MITP/13-089 12 July 2013 Conclusions of the MITP Workshop on T Violation and CPT Tests in Neutral-Meson Systems K. R. Schubert1,2, L. Li Gioi3, A. J. Bevan4 and A. Di Domenico5 Date of the Workshop: 15 - 16 April 2013 Participants: J. Bernabeu6, A. J. Bevan4, G. D’Ambrosio7, A. Denig2, A. Di Domenico5, H.-J. Gerber8, W. Gradl2, M. Heck9, T. Hurth2, J. S. Lange10, L. Li Gioi3, M. Neubert2, 4 T. Ruf11, K. R. Schubert1,2 and P. Villanueva-Perez6 1 0 The two-day workshop took place in the Institut fu¨r Kernphysik, Universit¨at Mainz in a very 2 lively and friendly atmosphere, and the participants thank MITP very much for providing the n frame for our presentations and discussions. Half of the time was used for discussions, and in fact a the workshop continued by a number of e-mail exchanges until summer 2013. This summary does J notcoverallcontributions,buttheyareavailableontheIndigopageoftheworkshop[1]. Thefour 7 parts of the summary are: 2 1. T Violation in Decays of Neutral B Mesons, K. R. Schubert ] x 2. T and CPT studies in B0B0 transitions with Belle, L. Li Gioi e - 3. Future Measurements of T violation in B and D decays, A. J. Bevan p 4. Direct tests of T and CPT symmetries in the entangled neutral Kaon system at a Φ factory, e h A. Di Domenico [ The first part covers the central point of the workshop, the interpretation of CP and T violation 1 in the interplay of B0B0 transitions and decays B →J/ψK. It also covers the continued and very v helpful discussions with H.-J. Gerber, T. Ruf, F. Martinez-Vidal, P. Villanueva-Perez and A. Di 8 Domenico. Parts 2 to 4 cover the prospects of future experiments with K, D and B mesons. 3 9 6 The most sensitive tests of CPT symmetry remain the Bell-Steinberger analyses of the K0K0 sys- 1. temusing unitaritywhich connectsthe CP-symmetrypropertiesofall observed KS and KL decay 0 modes with the CPT- and T-sensitive overlap (cid:104)K |K (cid:105). These analyses started in 1970 and have L S 4 reached the impressive sensitivity of |m(K0)−m(K0)| < 4·10−19 GeV at 95% C.L. in 2012, as 1 presented by G. D’Ambrosio at the workshop. An open question remains by how much invisible : v decays of neutral K mesons can influence the result. How well is unitarity tested experimentally? i How much does the product of lifetime and the sum of all measured partial decay rates deviate X fromunityforK andK decays? AndhowmuchwouldtheerrorsonRe(cid:15)andImδ increaseifthe r S L a invisible modes would have maximal CP violation? As long as this is not answered quantitatively by experiments, “direct” tests of CPT symmetry remain important. 1 Institut fu¨r Kern- und Teilchenphysik, Technische Universit¨at Dresden, Germany 2 Institut fu¨r Kernphysik, Johannes-Gutenberg-Universit¨at Mainz, Germany 3 Max-Planck-Institut fu¨r Physik, Mu¨nchen, Germany 4 Queen Mary, University of London, United Kingdom 5 Dipartimento di Fisica, Sapienza Universita di Roma, and INFN, Roma, Italy 6 IFIC, Universitat de Valencia-CSIC, Valencia, Spain 7 INFN, Napoli, Italy 8 IPP, ETHZ, Zu¨rich, Switzerland 9 Institutfu¨rExperimentelleKernphysik,KarlsruherInstitutfu¨rTechnologie,Karlsruhe,Germany 10 Justus-Liebig-Universit¨at, Giessen, Germany 11 CERN, Geneva, Switzerland 1 1 T Violation in Decays of Neutral B Mesons Klaus R. Schubert Abstract: The CP-violating observable Im λ with λ=qA(B0 →J/ψK0)/pA(B0 →J/ψK0) can be written as |A/A|·Imλ˜, where λ˜ = qA|A|/(pA|A|). In this product, |A/A| is CPT violatingandImλ˜isTviolating. Therefore,observationofImλ(cid:54)=0inthesin∆mttermofthe time-dependentrateofB0 →J/ψK decaysisaproofofT-symmetryviolation. CPTviolation S inthesedecayswouldleadtoanon-vanishingcos∆mttermintherate. Thefirstmeasurements of Imλ(cid:54)=0 in 2001 already prove T violation. The BABAR 