K_s⁰ and Λ production in quark and gluon jets at LEP

Article

Reference

K

and Λ production in quark and gluon jets at LEP

L3 Collaboration

AMBROSI, Giovanni (Collab.), et al.

Abstract

A measurement of the Ks0 and Λ inclusive production rates and momentum spectra in two- and three-jet events is presented. On the basis of about 3.1 million Z decays collected with the L3 detector at LEP, we observe that the production of these particles is well modelled by string fragmentation.

L3 Collaboration, AMBROSI, Giovanni (Collab.), et al . K

and Λ production in quark and gluon jets at LEP. Physics Letters. B , 1997, vol. 407, p. 389-401

DOI : 10.1016/S0370-2693(97)00864-2

Available at:

http://archive-ouverte.unige.ch/unige:113058

Disclaimer: layout of this document may differ from the published version.

1 / 1

ELSEVIER

4 September 1997

PHYSICS LETTERS B

Physics Letters B 407 (1997) 389-401

KS0 and A production in quark and gluon jets at LEP

L3 Collaboration

M. Acciarri ab, 0. Adrianiq, M. Aguilar-Benitez”, S. Ahlen k, J. Alcarazaa, G. Alemanni”, J. Allaby

A. Aloisioad, G. Alverson !, M.G. Alviggi ad, G. Ambrosi t, H. Anderhubay,

VP Andreev am, T. Angelescu m, F. Anselmo i, A. ArefievaC, T. AzemoonC, T. Azizj, P. Bagnaiad, L. Baksay as, R.C. Ball ‘, S. Banerjeej, SW. Banerjeej, K. BaniczaU, A. Barczyk ay*w, R. Barillere r, L. Barone at, P. Bartalini”‘, A. Baschirottoab, M. Basile i, R. Battistonai, A. Bay w, F. Becattiniq, U. Beckerr, F. Behner ay, J. Berdugo I, P. Berges p,

B. Bertucci”‘, B.L. Betev aY, S. Bhattacharyaj, M. Biasini’, A. Bilanday, G.M. Bilei ‘, J.J. Blaisingd, S.C. Blythq, G.J. Bobbink b, R. Bock a, A. BGhm a, L. Boldizsar n,

B. Borgiaae, A. Bouchamd, D. Bourilkovay, M. Bourquin ‘, D.

d, S. Braccini t, J-G. Branson ao, V. Brigljevic aY, I.C. Brock 4, A. Buffini 9, A. Buijs @, J.D. Burger P, W.J. Burger t, J. Busenitz =, X.D. Cai P, M. C~p~elli ‘Y, M. Cape11 P, G. Cara Romeo i,

G. Carlino ad, A.M. Cartacci 4, J. Casaus aa, G. Castellini 9, F. Cavallari ae, N. Cavallo ad, C. Cecchi t, M. Cerrada”“, F. Cesaroni ‘, M. Chamizo =, Y.H. Chaug ba, U.K. Chaturvedi s,

S.V. Chekanov af, M. Chemarin ‘, A. Chen ba, G. Chen g, G.M. Chen s, H.F. Chen “, H.S.

g, M. Chen P, G. Chiefari ad, C.Y. Chien e, L. Cifarelli an, F. Cindolo i, C. Civininiq, I. Clare P, R. Clare P, H.O. Cohn ag, G. Coignet d, A.P. Colijn b, N. Colino aa, V, Commichau”, S. Costantini h, E Cotorobai m, B. de la Cruz aa, A. Csilling “, T.S. Dai P,

R.

Alessandro q, R. de Asmundis ad, A. Degred, K. Deiters av, P. Denes &, E DeNotaristefaniae, D. DiBitonto *, M. Diemoz ae, D. van Dierendonck b, F. Di Lodovico”Y, C. Dionisia’, M. DittmaraY, A. Dominguezao, A. Doriaad, I. Domed, M.T. Dova ‘p4, J. Geralde, N. Gheord~escu m, S. Giagu &, S. Goldfarbw, J, Goldstein k,

Z.F. Gong”, A. Gougase, G. Grattaah, M.W. Gruenewaldh, V.K. GuptaBk, A. Gurtuj, L.J. Gutay au, B. Hartmann a, A. Hasan “, D. Hatzifotiadou i, T. Hebbeker h, A. HerveT,

W.C. van Hoek af, H. Hofer ‘Y, S.J. Hong =, H. Hoorani G, S.R. Hou ba, G. Hu e, V. Innocente r, H. Janssen d, K. Jenkes ‘, B.N. Jin s, L.W. Jones c, P. de Jong r, I. Josa-Mutuberria”“, A. Kasser w, R.A. Khan ‘, D. Kamradaw, Yu. Kamyshkovas, J.S. Kapustinsky y, Y. Karyotakisd, M. KaurS,5, M.N. Kienzle-Focacci’, D. Kim&,

D.H. Kima, J.K. Kim”‘, S.C. Kimm, Y.G. Kim=, W.W. Kinnisony, A. Kirkby&, D. Kirkby ah, J. Kirkby’, D. Kiss “, W. Kittel af, A. Klimentov Pvac, A.C. K&rig af, A. Kopp aw,

0370-2693/97/$17.~ 6 1997 Eisevier Science B.V. All rights reserved PII SO370-2693(97)00844-2

390 w C#~~~~~~~fionfP~~sics Lemm 3 407 (19971 x39-&?I

I.

