 , M.V.Castillo Gimenez 49 , A.Cattai 9 , F.R.Cavallo 5 , L.Cerrito 38 , V.Chabaud 9 , M.Chapkin 42 , Ph.Charpentier 9 , L.Chaussard 25 , J.Chauveau 23 , P.Checchia 36 , G.A.Chelkov 16 , R.Chierici 45 , P.Chliapnikov 42 , P.Chochula 7 , V.Chorowicz 9 , V.Cindro 43 , P.Collins 9 , J.L.Contreras 19 , R.Contri 13 , E.Cortina 49 , G.Cosme 19 , F.Cossutti 46 , H.B.Crawley 1 , D.Crennell 37 , G.Crosetti 13 , J.Cuevas Maestro 34 , S.Czellar 15 , E.Dahl-Jensen 29 , J.Dahm 52 , B.Dalmagne 19 , M.Dam 33 , G.Damgaard 29 , A.Daum 17 , P.D.Dauncey 37 , M.Davenport 9 , W.Da Silva 23 , C.Defoix 8 , G.Della Ricca 46 , P.Delpierre 27 , N.Demaria 35 , A.De Angelis 9 , H.De Boeck 2 , W.De Boer 17 , S.De Brabandere 2 , C.De Clercq 2 , C.De La Vaissiere 23 , B.De Lotto 46 , A.De Min 28 , L.De Paula 47 , C.De Saint-Jean 39 , H.Dijkstra 9 , L.Di Ciaccio 38 , F.Djama 10 , J.Dolbeau 8 , M.Donszelmann 9 , K.Doroba 51 , M.Dracos 10 , J.Drees 52 , K.-A.Drees 52 , M.Dris 32 , Y.Dufour 8 , F.Dupont 14 , D.Edsall 1 , R.Ehret 17 , G.Eigen 4 , T.Ekelof 48 , G.Ekspong 44 , M.Elsing 52 , J-P.Engel 10 , N.Ershaidat 23 , B.Erzen 43 , M.Espirito Santo 21 , E.Falk24, D.Fassouliotis32, M.Feindt9, A.Fenyuk42, A.Ferrer49, T.A.Filippas32, A.Firestone1, H.Foeth9, E.Fokitis 32 , F.Fontanelli 13 , F.Formenti 9 , B.Franek 37 , P.Frenkiel 8 , D.C.Fries 17 , A.G.Frodesen 4 , R.Fruhwirth 50 , F.Fulda-Quenzer 19 , H.Furstenau 9 , J.Fuster 49 , A.Galloni 22 , D.Gamba 45 , M.Gandelman 6 , C.Garcia 49 , J.Garcia 41 , C.Gaspar 9 , U.Gasparini 36 , Ph.Gavillet 9 , E.N.Gazis 32 , D.Gele 10 , J-P.Gerber 10 , M.Gibbs 22 , D.Gillespie 9 , R.Gokieli 51 , B.Golob 43 , G.Gopal 37 , L.Gorn 1 , M.Gorski 51 , V.Gracco 13 , E.Graziani 40 , G.Grosdidier 19 , P.Gunnarsson44, M.Gunther48, J.Guy37, U.Haedinger17, F.Hahn52, M.Hahn17, S.Hahn52, Z.Hajduk18, A.Hallgren 48 , K.Hamacher 52 , W.Hao 31 , F.J.Harris 35 , V.Hedberg 24 , R.Henriques 21 , J.J.Hernandez 49 , P.Herquet 2 , H.Herr 9 , T.L.Hessing 9 , E.Higon 49 , H.J.Hilke 9 , T.S.Hill 1 , S-O.Holmgren 44 , P.J.Holt 35 , D.Holthuizen 31 , M.Houlden 22 , J.Hrubec 50 , K.Huet 2 , K.Hultqvist 44 , P.Ioannou 3 , J.N.Jackson 22 , R.Jacobsson 44 , P.Jalocha 18 , R.Janik 7 , G.Jarlskog 24 , P.Jarry 39 , B.Jean-Marie 19 , E.K.Johansson 44 , L.Jonsson 24 , P.Jonsson 24 , C.Joram 9 , P.Juillot 10 , M.Kaiser 17 , G.Kalmus 37 , F.Kapusta 23 , M.Karlsson 44 , E.Karvelas 11 , S.Katsanevas 3 , E.C.Katsou�s 32 , R.Keranen 15 , B.A.Khomenko 16 , N.N.Khovanski 16 , B.King 22 , N.J.Kjaer 29 , H.Klein 9 , A.Klovning 4 , P.Kluit 31 , J.H.Koehne 17 , B.Koene 31 , P.Kokkinias 11 , M.Koratzinos 9 , V.Kostioukhine 42 , C.Kourkoumelis 3 , O.Kouznetsov 13 , P.-H.Kramer 52 , M.Krammer 50 , C.Kreuter 17 , J.Krolikowski 51 , I.Kronkvist 24 , Z.Krumstein 16 , W.Krupinski 18 , P.Kubinec 7 , W.Kucewicz 18 , K.Kurvinen 15 , C.Lacasta 49 , I.Laktineh 25 , S.Lamblot 23 , J.W.Lamsa 1 , L.Lanceri 46 , D.W.Lane 1 , P.Langefeld 52 , V.Lapin 42 , I.Last 22 , J-P.Laugier 39 , R.Lauhakangas15, G.Leder50, F.Ledroit14, V.Lefebure2, C.K.Legan1, R.Leitner30, Y.Lemoigne39, J.Lemonne2, G.Lenzen 52 , V.Lepeltier 19 , T.Lesiak 36 , D.Liko 50 , R.Lindner 52 , A.Lipniacka 19 , I.Lippi 36 , B.Loerstad 24 , M.Lokajicek 12 , J.G.Loken 35 , J.M.Lopez 41 , A.Lopez-Fernandez 9 , M.A.Lopez Aguera 41 , D.Loukas 11 , P.Lutz 39 , L.Lyons 35 , J.MacNaughton 50 , G.Maehlum 17 , A.Maio 21 , V.Malychev 16 , F.Mandl 50 , J.Marco 41 , B.Marechal 47 , M.Margoni 36 , J-C.Marin 9 , C.Mariotti 40 , A.Markou 11 , T.Maron 52 , C.Martinez-Rivero 41 , F.Martinez-Vidal 49 , S.Marti i Garcia49, F.Matorras41, C.Matteuzzi28, G.Matthiae38, M.Mazzucato36, M.Mc Cubbin9, R.Mc Kay1, R.Mc Nulty 22 , J.Medbo 48 , C.Meroni 28 , W.T.Meyer 1 , M.Michelotto 36 , E.Migliore 45 , L.Mirabito 25 , W.A.Mitaro� 50 , U.Mjoernmark 24 , T.Moa 44 , R.Moeller 29 , K.Moenig 9 , M.R.Monge 13 , P.Morettini 13 , H.Mueller 17 , L.M.Mundim 6 , W.J.Murray 37 , B.Muryn 18 , G.Myatt 35 , F.Naraghi 14 , F.L.Navarria 5 , S.Navas 49 , P.Negri 28 , S.Nemecek 12 , W.Neumann 52 , N.Neumeister 50 , R.Nicolaidou 3 , B.S.Nielsen 29 , M.Nieuwenhuizen 31 , V.Nikolaenko 10 , P.Niss 44 , A.Nomerotski 36 , A.Normand 35 , W.Oberschulte-Beckmann 17 , V.Obraztsov 42 , A.G.Olshevski 16 , A.Onofre 21 , R.Orava 15 , A.Ostankov 42 , K.Osterberg 15 , A.Ouraou 39 , P.Paganini 19 , M.Paganoni 28 , P.Pages 10 , H.Palka 18 , Th.D.Papadopoulou 32 , L.Pape 9 , C.Parkes 35 , F.Parodi 13 , A.Passeri 40 , M.Pegoraro 36 , L.Peralta 21 , H.Pernegger 50 , M.Pernicka 50 , A.Perrotta 5 , C.Petridou 46 , A.Petrolini 13 , H.T.Phillips 37 , G.Piana 13 , F.Pierre 39 , S.Plaszczynski 19 , O.Podobrin 17 , M.E.Pol 6 , G.Polok 18 , P.Poropat 46 , V.Pozdniakov 16 , M.Prest 46 , P.Privitera 38 , N.Pukhaeva 16 , A.Pullia 28 , D.Radojicic 35 , S.Ragazzi 28 , H.Rahmani 32 , J.Rames12, P.N.Rato�20, A.L.Read33, M.Reale52, P.Rebecchi19, N.G.Redaelli28, D.Reid9, P.B.Renton35, L.K.Resvanis 3 , F.Richard 19 , J.Richardson 22 , J.Ridky 12 , G.Rinaudo 45 , I.Ripp 39 , A.Romero 45 , I.Roncagliolo 13 , P.Ronchese 36 , L.Roos 14 , E.I.Rosenberg 1 , E.Rosso 9 , P.Roudeau 19 , T.Rovelli 5 , W.Ruckstuhl 31 , V.Ruhlmann-Kleider 39 , A.Ruiz 41 , K.Rybicki 18 , H.Saarikko 15 , Y.Sacquin 39 , A.Sadovsky 16 , G.Sajot 14 , J.Salt 49 , J.Sanchez 26 , M.Sannino 13 , H.Schneider 17 , M.A.E.Schyns 52 , G.Sciolla 45 , F.Scuri 46 , Y.Sedykh 16 , A.M.Segar 35 , iii A.Seitz 17 , R.Sekulin 37 , R.C.Shellard 6 , I.Siccama 31 , P.Siegrist 39 , S.Simonetti 39 , F.Simonetto 36 , A.N.Sisakian 16 , B.Sitar 7 , T.B.Skaali 33 , G.Smadja 25 , N.Smirnov 42 , O.Smirnova 16 , G.R.Smith 37 , R.Sosnowski 51 , D.Souza-Santos 6 , T.Spassov 21 , E.Spiriti 40 , S.Squarcia 13 , H.Staeck 52 , C.Stanescu 40 , S.Stapnes 33 , I.Stavitski 36 , K.Stepaniak 51 , F.Stichelbaut 9 , A.Stocchi 19 , J.Strauss 50 , R.Strub 10 , B.Stugu 4 , M.Szczekowski 51 , M.Szeptycka 51 , T.Tabarelli 28 , J.P.Tavernet 23 , O.Tchikilev 42 , A.Tilquin 27 , J.Timmermans 31 , L.G.Tkatchev 16 , T.Todorov 10 , D.Z.Toet31, A.Tomaradze2, B.Tome21, E.Torassa45, L.Tortora40, G.Transtromer24, D.Treille9, W.Trischuk9, G.Tristram 8 , A.Trombini 19 , C.Troncon 28 , A.Tsirou 9 , M-L.Turluer 39 , I.A.Tyapkin 16 , M.Tyndel 37 , S.Tzamarias 22 , B.Ueberschaer 52 , S.Ueberschaer 52 , O.Ullaland 9 , V.Uvarov 42 , G.Valenti 5 , E.Vallazza 9 , G.W.Van Apeldoorn 31 , P.Van Dam 31 , W.K.Van Doninck 2 , J.Van Eldik 31 , N.Vassilopoulos 35 , G.Vegni 28 , L.Ventura 36 , W.Venus 37 , F.Verbeure 2 , M.Verlato 36 , L.S.Vertogradov 16 , D.Vilanova 39 , P.Vincent 25 , L.Vitale 46 , E.Vlasov42, A.S.Vodopyanov16, V.Vrba12, H.Wahlen52, C.Walck44, A.Wehr52, M.Weierstall52, P.Weilhammer9, A.M.Wetherell 9 , D.Wicke 52 , J.H.Wickens 2 , M.Wielers 17 , G.R.Wilkinson 35 , W.S.C.Williams 35 , M.Winter 10 , M.Witek 9 , G.Wormser 19 , K.Woschnagg 48 , K.Yip 35 , F.Zach 25 , C.Zacharatou 24 , A.Zaitsev 42 , A.Zalewska 18 , P.Zalewski 51 , D.Zavrtanik 43 , E.Zevgolatakos 11 , N.I.Zimin 16 , M.Zito 39 , D.Zontar 43 , R.Zuberi 35 , G.C.Zucchelli 44 , G.Zumerle 36 1 Ames Laboratory and Department of Physics, Iowa State University, Ames IA 50011, USA 2 Physics Department, Univ. Instelling Antwerpen, Universiteitsplein 1, B-2610 Wilrijk, Belgium and IIHE, ULB-VUB, Pleinlaan 2, B-1050 Brussels, Belgium and Facult�e des Sciences, Univ. de l'Etat Mons, Av. Maistriau 19, B-7000 Mons, Belgium 3 Physics Laboratory, University of Athens, Solonos Str. 104, GR-10680 Athens, Greece 4 Department of Physics, University of Bergen, All�egaten 55, N-5007 Bergen, Norway 5 Dipartimento di Fisica, Universit�a di Bologna and INFN, Via Irnerio 46, I-40126 Bologna, Italy 6 Centro Brasileiro de Pesquisas F�isicas, rua Xavier Sigaud 150, RJ-22290 Rio de Janeiro, Brazil and Depto. de F�isica, Pont. Univ. Cat�olica, C.P. 38071 RJ-22453 Rio de Janeiro, Brazil and Inst. de F�isica, Univ. Estadual do Rio de Janeiro, rua S~ao Francisco Xavier 524, Rio de Janeiro, Brazil 7 Comenius University, Faculty of Mathematics and Physics, Mlynska Dolina, SK-84215 Bratislava, Slovakia 8 Coll�ege de France, Lab. de Physique Corpusculaire, IN2P3-CNRS, F-75231 Paris Cedex 05, France 9 CERN, CH-1211 Geneva 23, Switzerland 10 Centre de Recherche Nucl�eaire, IN2P3 - CNRS/ULP - BP20, F-67037 Strasbourg Cedex, France 11 Institute of Nuclear Physics, N.C.S.R. Demokritos, P.O. Box 60228, GR-15310 Athens, Greece 12 FZU, Inst. of Physics of the C.A.S. High Energy Physics Division, Na Slovance 2, 180 40, Praha 8, Czech Republic 13 Dipartimento di Fisica, Universit�a di Genova and INFN, Via Dodecaneso 33, I-16146 Genova, Italy 14 Institut des Sciences Nucl�eaires, IN2P3-CNRS, Universit�e de Grenoble 1, F-38026 Grenoble Cedex, France 15 Research Institute for High Energy Physics, SEFT, P.O. Box 9, FIN-00014 Helsinki, Finland 16 Joint Institute for Nuclear Research, Dubna, Head Post O�ce, P.O. Box 79, 101 000 Moscow, Russian Federation 17 Institut f�ur Experimentelle Kernphysik, Universit�at Karlsruhe, Postfach 6980, D-76128 Karlsruhe, Germany 18 High Energy Physics Laboratory, Institute of Nuclear Physics, Ul. Kawiory 