Precise determination of the lifetime of the charmed baryon Λc

  • Published on

  • View

  • Download

Embed Size (px)


<ul><li><p>Volume 218, number 3 PHYSICS LETTERS B 23 February 1989 </p><p>PRECISE DETERMINATION OF THE LIFETIME OF THE CHARMED BARYON A~ </p><p>ACCMOR Col laborat ion </p><p>Amsterdam-Br i s to l -CERN-Cracow-Mun ich -Ruther fo rd -Va lenc ia Col laborat ion </p><p>S. BARLAG a.l, H. BECKER a,2 T. BOHRINGER b.3, M. BOSMAN a, V. CASTILLO b.4, V. CHABAUD b, C. DAMERELL c, C. DAUM d, H. D IETL a, A. G ILLMAN c, R. G ILMORE e, T. GOOCH e, L. GORL ICH f, P. GRAS g, Z. HA JDUK f, E. H IGON g, D.P. KELSEY b,5, R. KLANNER a'6, S. KWAN b, B. LUCKING a, G. LOTJENS a, V. LUTH b,7 G. LUTZ ~, J. MALOS ", W. MA.NNER ", E. NEUGEBAUER ,,8, H. PALKA ~, M. PEPE ~.9, j . R ICHARDSON ~.lo, K. RYBICKI f, H.J. SEEBRUNNER b, U. ST IERLIN ", H.G. T IECKE d, G. WALTERMANN a, S. WATTS h, p. WEILHAMMER b, F. WICKENS c, L.W. WIGGERS d, M. WITEK f and T. ZELUDZIEWICZ f </p><p>Max Planck Institut fiir Physik, D-8000 Munich 40, Fed. Rep. Germany b CERN, CH-1211 Geneva 23, Switzerland ' RutherfordAppleton Laboratory, Chilton, Didcot OX l l OQX, UK </p><p>NIKHEF-H, NL-1009 DB Amsterdam, The Netherlands University' of Bristol, Bristol BS8 1 TL, UK </p><p>f Institute of Nuclear Physics, PL-30055 Cracow, Poland </p><p>Received l 9 December 1988 </p><p>We present the final result on the measurement of the lifetime of the At-baryon from a sample of At-decays obtained in the NA32 experiment at the CERN SPS using silicon microstrip detectors and charge-coupled devices for vertex reconstruction. We observe a total of 160 Ac in five different decay modes. A sample of 101 unambiguous decays of A~.-~pK-~ + (and charge conju- gate) above a background of 7 events is used for the Ac mass and lifetime measurement. The mean A,. lifetime is determined to be 1.96_+:~] 10 t3s and the Ac mass to be 2285.8 _+ 0.6 _+ 1.2 MeV/c 2. </p><p>1. Introduction </p><p>Present address: LAL, F-91405 Orsay, France. 2 Present address: Gesamthochschule, D-6600 Saarbrticken, Fed. </p><p>Rep. Germany. Present address: University of Lausanne, CH- 1015 Lausanne, Switzerland. </p><p>4 Present address: University of Valencia, Valencia, Spain. Present address: Rutherford Appleton Laboratory, Chitton, Didcot OXI 1 0QX, UK. </p><p>~' Present address: DESY, D-2000 Hamburg 52, Fed. Rep. Germany. </p><p>7 Visitor from SLAC, Stanford, CA 94305, USA. 8 Present address: Universitfit-GH Siegen, D-5900 Siegen, Fed. </p><p>Rep. Germany. ,7 Present address: CERN, CH-121 I Geneva 23, Switzerland. "~ Present address: University of Geneva, CH-1211 Geneva 4, </p><p>Switzerland. </p><p>The study of the decay properties of the weakly de- caying charm particles is a rich source of in format ion about the interplay of weak and strong interact ions at short distances. On the one hand the weak decays of heavy quarks offer the possibil ity to test predic- t ions of the SU(2) U( 1 ) standard model of elec- troweak interactions and on the other hand short- and long-distance QCD corrections to the weak decay hami l ton ian are of great importance for a complete understanding of the experimental results. </p><p>A major breakthrough for fixed target charm ex- per iments was the development of sil icon microstr ip detectors. In 1981 the ACCMOR col laboration at </p><p>374 0370-2693/89/$ 03.50 Elsevier Science Publ ishers B.V. ( Nor th -Ho l land Physics Publ ish ing D iv is ion) </p></li><li><p>Volume 218, number 3 PHYSICS LETTERS B 23 February 1989 </p><p>CERN was the first group to introduce silicon mi- crostrip detectors in an experiment (NA11 ) [ 1 ] and to complement these in 1985 with charge-coupled devices (CCDs) in the NA32 experiment [ 2 ]. </p><p>Many experiments contributed to a better under- standing of the decay properties of the three weakly decaying charmed mesons (15 ~, D + , D + ), whereas still very little is known about the four weakly decay- ing charmed baryons (A,?, -c-+-, (e~o_ c, (~o) . In con- trast to the mesons the baryons offer a unique possi- bility to study the importance of W exchange graphs in the weak decays of charm particles, since in the case of baryons these amplitudes are neither colour nor helicity suppressed [ 3 ]. </p><p>2. The experiment </p><p>During the year 1984 the NA32 experiment at the CERN SPS studied charm production in hadronic re- actions and measured the lifetimes of the charmed mesons [ 4 ], using an interaction trigger and a vertex detector consisting of a segmented silicon active tar- get and silicon microstrip detectors. We report here on the second phase of NA32 which was designed to be a dedicated Ac and Ds-experiment and had two data taking periods in 1985 and 1986. For this pur- pose a ~ p)K* and K + K ~ pair trigger was used, which was based on the FAMP (Fast Amsterdam Multi Processor) system [5]. This trigger had been used originally in the NA11 experiment in 1982 to collect 600 000 0's [ 6 ], but it had also proved to be a selec- tive charm trigger yielding a sample of D~-~0~ for mass determination and lifetime measurement [7]. In order to be more sensitive in the range of very short lifetimes the silicon microstrip vertex detector was upgraded with the addition of two CCDs [ 2 ]. These detectors have pixels of 22 jam 22 ~am and allow the measurement of space points with a precision of ~ 5 gm in two orthogonal coordinates. A two-track reso- lution of 40 gm was achieved. The two CCDs were located 10 and 20 mm downstream of the 2.5 mm long Cu target, allowing the observation of secondary vertices in vacuum immediately behind the target. They are followed by a setup of eight silicon micro- strip detectors ranging from 65 to 180 mm down- stream of the target. A beam telescope consisting of seven silicon microstrip detectors measured the tra- </p><p>jectory of the incoming beam particle with a preci- sion of 3 gm vertically and 7 gm horizontally at the position of the target. </p><p>The FAMP trigger reduced the rate of accepted events by a factor 12 compared to an interaction trig- ger. The enrichment factor of the recorded data sam- ple in A, .~pK-n + and Ac-,15K+n - decays is 7, tak- ing into account the ratio of acceptances between this trigger and an interaction trigger. However, the gain in sensitivity (defined as the number of events per unit time) for Ac decays was only a factor 5, due to the increase of the overall deadtime of the experiment. </p><p>The experiment was located in the H6 beam of the North Area at the CERN SPS and used a negative beam with a momentum of 230 GeV/c. Hadronic charm decays into charged particles were fully recon- structed with the large acceptance forward spectrom- eter [6], which consisted of two magnets and 48 planes of drift chambers. Three multicellular thresh- old Cerenkov counters were used to identify n, K, p in the momentum range 4-80 GeV/c. More details about the experiment, in particular about the trigger and the vertex detector, can be found in ref. [ 8 ]. </p><p>3. Data analysis </p><p>In order to ensure a high efficiency for finding charm decays two independent approaches to the data analysis have been developed. The comparison be- tween the two analysis programs during development was found useful to achieve a high efficiency. In the following we describe only the analysis on which the results presented in this paper are based. Details about the other approach can be found in ref. [ 8 ]. </p><p>The search for charm decay vertices is carried out in five steps. </p><p>First, tracks are reconstructed for all events in the drift chambers of the forward spectrometer. </p><p>In the second step, the beam track and all outgoing tracks are reconstructed in the beam and vertex tele- scopes independently of the drift chamber track in- formation. Then, tracks found in the drift chamber and in the vertex telescope are matched. Unmatched drift chamber tracks are used as a loose guidance to