137
REVIEW OF PEP EXPERIMENTS Gerson Gol dhaber* Department of Physics, and Lawrence Berkeley Laboratory , University of California, Berkeley, California USA
ABSTRACT Recent physics results from four PEP experiments : Mark II, MAC, DELCO, and TPC are presented herewith. The topics discussed deal with flavor tagging of charmed and bottom quarks, , and D0 lifetimes, Electroweak interference effects , searches for new particles and dE/dx measurements.
*Work supported in part by the U.S. Department of Energy under contract DE-AC03- 76SF00098. 138
I wi ll present the fol lowing data from PEP in my tal k:
l. A brief descri ption of the 4 PEP detectors from which data is presented here. 2. Results on fl avor tagging: Charm tagging via D*+ identification . Results from Mark II and DELCO . Charm and bottom quark tagging via semi leptonic decays . Results from Mark II and MAC.
3. T and D0 lifetime measurements with a Vertex Detector (Mark II). 4. Electroweak interference effects . Results from Mark II and MAC. 5. Search for new and pecul iar particles and speci fic decay modes ; Search for a µ* (MAC). + - Search for e e -+ '¥ + xand e+e- -+ T + x (Mark II). 6. Particle identifi cation in the TPC . l. The PEP Detectors I present herewi th summaries of 4 PEP Detectors as compiled by G. Gidal et al fo r the Particle Data Group at LBL. .1J A. Luminosity at PEP There has been a consideral:ile improvement in Luminosity at PEP. In particular peak Luminosity of 3.3 x 1031 cm- 2 sec-1 have been reached. Average running days have produced over 1000 nb-1. 139
MARK II LOCATION PEP e+e- storage ring SLAC, Stanford, CA, USA
MAGNET 4. 6 kG coil solenoid, 1.5 m radius (currently running at 2. 3 kG) Al
TRACKING Central drift chamber:
Active length = 2.64 m, inner radius = 0.41 m, outer radius = 1.45 m 6 axial layers, stereo layers ( ±3°) JO 50% ethane, 50% argon
= 200 u µ Vertex drift chamber:
Cylindrical drift chamber, 1.2 m long, inner radius = 10 cm, outer radius = 35 cm
Only axial wire layers ( 4 near r � 12 cm, 3 near r � 30 cm)
u = 100 µ Be beam pipe (0.006)(,,)
Combined (D.p/p)2 � (0.015)2 (0.0lp)2 + Tracks extrapolated to interaction pointwithin 100 µ.
SHOWER COUNTERS 8 modules of Pb-liquid argon ( 15)(,, each), arranged in octagon outside coil Covers 64% of 4" 2 Pb sheets separated by 3 mm liquid argon gaps mm 37 layers (0.4X,, sampling) are ganged to provide 6 samples in depth Readout in 3. 8 cm wide strips in u directions
D.E/E = 13%/ VE ¢, 8,
TIME OF FLIGHT 48 scintillation counters read out at both ends Cover 753 of 4"
1.50 m flight path at � 90°; � 340 ps u separation up to 81.3 5 GeV /c at level K," Ju
END CAPS 2 layers Pb-proportional chamber (5X,,) with 4 successive cathode strip readouts R-spiral, L-spiral) 503(8, argon, , 503 ethane
MUON DETECTION Proportional tubes interleaved with steel absorber ( 4 layers each for total thickness of m) I covering 553 of 41f
SMALL ANGLE TAGGING 6 planar drift chambers followed by shower counters LUMINOSITY Octagonal shower counters cover 22 mrad 80 mrad contain MONITOR 18 layers 1/4" Pb and 1/2" scintillator, read< 8 out < with BBQ wave shifter, front 5 layers separately from back 13
D.E/E � 15. 53/ 3 sets of scintillationVE counters
REFERENCES G.S. Abrams et al., Phys Rev. Lett. (1979) 477, and 481. I. 43 ibid 2. W. Davies White et al., Nucl. Instr. Meth. ( 1979) 227. & 160 3. G.S. Abrams et al., IEEE Trans. Nucl. Sci. NS ( 1978) 309, and NS ( 1980) 59. 25 I, ibid 27 4. J.A. Jaros, Proc. Int. Conf. on Instrumentation for Colliding Beam Physics, SLAC-250 (1982). 140
MARK II
,------Vacuum Chamber
Ve rtex Detector
Drift Chamber
me of Flight CountersTi
Solenoid Coi I
Liquid Argon -;--__/ Shower Counter
Im
Shower Counters
Co1I ��S:
End Cop
T1me-of-Fl1ght Sc1nt1llo11on Counters
'-- 4 -, 1 A
Fi g. l 141
MAC (MAgnetic Calorimeter) LOCATION PEP e+e- storage ring SLAC, Stanford, CA, USA
MAGNETS 5.7 kG solenoid, 7.5 cm thick coil Diameter = l m, length = 2.3Al m
17 kG iron toroids, m thick l TRACKING Cylindrical drift chamber 2. 2 m long, 12-45 cm tracking radius
Argon - 10% methane at I atm. 10 layers, double sense wires �5 points on tracks over As:J 95% of 4,,. = 3° stereo gives Az 4 mm = dE/dx to 15% ± up/P 6.5% p =
MUON DETECTION Muon tracking chambers 4 planes of I 0 cm diameter drift tubes surrounding magnetized iron toroids up/P = 30% = 97% of 4,,. An SHOWER DETECTORS Barrel: 14 of Pb - proportional chamber sandwich u /XoE 20%/ vE E = z-coordinate from charge division = 0.80, 1.30 "• <18 = Endcaps: 14 of Fe - proportional chamber sandwich u /XoE 45%/ E = <1>-coordinate fromvE cathode strips 1.50 "• = 20. "8 = Total: As:J 97% of 4,,. =
HADRON CALORIMETER 5. 5 of Fe - proportional chamber sandwich u /EAabs 75%/VE E = u8 2°; u0 I (barrel), 4° (endcaps) = = As:J = 97% of 4,,.0
TIME OF FLIGHT 144 scintillation counters (72 barrel, 72 endcaps) r = 1.3 m At = Ins = 97% of 4,,. An LUMINOSITY MONITOR 4 scintillator /shower counter telescopes at 32 mrad horizontally
REFERENCES
I. R.L. Anderson et al., IEEE Trans. NS (1978) 340. 2. W.T. Ford, SLAC- PUB-2894, March 198225, (Proceedings SLAC International Conference on Instrumentation for Colliding Beams). 142
MAC (MAgnetic Calorimeter)
MAC Detector Components :
CD - Central Drift Chamber EC - End-cap Shower and SC - Shower Chamber (Central) Hadron Cal orimeters TC - Trigger/TOF Scintillators MO , Ml - Muon Drift Chambers HC - Hadron Calorimeter (Central ) Co ils - Sol enoid and Toroid
XBL 831-7898
Fig. 2 143
DELCO
LOCATION PEP e+e- ring SLAC, Stanford, CA, USA
MAGNET Open-geometry (aperture I cosO < 0. 78) Pole-tip diameter 101 cm, separationI 125 cm B0 = 3.3 kG, fBdl = 1.8 kG- m
TRACKING Central (cylindrical) drift chambers: 94 cm maximum wire length, 12-49 cm radius Low mass (2.3% Xo) Depth measurement by narrow angle stereo 16 points(z) on tracks with I cosO I < 0.69
Outer (planar) drift chambers: 285 cm wire length, 160 cm
Multiple hit digital electronics ( 4 ns bin width)
"p/P = v(2%-iJ2--+-T6o/;)2
= I CERENKOV COUNTER I atm. isobutane threshold counter 9. 1) 36 cells each with (pTP-coated) 5" ()',RCA 8854 quantacon Radiator length 55- 110 cm,
SHOWER COUNTERS Barrel (IcosO < 0.62) : 48 Pb-scintillatorI counters, 6Xo