Korolko”, V. Koutsenko p*‘c, R.W. Kraemer d, W. Krenz a, A. Kunin Pyac, P Ladron de Guevaram, G. Landi 4, C. Lapoint P, IS. Lassila-Perini ay, P. ~urik~nen “,

M. Lebeau ‘, A. Lebedev P, P. Lebrun ‘, P. Lecomte Q’, P. Lecoq r, P. Le Coultre &Y, C. Leggett ‘, J.M. Le Goff F, R. Leiste aw, E. Leonardi al, P. Levtchenkonm, C. Li Us C.H. Lin ba, W.T. Lin hn, EL. Linde b,r, L. ListaadLd, Z.A. Liu s, W, Lohmannaw, E. Longoti,

W. Lu ah, YS. Lu 8, K, Ltibelsmeyer n, C. Luci &, D.

p, L. Luminari&,

W. Lustermann “, W.G. Ma”, M. Maityj, G. Majumderj, L. Malgeri &&, A, Malinin zLc, C. Mafia=, D. Mange01 af, S. Manglaj, P. Marchesini ay, A. Marin k, J.P. MartinZ, F. Marzanoa’, G.G.G. Massaro b, D. McNally ‘* S. Melead, L. Merulaad, M. Mes~hi~i~, W.J, Metzger af, M. von der Mey a, Y. Mi w, A. Mihul m, A.J.W. van Mil af, G. Mirabelli &, J. Mnich r, P. Molnar h, B. Monteleoni st R, Moorec, S. Morganti “, T. Moulikj, R. Mount ah9

S. Muller a7 F. Muheim’, A.J.M. Muijs b, S. Nahnp, M. Napolit~o~d, E Nessi-Tedald?, H. Newman ah, T. Niessen G, A, Nippea, A. Nisatia’, H. Nowakaw, Y.D, Oh=, H. Spitz”, G. Organtini af, R. Ost~~~n ‘, C. Palomares aa, D. Pando~las a, S. Pa&em&, P Paolucci ad,

H.K. Parki, I.H. Park”‘, 6. Pascale “, G, Passalevaq, S. Patricelliad, T, Paul [, M. Pauluzzi”‘, C. Paus a7 F. Pauss

Peach’, Y.J. Peia, S. Pensottiab, D. Perret-GaIlixd, B. Petersen nf, S. Petrak h, A. Pevsner e, D. Piccolo ad, M. Pieri 9, J.C. Pinto 4, PA. Piroue ak,

E. Pistolesi ab, V. Plyaskin”‘, M. Pohl ay, V. PojidaevaC*q, H, Posternap, N. Produitt, D. Prokofiev am, G. Rahal-Callot ay, N. Rajaj, PG. Rancoita&, M. Rattaggiab, G. Raven ‘O,

P. Razis ae, K, Readaa, D. Ren aYzyt M. Rescigno &, S. Reucroft”, T. van Rheeat, S. Riemann aw, K. Riles ‘_ 0. Rind ‘, A. Robohm BY, J. Rodin r, BP. Roe c, L. Romero aa9 S. Rosier-Lees d, Ph. Rosselet w, W. van Rossuma’, S. Roth”, J.A. Rubio’, D, Ruschmeier ‘>

H. Rykaczewski aY, J, Salicio r, E. Sancheza”, MI? Sandersaf, M.E. Sarakinos “, S. Sarkarj, M. Sassowsky 8, G. Sauvage d, C, Schsfer ‘, V. Schegelsky am, S. Schmidt-Kaerst a, D. Schmitz a, P. Schmitz a, M, Schneegans d, N. Scholz ay, H. Schoppera2, D.J. S~hot~~s af,

J, Schwenke”, G. Schwering a, C. Sciaccaad, D. Sciarrino’, L. Servoliai, S. Shevchenkoah, N, Shivarov”q, V. ShoutkoaC, J, Shuklay, E. ShumilovaC, A. Shvorobah, T. Siedenburga,

D. Sonw, A. Sopczakaw, V Soulimovad, B. Smithp, P Spillantiniq, M. Steuerp, DP, Stickland&, H, Stoneak, B. StoyanovaQ, A. Straessnera, K. Strauch’, K. Sudhakarj,

G. SultanovS, L.Z. Sun uI G.F. Susinno t, H. Suter”Y, J.D. Swains, X.W. Tang g, L, Tauscher f, L. Taylor !, Samuel CC. Tingr, S-M, Ting P, M. Tonu#ti ‘, SC. Tonwarj, J_ T&h n, C. Tully ak, H, Tuchscherer”, K.L. Tung s, Y. Uchidap, J. Ulbrichtay, U. User “,

E. Valente af, R.T. Van de Walleaf, G. Vesztergambi ‘, I. Vetlitsky ac, G. Viertelay, M. Vivargentd, R. Vlifkert aw, H. VogelaJ, H. Vogt aw, 1. Vorobieva’, A.A. Vorobyovam, A, Vorvolakos “, M. Wadhwaf, W. Wallraff a, J.C. Wangp, X.L. Wag”, Z.M. Wang’, A. Weber IL, F. Wittgenstein ‘, S.X. Wus, S. Wynhoff”, J. Xuk, Z.Z. Xu”, B.Z. Yang’, C.G. Yanga, X.71; Yao

J B. Ye “, SC. Yeh ba, J.M. Youq, An. Zalite ‘*, Yu. Zalite am,

P Zemp ay, Y.

