CLEO Contributions to Tau Physics
Total Page:16
File Type:pdf, Size:1020Kb
CORE Metadata, citation and similar papers at core.ac.uk Provided by CERN Document Server 1 CLEO Contributions to Tau Physics a Alan. J. Weinstein ∗ aCalifornia Institute of Technology, Pasadena, CA 91125, USA Representing the CLEO Collaboration We review many of the contributions of the CLEO experiment to tau physics. Topics discussed are: branching fractions for major decay modes and tests of lepton universality; rare decays; forbidden decays; Michel parameters and spin physics; hadronic sub-structure and resonance parameters; the tau mass, tau lifetime, and tau neutrino mass; searches for CP violation in tau decay; tau pair production, dipole moments, and CP violating EDM; and tauphysicsatCLEO-IIIandatCLEO-c. + + 1. Introduction and decay, e.g.,in[3]e e− Υ(nS) τ τ −. In addition to overall rate,→ one can→ search for Over the last dozen years, the CLEO Collab- small anomalous couplings. Rather generally, oration has made use of data collected by the these can be parameterized as anomalous mag- the CLEO-II detector [1] to measure many of the netic, and (CP violating) electric, dipole mo- properties of the tau lepton and its neutrino. Now + ments. With τ τ − final states, one can study that the experiment is making a transition from the spin structure of the final state; this is per- operation in the 10 GeV (B factory) region to the haps the most sensitive way to search for anoma- 3-5 GeV tau-charm factory region [2], it seemed lous couplings. We will return to this subject in to the author and the Tau02 conference organiz- section 9.1. ers to be a good time to review the contributions Measuring the production rate at 10 GeV is of the CLEO experiment to tau physics. We are only interesting if it is precise: < 1%. This has consciously omitting results from CLEO-I (data proven to be difficult at CLEO, for several rea- taken before 1989). The author has chosen to sons. take a semi-critical approach, emphasizing both First, it is desirable to do an inclusive selec- the strengths and weaknesses of tau physics at + tion of τ τ − final states, so that production rates CLEO: past, present, and future. don’t depend on decay branching fractions, which were not measured at the 1% level in the early 2. Tau Production at 10 GeV part of the last decade. It’s difficult to select taus + inclusively, because backgrounds from qq¯, e e−, We begin by discussing a topic on which CLEO + µ µ− and two-photon are less easily distinguish- has not published: tau pair production. + able from τ τ − than at LEP, and they depend on + In e e− collisions, one can study taus in pro- tau decay mode. It’s not really hard, but it’s hard duction and/or decay. The production reac- to get precise selection efficiencies. In the end, tau + + tion e e− γ∗ τ τ − is governed by well- → → selection at CLEO always had smaller efficien- understood QED. As such, it is not terribly inter- cies and/or larger backgrounds, and often also esting (or at least, nowhere near as interesting as larger systematic errors, than analogous analyses + ( ) + studying e e− γ∗/Z ∗ τ τ − at LEP I and → → at LEP. LEP II). CLEO has studied resonance production To measure the production cross-section, one needs a precise luminosity measurement, with an ∗Work supported by the US Department of Energy and National Science Foundation. error << 1%. At CLEO, we selected large angle 2 Table 1 3.1. One-prong problem Main decay modes of the τ lepton. In the early 90’s, the “tau one-prong” problem τ eνν 18% Br, univ, Michel was raging; the branching fractions for exclusively → τ ≈ τ µνν 17% Br, univ, Michel reconstructed tau decays didn’t add up to 1 [6]. → τ ≈ τ πν, Kν 12% Br, univ The resolution of this discrepancy required preci- → τ ≈ τ ππν 25% Br, ρ, ρ0,CVC,Π sion (sub-1%) branching fractions measurements. → τ ≈ τ Kπν 1.4% Br, K∗, K∗0 CLEO-II was a new detector; acceptances, → τ ≈ τ 3πν 18% Br, a , a0 , h in/efficiencies, and detector simulation needed to → τ ≈ 1 1 ντ τ Kππντ 0.8% Br, K1, K1b,W-Z be understood very well. Important issues in- → ≈ 0 τ 4πν 5% Br, ρ0,CVC cluded the detection of π ’s: they rarely merged → τ ≈ τ rare 2% 5π,6π, ηππ, ... into one shower, and CLEO obtained great m → ≈ γγ τ ηπν, b ν 1% 2nd-class currents resolution ( 6 MeV) with its CsI calorime- → 1 τ τ forbidden 1% neutrinoless decays