Rare Decays at Tevatron

Rare Decays at TevatronRare Decays at Tevatron

Marco RescignoMarco RescignoINFN/RomaINFN/Roma

for the CDF and DØ for the CDF and DØ coll.coll.

CKM06, Nagoya CKM06, Nagoya Dec 14Dec 14thth 2006 2006

Topics: 1)BTopics: 1)Bsearch at search at TevatronTevatron

2) b2) bslsl++ll- - at Tevatronat Tevatron

M.Rescigno - Nagoya12/06M.Rescigno - Nagoya12/06 22

Main Injector & Recycler

Tevatron

Chicago

p source

Booster

p

p

p p

1.96 TeV

CDF

DØ

TevatronTevatron


Tevatron LuminosityTevatron Luminosity

>2 fb>2 fb-1-1 delivered!delivered!

1.5 fb1.5 fb-1-1 good good data on tape data on tape nownow

Analysis Analysis shown today shown today 0.7 – 1 fb0.7 – 1 fb-1-1


DO

CDF

CDF & DØ CDF & DØ detectorsdetectors Good muon coverage and Good muon coverage and

triggeringtriggering DØ: |DØ: ||<2.2|<2.2 CDF: |CDF: ||<1|<1

Good momentum Good momentum resolution, tracking:resolution, tracking: CDF:CDF:(M(MBB)~25 MeV/c2 )~25 MeV/c2 reduce combinatorial and reduce combinatorial and

BBddhh’ hh’ B Bss search window search window contaminationcontamination

Good Vertexing Good Vertexing CDF: L00 CDF: L00 (r(rinnerinner~1.4cm)~1.4cm) DØ: L0 upgrade DØ: L0 upgrade

(r(rinnerinner~1.6cm)~1.6cm)


Triggers and Data SampleTriggers and Data Sample CDF:CDF:

di-muon triggered datadi-muon triggered data Two separate search channelsTwo separate search channels

Central/central muons Central/central muons (CMU-CMU)(CMU-CMU) Central/forward muons Central/forward muons (CMU-CMX)(CMU-CMX)

CMU |CMU ||<0.6, CMX 0.6 < ||<0.6, CMX 0.6 < || <1| <1 ppTT(()>1.5 GeV/c)>1.5 GeV/c

780 pb780 pb-1-1:: B Bss(B(Bd d ))limitlimit 920 pb920 pb-1-1:: Search for B Search for Bu,d,su,d,s hh

DØ:DØ: First 300 pbFirst 300 pb-1-1 di-muon triggered di-muon triggered

datadata with box opened with box opened limit limit 400 pb400 pb-1-1 data still blinded data still blinded Combined sensitivity for 700 pbCombined sensitivity for 700 pb -1-1

of recorded data (300 pbof recorded data (300 pb-1-1 + 400 + 400 pbpb-1-1 ) )

S/B is expected to beextremely small. Effectivebkg rejection is the keyto this analysis!!

B search regionBh search region


Trigger @ high lumi?Trigger @ high lumi? Clever and Clever Clever and Clever

triggering as inst. triggering as inst. Luminosity increasesLuminosity increases

Keep high purity triggers Keep high purity triggers alive at high luminosityalive at high luminosity

Full use of available Full use of available bandwidth at lower inst. bandwidth at lower inst. Luminosity with Luminosity with dynamically adjusted dynamically adjusted prescalesprescales

Hardware upgrade on the Hardware upgrade on the level 1 track trigger level 1 track trigger processor processor reduce fake rate by reduce fake rate by

reconstructing track reconstructing track segments also in the stereo segments also in the stereo layerlayer

~540 Hz @ 200 E30

~320 Hz @ 200 E30

BB search at Tevatron search at Tevatron


MotivationMotivation Standard Model prediction very suppressedStandard Model prediction very suppressed

Sizeable New Physics enhancement predicted in Sizeable New Physics enhancement predicted in many scenarios, e.g. high tanmany scenarios, e.g. high tan SUSY: SUSY:

910)9.05.3()( sBBR (Buchalla & (Buchalla & Buras, Buras, Misiak & Urban)Misiak & Urban)

Any signal @ Tevatron Any signal @ Tevatron would be New Physics !would be New Physics !