Zhang g, Z.P. Zhang uI B. Zhou k, Y. Zhou ‘, G.Y. Zhu &t

R.Y. Zhu ah, A. Zichichi i,r,s, F. Ziegler aw

L3 CoiZaboration /Physics Letters B 407 (I 9971389-401 391

il I. Physikalisches Institut, RWTH, D-52056 Aachen, FRG ’ III. Physikalisches Institut, RWTH, D-52056 Aachen, FRG ’

b National Institute for High Energy Physics. NIKHEF and University of Amsterdam, NL-1009 DB Amsterdam, The Netherlands E University of Michigan. Ann Arbor MI 48109, USA

d ~boratoire d’Anne~y-ie-~eux de Physique des Pa~i~uies~ LAP~INZF3-~NRS, BP 110, F-74941 Anne~y-ie-beg CEDEX, France c Johns Hopkins Universal, 3aitimore, MD 21218, USA

I Institute of Physics, University of Basei, CH-4056 Basei, Switzerland f Institute of High Energy Physics, IHEP 100039 Beijing, China6

h Humboldt University, D-10099 Berlin, FRG ’

i University of Bologna and INFN-Sezione di Bologna, I-40126 Bologna, Italy .i Tam Institute of Fundamental Research, Bombay 400 005, India

k Boston University, Boston, MA 022I5, USA i N[~rtheastern University Boston, MA 02115, USA

m lnsti~te of At~~rni~ Physics and Universi~ ctf Bucharest, R-769~ Bu~~rest, Ro~nia

n Central Research institute for Physics of the Hungarian Academy of Sciences, H-152s 3udapest 114. Hungary2 0 Harvard University, Cambridge, MA 02139. USA

v Massachusetts Institute of Technology, Cambridge, MA 02139, USA q INFN Sezione di Firenze and University of Florence, l-50125 Florence, Italy r European Laboratory for Particle Physics, CERN, CH-1211 Geneva 23, Switzerland

s World Laboratory, FBLJA Project, CH-1211 Geneva 23. Switzerland

’ Un~versi~ of Geneva, CH-I211 Geneva 4, Switzeria~

” Chinese Vn;versi~ of Science and Te~hnoiogy, VSTC, Hefei. Anhui 230 029, China6 y SEFL Research Institute for High Energy Physics, PO. Box 9, SF-00014 Helsinki, Finland

w University of Lausanne, CH-1015 Lausanne, Switzerland

X INFN-Sezione di Lecce and Vniversitd Degli Studi di Lecce, I-73100 Lecce, Italy Y Los Alamos National Laboratory, Los Alamos, NM 87544, USA

z Institut de Physique Nucieaire de Lyon, IN2P3-CNRSVniversite Claude Bernard, F-69622 Vilieurbanne, France iw Centro de Investigaciones Energeticas, Medioa~ientaies y Tecnologicas, CIEMAI: E-28040 Madrid, Spain 3

* INFN-Sezione di M~lano, I-20133 Milan, Itaiy

as fnst~tlf fe of Theoretical and ~~~er~rn~~~~~ Physics, ITEP, Moscow, Russia ari INFN-Sezione di Napoii and University of Naples. I-80125 Naples, Italy Be Department of Natural Sciences, University of Cyprus. Nicosia. Cyprus

‘t University of Nljmegen and NIKHEF NL-6525 ED Nljmegen, The Netherlands se Oak Ridge National Laboratory, Oak Ridge, TN 37831, USA ah California Institute of Technology, Pasadena, CA 91125, USA

Bi INFN-Sezione di Perugla and Vniversitd Degii Studi di Peru&a, l-06100 Perugia, Italy

‘i Carnegie Mellon Universi~, Pittsburgh, PA 15213, USA

& Princeton Univer.~i~~ Princeton. NJ 08544. USA

” INFN-Sezione di Roma and University of Rome, “La Sapienza”, 1-00185 Rome, Italy am Nuclear Physics Institute, St. Petersburg, Russia

an University and INFN, Salerno, I-84100 Salerno, Italy a’ University of CalQornia, San Diego, CA 92093, USA

av Dept. de Flsica de Particulas Elementales, Vniv. de Santiago, E-15706 Santiago de Compostela. Spain q Bulgarian Academy of Sciences, Central Lab. of Mechatronics and l~trumentation, BU-1113 Sofia, Bulgaria

a= Center for High Energy Physics, Korea Adv. Inst. of Sciences and Te~hnoiogy, 305-701 Taejon, South Korea ilF Unive~i~ of Aiaba~, Tus~ai~~osa, AL 35486. USA

” Vtrecht University and NIKHEE NL-3584 CB Vtrecht, The Netherlands a’ Purdue University, West Infayette, IN 47907, USA a’ Paul Scherrer Institut, PSI, CH-S232 tiiligen, Switzerland Bw DESY-Institut fu’r Hochenergiephysik, D-15738 Zeuthen, FRG

‘y Eidgeniissische Technische Hochschule, ETH ZSirich, CH-8093 Zrich, Switzerland a’ University of Hamburg, D-22761 Hamburg, FAG

Iw. High Energy Physics Group, Taiwan, ROC Rece+d 7 May 1997

Editor: K. Winter

392 L3 Collaboration/Physics Letters 3 407 (1997) 389-401

Abstract

A measu~ment of the Kf and A inclusive prodaction rates and momentum spectra in two- and three-jet events is presented, On the basis of about 3.1 million 2 decays collected with the L3 detector at LEP, we observe that the production of these particles is well modelled by string fragmentation. @ 1997 Elsevier Science B.V.