26a, PL-30055 Krakow 30, Poland 19 Universit�e de Paris-Sud, Lab. de l'Acc�el�erateur Lin�eaire, IN2P3-CNRS, Bat 200, F-91405 Orsay Cedex, France 20 School of Physics and Materials, University of Lancaster, Lancaster LA1 4YB, UK 21 LIP, IST, FCUL - Av. Elias Garcia, 14-1 o , P-1000 Lisboa Codex, Portugal 22 Department of Physics, University of Liverpool, P.O. Box 147, Liverpool L69 3BX, UK 23 LPNHE, IN2P3-CNRS, Universit�es Paris VI et VII, Tour 33 (RdC), 4 place Jussieu, F-75252 Paris Cedex 05, France 24 Department of Physics, University of Lund, S�olvegatan 14, S-22363 Lund, Sweden 25 Universit�e Claude Bernard de Lyon, IPNL, IN2P3-CNRS, F-69622 Villeurbanne Cedex, France 26 Universidad Complutense, Avda. Complutense s/n, E-28040 Madrid, Spain 27 Univ. d'Aix - Marseille II - CPP, IN2P3-CNRS, F-13288 Marseille Cedex 09, France 28 Dipartimento di Fisica, Universit�a di Milano and INFN, Via Celoria 16, I-20133 Milan, Italy 29 Niels Bohr Institute, Blegdamsvej 17, DK-2100 Copenhagen 0, Denmark 30 NC, Nuclear Centre of MFF, Charles University, Areal MFF, V Holesovickach 2, 180 00, Praha 8, Czech Republic 31 NIKHEF-H, Postbus 41882, NL-1009 DB Amsterdam, The Netherlands 32 National Technical University, Physics Department, Zografou Campus, GR-15773 Athens, Greece 33 Physics Department, University of Oslo, Blindern, N-1000 Oslo 3, Norway 34 Dpto. Fisica, Univ. Oviedo, C/P. P�erez Casas, S/N-33006 Oviedo, Spain 35 Department of Physics, University of Oxford, Keble Road, Oxford OX1 3RH, UK 36 Dipartimento di Fisica, Universit�a di Padova and INFN, Via Marzolo 8, I-35131 Padua, Italy 37 Rutherford Appleton Laboratory, Chilton, Didcot OX11 OQX, UK 38 Dipartimento di Fisica, Universit�a di Roma II and INFN, Tor Vergata, I-00173 Rome, Italy 39 Centre d'Etude de Saclay, DSM/DAPNIA, F-91191 Gif-sur-Yvette Cedex, France 40 Istituto Superiore di Sanit�a, Ist. Naz. di Fisica Nucl. (INFN), Viale Regina Elena 299, I-00161 Rome, Italy 41 C.E.A.F.M., C.S.I.C. - Univ. Cantabria, Avda. los Castros, S/N-39006 Santander, Spain, (CICYT-AEN93-0832) 42 Inst. for High Energy Physics, Serpukov P.O. Box 35, Protvino, (Moscow Region), Russian Federation 43 J. Stefan Institute and Department of Physics, University of Ljubljana, Jamova 39, SI-61000 Ljubljana, Slovenia 44 Fysikum, Stockholm University, Box 6730, S-113 85 Stockholm, Sweden 45 Dipartimento di Fisica Sperimentale, Universit�a di Torino and INFN, Via P. Giuria 1, I-10125 Turin, Italy 46 Dipartimento di Fisica, Universit�a di Trieste and INFN, Via A. Valerio 2, I-34127 Trieste, Italy and Istituto di Fisica, Universit�a di Udine, I-33100 Udine, Italy 47 Univ. Federal do Rio de Janeiro, C.P. 68528 Cidade Univ., Ilha do Fund~ao BR-21945-970 Rio de Janeiro, Brazil 48 Department of Radiation Sciences, University of Uppsala, P.O. Box 535, S-751 21 Uppsala, Sweden 49 IFIC, Valencia-CSIC, and D.F.A.M.N., U. de Valencia, Avda. Dr. Moliner 50, E-46100 Burjassot (Valencia), Spain 50 Institut f�ur Hochenergiephysik, �Osterr. Akad. d. Wissensch., Nikolsdorfergasse 18, A-1050 Vienna, Austria 51 Inst. Nuclear Studies and University of Warsaw, Ul. Hoza 69, PL-00681 Warsaw, Poland 52 Fachbereich Physik, University of Wuppertal, Postfach 100 127, D-42097 Wuppertal 1, Germany 1 1 Introduction The �b baryon was �rst observed in the exclusive decay �b ! �J= by the UA1 experiment at the Sp�pS collider [1]. Evidence for its production in Z hadronic decays has been reported by the LEP experiments [2,3]. They attributed the observed correlation between �'s and leptons (`'s) to �b decays. Measurements of the average b-baryon lifetime have been recently published [3,4]. Its precise determination tests the theory of heavy quark decays and the simple quark-spectator model. This is of particular interest for the beauty quark [5] where, due to the high b-quark mass, the theoretical predictions based on perturbative expansions are less uncertain than those for charm decays. This paper extends the previous analysis [3] and adds two new semileptonic decay channels, based on the detection of a �c or a fast proton (p) in the same jet as a high transverse momentum lepton. The �` channel provides a clear signature for b-baryon production but the position of the b-baryon decay vertex is precisely determined with relatively low e�ciency. The �c` channel provides the purest b-baryon sample. Finally the p` channel relies on the particle identi�cation capabilities of DELPHI. 2 The DELPHI Detector The DELPHI detector has been described in detail elsewhere [6]. Both charged particle tracking through the uniform axial �eld and particle identi�cation are important in this analysis. The detector elements used for tracking are: the Vertex Detector (VD), the Inner Detector (ID), the Time Projection Chamber (TPC) and the Outer Detector (OD). The other important detectors are: the the Ring Imaging Cherenkov detector (RICH) for hadron identi�cation, the barrel electromagnetic calorimeter (HPC) and the muon chambers for lepton identi�cation. The ionization loss dE=dx measurements in the TPC are also used for particle identi�cation. The VD, consisting of 3 cylindrical layers of silicon detectors (radii 6, 8 and 11 cm), provides up to 3 hits per track (or more in small overlapping regions) in the polar angle range 43o < � < 137o. The intrinsic resolution of the VD points is �8�m, measured only in the plane transverse to the beam direction (r� plane). The precision on the impact parameter with respect to the primary vertex of a track having hits associated in the VD is �26�m, measured in dimuon Z events. Charged particle tracks were reconstructed with 95% e�ciency and with a momentum resolution �p=p < 2:0 � 10�3p (GeV/c). The primary vertex of the e+e� interaction was reconstructed on an event-by-event basis using a beam spot constraint. The position of the primary vertex could be determined in this way to a precision of about 40�m (slightly dependent on the avour of the primary quark-antiquark pair) in the plane transverse to the beam direction. In this plane secondary vertices from beauty and charm decays were reconstructed with a precision of �300�m along the ight direction of the decaying particles. The � ! p� decays could be reconstructed if the distance (in the r� plane) between the � decay point and primary vertex was less than 90 cm. This condition meant that the proton and pion had track segments at least 20 cm long in the TPC. Hadron identi�cation relied on the speci�c ionization in the TPC and on the RICH detector. The dE=dx measurement had a precision of �7% in the momentum range 4 < p < 25 GeV/c. The RICH detector [7] consisted of a liquid radiator which provided p=K=� separation in the intermediate momentum region 2{8 GeV/c, and a gas radiator which worked in veto mode for proton selection in the region 8{15 GeV/c and separated protons from kaons for momenta less than 30 GeV/c. 2 The barrel electromagnetic calorimeter (HPC), covered the polar angle region 46� < � < 134�, and detected electrons with an energy precision �E=E = 0:25= p E(GeV ). Two planes of muon chambers covered the polar angle region 20� < � < 160�, except for two regions of �3� around � = 42� and � = 138�. The �rst layer was inside the return yoke of the magnet, after 90 cm of iron, while the second was mounted outside the yoke, behind a further 20 cm of iron. 3 Lepton selection and hadron identi�cation Hadronic events from Z decays were selected by requiring a charged multiplicitygreater than 4 and a total reconstructed energy greater than 0.12 p s; charged particles were required to have a momentum greater than 0.4 GeV/c and polar angle between 20o and 160o. The overall trigger and selection e�ciency was 0:950�0:011 [8]. Lepton candidates in these events were used in the analysis if their momentum was greater than 3 GeV/c. 3.1 Electron Identi�cation The probability of any track being due to an electron was calculated using the spatial separation between the extrapolated position of a track at the HPC and the position of the nearest electro-magnetic shower, a match between the measured energy and the track momentum and a successful �t to the longitudinal pro�le of the shower in the 9 HPC layers [9]. Tracks with a �2 probability greater than 4% for this electron hypothesis were retained for further analysis. The �nal electron sample was obtained by using additional information from the TPC and RICH. The dE=dx measurement in the TPC was used to check that the speci�c ionization for the track was consistent with that expected from electrons with a probability of at least 2%. Also, when the gas RICH was sensitive it was required to show at least one associated photoelectron at the correct angle for the electron hypothesis. Electrons arising from photon conversions were removed by a vertex �t to pairs of electron candidates. If the e+e� invariant mass was reconstructed to be less than 20 MeV/c2 the pair was assumed to be a converted photon. Using this procedure the electron identi�cation e�ciency in the HPC �ducial volume was found to be (62 � 1)%, with a hadron misidenti�cation probability of (1:5 � 0:4)%. 