VIEWS: 6 PAGES: 76 POSTED ON: 9/24/2012
Southampton July 1999 Future of Heavy Flavour Physics: Experimental Perspective Sheldon Stone Syracuse University 1 Future Physics Goals: Introduction Our goals are to make an exhaustive search for physics beyond the Standard Model and to precisely measure SM parameters. Here we ask what studies need to be done, not just what studies can be done. Measurements are necessary on CP violation in Bo and Bs mesons, Bs mixing, rare b decay rates, and mixing CP violation and rare decays in the charm sector. These quarks were present in the early Universe. There is a connection between our studies and Cosmology. 2 The CKM Matrix We need to connect the weak eigenstates of quarks with the mass eigenstates weak eigenstates VCKM mass eigenstates If n’s have mass there will be analogous matrix 3 Proper Formulation of CKM Matrix d s b 1 1 2 u 1 l2 l Al r ih1 l 3 2 2 V c l 1 2 1 l ihA 2 l4 2 2 Al 1 ihl 2 2 t Al 1 r ih Al2 1 Good l3 in real part & l5 in imaginary part We know l=0.22, A~0.8; constraints on r & h Due unitarity there are 6 CKM triangles 4 The 6 CKM triangles “ds” - indicates rows or columns used There are 4 independent phases, which can be used to construct entire CKM matrix 5 The 4 CKM Phases Vtb Vtd * Vub Vud * V V* arg arg * V V cb cd cb cd V Vcb * V Vus * arg VV cs arg ud * V V tb * ts cs cd & probably large, small ~0.02, smaller 6 Usual Triangle (for ref.) Real a, & must sum to 180o Therefore, any two will do IF we are really measuring the intrinsic angles New physics can hide, only these angle measurements not sufficient. Ex: Suppose there is new physics in Bo-Bo mixing (q) & Two sides |Vub/Vcb| & we measure CP in yKs and pp, |Vtd/Vts| important. then 22q, 2a2aq, & BTeV can measure the 22a22a, new physics latter from Bs mixing but a 180o 7 Ambiguities Suppose we measure sin(2) using yKs, what does that tell us about ? Ans: 4 fold ambiguity- , p/2, p, 3p/2 Only reason h>0, is Bk>0 from theory, and related theoretical interpretation of e 8 Problems with measuring a using Bopp Using Bopp would be nice, but large Penguin term, CLEO: B(Bo p+p-) < 0.84 x 10-5, while B(Bo K+p-) =(1.4±0.3±0.2)x10-5 The effect of the Penguin must be measured in order to determine a. Can be done using Isopsin, but for s get K- requires a rate measurements of ppo and popo (Gronau & London). However, this is daunting. 9 Measuring a using Borp pppo A Dalitz Plot analysis gives both sin(2a) and cos(2a) ( Snyder & Quinn) CLEO has measured the branching ratios B(Brop = (1.5±0.5±0.2)x10-5 B (Borp + rp Snyder & Quinn showed that 1000- 2000 tagged events are sufficient = (3.5±1.0±0.5)x10-5 10 Comparison of rp modes Final State: rp rp rpo rop r op o CLEO(10-5): 3.5±1.2 <7.7 1.5±0.5±0.2 <1.8 Ciuchini et al.: 1.0-7.5 0.2-1.9 0.3-2.6 0.5-1.1 0.00-0.02 Ali et al.: 2.1-3.4 0.6-0.9 1.1-1.6 0.1-0.7 0.00-0.02 B(B+ roK+) = < 2.2x10-5 @ 90% c.l. (CLEO) {To measure the three neutral rp modes, requires triggering on hadronic decays with high efficiency, RICH particle ID and high quality photon detection} 11 More on a Using rp Dalitz Plot analysis, sin(2a) and cos(2a) are measured, only one ambiguity remains a & pa To remove this use Bopp.This works because of large Penguin rate (Grossman & Quinn) a(pp)-a(rp)=-2(AP/AT)cos(dP-dT)[cos(2a)sin(a)] Usefactorization to get sign of AP/AT (-) Theory says cos(dP-dT) is small 12 Ways of measuring May be easier to measure than a There are 4 ways of determining Time dependent flavor tagged analysis of BsDsK Measure rate difference between B-DoK- and B+DoK+ Rate measurements in Kop and Kp (Fleisher-Mannel) or rates in Kop & asymmetry in Kpo (Neubert-Rosner) . Has theoretical uncertainties but can be useful. Use U spin symmetry ds: measure time dependent asymmetries in both Bopp& BsKK (Fleischer). Ambiguities here as well but they are different in each method, and using several methods can resolve them. 13 BsDsK Decay processes ± Diagrams for the two decay modes, B ~ 10-4 for each 14 B-[K+p-]K- Decay processes B ~ 10-6 B ~ 2x10-7 15 What do we need to measure involving ? Phase of Bo-Bo mixing, using yKs (done soon? precision?) Measure using other modes: fKs,hKs, ypo Remove ambiguities if possible Measure interference in yK*. However, either must make theoretical assumption (factorization) to guarantee that fsi don’t change sign of strong phase-shift or measure Bs yf & use SU(3) Kayser: Measure time dependence using yKs, Kspln, has a cos(2) term (loss of rate) 16 Decay Widths for Bo yKs, Kspln t B , t K e B t B {e s t K [1 sin(2) sin(m B t B )] cos(2)>0 e s t K [1 sin(2) sin(m B t B )] cos(2)<0 1 s L t K ( )2e 2 [cos(m B t B ) cos(m K t K ) cos(2) sin(m B t B ) sin(m K t K )]} top sign for Bo, bottom for Bo 3rd line 1st pair for p-l+n (K), 2nd pair for p+l-n (K) t K integrated over t B CPT tests? 17 A Critical Check using Silva & Wolfenstein, (Aleksan, Kayser & London), propose a test of the SM, that can reveal new physics; it relies on measuring the angle . (This is the sine qua non Unitarity check) BTeV can use CP eigenstates to measure , for example Bsyh, hr Can also use yf, but need complicated angular analysis The critical check is sin sin sin l 2 sin( ) Very sensitive since l 0.2205±0.0018 Since ~ 0.02, need lots of data 18 More checks using Other checks using |Vtd/Vts|, |Vub/Vcb| possible 2 Vub sin sin( ) sin Vcb sin() 2 Vtd sin sin( ) sin Vts sin( ) might provide best measurements of Ultimately, these CKM ratios! 19 Other critical CKM measurements Bs mixing Use Bs Ds p to determine xs Find by comparing lifetime for yh or K+K- with Ds p If is large enough ~10%, then other interesting measurements are possible |Vub|, how do we do this with minimal theoretical error? 20 Rare b Decays Exclusive Rare Decays such as Br, BK*l+l Inclusive Rare Decays Possibilities for New Physics such as inclusive bs, W- in loop is replaced by bd, bsl+l. Can other charged object such hadron machines do as H-, X - this? Probably using new fermion like objects replace t same technique as CLEO. 21 Charm decays Predictions of the Standard Model contribution to mixing and CP violation in charm decay are small. Thus, this provides a good place to search for New Physics. Currentexperimental limit on mixing, rD<5x10-3, SM expection rD~10-8 CP current limit is ~10%, SM expectation is 10-3 22 Summary of required measurements for b’s Physics Decay Mode Vertex K/p det Decay Quantity Trigger sep time sin(2a) Borppppo sin(2a) Bopp & BsKK cos(2a) Borppppo sign(sin(2a)) Borp & Bopp sin() BsDs K sin() BoDo K sin() BK p sin(2) BsJ/yh, J/yh sin(2) BoJ/yKs cos(2) BoJ/yK* & BsJ/yf xs BsDsp for Bs BsJ/yh, KK, Dsp 23 Summary of Standard Model Tests Check that a = 180o, after removal of ambiguities; necessary if we have properly measured these quantities Check that the Silva-Wolfenstein test has been met: 2 sin sin sin l sin( ) Check that magnitude ratios |Vtd|/|Vts| and |Vub| / |Vcb| are consistent with determined a, , , Search rare decays for anomalous rates, or dilepton polarizations 24 Current Status: Brief Summary We know xd and have limit on xs We know |Vcb| & |Vub|, but how accurately? What is the meaning of a theoretical error? Is it Gaussian distributed? Does statistics work in combining undefined errors? Whatis the error on Vcb? What is the error on Vub? We know lifetimes 25 We know lifetimes (still some mysteries) total = 1/t 2 1.9 1.8 1.7 B lifetimes (ps) 1.6 1.5 B+ Bo 1.4 Bs “Low” 1.3 1.2 1.1 Lb 1 26 Heavy Quark Effective Theory HQET tells us that in first order when a b quark transforms to a c quark with the c going at the same