1. Introduction

Hadronic 2 decays provide important data to study the process of hadronization, which is too complex to be calculated in detail by perturbative QCD. Presently, it is only described by phenomenologicat models. Re- cently, we have shown experimental evidence for dif- ferences in hadroni~tion of quarks and gluons: The production of v mesons was found to be harder than expected in jets originating from the fragmentation of

gluons [ 1 ] .

In this paper, we measure the production of e mesons and A (A) 7 baryons in two-jet events and in quark and gluon jets from three-jet events. Compar- ison of Kt, A and charged particles may reveal dif- ferences in hadronization due to the strange quark as welt as to the mesonic and baryonic nature of these particles.

imbalances, EL and Et!, must satisfy the following conditions:

0.5 < 5 < 2.0; EL

7 < 0.5 and IElI ^I

fi ViS

F < 0.5.

“IS

About 3.1 miltion events are accepted. The trigger efficiency for these events is 99.9%.

The JETSET 7.3 program [3] is used to generate Monte Carlo events. The generated events are passed through the full detector simulation [4] which takes into account the effects of energy loss, multiple scat- tering, interactions and decays inside the detector ma- terials. The efficiency to accept hadronic Z decays is found to be 99.0%.

For comparison with another hadronization model, the HERWIG 5.9 event generator [ 51 is used. Its main difference with JETSET is that the hadroni~tion of partons is modelled by cluster fragmentation instead of string fragmentation.

2. Data and Monte CarIo samples

3. Measurement of production rates The analysed data, collected by the L3 detector [ 21

at LEP ( fi = 9 1 GeV) , correspond to an integrated luminosity of 112 pb-‘. The selection of hadronic Z decays is based on the energy measured in the efectro- magnetic and hadron calorimeters. Events must have more than 12 calorimetric clusters. The total visible energy, &is, the transverse and longitudinal energy

t Supported by the German ~uudesministe~um fur Bildung, Wis- senschaft, Forschung und Technoiogie.

* Supported by the Hungarian OTKA fund under contract num- bers T 14459 and T240 1 I.

3 Supported ako by the Comisidn Interministerial de Ciencia y Technologia.

4 Also supported by CONICET and Universidad National de La Plats, CC 67, 1900 La Plats, Argentina.

5 Also supported by Panjab University, Chandigarh- 16M)I 4, ^India.

’ Supported by the National Natumi Science Foundation of China.

’ In the following, A will refer to both A and k

The charged particle reconstruction is based on a Time Expansion Chamber surrounded by a Z Cham- ber and, since 1994, on a Silicon Microvertex Detec- tor [ 6,7]. To be selected, tracks must reach the ten- tral part of the detector (40” < 8 <140’) and have a transverse momentum greater than 150 MeV.

3.1. Secondary vertex selection

For the reconstruction of secondary vertices, V”, selected tracks are accepted according to the following criteria:

- The distance of closest approach to the nominal beam position of each track, projected onto the transverse plane, must be greater than 0.5 mm. This value is large enough to remove tracks originating from short-lived particle decays and small enough

W Collaboration/Physics Letters B 407 (1997) 389-401 393

to retain efficiency for tracks resulting from sec- ondary vertices of K! or A. The beam position is determined on a fill-by-fill basis.

- The transverse distance of flight, defined by the sec- ondary vertex position with respect to the beam axis, must be greater than 5 mm. This removes back- ground from short-lived particles.

- The angle between the direction of the transverse momentum of the track pairs and the transverse flight direction is required to be smaller than 200 mrad. This eliminates combinations of tracks not pointing to the beam axis.

3.2. I!$ and A measurement

The @ candidates are r~onstructed by calculating the invariant mass of oppositely charged tracks, assum- ing each track to be a pion. For the A reconstruction, the proton mass is assigned to the track with the high- est momentum. KY candidates with a scaled momen- tum xP smaller than 0.005, where x,> is defined as the reconstructed momentum divided by the beam energy, are rejected. This threshold is increased to 0.01 for A baryons to exclude photon conversions. The combina- torial background distributions are obtained by com- bining tracks with the same charge. These distribu- tions are corrected on an event-by-event basis for the difference between the number of like and unlike sign combinations. Without this correction, the combina- torial backgrounds would be underestimated by about 3% and the results overestimated by 0.4%.

The ?r+rr- and p7r- (prr’ f mass distributions, af- ter subtraction of the combinatorial background, are shown in Fig. 1, both for the data and simulation. The hatched histograms in Fig. 1 show the expected con- tributions of e, A and y + e+e- . These mass dis- tributions are mainly populated by secondary vertex candidates. Other contributions to the background are found to be less than 0.5%

As shown in the Monte Carlo distributions of Fig.

la, the ST+-- mass distribution includes mainly KY but also has some contribution from A. The prr mass distribution in Fig. lb shows, besides the A signal, a strong @ contribution and a small contribution from y --+ e+e- conversions.

