CP Violation in Kaon Decays (II) Taku Yamanaka, Osaka University To appear in the proceedings of the 50 years of CP violation conference, 10 { 11 July, 2014, held at Queen Mary University of London, UK. 1 Introduction Major progress has been made in kaon physics in the past 50 years. The number of observed + − KL π π events has increased by 6 orders of magnitude, and the observed CP violation ! was experimentally proven to be caused by a complex phase in the CKM matrix. This mechanism is now a fundamental piece of the standard model. Recent kaon experiments are now even searching for new physics beyond the standard model with K πνν decays. ! The branching ratios of K πνν decays are 7{8 orders of magnitude smaller than the !+ − 0 0 branching ratio of KL π π , and CP -violating KL π π decay is now a major 0! ! background for KL π νν experiments. ! This paper reviews the progress of kaon experiments in the US and Japan as requested by the conference organizer, and how the 6{7 orders of magnitude improvements were possible in the past 50 years.1 2 Quest for 0/ + − Soon after the discovery of CP violation [1], the KL π π decay was explained to be ! caused by an admixture of a CP -even component in the KL [2]: KL Kodd + Keven ; (1) j i ∼ j i j i where this CP -even component was decaying to a CP -even π+π− state. This admixture is introduced by a complex phase in the K0 K0 mixing. The next question was whether the − CP -odd Kodd can directly decay to a CP -even ππ state. Such process is called direct CP violation. If the direct CP violation exists, the ratios between decay amplitudes: + − + − 0 η± A(KL π π )=A(KS π π ) = + and (2) ≡ ! 0 0 ! 0 0 0 arXiv:1412.5906v1 [hep-ex] 18 Dec 2014 η00 A(KL π π )=A(KS π π ) = 2 (3) ≡ ! ! − can be different due to isospin. The existence of the direct CP violation can thus be tested by checking whether the double ratio: + − + − BR(KL π π )=BR(KS π π ) R ! 0 0 ! 0 0 (4) ≡ BR(KL π π )=BR(KS π π ) 2 ! ! η± = (5) η00 1 + 6Re(0/) (6) ' 1Unless noted, the years are given in published years. 1 deviates from 1 or not. The superweak model [3] explained that a very weak unknown interaction that changes the strangeness by 2 brings in the phase. However, the superweak model cannot violate ± CP in the KL ππ decay process because it cannot contribute to such a ∆S = 1 ! ± transition. 2.1 Advancement in Experimental Technologies To measure the double ratio R, high statistics is required. This means that both a higher kaon flux and detectors capable to collect data with higher rates are needed. 2.1.1 Production Target To get a higher kaon flux, the advancement of accelerators was essential, but there was also a change in production targets. In 1964, the experiment that first discovered CP violation + − with 35 KL π π events used an \internal target" in the accelerator ring as shown in ! Fig. 1. The target was a Be wire with 0.5 mm in diameter [4]. In 1969, at the CERN PS, the proton beam was extracted from the accelerator to + − bombard a 72-mm-thick tungsten target. About 400 KL π π events were collected [5], ! ten times more than in the first experiment. B 76 CHRIS TENSON RON I N FIT AN D TURLAY d 'Ca qW y, g{' FIG. 1. Plan vievr oof eexperiment. M~~ andn 3f2 are the spectrometer magnets. " Figure 1: Planseparated viewby, of-in. thegaps. neutralEach banka presented kaon experimenta tota1a pooroor geometry"eo at BNLand are AGS.thus attenuate For the experiment that sorption cross section 0~0 an d not the totao a cross section discovered CP violation, the regenerator was removed, and a He bag was installed between in ows. Adetailed "e p apparatus In a ition to the studies of coherent e sew ere.' report dirt hi r h as used to search the collimator and' ' spectrometers' [4]. c 2014 The Americanp p Physical Society. A scintillation countern er folloil e , covere the exit a e den e sp rk h ti"' "' 'nt 140000 events were recorded.re anticoincidence with III. DATA REDUCTIO N the other countersun ers guaranteed that thee Eo meson e anticoincidenc e counter. 