Rare Kaon Decays Experiments

1
Rare Kaon Decays Experiments

• Augusto Ceccucci/CERN
• LNF Spring School, May 17-18, 2004

2
Why study Rare Kaon Decays

• Search for explicit violation of Standard Model
• Lepton Flavour Violation
• Probe the flavour sector of the Standard Model
• FCNC
• Test fundamental symmetries
• CP,CPT
• Study the strong interactions at low energy
• Chiral Perturbation Theory, kaon structure

• I will give a review of recent experimental
  results
• In addition to KL and K, also KS rare decays
  (BR10-8) start to be studied
• I will also briefly review the new initiatives
  that should lead to significant advance in the
  field by the end of this decade

3
Outline

• Search for Lepton Flavour Violation
• BNL E871, E865
• Standard Model and CP-Violation
• K0L?p0 ee, K0L?p0 mm KTeV-E799 II
• K0S?p0 ee, K0S?p0 mm NA48/1 CERN
• K0L?p0 nn KTeV
• K?p nn BNL-E787 / E949
• Other tests of CP Violation
• Chiral Perturbation Theory, Kaon structure
• K?p ee, K?p mm
• KS,L?gg NA48/1, KLOE
• KS,L?p0gg NA48, NA48/1
• Kaon Dalitz Decays KTeV
• Proposed/Future Experiments
• BNL/J-PARC/FNAL/CERN

Lecture I
Lecture II
4
Kaon Mesons
Particle Mass (MeV) Lifetime (ns) ct (cm) ct _at_E100 GeV
K 493.677 0.013 12.384 0.024 371 750 m
K0 497.672 0.031
K0L 51.7 0.4 1551 3.1 Km
K0S 0.08937 2.68 5.4 m
Main Decay Modes
K mn pp0 ppp- pp0p0 p0mn p0en 63 21 6 2 3 (called Km3) 5 (called Ke3)
K0S pp- p0p0 69 31
K0L p0p0p0 pp-p0 pmn pen 21 13 27 (called K0m3) 39 (called K0e3)
5
Bag of tricks for the Rare Kaon Decay Practitioner

• Identify an intense source of kaons
• Proton Accelerators
• ee- storage ring (phi factory)
• Prepare homogeneous kaons beams
• Separated beams (when possible)
• Well collimated beams (Constraint on transverse
  momentum)
• Ideal beam ? ? K0S K0L
• Construct a fast, large acceptance detector
• Design a very selective and efficient trigger
• Rejection factors of 100-1000 are not unusual
• Find an appropriate normalisation channel
• With topology as similar as possible to the
  signal one
• Protect Nature from human bias
• Blind Analysis
• Be paranoid about backgrounds
• Consider all possible sources
• Measure from data as much as you can inverting
  cuts

6
Proton Accelerators
Machine Energy (GeV) Current (mA) Rate (Hz) ppp (1013) Kaon Experiments
CERN-PS 26 1.6 0.5 2 ----
CERN-SPS 400 0.8 0.07 7 NA48
BNL-AGS 24 5 0.3 10 E871,E865,E787, E949,KOPIO
FNAL-MI 120 1.6 0.33 3 KTeV,CKM
J-PARC 50 10 0.16 40 E391
KEK-PS 12 0.16 0.25 0.4 E391a
IHEP-U70 70 0.32 0.11 2 OKA
Foreseen used the 800 GeV TeVatron beam
ee- Storage rings (? factories)
DA?NE (KLOE), VEPP II (CMD2, SND)
Planned Kaon experiment ? 2003
7
SEPARATED K BEAMS
Electrostatic separators D? (m12-m22)cEL/2p3
(Viable only up to a few GeV) Radio-Frequency
separators
DF 2p (L f / c) (b1-1 b2-1) with b1-1 b2-1
(m12-m22)/2p2
8
DFpp
14
9
Master Formula
10
Forbidden DecaysLepton Flavour Violation
11
Lepton Flavor Violation