StrategyStrategy Blind optimization using signal Monte Carlo Blind optimization using signal Monte Carlo

sample and sideband datasample and sideband data Normalize to known BNormalize to known B++J/J/KK++

Reconstruct Normalization mode in the Reconstruct Normalization mode in the same data, applying same criteria same data, applying same criteria reduce systematicsreduce systematics

Only ratio of efficiency mattersOnly ratio of efficiency matters Evaluate expected background and then Evaluate expected background and then

open the box and calculate BR or limitopen the box and calculate BR or limit

)/()/()(

JBRKJBBR

f

f

N

NBBR

Bsb

BbtotalBsBs

totalBB

B

Bss


Preselection CutsPreselection Cuts CDF:CDF: ppTT(()>2.0 (2.2) GeV/c CMU (CMX))>2.0 (2.2) GeV/c CMU (CMX) ppTT(B(Bss cand.)>4.0 GeV/c cand.)>4.0 GeV/c |y(B|y(Bss)| < 1)| < 1 4.669 < m4.669 < m< 5.969 GeV/c< 5.969 GeV/c22

muon quality cutsmuon quality cuts good vertexgood vertex 3D displacement L3D displacement L3D3D between between

primary and secondary vertexprimary and secondary vertex (L(L3D3D)<150 )<150 mm proper decay length 0 < proper decay length 0 < < 0.3cm < 0.3cm Pre-selection DØ:Pre-selection DØ:

ppTT(()>2.5 GeV/c)>2.5 GeV/c (()| < 2)| < 2 ppTT(B(Bss cand)>5.0 GeV/c cand)>5.0 GeV/c 4.5 < m4.5 < m< 7.0 GeV/c< 7.0 GeV/c22

muon quality cutsmuon quality cuts good di-muon vertexgood di-muon vertex

38k events after pre-selection

300 pb300 pb--

11


++-- mass mass ~±2.5~±2.5 mass window (60 MeV/c mass window (60 MeV/c22))

B vertex displacement:B vertex displacement: CDF CDF

D0 D0

Isolation Isolation (Iso)(Iso)::

(fraction of B(fraction of B p pTT within within R=(R=(22++22))1/21/2=1 =1 cone)cone)

““pointing pointing (())”:”:

(angle between B(angle between Bss momentum and decay axis) momentum and decay axis)

P

P P

L3D

x

y

BBss BBss R < 1 (< 57o)

z

)(3

Bp

McL D

i iiTT

T

RpBp

BpIso

)1()(

)(

))(( 3DLBp

LxyxyL /

BB signal discrimination signal discrimination


Search Optimization @ CDFSearch Optimization @ CDF CDF construct a CDF construct a

lilkelihood ratio Llilkelihood ratio LRR using using Iso Iso

ii

iR

xPxPxP

ibis

isL

)()(

)(

Optimize LR cut on the Optimize LR cut on the expected expected a-prioria-priori 90% 90% C.L. limitC.L. limit

Background PDF from Background PDF from data sidebands: data sidebands: 4.669<M4.669<M<5.169 GeV/c<5.169 GeV/c2 2

U 5.469<MU 5.469<M<5.969<5.969


Search Optimization @ CDFSearch Optimization @ CDF

Optimal value LOptimal value LRR > 0.99 > 0.99 Signal efficiency Signal efficiency BBss))

35% 35% Background Rejection Background Rejection

O(10O(1033))

LR distributions in LR distributions in signal and sidebands signal and sidebands matchmatch

A posteriori check...A posteriori check...