2.1.2 ChargedVarious n Spectrometers d o . -mrad resolution. Of eliminated. f15 ' 35 nize d as events and measuredre Thee spectrometerss wereere plaaced syrnmetricall good by There was alsothe beam a changewith in detectorhe re technologies.generators the scanners. TheD 1 standard tracking devices in the e ucial hnes were calculate easure 1960s were spark chambers. They had an advantageangles on the that61m W' onlyith the ai the tracks of interest can be ' in in. of the re e i16d I e sp~c~ coo~d~~at~s made visible bya mean applyingm HV pulses/ for triggeredojtho e events. Mirrorsermmed all were arrangeda ion to capture fth t 1 vec tor momen en a eter- sparks in multiple spark chambers of~ each eventmined inthe ar single photograph. For the first CP e re ative E2' beam 0ux was Gloo d iion x ~ beam 1{-, tectining a fraction of th along the , an d the invv o t etwo- four-coun ace u stream body system ~ During processing the even ' ~ - 2 oft t d dt th tb om a ~~m~~~ origin and t c es not is monitor roug h' ' the spectrom eters. A total of 67 of the measurred y ssuitable when thee t ickness of the z ~ were 11 ried since the varyingvar 'n attentuatio ~~~~ly vents succes sfu y rocess The eventsn scovs covcover a mass ran e of 350-550M' have . However these E' ecays are seen in g ~ ~ H. Christenson * n IREE H. Christenson Cronin J. ' n3 . , . Croni, J. J. %. V urlay, Trans Nug S . 11,1 0 (i~), . R . L tt 1$, 138 (1964 0 violation experiment, tracks in the photographs were scanned by human \scanners" with digitized angular encoders [4]. 44 IEEE TRANSACTIONS ON NUCLEAR SCIENCE June In 1969, the experiment at the CERN PS [5] used an automatic \Luciole Flying Spot Digi- tizer" which could scan 1000 frames per hour. This used a CRT instead of a mechanical sys- tem to move a bright light spot quickly across a film and record the intensity of light passing through the film with a phototube [6]. Figure 2 VIEWING shows a similar spot scanner with a CRT, made SCREEN by the Univ. of Michigan. In the late 1960s, experiments started to Fig. 5-From reference 43. The light spot on the move away from using films. Experiments at Figure 2:face Aof the spotCRT scanneris controlled usingin position CRTby as a PDP-1 computer. It is focussed onto the CERN and BNL used ferrite-core readout sys- a movingfilm, lightof which source.the local transparency The lightis spot measured by the detector photomultiplier. tems to read out spark positions and recorded was controlledUse of optical by abeam PDP-1splitters computer,provides for anduse them on tape directly [7, 8, 9]. This readout was focusedof a reference ontophotomultiplier, the film, offor whichscanning the another film in parallel, if desired, and for system allowed the BNL experiment to collect local transparencydirect observation wasof the film measuredon a viewing by a + − screen. 9400 KL;S π π events, 300 times the first photomultiplier (DET. PM). @2014 IEEE. ! CP violation experiment. Reprinted, with permission, from [6]. In the early 1970s, experiments started to use multi-wire proportional chambers (MWPC). For example, the experiment that observed the + − first KL µ µ decays [10] used MWPCs with(0 5000) wires that had a 2-mm wire(b) spacing. ! + − I With the same detector, the experiment collected 2 M KL;S π π events [11]. ! I The geometry of spectrometers has also changed. The CP violation experiment in 1964 <- H `4~ used a double-arm spectrometer with two sets of quadrupole magnetsI and spark chambers V located at 22◦ from the neutral beam line as shown in Fig. 1. This geometry was optimized ± Cl for the 1.1-GeV/c average KL momentum. C3 Thin iron A later experiment at CERN [12]160 - inCMeV 1965iHn used al forward spectrometer,TargetI| with one dipole magnet sandwiched between four spark- - chambers1 --L-- asL shown in Fig. 3 to have a high accep- tance for pions from KL's with the average momentumLl L2 11 GeV/c. It also had a Cerenkovˇ _tr torIS counter between the downstream spark chambers for a particle identification. The forward--toe C2 L5 VS spectrometer became the standard for the experiments that followed. topper ~ >Cl C6 I Fig. 6-Schematic diagrams of uses of decision-making spark chamber. (a) From reference 47. Protons produced in the lithilm target by the reaction ir+ + Li6 -i He4 + p + p are stopped in the range chambers RI and R2. C1 through C6 are scintillators. The function of the 'decision-making" current-ratio chambers L1 through Ls is described in the text. (b) From reference 48. Elec- Figure 3: Plan view of an experimentaltrons apparatusfrom the decay ati- CERN-- e- + v + which1 are analyzed usedfor apolarization forward spec-by scattering in a magnetized iron foil. Similarly, positrons are obtained from ir + -* v + M +-* vl + + v-+ e+. H and V are trometer. SCi are spark chambers, Ni andrespectively Pi arehorizontal scintillators,and vertical Bscintillator is a bendinghodoscopes; magnet,C1, C2, andand C3 are scintillators; L2 is a thin-aluminum-electrode chamber; and L3 is a range chamber.