• Puzzling replication of generations.
• Foreseen in many extensions of SM
• Generation-Changing gauge interactions
• Left-Right symmetry
• Technicolor
• Compositeness
• Super-symmetry

Neutrino oscillations not discussed
12
Processes mediated by Generation-Changing Bosons
Cahn, Harari (1980)
m-e conversion
x
m? e g
x
x
x
K?pme-
Dg1
D g 0
D g 1
x
x
m? eee
KL?me
13
AGS-871 K0L? me

• Searched for K0L? me (Lepton
  Flavour Violation)
• Measured very precisely K0L?mm-
• Observed the rarest (hadronic)
  particle decay
• BR(K0L? e
  e-) 9 6 -4 10-12
• Hottest neutral kaon beam
• 20 KW (1.5 1013 /3.2 sec, 24 GeV protons)
• 2 108 K0L per AGS spill
• 15 MHz Kaon decays
• Beam stopper after the first two tracking
  stations
• n/K ratio 8
• Benchmark for high intensity detector
• Straw chamber tracker
• Rate up to 750 KHz/wire (K. Lang et al. NIMA522
  2004)
• Double Magnetic Spectrometer (PT kicks 418 and
  216 MeV/c)
• Provides redundant momentum measurements
• Allows one to trigger on parallel tracks from
  kaon two body decays
• Muon Range Stack
• 26 V 26 H hodoscopes (prop. tubes) located
  between steel, Al, and marble absorbers (5
  increments of muon range)

14
AGS-871 K0L? me
PmeTmax 238 MeV/c
Steel, Marble, Al absorbers 5 m range increments
Redundant electron identification
15
AGS-E871 K0L? me

• Background from semi-leptonic decays K0L? p e
• with p? mn decay in flight Mmax(me)489.3MeV
• Excellent mass resolution ( 1 MeV) Required
• Scattered on vacuum window and first
• tracking station
• Irreducible background _at_10-13

16
AGS-E871 K0L? me
Excluded Region
Search Region
BR(K0L? me) lt 4.7 ? 10-12 90 CL
17
AGS E865 K? pme-

• Searched for Lepton Flavour Violation
• Measured K? pee- and K? pmm-
• Measured K? pp-e n (Ke4)
• 6 GeV/c un-separated positive beam
• 1013 p/ cycle on 10 cm long Cu target
• Double magnetic spectrometer
• S(P) 0.003 P2 GeV/c
• Redundant e PID 2 Cherenkov Shashlyk-like
  calorimeter
• Backgrounds from
• K? p p0 ? pe- e g (p0 Dalitz decay)
• K? ppp- (t decays)
• Accidentals (time-wise overlap of different kaon
  decays)

18
AGS E865

• 108 K /pulse
• 109 p/p /pulse

19
AGS E865 K? pme-

• Likelihood incorporating information from
• Vertex and track quality
• Reconstructed beam parameters
• PID and timing
• Three-track invariant mass

K? ppp-
10 prob. of finding smaller Likelihood
20
AGS E865 K? pme-
Data
MC
Monte Carlo
B(K? pme-) lt 2.8 ? 10-11 90 CL
21
Lower limits on horizontal bosons (gX/gW)1
gt150 TeV/c2
gt60 TeV/c2
gt37 TeV/c2

• Further progress on LFV expected in the muon
  sector
• m?eg PSI
• m- N ?e- N MECO_at_AGS, J-PARC?