3.2 Muon Identi�cation The identi�cation of muons relied on the muon chambers. Tracks were extrapolated to the muon chambers and a global �2 of the track was used to de�ne a re�tting procedure which took into account the multiple scattering between the inner tracking devices and the muon chambers. At least 1 hit in the chamber layer outside the iron yoke and a �2=ndof < 5 were required (< 6 in the forward region). The corresponding muon identi�cation e�ciency was (81 � 1)% in the barrel and (82 � 2)% in the end-caps, with hadron misidenti�cation probabilities of (1:01 � 0:05)% and (1:15 � 0:08)% respectively. 3.3 Hadron Identi�cation using the RICH Particle identi�cation using the DELPHI RICH detector has been described in detail elsewhere [10]. The three analyses presented in this paper used protons with momentum range well above the pion threshold in the gas radiator of 2.5 GeV/c. Above this threshold, the gas radiator worked in veto-mode for p=� separation up to 16 GeV/c, with 75% 3 e�ciency and a pion rejection factor of 15. A K=p separation with the same background rejection power was ensured in this mode of operation between 8.5 GeV/c, the gas radiator threshold for kaons, and 16 GeV/c. Above this energy, identi�cation was provided by the measurement of the Cherenkov angle of the detected photons using a \ring identi�cation mode" algorithm [10], with 80% e�ciency and rejection factors 5-10. This algorithm was also applied to the liquid radiator data, which provided complementary information for K=� and K=p separation in the momentum range 1-7 GeV/c. The RICH was operational for 25% of the 1991 data, 60% of the 1992 data and nearly 90% of the 1993 data sample (gas radiator only). 4 �` Channel The analysis of events with a � and a lepton is based on about 1.7 � 106 Z hadronic decays collected in the years 1991 - 1993. Decays of b-baryons with a �` pair in the �nal state originate mainly from the decay chain: b-baryon ! �c`�X;�c ! �X. These decays have the following properties: the lepton has high transverse and longitudinal momentum, the � has a harder momentum spectrum than the � produced in light quark fragmentation and the �` pair has the right sign i.e. p`� rather than p`+, where p is the proton from the � decay. In the following the lepton transverse momentum (pT ) is computed, if not otherwise stated, with respect to the jet axis de�ned including the lepton in the jet. Charged particles are clustered into jets using the LUND jet �nding algorithm [11] (routine LUCLUS) with a clustering mass parameter equal to 2.5 GeV/c2. Semileptonic B meson decays, such as B ! �c �N`��X (where �N is an antibaryon), can also contribute to an excess in right sign pairs. This was estimated to be negligible, under the conservative assumption that 100% of b quarks hadronize to a B meson, using the 90% CL upper limit BR( �B ! p`��X) < 1:6 10�3 [12] and the CLEO result that 30% of the protons produced in B decays come from � particles [13]. This conclusion takes into account the fact that the e�ciency of the selection cuts described below (section 4.2) for this channel is smaller by a factor 3 than for the b-baryon decay. Background events from direct c-quark production through the c ! �c ! �`�X decay chain have protons and the leptons of the same sign; in addition, the lepton pT spectrum is softer. A quantitative analysis of the background based on detailed simulation of Z hadronic events is discussed in section 4.3. 4.1 � selection In the search for � ! p� decay all pairs of opposite sign charged particles with momentum 0:1 < p < 30 GeV/c were considered. A candidate � vertex was formed if the minimum separation in the r� plane of the two tracks was less than 3 mm and if their perigee separation in the beam direction was less than 5 mm. If the same track was associated with more than one vertex only the vertex with the largest decay length (in the r� projection) was used. For decays inside the beam pipe at least one vertex detector hit was required per track. Only combinations where the vertex was closer to the primary vertex than the starting point of both tracks were kept. Particle identi�cation greatly improved the background rejection with negligible loss in e�ciency. The identi�cation criteria using the dE=dx measurement in the TPC and the selections for rejection of conversions and K0 decays are described previously [3]. If the extrapolation of the track of charged particle with highest momentum (assumed to be the proton) to the RICH was in the sensitive volume of the detector and the RICH was 4 operational, the identi�cation algorithm described in section 3.3 was used. To improve the signal-to-noise ratio further, the following kinematic selection criteria were applied: the angle in the r� plane between the line of ight and the reconstructed � momentum was required to be smaller than 2� and the probability for the lifetime of the � decay candidate to be greater than that observed was required to be greater than 4%. Figure 1a shows the p� invariant mass distribution for the remaining candidates with momentum greater than 4 GeV/c. In this sample the �tted � signal was 22793 � 556 decays, with a � mass mean value of 1114:9 � 0:1 MeV/c2 and a measured width of 4:1 � 0:6 MeV/c2. The momentum distribution for the reconstructed � candidates with the background subtracted is shown in �gure 1.b for the mass range from 1106 to 1126 MeV/c2. It is compared with the prediction of the DELPHI simulation program using the JETSET 7.3 model [14] with the results analysed using the same programs as the real data. The � ! p� reconstruction e�ciency from the simulation, shown in �gure 1c, was (20�1)% for p > 4 GeV/c. This increase in e�ciency compared with the previous DELPHI publication [3] is due to improved pattern recognition. 4.2 �` Correlations To select � and leptons coming from the �b decay chain, the following criteria were applied: the momentum of the �candidate was required to be greater than 4 GeV/c and the momentum of the lepton greater than 3 GeV/c. The lepton was only used if it was in the same jet as the � and its pT was greater than 0.5 GeV/c. The mass of the �` combination was required to lie in the range 1.9 to 5.0 GeV/c2 and the �` pairs were only selected for analysis if their total momentumwas greater than 9 GeV/c. In the simulation the above procedure reduced background sources of �` pairs by more than two orders of magnitude [3] and selected �b ! �`�X decays (provided the � was reconstructed) with an e�ciency of (50 � 3)%. The p� invariant mass spectra in the data for the right and wrong sign �` pairs are shown by the dots in �gures 2.a and 2.b, together with the result of a �t to the data using a Gaussian function and a polynomial background. The �t gives a signal of (234 � 20) �'s in the right sign pairs and (112 � 19) �'s in the wrong sign pairs. The histograms show the corresponding distributions from simulation normalized to the total number of hadronic Z events. The yield of genuine �'s predicted by the simulation is shown by the single hatched area; the double hatched areas show the simulation prediction for the � coming from a b-baryon decay. The simulation assumed a �b production rate f(b ! �b) � Br(�b ! �`�X) = 0:3% and a combined �b and �b production rate of 0.03 %. It also predicted a small signal in the wrong sign pair combinations, due to �c ! �`�X decays and to the associated production of �b + �� in which the �� was reconstructed and associated with the lepton. 4.3 Branching Ratios As shown in �gures 2.a,b, the simulation included a large number of �'s coming from sources other than b-baryon decays, in both right and wrong sign combinations. The absolute value was model dependent and was not used in this analysis. However the ratio (R = 1:0 � 0:1) of the background level of �'s in the two distributions in �gures 2.a and 2.b was assumed to be correct. The statistical error of 0.1 on this ratio was 5 included in the systematic error on the production rate. Moreover, a small b-baryon signal (15 � 5% of the signal in the right sign sample) was predicted in the wrong sign pair sample. Thus, to estimate the b-baryon yield in the right sign sample, the � signal in wrong sign combinations was subtracted from the signal in �gure 2a and the result scaled by the correction factor C = 1=(0:85 � 0:05). This led to a total b-baryon signal of 144 � 33(stat:) � 14(syst:) events. For the analysis of the �� pairs, a hadronic data sample in which the TPC and the barrel and forward muon chambers were more than 90% operational was used. This selected 1,620,000 Z events. The overall e�ciency for the �� channel was (4:4 � 0:4)%. The estimated number of b-baryons in this sample (118 � 27 � 12) leads to a production rate: f(b! b-baryon) � Br(b-baryon! ���X) = (0:36 � 0:07+0:05 �0:04)%. For the analysis of the �e pairs, the hadronic data sample in which the TPC and HPC