velocity as the b, the form factor is 1 in first order AND the corrections to 1 can be calculated The form-factor therefore known to be 1- correction, at maximum q2, called w=1, where M2 M2 * q 2 w B D 2M BM D* 27 |Vcb| from BD*l n Use BD*l n because the decay rate is largest for and the corrections are better determined. In HQET there is one “universal” form-factor function, so we don’t have to deal with 3 form- factors To find Vcb measure value at w=1, here D* is at rest in B rest frame m m 2 1 2wD* D* d(B D*ln) G2 mB m 2 F 3 Vcb F2 (w)(m B m D* ) 2 m 3 * w2 1 4w(w 1) 2 B dw 48p D 2 m D* 1 mB 28 CLEO Measurement background background 29 Vcb results, an example To get results fit using shape proposed by Caprini et al, or Boyd & Grinstein, or in CLEO case by Stone Use F(1)=0.91±0.03, from Caprini, Uraltsev….. Results DELPHI: (41.2±1.5±1.8±1.4)x10-3 ALEPH: (34.4±1.6±2.3±1.4)x10-3 OPAL: (36.0±2.1±2.1±1.2)x10-3 CLEO: (39.4±2.1±2.2±1.3)x10-3 World Average 0.0381±0.0021 by adding theoretical error in quadrature with exp error. 30 Theoretical Value of F(1) Lim F(1) = 1 as mb , F(1)=1+O(as/p)+d1/m2+d1/m3 (no d1/m , Lukes thrm) F(1)=0.91±0.03, from Caprini, Uraltsev….. F(1)=0.89±0.06, from Bigi Can we get an accurate non-quenched value from the Lattice? The errors are not consistent. What do the errors mean? Bigi: “In stating a theoretical error, I mean that the real value can lie almost anywhere in this range with basically equal probabilty rather than follow a Gaussian distribution. Furthermore, my message is that I would be quite surprised if the real value would fall outside this range. Maybe one could call that a 90% confidence level, but I do not see any way to be more quantitative.” 31 QCD Sum Rules for |Vcb| Using Operator Product Expansion & Heavy Quark Expansion, in terms of as(mb), L, and the matrix elements l1 and l2, we can accurately determine Vcb. These quantities arises from the differences l1 3l 2 l l2 mB mb L , m B* m b L 1 , 2m b 2m b From B*-B mass difference, l2 = 0.12 GeV2 a L a 1 1.54 s 1.65 1 0.87 s p p 2 2 5 G F Vcb m B mB sl 0.369 192p 3 L2 l1 l2 0.95 2 3.18 2 0.02 2 mB mB mB 32 Measurement of sl & moment analysis Use total Branching Ratio Measurement CLEO using lepton tags (10.490.170.43)% Lifetime 1.6130.020ps sl= 65.03.0 ns-1 (note LEP 68.61.6 ns-1 ) “Moment Analysis” of BXln, Mx and El 33 Result for |Vcb| Using Moment Analysis Discrepancy between hadronic mass moments and El moments Theoretical estimates favor Mx moments Taking Mx estimates only: L = 0.330.020.08 GeV l1=-0.130.010.06 GeV2 Ligeti claims: |Vcb| = 0.04150.0012, but Is this an experimental problem or an inherent problem in OPE? it would be foolish to use it 34 Vub from lepton endpoint B & value of Vub depends on model, Y(4S) data bcln since fraction continuum of leptons in signal region depends on model! 35 Vub from pln and rln pln p0ln bu backrounds & cross-feeds rln r0ln bc backrounds woln bu backrounds & cross-feeds 36 New CLEO form-factor Analysis Find rln r0ln woln as function of El using likelihood method to fit M(pp) & E distributions, where E = Er+El+|pmiss|-Ebeam El >2.3 GeV/c r modes with E cut r modes with M(pp cut r modes with both cuts buln bkgrd bcln M(pp) E El 37 Form-factor Results In general 3 form-factors for 0- 1- transitions, but we do not have enough precision to disentangle them Data shows the need for more data Combining with older result: |Vub|=(3.250.14 0.29 0.55)x10-3 +0.21 38 Summary of |Vub| Results LEP measurements use 8% theoretical error as given by Uraltsev. However Jin’s similar calculation claims a 10% error but differs by 14%. I use a 14% theory error here Since the LEP Monte-Carlo calculations are highly correlated, I take a common 14% systematic uncertainty The exclusive channels rule out the Korner & Schuler (KS) model (gets the wrong V/P ratio), but have large errors The CLEO endpoint results have the best statistical error. Hard to estimate the theoretical error. I take 14%. 