22
Charged LFV limits versus time
1

















10-2
?- N ? e-N ? ? e? ? ? e e e-
W. Molzon
10-4
10-6
10-8
10-10
E871
10-12
K0?? ?e- K?? ? ?e-
10-14
SINDRUM2
10-16
MECO Goal ?
1940 1950 1960 1970
1980 1990 2000 2010
23
Rare K decays and Standard Model
24
CP-Violation in SM
A phase in the quark-quark current leads to
CP-Violation (Kobayashi, Maskawa, 1973)
Ng2 Nphase0 ? No CP-Violation Ng3
Nphase1 ? CP-Violation Possible
6 unitarity relations (triangles in the
complex plane)
lt VtsVtd Im lt ? 0 CP-Violation
Paradigm shift After the demonstration of the
existence of direct CP-Violation, e/e ? 0,
(NA48, KTeV) and of CP-Violation in B mixing
(BaBar, Belle) one is searching for
inconsistency in the CKM model
25
Kaon Rare Decays and the SM
(holy grail)
CP-Violation
CP-Conservation
Kaons provide quantitative tests of SM
independent from B mesons
26
K0L?mm- Motivation
Short distance contributions to KL ?mm sensitive
to Re lt
Buchalla Buras 1994
Long distance contributions

• Absorptive part (Im Agg) is dominant (2 real
  photons)
• Dispersive part (Re Agg) depends on the
• K-gg() form factors that can be studied with
• the Dalitz decays
• Is much less under control
• Can interfere with the SD piece

27
AGS-E871 K0L?mm-
6200 events

Limits on r limited by the error on Agg and gg
form factors
28
Progress on K0L? p0ee and K0L? p0mm
29
K0L?p0ee (mm) Motivation

• Study of direct CP-violation
• Direct CPV is expected to be sizeable in KL ?
  p0ll-
• But Indirect CPV and CPC Contributions have to be
  addressed

Direct CPV
Indirect CPV
CPC
0, 2
30
K0L,S?p0ee (mm) Experimental Consideration

• The recent progress on K0L,S?p0ee (mm) has been
  possible thanks to a new round of experiments
  (FNAL-KTeV and CERN-NA48) approved to measure
  Direct CP-Violation in two pion decays of the
  neutral kaons (Re e/e)
• To measure Re e/e , these experiments were
  equipped with state-of-the-art electro-magnetic
  calorimeters to reconstruct the 2p0 decays. These
  calorimeters are essential for the study of rare
  kaon decays with photons in the final state.
• Irreducible background K0L? eegg (Greenlee,
  1990)
• - The K0L? p0 ee mode is background
  limited
• Same final state as the signal, only gg mass
  resolution and kinematics are available to
  suppress these backgrounds
• To keep it to the 1 event level the
  acceptance is quite reduced
• Possible future searches will be background
  dominated

31
KTeV-E799-II (FNAL)
Pure CsI mgg1 MeV
TRD
KL flux 7 1011
32
FNAL-KTeV
33
Fermilab - KTeV
34
NA48 Detector Data Taking
1996
Total 5.3M KL??0?0
Magnetic spectrometer Liquid krypton EM
calorimeter
NA48/2 K?
2004
35
Liquid Krypton Calorimeter
9 m3 of Lkr (13212 cells) 1.25 m deph (27 X0)

s(E)/E 3.2/?E ? 9 /E ? 0.42 s(mgg)1 MeV/c2
s(t) 300 ps
36
Re e/e measurements versus time
37
Re(?/?) Results
?26.2/3
Final result (1997-2001)
Half statistics (1997)
Direct CP violation proved at gt7? levelafter
36 years!
38
NA48 Re e/e14.7 2.2 ?10-4