2012 analysis demonstrates that Imλ (cid:54)= 0 leads to a “motion-reversal” difference in the rates for the transitions B0,B0 → B ,B and B ,B →B0,B0. + − + − The notion “Time Reversal Violation” has two different meanings, either breaking the symme- try of the transformation t → −t in the fundamental laws of an interaction, or unequal motions (in classical mechanics) or evolutions (in quantum mechanics) for two systems when exchanging initial and final state and reversing velocities and spins [2, 3]. I will use the term “T-symmetry violation” for the first and “motion-reversal violation” for the second case. In particle physics we use motion-reversal experiments, like the comparison of the two time-dependent rates for the transitions K0 →K0 and K0 →K0, as a method to test T-symmetry of the weak interaction [4]. Since the recent BABAR experiment [5, 6] on motion reversal in the transitions B0 → B and − B →B0 has been a central discussion point in this workshop, this note summarizes its relevance − for demonstrating T-symmetry violation. B0B0 transitions are well known to be sensitive to the symmetries CP, T and CPT. In the Weisskopf-Wignerapproximation,theevolutionofthetwo-dimensionalB0 stateΨ=ψ B0+ψ B0 1 2 is given by iψ˙ =Λ ψ , i ij j whereΛ =m −iΓ /2withtwotime-independenthermitean2x2matrices. Thegeneralsolution ij ij ij is ψ (t)=U (t)ψ (0) , i ij j where the time-dependent matrix U (t) follows unambiguously from Λ [7, 8]. The transition ij ij rates |(cid:104)ψ |U|ψ (cid:105)|2 are determined by six real observables ∆m = m −m , Γ = (Γ +Γ )/2, f i H L H L ∆Γ=Γ −Γ , H L Reδ+i Imδ = (m22−m∆11m)/2−−i∆i(ΓΓ/222−Γ11)/4 and (cid:12)(cid:12)(cid:12)(cid:12)pq(cid:12)(cid:12)(cid:12)(cid:12)=(cid:12)(cid:12)(cid:12)(cid:12)(cid:12)(cid:114)ΛΛ2112(cid:12)(cid:12)(cid:12)(cid:12)(cid:12) . A non-zero phase between m and Γ leads to |q/p| (cid:54)= 0, violating CP and T; Λ (cid:54)= Λ leads 12 12 11 22 to δ (cid:54)= 0, violating CP and CPT. All six observables have been measured [9], and ∆Γ, Reδ, Imδ and |q/p|−1 are compatiblele with zero. Within experimental errors, the Hamiltonian for B0B0 transitions is invariant under all transformations T, CPT and CP. DecaysofneutralBmesonsintoJψK andJψK aredeterminedbyonlythreemorerealobserv- S L ables, if they are dominated by a single weak amplitude, |A|=|(cid:104)JψK0|D|B0(cid:105)| , |A|=|(cid:104)JψK0|D|B0(cid:105)| , and Imλ, where λ is defined in a phase-convention independent way by qA λ= . (1) pA The single-amplitude condition (no “direct T violation”) is assumed in this summary and can be tested in the data, as presented later. The operator D is the Hamiltonian for the decay, and 2 |A|2/Γ is the fraction of B0 → J/ψK0 decays. The relevance of |A/A| and Imλ is easily seen by introducing the factorization (cid:12) (cid:12) λ=(cid:12)(cid:12)A(cid:12)(cid:12)·λ˜ with λ˜ = qA|A| . (2) (cid:12)A(cid:12) pA|A| The observables |A/A| and λ˜ describe the CPT and T symmetry properties of the decays. If the matrix element (cid:104)JψK0|D|B0(cid:105) is given by a single weak amplitude, then CPT invariance of D requires |A/A| = 1 [10, 11], T invariance requires Imλ˜ = 0 [12, 11], and CP invariance requires |A/A|=1 and Imλ˜ =0. Decays of neutral B mesons into flavour-specific states, e. g. into (cid:96)±νX, are given by only two more observables, |A |=|(cid:104)(cid:96)+νX|D|B0(cid:105)| and |A |=|(cid:104)(cid:96)−νX|D|B0(cid:105)| . (cid:96) (cid:96) The equivalent to Imλ˜ is not present if the “∆Q=∆b” rule is strictly valid, i. e. (cid:104)(cid:96)−νX|D|B0(cid:105)=(cid:104)(cid:96)+νX|D|B0(cid:105)=0 . The following text consists of two parts. In the first part I present the consequences of CPT and T symmetry for the transitions B0,B0 → J/ψK ,K and B ,B → B0,B0 as derived by S L − + H.