The number of e and A in the Monte Carlo dis- tributions, inside the mass windows 0.3-0.8 GeV and 1.07- 1.17 GeV of Fig. 1 a and Fig. 1 b respectively, are

l DATA L3 iiCt<t

!ZBMCA MC y-+e+e-

y.f. 25

0 0.4 0.6

hf- mass (GeV)

4

. IMCA

ES~MCK:

MC y+e+e

pn+iin’

mass (GeV)

5

Fig. 1. a) Background subtracted v+?r- mass distributions for

simulation and data. b) Background subtracted pr- + pa+ mass distributions for simulation and data.

resealed to fit the data. The fractions of expected @ and A in these distributions, as well as the magnitude of other background contributions, mainly photon con- versions, are kept fixed. The scale factors so obtained are used to calculate the production rates of e and A in the data. The resulting total numbers of particles recons~uct~ in the a~ve-mentions mass intervals are (560.0f1.5) x lo3 e and (88.0&0.9) x 16 A.

Including correction for neutral decays, the total ac- ceptances are 18% and 8% for g and A, respectively.

394 W Collaboration/Physics Letters B 407 (1997) 389-401

KQ

I

b ‘c

I ^{I I t ‘*.}

0 0.2 0.4 0.6 0.8 cl

a

+

*

1

*~=-&xp)

4

4 A L3

*

I . . . . I . . ..I...

1

2~=lo~(xp) 4

0 0.2 0.4 0.6 I

X P

Fig, 2. a) The differential inclusive production rate for Kf in the total event sample. b) The corresponding spectrum for A. c) and d) The distribution of the variable 5 for KY and for A respectively. The total hadronic cross section is denoted by CQ. Error bars include statistical and systematic uncertainties.

3.3. Kf and A momentum spectra

The procedure described in Section 3.2 has been repeated for different q and A momenta. The result- ing differential inclusive production rates are shown in Fig. 2a for the Kf , and Fig. 2b for the A. The shape for Kf is in reasonable agreement with both Monte

Carlo expectations while the A production rate at low momentum is under~timat~ by JETSET and overes- timated by HERWIG. The inclusive production rates have also been measured as a function of the vari- able 5 = - log( xP). These distributions are shown in Fig. 2c for KY and Fig. 2d for A. They have been fitted in the indicated region using the distorted Gaussian

Table I

Contributions to the total systematic uncertainty on average mul- tipficities.

K’,’ A

Tracking acceptance and efficiency 1.3% 1.5%

Secondary vertex selection and background 1.5% 2.9%

A reflection, KY reflection, y -+ e+e- 0.6% 3.3%

Monte Carlo Stutistics 0.3% I .3%

Total 2.1% 4.8%

function expected in the bodied Leading Log Ap- proximation (MLLA) [S] . The 63-iosen region is rhe largest with a smalf x”. We find a maximum at <i& = 2.7610.04 for Kf and 5; = 2.78f0.05 for n baryois.

The spectra have been integrated in order to obtain the total inclusive production rates. After extrapolation to the full phase space, we obtain the following average

multiplicities per event:

= 1.012 Z!Z O.O03(stat.) * 0.021 (syst.) KY/event, (&@I

= 0.364 f O.O04(stat.) f O.O17(syst.) A/event.

Both muitipl~city and p measurements are in good agreement with our previous measurement f9] and with results of other LEP experiment [ to-121.

7%ese production rates can be compared with the Monte Cario predictions which are respectively

1.024 qlevent and 0.347 A/event with JETSET, and 1.041 K!/event and 0.378 A/event with HERWIG.

The expected values for r from JETSET ( HERWIG) are SC, = 2.71 (2.77) and 6: = 2.60 (2.83).

The quoted systematic uncertainty on the presented average multiplicities is the quadratic sum of different contributions as shown in Table 1. The main contribu- tion to the systematic error arises from the selection procedure for tracks and secondary vertices. The mea- surements have been repeated varying the selection criteria described in Section 3.1. Changing the charged particle momentum threshold and extending the polar angle acceptance introduces an error of 1.3% ( lS%) for IQ {A). Small errors in the ~xtr~~at~on to the fult phase space are also included. Bhabha events have been used to correct Monte Carlo tracking efficiency.

Removing one by one the cuts to accept secondary

vertex candidates increases the background by more than a factor two, whereas the measured rate does not change by more than 1.4% for Kz and 2.7% for A. In- cluding an uncertainty of 0.5% ( 1.1%) found by com- paring total rates with the sum of those obtained in two- and three-jet events (see next section) where the event topology is different, we get a total uncertainty of 1.5% (2.9%). Tests with Monte Carlo simulation have shown that the measurements are not sensitive to a wrong number of generated charged particles, q or A.

In the case of the Kt, the uncertainty in the background originating from the measured R production rate is 0.6%. For the A, the ~o~~nding dainty due to Ki is 3.3%. An un&e~~nty of IO% in the rate of pho- ton conversions has a n~Iigible effecr on the results.

4, Production rates within jets 4.1. Jet reconstruction

The LUCLUS jet finder algorithm [3,13] is used to reconstruct jet energies and directions. Calorim&c clusters of energy and direction Ei and Ej are joined to a single cluster with energy and direction Ei+Ej if the distance dij defined by

d. gj = 2 EiEj sin(#J2) Ei + Ej

is smaller than a given value &in. The angle between the two clusters is denoted by @G. This procedure is repeated recursively to join alI possible pairs of clus- ters. The final resulting clusters are called jets. The parameter djoin is fixed to 7 GeV. The fraction of two- jet events is found to be 70.6% in the data and 69.8%

(70.5%) in the simulation based on JETSET (HER- WIG). Events with more than two jets are classified as three-jet events.