ter. However,∼ soft photons could get lost; and → “splitoffs” from hadronic showers could fake pho- tons. Overall, ensuring reliable detection of π0’s resulted in a detection efficiency 50(1 0.03)%. Also, CLEO had poor K/π separation∼ over± most + + + Bhabhas (e e−), µ µ−, e e− γγ with rather of the interesting momentum range. We made high statistics. To get the luminosity,→ we need 0 progress using KS, with detection efficiency accurate predictions of the cross-section times se- 50%. Because of all this, it was hard to know the∼ lection efficiency from precision Monte Carlo sim- overall detection efficiency, after backgrounds, to ulations, incorporating accurate QED radiative better than 1%. corrections. Much less effort has gone into this at Ultimately, CLEO’s branching fraction mea- 10 GeV than at LEP! Each of these measurements surements were limited by knowledge of luminos- has small statistical errors, but the systematic er- ity (1%), cross-section (computed to order α with rors are of the order of 2%. CLEO got agree- KORALB [7], 1%), and knowledge of the de- ∼ ment between the 3 QED processes at the level tection efficiency∼ and backgrounds ( 1 2%). ∼+ − of 1% but not much better; the discrepancies are However, with millions of produced τ τ −,what likely to be in the QED Monte Carlos. CLEO we lacked in efficiency we made up for in statis- quotes a 1% systematic error on luminosity [4]. tics. The moral is that further progress in this By 1995, we made (1 2)% measurements of ∼ − 0 topic requires precision QED Monte Carlos. The branching fractions to eνντ , µνντ , π/Kντ , ππ ντ , 0 0 KKMC program [5] promises precision results, πnπ ντ ,3πντ ,3ππ ντ , etc.[8–11]. These mea- but it must be validated with careful computa- surements reduced the “tau one-prong” problem tional and experimental cross-checks. Consistent to insignificance by PDG 1996 [12]. We had also + + + results for e e−, µ µ−, e e− γγ are a good tested e/µ/τ charged-current coupling universal- → first step. Further, measurements of anomalous ity at the 1% level [8]. By then, LEP was mea- moments make use of the spin correlations in tau suring branching fractions with total errors much pair production at 10 GeV, so one must validate smaller than 1%. This came as quite a shock to the correct treatment of spin-dependence in the CLEO! It helped that the LEP experiments knew Monte Carlo. Nττ = σ quite well, as a by-product of their in- crediblyL successful electroweak program. 3. Tau decay physics at CLEO 4. Rare semi-hadronic decay modes CLEO pursued a systematic study of all tau de- cays during the 1990’s, learning much about the Rare semi-hadronic decay modes of the tau pro- tau, its neutrino, and low-energy meson dynam- vide unique laboratories for low-energy meson dy- ics. The main decay modes are listed in Table 1. namics, and tests of conservation laws. With the 3 Table 2 The ηπ− mode is a signature for second-class Measurements of branching fractions for rare (isospin-violating) currents. The ηK− mode pro- semi-hadronic τ decay modes by CLEO-II. 0 ceeds by SU(3)f violation. The ηπ−π mode proceeds by the Wess-Zumino anomalous current. + 0 3 The η3π signals, which are rich in sub-structure, (2π−π 2π ντ )=(5.3 0.4) 10− B ± × areshowninFig.2. (3π 2π+ν )=(7.8 0.6) 10 4 − τ − CLEO also observed the radiative decay modes B + 0 ± × 4 (2π−π 3π ν )=(2.2 0.5) 10− B τ ± × τ − e−ντ γ and τ − µ−ντ γ [22], and made the + 0 4 → → + (3π−2π π ντ )=(1.7 0.3) 10− first observation of the decays τ − e−e e−ν¯ ν B ± × → e τ 0 4 (5 events) and τ µ e+e ν¯ ν (1 event) [23]. (π−2π ων )=(1.5 0.5) 10− − − − µ τ B τ ± × → + 4 (2π−π ωντ )=(1.2 0.3) 10− B + 0 ± × 4 (3π−2π 2π ντ ) < 1.1 10− B × 50 100 0 6 45 (7π±(π )ντ ) < 2.4 10− . B × 40 80 35 60 30 25 40 20 Events/(50 MeV) Events/(50 MeV) Events/(50 15 Table 3 20 10 Measurements of τ decay modes involving kaons 5 0 0 from CLEO-II. 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 M(2π -π + 3π0) (GeV) M(3π-2π+π0) (GeV) 4 (ν ηπ−) < 1.4 10− at 95% CL B τ × (ν ηK )=(2.6 0.5) 10 4 τ − − Figure 1. Mass distributions from τ 6πντ .The B 0 ± × 3 → (ν ηπ−π )=(1.7 0.3) 10− data are shown as data points with error bars, the B τ ± × + 4 lower histograms show various background mod- (ντ ηπ−π π−)=(3.4 0.8) 10− B 0 0 ± × 4 els, and the open solid histogram shows the model (ν ηπ−π π )=(1.4 0.6) 10− B τ ± × of the background plus tau decay signal.