-J. Gerber [11] in collaboration with M. Fidecaro and T. Ruf. In the second part I discuss the consequences for the 2001 analyses of BABAR [13] and Belle [14] and for the 2012 analysis of BABAR [5]. I will use some additional simplifying assumptions which have no influence on the main conclusions, |q/p|=1, δ =0, ∆Γ=0, √ √ K =(K0+K0)/ 2, K =(K0−K0)/ 2 , (3) S L and validity of the “∆Q = ∆b” and “∆S = ∆b” rules in flavour-specific and J/ψK decays, respectively. This leads to √ √ A =A(B0 →JψK )=A/ 2 , A =A(B0 →JψK )=A/ 2 , S S S S √ √ A =A(B0 →JψK )=A/ 2 , A =A(B0 →JψK )=−A/ 2 . L L L L The evolution operator U(t) is given by (cid:18) (cid:19) cos(∆mt/2) isin(∆mt/2)·p/q U (t)=e−Γt/2 , (4) ij isin(∆mt/2)·q/p cos(∆mt/2) and ∆S =∆b means that the decay matrix is diagonal, (cid:18) A(B0 →J/ψK0) A(B0 →J/ψK0) (cid:19) (cid:18) A 0 (cid:19) D = = . (5) A(B0 →J/ψK0) A(B0 →J/ψK0) 0 A The decay rate R=R(B0 →J/ψK |t) of a B0 at time t=0 into the final state J/ψK at time t S S is R= 1e−Γt(cid:12)(cid:12)(cid:12)(cid:0) A A (cid:1)(cid:18) cos(∆mt/2) isin(∆mt/2)·p/q (cid:19)(cid:18) 1 (cid:19)(cid:12)(cid:12)(cid:12)2 2 (cid:12) isin(∆mt/2)·q/p cos(∆mt/2) 0 (cid:12) e−Γt =|A|2· [1+κ−κ·cos(∆mt)−Imλ·sin(∆mt)] , (6) 2 with the CPT-violating parameter |λ|2−1 κ= (7) |λ|2+1 and Imλ from eq. 1. I also assume κ (cid:28) 1, i. e. |λ|2 = 1+2κ. If CPT is conserved, κ = 0 and the cosine term has to vanish. If T is conserved, Imλ = 0 and the sinus term has to vanish. If CP is conserved, the rate has to be a pure exponential. With good statisics and well-understood systematics, tests of CP, T and CPT can be made with a single time-dependent measurement of 3 the rate for B0 →J/ψK0 decays. Additional rate measurements of B0 →J/ψK0, B0 →J/ψK0 S L S and B0 →J/ψK0 improve statistics as well as systematic uncertainties; but the full physics infor- L mation is already contained in the time dependence of B0 →J/ψK0 decays alone. S Thederivationoftherateineq.6allowsasimpleargumentforthefactthatIm(qA/pA)(cid:54)=0proves T violation of the Hamiltonian [15]. Im(qA/pA) (cid:54)= 0 violates CP symmetry which implies CPT and/or T violation. CPT symmetry requires only |A/A| = 1 and nothing for the phase of A/A. Since Im(qA/pA)(cid:54)=0 does not contradict CPT symmetry, it must violate T. The same conclusion that Imλ(cid:54)=0 is CP- and T-violating is found by P. Villanueva-Perez [16]. The rates for B0 →JψK0, B0 →JψK0 and B0 →JψK0 follow from replacing (A,A) and (1,0) L S L in eq. 6 by (A,−A) and/or (0,1). All four rates are given by the same expression e−Γt R=|A|2· [1+κ+C·cos(∆mt)+S·sin(∆mt)] , (8) 2 with different values for the coefficients C and S as given in Table 1. Table 1: Coefficients C and S for the rates in eq. 8. C S B0 →JψK0 −κ −Imλ S B0 →JψK0 −κ +Imλ L B0 →JψK0 +κ +Imλ S B0 →JψK0 +κ −Imλ L TheTviolationinallfourratesisgivenbythesameparameterImλ, theCPTviolationinallfour by κ. From entangled B0B0 pairs with tagging by flavour-specific decays, the B0 rates obtain an extra factor |A |2, the B0 rates |A |2. The four rates are only equal, up to the signs in the table, (cid:96) (cid:96) if |A |=|A |, i. e. CPT symmetry in flavour-specific decays. (cid:96) (cid:96) With the Kaon sign-conventions in eq. 3, the states B and B [17, 18, 19], + − B =N(AB0−AB0) , B =N(AB0+AB0) , N−2 =|A|2+|A|2 =2|A|2(1+κ) (9) + − have the properties that B decays only into J/ψK0, not into J/ψK0, and B only into J/ψK0, + L S − S not into J/ψK0. Since the amplitudes in eq. 5 are defined as A = (cid:104)J/ψK0|D|B0(cid:105) and A = L (cid:104)J/ψK0|D|B0(cid:105),thestatesineq.9areingoingstates|B (cid:105)and|B (cid:105)andnottheoutgoingstatesin + − the transitions B0,B0 → B ,B . With direct CPT violation, the two states are