Optimize cuts on three Optimize cuts on three discriminating variables:discriminating variables: Pointing anglePointing angle 2D decay length significance2D decay length significance IsolationIsolation

Maximize S/(1+sqrt(B))Maximize S/(1+sqrt(B)) B from sidebandsB from sidebands

Search optimization @ DØSearch optimization @ DØ

Similar efficiency and Similar efficiency and background rejection background rejection

BBss)) 35% 35%


Background predictionBackground prediction Combinatorics:Combinatorics:

linear extrapolation linear extrapolation from sidebands into from sidebands into ±60 MeV/c±60 MeV/c22 signal signal window for CDF (± window for CDF (± 180 @ D0)180 @ D0)

BBhh’:hh’: expected signal expected signal convoluted with muon convoluted with muon fake rate (CDF)fake rate (CDF)

Cross check predictions using Cross check predictions using independent background independent background enriched samplesenriched samples Same Sign di-muonsSame Sign di-muons Opposite Sign di-muons with Opposite Sign di-muons with

LLxyxy<0<0 Fake muonsFake muons


Examining Signal BoxesExamining Signal Boxes

ExperimentExperiment BB00ss search search BB00

dd search search

ExpectedExpected Obs.Obs. ExpectedExpected Obs.Obs.

CDFCDF

780 780 pbpb-1-1

CMU-CMUCMU-CMU 0.88±0.30.88±0.300

11 1.86±0.341.86±0.34 22

CMU-CMXCMU-CMX 0.39±0.20.39±0.211

00 0.59±0.210.59±0.21 00

D0D0Old DataOld Data 4.3±1.24.3±1.2 44

New DataNew Data 2.2±0.72.2±0.7 --


Normalization: BNormalization: B++J/J/KK++

Need BNeed B++J/J/KK+ + yield to extract limitsyield to extract limits CDF (780 pbCDF (780 pb-1-1) : 4200 (CMU-CMU) + 1550 ) : 4200 (CMU-CMU) + 1550

(CMU-CMX)(CMU-CMX) D0 (400 pbD0 (400 pb-1-1): 900 ): 900


ResultsResults No signal found…No signal found…

CDF BCDF Bss limit (780 pb limit (780 pb-1-1)) BR(BBR(Bss) < 8 ∙ 10) < 8 ∙ 10-8-8 (10) @ 90% (95%)C.L. (10) @ 90% (95%)C.L.

DØ average expected limit (700 pbDØ average expected limit (700 pb-1-1)) BR(BBR(Bss) < 19 ∙ 10) < 19 ∙ 10-8-8 (23) @ 90% (95%)C.L. (23) @ 90% (95%)C.L.

CDF BCDF Bdd limit (780 pb limit (780 pb-1-1), world best), world best BR(BBR(Bdd) < 2.3 ∙ 10) < 2.3 ∙ 10-8-8 (3) @ 90% (95%) C.L. (3) @ 90% (95%) C.L. compare Babar (hep-ex/0408096, 110 fbcompare Babar (hep-ex/0408096, 110 fb-1 -1 ))

BR(BBR(Bdd) < 8.3 ∙ 10) < 8.3 ∙ 10-8-8 @ 90% C.L. @ 90% C.L.


Impact on New PhysicsImpact on New Physics Current Current

constraints from constraints from mmss and B and Bss are differently are differently effective in the effective in the new physics new physics parameter parameter phase space phase space

Improved limits Improved limits on Bon Bss can can further further constraint SUSY constraint SUSY at large tanat large tan Foster,Okumura,RoszkowsFoster,Okumura,Roszkows

ki Phys.Lett. B641 (2006) ki Phys.Lett. B641 (2006) 452 452


Tevatron Expected ReachTevatron Expected Reach

Based on current Based on current analysis analysis might be might be

conservativeconservative Can exclude region Can exclude region

of low 10of low 10-8 -8 with full with full Run II statisticsRun II statistics

Significantly Significantly improved analysis improved analysis will appear soon…will appear soon… Integrated Lumi/experiment (fb-

1)


Area of ImprovementArea of Improvement Improved muon selection based Improved muon selection based

on additional information:on additional information: Energy deposition in the Energy deposition in the

calorimetercalorimeter dE/dx in the drift chamberdE/dx in the drift chamber

Significant reduction in fakes Significant reduction in fakes expected.expected.