were more than 90% operational was used; this requirement selected 1,589,000 Z events. The overall e�ciency for the �e channel was (2:0 � 0:3)%. The estimated number of b-baryons in the sample was (26 � 19 � 3), giving a production rate: f(b! b-baryon) � Br(b-baryon! �e�X) = (0:18 � 0:12+0:03 �0:02)%. Assuming lepton universality, these results may be averaged to give: f(b! b-baryon) � Br(b-baryon! �`�X) = (0:30 � 0:06 � 0:04)%. Table 1 shows the contributions from di�erent sources to the total systematic uncer- tainty. The e�ciency of the selection de�ned by the kinematic cuts discussed in section 4.2 was dependent on the momentum spectrum, the polarization and the decay model assumed for the b-baryon. The polarization value quoted in the table is the central one of the allowed range [-0.936, 0.0], where the lower limit is the Standard Model predic- tion for the polarization of the original b quark, assuming sin2�W = 0:23. The b-baryon semileptonic decay was simulated in the framework of Heavy Quark E�ective Theory [15] using the following parameterization of the Isgur-Wise function: �(!) = exp[aIW(1 � !)]; where ! = v�b � v�c and v�b (v�c) is the b-baryon (c-baryon) 4-velocity. A further e�ect arose if resonant and non-resonant �b ! �cn�`� decays were an important fraction of the total width, where n is a positive integer. Finally, di�erent assumptions about the �c ! �X branching fractions gave negligible e�ects on the overall e�ciency. As can be seen from the table, the dominant contribution to the systematic error comes from the background subtraction procedure used to eliminate accidental �` correlations. The above result can be compared with the previous determination by DELPHI [3]: f(b ! �b) � Br(�b ! �`�X) = (0:41 � 0:13(stat:) � 0:09(syst:))%. Figure 3a shows the right-sign � momentum spectrum after the subtraction of the wrong sign �sample for the data (dots); the superimposed histogram, showing the simu- lation prediction for the momentum of reconstructed � originating from a b-baryon, was in good agreement with the observed spectrum. Similar plots for the lepton pT spectrum, the sum of the lepton and � momenta and the �` invariant mass are shown in �gures 3.b-d. 6 Table 1: Contributions to the total systematic uncertainty on the b-baryon production rate times its branching ratio to �`�X. Source variation level Syst:uncertainty(�102) lepton identi�cation e�ciency �2% �0:006 � reconstruction e�ciency 0:20 � 0:01 �0:016 background subtraction � �0:032 < Eb > =Ebeam 0:70 � 0:03 �0:009 �(!) = exp[aIW (1 � !)] aIW = 1:7+3:3�1:7 �0:008 �b polarization �0:47 � 0:47 �0:013 BR(�b ! �c`�n�)=BR(�b ! �c`�) 0 ! 0:3 +0:020 total syst: uncertainty � +0:045 �0:040 4.4 Measurement of b-baryon lifetime The analysis followed the method previously used [3] and was based on the muon sample only. Since the extrapolation of the � ight direction to the interaction region was not precise enough to separate secondary from tertiary vertices in the b-baryon decay chain, a unique secondary vertex was reconstructed using the �, the correlated high pT muon and an oppositely charged particle (assumed to be a pion) with momentum greater than 0.4 GeV/c . The muon and the candidate pion were required to have at least 2 associated hits in the microvertex detector. To reduce the combinatorial background, the (���) invariant mass was required to be less than 5.6 GeV/c2 and the (��) invariant mass to be less than 2.4 GeV/c2. Furthermore, the contribution of the muon and pion track to the �2 of the vertex was required to be less than 3.5 and the contribution of the � ight path less than 5. In case of more than one reconstructed vertex, the vertex with the pion of highest momentum was chosen. Out of 240 right sign �� events with 1:106 < M(p�) < 1:126 GeV/c2, 63 decay vertices were reconstructed. This procedure selected b-baryons in which the subsequent charmed particle in the decay chain had a small decay length with respect to the resolution of secondary vertices. In simulated data this did not introduce any sizeable bias in the decay length distribution of the b-baryon; the e�ciency was 40%, and in 90% of the cases the candidate pion associated with the vertex originated from the �b decay chain. The b-baryon purity of the sample after the vertex reconstruction, Fs, was determined from the data by a �t to the mass plots for the right and wrong sign correlations (�gures 4.a,b). Assuming an equal number of background events in both samples, the �t gave Fs = (61 � 7)%. Background events came from fake vertices, whose lifetime distribution had an average value of zero and a Gaussian spread determined by the detector resolution, and from sec- ondary vertices originating from charm baryon and B meson decays (0flying background0 component). The latter component was predicted by the simulation to be (80 � 10)% of the background, both in the right and wrong sign pairs. Its average lifetime was deter- mined from the data using a larger sample of candidate decays reconstructed in the high pT muon events, as described in [3]. The b-baryon momentum was estimated from the total momentum ptot of the decaying particles using the residual energy technique. The residual energy was computed by subtracting the energy associated with the b-baryon candidate (the �, the muon and the pion energy) from the total energy associated with charged particles in the hemisphere 7 containing the � and the lepton, de�ned by the plane perpendicular to ptot. The b- baryon energy was estimated by subtracting this residual energy from the beam energy. The energy associated with all neutral particles in the hemisphere was by de�nition associated with the b-baryon by this method. The charged pions from the b-baryon decay chain may be wrongly included in the residual energy computation. As discussed in [3], the two e�ects nearly compensate, the correction factor computed in the simulation to reproduce the generated spectrum being on average 0:97 for unpolarized b-baryons. Sources of systematic error on this factor are the uncertainties on the b-baryon mass and polarization, its momentum spectrum and semi-leptonic decay modes. Their e�ect on the �nal lifetime result is listed in table 2. The resolution of the b-baryon momentum predicted by the simulation was 11%, as shown in �gure 5. The e�ect of the non-Gaussian tails of the distribution on the �nal result of the lifetime �t was found to be negligible (see below). A maximum likelihood �t was performed simultaneously to the lifetime distribution of the 63 events of the signal sample and to the one of the background vertices described above (300 events) with the likelihood function [3]: L = ��i ln[f(ti;�i;�;�bck)]; with f(ti;�i;�;�bck) = Fse (�2 i =2�2�ti=�) � erf � �i=� � ti=�i)= p 2 � =2� + (1 � Fs)�� Ffbe (�2 i =2�2 bck �ti=�bck) � erf � �i=�bck � ti=�i)= p 2 � =2�bck + Fnfe �t2 i =2�2 i � where � and �bck are the signal and background lifetimes; �i is the error on the measured decay time ti; the normalization constant Fs for the signal fraction was �xed to the �tted value of the b-baryon purity discussed above; �nally, Ffb was the normalization constant for the background fraction from B;D meson decays and Fnf = 1 �Ffb is the fraction of \non- ying" background. The three parameter �t to the 63 decays in the ���X channel, gave the result: �(b-baryon)= 1:12+0:30 �0:23 ps with a background lifetime �bck = 1:62 +0:14 �0:10 ps and Ffb = 0:79 � 0:03, in agreement with the simulation. The lifetime distributions for the signal events and for the background, together with the probability functions resulting from the �t, are shown in �gures 4.c,d. The uncertainties on the magnitude of the ying background and on its lifetime are accounted for in the statistical error of the �t result. The correlation matrix is shown in table 3, where the small anticorrelation between the signal and background lifetimes is quanti�ed. The di�erent contribution to the systematic uncertainty are listed in table 2. The �rst comes from the uncertainty on the sample composition, while the others a�ect the estimation of the b-baryon momentum. The assumed value of the average b- baryon mass, Mbar, was shifted with respect to the measured mass of the �b, M(�b) = 5640�50MeV=c2[1], to take into account the contribution to the observed decay channel of the production of �b particle (measured to be 5 times smaller than �b production [16]), whose mass is expected to be 250 � 50MeV/c2 higher than the �b mass. The same �tting procedure applied to the Monte Carlo simulation sample gave: �bck = 1:74 +0:10 �0:08 ps and �(b-baryon)= 1:52 +0:24 �0:14 ps, compatible with the generated av- erage b-baryon lifetime of 1.56 ps . In the simulation, di�erent samples of b-baryons were generated with average lifetimes varying in the range 0:75 � 2:25 ps and added in turn 8 Table 2: Contributions to the systematic error on the average b-baryon lifetime measured using �� correlations. Error source variation level Syst:error(ps) b�baryon purity 0:61 � 0:07 �0:04 �c decay mode uncertainty one st:dev: [12] �0:02 < Eb > 0:70 � 0:03 �0:01 Mbar 5670 � 70MeV �0:015 �b polarization �0:47 � 0:47 �0:01 �(!) = exp[aIW (1 � !)] aIW = 1:7+3:3�1:7 �0:01 BR(�b ! �c`�n�)=BR(�b ! �c`�) 0 ! 