not orthogonal, + − (cid:104)B |B (cid:105)=κ. The orthogonal states to B and B are + − + − B =N(A∗B0+A∗B0) and B =N(A∗B0−A∗B0) , (10) +⊥ −⊥ respectively,andthenormalizationfactorN isthesameasineq.9. IndependentofCPTinvariance, an Υ(4S) meson decays into the entangled two-body state B0B0−B0B0 =B B −B B =B B −B B , (11) + +⊥ +⊥ + − −⊥ −⊥ − where the first B in this notation moves in direction p(cid:126) and the second one in direction −p(cid:126). With CPT invariance, B = B and B = B . With κ (cid:54)= 0, B decays into both JψK and +⊥ − −⊥ + +⊥ S JψK . But if the first B decay of the entangled pair is JψK , then the remaining B is in L S the state B . The state B is not used in the following motion-reversal discussion, its only + +⊥ purpose is the preparation of the state B . The analoguous argument holds for B . The decays + −⊥ Υ(4S) → B0B0 → (J/ψK )(J/ψK ) and (J/ψK )(J/ψK ) are forbidden in accordance with S S L L 4 Bose statistics. As a side remark, not relevant for T violation, the two states B and B are well + − defined physical states like the mixing eigenstates B and B , free of phase conventions, but all H L four are not CP eigenstates. The decay rates of B and B are + − Γ =Γ(B →JψK )=Γ(B →JψK )=|A|2(1+κ) , (12) ψK − S + L where the factor 1+κ originates from the normalization N in eq. 9. Dividing the rates R(t) in eq. 6 by this decay rate leads to e−Γt R (t)= [1+C ·cos(∆mt)+S ·sin(∆mt)] , (13) 1 2 1 1 with the coefficients C and S as given in Table 2, using Imλ˜ =Imλ/(1+κ)=Imλ(1−κ). Note 1 1 that these rates R (t)=R(t)/Γ are not the rates for the transitions B0,B0 →B ,B because 1 ψK + − of the difference between in- and outgoing states. Calculation of the time-dependent rates R (t) for the transitions B0,B0 → B ,B requires the 2 + − outgoing states (cid:104)B |=N(A∗(cid:104)B0|−A∗(cid:104)B0|) , (cid:104)B |=N(A∗(cid:104)B0|+A∗(cid:104)B0|) , (14) + − leading to e−Γt R (t)= [1+C ·cos(∆mt)+S ·sin(∆mt)] , (15) 2 2 2 2 with C and S in Table 2. The motion-reversed transitions B ,B →B0,B0 require the ingoing 2 2 + − states of B and B . Their rates R (t) are found to be + − 3 e−Γt R (t)= [1+C ·cos(∆mt)+S ·sin(∆mt)] , (16) 3 2 3 3 with C and S in Table 2. 3 3 Table 2: Coefficients C and S , i = 1 for the rates R(B → J/ψK)/Γ in eq. 13, i = 2 for the i i ψK transitions in eq. 15, and i=3 for the transitions in eq. 16. C S C S C S 1 1 2 2 3 3 B0 →J/ψK −κ −Imλ˜ B0 →B +κ −Imλ˜ B →B0 +κ +Imλ˜ S − − B0 →J/ψK −κ +Imλ˜ B0 →B +κ +Imλ˜ B →B0 +κ −Imλ˜ L + + B0 →J/ψK +κ +Imλ˜ B0 →B −κ +Imλ˜ B →B0 −κ −Imλ˜ S − − B0 →J/ψK +κ −Imλ˜ B0 →B −κ −Imλ˜ B →B0 −κ +Imλ˜ L + + The time-dependent motion-reversal asymmetries like between B0 →B and B →B0, − − R (t)−R (t) C −C S −S A = 3 2 = 3 2 ·cos(∆mt)+ 3 2 ·sin(∆mt) , (17) MR R (t)+R (t) 2 2 3 2 are sensitive to only T-symmetry violation; (C −C )/2=0 and (S −S )/2=S =±Imλ˜. The 3 2 3 2 3 time-dependent quasi-motion-reversal asymmetries like between B0 →J/ψK and B →B0, S − R (t)−R (t) C −C S −S A = 3 1 = 3 1 ·cos(∆mt)+ 3 1 ·sin(∆mt) , (18) QMR R (t)+R (t) 2 2 3 1 are sensitive to both CPT- and T-symmetry violation; (C −C )/2=C =±κ and (S −S )/2= 3 1 3 3 1 S =±Imλ˜. 3 5 The CP-violation analyses [13, 14] have combined the transitions B0,B0 → JψK ,K and S L B ,B →B0,B0 byusingeventswithbothsignsof∆t=t(decay toJ/ψK )−t(decay to(cid:96)±X). − + S,L Recalling that ∆t = +t and ∆t = −t for the events with i = 1 and i = 3 in Table 2 respectively, andinspectingthesignsofC andC , showsthatthedeterminationof|λ|intheseanalysesisalso 1 3 a determination of κ. Accidental cancellations between direct T violation (two decay amplitudes with different phases giving leading to |λ|=(cid:54) 1) and direct CPT violation (κ(cid:54)=0) are very unlikely. Therefore, the analyses determine Imλ˜ and κ and show T-symmetry violation with a significance of about 4σ independent of any assumption on CPT [11]. The recent BABAR analysis [5] has for the first time separated the CP-violation data into events with ∆t > 0 and ∆t < 0 and has determined eight uncorrelated coefficients C and S for the i i cos∆mt and sin∆mt terms. All C values are compatible with CPT symmetry, and all S values i i proveT-symmetryviolation. Sinceearlieranalysescanleadtothesameconclusion,themainmerit oftheanalysisinref.