Pole-tip (0.79 < lcosO I < 0.98) : 36 Pb-scintillator BBQ counters, 5Xo
TIME OF FLIGHT 52 counters 324 cm length, 180 cm
' = " 350 ps Acceptance I cosO I < 0.67
LUMINOSITY MONITOR 12 Pb-scintillator BBQ counters, 16Xo Acceptance 25-68 mrad relative to beam axis
REFERENCES
I. W. Bacino et al., Phys. Rev. Lett. ( 1978) 67 1. 2. W.E. Slater et al., Nucl. Instr. Meth.40 (1978) 223. Ouimette et al., IEEE Trans.& NS No.154 I ( 1982) 290. 3. D. 29, 144
DELCO
CYLI NDRICAL DRIFT CHAMBERS
SHOWER COUNTERS
TOF COUNTERS PHOTO TUBE � _r:: MAGNET5" COIL ==�� �=i������ii���� r=
CYLINDRICAL INNER DRIFT CHAMBERS
POLE TIP SHOWER COUNTER
MIRROR 10-82 438988 XBL 831-7897
Fig. 3 145
TPC
2 LOCATION PEP e+e- ring, Interaction Region (IR) SLAC, Stanford, CA. USA
1982-83: 4 kG coil solenoid ( 1.32X., coil package) MAGNET Al 1984: 14.5 kG superconducting coil (0. 86X., package)
Diameter = 2. 15 m, length = 3.0 m
TRACKING Time Projection Chamber (TPC) 2.0 m long (in z) at 20 to 100 cm radius (r) Argon-methane ( &Oo/c-20%) at 8. 5 atm. Max. drift l.O min 20 µSec, 75 kV /m drift electric field 183 proportional wire hits on tracks with !cos 0.71, each wire gives r,z and amplitude for six 60° sectors at each end and provides dE/dx meas. by multiple81 < ionization sampling 15 3-dim. space points from induced cathode signals on several of 13,824 channels to give r, ¢, and z (from the drift time), for I cos 0.71 ?.2 3-d points and ?.1581 POLE-TIP Gas, proJXlrtional mode, sampling Pb-laminate calorimeter CALORIMETER 2 modules, 13. 5Xo deep, at z m, covering 18% of 411" = I.I Argon-methane ( 80o/o--203) at 8. 5 atm; total of 51 samples Three 60° stereo views, each with 13 and 4 samples in depth Projective strip geometry with 8 mrad angular segment •E/E � ± 11%/YE, below JO GeV ±6.0% for Bhabhas at 14.5 GeV HEXAGONAL Gas, limited Geiger mode, sampling Pb-laminate calorimeter CALORIMETER 6 modules, lOXo deep, 4.2 m long at 1.2 m radius Argon-ethyl bromide (96%-4%) at atm. I Solid angle coverage of 75% (90% including PTC) 3 correlated 60° stereo views using wire and cathode signals in 40 samples (27 and 13 samples in depth) Projective strip geometry with 9 mrad angular segment •E/E � ± 14%/ YE, below GeV I ±12% for Bhabhas at 14.5 GeV MLON Magnet flux return 2 layers iron, total 810 g/cm + DETECTOR Triangular, double layer, extruded proportional tubes Al Argon-methane ( 80o/o--20%) at l atm. 3 layers with axial wires and 4th layer at 90 deg. Endcap with 3 layers provides 98% of 411' sterad coverage Resol ution cm, expect 3 mm when operated as drift tube = I TRIGGER 2:2 charged over 85% of 411'" sterad; neutral energy of 2:4 GeV, or energy in two or more calorimeter modules of 2:1.5GeV; 2:1 charged and neutral energy of 2:750 MeV or energy in two or more calorimeter modules of 2:1.5GeV REFERENCES H. Aihara et al., IEEE Trans. NS ( 1983). I. TPC: 30 2. JDC: Gorn et al., I.EEE Trans. NS ( 1979) 67. W. 26 3. HEX: H. Aihara et al., IEEE Trans. NS ( 1983). 30 4. MLON: Bakken et al., IEEE Trans. NS ( 1983). J. 30 TRIG: M. Ronan et al., IEEE Trans. NS (1982\ 421 5. 29 f-"' ... O'I ni:1>.Mf MAGNET POLE BASE SUPPORT BLOCK OUTER ORIFT CHAMBER MAGNET POLE TIP SUPER-CONDUCTING COIL l TIME PROJECTION CHAMBER POLE TIP CALORIMETER L., 1� _...... - ...... _, � -0 u:i t"'j -!'> 0 CQ!l CRYOGENIC t INES o' I�I PEP-4 FACILITY 147 29 GeV PEP Data 3 2 II Mark «> = 0.59 :': 0.05 N > 0 Q) (.9 (b) rt.I ..0 4 bottom ::i. :co 0 3 ""Obl N -0 en j 2 . 0 0 0 2 0.4 0.6 0.8 1.0 0. 0.4 0.6 0.8 = 2 z ( 2EoH /Ec.m.) 430!lA3 l 15 XBL 836-196 Fi g. Fi g. 17 1.00-r--�- 0.75 0 50 0.25 - _____J______J_____J_ __L_ 0.00 0��-�-�-� .__l_05 .• x=l\./E� XBL 835-9661 ri g. 16 Fi g. 15 Channed quark to DH fragmentation function. 0.59 + 0.05. - 2. Fl avor Tagging There are two fortunate circumstances which allow flavor tagging even though complete particle identification is not always available. These are charm tagg ing via D*+ identification and charm and bottom tagging via lepton identification from semileptonic decays. Charm Flavor Tagging via D*+ Identification D*+ identification is possible even without K-identification due to the very n+ tight kinematical constraint in the decay D*+ + n+D0 where the bachelor has a kinetic energy of 5.7 MeV in the Drl c.m. sy t �+ This feature is expl ited t �� � by plotting the mass difference � = M{D*+) - M(D0} which gives a peak centered at · K-n+ � = 145.4 + 0.2 MeV. Here the D0 is primarily observed through its D0 + (Br = 3.0 0.6%) decay mode, and if K-identifi cation is available (DELCO} through � the D0 K-n+n+n- (Br = 8.5 2.1%) decay mode. Furthermore a charm associated ±. "satelli+te enhancement" S0, discussed below, can also serve in D*+ tagging. A Charm Associated Enhancement As illustrated by the Mark II data from SPEAR in a K-n+ mass plot aside from - the D0 peak at M(K n+) = 1.863 GeV there is a second broader charm associated "satellite" peak S0 with M:: l.61 GeV and 0.12 GeV .(see Figure 5). . :::_ These val ues are approximate since the S0 renhance ment is skewed towards low mass val ues. The mass interval 1.48 - l.68GeV contains about 90% of the enhancement. Figure 6 shows that the S0 enhancement persists at the various SPEAR energy 1 ', regions: 3.77 GeV the o/ 3.9 - 4.5 GeV and 4.5 - 5.5 GeV. Furthermore it is 1 absent at the o/ and o/ energies . This indicates that the S0 satellite enhance ment is charm associated and we interpret it as primarily due to the more copious decay mode D0 K- n+n° , B = 9.3 2.8% with perhaps a small contri bution from the + r ±. D+ K-n+n+ decay mode. The way this comparatively sharp S0 enhancement arises is +ill ustrated in Figure 7 which gives the features of the Dalitz plot for the K-n+n°. decay D0 + We note that this decay mode proceeds via the two intermediate state {pseudo scalar meson + vector meson) channels: and D0 + K*-n+ as wel l as a third channel D0 K*0n° (l) K-no + K-n+ 4 4 whkh is of no relevance here and is not shown on the Dal itz plot. In view of P - P - P the fact that we have the decay of J = 0 {the D0 ) to a J = 0 and J = , state it must proceed via LP = , - in relative orbital angular momentum between the pseudo scalar and vector meson. As a consequence of the angular momentum addition + (V) = 0 the vector meson is produced ful ly al igned and hence with L J a cos2e distribution in the vector meson c.m. This expresses itself as a mass squared distribution along the p+ and K*- bands which peak at the ends of these 149 100 3.77 GeV 2.0 3.77-6.7GeV data (a) r0 60 Qx 1.0 0 2.0 -20 r0 300 :::;:..Q) Qx 39-4.5 GeV 1.0 � 200 0 N ...... 