In three-jet events, jets are sorted according to their energy ( EI > E2 3 Es). The reconstructed jet en- ergy distributions are shown in Fig. 3. We observe good agreement between data and simulation. Using the JETSET program in Matrix Element mode, the probability that the least energetic jet originates from gluon fragmentation is found to be 72%, while the probability that the most energetic jet originates from one of the primary quarks is kger than w%. Hence,

396 L.3 Coilaboration/Ph~sics Letters B 407 (1997) 389-401

0 80

Fig. 3. The reconstructed jet energy distributions for the three species of jets in three-jet events.

the two most energetic jets are referred to as quark jets while the remaining jets are called gluon jets.

The flow of the sum of Kz and A (V”) and of charged particles in three-jet events, is compared with the simulation in Fig. 4. The charged particles used in this comparison are selected with the same track se- lection described in Section 3, The number of charged particles includes tracks from decays of neutral parti- cles like Kf, A and n”” Dalitz decays, but it has been corrected for electrons coming from photon conver- sions. The comparison is performed in the event plane of the three-jet events. This plane is defined by the axis of the quark jets, in their rest frame, and the di- rection of the third jet. The distribution of the angle, defined such that jet 1, the most energetic jet, is at O’, jet 2 at 180* and jet 3 between 180’ and 360”, of the particle trajectories projected onto this plane is shown in Fig. 4a for the sum of IC! and A, and in Fig. 4b for charged particles. The particle distributions within jets are well reproduced by our simulations.

Slight differences are observed in the region between the two quark jets opposite to the third jet, which is sensitive to the “string effect” [ 141. In this region, the data for charged particles and for the sum of g and A are consistent with each other. One also notes that the enhancement co~esponding to gluon jets is more marked for neutral particles associated to V” than for charged particles. According to JETSET, this is due to A production.

0 90 180 240 3

angle (degree) charged L3 n ^{l DATA}

-lo47

0 90 180 240 2 i0

angle (degree)

Fig. 4. a) Angular flow distribution in three-jet events for neutral particles decaying in a secondary vertex, V”. The angle goes from the most energetic jet to the second one (at 18P) then, from the second one to the third and finally back to the first one. The region sensitive to the “string effect” is around 135’ and the gluon jet region is found around 200°. The continuous lines stands for the JETSET expectation. The dashed line corresponds to the expected Kz contribution. b) Angular flow distribution for chatged particles, The continuous (dashed) line stands for the JETSET (HERWIG) prediction.

4.2. Measurements within jets

~oduction rates in two- and three-jet events are measured using the same procedure as for the total event sample in Section 3. The differential production rates for e and A are shown in Fig. 5. For both parti- cles, the dis~butions show harder momentum spectra in two-jet events than in three-jet events. The ratios,

R32, of the total rates measured in three-jet events over that the total rates in two-jet events are found to be RX(~) = 1.30f0.02 and R32 (A ) = 1.40f0.04. A

L3 Collaboration/Physics Letters B 407 (1997) 389-401 397

a I t

0

X P

4 A L3

1 DATA

I 0 Z-jet events

S

0 3-jet events 3 ld

i . b ‘f

ld

- JETSET I ,rn,rr,rr

“i t-

--.-._.. ^rlcnvvlu

02 * 04 * 06 . (

X P

0.4

I

L3

%

i3

To.2

‘c b

DATA

l 2-jet events O 3-jet events - Fit

I *++%

0 O0

0 . 0

0

I*“‘.“‘- 11.. . .I -.,

O

2~=-lo;(xp)

4

%O.l

2 .

m

‘c b

3

O

4 A L3

DATA

l 2-jet events O 3-jet events - Fit

Fig. 5. a) The differential inclusive production rates for Kf in two-jet and three-jet events. b) The corresponding spectra for A. c) and d) The distribution of the variable ( for Kf and for A respectively. The total hadronic cross section is denoted by CJ,. Error bars include statistical and systematic uncertainties.

similar increase is also observed for charged particles for which Rs2fNct,) is 1.36 i 0.01. Higher particle pr~uctioR is expected from QCD for gluon jets be- cause the colour charge of gluons is larger than that of quarks. This effect is implemented in JETSET as well as in HERWIG. The Monte Carlo expectations from JETSET are Rjz(Kf) = 1.26, R32(R) = 1.42

and R32( Nch) = 1.30. They are in good agreement with our measurements. The charged particle produc- tion ratio is reasonably well reproduced by HRRWIG with Rs2(Nct,) = 1.30, but its predictions for e and A,Rsz(Kf) = 1.12andRg2(A) = 1.19,disagreewith the data. The expected higher multiplicity in tbree-jet events leads to lower average particle momentum and

398 L3 Collaboration/Physics Letters B 407 (1997) 389-401

. . . . lir'

- JETSET . ^.._--1.

l Quark Jets 1+2 A Gluon Jets

X P

0.08

%

9 .