39 CKM Plot 68% c.l. 95% c.l. From Parodi, Roudeau & Stocchi Several authors have done these Dominant error in e band is Vcb fits: Mele, (long ago Rosner). Question: If we find a region well 1 errors, not to be believed outside of contour after measuring good art? angles is SM ruled out? 40 Short Term Perspective e+e- B factories at Y(4S) will turn on: CLEO III, BaBar, Belle Since CLEO III didn’t have an upgrade talk, I will show you some pictures. Detector installation is occurring now. 41 CLEO III (not yet shown here) Yes Virginia CLEO will continue with symmetric beams and higher L Illustrative lesson: Best CESR Performance “These are the highest values measured during normal High Energy Physics running with a beam energy of 5.3 GeV. They did not necessarily occur simultaneously: Peak luminosity 8.0 x 1032 cm-2/sec Best integrated luminosity 40 pb-1 per day, 750 pb-1 per month, 4.4 fb-1 per year” 42 CLEO II CLEO III Replace everything inside the magnet coil to allow for interaction region quads to move closer (higher L) and allow for new particle identification system (RICH) 43 Some CLEO Pictures A Silicon Ladder The RICH Radiator 44 More CLEO III RICH photos The Photon Detectors Mating the Two Cylinders 45 Expected Results from e+e- B factories Babar & Belle can & should measure sin(2) For everything other than CP violation via mixing CLEO III will be directly competitive Rare decays CP violation in rare decays Better understanding of Vcb, especially using D*oln at maximum q2 Better understanding of Vub using moments and higher statistics data on exclusive decays Lattice calculations may help 46 Short Term Competition Hadron colliders HERA-b has the potential to also measure sin(2) CDF already has already taken the first steps toward measuring sin(2). Next run will start ~Aug. 2000. CDF could also measure xs. 47 Why do b & c decay physics at hadron colliders? Large samples of b quarks are available, with the Fermilab Main Injector, the collider will produce ~4x1011 b hadrons per 107 sec at L = 2x1032 cm-2s-1. Rates are potentially 5 x larger at LHC. e+e- machines operating at the Y(4S) at L of 3x1033 produce 6x107 B’s per 107 s. Bs & Lb and other b-flavored hadrons are accessible for study. Charm rates are ~10x larger than the b rate 48 Main detector challenges Problems: b/tot~ 1/500 at Fermilab, 1/100 at LHC Background from b’s can overwhelm “rare” processes Large data rate just from b’s - 1 kHz into detector Large rates cause Radiation damage to EM calorimeter; photon multiplicities may obscure signals Solutions for BTeV: Use detached vertices for trigger and background rejection Have excellent charged particle identification & lepton id Dead-timeless trigger and DAQ system capable of writing kHz of events to tape Use PbWO4 crystal calorimeter 49 Fundamental Detector Principles Necessary to trigger efficiently on purely hadronic final states - detached vertex trigger Necessary to reconstruct final states with excellent decay time resolution, good efficiency and mass resolution Necessary to detect final states with or po efficiently with good energy resolution Necessary to able to identify p/K/p 50 Characteristics of hadronic b production The higher momentum b production peaks at large b’s are at larger h’s angles with large bb correlation h 51 Long Term View All measurements I have discussed, must be done Hadron machines produce enough b’s, & Bs We will have LHCb, Atlas, CMS & possibly BTeV Atlas & CMS work in central region, they lack the ability to trigger on purely hadronic final states and the lack particle id. Atlas also