0 +-Cf) c 100 (c) Q) 600 SUBTR. > N u BKGD. w ...... � 400 so � 0 C\J 45-5.5 GeV ...... If) 200 1 50 cQ) > 0 w 0 50 1.2 1.4 1.6 1.8 2.0 2.2 .. M(K- rr + ) GeV/c 2 14 1.6 1.8 2.0 22 XBL833-1329 Fi g. 5 M (K-7T+) GeVXB L 833-1 328 Fi g 6 . Fi g. The resonance and S0 enhancement. The signal s correspond to 1340 5 - events0° and 1470 S0 events. The S0 signal thus more then doubles the0° number of charm tags for the decay mo de. (a) The curves is a fit to background and the si9K-n+nal . (b) The curve is a fi t to background and the S0 signal 0°. (c) The data with background (as determined from the above fits) subtracted. Mark II at SPEAR. Fig. 6 The D0 and S0 signal s for 3 energy regions as noted with - backar· ound subtracted. Mark II at SPEAR. 150 + 0 p _.,,.+ if N " Sate! lite" (.'.)>w 2.0 \ l .l48f BGeV IY'. 1 .5 =S0 enhancementI N::;;: 1.0 0.5 0.5 LO L5 2.0 2.5 3.0 3.5 4.0 M2 (K"j,-+) GeV2 0.7 LO 1.4 L5 1.6 L7 M(K7T) GeV Fi g. 7 The Dalitz plot boundary for the reaction sho�iing the - and bands. The enhancement corresponds to the high end of these resonance bands5° . Both M2 and'[i�'":'." scalK-n+n°es are shown . p+ K*- M K-n+ 8 Mark 11 spear o0-.K-P+ 300 Ecm = 3.7 - 6.7 GeV L.. 1T'o 1T "' 4 ---67% p++33% K'- .µ 200 Monte Carlo <: :::> >, 0 S- "' S- .µ 100 1.2 f o0- 7T+K� <( l+.K- 1To V1 I- .8 z w > w 1.3 14 1.6 1.7 1.5 .4 (K-7T+) GeV/c2 M Fig. 9 o0.8u_�---'��---'-��..__'----'-----'--'--� 1.0 1.2 1.4 1.6 1.8 M GeV (Kn-•) Fig. 8 XBL 835-195 Fig. 8 The calculated shapes of the mass projections of tile and - bands. The observed enhancements are the result of the vector meson alignment from 0° decay.K-n+ p+ Fig. TheK*- signal fi tted by a combination of the and distributions. - g Mark5° II at SPEAR. p+ K*- 151 bands. Fi gure 8 shows these as reflected into M(K-TI+) evaluated by a Monte Carlo calculation. Figure 9 shows a possible fit to the 5° peak with 2/3 of the intensi ty ascribed to the p+ band and 1/3 to the K*- band with no account taken of possi ble interference between the two bands. Thus the 5° peak is explained as an en hancement in M(K-TI+) , primari ly due to the D0 ->- K-TI+Tio decay mode. 2 •3] The associ ated in the relevant portion of the Dalitz plot is of rather TI0 low momentum in the D0 c.m. system. This implies a rather low TI0 detection ef ficiency and procl uded the direct observation of the 5° enhancement in the Mark II study of the D0 ->- K-TI+Tio Dalitz plot. 3J What is important in what follows is that since the 5° peak corresponds to a 0 low momentum TI0 (in the 0° c.m.), mass difference Lio = M(·rr+o ) - M(D0) which has an experimental full width of "' 2 MeV in the Mark II at PEP, is still + appl icable as Lis = M(TI S0) - M(S0) although the full width now broadens to "' 10 MeV. Figure 10 shows a Monte Carlo calculation at 5.2 GeV incl uding the detector resol ution. The cal cuation at 29 GeV is very sim'ilar and is in good agreement with the observed experimental di stributions. D*+ Studies at Figure 11 shows .P-EPthe Li distribution for the Mark II data at 29 GeV, here a cut on M(K-TI+} of l .76 - 1.96 GeV was appl ied to select the 0° band. The Li distributions are shown for the values 0.4 and the more restrictive i! i! > 0.6. Here = E(D*)/E Figure 12 showsi! the > M(K-TI+) distributions for a Li i! beam· interval of 143 - 147 MeV. Physics Results from D*+ Studies There are four physics results which are obtained from the identified D* mesons. (i) The charmed quark fragmentation function. One finds that the charmed quark fragmentation into D*'s occurs at considerably higher z values then for the light quarks. 0.6. (Mark II and DELCO.) {ii) The cross section for D*+ produc R = 3 4 1 (2) C Cg + 9) for charmed and bottom quark pairs (assuming the latter primari ly proceed via charm decay). Here, 2 Rc = 2 x = 3.33 is expected fo r the incl usi ve charm production R val ue (both� charges and all charmed mesons and Baryons). (Mark II and DELCO.) 152 2 15 16 6=143-147 MeV/C 14 Z> 0.4 - 16 (a) 12 14 z>0.4 10 12 VI JO 8 10 ·E::::J 6 5 68 � 4 C\J _g C\Ju 4 <{ '-' 2 :e 2 0 �"' � 'a; 15 � 8 ' 111 11I1�1� 1 � 1 � (b) 11�z>0.6 1 �I 11 0 8 Z>0.6 ' ';;:; 7 -E"' 7 VI 6 6 10 tij i "E"' Il> w> 5 5 w 4 4 5 3 3 2 2 I I 0 0 140 150 160 0.5 6.MeV 0.13 XIJL 835-197 XBL833-T331 XBL 833-1330 Fig. 10 Fig. 11 Fig. 12 Fig. 10 (a) The � signal from D*+ decays with M(K-n+ ) in the 0° region. - (b) The � signal from D*+ decays with M(K-n+) in the 5° region. Monte Carlo calculations at E = 5.2 GeV. Results at 29 GeV are essentially the same. cm Fig. 11 Experimental � distri butions at E 29 GeV, for M(K-n+) M - c 0o l GeV/c2 and 2 cuts on Mark m = at PEP. 0. 2. :t:_ 1I Fig. 12 Experimental M(K-n+) distribution E 29 GeV for 2 cuts on The - 0° signal is cl ear while the 5° is smal= l since a narrow � region2. , 143-147 MeV, is shown here. Mark II at PEP. 153 Gev M(D•-DO) XBL Fig. 13 (b) 835-9650 + = Fig. 13 - Scatter plot of sin en* ' vs . for M(K-n ) M00 0.3 GeV/c2. This data thus contains botnnR 0° and6 5° signals. and/or� by Cere�kov identification: (a) 11right sign" (b)K 11wrong nsi gn11 i.e., background. (DELCO). n8 ; n8 ; cr(D*+} = 0.18 0.04 0.06 nb from these data. � � 154 M0• -M00 K3n Mode(Right Sign 7T , ) 0 20 + + 0 15 0 10 0 05 0 DD 0 15 Gev M(D•��DO) XBI 835-9655 a) ( D• 00 M -M K3rr Mode( Wrong Ci1gn rr, ) 0 20 + + 0 15 0 10 "'i � � � i1 0 05 I ___)11 OOO 0 15 02 Gev L M(Dll<-DO) XBL 835-9665 ( b) - Scatter plot of sin 8 vs . for -* K-rr+rr-rr+ Fi g. 14 D* , t:. decay mode. ( DELCO) ns 0° + a(o* ) = 16 .:!:_ 0.06 0.09 nb from these data . o. .:!:_ 155 (iii) The D*+ gives tagged D0 events which are used for a D0 lifetime measurement in the Mark II Vertex detector. (iv) The angular distribution and asymmetry of the charged quark distribu tion relative to the e+ direction (so far only results from TASSO are available, see talk by Yamada at this conference). The mass constraint [;,which characterizes D* events, can be expressed differently as well. The kinetic energy of the bachelor pion 11� in the decay DH _,_ 118 D0 has a kinetic energy of T; � 5. 