Neural Net based final Neural Net based final discriminant with additional discriminant with additional background suppression powerbackground suppression power

Use the 2-dimensional dimuon Use the 2-dimensional dimuon mass-discriminant plane to mass-discriminant plane to evaluate signal/limitevaluate signal/limit

Stay tuned for an updated Stay tuned for an updated results at winter conferences…results at winter conferences…

Current muon fake rateCurrent muon fake rate Determined for Kaon and pions Determined for Kaon and pions

of each charge from high of each charge from high statistics D*statistics D*DD00KK--++ sample sample

bbslsl++ll-- decays at Tevatron decays at Tevatron


GoalsGoals Sensitive to New Physics (Rates and Asymmetries)Sensitive to New Physics (Rates and Asymmetries)

BBdd and B and B++ modes established at B-factories modes established at B-factories BR(BBR(B++ KK++)=0.34)=0.34+0.19+0.19

-0.14 -0.14 x 10x 10-6 -6 (PDG 06)(PDG 06) BR(BBR(Bdd K*)=1.22K*)=1.22+0.38+0.38

-0.32 -0.32 x 10x 10-6 -6 (PDG 06)(PDG 06)

Re-establish signals in Tevatron data and Re-establish signals in Tevatron data and “discover” unseen “discover” unseen BBss decays decays BR(BBR(Bss )=1.6x10)=1.6x10-6-6 C. Geng and C. Liu, J. Phys. G 29, C. Geng and C. Liu, J. Phys. G 29,

1103 (2003)1103 (2003)

CDF new results with 0.92 fbCDF new results with 0.92 fb-1-1

D0 published a BR(BD0 published a BR(Bss limit with 0.4 fb limit with 0.4 fb-1 -1

PRD 74, 031107 (2006)PRD 74, 031107 (2006)


StrategyStrategy Similar to the BSimilar to the B case case Normalize signal to analogous BNormalize signal to analogous BJ/J/h h

(J/(J/) decays) decays

Blind optimization Blind optimization Exclude J/Exclude J/ and and ’ region’ region Sideband data for optimization and Sideband data for optimization and

background estimatebackground estimate Monte Carlo and data for efficiency ratios Monte Carlo and data for efficiency ratios

with normalization modewith normalization mode

)/()/(

)( /

/

JBRN

N

hJBBR

hBBRtotalh

totalhJ

hJ

h


Signal Selection Signal Selection OptimizationOptimization

CDF/DØ similar analysis:CDF/DØ similar analysis: CDF optimize Nsig / sqrt(NCDF optimize Nsig / sqrt(Nsigsig+N+Nbkgbkg) ) DØ optimize Nsig / (1 + sqrt(NDØ optimize Nsig / (1 + sqrt(Nbkgbkg) )) )

)( XY

XYL

L

i i

iTT

T

RpBp

BpIso

)0.1()(

)(|| vtxB


BBu,d u,d ResultsResults For all modes:For all modes:

pT(B)>4.0 GeV/cpT(B)>4.0 GeV/c pT(h)>1.0 GeV/cpT(h)>1.0 GeV/c

|m|mKK-m-mK*K*|<50 MeV/c|<50 MeV/c22

|m|mKkKk-m-m|<10 MeV/c|<10 MeV/c22

Counting events in 2Counting events in 2 window window around B mass, excesses seen in around B mass, excesses seen in all modesall modes

Background from 3-9 Background from 3-9 sideband sideband extrapolated to signal window extrapolated to signal window Fit shown for illustration purposeFit shown for illustration purpose


BBss Results Results CDF (920 pbCDF (920 pb-1-1):):

11 candidates found11 candidates found 3.5±1.5 expected 3.5±1.5 expected

backgroundbackground 2.4 2.4 significance significance

DØ (400 pbDØ (400 pb-1-1)) 0 observed 0 observed 1.6±0.6 expected1.6±0.6 expected


BR (BBR (Bh) h)

