Prospects for Beauty Physics at Hadron Colliders K. T. McDONALD Joseph Henry Laboratories Princeton University Princeton, New Jersey 08544 INTRODUCTION N a t u r e wants us to do B physics! T h e B lifetime is - 1.2 psec - can observe the decay vertex separated from the production vertex. T h e B's can be isolated even when produced in hadronic interactions. 0 B o - Eo mixing is large - phenomenology rich in quantum-mechanical effects. T h e cross-sections for g g - B B a r e large a t collider energies.'T2 R H I C TEV I ssc 0.5 2 20 10 1/4000 40 1/1000 5 00 1 /200 @ T h e signal to noise for beauty a t the S S C is comparable to that for charm in B-decay products a r e soft: 90% of all tracks have P, < 5 GeV/c. * C a n follow a classic prescription for detailed study of particles: measure the CP violation is accessible. fixed-target photoproduction. decays to few-body, all-charged final states. - A 3-u signal for a CP-violating asymmetry of 0.1 requires 1000 reconstructed events. - need 10' reconstructible pairs if branching fraction = lo-'. - need 1O'O pairs produced if reconstruction efficiency = 1%. - T h e luminosity needed to produce such a sample in lo7 sec is then L = lo3* cm-2sec-' a t R H I C L = 2.5 x lo3' cm-2sec-' a t TEV I L = 2 x 1030 cm-'sec-l a t S S C - How to survive in high multiplicity. T h e crispness of the CP-violation physics encourages us to tackle difficult issues. 100 tracks/event; 100 samplesltrack - - lo4 wordslevent. Can readily find patterns if detector occupancy is - need - l o 7 detector elements (- 105/layer). - How to survive a t high rate. 215 216 ANNALS NEW YORK ACADEMY OF SCIENCES Have 5-10 M H z event rate a t L = Need sparse readout and buffering of the lo4 elements struck per event. - 50-100 Gword/sec raw event rate! Front-end analog triggers reduce event rate to < l o 0 kHz. Numeric-processor farm reduces rate to 1-2 kHz. Archival data rate - 100 Mbytelsec to video tapes. cm-2sec-'. * T h e same detector concept applies to RHIC, TEV I, and the SSC. FIGURE 1 shows the Beauty spectrometer as conceived a t the 1988 Snowmass Workshop. PREJUDICES ABOUT CP VIOLATION IN THE B-B SYSTEM 0 The best signal for CP violation is an asymmetry: r(B--f) - l?(B-f) I'(B--f) + l?(B-f) ' A = 0 The theoretical interpretation of A is clean only w h e n f = f = CP eigenstate. - must tag the other B of the B - i p a i r . - 'self-tagging' modes like B o - K - r - or D + T - might be more accessible experimentally, but a r e less useful theoretically. (Like e ' / e in KO - TT these depend on penguin diagrams. . .) Large A - small r, in most models. To reach S standard deviations in a measurement of A for mode B - f with branch I', need N produced B s , where Example: S = 3, A = 0.1, r = - N = 9 x 10'. -. SSC BEAUTY SPECTROMETER (Snowmass '88) FIGURE 1. A beauty spectrometer concept. MCDONALD PROSPECTS FOR BEAUTY PHYSICS 217 P I - " t s FIGURE 2. The unitarity triangle. CP violation and the KM matrix (Wolfenstein): vud vus vub 1 wI.x3(p - ilt) v K M = [ ; ; ; ;I=[ - A 1 FA2 1 1. p ~ 3 ( 1 - - ilt) -& X - the Cabibbo angle. p is known via the B lifetime. 4 # 0 - CPviolation. But, p , 7 a r e not well determined from the KEsystem. Bjorken's trick: unitarity of V K M - v r d + XVrs + v,*, = 0. - T h e B-Bsystem is the place! - these 3 vectors form a closed triangle. On dividing the lengths by pX3, we obtain FIGURE 2. Since the base is known, the experimental challenge is equivalent to measuring the 3 interior angles 4,, &, &. 0 For B --fwithfa CP eigenstate, the asymmetry A depends only on sin @ A - sin 24, for B - + K ~ , DDK,, +?T?T, DO, D O ? T + T - , . . . A - sin 2$* for B - ?T+?T-, p p , . . . A -sin($, - 4,) for B - E*' K s , . . . 0 An overconstrained study of CP violation in the KM matrix requires the observation of many decay modes. =) need a general purpose detector with particle identification. - need tagging of the second B. - an extensive program for the 1990's! PRE-CP PHYSICS IN THE B-B SYSTEM o B ~ - gluon structure functions a t collider energies. Bh&and B,-& mixing. T h e latter is large but surprisingly hard to measure owing to rapid oscillations. 218 ANNALS NEW YORK ACADEMY OF SCIENCES 0 r ( B --f) for likely CP-violating modes - a rich program even before CP violation. RECENT DETECTOR PROPOSALS 1 . March 1987, Letter of Intent to F N A L by Volk, Reay et uL3 2. March 1987, Letter of Intent to F N A L by Van Berg, Lockyer et al.4 3. July 1987, SSC Beauty Spectrometer a t the Berkeley Workshop.s 4. November 1987, Dipole Beauty Spectrometer a t the F N A L Beauty Work- 5. June 1988, Status Report of the Fermilab B Collider Study Group.' 6. July 1988, SSC Beauty Spectrometer at the Snowmass Workshop.* shop! E l e v a t i o n V i e w FIGURE 3. Overview of the BCD detector. 7. July 1988, Solenoid Beauty Spectrometer a t the R H I C Workshop.' 8. August 1988, Letter of Intent to C E R N by Brandt, Schlein et a[." 