7 MeV in the D* rest system (or a momentum of p* "' 40 MeV/c). This irnp ies that p* is the maximum transverse TI � momentum 118 ca have in the laboratory system. i.�e. , sin * where is � p 11B -< p11B the laboratory angle between the D* direction and the directG ion. G 1[B This feature was emphasized by the DELCO experiment. In Figures 14 and 1 5 show a scatter plot from DELCO of sin vs. [;,. Typical p 11B values a re " 400 MeV/c hence all D* events should occur forG sin e� 0.1. As can be noted from Figures 14a and b this is indeed the case. In the DELCO experiment the Cerenkov counters were used for and/or 11 identification and thus they were able to separate "right sign 118"K candidates, i.e., 118D0 from "wrong sign 118"events i.e., 116-D0• The latter can be considered as a measure of the background. Furthermore in view of the Cerenkov identifications the DELCO experiment was also able to study DH candidates with the D0 decay mode D0 _,_ K-11-·/11+, see Figures 15a and b. The Charmed Quark Fragmentation Function The results that emerged from the Mark II experiment 4 ] are that the heavy quark fragmentation function D peaks at considerably higher z values then those for light quarks. The qualitati� ve features are, that just like in the case of 11 ° decay for example, the heavy particle (proton) carries a large fraction of the 11 ° momentum. In our case the charmed meson carries off a large fraction of the charmed quark momentum. This idea has been cast into a quantitative form by Peterson et al. 5J who suggest the expression: DH = A (3) Q z [l - l/z - EQJ\l - z)J2 where A is a normalization constant and Eq is a constant characteristic of the individual quark mass. Ec ":. 0.25 g·ives a reasonable fit to the data for charmed quarks. The results from the DELCO experiment are very similar. See Figures 15, 1 6 and 17. 156 Semi�tonic Charm and Bottom Decays Prompt lepton production in hadronic events from high energy e+e- annihila tion provides a tag for the presence of hadrons containing charm (c) or bottom (b) quarks . The production rates and momentum spectra of such leptons depend on the semileptonic branching ratios and the momentum spectra of the parent hadrons. Furthermore the momentum spectra of these hadrons provide information on the fragmentation properties of c and b quarks. As in the case of charmed quark fragmentation the b quarks fragment predominantly into high momentum hadrons. A. Studies in the Mark II Detector In the Mark II analysis the total momentum and transverse momentum spectra of prompt electrons in hadronic events are measured. The transverse momentum p . .L is measured with respect to the thrust axis defined by all the charged particles in the event. The hard p� distribution of electrons from bottom decays relative to chann decays allows one to separate the contributions of b and c quarks to the prompt electron signal . The Mark II data, corresponds to an integrated luminosity of 35 pb-1. The electron-hadron separation algorithm is based on measurements of the ratio ri _ Ei/p, where Ei is the energy deposition in one of three groupings , i = l - 3, of layers in the Mark I I Liquid Argon-Lead calorimeter. Each of these groupings combines all layers in the first 8 radiation lengths which have the same strip orientation. To minimize the effects of neighboring particles , particularly photons, the energy deposition Ei is taken from a narrow lateral region, comparable in size to one stri p width 4 cm) , centered about the extrapolated particle (:::_ trajectory. The algorithm demands that each value of ri and that i:ri be greater than an appropriate rninimum val ue. The electron identi fi cation efficiency was detennined with electrons from Bhabha events and photon conversions. This efficiency varies from 78% at l GeV/c to 93% at the highest momenta. The probability , as a function of and pL , that a hadron wi ll be misidenti fied as an e 1 ectron was determined from:p Hadron nteracti ons in the ca 1 ori meter ·j \ exposed in pion test beam as well as with pions from the decay '!'_,. 2( r- )11° in a 11 data from SPEAR. Accidental overlap with nearby photons was estimated from jet studies. The misidentification probabilities are typically 0.5%, but can be as large as 35; for a track momentum 1 GeV/c in the core of a jet. A hadronic event sample was selected by requiring Nch 5 and total detected energy E-+ E0 E /4. There were a total of 10691 such eve."._ nts. To eliminate + > cm real electrons which are not par·t of the prompt signal , a visual scan was per formed to remove a small number of electron candidates arisinn frnm: 157 D b-c -e± Background 0 c�e± b-e± D • lb) 0.5 < < 1.0 (a) (bl p� 4 120 P{0.5 GeV/c VJ f z � 80 0 w 40 uJ{;: VJ co 0 (d) p�, 1.5 0 0 LJllliili!ii:i..,..iliiil,.l___..=o1�•._•ll.i..J4 2: 6 IGeV/c) 4 6 Fig. 18 0 0 2 4 60(GeV/c) 2 4 6 p Fig. 19 (b) (al GeV/c p>2 4 u"- bu 2 + + + _ 0 L_J__l_j__L_L__i::.='="=J 0 l__L__l____j___J _b>,=<::lt::1 0 2 0 2 4(Ge V/c)6 8 (GeV/c) 3 p p l_ Fig. 20 Fig. 18 - Observed prompt electron momentum spectra in 4 regions of transverse momentum p .L . Background contributions unshaded. (Mark I I) Fi g. 19 - Prompt electron momentum spectra in 4 regions of transverse momentum p.L . Two sets of error bars are shown for each data point. The smaller ones are statistical only. The larger ones are the statis tical and systematic errors added in quadrature. The highest momentum bin incl udes all momenta 6 GeV/c. The histograms show the results of the fit. The three con> tributions shown are (i) b primary (solid), (ii) c secondary (di agonally hatched), and (iii) c primary (unshaded). (Mark I I) Fig. - Differential cross sections for prompt electrons with p 2 GeV/c: 20 (a) total momentum and (b) transverse momentum. Two sets> of error bars are shown for each data point as in fi gure. (Mark II) 158 pair production ; T 0 beam-gas interactions; the process e+e- e+e- hadrons; � + background electrons from photon conversions and Dal itz decays