0:3 �0:06 total syst:error � +0:05 �0:08 Table 3: Correlation matrix between the variables of the lifetime �t. � � �bck Ffb � 1:00 �bck �0:12 1:00 Ffb �0:07 �0:18 1:00 to hadronic Z events in which all the other sources of ying background were kept with constant lifetimes. The number of b-baryons in the sample was chosen to reproduce the purity observed in the data. The response of the �tting procedure was linear, without any bias over the whole time interval considered. Summing the systematic uncertainties listed in table 2 in quadrature gives an overall systematic uncertainty of +0:05 �0:08 ps , much smaller than the statistical uncertainty from the �t. 5 �c` channel In this section a study of �b semileptonic decays using fully reconstructed �c is pre- sented, based on the data collected in the 1991 and 1992. Possible sources of �c ( ��c) `� (`+) in the same jet are �b semileptonic decays, B meson semileptonic decays and accidental correlations of a �c and a lepton. The �c` combinations from �b decays are characterized by higher invariant mass and higher transverse and longitudinal momentum of the lepton than the background pairs from accidental correlations. The contribution of the B meson semileptonic decay to a �c was estimated to be negligible, by an argument similar to that used in section 4. 5.1 � c selection The �c was reconstructed via the decay �c ! pK�. This is the most abundant decay mode but it is accompanied by a large combinatorial background. In order to enhance the signal, kinematic selection criteria on the �c candidates were optimized using the simulation. The �c was only accepted if the candidate's momentum was greater than 10 9 GeV/c and if the proton momentum was greater than the � momentum and also greater than 5 GeV/c. The protons and kaons were identi�ed by the RICH or by requiring that their dE=dx measurements be within 2 standard deviations of the expected values. In addition, all three tracks were required to have at least 2 hits in the VD, the �2 probability of the 3-prong �tted vertex was required to exceed 0:01 and the ight distance in the r� plane, LT, was required to be greater than 350�m. Figure 6 shows the pK� invariant mass distribution obtained. A �t to the pK� invariant mass distribution using a Gaussian distribution superimposed on a linear background yields a signal of 137 � 30 events. 5.2 � c ` correlations To improve the �c e�ciency in events with an identi�ed lepton, the cut described above on the ight distance of the �c candidate was relaxed, requiring only LT > 0: The �c candidates were paired with identi�ed leptons with momenta greater than 3 GeV/c within a cone of 45� around the �c direction. The lepton was required to have a pT greater than 0.6 GeV/c. The total momentum of the lepton and of the �c was required to be greater than 18 GeV/c and the invariant mass of the �c � (�c e) pair was required to exceed 3.5 GeV/c2 (3.3 GeV/c2). The M(pK�) invariant mass spectrum of �+c (� � c ) candidates associated with a `� (`+) in the same jet is shown in �gure 7a. A signal of 29:1�7:5 events (18:5�5:7 �c� and 10:6�4:4 �ce events) around the nominal �c mass is visible. No peak was found in the pK� mass distribution for �c candidates with a lepton of the same sign in the same jet (�gure 7b). The signal in �gure 7a was interpreted as coming from b-baryon! �cl�X decays. The contribution to the right sign sample from accidental combinations of a �c and a lepton and from �c-lepton pairs from B meson decay was estimated to be negligible. No contribution from the �c signal could be attributed to a kinematical re ection of a D + decaying into K�� or a D+s decaying into KK�. The simulation of the decay �b ! �c�� gives an overall e�ciency of selection and reconstruction of (7:2 � 0:6)% in the decay mode �c ! pK�. If one or more pions are produced in the �b semileptonic decays, the e�ciency becomes (3:07 � 0:26)% (assuming up to a maximum of 30% of decay modes with 1 or 2 pions, in equal amounts) due to the softer spectrum of the �c and of the �. This e�ect was included in the systematic uncertainties. Using the measured rate Br(�c ! pK�) = (4:4�0:6)% [12], this leads to a production rate: f(b! b-baryon)�Br(b-baryon! �c��X) = (1:19 � 0:34+0:31�0:21)%. The overall simulated reconstruction e�ciency of (4:6�0:6)% for the decays �b ! �ce� gives a production rate: f(b! b-baryon)�Br(b-baryon! �ce�X) = (1:15 � 0:44+0:31�0:21)%. Assuming lepton universality: f(b! b-baryon) � Br(b-baryon! �c`�X) = (1:18 � 0:26+0:31�0:21)%. Table 4 summarizes the di�erent contributions to the systematic error. 10 Table 4: Contributions to the total systematic uncertainty on the b-baryon production rate times the branching ratio to �c`�X. source of uncertainty variation level Syst:uncertainty(�102) �b sel: + rec: e�ciency (7:2 � 0:6)% �0:09 �c branching fraction (4:4 � 0:6)% �0:15 �b polarization �0:47 � 0:47 �0:09 �(!) = exp[aIW (1 � !)] aIW = 1:7+3:3�1:7 �0:06 Br(�b ! �c`�n�)=Br(�b ! �c`�) 0 ! 0:3 +0:23 total syst:uncertainty � +0:31 �0:21 5.3 Measurement of b-baryon lifetime In the �c`�X channel, b-baryon candidate vertices were reconstructed using the trajec- tories of the �c and the lepton to �t a common vertex. The �b momentum was estimated with the missing energy technique: E�b = Ebeam � Evisible + E�c + E` where Evisible was the sum of the energies of both charged and neutral particles in the same hemisphere as the �c. The quantity Ebeam�Evisible measured the neutrino energy in the b-baryon semileptonic decay (this was not true in the �` analysis, where the �c decay was not fully reconstructed), provided that only the 3-body �c`� decay mode was present. In this case, the simulation showed that the momentum used must be scaled by the factor 0:950 � 0:015, where the uncertainty was due to the �nite statistics available. If one or two additional pions were produced in the �b decay, the estimator gave a �b energy that was on average respectively 3.5 or 6 GeV too low, but this e�ect was reduced by the lower e�ciency of the many-� modes with respect to the 0-� mode. A sample of 28 signal vertices was selected using right sign �c` pairs with 2:260 < M(pK�) < 2:310 GeV/c2. The b-baryon purity of this sample was determined from a �t to the data to be (60 � 20)%. In a similar way, a sample of 139 background vertices was selected with wrong sign pairs with 2:085 < M(pK�) < 2:485 GeV/c2 and sideband right sign pairs (2:085 < M(pK�) < 2:240 GeV/c2 and 2:330 < M(pK�) < 2:485 GeV/c2). The reconstructed �c track and the lepton were �tted to a common secondary vertex (the b-baryon candidate decay vertex); the proper time distributions of the signal and background samples, shown in �gure 7c and 7.d respectively, were �tted with the same technique used for the study of the �` channel. The result is: �(b-baryon)= 1:33+0:71 �0:42 +0:08 �0:09 ps (�c`�X channel, 28 decays). with a ying background lifetime of 1:52+0:28 �0:21 ps; the correlation matrix of the �t parameters is shown in table 5. The �tted ying background fraction was 0:63 � 0:05. The di�erent contributions to the systematic error are shown in table 6. The e�ects of the �b polarization have been studied with the simulation and found to be negligible. 6 Muon-proton channel In the analysis of this channel, semileptonic decays of b-baryons were selected by the presence of a muon and a proton of high momenta and opposite charges in the same jet. About 500,000 hadronic events recorded in 1992 with the barrel gas RICH operational 11 Table 5: Correlation matrix between the variables of the lifetime �t in the �c` sample. � �bck Ffb � 1:00 �bck �0:14 1:00 Ffb �0:04 �0:27 1:00 Table 6: Contributions to the systematic error on the average b-baryon lifetime measured using �c` correlations. Error source variation level Syst:error(ps) b�baryon purity 0:60 � 0:20 �0:08 Monte Carlo statistics � �0:02 Mbar 5670 � 70MeV �0:015 Br(�b ! �c`�n�)=Br(�b ! �c`�) 0 ! 0:3 �0:04 total syst:error � +0:08 �0:09 were used. Proton selection used the measurement of the speci�c energy loss in the TPC (dE=dx) and the detection of Cherenkov photons in the RICH. The proton is thought to come predominantly from the chain decay b-baryon ! ����c-baryon, c-baryon ! pX. It is noted that the ight distance of the secondary charm baryon is, on average, much less than that of its parent, and that the fast proton follows its direction. To allow for a precise determination of the b-baryon decay vertex, which is essential for the present analysis, the proton and muon candidates were required to have at least two associated hits in the Vertex Detector. Detailed simulations showed that 70% of �b ! ����pX decays gave rise to a reconstructed three-dimensional �-p vertex. These vertices were distributed around the simulated �b decay vertex with a precision of �300 �m in the r� plane. The requirement of the detection of the proton in the VD and secondary muon- proton vertex reconstruction substantially reduced backgrounds due to tertiary protons: only (16 � 7)% of the signal was estimated to be protons from non-charmed hyperon decays in the b-baryon decay chain. This results in an overlap smaller than 5% between this sample and the �� sample discussed above. 