• Top 10 articles from Physics Letters B
• 1. The hierarchy problem and new dimensions at a
  millimeter http//dx.doi.org/10.1016/S0370-2693(98
  )00466-3 Physics Letters B, Volume 429, Issues
  3-4 , 18 June 1998, Pages 263-272 Nima
  Arkani-Hamed, Savas Dimopoulos and Gia Dvali
• 2. A precision measurement of direct CP
  violation in the decay of neutral kaons into two
  pions http//dx.doi.org/10.1016/S0370-2693(02)0247
  6-0 Physics Letters B, Volume 544, Issues 1-2 ,
  19 September 2002, Pages 97-112 J. R. Batley et
  al. (NA48 Collaboration)
• 3. Has the GZK suppression been discovered?
  http//dx.doi.org/10.1016/S0370-2693(03)00105-9
  Physics Letters B, Volume 556, Issues 1-2 , 13
  March 2003, Pages 1-6, John N. Bahcall and Eli
  Waxman
• 4. Testable scenario for relativity with minimum
  length http//dx.doi.org/10.1016/S0370-2693(01)005
  06-8 Physics Letters B, Volume 510, Issues 1-4 ,
  21 June 2001, Pages 255-263 Giovanni
  Amelino-Camelia
• 5. Role of effective interaction in nuclear
  disintegration processes http//dx.doi.org/10.1016
  /S0370-2693(03)00801-3 Physics Letters B, Volume
  566, Issues 1-2 , 24 July 2003, Pages 90-97 D. N.
  Basu
• 6. Determination of solar neutrino oscillation
  parameters using 1496 days of Super-Kamiokande-I
  data http//dx.doi.org/10.1016/S0370-2693(02)02090
  -7 Physics Letters B, Volume 539, Issues 3-4 , 18
  July 2002, Pages 179-187 S. Fukuda et al.
• .

39
KTeV K0L? p0ee
WC
CsI
Charged Vertex
e
g
d12
g
e
D
Neutral Vertex D 1/m(p0) sqrt(Eg1Eg2) d12
40
KTeV KL? p0ee
KL?p0p0p0D, penp0acc
KL?eegg background crossing the signal box
(BR6 10-7 !!)
DATA 1999
M(eegg) computed using the neutral vertex,
assuming m(gg)m(p0) M(gg) computed from
charged vertex
41
KTeV p0ee vs eegg kinematics
KL ?eegg radiative distribution
Minimum angle between any e and g
p0 ?gg decay asymmetry
p0 ?gg isotropic distribution
Expected background 0.99 0.33 events
cosqp0
42
KTeV KL? p0ee
1999 data

• One candidate in the signal box
• Combining 1997 and 1999

BR(KL ? p0 ee ) lt 3.5 10-10 _at_90CL
BR(KL ? p0 ee ) lt 2.8 10-10 _at_90CL
43
KTeV K0L ? mmgg

4 events seen
First observation
PRD 62, 112001 (2000)
44
KTeV K0L ? p0mm
2 events in signal region

Data
Background MC
45
Interpretation of KL? p0ee(mm)

• Before interpreting the results one has to
  address the
• CP-Conserving amplitude (see later)
• CP-Violating from K0-K0bar mixing
• This was exactly the goal of NA48/1 at CERN
• The results will be described in some detail

46
First Observation of K0s? p0ee and K0s? p0mm
47
CERN-NA48/1 High Intensity K0S
Converges to KLaxis at -0.6 mrad
48
NA48/1 K0S ?p0 ee
Example of calculation of charged vertex for
K0S ? p0 p0D

• The charged vertex is used to compute M(eegg)
• The neutral vertex is computed imposing the Kaon
  mass

49
NA48/1 KS ? p0 ee
KS ? p0 p0D?ggee(g)
KS ?p0 ee
MC
MC

• To reject the KS ? p0 p0D decays that may mimic
  KS ?p0 ee if a g is lost, a cut meegt0.165 GeV/c2
  is applied

50
NA48/1 KS ?p0 ee
e-e- (Odd Sign) DATA
e-e- DATA vs. MC
Blind Control Signal regions
meegg (GeV/c2)
mee (GeV/c2)
51
NA48/1 KS ?p0 ee
e-e- (Same Sign) DATA
e-e- DATA vs. MC
Search region
meegg (GeV/c2)
mee (GeV/c2)
52
NA48/1 KS ?p0 ee
Background from KL,S?eegg measured using NA48
KL data from 2001 N(KL?eegg, 2001) ? 10 N(KL,S
?eegg, 2002)
meegg(GeV/c2)
mgg(GeV/c2)
53
NA48/1 KS ?p0 ee
Accidental backgrounds
meegg (GeV/c2)