were removed by a pai r finding algorithm, which removes about 70% of the electrons. After removal of these backgrounds on an event by event basis, 930 electron candidates with p 1 GeV/c remained. These were partitioned into 24 p, p.L > Background contributions �� bin from misidentified hadrons, and electron lQ_ pairs was calculated. In Figurel8 the observed electron spectra are shown. In three of the lowest momen tum and transverse momentum bins [Figure 18a: p 3.0 GeV/c and Figure 18b: p 2.0 GeV/c] backgrounds account for almost 75% of< the observed signal . In the rem< aining bins, however, the background level s are much less severe. Exel uding these three bins, the background fractions for Figure 19 are 46%, 36% , 30%, and 23% respectively. Figure 20 shows the total momentum and transverse momentum differential cross sections for prompt electrons with p > 2 GeV/c. Two sets of errors are plotted for each point. The smaller error bars represent the statistical errors. The larger ones show the quadratic sum of the statistical and systematic errors. The systematic errors are dominated by the uncertainty the hadron misidentification probabilities. The total inclusive cross section"in fo r this sample is 14.4 1 .6 5.2 pb. ±_ ±_ A maximum likel i hood fit to the observed populations in the various p , bins was performed leaving out the 3 high background bins and accounting forP theL signal above background in terms of the fo llowing contributions : (i) Bottom decays in bo events (b primary) (ii) Charm decays in bo events (c secondary) (iii) Charm decays in c�c events (c primary) To represent contri butions (i) - (iii) a Monte Carlo simulation with a Feynman 6 Field hadronization model and gluon radiation as incorporated by Al i et al . ] was performed. The semi leptonic electron spectra produced by the heavy meson decay models in the Monte Carlo agreed satisfactorily with measured spectra from the DELCO (charm) and CLEO (bottom) experiments. + In the fit sc was fixed at 0.25 in accordance with the D* results given above. The remaining parameters were : Be(b) = (11.6 ±_2.l _:!:1. 7)% Be(cl 1.2 2.1)% and = (6.3+ !0. 032 ! 0.023 = 0 030 + Sb ' - 0.018 - 0.014 159 The histograms in Figures 18 and 19 show the results of the fit and the relative contributions of (i) - (iii) to the prompt electron signal in each p, pi bin. The x2/D. F. of the fit is 14.0/18. Here we must note that the Be(c) and Be(b) val ues are averages over charged and neutral branching ratios wei ghted by the corresponding experimental production ratios. Thus in the case of charmed particl es, since D+ and 0° are primarily produced via D*'s and since only the D*+ feeds the D+ mesons a o+ / 0° production ratio of to is expected. A different "' parameterization was also studied namely the form i!a(;!-1), which gives qualita tively simi lar results. For either of these parametej rizai- tions, the average value of z is 30 150 (f) r- 20 z w 100 > w 10 - /-� � - _ (zb)"0.45 0 � 2 4 6 8 10 P� IGeV/c ) p (GeV/c) Fi g. 21 Fig. 22 10 8 4 2 o L.1.-.-�=±=±!':i::�=._�i___...J__.J 0 0.2 0.4 0.6 0.8 1.0 Fig. 23 Fig. 21 - P.1. spectrum of muons with bb (dashed curve), cc (dot-dashed), back ground from decay and punch-through (dotted), and total (solid curve) predictions. (MAC) Fig. 22 - Total momentum of muons with p.1- 1.5 GeV/c. Dashed curves are the be5t fits obtained with the b > fragmentation fixed to a narrow range of The sol id curve is the best overall fi t. (MAC) z. Fig. 23 - The shaded region is the enve l ope of acceptable b quark fragmentation functions. The curve represents the Peterson et al. function for E 0.008. (MAC) b = 161 A Monte Carlo model was used to fi nd the level of background remaining in the sample from � and K decays. This was found to be (23 1)%. � The punch-through background was determined empirical ly from hadronic decays of tau leptons. From a sample of 1600 tau pair events, roughly 3 muon candidates in hadronic jets attributable to punch-through were found. Differences between the energy spectra of tau and multihadron jets were accounted for by taking the ratio of the energy deposited in the outermost calorimeter layer in the two sampl es. From these studies, the punch-through fraction in the incl usive muon sample was fo und to be (9 � 7)%. Figure 21 shows the spectrum of the observed muons, along with Monte Carlo predictions for muonsP J.. from b and c quarks, and the overall {decay pl us punch-through) background prediction. The background is concentrated at low val ues of pJ.. , and is wel l separated from the bb predicted spectrum. Figure 22 shows the momentum spectrum for muons with > l.5 GeV/ c, the regions containing the highest fraction of bb events. ThePJ.. da shed lines illustrate the effect of fixing the b fragmentation function at particul ar val ues of and z, allowing the c fragmentation function and the semileptonic branching fractions of both quarks to vary to obtain best fi t to the p by pJ.. array. Low val ues of The solid curve is the predicted spectrum for the Cylindrical Aluminum Shell Beryl lium Vacuum Chamber Inner Layers Outer Layers4 3 Fig. 24 - The Mark II Vertex detector. MARK II-PEP ,,, lb) ,,, + - - 25 - (a) e e ,+, as seen the Mark II detector at PEP ; (b) close-up Fig. of the same+ event in the "invertex detector; (c) extreme close-up of the same event showing particle trajectories extrapolated to the vicinity of the interaction point. 