6.1 Signal and background characteristics The signal muon-proton pairs have the following properties: the muon has hard mo- mentum (p�) and transverse momentum (pT ) spectra, the proton has a hard momentum (pp) spectrum, the muon and proton form a secondary vertex and they have opposite charge. The background is due to genuine protons which do not come from from b-baryon decays and to pions and kaons misidenti�ed as protons, as well as charged hadrons faking muons. The background involving genuine protons was almost completely eliminated by re- quiring the proton momentum to be above 8.5 GeV/c and the muon momentum above 4 GeV/c. The background involving fake protons is dominated by charged kaons. At low pT the muon-kaon pairs are predominantly of opposite charge whereas at high pT the background is mostly same sign pairs. This ip in the charge correlation of the back- ground involving kaons is caused by semileptonic b-hadron decays b ! c����� followed by 12 a c ! sX transition dominating at high pT and semileptonic decays of primary and sec- ondary charm hadrons c ! s�+�� dominating at low pT . Because of this, the procedure of removing events below a given transverse momentum of the muon and subtracting the wrong charge correlation, used in the analyses of �` and �c` channels, was not followed. Instead, a global �t (see section 6.3) to the muon pT spectrum, the hadron dE=dx distribution and the proper time distribution of reconstructed muon-hadron vertices was applied to the separate samples of muon-hadron pairs enriched in protons, kaons and pions simultaneously. These samples were obtained with the use of the RICH as explained in the next section. In this way the yield of b-baryon signal and its average lifetimewas extracted using all charged hadron identi�cation information and minimizing the dependence on the simulation. 6.2 Sample De�nition 6.2.1 Hadron Identi�cation Hadronsy were selected in a momentum range where energetic kaons and protons could be separated by the gas radiator of the RICH, namely p � 8:5 GeV/c. In this range, the expected mean number of Cherenkov photons detected for a kaon by the RICH was greater than 1.5. Protons up to 16 GeV/c are below the Cherenkov threshold. K/p separation was e�ective up to 30 GeV/c and covered most of the high momentum part of the spectrum of the signal protons. Using the information provided by the RICH, four separate samples of energetic charged hadrons were de�ned: � the proton sample. This contained tracks whose proton hypothesis probability ex- ceeded 90%. This cut suppressed kaons and pions su�ciently to make the p:K:� ratio approximately 1:1:1. � the kaon sample. This contained tracks whose kaon hypothesis probability exceeded 80%. This cut removed all protons and gave a K/� ratio greater than 2. � the pion sample. This required that the pion hypothesis probability exceeded 25% and that more than 5 Cherenkov photons were compatible with the pion hypothesis. All protons and kaons in this data set were suppressed by this cut. � the unresolved hadron sample taking all tracks not accepted in the previous three samples. The composition of these samples was determined using dE=dx measurement from the TPC. In the momentum range above 8.5 GeV/c pions, kaons and protons are on the relativistic rise of the dE=dx. The mean values of their energy loss di�er by approximately constant amounts from � 4 GeV/c up to � 25 GeV/c. Requiring at least 30 hit TPC wires to analyse a track, the ratios dE dx =Tj(p) of the measured mean energy loss to the momentum dependent theoretical values Tj(p) (j = p,K,�) have Gaussian distributions with a common precision of �7%. The consistency between the theoretical and observed speci�c ionization was checked on the four samples described above. This ensured a very good parameterization of the speci�c ionization measurement, independent of the simulation. 6.2.2 Muon-Hadron Selection The selection procedure consisted of three sets of cuts, which will be referred to in the determination of the selection e�ciency (section 6.4): y In what follows hadron stands for a charged particle not identi�ed as a muon. 13 1. Event and muon selection: In addition to the hadronic event selection described in section 3 a successfully reconstructed primary vertex was required, formed by at least three charged tracks with the �2 probability of the vertex �t greater then 1%. The muon candidate selection (section 3.2) was complemented by the requirement that the muon candidate had at least two associated hits in the Vertex Detector and a momentum above 4 GeV/c. These cuts de�ned the event sample used for the determination of the number of muons from b decay (see section 6.4) and to which the muon-hadron vertex search was applied. 2. Hadron track quality cuts: hadrons were accepted when the information from the RICH was available for the hadron track, when the hadron track had at least two associated hits in the Vertex Detector, and more than 30 wires used for the dE=dx measurement. 3. Muon-hadron vertex de�nition: muon-hadron pairs were accepted when the hadron had a momentum above 8.5 GeV/c, when the muon and the hadron were in the same jet, when the muon-hadron secondary vertex had a probability greater than 1%, and the error on the distance �V between the primary and the secondary vertices was smaller than 1mm. Combining these three sets of cuts with the RICH selection described in the previous section, four samples of muon-hadron pairs were obtained: the muon-proton sample (�p) and the muon-kaon (�K), muon-pion (��) and muon-unresolved (�X) control samplesz. 6.3 b-baryon Lifetime 6.3.1 Global Fit Procedure A maximum likelihood �t was used to estimate the number of muon-proton pairs from b-baryon decays and the average lifetime of b-baryons. For each �-hadron event, the dE=dx, the signed muon transverse momentum p (S) T = S � pT (where S = +1 for the right sign and S = �1 for the wrong sign correlation), and t = �V =(pbar=Mbar), where �V is the distance of the �-hadron vertex from the primary vertex, were considered as a set of three independent measurements. The last quantity estimated the b-baryon proper time assuming the event belongs to the signal. To compute it, the b-baryon momentum pbar was evaluated using a linear relationship with respect to j~p� + ~ppj obtained in the simulation (� �16% accuracy at 13 GeV/c and � �6% at 35 GeV/c). Six classes of events were distinguished: (1) the signal, the backgrounds involving (2) protons, (3) kaons from b-hadron decays, (4) other kaons, (5) pions from b-hadron decays and (6) other pions. Each class had its own probability density function (pdf ) being the product of the three pdf 's associated to each of the measured quantities: P(p(S)T ; dEdx ;tjclass) = P?(p (S) T jclass) � PdE dx (dE dx ;�dE dx jclass) � Pt(t;�tjclass) The P? probability density functions were taken from the simulation. In this pdf the distinction between di�erent 'kaon' classes (3 and 4) and 'pion' classes (5 and 6) were preserved to allow for variations in the muon transverse momentum distributions resulting from the two components of the backgrounds of kaons and pions. z In the following, the notation �i will be used to refer generically to one of these four samples. 14 The PdE dx probability density function was taken to be: PdE dx (dE dx ;�dE dx jj) = 1p 2��dE dx exp 0 @�( dE dx =Tj � 1)2 2�dE dx 2 1 A where Tj were momentumdependent theoretical mean values of the dE=dx for the hadron from the class j. The signal Pt probability density function was parameterized as a convolution of an exponential decay probability density function of mean ��p and a Gaussian resolution function. The kaon and pion background Pt probability density functions were taken as linear combinations of a ying part (fraction fBGD(K or �) described by a convolution of an exponential decay of e�ective lifetime �BGD(K or �), and a resolution function) and a 'non ying' part (fraction 1-fBGD described by a resolution function alone). These four parameters, �BGD(K), �BGD(�), fBGD(K) and fBGD(�), were determined by the �t. For the Pt probability density function of the proton background two extreme parameterizations were used: the pion one and a Gaussian parameterization. The �nal results were obtained by averaging the results of the �ts performed with these two parameterizations of the proton background Pt pdf , taking half of the di�erence as a contribution to the systematic error. The following negative log-likelihood function was minimized by the �t: L = � X �i N�iX n=1 ln 0 @ 6X class=1 F(classj�i)P([p(S)T ; dEdx ;t]njclass) ; 1 A where N�i was the population of sample �i and F(classj�i) was the fraction of events in sample �i coming from the given class. The 24 composition parameters F(classj�i) were constrained by four normalization conditions (one for each sample): P 6 j=1 F(jj�i) = 1. Moreover, the relative contents F(jj�i)=F(j + 1j�i) of the proton classes (j = 1), the kaon classes (j = 3) and the pion classes (j = 5) were the same in each sample. This left 11 independent fractions to be determined by the �t. The proton content in the three control samples and the kaon content in the �� sample were found by the �t to be compatible with zero and were �xed to zero in the �nal �t, leaving only seven composition parameters to be determined. The systematic e�ect introduced by this assumption was taken into account in the contributions from the background composition. 