[5]isitsdemonstrationofT-symmetryviolationbymotion-reversalviolation. The data can only use quasi-motion-reversals, comparing e. g. B0 → J/ψK and B → B0, and S − not motion-reversals with B0 → B and B → B0. However, the null results for the eight − − determinations of κ justify this approximation. References [1] https://indico.cern.ch/conferenceDisplay.py?ovw=True&confId=244839 [2] R. G. Sachs, “The Physics of Time Reversal”, The University of Chicago Press (1987) [3] K. R. Schubert, presented at this workshop, see slides at https://indico.cern.ch/conferenceDisplay.py?confId=244839 [4] T. Ruf, presented at this workshop, see slides at https://indico.cern.ch/conferenceDisplay.py?confId=244839 [5] J. P. Lees et al (BABAR), Phys. Rev. Lett. 109, 211801 (2012) [6] J. Bernabeu, presented at this workshop, see slides at https://indico.cern.ch/conferenceDisplay.py?confId=244839 [7] G. Branco, L. Lavoura and J. P. Silva, “CP Violation”, Clarendon Press Oxford (1999) [8] M. Fidecaro and H. J. Gerber, Rep. Prog. Phys. 69, 1713 (2006) [9] Particle Data Group, J. Phys. G 37, 075021 (2012) [10] T. D. Lee, R. Oehme and C.-N. Yang, Phys. Rev. 106, 340 (1957) [11] H. J. Gerber, presented at this workshop, see slides at https://indico.cern.ch/conferenceDisplay.py?confId=244839 [12] C. P. Enz and R. R. Lewis, Helvetica Physica Acta 38, 860 (1965) [13] B. Aubert et al (BABAR), Phys. Rev. Lett. 87, 091801 (2001) [14] K. Abe et al (Belle), Phys. Rev. Lett. 87, 091802 (2001) [15] H. J. Gerber and T. Ruf, private communication [16] P. Villanueva-Perez, presented at this workshop, see slides at https://indico.cern.ch/conferenceDisplay.py?confId=244839 [17] M. C. Banuls and J. Bernabeu, Phys. Lett. B 464, 117 (1999) [18] E. Alvarez and A. Szynkman, Mod. Phys. Lett. A 23, 2085 (2008) [19] J. Bernabeu, F. Martinez-Vidal and P. Villanueva-Perez, arXiv:1203.0171v2 (2012) 6 2 T and CPT studies in B0B¯0 transition with Belle Luigi Li Gioi The Belle collaboration, having performed experiments at the KEKB B-factory since 1999, made the essential observations of the CP violation, proving the differences in the decays of B mesons compared to their anti-particles B¯0. The KEKB B-factory is currently under upgrade to a new generation of super flavor factory (Su- perKEKB)whichaimstodelivermorethan50ab−1bytheendof2022. Togetherwiththemachine, the Belle detector upgrade is ongoing (Belle II) [2]. The main redesign goals are to cope with the muchhigherphysicsratesandthemuchlargerbackground,aswellasimprovingtheoverallphysics performance. For T and CPT violation studies the new tracking system plays a key role. The tracking system in the former Belle detector consisted of 4 layer of Si strip vertex detector (SVD), followed by a Central Drift Chamber (CDC). For the new tracking system a two-layer pixel detec- tor (PXD) for the innermost Si layers will be implemented. The SVD will be replaced entirely as well as the CDC: due to the harsh backgrounds the inner radius of the CDC has to be moved out andthetwoouterlayersofthenew SVDwillcoverthegap. Themomentum resolutionofcharged particles will be improved by extending the CDC to a larger radius. The impact parameters: d and z , defined as the projections of the distance from the point of 0 0 closest approach to the origin, are a good measure of the overall performance of the tracking systemandassuchareusedtofindtheoptimaltrackerconfiguration. Animprovementofroughly a factor two is expected on the impact parameter resolution. The introduction of a new vertex fitter together with the improvement of the alignment procedure will thus sensibly improve the systematic error of any time dependent measurement. 