• DC proton beam
• Read out window 200ns
• Use time side band to measure background from
  time-overlapping fragments from different decays
• Major component
• epn p0p0(p0)
• Confirmed relaxing E/P cuts

mgg (GeV/c2)
54
NA48/1 KS ?p0 ee
SUMMARY OF BACKGROUNDS
Source Control Region Signal region
KS?p0Dp0D 0.03 0.007
KL,S ? eegg 0.11 0.075
pen2p0(p0) 0.19 0.069
Total 0.330.18-0.11 0.150.05-0.04

• Many other sources investigated and found to be
  negligible (e,g neutral cascade decays)
• Blind analysis Control and signal region
  remained masked until the study of the background
  was finished

55
NA48/1 KS ?p0 ee

• 7 candidates in the
• signal region
• 0 in control region
• Background 0.15
• The probability that all 7
• events are background
• is 10-10

meegg (GeV/c2)
mgg (GeV/c2)
First observation of KS ?p0 ee
56
The 7 KS?p0ee candidates
Meegg (GeV/c2)
t/tS
Mgg (GeV/c2)
Mee (GeV/c2)
57
NA48/1 KS ?p0 ee
PL B576 (2003) hep-ex/0309075
BR(KS?p0ee, meegt165 MeV/c2) (3.01.5-1.2(stat)
0.2(syst))10-9

• Assuming vector interaction and unity form
  factor
• In remarkable agreement with L. Sehgal
  prediction 5.5 10-9 NP B19 (1970)

BR(KS?p0ee) (5.8 2.8-2.3(stat)
0.8(syst))10-9
58
Search for KS ?p0 mm

• The considerations made before for
• KL ?p0 ee apply also to KL ?p0 mm but the
  CP-conserving contributions need more attention
• A measurement of KS ?p0 mm is quite
  complementary to the KS ?p0 ee one
• Different backgrounds
• Larger acceptance (no Dalitz background)
• In principle one can relate KS ?p0 mm and KS
  ?p0 ee to extract the form factor (for example
  the aS and bS parameters)

59
NA48/1 KS ?p0 mm

• Study of backgrounds from KL? p0pp- ? p0
  (mn)(mn)
• MC 22 times the data
• None of the simulated events falls in the signal
  box

60
NA48/1 KS ?p0 mm

• Backgrounds from Neutral Cascade decays are
  rejected using a Momentum ratio cut
• Radiative (mmgg) backgrounds
• 0.04 0.04 events

61
NA48/1 KS ?p0 mm
Study of accidental
backgrounds

• KL ?p m n KS ?p0(p0)
• KS?pp- KS ?p0(p0)
• Studied in time side band
• 6 events in 125 ns
• 0.18 0.18-0.11 background events expected in the
  in time signal region

62
NA48/1 KS ?p0 mm

• BR(KS?p0mm) ? 109 2.9 1.4-1.2(stat)
  0.2(syst) La Thuile 2004

First Observation! 6 events Expected
back. 0.22019-0.12
63
Dilepton invariant mass distributions
NA48/1
NA48/1
64
Extraction of form factor

• VMD bs/as 0.4
• Data compatible with VMD
• Too low statistics to
• extract the two parameters

• Assuming VMD
• BR(KS ?p0 ee) 5.2 as2 ? 10-9
  aspee 1.06 0.26 -0.21
  0.07
• BR(KS ?p0 mm) 1.2 as2 ? 10-9
  aspmm 1.55 0.38 -0.32
  0.05

65
K0L,S?p0ee (mm) Phenomenology
66
K0L?p0ee (mm) CP-ConservingContributions
Isidori, Unterdorfer, Smith, hep-ph/0404127
estimate Scalar 0 Using
Tensor 2
G.Buchalla, G.DAmbrosio, G.Isidori,
Nucl.Phys.B672387 (2003)
67
Interference between K0L?p0ee and K0s?p0ee