6.3.2 Results of the Fit The �t was performed with 125 events of the �p sample, 243 events of the �K sample, 295 events of the �� sample and 369 events of the �X sample. The projections of the �t space onto the p (S) T , � = ( dE dx =Tp � 1)=�dE dx and t axes are shown in �gures 8, 9, and 10 respectively. The purity of the signal can be read from �gure 11 where additional cuts on p (S) T > 0:7GeV/c and � < 1:5 were applied to the �p sample. The number of signal events present in the muon-proton sample was estimated to be N(�p from b-baryon) = 28:9+6:7 �6:2 +1:8 �2:6 +0:9 �0:4 : The average lifetime of b-baryons was estimated to be ��p = n 1:27+0:35 �0:29 � 0:09(syst:exp:) � 0:02(syst:theory) o ps : The �rst systematic error was due to the measurement procedure, whereas the second represents the in uence of unknown b-baryon properties. 15 The estimates of the seven composition variables chosen as �t parameters, together with the �ve variables involved in the lifetime part of the likelihood function are reported in table 7. The correlation matrix for the variable parameters is given in table 8. The \compo- sition" parameters (P1 � P7) and the \lifetime" parameters (P8 � P12) are practically uncorrelated. There was no parameter correlated to the mean b-baryon lifetime (P8) by more than �12%. Table 7: The result of the maximum likelihood �t of the average b-baryon lifetime and the composition of the selected samples. The �rst error quoted comes from the �t, the second is half the di�erence between the results corresponding to the two di�erent proton background parameterizations. Parameter Result P1: fraction of signal in the �p sample 0.233 + � 0:054 0:050 � 0.002 P2: ratio (signal)/(all p) 0.75 � 0.15 � 0.02 P3: ratio (K from b)/(all K) 0.557 + � 0:053 0:051 � 0.001 P4: ratio (� from b)/(all �) 0.453 + � 0:057 0:056 � 0.001 P5: fraction of kaons in the �p sample 0.354 + � 0:105 0:095 � 0.009 P6: fraction of kaons in the �K sample 0.858 + � 0:039 0:038 � 0.001 P7: fraction of kaons in the �X sample 0.467 + � 0:044 0:042 � 0.001 P8: average lifetime of b-baryon��p 1.27 + � 0:35 0:29 � 0.03 ps P9: �BGD(K) 1.51 + � 0:29 0:25 � 0.002 ps P10: �BGD(�) 1.84 + � 0:19 0:17 � 0.001 ps P11: fBGD(K) 0.64 � 0.09 � 0.001 P12: fBGD(�) 0.731 + � 0:047 0:050 � 0.000 6.3.3 Fit Systematics The parameters P3 and P4 describe the relative amount of true kaons and pions arising from b-hadron decays among all kaons and pions. To examine relevant systematic e�ects, three approaches were taken: (1) three di�erent de�nitions of these parameters were used (a) K (�) from b all K (�) , (b) K (�) and � from b all K (�) , (c) K (�) and direct � from b all K (�) ; (2) these fractions were �xed to the Monte-Carlo prediction; (3) �K or �X samples were excluded from the �t. The maximal variation of the �t results was taken as a contribution (\K, � bkg composition" in table 9). To evaluate possible systematics related to the parameterization of the P? probability density function of the proton background class, this class was divided into four groups characterized by very di�erent p (S) T spectra of the accompanying muon: (1a) right sign muons from b-hadron decays, (1b) wrong sign muons from b-hadron decays, (2a) other right sign muon candidates (2b) other wrong sign muon candidates. From the set of these four groups, 14 non-trivial subsets can be chosen (4 containing one group, 6 containing two groups and 4 containing three groups). The �t was performed 14 times with the P? probability density function of the proton background sample determined after the chosen subset was scaled up by a factor of 2. The maximal variation was taken as an estimate of the systematic e�ects (\p bkg composition" in table 9). The results quoted were obtained with pT calculated including the muon candidate in the jet. To evaluate systematic errors, pT was replaced (1) by pT out calculated excluding 16 Table 8: Correlation matrix for the �t parameters. The de�nitions of the parameters are given in table 7. P1 .40 P2 .06 .05 P3 -.04 -.01 -.03 P4 -.19 .62 .00 .03 P5 -.01 -.01 -.11 .08 .00 P6 -.01 -.01 -.13 .07 .00 .02 P7 -.01 .09 .00 .01 .09 -.00 -.01 P8 .02 -.01 -.01 -.01 -.00 .01 .03 -.12 P9 .01 .01 -.00 -.01 .02 .00 .01 -.02 -.12 P10 -.05 -.03 -.02 -.02 .04 -.00 .05 -.04 -.22 0.05 P11 -.01 .01 .00 .01 .00 .02 -.01 -.00 .02 -.20 -.22 P12 the muon candidate from the jet and (2) by the quadratic sum q (pT 2 + (p�=10)2). All three de�nitions were tested with several binnings. The maximal variation was taken as a contribution to the systematic error (\pT binning/de�nition" in table 9). In the likelihood function, the Pt probability density function was used only for the right sign muon-proton sample, and optionally for the part above some pT cut. Outside this sample Pt pdf's were �xed to a constant value for all classes. The result was found to be stable within 2% in the range of pT cut from 0 GeV/c (no cut) to 0.7 GeV/cas can be seen in the �gure 12. For higher cut values, the b-baryon lifetime begins to uctuate within increasing statistical error. The �t procedure was tested in the following way. From the available statistics of signal muon-proton pairs in simulated b-baryon decays passing all the selection cuts, di�erent sets of 28 pairs each were randomly chosen; from each of them a larger test sample of muon-hadron pairs was formed by adding a number of muon-unresolved hadron pairs randomly chosen from real data, in such a way as to reproduce in the test sample the signal fraction 0.23 observed in the data. The generated lifetime of the b-baryon in the simulation was 1.3 ps. The whole �t was repeated several times with the data muon- proton sample replaced by one of the test samples described above. The distributions of the �t result for the proton signal fraction P1 and for the average b-baryon lifetime P8 are shown in �gures 13a,b respectively; their average values reproduced the known input values of the parameters, with a spread in agreement with the average �t error. 6.4 Branching Ratio The number of signal events found by the �t was used for the calculation of the following production rate: f(b ! b-baryon) � BR(b-baryon ! p����X) = (0:49 � 0:11 � 0:06+0:14�0:09)%. The �rst systematic error is due to the measurement procedure, whereas the second represents the in uence of unknown b-baryon properties. Systematic e�ects are sumarized in table 10. The total experimental systematic error results from the following sources listed in the table: � N(�p from b-baryon) is the number of signal events found in the previous section. 17 Table 9: Systematic uncertainties in the �t. source of variation variation level resulting variation of N(�p from b-baryon) ��p [events] [ps] Experimental systematics dE=dx normalization one stand. dev. �1:2 �0:03 pT binning/de�nition see text +1:2 �2:3 �0:08 K, � bkg composition see text �0:2 �0:01 p bkg composition see text +0:6 �0:3 �0:01 p background Pt pdf see text �0:3 �0:03 boost estimate one stand. dev. | �0:03 total systematic error (measurement) +1:8�2:6 �0:09 Systematic uncertainty due to unknown b-baryon properties b-baryon polarization �0:47 � 0:47 �0:3 �0:01 �b ! �c���� decay form factor �(!) = exp[aIW (1 � !)] aIW = 1:7+3:3�1:7 �0:1 �0:01 hE(b-baryon)i=E(beam) 0:70 � 0:03 �0:1 �0:01 BR(�b ! �c`��)=BR(�b ! `��X) 1:0 ! 0:7 +0.9 +0.02 total systematic uncertainty (theory) +0:9 �0:4 �0:02 Table 10: Contributions to the total systematic uncertainty of the b-baryon production rate times its branching ratio into p�X. (For de�nitions of the e�ciencies �1, �2, �3, �R and the correction C� see text.) quantity value contribution �102 Experimental systematics N(�p from b-baryon) 28.9 +1:8 �2:6(syst.) +0:03 �0:04 �1 0.376 �0:011 �0:02 �2 0.308 �0:009 �0:02 �3 0.195 �0:012 �0:03 �R 1.0 �0:06 +0:03 Total Systematic Uncertainty (measurement) �0:06 Systematic uncertainty due to unknown b-baryon properties C� 0.84 �0:07 �0:04 b-baryon polarization �0:47 �0:47 �0:07 �b ! �c���� decay form factor �(!) = exp[aIW (1 � !)] aIW = 1:7 +3:3�1:7 +0:09�0:03 hE(b-baryon)i=E(beam) 0.70 �0:03 �0:03 BR(�b ! �c`��)=BR(�b ! `��X) 1:0 ! 