2.1 CPT Violation InthepresenceofCPTviolation,theflavorandmasseigenstatesoftheneutralBmesonsarerelated by |B >= p(cid:112)(1−z)|B0 > +q(cid:112)(1+z)|B¯0 > and |B >= p(cid:112)(1+z)|B0 > −q(cid:112)(1−z)|B¯0 > L H where |B > and |B > are the light and heavy mass eigenstates. Here z is a complex parameter L H accounting for CPT violation; CPT is violated if z (cid:54)= 0. Then, the time dependent decay rate of the two B mesons generated from the Υ(4S)→B0B¯0 is given by [3]: Γ |η |2+|η |2 (cid:18)∆Γ (cid:19) (cid:18)∆Γ (cid:19) P(∆t,f ,f )= de−Γd|∆t|[ + − cosh d∆t −Re(η∗η )sinh d∆t rec tag 2 2 2 + − 2 |η |2−|η |2 + + − cos(∆m ∆t)+Im(η∗η )sin(∆m ∆t)] (1) 2 d + − d where η =A A¯ −A¯ A , η =(cid:112)(1−z2)(p/qA A −q/pA¯ A¯ )+z(A A¯ +A¯ A ), A and A + 1 2 1 2 − 1 2 1 2 1 2 1 2 1 2 are the decay amplitudes of the reconstruction and tag side B mesons to f and f final states. 1 2 The Belle collaboration measured the CPT-violating parameter z and the normalized total decay- width difference ∆Γ /Γ [4] in B0 → J/ψK0 (K0 = K ,K ), B0 → D(∗)−h+ (h+ = π+ for D− d d S L and π+,ρ+ for D∗−), and B0 → D∗−l+ν (l+ = e+,µ+) decays. Most of the sensitivity to Re(z) l and∆Γ /Γ isobtainedfromneutralB-mesondecaystof ,whileIm(z)isconstrainedprimarily d d CP from other neutral B-meson decay modes. The results are based on a data sample of 535×106BB¯ pairs collected at the Υ(4S) resonance: Re(z) = [+1.9±3.7(stat)±3.3(syst)]×10−2,Im(z) = [−5.7±3.3(stat)±3.3(syst)]×10−3 and∆Γ /Γ =[−1.7±1.8(stat)±1.1(syst)]×10−2,allofwhich d d are consistent with zero. This is the most precise single measurement of these parameters in the neutral B-meson system to date. The dominant contributions to the systematic uncertainties are fromvertexreconstructionandthetag-sideinterference(TSI)[5];thenextlargestcontributionsare from fit bias. Using the new tracking system of the Belle II detector and refining the analysis will then permit to sensibly reduce the amount of the systematic uncertainty and to fully exploit the 7 larger data sample that will be collected by the Belle II experiment. Assuming a final integrated (cid:112) luminosity of 50 ab−1, an improvement of 50/0.5=10 is expected. 2.2 T Violation The CPLEAR collaboration reported on the observation of time-reversal symmetry violation throughacomparisonoftheprobabilitiesofK0 transformingintoK¯0 andK¯0 intoK0 asafunction of the neutral-kaon eigentime t. An average decay-rate asymmetry < A >= [6.6±1.3(stat)± T 1.0(syst)]×10−3 was measured over the interval 1τ <τ <20τ [6]. S S In the case of the B0B¯0 mesons the asymmetry is then expected to be close to zero; a value sig- nificantly larger than 10−3 would be an indication of new physics [7]. TheBellecollaborationperformedthismeasurementusingthesemileptonicdecaysoftheneutralB meson[8]: apossibledifferencebetweenthetransitionsB0 →B¯0 andB¯0 →B0 canmanifestitself as a charge asymmetry in same-sign dilepton events in Υ(4S) decays when prompt leptons from semileptonic decays of neutral B mesons are selected. Using a data sample of 78 fb−1 recorded at the Υ(4S) resonance and 9 fb−1 recorded at an energy 60 MeV below the resonance, Belle measures A =[1.1±7.9(stat)±8.5(syst)]×10−3. The dominant contributions to the systematic sl uncertainties are from Track finding efficiency and Continuum subtraction; they can be reduced