• Two independent theoretical analyses find that
  the
• interference term is constructive
• Buchalla,Isidori,DAmbrosio hep-ph/0308008, NP
  B 672 (2003)
• Friot, Greynat, de Rafael hep-ph/0404136

68
Predictions
Results Experiments




Friot, Greynat, de Rafael
69
K0L?p0ee (mm) SM Branching Ratios
(Isidori, Unterdorfer, Smith)
Constructivenow favored by two independent
analyses
Destructive

G. Buchalla, G. DAmbrosio, G. Isidori,
Nucl.Phys.B672,387 (2003) S. Friot, D. Greynat,
E. de Rafael, hep-ph/0404136
70
K0L?p0ee (mm) Sensitivity to New Physics
Isidori, Unterdorfer,Smith
Buras et al Ratios of B ? Kp modes could be
explained by enhanced electroweak penguin and
enhance the BRs
A. J. Buras, R. Fleischer, S. Recksiegel, F.
Schwab, hep-ph/0402112
71
K0L?p0ee(mm) Perspectives

• Detector s(gg) 2
• Very ambitious, KTeV/NA48 already state of the
  art
• KS-KL time dependent interference 2
• Position experiment between 9 and 16 KS lifetimes
  
• (hep-ph/0107046)
• KS-KL time independent interference 3
• Assume constructive interference (theoretically
  preferred)
• Data Taking 5
• Run in factory mode. After all E799-II run
  only for a few months to collect 7 1011 KL
  decays
• Beam intensity 4
• Need 1012 protons/sec, slowly extracted, high
  energy, DC
• Tot 240 ? sens on BR 15 (on Im lt
  4-15)
• close the gap between current upper limit and SM
• NA48/KTeV resolutions with E871 beam intensities
  !!

72
Towards the holy grail K0L? p0nn and K? pnn
73
K0L ? p0 n n

• Purely theoretical error 2
• Purely CP-Violating (Littenberg, 1989)
• Totally dominated from t-quark
• Computed to NLO in QCD ( Buchalla, Buras, 1999)
• No long distance contribution SM3 10-11

• Experimentally 2/3 invisible final state !!
• Best limit from KTeV using p0?eeg decay

BR(K0 ? p0nn) lt 5.9 10-7 90 CL
Still far from the model independent limit
BR(K0 ? p0nn) lt 4.4 BR(K ? pnn) 1.4 10-9
Grossman Nir, PL B407 (1997)
74
KTeV K0L ? p0nn
1997 Data Dalitz Analysis
75
K?p nn Theory

• The hadronic matrix element can be extracted from
  
• the well measured K? p0 e n
• No long distance contributions

QCD NLO Buchalla, Buras 1999
Prediction (CKM Workshop)
BR(K ? pnn) 8.0 1.1 10-11

76
K?p nn Experiment
Region I 211 lt P lt 229 MeV/c
Region II 150 lt P lt 195 MeV/c
p
in the K CM
77
AGS-E787
78
AGS-E787/E949

• 500-700 MeV/c Kaons from a separated beam come to
  rest in a active target
• To avoid backgrounds from p scattering or decays
  in flight, one waits gt2ns for the kaon to decay
• Redundancy the outcoming p is measured by
• Magnetic spectrometer (Momentum)
• Range (cm)
• Kinetic energy (In the range stack)
• The decay chain p ? mn , m ? enn is followed with
  transient recorders
• Biforcated analysis the rejection power is
  measured on data samples selected inverting cut

79
AGS-E787/E949
80
AGS-E787
81
AGS-E787/E949 K?p nn
hep-ex/0403036
BR(K ? p nn ) 1.471.30-0.89 10-10

82
AGS E787/E949
Projection
83
Current Status
G. Isidori
STILL A LARGE WINDOW OF OPPORTUNITY EXISTS