lacks an excellent EM cal Although LHCb & BTeV can do all the measurements that Atlas & CMS can do, there is a class of measurements involving J/y decays, for which they can compete 52 The BTeV Detector Inside the beam pipe -PbWO4 crystals 53 The C0 Interaction Region Construction finished BTeV is designed to be compatible 54 The LHCb Detector 55 The BTeV Pixel Detector Pixels necessary to eliminate ambiguity problems with high track density; Essential to our detached vertex trigger Crucial for accurate decay length measurement Radiation hard Low noise 56 Pixel Trigger Description Triplets used to get space point & mini-vector, called a ‘station hit’ Station hits are organized into f-slices Tracks are found in these f-slices full pattern recognition is performed Minimum track p cuts are applied Event level processors then find primary vertices & detached tracks 57 Detached Vertex Trigger Level I Trigger uses information from the Pixel Detector to find the primary vertex and then look for tracks that are detached from it The simulation does the pattern recognition. It uses hits from MCFast including multiple scattering, bremsstrahlung, pair conversions, hadronic interactions and decays in flight Detailed studies of efficiency and rejection for up to an average of three interactions/crossing 58 BTeV Trigger Performance For a requirement of at least 2 tracks detached by more than 4, BTeV triggers on only 1% of the beam crossings and achieve the following efficiencies for these states: State efficiency(%) State efficiency(%) B p+p- 55 Bo K+p- 54 Bs DsK 70 Bo J/y Ks 50 B- DoK- 60 Bs J/yK* 69 B- Ksp- 40 Bo K* 40 59 Ring Imaging Cherenkov’s Both LHCb & BTeV have excellent p/K/p using gas (C4F10) & possibly also aerogel radiators Visible photons detected use phototubes or HPD’s 60 EM calorimeters BTeV uses 22 cm long LHC-b uses a Shaslik PbWO4 crystals developed EM cal with scintillating by CMS fibers and lead 20k-40k crystals There is also a Crystals are radiation hard preshower detector Scintillation is fast, 99% This is very radiation of light emitted < 100 ns hard BTeV will use phototube readout since calorimeter is not in a magnetic field 61 Expected EM Calorimeter Performance BTeV LHCb Energy resolution: E 1.6% 2 0.55% 10% 1.5%2 2 E 2 E E E in (GeV) Position resolution: x (m) 35002 2202 E 62 Physics Simulations: Some Lessions BTeV simulations do have pattern recognition (except for trigger); smears hits and refits the tracks using “Kalman Filter,” has multiple scattering, bremsstrahlung, pair conversions, hadronic interactions and decays in flight included Parameterized shower energy deposits Next, follow examples 63 The CP asymmetry in Bop+ p- The average decay distance and the B momentum uncertainty in the average decay distance are functions of B momentum: <L> = 480 m x pB/mB Decay distance error 64 Bop+ p-: L/ distribution L/ = Decay length/error is very important in rejecting background both at trigger level and in analysis Much better in Forward (BTeV) geometry than Central geometry because b’s are moving faster L/ 65 Bop+ p- analysis: the importance of particle identification Require that each p be properly identified in the RICH. Otherwise the measurement is probably impossible. 