with a better knowledge of the tracking system and a refinement of the analysis. Recently the BaBar collaboration reported a measurement of T-violating parameters in the time evolution of neutral B mesons [9] using the decays of entangled neutral B mesons into definite flavor states (B0 or B¯0), and J/ψK or cc¯K final states with the comparisons between the prob- L S abilities of four pairs of T-conjugated transitions as a function of the time difference between the two B decays [10]. The results obtained by the BaBar collaboration are in agreement with CPT conservation. At this point the Belle collaboration would not expect that repeating the BaBar analysis would improve the results significantly. However there is an interest in these results as a test of CPT. The experimental method consists in dividing the very same ∆t distribution used in the standard CP analysis in two parts: ∆t > 0 and ∆t < 0, and fitting them using the same function used for the CP violation analysis: gα±,β(∆τ)=e−Γd∆τ{1+Sα±,βsin(∆md∆τ)+Cα±,βcos(∆md∆τ)} (2) where ∆Γ = 0 is assumed, the indices α = l+l− (flavor state), β = K ,K (CP states) and the d S L symbol+or−indicateswhetherthedecaytotheflavorfinalstateαoccursbeforeorafterthede- caytotheCPfinalstateβ. Γ istheaveragedecaywidthand∆m isthemassdifferencebetween d d the neutral B mass eigenstates. Since, using the data sample collected by the Belle experiment in the standard CP violation analysis [11], no tension is observed in the fit of the ∆t distributions between the left and right sides of the distributions and between J/ψK and cc¯K , this analysis L S could not yield results compatible with CPT violation. The measurement of the T-violation parameters could become important for the Belle II collab- oration, when a very high experimental precision will be achieved. At that time also a number of measurements should be accessible to test T symmetry invariance in b→u,d and s transitions as well as in the charm sector to test c → d and c → s transitions [12]. It is also considered an advantagethatanyCPandTmeasurementshouldyieldaboutthesameresultifCPTisconserved. This would allow a cross check of any CP measurement. 8 References [1] A.Abashianetal.(BelleCollaboration),Nucl.Instrum.MethodsPhys.Res.Sect.A479,117 (2002);seealsodetectorsectioninJ.Brodzickaetal.,Prog.Theor.Exp.Phys.(2012)04D001. [2] I. Adachi et al., sBelle Design Study Report, KEK Report 2008-7 (2008). [3] B. Aubert et al. (BaBar Collaboration), Phys. Rev. D 70, 012007 (2004); B. Aubert et al. (BaBar Collaboration), Phys. Rev. Lett. 92, 181801 (2004). [4] T. Higuchi et al., (Belle Collaboration),Phys. Rev. D 85, 071105 (2012); [5] O. Long, M. Baak, R. N. Cahn, and D. Kirkby, Phys. Rev. D 68, 034010 (2003). [6] A. Angelopoulos et al. (CPLEAR Collaboration), Phys. Lett. B 444, 43-51 (1998). [7] S.Laplace,Z.Ligeti,Y.Nir,andG.Perez,Phys.Rev.D65,094040(2002),referencestherein. [8] E. Nakano et al., (Belle Collaboration), Phys. Rev. D 77, 112002 (2006). [9] J. P. Lees et al. (BaBar Collaboration), Phys. Rev. Lett. 109, 211801 (2013). [10] J. Bernabu, F. Martnez-Vidal and P. Villanueva-Prez, JHEP 08, 064 (2012). [11] I. Adachi et al., (Belle Collaboration), Phys. Rev. Lett. 108 171802 (2012). [12] Bevan, Inguglia, Zoccali arXiv:1302.4191. 