‘32 b

3 0

DATA

* Quark Jets 1+2 A Gluon Jets - Fit

+

:.fij

1

25=-lo&p) ⁴

4 A

DATA

l Quark Jets 1+2 A Gluon Jets - Fit

Fig. 6. a) The differential inclusive production rates for Kt in three-jet events for quark and gluon jets separately. b) The corresponding spectra for A. The result labelled as quark jets is the sum of the rates measured in the two most energetic jets. The difference between the two quark jets were not significant. c) and d) The dis~bution of the variable 5 for Kt and for A respectively. The total hadronic cross section is denoted by rrh. Error bars include statistical and systematic uncertainties.

hence to softer spectra and higher value of t*. This production rates are presented in Fig. 6a and Fig. 6b.

is indeed observed (see Table 2) and reasonably well For Kf, the shapes of the measured spectra are in good reproduced by both JETSET and HERWIG. agreement with JETSET for both types of jet. Slightly For three-jet events, we compare production rates softer distributions are again observed for A, except in gluon jets with those in quark jets. To measure for the gluon jets, where the rate is in better agreement the rates within jets, particles are associated to the at low x,, than for quark jets. With IERWIG for both closest jet in the laboratory frame. The differential &y and A, we observe genera1 agreement in quark jets,

Table 2

W Coliaboration / Physics Letters B 407 (1997) 389-401 399

Position of the maxima 6’ in the different event configumtions for K$ and A. Uncertainties am statistical and systematic.

DATA JETSET HERWIG DATA JETSET HERWIG

All events 2.76f0.04 2.71 2.77 2.78f0.05 2.60 2.83

Two-jet events 2.60f0.04 2.50 2.54 2.57zkO.13 2.35 2.78

Three-jet events 2.99f0.04 2.99 3.09 2.90f0.05 2.70 2.86

- Quark jets 2.8lf0.04 2.69 2.80 2.81f0.07 2.45 2.67

- Gluon jets 3.28f0.03 3.28 3.47 3.01 zbo.05 2.92 3.06

Table 3

Measured relative rates for K’,‘, A and charged particles observed in different species of jets and events with their respective Monte Carlo expectations, Un~~ainties are statistica and systematic added in quadrature.

(K$erjet ^(Nprjet

(Ki’ht)lota~ 6%hlaI

DATA JETSET HERWIG DATA

hJpjet

~Nchb-d

JRTSET HERWIG DATA JETSET HERWIG

Two-jet events ~.#O~O.O# 0‘464 0.48 I 0.45110.01 I 0.444 0.470 0.460~0.~1 0.460 0.455 Three-jet events 0.39910.005 0.39 I 0.360 0.42 1 f0.013 0.423 0.375 0.4~~0.~1 0.398 0.395 - Quark jets 0.417f0.009 0.413 0.400 0.436f0.014 0.426 0.400 0.417f0.001 0.416 0.412 - Gluon jets 0.36 I f0.007 0.347 0.28 I 0.403*0.015 0.417 0.326 0.363f0.001 0.358 0.361 , _ Rat~I”Gl”onJct

(13&3)% 16% 29% (S&5)% 2% 19%

Ratet~~l (13.0f I)% 14% 12%

but the spectra are softer in the gluon jets.

Measured as a function of 5, the dis~ibutions, shown in Fig. 6c and Fig. 6d, have a maximum 5 (see Table 2) shifted toward higher values in gluon jets relative to quark jets. This is qualitatively re- produced by both IETSET and HERWIG. The value of 5” for gluon jets is much greater than for quark jets or than all events although the energy is much lower. This is in contrast to the relationship [* cx fi experimentally observed for all events [ 81.

The spectra are integrated to determine the rates of e and A per jet. Then, the rates are divided by the total inclusive production rates measured in Section 3. These relative production rates, i.e, the ratio of the rate per jet to the rate per event in the total sample, are more accurate because many of the systematic un- certainties cancel in the ratio. These relative rates are given in Table 3. Good agreement is found between data and JETSET expectations. However, HERWIG differs markedly from the data.

The corresponding rates for charged particles are given in the last columns of Table 3 with their respec-

tive Monte Carlo expectations. The data agree with the predictions of both JETSET and HERWIG.

In three-jet events, we can compare rates in gluon jets with those measured in quark jets by forming the relevant ratios. For K$ we observe that the rate in gluon jets is about ( 13&3)% lower than in quark jets. The same reduction, ( 13fl) %, is measured for charged particles. A reduction of (8&S)% is found for A. These suctions are well reproduced by JET- SET. HERWIG predicts larger reductions for e and A, namely 29% and 19%, respectively.

4.3. Comparison with the r&es of charged paltticles The observed reduction of particle production in gluon jets is expected to be mainly due to the reduced phase space available during the fragmentation. How- ever, this reduction is partially compensated by an in- crease of particle production from gluon fragmenta- tion relative to quark catenation due to the higher colour charge of gluons. Hence, in order to investigate a possible difference in the g and A production in

400 L3 ~~~~~~~uti~n /Physics Let!ers B 407 (I 9971389-401 Table 4

The relative production rates, Rf Kf) and R(A), observed in the different species of jets and events with their Monte Carlo expectntioas.

Uncertainties are statistical and systematic added in quadnture.

Two-jet events Three-jet events - Quark jets - Glrron jets

gluon jets related to the nature of these particles rather than to the jet energy, a comparison is made with the rate of charged particles.

Table 4 shows the ratios, R(

Kf

The measured ratio R(Kf) is similar in gluon jets and in quark jets, whereas R(A) is greater in the gluon jets. Tbis is reproduced by JETSET but not by HER-

WIG.