0:7 +0:06 Total Systematic uncertainty (theory) +0:14 �0:09 18 � �1 is the e�ciency of the \event and muon" selection (the �rst item in the Section 6.2.2). � �2 is the e�ciency of the hadron track quality cuts (the second item in the Section 6.2.2). This e�ciency was found in the data. � �3 is the e�ciency of the additional selection de�ned in the third item of the Section 6.2.2. This e�ciency was found using simulation. � �R is the e�ciency of the selection of the �p sample with the RICH. This e�ciency was found by the �t (before �xing to zero proton contents in the control samples). Entries for the theoretical systematics are similar to those described in the analysis of the �` channel. C� is the correction due to the residual presence of protons from the chain decay b-baryon! c-baryon!hyperon!proton. 7 Conclusions The production and lifetime of the b-baryon has been studied with three di�erent and complementary methods, relying on the detection of a fast �, a �c and a fast proton in the same jet as a high pT lepton. The following semi-exclusive branching ratios have been measured: f(b ! b-baryon) � BR(b-baryon ! �`��`X) = (0:30 � 0:06 � 0:04)%, f(b ! b-baryon) � BR(b-baryon ! �c`��`X) = (1:18 � 0:26+0:31�0:21)%, f(b ! b-baryon) � BR(b-baryon ! p����X) = (0:49 � 0:11+0:15�0:11)%. From partially reconstructed b-baryon decay candidates in these three di�erent semi- leptonic channels, the following values for the average b-baryon lifetime have been mea- sured: �(b-baryon) = 1:12+0:30 �0:23 +0:05 �0:08 ps (63 decays, �����X channel), �(b-baryon) = 1:33+0:71 �0:42 +0:08 �0:09 ps (28 decays, �c`���X channel), �(b-baryon) = 1:27+0:35�0:29 � 0:09 ps (47 decays, p����X channel). The above lifetime determinations rely on completely independent event samples. This was checked on an event by event basis for the �-proton and �-�c samples, where a small overlap could not be excluded a priori by the selection criteria discussed above. The overlap between the � � � and the �-proton samples was found negligible by the simulation, as discussed in section 6. The common systematics, due to the modelling of the b-baryon production and decay properties, can be inferred from tables 2, 6 and 7. Averaging the three results, under the assumption that the di�erent b-baryon species enter in the same proportion in the decay channels considered (all of them are expected in fact to be largely dominated by the �b baryon), gives the mean b-baryon lifetime: �(b-baryon) = 1:21+0:21�0:18 � 0:04(exp:syst:)+:02�:07(th:syst:) ps. Acknowledgements We are greatly indebted to our technical collaborators and to the funding agencies for their support in building and operating the DELPHI detector, and to the members of the CERN-SL Division for the excellent performance of the LEP collider. 19 References [1] UA1 Collaboration, C. Albajar et al., Phys. Lett. B273 (1992) 540. [2] ALEPH Collaboration, D. Decamp et al., Phys. Lett. B278 (1992) 209. OPAL Collaboration, P. D. Acton et al., Phys. Lett. B281 (1992) 394. [3] DELPHI Collaboration, P. Abreu et al., Phys. Lett. B311 (1993) 379. [4] ALEPH Collaboration, D. Buskulic et al., Phys. Lett. B297 (1992) 449; OPAL Collaboration, R. Akers et al., Phys. Lett. B316 (1993) 435. [5] I.I. Bigi and N.G. Uraltsev, Phys. Lett. B280 (1992) 271; G. Altarelli and S. Petrarca, Phys. Lett. B261 (1991) 303. [6] DELPHI Collaboration, P. Aarnio et al., Nucl. Instr. Meth. A303 (1991) 233. [7] E.G. Anassontzis et al., Nucl. Instr. Meth. A323 (1992) 351. [8] DELPHI Collaboration, P. Abreu et al., Nucl. Phys. B418 (1994) 403. [9] C. Kreuter, Ph.D Thesis, Karlsruhe University, IEKP-KA/93-9. [10] W. Adam et al., \Analysis techniques for the DELPHI RICH", Contributed paper GLS0188, to the 27th International Conference on High Energy Physics, Glasgow 1994. [11] T. Sjostrand et al., Comput.Phys.Commun. 39 (1986) 346; 43 (1987) 347. [12] The Particle Data Group, Phys. Rev. D50 (1994) 1173. [13] CLEO Collab., G. Crawford et al., Phys. Rev. D45 (1992) 752. [14] T. Sj�ostrand, Comp. Phys. Comm. 82 (1994) 74. [15] N. Isgur and M.B. Wise, Phys. Lett. B232 (1989), 113; N. Isgur and M.B. Wise, Phys. Lett. B237 (1990), 527. [16] DELPHI Collaboration, \Production of strange B-baryons decaying into �� � `� pairs at LEP", CERN-PPE/95{29, to be published in Zeit. f. Phys. C. 20 Figure 1: a) p�� invariant mass distribution for � candidates with p > 4GeV=c; the curve is the result of a �t using a Breit-Wigner function, which takes into account the variation of the mass resolution with the momentum of the decaying tracks, and a polynomial background. b) Background subtracted � momentum spectrum (dots: data; histogram: Monte Carlo simulation) ; c) � ! p�� reconstruction e�ciency computed in the simula- tion. 21 Figure 2: Distribution of p� invariant mass for �candidates correlated to high pT leptons in the same jet: a) right-sign pairs; b) wrong-sign pairs. The data are shown by the points; the simulation, normalised to the total number of hadronic Z decays, as a histogram: the contribution from b-baryon decay is shown double-hatched, while the background from fragmentation �'s is shown single-hatched. The curves show the result of the �t described in the text. 22 Figure 3: Subtracted spectrum (right sign - wrong sign) in the data (dots) for: a) � momentum, b) lepton transverse momentum, c) the sum of the � and lepton mo- menta and d) �` invariant mass. The histograms show the simulation for the b-baryon signal. 23 Figure 4: � signal for reconstructed ��� vertices of a) right sign and b) wrong sign respectively; c) lifetimedistribution for 63 b-baryon candidates (hatched area in the �mass plot); the full lines represent the result of the �t described in the text; the dotted-dashed line is the estimated b-baryon contribution, the dashed and dotted lines represent the ying and not- ying background respectively, determined from d) the lifetimedistribution of the background sample, ��� vertices with wrong sign or p� mass outside the above range. 24 Figure 5: Ratio between estimated and generated �b momentum predicted by the simu- lation for the decay channel �b ! �c��. The curve is the result of a Gaussian �t to the distribution. 25 Figure 6: �c inclusive signal for reconstructed pK� vertices. The curve is the result of a �t to the distribution using a Gaussian superimposed on a linear background. 26 Figure 7: a,b) pK� invariant mass distribution for �c` pairs of opposite sign and same sign respectively; c,d) proper time distributions for b-baryon signal and background sample. The curves are as in �gure 4. 27 Figure 8: Projection of the data distribution onto the p (S) T axis (where p (S) T is signed transverse mo- mentum of the muon. Its positive values corre- spond to the right sign combination (muon and hadron have opposite charges), whereas negative values to the wrong sign one (same sign �{hadron pairs). Points with error bars (data) are com- pared to the �t (uppermost curve) decomposed into six classes shown with di�erent hatching. The four plots shown correspond to the four sam- ples used in the �t. 28 Figure 9: Projection of the data distribution onto the � axis. Points with error bars (data) are compared to the �t represented by the uppermost curve. This curve is the sum of the p, K, � contributions, shown with Gaussians centered at � = 0:0; 1:1; 3:05 respectively. The four plots correspond to the four samples used in the �t: a) �p sample (the signal content is hatched); b) �K sample; c) �� sample; d) �X sample. 29 Figure 10: Projection of the data distribution onto the proper time axis for p (S) T > 0:7GeV/c. The data are represented by points with the error bars, the �t is shown with up-most continuous lines: a) �p sample { the signal is shown double hatched; b) �K sample { the kaon content is shown single hatched; c) �� sample; d) �X { the kaon content is shown single hatched. 30 Figure 11: Projections of the �p sample onto (a) p (S) T axis with a cut � < 1:5, (b) � axis with a cut p (S) T > 0:7GeV/c and (c) proper time axis with both cuts applied. The signal content is shown double hatched (the hatching on the �gure (a) is explained in the caption to the �gure 8). 31 muon transverse momentum cut (GeV/c) b -b a ry o n l if e ti m e ( p s ) 0 0.6 1.2 1.8 2.4 3 0 0.35 0.7 1.05 1.4 Figure 12: Study of the stability of the b-baryon lifetime determination with respect to a given muon transverse momentum cut. The width of the double hatched area shows the uncertainty due to parameterization of the proton background Pt e�ective lifetime (the lower border was obtained with the pion parameterization, the uper one with zero e�ective lifetime). Vertical bars shows the symmetric error of the �t. 32 Figure 13: Results of the toy Monte Carlo simulation described in the text { points with error bars { are �tted with Gaussians for a) estimated signal yield (input value = 28, output mean value = 26, RMS = 6); b) estimated b-baryon lifetime (input value = 1.3 ps, output mean value = 1.29 ps, RMS = 0.24 ps).