163 at about 30 cm. To keep multiple scattering to a minimum the chamber has been built directly around a beryllium beam pipe 0.6% of a radiation length thick. The beam pipe serves as the inner gas seal for the chamber. The average resolu tion per layer in hadronic events is about 100 µ. A study of the measured distance between the two tracks in Bhabha events after extrapolation to the origin give a distribution of the distance between them - with An integrated luminosity of 20 pb 1 was accumulated at E = 29 µ. cm GeV witha� thi10 0 s chamber. T Lifetime The lifetime provi des a sensitive check of the standard model of weak interactiTo ns. With the assumption that decay proceeds in direct analogy wi th muon decay, and that neutrino is massless.T The tau lifetime can be related to the muon lifetime, namT ely: = (4) T T Experimental ly B(• evv) 17.6 1.1 % (5) + = .:!:_ so 2.8 + 0.2 l0-13 (6) T = X S T - where the error refl ects the uncertainty in the electronic branching ratio. A number of models have been considered under which this simple prediction could fail8,9,lOJ . Furthermore the tau lifetime would be extended if the tau neutrino were massive enough so as to significantly limit the phase space of the decay. Tau leptons are pair-produced in e+e- annihilations, so that each tau has the beam energy. Thus the lifetime can be measured by determining the average decay length of the taus. At PEP with E 29 GeV, it is expected to be about 700 cm µ. The decay length can be measured when= the tau decays in the three-charged-prong topology. It is the distance between the production point, i.e., the beam position, and the position of the decay vertex. Events are selected in which at least one of the taus has decayed in the three-charged prongs topology and the total charged of the prongs is zero. The three particle invariant mass is required to be in the range 0. 7 m 1.5 < 3'IT < GeV/c2, and tau pairs produced by the two-photon process are rejected. Fi gure 25 164 ___ I 5 ,------,,------,-- ---,--,--T ' 30 60 EE 1 0 (f) l !zw 20 40 > w 10 10 - 1.0 6x 0 1.0 -1.5 -1.0 6y 0 1.0 1.5 0 Imm) Imm)� J 0 I 3 o ERROR (mm) DECAY LENGTH Fig. 26 Fi g. 27 6 � � ' 4 g '"'3 a � -4 -2 0 0 DECAY LENGTH (mm) 0 0.1 0. m 0.3 0.4 2 v, 0.5 0.6 (GeVlc2) Fig. 28 Fig. 29 Fig. 30 Fi g. 26 - Hori zontal and verti cal hadronic vertex po sitions rel ative to the beam position. (MARK II) Fi g. 27 - Calculated error in the decay length. (MARK II) Fi (). 28 - Measured decay lengths. (MARK II) 29 Fi g. - Va riation of the T lifetime V1ith the mass of the T neutrino. (MARK II) Fig. 30 - Measured proper times for the seven events. (MARK II) 0° 165 shows such an event in the Ma rk II detector. The rms beam size at PEP is 500 horizontally and about 50 vertically. µ µ The average beam position is remarkably stable from one fill to the next. This was measured by finding the average intersection point for an ensemble of well measured tracks. As a cross check, this determination of the beam position was compared to the vertex position measured in hadronic events. Figure 26 shows that these methods agree. The width of the �x distribution is consistent with the known beam size, and the width of the �Y distribution is consistent with the vertex resol ution, indicating that the beams are stable. The decay vertex position and its error ell ipse are determined from the three pion trajectori es and their associated errors with a chi-square minimization procedure. Events with a vertex chi-squared per degree of freedom greater than 6 were excl uded. The best estimate for the projected decaylength is then given in terms of the decay vertex position relative to the beam position (xv,yv), the sum of the beam and vertex error matrices (cr ), and the direction cosines (t ' ij T x ty) by the following expression : (7) The tau di rection is wel l approximated by the direction of the 3TI system. Then the decay length is: (8) where is the total momentum of the three pion system. p31T Fi gure 27 shows the cal culated error in the decayl ength, which depends on the opening angles and orientation of the decay, the tracking errors, and the beam size. The measured decay lengths are shown in Figure 28, where only those events with decaylength errors less than 1.5 mm are included. The mean of the distri bu tion is obviously positive and its shape is asymmetric. The d·istribution is fitted with a maximum likel ihood technique which takes the decaylength error into account event-by-event. The fitting function is the convolution of the Gaussian decay length error with an exponential decay distribution. The average decay 166 length is 710 120 2:. µ. After correction for hadron contamination and initial state radiation, this = + 0.57 + 0. 60 10-13 yields: 3.31 - x s, where the first error is the statisti- T - cal error Tand the second is the systematic. The measurement is compared to the other measurements which have appeared in the literature in Table The number of decays studied and the average decay 1 ength error are al so shownI. for comparison. Table I Measurements. T Le[>ton Lifetime Average Decay (1 Experiments Number of Length Error T s) Deca s mm T o- 13 TASSO 5gg 10 0.8 + 2.2 MARK 126 4 4.6 + 1.9 MAC II 280 4 4.1 +1.2 + 1.l � 3.9 78 6 4 . 7 CELLO - 2.9 MARK II 71 0. 9 3.31 + .57 .60 Vertex Detector + Theory (Universal coupl ing strength) 2.8 + 0.2 All of the measurements are consistent with the expected lifetime. The present experiment confirms that the tau couples to the charged weak current with universal strength wi thin the errors, 0.92 0.086 0.090 (9) gT/ge = + + If one assumes that the coupling has the universal strength , one can set a limi t on the mass of the tau neutrino. Figure 29 shows how the lifetime varies as a function of the tau neutrino mass11] al ong with the 90% confidence level upper 0.49 limit set by this measurement. One finds mv GeV/c2, which is compatible 1•ith the 0.25 GeV/c2 limit deduced earl ier ftom< the shape of the decay lepton 12] spectrum . 167 Lifetime As0° Kalmus13] reported at the Paris conference, charmed meson lifetimes seem to be stablizing as follows : The lifetime is about 4 x 10- 13 s, about half the o+ lifetime of 9 x lo- 13 s. 0° The lifetime was measured wi th the Mark II vertex detector using a sample of D0's wh0°ich had been cleanly identified by the main tracking chamber. At the moment the main virtue of this measurement is that it is subject to very different systematic errors from the earlier measurements. Like the other, measurements however it also suffers from low statistics. integrated luminosi ty of 17 pb- 1 gave seven wel l-measured decays. An 0° The clean event sample comes about by selecting D0 's from D* decays. For the lifetime analysis below, events with 0.6 were selected. They seem to be essenti ally free of background. The high mom� > entum cut also makes it unli kely that these D*' s have originated from decays. B The decay vertex is determined