In Fig. 7, we plot the dependence of R on the

SC&~ jet energy (2~j~~/~~i~) for Kz and A in three- jet events. The distributions for Kz and A are differ- ent, Whereas R(q) is almost independent of energy, R(A) increases signi~cantly at both low and high en- ergies, Data and JETSET are in reasonable agreement for both g and A, but HERWICI fails to describe R( Kf ) and R( A ) for the tow energetic jets, where the

~~richement in gfuon jets is the highest. Similar obser- vations have recently also been made by the DELPHI collaboration I 15 1.

These relative rate variations seen for A are well reproduced by JETSET. This agreement is obtained by the presence of different processes which con- tribute to the baryon production. The increase of A baryon production in the less energetic jets with re- spect to charged particles is modelled by JETSET with a larger diquark production relative to quark produc- tion at lower energy, This is obtained with a diquark segmentation function which is softer than the one

Fig. 7. The v&es of R(KI,‘) and R(A) as function of the scaled

jet energy (2Ejer/EVis) in three-jet t~ant~, The erm ba*; include statistical and systematic ao~~aia~~s.

W Collaboration /Physics Letters B 407 (1997) 389-401 401

for quarks. No increase is expected from the HER- WIG program in which no specific process such as diquark production is implemented. In the JETSET program, the increase of quark production with en- ergy also favours “popcorn” processes [ 31. This qual- itatively explains the increase of the relative rates ob- served in highly energetic jets. In HERWIG, cluster fragmentation makes the particle production mainly dependent on the available phase space. Hence, the production of the lightest particles is favoured in the least energetic jets.

5. Conclusion

We have measured production of Kg and A in hadronic Z decays, and their production in two-jet events and within three-jet events have been com- pared. The overall agreement between data and sim- ulation based on string fragmentation is good, both for quark jet and gluon jet samples. For Kt mesons, this illustrates that the production of strange quarks relative to other light quarks does not differ signifi- cantly between quark and gluon fragmentation. For A baryons, differences are observed. They can be explained by the specific energy dependence of the production processes for baryons as implemented in the JETSET model. There is no need to invoke differ- ences in quark and gluon fragmentation beyond those implied by string fragmentation.

Acknowledgements

We wish to express our gratitude to the CERN ac- celerator divisions for the excellent performance of the LEP machine. We acknowledge with appreciation the effort of all engineers, technicians and support staff who have participated in the construction and main- tenance of this experiment. Those of us who are not from member states thank CERN for its hospitality and help.

References

[ I ] L3 Collab., M. Acciarri et al., Phys. Len. B 371 (1996) 126.

121 L3 Collab., B. Adeva et al., Nucl. Instr. Meth. A 289 ( 1990) 35;

L3 Collab., B. Adriani et al., Phys. Rep. 236 ( 1993) I.

131 JETSET Monte Carlo program.

T. Sjiistrand, Comp. Phys. Comm. 82 ( 1994) 74;

T. Sjiistrand, Lund and CERN Preprints LU TP 95- 20/CERN-TH.7112/93 (revised, 1995).

The parameters of JETSET have been tuned in order to reproduce event shape variables as seen by the L3 detector.

The production of particles (tensor mesons) with orbital angular momentum is enabled. For the tuning see:

L3 Collab., B. Adeva et al., Z. Phys. C 55 (1992) 39.

[4] The L3 detector simulation is based on GEANT

I51 HERWIG Monte Carlo program:

G. Marchesini and B. Webber. Nucl. Phys. B 310 (1988) 461;

161

[71

I.G. Knowles, Nucl. Phys. B 310 (1988) 571;

G. Marchesini et al., Comp. Phys. Comm. 67 ( 1992) 465.

The parameters of HERWIG have been tuned in order to reproduce event shape variables as seen by the I.3 detector.

The used parameters are: CLMAX = 3.052, QCDLAM = 0.183, VPCUT = .4 and CLPGW = 2.221.

F. Beissel et al., Nucl. Instr. Meth. A 332 (1993) 33.

L3 Collab., M. Acciarri et al., Nucl. In&. Meth. A 351 (1994) 300.

181 L91 1101 [Ill

Y.L. Dokshitzer, V.A. Khoze and V.A. Troyan, Z. Phys. C 55 (1992) 107.

L3 Collab., M. Acciarri et al., Phys. Lett. B 328 ( 1994) 223.

ALEPH Collab., D. Buskulic et al., Z. Phys. C 64 (1994) 361.

DELPHI Collab., P Abreu et al., Z. Phys. C 65 ( 1995) 587.

[ 121 OPAL Collab., A. Ackers et al., Z. Phys C 67 (1995) 45;

Z. Phys C 73 (1997) 569.

131 T. Sjiistrand, Comp. Phys. Comm. 28 (1983) 229.

141 Y.I. Azimov, Y.L. Dokshitzer, V.A. Khoze, S.I. Troyan, Phys.

Lett. B 165 (1985) 147.

I51 DELPHI Collab., P Abreu et al., CERN-PPE/96-193, submitted to Phys. Iett. B.

see R. Brun et al, GEANT 3, CERN DD/EE/84-1 (revised, Sept 1987).

The GHEISHA program is used to simulated hadronic interactions see H. Fesefeld, RWTH Aachen report PITHA 85/02( 1985).