recover tracks out of complex clusters of signals in the vertex telescope. Particles are identified using the information of the Cerenkov hodoscopes. The recon- </p><p>375 </p></li><li><p>Volume 218, number 3 PHYSICS LETTERS B 23 February 1989 </p><p>struction of the primary vertex is performed using all tracks except for tracks with a large impact parame- ter (the distance of closest approach in space of a track to the primary vertex), i.e. tracks with a X 2 probabil- ity of less than 1% for originating from the primary vertex. Fig. 1 shows the calculated error distributions of fitted coordinates of the primary vertices (x refers to the horizontal, y to the vertical coordinate and z points along the beam axis). We checked that all Z 2 distributions are well normalized. Fig. 2 shows the measured impact parameter distribution of all tracks. The impact parameter of a track is measured with a precision ofae~ad,,~ [52+ (18/p) 2 ] 1/2 gm (dx, dy: minimum projected distance of a track from the pri- mary vertex) where the momentum dependent term </p><p>3 .10 1.2 a </p><p>- - I I I. I </p><p>E Ji-I &lt; o-~&gt; ~2.3/.zm </p><p>0.4 ]1 &lt; y&gt; 2"O/'zm </p><p>~o 0 .8 - d </p><p>0.6 - </p><p>t~ I -- o', </p><p>0.2 </p><p> 0 2 4 6 (~) </p><p>200 b </p><p>175 </p><p>~ 150 125 </p><p>100 </p><p>75 </p><p>w 50 </p><p>25 </p><p>0 0 80 160 240 </p><p>a, (ffm) </p><p>Fig. 1. Error distributions of the fitted primary vertices (z-axis is beam direction ). </p><p>.10 </p><p>1.6 </p><p>1.4 E :::l. 1.2 </p><p>1. </p><p>o.8 to </p><p>2 0.6 b- </p><p>0.4 </p><p>0.2 </p><p>O. 0 </p><p> =7.9,u.m </p><p>10 20 30 40 i.p. (,u,m) </p><p>Fig. 2. Impact parameter (i.p.) distribution of all tracks with re- spect to primary vertex. </p><p>(p is measured in GeV/c) is the contribution from multiple scattering. </p><p>The third step is a loose preselection on charm de- cays. All events are selected which have a primary vertex inside the Cu target and at least two tracks not originating from the primary vertex (as defined un- der step two) or one such track and a K or A recon- structed in the drift chambers. This reduced the orig- inal data sample by about a factor seven. </p><p>In the fourth step a secondary vertex search is per- formed. First the tracks which do not originate from the primary vertex are used to fit one or more sec- ondary vertices. Then tracks which were originally used in the fit of the primary vertex are checked for compatibility with any of the secondary vertices. Fre- quently, this leads to multiple assignments of the tracks all of which are used in the final analysis. Fig. 3 shows the measurement error on the distance be- tween fitted primary and secondary vertices. The se- lection of all events with at least one good secondary vertex outside the Cu target reduces the remaining data sample by another factor four. </p><p>In the last step, the secondary vertices are scanned for fully reconstructed charm decays by checking all effective mass combinations compatible with the particle identification. Also "wrong sign" combina- tions are formed for background studies. This ap- proach based on the purely topological search for sec- ondary vertices enables us to detect some rare decay modes which have not been observed before [9 ]. In </p><p>376 </p></li><li><p>Volume 218, number 3 PHYSICS LETTERS B 23 February 1989 </p><p>140 </p><p>E120 = 190/.