1. KTeV Results on Chiral Perturbation Theory
Patricia McBride, for the KTeV Collaboration
Fermilab
Joint Meeting of Pacific Region Particle Physics Communities
Hawaii, October 2006

2. Patricia McBride, KTeV Collaboration
Outline
The KTeV experiment
Probes of ChPT at O(p6)
KL0
KL0e+e-g
KL00
Summary
Other KTeV results at this meeting: See talks by Bob Tschirhart, Ron Ray and Erin Abouzaid.
Hawaii, October 2006
Patricia McBride, KTeV Collaboration

3. The KTeV Detector (E799 configuration)
Pure CsI Calorimeter:(Energy resolution  1% at Eg  = 10GeV; (p/e rejection of  700)
Four drift chambers: resolutions: ~100m
Transition radiation detectors: (p/e rejection of  200) [E799]
Intense beams: 5 x 1012 protons on target per spill  5 x 109 kaons/spill
For EK ~ 70 GeV: KS: c ~ 3.5m
KL: c ~ 2.2 km
Hawaii, October 2006
Patricia McBride, KTeV Collaboration

4. Patricia McBride, KTeV Collaboration
The KTeV Experiment
The KTeV Collaboration:
Arizona, Campinas (Brazil), Chicago, Colorado, Elmhurst, Fermilab, Osaka (Japan), Rice, Rutgers, San Paulo (Brazil), UCLA, UCSD, Virginia, Wisconsin
Two KTeV Goals:
E832: Measure 
E799: Investigate rare decays of the kaon
Hawaii, October 2006
Patricia McBride, KTeV Collaboration

5. Patricia McBride, KTeV Collaboration
KL ® 0g    
Chiral perturbation calculations of the branching ratio
O(p4) calculations predict BR(KL0= 0.6 x 10 -6
Earlier measurements have shown that this prediction is low by a factor of 2-3
O(p6) calculations increase this rate.
The variable av parameterizes the contribution of vector meson exchange to the decay amplitude.
Results can be used to predict the relative contributions of the CP conserving and direct CP violating components in the decay KL0e+e-
Hawaii, October 2006
Patricia McBride, KTeV Collaboration

6. KL ® 0g event selection
signal topology: KL  0
Require 4 electromagnetic clusters in the calorimeter, and
no charged tracks.
Select best 0 mass combination from the 4 photon candidates.
The background comes from KLp0p0 and KLp0p0p0. The largest and most troublesome background originates from KLp0p0p0 decays with missing/merged photon clusters.
The normalization mode is KLp0p0.
KTeV E832 configuration
Hawaii, October 2006
Patricia McBride, KTeV Collaboration

7. Patricia McBride, KTeV Collaboration
KL ® 0g backgrounds
The major background to 0 comes 30 decays where only 4 clusters are detected. These events form a vertex downstream of the true KL vertex.
KL decay vertex region
Signal events
background
Z vertex position
Hawaii, October 2006
Patricia McBride, KTeV Collaboration

8. KL ® 0g background in published result
We remove overlapping showers with a photon shape variable.
Disagreement between data and 30 MC at low values of photon shape variable - the region where there should be no overlapping photons.
Result: underestimate of background from 30 events.
Photon Shape Variable
default library
0 mass for candidate events with low photon shape variable
Hawaii, October 2006
Patricia McBride, KTeV Collaboration

9. KL ® 0gestimating the background
For this analysis we improved the simulation to get a better match for the photon shape variable.
The background levels increased.
Better agreement in m tails.
Hawaii, October 2006
Patricia McBride, KTeV Collaboration

10. Final results: KL ® 0gdecays
From the full combined data sample, 1982 KL0 candidate events are observed.
The background is estimated to be 601 events coming from primarily from KLp0p0p0 events.
m
The normalization mode is KLp0p0.
m
BR(KL0) =  0.03(stat) X 10-6
Preliminary result
Hawaii, October 2006
Patricia McBride, KTeV Collaboration

11. KL ® 0g: systematic error on BR
Source of Uncertainty
Uncertainty
av dependence
1.5 %
30 background
1.3 %
MC statistics
1.0 %
Normalization
0.9 %
Photon Shape
1.1 %
Tracking Chamber
20 branching ratio
Photon vetoes
Kaon Energy
0.7 %
Decay Vertex
0.4 %
Total
2.9 %
0 branching ratio
BR vs Run period
Preliminary
BR(KL0) = ( 1.30  0.03(stat)  0.04(sys) ) X 10-6
Hawaii, October 2006
Patricia McBride, KTeV Collaboration

12. Updated BR for KL ® 0gdecays
BR(KL0) = ( 1.30  0.03(stat)  0.04(sys) ) X 10-6
KTeV Preliminary
0 branching ratio
This measurement is consistent with chiral perturbation theory with the addition of vector meson exchange through the parameter av.
The extraction of av should be available soon.
Previous results:
KTeV: PRL 83, 917 (1999)
NA48: PL B536, 229 (2002)
Hawaii, October 2006
Patricia McBride, KTeV Collaboration

13. A related decay: KL ® 0e+e-g
The rare decay KL0ee- is related to the decay KL0via the internal conversion of one of the photons.
KL0ee-provides another handle for untangling the CP conserving amplitude in the decay KL0ee-.
The rate for KL0ee-is expected to be several orders of magnitude higher than KL0ee-.
First observation of this decay was made by KTeV and published in PRL. A.Alavi-Harati et al.,PRL 87, (2001)
This new analysis uses the full KTeV (E799) data set.
Hawaii, October 2006
Patricia McBride, KTeV Collaboration

