Tevatron Run II

Charm physics in CDF

Sandro De Cecco, INFN Roma1 for the CDF Collaboration

Incontri di Fisica delle Alte Energie

Festa della liberazione, 2003, Lecce

Tevatron Run II

Tevatron

DØ CDF

Chicago

↓

Booster

Run II: proton-antiproton collisions at √s=1.96 TeV

C D F

Tevatron pp collider

9

9

(36x36 bunches) Æ ~ 2M collisions per second

9

Current Tevatron status

Initial Luminosity (10³⁰ cm^-2 s^-1)

Today:

4.3x10³¹cm^-2s^-1

4 to 7 pb^-1 /week delivered

Goal:

5-8 x 10³¹ cm^-2s^-1 2 fb-1 in run IIa

Outline

•CDF detector & trigger overview

•Charm Physics topics

– J/ψ production cross section

– Charm production cross section – Mass: D

and D

– Cabibbo suppressed D

decay – CP violation

– D

mixing

– Rare Decays, D

Jµµ

Tracking: Si strips + drift chamber (in 1.4T ) Muons:

•Central:

•Fwd:

Time of flight

Scintil. PID (p,K,π) 100 ps @ 140 cm

EM + HAD calor.

• Central:

• “Plug” :

Trigger

2D-silicon tracks at Level2

SELECT:

b/c events from background

reconstructing a decay vertex

Integrated Luminosity

Mar 02

Data used for results : Mar 02 – Jan 03

130 pb^-1 (delivered) 100 pb^-1 (to tape)

After: good run, Silicon conditions Æ B/Charm: ~ 70 pb^-1

Jan 03

commissioning

CDF CLEO

Beauty and Charm physics at pp collider

• Total inelastic x-section (70 mb)

~ 10³ - 10⁴ × σ(bb/cc)

9 x-section bb/cc is O(10⁵)/O(10⁶) larger than e⁺e^- @ ϒ(4S) or @ Z⁰ 9 Open wide spectrum of B/D hadrons:

B^±, B⁰, B , B , Λ , Ξ ……

Strategy is:

to TRIGGER

on displaced tracks

with SILICON VERTEX

Æ

The S

V

T

d

XFT

COT

hits

d

, Φ

, P

SVT

hits

Detector Raw Data

Level 1 storage pipeline:

42 clock cycles

Level 1 Trigger

L1 Accept

Level 2 Trigger

L2 Accept

L3 Farm

Level 1

•7.6 MHz Synchronous Pipeline

•5544 ns Latency

•50 KHz accept rate

Level 2

• Asynchronous 2 Stage Pipeline

•20 µs Latency

•300 Hz accept rate 7.6 MHz Crossing rate

132 ns clock

20µs !!!

SVT performances

•Level 2: Silicon Vertex Trigger – Impact Parameter resolution:

~ 50 µm

(35µm beam size + 35µm SVT)

•Increase physics sensitivity of the Run II CDF:

– CDF is a “Charm Factory”

•> Millions of D’s per 100 pb^-1

– Collect Hadronic B/D sample:

•No Lepton required in final state

•B_s physics (mixing in D_sπ)

Select ON-LINE displaced vertices

with IP parameter cut on tracks

HF Triggers and data samples

Displaced trk + lepton (e, µ)

IP(trk) > 120µm Pt(lepton) > 4 GeV Semileptonic modes

2-Track Trig.

Pt(trk) > 2 GeV IP(trk) > 100 µm

fully hadronic modes

Di-Muon (J/ψ)

Pt(µ) > 1.5 GeV

J/ψ modes down to low Pt(J/ψ) (~ 0 GeV)

Larger yield: lower Pt threshold wrt RunI: e (µ): 8 (2.2) Æ4 (1.5) GeV Better S/N Æ trigger on long-lived decays (displaced tracks)

-Quarkonia, rare decays

- CP violation

- Masses, lifetimes

- High statistics lifet.