Good agreement & similar uncertainty with: Good agreement & similar uncertainty with: Babar PRD 73, 092001 (2006) (208 fbBabar PRD 73, 092001 (2006) (208 fb -1 -1 ~10 ~10 KK++, ~15, ~15K*K*00)) Belle hep-ex/0410006 (250 fbBelle hep-ex/0410006 (250 fb-1 -1 ~40 ~40 KK++, ~40, ~40K*K*00))

BR(BBR(B++KK++) = [0.72 ) = [0.72 ±± 0.15(stat.) 0.15(stat.) ±± 0.05(sys.)]x10 0.05(sys.)]x10-6 -6 (45 ev.)(45 ev.) BR(BBR(B00K*) = [0.82 K*) = [0.82 ±± 0.31(stat.) 0.31(stat.) ±± 0.10(sys.)]x10 0.10(sys.)]x10-6 -6 (20 ev.)(20 ev.)

BR(BBR(Bss) <2.4 x10) <2.4 x10-6 -6 @ 90% C.L. @ 90% C.L. = [1.16 = [1.16 ±± 0.56(stat.) 0.56(stat.) ±± 0.42(sys.)]x10 0.42(sys.)]x10-6-6

Improve upon DØ limit (400 pbImprove upon DØ limit (400 pb-1-1) BR(B) BR(Bss) < 3.3 x 10) < 3.3 x 10-6 -6 @ 90% C.L. @ 90% C.L.


SummarySummary CDF/DØ analyzed ~800 pbCDF/DØ analyzed ~800 pb-1 -1 of Run II data of Run II data

searching for Bsearching for B signal: signal: Current limits in the 10Current limits in the 10-8-8 territory territory No major obstacle in pushing down limits with No major obstacle in pushing down limits with

increasing exposureincreasing exposure Significantly improved analysis with >1 fbSignificantly improved analysis with >1 fb-1 -1

data sampledata sample Constraining more & more New Physics…Constraining more & more New Physics…

CDF/DØ entering the bCDF/DØ entering the bsll arena:sll arena: New solid BNew solid B++K signal from CDF K signal from CDF A 2.4 A 2.4 excess in the B excess in the Bss reported from CDF reported from CDF

U.L. limit close to SM prediction U.L. limit close to SM prediction

1.5 fb1.5 fb-1-1 on tape: more to come… on tape: more to come…


BACKUP


MB vs Likelihood RatioMB vs Likelihood Ratio


Expect 2.2 ± 0.7 background events

Cut Values changedonly slightly!

Used in previous analysis

Used now in addition! • Obtain a sensitivity (w/o unblinding) w/o changing the analysis

• Combine “old” Limit with obtained sensitivity

D0 SENSITIVITY FOR 700 pb-1

(400 pb-1)


Signal mSignal m Spectra Spectra

Mass spectra not corrected for efficiency/backgroundMass spectra not corrected for efficiency/background Resonance veto:Resonance veto:

2.9<m2.9<m<3.2 (J/<3.2 (J/)) 3.6<m3.6<m<3.75 (<3.75 (’)’) D (D+,Ds) any 2(3) track combination within 25 MeV of PDG D (D+,Ds) any 2(3) track combination within 25 MeV of PDG

massmass


NORMALIZATION MODES

450 pb-1

Clean samplesof norm events

Apply similar pre-selection requirementsas B analysis

NB+ = 6246

NB0 = 2346NBs = 421


Crossing: 396 ns: 2.5 Crossing: 396 ns: 2.5 MHzMHz

Level 1: hardwareLevel 1: hardware Calorimeter, Muon, Calorimeter, Muon, TracksTracks 30kHz (reduction ~x100)30kHz (reduction ~x100)

Level 2: hardware + CPULevel 2: hardware + CPU Cal cluster, Cal cluster, Silicon tracks Silicon tracks 900 Hz (reduction ~x60)900 Hz (reduction ~x60)

Level 3: Linux PC farmLevel 3: Linux PC farm ~ Offline quantities~ Offline quantities 100 Hz (reduction ~ x5)100 Hz (reduction ~ x5)

CDF trigger architectureCDF trigger architecture

Documents

Rare Decays at Tevatron