~m </p><p>2 lOO </p><p>8O </p><p>6o </p><p>t~ 4o </p><p>20 </p><p>0 0 200 400 600 </p><p>~ (/~rn) </p><p>Fig. 3. Distribution of errors on decay lengths. </p><p>order to be kept as a charm candidate the total mo- mentum vector of the decay has to have a Z 2 proba- bility to originate from the primary vertex greater than 1%. </p><p>4. Results </p><p>The results presented are based on the analysis of the full data sample of 16.7 106 triggers. We observe clean Ac decays in five different decay channels: Ac~pKn (~ 135 events), Ac~I~pn+n (~ 10 events), Ac~pK-n+rt -n + ( ~ 4 events), Ac~p ( ~ 3 events) and Ac~Z+n+n - (~ 11 events) w i the + de- caying to pn [ 9 ]. All event numbers given are back- ground subtracted. For the determination of the Ac lifetime we only retain the pK-n + (and c.c.) com- binations where we observe a signal of 135 events above a background of 31 events. The rather loose cuts of the general data analysis are further tightened to obtain a bias-free measurement of the lifetime of the A~ for which the removal of background is essen- tial. A serious source of background is reflections from the decay channels D+~K+K-n and D - , K + K - rt +- which occur due to ambiguities in the iden- tif ication of protons and kaons. These decays are re- moved from the sample. Reflections from D --,K;rt-+r~ +- decays are negligible because of the small probabil i ty of ambiguity between proton and pion. To improve the s ignal /background ratio even further, the following addit ional cuts are imposed: </p><p>( i) we require impact parameters of more than 3 a </p><p>with respect to the primary vertex (see fig. 2) for at least two of the decay tracks, and 1 a for the third decay track, </p><p>( i i) we require a min imum separation of 3 aA_-be- tween primary and secondary vertex (see fig. 3 ), </p><p>( i i i ) we l imit decay volume to the posit ion of the second CCD, which is located 20 mm downstream of the end of the target. The probabil ity for a Ac with a proper lifetime of 2 X 10- ~3 s and a momentum of 100 GeV/c to decay downstream of this decay vol- ume is below 10 -4 . </p><p>Finally we end up with 108 events within 3 ao f the A~ mass (where a is the calculated error on the mass) and a background of 75 events in the mass interval from 2.1 to 2.5 GeV. The estimated background be- low the signal is 7 events. Fig. 4 shows the invariant mass distribution. </p><p>For the extraction of the lifetime from the data, we have to correct for the acceptance of the selection cri- teria [ 1 ]. For each event we determine the min imum and maximum detectable lifetime, tmin and t . . . . re- spectively. This is done by moving the decay vertex along the straight line connecting the primary and secondary vertex backwards and forwards keeping all other parameters of the event unchanged. The mini- mum and maximum distances from the primary ver- tex, /min and/max respectively, where the event fails the imposed cuts, determine tmin and tmax. Here tmax is defined by the end of the decay volume, i.e. the posit ion of CCD2. As discussed in ref. [ 1 ], we fit the distr ibution of corrected lifetimes, tcor/r= (t . . . . - - tmi)/r, where z is the mean lifetime. We perform a </p><p>40 </p><p>~u 35 </p><p>30 &gt; </p><p>25 t~ 2O </p><p>15 _+.., </p><p>E 10 Q) &gt; </p><p>t~ 5 </p><p>0 </p><p>m </p><p>m </p><p>m </p><p>m </p><p>m </p><p>L . . . , . . . . . . ~ . . . . . ~ .~.~ </p><p>2.1 2.2 2..3 2.4 2.5 </p><p>M(pK-'n -+) (GeV/c') </p><p>Fig. 4. Invariant mass distribution of the pK ~+ and 15K+~ - sys- tem after all cuts for determination of the Ac lifetime. </p><p>377 </p></li><li><p>Volume 218, number 3 PHYSICS LETTERS B 23 February 1989 </p><p>combined maximum likelihood fit to the mass and corrected lifetime distributions for signal and back- ground, where the lifetime distribution of the back- ground is also taken to be exponential. For the mass spectrum we assume a gaussian signal above a linear background. The width is obtained from the calcu- lated mass resolution of each event. Figs. 5 and 6 show the decay length and the corrected lifetime distribu- tions for the A~ signal and the background respec- tively. The fit yields a mean A~ lifetime of </p><p>1 96 +0.23 "fA = . _0 .20X I0 -13S </p><p>and a mass of </p><p>mA, =2285.8 + 0.6 MeV/c 2 . </p><p>f - </p><p>55 a ~- b </p><p>25 </p><p>~" 20 ,~1o </p><p>~&gt; 5 I o I ]R I rl I I 1 </p><p>0 8 16 24 0 4 8 12 Decoy length (mm) Corrected Lifetime (10-'~S) </p><p>Fig. 5. (a) Decay length and (b) corrected lifetime distrib...</p></li></ul>


View more >