- Sample for tagging studies

-B_S mixing

-Hadronic charm & beauty - CP asymmetries

…{Thanks to SVT trigger}

“a Classic”

Charming Physics analysis

I will show status and prospects of a few selected

on going

•Studies of QCD

– Onium Production (J/ψ)

• Cross section

• Polarization

– Charm production

• Cross section

• D** Production

•Rare Decays – D Jµµ, …

•Masses and Lifetime – D⁰, D⁺, D_s, Λ_c, …

•CKM studies & New physics –Mixing

• In D⁰ with lifetime diff. ∆Γ

• In D⁰ time dependent analysis –Direct CP violation

• D⁰JKK,ππ

(1) Dimuon J/ψJµµ dataset

• (1) Dimuon dataset:

– 2 central muons p_T > 1.5 GeV

• Run I : > 2 GeV

–Trigger on J/ψJµµ decays

• Collect ~ 70 pb^-1

– ~ 0.5M J/ψJµµ signals

(2) Lepton + displaced track dataset

•(2) Lepton + Track

– 1 muon/electron p_T> 4 GeV – 1 other track with

• p_T > 2 GeV, SVT IP > 120 µm –M(l-Track) < 5 GeV

•Collect ~70 pb^-1 of data – ~ 0.5M B J lX signal

•(3) Two Track Trigger – 2 Tracks with

•p_T>2GeV

•SVT IP > 120 µm – p_T1+p_T2 > 5.5 GeV

•Collect ~70 pb^-1 of Data – ~ 0.5M D⁰JKπ signal

(3) Two displaced tracks dataset

Run I J/ψ Production Cross Section

•Run I Measurement:

– LO calculation: 1/ 100 x CDF

•Non-relativistic QCD

– Include color octet states – Theory doesn’t predict the absolute normalization

•Fitting the CDF data

•Prediction

– J/ψ production is dominated by the color octet mechanism

– J/ψ is polarized at high p_T

•Some discrepancy (~ 2σ) between the Run I polarization measurement and NRQCD

CDF Run I

J/ψ J µµ Cross Section (Run II)

•1.5x2 = 3 < M(J/ψ) – 2xM(µ) – Trigger on stopped J/ψ

•We can measure cross section down to p_T = 0

– σ(ppJJ/ψ; p_T>0; |η|<0.6)

•Dimuon Mass distribution for the lowest pT bin (0-250 MeV)

Background is subtracted

J/ψ J µµ Cross Section (Run II)

σ(ppJJ/ψ; p

>0; |η|<0.6)

BR = 240 +-1 (stat) +35 -28 (syst) nb

Production Cross Section: Charm

•Run I Measurement:

– D* J D⁰π: D⁰JµνKX

•muon with p_T > 8 GeV

–Slightly higher than theory expectation

•Run II

– Use SVT sample ÆPt as low as 5.5 GeV

– Early Run II data (~6 pb^-1)

•enough statistics for counting experiment, D⁰, D⁺, D*⁺, D_s CDF Run I (unpublished)

Production Cross Section: Charm

Major issues:

9bb fraction of the sample

9Trigger acceptance (MC):

•With O(10 pb^-1) Æ

X-sec

D⁰/D*/D^±/D_s/Λ^c

In the near future ccbar correlation

studies will be possible

-

FIRST TIME at Collider !

Production Cross Section: Charm

•For measuring the Charm cross

section, we need to separate direct D and BJD decays

– Use Impact parameter of D

– D meson from B decay has larger impact parameter

•Direct Charm fraction

– D⁰: 86.6 ± 0.4 ± 3.5 % – D^*+: 88.1 ± 1.1 ± 3.9 % – D⁺: 89.1 ± 0.4 ± 2.8 % – D_s⁺: 77.3 ± 3.8 ± 2.1 %

B D

K

π X

P.V.

BÆD⁰: 16.4-23.1%

Production Cross Section: Charm

Production Cross Section: Charm, results:

σ

= ½N

* f

/ (L * ε

* BR

) (f

L

.; ε

)

based on 5.7 pb-1 of SVT data (|y|<1):

• σ(D

, p

>= 5.5 GeV) = 13.3±0.2±1.5µb

• σ(D

, p

>= 6.0GeV) = 5.2±0.1±0.8µb

• σ(D

, p

>= 6.0 GeV) = 4.3±0.1±0.7µb

• σ(D

, p

>= 8.0 GeV) = 0.75±0.05±0.22µb

Comparison with NLO calculations

from M. Cacciari and P. Nason by private communication calculations of the direct D meson cross-section.

