Recent CDF Results on Bs Flavor Oscillations

1
Recent CDF Results on Bs Flavor Oscillations

• The CDF Bs Mixing Group
• CDF Executive Board Meeting
• 13 April 2006

J. Kroll (Penn), F. Bedeschi(INFN Pisa), M.
Herndon (Wisconsin), Ch. Paus (MIT), K. Pitts
(Illinois)
2
The CDF Bs Mixing Group
80 people, 27 Institutions Big effort with many
separate important contributions
3
Made Possible By the CDF Detector

• Trigger
• high bandwidth tremendous flexibility
• XFT and SVT
• Exquisite charged particle tracking
• excellent pattern recognition mass resolution
  (COT, solenoid)
• precise production decay point reconstruction
  (L00, SVX)
• Particle identification
• Time of Flight (TOF) and dE/dx in COT
• Extremely dedicated collaborators for operations
• a lot of sleepless nights for analysis but many
  more for operations
• many of the analyzers spent sleepless nights
  working on critical detector components too

4
Sequence of CDF Run II Results on Bs Mixing

• Fall 2001 1st heavy flavor signal from TTT data
• Summer 2002 1st lifetime measurements
• Winter 2003 1st fully reconstructed Bs! Ds?
• Summer 2003 1st blessed opposite side flavor tag
  (muon)
• Summer 2004 1st B0 ?md result

A lot of ideas and development in Run I
5
Sequence of CDF Run II Results on Bs Mixing

• 1st result March 2005 presented at Moriond QCD
• hadronic modes from two-track trigger
• semileptonic modes from lepton SVT trigger
• opposite-side flavor tags muons, electrons,
  jet-charge
• combined sensitivity ?ms gt 8.4 ps-1
• 1st update October 2005 presented at PANIC
  2005
• substantial increase in semileptonic signal from
  TTT
• several analysis improvements, e.g.,
• neural net based jet-charge opposite side flavor
  tag
• improved opposite side flavor tag calibration
• improved boost resolution on semileptonic decays
• combined sensitivity ? ms gt 13.0 ps-1

These results based on 360 pb-1 (0d data)
6
Current CDF Bs Oscillations Analysis

• This winter our focus shifted to three main areas
• adding same-side kaon tagging (SSKT)
• blessed on 23 February 2006
• adding 0h and 0i data (an additional 640pb-1)
  (done see later)
• using partially reconstructed hadronic decays
  (still in progress)
• 1st step add SSKT to 360pb-1 hadronic analysis
• Schedule had us unblinding on 13 March 2006
• DØ announcement at Moriond EWK on 12 March 2006
• CDF unblinded hadronic scan on 13 March 2006 as
  foreseen
• found 3? excess _at_ ? ms 17.3 ps-1
• 0d hadronic-only result already much more
  significant than DØ
• preblessed at B meeting on 16 March 2006
• decided to add 0h 0i data asap for both
  hadronics semileptonics

7
DØ Result
Presented at Moriond EWK on 3/12/06
Amplitude peaks _at_ ? ms 19ps-1
2.5? discrepant with A0 (1)
10 compatible with A1
17 lt ? ms lt 21 ps-1 _at_ 90 CL
Probability that random fluctuation produces this
signal p-value (5.0 0.3)
8
CDF Hadronic Data (355pb-1)
We were on schedule to unblind 1st part of
hadronic data on 3/13/06
caution arbitrary offset
90
This result was comparable in sensitivity to DØ
result
We had a lot more data, so we decided to add it,
and fast
9
Rapid Progress from 3/13 to 3/31

• This was a well-tuned analysis extensively
  reviewed
• Winter 2005 had two independent semileptonic
  analyses
• included independent fits and data sets, which
  were fed to each other
• direct comparison of amplitude scans
• also independent lifetime measurement in Bs! Ds?
• Emphasis on adding new data was validation
• signal invariant mass distributions
• signal lifetime distributions
• particle identification (TOF and dE/dx)
• proper time resolution
• Had to understand how to move from limit to
  measurement
• required evaluation of probability of background
  fluctuation
• determination of ?ms including systematic
  uncertainties
• Unblinded amplitude scans after B meeting on
  3/30/06
• presented at Joint Physics Meeting on 3/31/2006

10
Result Amplitude Scans in 1 fb-1
Hadronic Decays
Semileptonic Decays
Single experiment sensitivity gt combination of
previous results
11
Result -ln(likelihood) ?ms, Vtd/Vts
A 1.01 0.29 (3.5?)
p value 0.5
Limited by theoretical (lattice) uncertainties
12
Impact on CKM Unitarity Triangle
EPS05
EPS05 CDF
Figures courtesy of Z. Ligeti (LBNL)
13
Since Blessing on 6 April 2006

• Special Wine Cheese on Monday 10 April 2006
• 3 rehersals Friday Sunday afternoon, Monday
  morning
• 1 West was standing room only
• spectacular performance by speaker Ivan Furic
• FNAL press release on Tuesday 11 April 2006
• 1st Conference presentation at FPCP 2006,
  Vancouver, BC
• CDF contingent flew Monday night
• Rehersal in hotel room Tuesday morning
• Excellent presentation by Guillelmo Gomez
  Ceballos
• Received with great enthusiasm by Flavor
  community
• Several requests for seminars from National Labs
• ANL, BNL, DESY, LBNL, SLAC

14
Public Documentation

• PRL in preparation
• 1st draft Ch. Paus, JK
• Godparents Ristori (Chair), Glenzinski,
  Wuerthwein
• Heinrich, Punzi are statistics committee
  consultants
• plan for accelerated review process
• Public Document (CDFNOTE 8204)
• K. Pitts editor, many contributors
• circulating in collaboration now please read
  and send comments
• would hope to make public by next week
• could serve as PRD draft
• Web page
• link on main CDF public page to main plots
  talks
• http//www-cdf.fnal.gov/physics/new/bottom/060406.
  blessed-Bsmix/

15
Whats Next?

• Analysis improvements (all already well underway)
• add satellite peaks (Ds?, Ds?) in hadronic modes
• neural networks for hadronic signal selection
• use all semileptonic trigger paths (lepton SVT,
  SUSY dilepton)
• neural networks for opposite side tags
• improve SSKT by combining kinematics and particle
  id
• use particle id in signal selection (Ds! KK)
• Technical improvements
• reprocess all data offline version 6
• at least 1st 360 pb-1
• improvements to fitter implementation
• e.g., replace ascii input file with ntuple
• Related physics results (e.g., lifetimes)

Goal 5? observation for ICHEP 2006