9. October 1988, Letter of Intent for a Bottom Collider Detector for the Fermilab Tevatron." OVERVIEW OF THE BOTTOM COLLIDER DETECTOR We will use the recently proposed Bottom Collider Detector" as an example. A schematic side view is shown in FIGURE 3. The design is driven by the need for large angular acceptance, good momentum resolution for low-momentum tracks, precision vertexing, and particle identification. Calorimetry is not important except for electron McDONALD. PROSPECTS FOR BEAUTY PHYSICS 219 identification. The basic character of the detector is ‘central,’ but with greater emphasis on the angular region 2 O < 6 < 30° than in present detectors designed for W and Z physics. a A dipole magnet is chosen to optimize the detection of tracks produced between 2 O < 6 < 178O (pseudorapidity: -4 < 7 < 4). The kinematics of B - B production a r e such that both forward and central tracks must be measured well to have a high geometric acceptance, while transverse momenta greater than 5 GeV/c are seldom of interest. A dipole magnetic field oriented perpendicular to the beams is the best and simplest solution. A low-field ( 1 Tesla), large-diameter magnet is preferable for pattern recognition of low-momentum tracks. 0 T h e magnet design calls for circular pole tips of 4-m diameter, separated by a 4-m gap. This large gap permits a tracking system with 75-100 samples per track, the minimum acceptable number in a high-multiplicity event, while still accommodating the EM calorimeter, TRD, and R I C H counters inside the gap. 0 T h e vertex detector is designed to find the secondary vertices of B particles with high accuracy and efficiency, thereby isolating the B from the rest of the event. T h e resulting ability to study the time evolution of the states is particularly advantageous for CP studies. Extensive Monte Carlo simulation indicate that 3-d vertex reconstruc- tion is necessary to achieve good pattern recognition, and that the system should have a worst-case impact-parameter resolution of <20 pm. All tracks should intersect a t least 3 planes with an angle of incidence t4S0. These requirements, along with the length of the interaction region, led to a hybrid design of barrels and planes of double-sided silicon detectors. These devices are located outside the beam pipe, beginning a t 1.5-cm radius, to minimize the effects of multiple scattering. Monte Carlo studies of this design, where tracking efficiencies were loo%, gave an efficiency of finding B vertices of -45%. A preliminary mechanical model of this detector design has been constructed. * T h e vertex detector relies on the gas tracking system for most of the pattern recognition. The tracking system is designed for efficient and rapid 3-d pattern recognition of tracks over the full angular range. There are 75-100 samples along each track. The technology used is thin-walled straw tubes, pressurized to 3 atmospheres and arranged in superlayers. Such straw tubes permit a measurement error of 40 pm per hit. This high precision will allow a mass resolution of 20 MeV/c2 and an extrapolation error into the silicon vertex detector of 50 pm. Good mass resolution is desirable to separate Bd from B,, and to set a narrow mass window around the B as a rejection against combinatoric background. Particle identification is important in reducing the combinatoric background, especially for modes such as B - KT and B - p p . Electron identification is required for triggering and tagging the particle-antiparticle nature of the B . The design incorporates TRD’s R I C H counters, and an electromagnetic calorimeter over the full detector acceptance and for the full momentum range of the B-decay products. 0 T h e trigger and data-acquisition system is designed to handle a luminosity of lo3’ cm-2sec-’ and data-flow rates of GigaBytes per second. The trigger philosophy is to assemble the full event as soon as possible and pass it to a farm of numeric processors where a variety of trigger algorithms can be implemented in software. T h e system is based on the latest communications-industry technology, which represents a new approach for high-energy physics experiments and is suitable for SSC data rates. 220 ANNALS NEW YORK ACADEMY OF SCIENCES - - - t - 1 I 1 I An initial cost estimate for the detector is $30M. I n the space remaining we elaborate slightly on two issues, a possible ‘topology trigger,’ and the data-acquisition system based around a ‘barrel switch.’ TOPOLOGY TRIGGER A first-level trigger is desired that might be straightforward to implement while reducing the event rate by 20-50. T h e ‘topology trigger’ simply counts the number of x 510 a c 0 2 2 210 I $10 0 0 z I 5 5 ’ a m d 1 TRACK 0 2 TRACKS X I m 1 TRACK 0 2 TRACKS BACKGROUND M N T S FROM U.UB.D.DB.S.SB.C.CB TWOJET fNTE%T$g’c’ FIGURE 4. (upper) T h e number of charged tracks above a given P, for events containing a B - T+K- decay, according to an ISAJET simulation; (lower) t h e same for events without a n y B’s. MCDONALD PROSPECTS FOR BEAUTY PHYSICS 221 charged tracks in the event above a given cut in P p This is illustrated in FIGURE 4, based on an ISAJET simulation. If a cut on only a single track could be made reliably, we see that a cut a t PT = 2.5 GeV/c would yield a factor of 50 reduction in the event rate a t the expense of only a 10% loss of events with a 2-body decay such as B - a+a-. Or, if two tracks a r e each required to be above 2 GeV/c, the desired reduction of the total event rate could be achieved with 40% loss of 2-body B decays. The efficiency for many-particle B decay is, of course. less. DATA ACQUISITION The architecture of the data-acquisition system takes advantage of several new 0 Digital transmission over fiber-optic cables. A simple ‘barrel shifter’ switch for event building a t 100 kHz. A farm of -2000 numeric processors, especially suited to tracking problems, approaches: in particular: which implement the second-level triggers, employing full reconstruction algorithms for some detector systems. A block diagram of the proposed data-acquisition system is shown in FIGURE 5. The data flow is from top to bottom. Prompt triggers (such as the topology trigger) will first reduce the data rate by a factor of 20-50, using a subset of the detector elements labeled A-Z in FIGURE 5. From this point on there is no other specially built logic for triggering. If the event is accepted by the first-level trigger, the data fragments are transmitted over fiber-optic cable to the Event-Builder Switch. There can be any number of data sources from each detector system. The Event-Builder Switch is based on the high-speed technology of a telephone exchange. The ‘barrel-switch’ technique can accept parallel inputs from a large number of sources and reorganize these into a set of output streams, each stream containing only information from a single event. Data rates of tens to a few hundreds of GigaBytes/sec are possible. The Receiver/Formatters format the data into structures suitable for high-level language programs, and then pass the events into one member of the large farm of numeric processors. Each of these processors will be the equivalent of about 40 VAX 780’s and will implement the second-level trigger in software. They should reduce the event rate by another factor of 100, leaving a rate of -1 kHz a t a luminosity of cm-2sec-l, which can be archived to tape on video cassettes. SUMMARY The cross-section for gg - B B is relatively high a t collider energies, so that a one-year r u n a t RHIC or TEV I might yield > l o l a B-Bpairs, and pairs a t the SSC. The challenge to the experimenter is to trigger on and reconstruct a significant fraction of this sample. Detectors are being proposed that make extensive use of silicon vertexing, VLSI readout, and massive online numerical processing with the goal of 222 ANNALS NEW YORK ACADEMY OF SCIENCES Event FragmSfltS Event Fragments From M Sources From N Sources From Deleclor Element A From Detector Element Z Transmitted Via Transmitted Via Fiber Optic Cables Fiber Optic Cables Event Builder (Switch) 6 'Barrel Shiner With Very Linle Memoiy I I Unformatted Event X+i fiber Optic Cables C- TransmnedWa Unformatted Event X Transmined Via Fiber optic Cables Recelverl Receiver1 Formatier Formatter Formaned Events Copper Cables Procassor(s) 7 Processor(s) - Processor(s) t""l Processor(s) P FIGURE 5. Block diagram of the proposed data-acquisition system. maintaining a 1% efficiency for few-body decays to all-charged final states. If achieved at the SSC for L = lo3' cm-'sec-', this would be equivalent to an e + e - - B factory operating at L = cm-'sec-' and 100% reconstruction efficiency. Even at RHIC or TEV I with lo8 reconstructible B ' s , the strongest signals for CP violation in the B - B system would be accessible. REFERENCES 1 . 2. BERGER, E. L. ANL-HEP-PR-88-26. NASON, P., S. DAWSON & R . K . ELLIS, FermiIab-PUB-87/222-T. MCDONALD PROSPECTS FOR BEAUTY PHYSICS 223 3 . 4. 5 . 6. 7. 8. 9. 10. 1 I . VOLK, J. T., N. W. REAY, et al. Letter of intent for a Tevatron collider beauty factory V A N BERG, R., N. S. LOCKYER, et a/. Proposal for a bottom collider detector (March FOLEY, K . J., et al. A beauty spectrometer for the SSC. In Proceedings of the 1987 SSC REAY, N . W., et al. In Proceedings of the Workshop on High Sensitivity Beauty Physics a t Status report of the Fermilab B collider study group. Princeton University preprint In Proceedings of the 1988 Snowmass Workshop on Physics in the 1990’s (July 1988). I n LOCKYER, N. S., et al. In Proceedings of the R H l C Workshop (July 1988). I n press. BRANDT, A,, P. SCHLEIN, et al. Letter of intent to the SPCS for a study of beauty and charm CASTRO, H., et al. Letter of intent for the BCD, a bottom collider detector for the Fermilab (March 1987). 1987). Workshop (Berkeley, July 1987) p. 759. Fermilab (November 1987). DOE/ER/3072-45 (June 1988). press. physics at the SPS collider (August 1988). Tevatron (October 1988).