14. KL ® 0e+e-g event selection
Require 2 charged tracks and 5 clusters in the calorimeter.
Require each charged track to point to a cluster and form a good vertex.
Select best 0 mass combination from photon candidates.
signal topology: KL  0e+e- g
Hawaii, October 2006
Patricia McBride, KTeV Collaboration

15. Background: KL ® 00, 0 ® e+e-g
Dalitz decay
There are 3 ways to combine the photons to form a 0 candidate.
Most KLp0p0D events can be removed through a p0 mass cut on the best e+e- combination.
The remaining 20 background from wrong 0 combinations is removed using a neural network analysis. (m0 and m e+e-for the 2nd and 3rd best combinations.)
The NN variable for 2πo and πoe+e−γ MC events.
Events for which the best e+e−γ combination has the πo mass have been removed.
KLp0p0D
KL0e+e-g
Neural Net Variable (NN)
Hawaii, October 2006
Patricia McBride, KTeV Collaboration

16. Backgrounds from 30 Dalitz decays
These events have missing photons.
We require that each photon candidate has a small value for the shape variable to eliminate events with overlapping photons.
cut
30 background
KLp0p0p0 , p0e+e-
Signal MC
KL0e+e-g
Missing momentum versus the m for 30 events
Hawaii, October 2006
Patricia McBride, KTeV Collaboration

17. Patricia McBride, KTeV Collaboration
KL ® 0e+e-g results
From the combined data sample, 139 KL0e+e-g candidate events are observed.
The background is estimated to be 14.4 events coming from KLp0p0D and KLp0p0p0D.
The normalization mode is KLp0p0D. (e+e-g)
m e+e-
BR(KL0ee) = ( 1.90  0.16(stat) ) X 10-8
preliminary
Hawaii, October 2006
Patricia McBride, KTeV Collaboration

18. KL ® 0e+e-: preliminary result for BR
Source of Uncertainty
Uncertainty
MC statistics
4.2 %
av dependence
3.8 %
KL and 0 BR
2.8 %
3 0 background
0.8 %
acceptance
0.4 %
20 background
0.1 %
Total
6.4 %
KTeV PRL 99 97
0e+e- branching ratio
BR(KL0e+e-) = ( 1.90  0.16(stat)  0.12(sys) ) X 10-8
Preliminary result
Hawaii, October 2006
Patricia McBride, KTeV Collaboration

19. Patricia McBride, KTeV Collaboration
KL ® 00g
Select events with a Dalitz decay of one of the 0.
Choose the best 0 mass combinations.
Main background is from KLp0p0p0D.
The normalization is made with KLp0p0p0D decays that are not fully reconstructed. One photon passes through one of the beam holes in the calorimeter.
signal topology: KL  0 0 g
Hawaii, October 2006
Patricia McBride, KTeV Collaboration

20. Patricia McBride, KTeV Collaboration
KL ® 00gMonte Carlo
97 MC
99 MC
pt2 mK mK
Signal MC
Background MC:
Generated KLp0p0p0D MC with statistics 4x data
pt2 mK
Hawaii, October 2006
Patricia McBride, KTeV Collaboration

21. Patricia McBride, KTeV Collaboration
KL ® 00g Results
99 data
97 data
Blind analysis: One event seen 99 data, none in 97
Probability that event comes from background is 10%
Limit: BR(KL00g ) < 2.32 X KTeV prelimiinary
Hawaii, October 2006
Patricia McBride, KTeV Collaboration

22. KTeV tests of Chiral Perturbation Theory
The measured branching ratio results for KL0 and KL0e+e- are inconsistent with O(p4) predictions of chiral perturbation theory, but are consistent with O(p6) calculations.
Decay Mode Publ # Events Branching Ratio
KL  p0gg Prel (1.30  0.03  0.04) x 10-6
KL  p0e+e-g Prel (1.90  0.16  0.12) x 10-8
KL  p0p0g Prel < 2.32 x 10-7
Hawaii, October 2006
Patricia McBride, KTeV Collaboration

23. Summary and Conclusions
The branching ratio results are preliminary, but represent results from the full KTeV data sample.
Fits for av from the combined data sets should be available soon.
KTeV results in these channels agree with O(p6) calculations in chiral perturbation theory that include contributions from vector meson exchange through the parameter av.
Hawaii, October 2006
Patricia McBride, KTeV Collaboration

24. Patricia McBride, KTeV Collaboration
Additional slides
Hawaii, October 2006
Patricia McBride, KTeV Collaboration

25. Patricia McBride, KTeV Collaboration
Search for KL ® p0e+e-
KL>  KODD> + KEVEN>
0*
0e+e-
Indirect
CP Violation
0Z*
0 W+* W-*
Direct
VtdVts* Vtd*Vts
 (VtdVts*)  A25
0**
CP conserving
Helicity suppressed
The decay KL ® p0e+e- decay is of interest because it is expected to have a large CP-violating component.
The Standard Model prediction for the
Branching Ratio is (3-10 x 10 –12).
The major background is from the
radiative Dalitz decay of the kaon
KL  eegg.
Hawaii, October 2006
Patricia McBride, KTeV Collaboration

26. Backgrounds from 00 Dalitz decays
KLp0p0 , p0e+e-
KL0e+e-g
Before NN
After NN
m0
m e+e-
2nd best combination
Signal
20 background
Hawaii, October 2006
Patricia McBride, KTeV Collaboration