by vertexing the K and n from the The decay length is then determined with the same technique used in the tau0°. analys is above. The average decay length is about 500 µ, and the average vertex error about 700 µ. Figure 30 shows each of the 7 measurements with its error. A simi lar maximum likelihood fit as for the ' above gives a lifetime of 0° 3.7 1.0 x l0- 13 s (10) •oo ±. ±. �:� This result is in agreement with the current world average. Limits on the Meson Lifetime 8 If their decays were not suppressed , the hadrons containing the b quark would be expected to decay with a lifetime m . µ • = lo- 1s s 'µ s (11) ,o "' m l9 B ( b ) The (qu�l itative) factor 9 comes from the number of available final states . The meson lifetime is expected to be longer, however, since the decay is suppressed Bby the presumably small mi xing between the second and third quark generations. The meson lifetime thus gives a measure of the quark mixing, at least if one assumesB universal weak coupl ing strength14J. 168 Two experiments have obtained upper limits for the B lifetime by measuring the average impact parameter of energetic muons. Events are sel ected which have high mass jets and an identified high-momentum muon as discussed above. Using the thrust axis as an estimate of the B meson di rection , a signed impact para meter can be determined from the lepton trajectory and the known beam position. The upper limits obtained are as follows: The JADE experiment finds T < B 1.4 x 10- 12 s and the MAC experiment finds ' 3.7 x 10-12 s, both limits at the B < 95% confidence level . Thus there is as yet no evidence for a B lifetime accessi ble to current measurements. A Mark II measurement is in progress on the electron tagged events discussed above. When compl eted , this should have a sensitivity of 3 x 13 sec. lo- 4. Electro Weak Interferen ce Effects . As shown in Yamada 's tal k at this conference the results from the five PETRA Groups : TASSO, JADE , MARK J, CELLO, and PLUTO make a very convincing case for Electro-Weak interference. In this respect the higher energy available at PETRA, mainly 34 GeV, compared to the energy at PEP, 29 GeV, hel ps . Tables II and III show the corresponding results . Here we must note that Mark II uses a more limited interval in cose viz. lcos el < 0.7 wh ile MAC uses nearly ful l co s e range. Figures 31 to 34 show the rel evant angul ar distributions. The above experiments compared thei r data to the QED cal culation wi th a3 1 5J corrections according to Behrend and Kl eiss . TABLE II The Data Sets Electro-Weak Interference Effects Luminosity Number of Bkgrnd. % Detector Particles pb- 1 Events Mark II 50.0 2703 2.4 jcos 0.7 µµ EJI < 46.6 1607 14.8 TT ee 50.0 40,989 0.06 MAC 0.94 µµ 40.0 3067 lcos el < 4.2 ee 40.0 169 60 · ...... · . .. ·· ·.... 40 · . . --"""'-"� .c::;" * �· · 5 0 20 ' -li. " 1I 0 L-L.��-'-�--L�L__--'-.�-'-� -1 -0 5 0.5 -�1 cos 6 • 29.0 CEV ECM lll. 135·9151 Fig. 31 - Angular distribution for µ pairs. Solid curve is fit to data. Asym (-4.4 -+ 1.9)%. Dotted curve is QED with corrections. (MARK= II) a 3 60 ::; 40 .c" 0 0 " 20 � 0 -1 cos (J • GEY Et'M 29.0 lBL 835-965Z Fig. 32 - Angular distribution for , pairs. Solid curve is fit to data. Asym (-4. l 2.7)%. Dotted curve is QED with correction. (MARK = I I) .:!:. a 3 170 80 N 60 l? :;-;I D c 40 "' 20 0 �-�----'---L__ -1.0 -0.5 0 0.5 ____J 1.0 cose Fig. 33 - Angular distribution for µ pairs. Sol id curve is fit to data. 6 Asym = ( - 7 . 1.8)%. .:!:. (MAC) du s an 103 (nb GeV1) ( b) 1.1 2 sin 8,, := 0.24 du exp 1.0 � auQ ED + 0.9 0 0.2 !co0.4s(O)! 0.6 O.B XBL 835-9660 Fi g. 34 - (a) Angular distribution for Bhabha scattering. (b) Ratio of experimental to QED distributions. Curve corresponds to fit with 2 sin ew as given. 171 TABLE I II Results From Electroweak Interference Effects. Results Mark II MAC e µ ga g a 0.25 ±.o.11 ±. 0.02 0.31 ±. 0.08 e gagTa 0.23 ±. 0.15 ±. 0.02 g2 0.03 0.05 ±. 0.03 v ±. sin20 0.22 ±. 0.06 0.24 0.085 w ±. g2 (al 1) 0.22 0.07 0.02 a ±. ±. 5. Searches for New and "Peculiar Particles ." Many searches which have been made for new particles have been presented at the 1982 Paris conference or have been publ ished earlier. Here I will only touch on a few new or augmented results. Study of µµy and Events µµyy+ - µ*µ and e+e- µ+*µ-* the MAC Group has presented In a search for e e + + n from these m(µy) and coseµ distributions, see Figures 35 to 37. The conclusio measurements is that the results are consistent with 3rd and 4th order QED (the curves in the figures). It is interesting to note that due to interference between initial and final state radiation effects purely QED processes give rise to very considerably asymmetdes in the µ angular distributions. Figure 37 shows a plot of Mµ-y versus Mµ+y for the µµyy events. There µ*µ* production would show up as an enhancement along the 45° axis of this fi gure. The dotted lines show the experimental uncertainty in this region. This places a limit on - cr(µ*µ*)/o(µµ) 2 x 10 3 for 2 < Mµ* 14 GeV/c2 at the 90% CL. < - < Search for e+e '¥ + x and e+e- + x. A search for '¥ and+ production via+ Tdecays into lepton pairs was made in the Mark II detector. This isT based on a total luminosity of 35 pb-1 yielding 6276 hadron events which were restricted to have their thrust axis within Jcosef 0.7. Lepton pai rs were selected as fol lows : < At least one lepton had p 1 GeV/c relative to the thrust axis; Both lepton have momenta ran> ging between 1 .4 - 11 GeV/c; Only pairs of ee or µµ were accepted. With these criteria the efficiency for '¥ detection is estimated at 1.5%. 172 60 30 20 40 N' '-'.:: 10 20 � -.::::-0(.'.) 0 �-·�-�-�-�-�-._, 0 ______"" 8 1--- -_J__ _J__ _L_ U) - f- 6 z w 6 > w 4 4 2 30 10 20 -0.5 0 0.5 LO M/LY (GeV/c2) Fi g. cose Fi g. 36 35 30 I I '.il'.l - I1\ I (a) g I II . . 4 I I 20 -· .ci' /I I / "' 0E 2 % . . u -- \ (.'.) . . 0 _.,I / <1.0 2.6 3.0 3.6 3.8 4.2 4.6 / "'j_ 10 '-- / ' . . / '� :;> . / '.i /. , \{)"' / [ M (ee,µµi (b) 1+r / .-, 4 / . f! •//. .