9 3 Future Measurements of T violation in B and D decays Adrian J. Bevan Abstract: The potential for future measurements of T violation in B and D decays is sum- marisedhere. Thisdiscussionconsiderspossiblequarkflavourchangingtransitionsfromband c quarks to all kinematically accessible final states via tree and loop topologies. There is scope to extend the tests of the T symmetry using quantum entangled pairs of neutral mesons outlined in Refs [1, 2, 3], and performed recently by BABAR [4], to other CP filter final states of B decays as well as for some charm decays. These measurements can be used to test the nature of the T symmetry under quark transitions at tree and loop level in b → c, s, u, and d transitions, as well as c → d and s transitions. The loop transition c → u is experimentally inaccessible given that the corresponding loop is Cabibbo suppressed (i.e. by a factor of |V | ∝ λ3) and more copious loop transitions will dominate any attempt to extract ub that term. There are two categories of orthonormal CP basis filters that can be used for such measurements: (i)theapproximatelyorthonormalpairofT-conjugatedecaysX+K0 andX+K0, S L and (ii) T self-conjugate decays of a pseudoscalar to two spin-one particles where a transversity analysis of the final state allows one to experimentally distinguish between CP even and CP odd parts of the decay. This summary is based on Ref. [5] and naive numerical estimates prepared for this workshop. These estimates are obtained using existing experimental results on time- dependent CP asymmetry measurements from the B Factories. Only uncertainties are quoted as it would be incorrect to extrapolate central values for the CP asymmetry parameters in terms of the T asymmetry parameters ∆S± without re-analysing the data. T 3.1 B decay measurements at the Υ(4S) The BABAR and Belle experiments already have data that can be used to perform a number of T violation tests beyond the b → c transitions described in Ref. [4]. In the near future Belle II is expected to accumulate 50ab−1of data which will enable one to perform precision measurements of time-dependent T-violating asymmetries that can be related to the CKM matrix description of quark flavor changing transitions. T violation has been established in entangled B meson decays to combinations of flavor tagged final states denoted by (cid:96)±X and CP filter final states B and + B , taken to be ccK . The presence of two sets of orthonormal filters enables one to test T , − S,L CP and CPT via the different pairings of events as a function of proper time difference ∆τ as outlined by Bernabeu et al. In analogy with the b→c transitions proposed by Bernabeu et al., it ispossibletostudyT -violationintheinterferencebetweenmixinganddecayamplitudesinb→s transitions involving η(cid:48)K , φK and ωK CP filter pairs, however the B Factoriesonly have S,L S,L S,L sufficient data for the first two measurements as B → ωK0 has not yet been observed. Nonethe- L lessonecanestimatetheanticipatedprecisionattainableonatestofT usingtheωK CP filters. S,L ThesetofB andB involvingK0 andK0 isapproximatelyorthonormal(thisisagoodapprox- + − S L imation given current levels of experimental precision). As mentioned above it is also possible to define an exactly orthonormal basis pair for the B and B filters as illustrated in the following. + − The b→c transition B →J/ψK∗0 is composed of three P wave parts, one is a longitudinal com- ponent that is CP even and there are two transverse components: the perpendicular (CP odd) and parallel (CP even) parts. As a result it is possible to define an exactly orthonormal basis of B and B decays into the J/ψK∗0 final state (the same is true for other decays of a neutral + − pseudo-scalar meson to two spin-one particles). In analogy with this discussion, one can also test the T symmetry using the CP even and odd parts of other modes, such as the decay B →D∗D∗ (a b → d quark transition), B → φK∗ (a b → s quark transition), and B → ρ0ρ0 (a b → u quark transition) as a basis of CP filters. Estimates of the experimental precisions attainable for these modes, where possible to compute, canbefoundinTable1. ItislikelythatBelleIIwillbeabletoobserveT violationatasignificance 10