66 A sample calculation: Bop+p - BTeV LHCb Cross-section 100 µb 500 µb Luminosity 2x1032 2x1032 # of Bo/Year (107 s) 1.4x1011 7x1011 B(Bo p+p-) 0.75x10-5 0.75x10-5 Reconstruction efficiency 0.06 0.032 Triggering efficiency (after all other cuts) 0.50 0.17 # (p+p-) 34,000 28,560 eD2 for flavor tags (K±, l±, same + opposite side jet tags) 0.1 0.1 # of tagged p+p- 3,400 2,900 Signal/Background 0.6 1 Error in p+p- asymmetry (including bkgrd) ±0.023 ±0.019 67 Measuring a Using Bo r p p+p-po BTeV LHCb B (x10-5) 4 4 efficiency 1.0x10-2 1.7x10-4 BTeV # found 28,000 2,400 # tagged 2,800 240 Backgrounds not yet determined BTeV has more than enough for Dalitz plot analysis 68 xs Reach Both LHCb & BTeV have excellent xs reach using BsDsp- LHCb gets a 5 signal for ms < 48 ps-1 (xs < 68) 69 Comparisons Here I compare BTeV with LHCb and with other experiments that posses the all the necessary elements for a state of the art heavy quark experiment: Ability to trigger efficiently on purely hadronic final states Ability to detect final states with or po efficiently with good energy resolution Ability to identify p/K/p We are left with e+e- B factories 70 Comparisons of BTeV With e+e- B factories Number of flavor tagged Bop+ p - (B=0.75x10-5) L (cm-2s-1) #Bo/107s e eD2 #tagged e+e- 3x1033 1nb 3.0x107 0.4 0.4 46 BTeV 2x1032 100b 1.4x1011 0.03 0.1 3400 Number of B-Do K - L(cm-2s-1) #Bo/107s e # e+e- 3x1033 1nb 3.0x107 0.5 2 BTeV 2x1032 100b 1.4x1011 0.015 320 Bs , Bc and Lb not done at Y(4S) e+e- machines Number of tagged, reconstructed Bo decays to rp is a factor of at least 10 higher for BTeV. 71 Comparisons of BTeV with LHCb Advantages of LHCb b 5x larger at LHC, while t is only 1.6x larger The mean number of interactions per beam crossing is 3x lower at LHC, when the FNAL bunch spacing is 132 ns LHCb HAS BEEN APPROVED! Advantages of BTeV (machine specific) The 25 ns bunch spacing at LHC makes 1st level detached vertex triggering more difficult. The 7x larger LHC beam energy causes problems: much larger range of track momenta that need to be analyzed and large increase in track multiplicity, which causes triggering and tracking problems The long interaction region at FNAL, =30 cm compared with 5 cm at LHC, somewhat compensates for the larger number of interactions per crossing, since the interactions are well separated 72 Comparisons with LHCb II Advantages of BTeV (detector specific) BTeV is a two-arm spectrometer (gives 2x advantage) BTeV has vertex detector in magnetic field which allows rejection of high multiple scattering (low p) tracks in the trigger BTeV is designed around a pixel vertex detector which has much less occupancy, and allows for a detached vertex trigger in the first trigger level. for accumulation of large samples of rare hadronic decays and Important charm physics. Allows BTeV to run with multiple interactions per crossing, L in excess of 2x1032 cm-2 s-1 BTeV will have a much better EM calorimeter 73 The Status of BTeV BTeV is an approved R&D project at Fermilab, E897, whose purpose to generate a full proposal for a heavy quark decay experiment at the Tevatron collider by May 2000. BTeV has submitted a preliminary technical design report in May of 1999. BTeV has been asked to submit a full proposal in May of 2000. 74 Conclusions A complete program to test the Standard Model and see beyond it requires the measurement of CP violation and rare decays in the b & c sectors. A complete experiment requires: large b rates to measure small B’s & asymmetries in B & Bs ability to trigger on purely hadronic final states excellent mass & decay time resolution ability to identify leptons p/K/p ability to use final states with and po 75 Conclusions II Short term We will see a statistically significant measurement of sin(2) from e+e- B factories and CDF, HERA-b Perhaps a measurement of xs from CDF Beware of fit confidence levels where theoretical errors dominate Long term precision measurements of a, , , & ambiguity removals from BTeV & LHC-b with some important contributions from Atlas & CMS Ultimate tests will couple magnitudes l, |Vub/Vcb|, |Vtd/Vts| with the phase measurements where is essential Moreover, BTeV & LHCb are powerful enough experiments to do physics beyond that mentioned here which may become much more interesting in the future, ex: CPT tests 76