- ' ' 2 0 /• / . .ci � . . I E0 0 10 20 30 u 0 7 9 II Mµ + y ( GeV/c2) 13 15 17� Fig. 37 4-83 M(+Q- (GeV) \ 44T!A3 XBL Fi g. 38 835-9664 Fi g. 35 - (a) The m(µy) distribution for the µµy final state (282 events). The sol id curve is the order QED predi ction . (b) Distri bution of m(µy) for the µµyy fia3nal state (21 events ). The curve is the QED cal culation. (MAC) a4 Fig. 36 - (a) Polar angle distribution of muons for the µµy final state Asym = (-21 .6 + 4.1)%. The solid curve is the order QED prediction Asym(QED) = (-21 .1 + 1.3)%. (b) Muona3 angul ar distri bution for the µµyy final state Asym ;- (-38 + 14)%. The curve is the order QED calculation Asym(QED) = (-36. 6 4.8)%. (MAC) a" .:1:_ - F·ig. 37 - Scatter plot of the invariant masses of µ+y and µ y combinations for the µµyy final state. Dashed lines indicate a range of two standard deviations of the µr ma ss resolution about the 45° line. (MAC) Fig. 38 - - Search for e+e -> x and e+ e x at E = 29 qi + - + GeV. (Mark II) -> T crn The shaded events correspond to the expected invar.i ant mass distribu tiqn from a Monte Carl o calculation with 5 quark fl avors. The dotted curve is the signal expected from the "non-perturbative model ." qi 173 Perturbative QCD calculations give levels "' lQ-3 to 10-4 of for a(µµ) a(e+ + x) which are too small to observe with present statistics. However e- + '¥ some non-perturbative models16] suggest a(ee 'l'x) 1 0- 1 ) lOpb (12) + a(µµ A study of µµ and ee pairs in haadronic events yielded the data in Figure 38 for both '¥ and T(9.4) region. Also shown is the '¥ signal expected on the above non perturbative model . No candidates were found which places a 90% CL upper limit on a(ee + '¥ + x) < 4.4 pb, and a(ee + x) < 4. 7 pb. This same search corre sponds to a sample of 940 bb decays . TThe + fl avor changing neutral currents in b - - decay for M(9,+9, ) > l.6 GeV gives the limit Br(b + 9,+9, x) < 0.8%. Furthermore a limit Br(b + '¥ + x) < 4.9% at the 90% CL is also obtained. However in the latter case we heard of the observation of such a decay mode by the CLEO Group presented by Thorndyke at this conference. They find Br(b x) = (0.64 0.23)%. + '¥ + .:':_ 6. Particle Identification with the TPC. So far the TPC Group have analyzed 5 pb- 1 of PEP data and some early results are shown herewith. The dE/dx Capabi lity. Figure 39a shows the dE/dx measuremen ts on a single cosmic ray muon as seen in some 180 wires. Note the high pulses due to the Landau fluctuations! Fi gure 39b gives the corresponding pulse height distribution and shows the location of the pul se height cutoff at the 65% level . This still leaves signals from 167 wires and yields the "truncated mean dE/dx" value. Figure 40 gives this truncated dE/dx distribution vs . momentum for hadronic events and Fi gure 41 gives corresponding Monte Carlo results. P'ions, kaons, protons, and electrons are clearly separated below the minimum ionizing region. As an example, Figure 42 gives the dE/dx ratio to that for pions for 0.45 < p < 0. 74 GeV/c. Figure 43 gives the corresponding results for 2. 7 < p 4.1 GeV/c and shows partial K,p separation in the relativistic ri se region . < Figure 44 gives the rr, K, and p fractional distributions obtained from the above. Finally the TPC Group presented a prompt rate in hadronic events. Fi gure µ 45 shows their results compared wi th results from several PETRA Groups. For 174 •• 20 •o eo 100 120 140 110 1ao •Ir• I Pul 1e·helchl. per wire .. .: ... ·� �------� �l�.,, ., .... , I .. (b) 30 .. 10 100 "fOO '" put kelllal... Pu l1e bela:ht. dlstrlbul lo..n X8L 835-9658 Fig. 39 XBL BJS-9659 Fig. 40 8.5 ATM BO�AR 20�CH4 �------· 1.4: :. ------��/ 1.S& �--�· .. ·;� / .·· /. · / /.· , 1.2 .·· .'! ' . I , • 1.15 . . /, J 1.1 · / .1 1.015 ·' I MOMENTUM (GEY} X8L 835-9662 Fig. 41 Fig. 39 - (a) dE/dx measurements on a single cosmic ray muon. Up to 180 individual dE/dx measurements are made on the track. (b) The corresponding pulse height distribution. The truncated mean value is determined by a cut selecting the lowest 65% of the signal s. (TPC) Fig. 40 Truncated mean dE/dx distributions for particles from hadronic - events at PEP. (TPC) Fig. 41 Monte Carlo calculations of truncated mean dE/dx distributions - for the TPC running conditions at PEP. 175 Hadronic Eventg • !� 0 Pion . o.• 0 Kaon '" ..� A Proton • D.I ...�• :i: D.? ...• D.I �100 . "':' u " � 110 D.I I- o. 4 .. D.S o.e RATIO ... 0.1 Fi g. 42 • to- 10 P (O.Y/o) Hadronic Events Fi g. 44 XBL 835-9663 .. x TPC 8 +PLUTO x llARK J o JADE • LUND llONTE CARLO ,. o LUND llONTE CARLO 10 0.7 0.D 1.1 1.3 ... RATIO 2 Fig. 43 O '-'--'-'-'-'-.._.....__.._._.._._._...... __,_,_._.._.....__.._._.._.._._�.�_._.j 10 15 20 25 30 Ii35 40 E (GEVj11 1 1 �·i"= Fi g. 45 XBL 835-9654 Fi g. 42 - Ratio of measured dE/dx to expected dE/dx for pions. Momentum 0.45 0.740 range < p < GeV/c. (TPC) F1g. Ratio of measured c.JE/cix to expected dE/dx for pions. Momentum 43 - 2. 7 4.1 range < p < GeV/c. (TPC) Fi g. 44 Charged hadron fraction for K, and p. Hadronic events at - PEP. (TPC) ;r, Fig. 45 Percentage of prompt muons in hadronic events at PEP. - TPC resu compared to PETRA results. lts 176 fo r 2 GeV/c. Also shown is the expected rate from Monte Carlo calculation µ with_p and > without the presence of a top quark at the PEP energy (29 GeV) . REFERENCES l] G. Gidal , B. Armstrong, and A. Rittenberg. LBL-91 Supplement UC-37, March 1983. 2] R.N. Cahn , M.S. Chanowitz , and J.D. Jackson. Private communication. 3] R.H. Schindler et Phys . Rev. 024, 78 (1981 ). .tl_. , 4] J.M. Yelton et £}_. , Phys . Rev. Lett. 49, 430 (1982). 5] C. Peterson et £}_. , Phys. Rev. 027, 105 (1983). 6] A. Ali et Phys. Lett 93B, 155 (1980) . gl. , 7] J.A. Jaros, SLAC Report 250, SLAC-PUB-3044, Stanford, California (1982). 8] X. Li and E. Ma , Phys. Rev. Lett 47, 1788 (1981 ). 9] D. Mi ller, Phys. Rev. 023, 1158 (1981 ). 10] G.L. Kane, SLAC-PUB-2326 (1979). 11] S. Narison, CNRS 78/P.997 (1978). 12] W. Bacino et Phys . Rev. Lett 42, 749 (1979). _tl. , 13] G. Kalmus , Rapporteur's tal k at XXI International Conference on High Energy Physics, Paris, 1982. 14] H. Harari , SLAC-PUB-2234 (1978); M. Gai llard and L. Maiani, LAPP-TH-09 (1979). P.H. Gir.sparg, S.L. Glashow and M.B. vlise. Phys. Rev. Lett. 50, 141 5 ( 1 983) . 15] H.J. Behrend UESY 92/056 (1982). tl ill_ ., 16] G.L. Kane et Phys. Lett B90, 164 (1980). ill_.,