These calulations have not yet been published, but they follow the same prescription (called FONNL) as what was used for their paper 'IS THERE A SIGNIFICANT EXCESS IN BOTTOM HADROPRODUCTION AT THE TEVATRON?' (Phys.Rev.Lett.89:122003,2002, hep-ph/0204025)

Mass: ∆ Μ(D _s – D ⁺ )

•D_s, D⁺ J φπ ; φ JKK

– Same final state, almost identical kinematics

– 10 pb^-1 of two track trigger

•Measure mass difference – Systematics are reduced:

•Result: M(D_s) – M(D⁺) =

99.41 + 0.38 + 0.21 MeV/c

(PDG: 99.2+0.5 MeV/c²)

FIRST CDF RUN II PAPER !

High precision measurement

allows stringent test of HQ

effective theories

After correction for relative acceptance of SVT trigger & reconstruction for the 3 decays

⇓

K π

^KK

56320±490 5670±180 2020±110

Γ(D→KK)/Γ(D→Kπ) = 11.17 ± 0.48(stat) ± 0.98 (syst) % Γ(D→π π )/Γ(D→Kπ) = 3.37 ± 0.20(stat) ± 0.16(syst) %

WORLD BEST MEASURES: CLEO2 (PDG 2002)

•Γ(D→KK)/Γ(D→Kπ) = 10.40 ± 0.33 ± 0.27 %

•Γ(D→π π )/Γ(D→Kπ) = 3.51 ± 0.16 ± 0.17 %

Already competitive measurements in Charm with The first statistic collected with the new SVT trigger.

9.6 pb-1 SVTdata

NOW : huge sample of D mesons Æ attempt for D0 mixing & CPV in D decays

ππ

Cabibbo Suppressed D

Decays (@ICHEP 2002)

D* tagging

•Very high purity D⁰ signal using “D* tag”

technique

–D*⁺JD⁰π: Q =39 MeV –M(D*)–M(D⁰):

• σ(M_D) ~ 10 MeV

• σ(∆M) ~ 0.6 MeV

– 20% of the D⁰ : D* tagged

From 0.451M D⁰ Æ 78160 D* tagged D⁰

•Eliminate the “reflection”

background (D⁰JKπ and πK)

•No PID applied yet

•Initial flavor of D⁰ is known

– D*⁺ J D⁰ + π⁺ / D*^- J D⁰ + π^-

– The best place to study D⁰ mixing and direct CP violation

σ~0.6 MeV ^{W/o D* tag}

with D* tag

NOW: 65 pb-1:

Cabibbo Suppressed D ⁰ Decays

•Summer 2002 (10 pb^-1): no D* tagging

Br(D⁰JKK)/Br(D⁰JKπ)=(11.17+0.48+0.98)%

Br(D⁰Jππ)/Br(D⁰JKπ) = (3.37+0.20+0.16)%

– main systematics: background subtraction

•Spring 2003 (65 pb^-1): with D* tagging – Repeat the relative BR measurement

with δ(rel.BR) ≈ 1 %

(without D*: ≈ 10% for KK)

c

u u

d u d W

c

u u

u s W

V

s

V

N = 8320+- 140

N = 3697+- 69

• Can be observed through an A^CP between DÆf and its CP conjugate

• Need two weak amplitudes to interfere (Cabibbo allowed decays tree amp.only)Æ possible in Single Cabibbo suppressed (tree + penguin)

• A^CP expected to be <O(10-3) in Cabibbo suppressed modes like KK and ππ

CP violation in D

decays:

) (

2

) (

2 )

( )

(

) ( ) (

2 1

* 2 1 2

2 2

1

2 1

* 2 1

δ δ δ δ

− +

+

= − Γ

+ Γ

Γ

−

= Γ

cos A ReA A

A

sin A ImA f

f

f A_CP f

Γ ( D f ) ≠ Γ( D f ^_ ) ^-

A₁e^iδ1

A₂e^iδ2 A^*₂e^iδ2 A^*₁e^iδ1

Best single measurement at present comes from CLEO 2:

ACP(KK) = (0.0 +- 2.2 +- 0.8)% with ~ 3000 KK ACP(ππ)= (1.9 +- 3.2 +- 0.8)% with ~ 1100 ππ

With 65 pb-1 we expect to upgrade soon (summer 2003) previous measurements:

δA (KK,ππ) < 2% and δA (KK,ππ)

D ⁰ mixing

•D mixing in the SM ~ O(10^-3)

•New Physics can enhance it

•For CDF:

- High statistics

- excellent S/B (purities ~ 95%) (with D*) - High mass and proper time resolution

c

u

u c

d,s

W^-

W⁺

V_cd,cs V_ud,us

V*_ud,us V*_cd,cs

 D

〉 _ _  D _

〉

Two main way to study it:

– lifetime ratio: y = ∆Γ/2Γ = τ^mix/ τ^CP - 1

where CP mixed final state can be Kπ, and CP eigenstates can be KK and/or ππ

– Measuring wrong sign decays in D⁰ÆKπ and D⁰ÆKlν: (x²+y²) time-dependent analysis to deconvolve DCS decays

…News soon from CDF

Rare Decay Search: D ⁰ J µµ

•BR (D⁰ J µµ )

– SM expectation : ~ 10^-13

– Possible enhancement by new physics

• R-parity violating SUSY: ~ 10^-6

– Current best limit: < 4.1x10^-6 (90% CL) ( E777, Beatrice )

•Analysis

–Use D* tagged D⁰

–Use D⁰Jππ signal for normalization mode

•Almost identical kinematics –Br(D⁰Jππ) ~ 1.5x10^-3

•300 D⁰Jππ J ~1 D⁰Jµµ signal (Br=4.1x10^-6)

–Major backgrounds: D⁰Jππ, and muon fake rate from D⁰JΚπ

Rare Decay Search: D ⁰ J µµ

Muon detector fiducial

•Expected background after the optimized selection cuts:

– 1.7+0.7 events

•Fake: 0.22+0.02

•Combinatorial: 1.5+0.7

•New best limit

BR < 2.4 x 10^-6 at 90% CL

{ < 4.1x10^-6 (90% CL) ( Beatrice ) } With much higher integrated luminosity:

further background study D⁰J eµ, ee and D⁺Jµµπ

Conclusions: a lot of Charm to come from CDF

• Run II CDF collected ~100 pb^-1 of data

– ~70 pb^-1 for B/Charm physics in the SVT trigger

• CDF is an high statistics Charm experiment: O(10⁷)D⁰ in 2fb^-1

•J/Ψ production x-sec from Pt=0

•First time D meson production x-sec measured at collider.

•World best Ds-D+ Mass difference measurement (first CDF paper is on Charm)

•In D0 ACP in Single Cabibbo suppressed modes very promising and D0 mixing Æ

CDF will in principle, perform competitive measurements before CLEO-C New limit (x 2 better) on rare D0Ƶµ decay BR.

And (not included in this talk) other charmed particles studies D**.

…Backup slides

CDF Trigger System Overview

• Crossing: 396 ns: 2.5 MHz

• Level 1: hardware

– Calorimeter, Muon, Track – 15kHz (reduction ~x200)

• Level 2: hardware + CPU – Cal cluster, Silicon track – 300 Hz (reduction ~x5)

• Level 3: Linux PC farm – ~ Offline quantities – 50 Hz (reduction ~ x6)

•Statistical uncertainty for tagging efficiency –A typical tagging: ε=0.1,D=0.4,εD²=1.6%

–1000 events: εD² =1.6+0.7% (44%) –100K events: εD²=1.60+0.07% (4.4%)

•We can’t study/optimize the flavor tagging with ~O(1000) events of the B signal events

– B J J/ψK: ~ 1000 events/100pb^-1 – B J Dπ: ~ 500 events/100pb^-1

•Our solution: Use Semileptonic B decays in the lepton + track dataset

– ~200K semileptonic B signal events – High B purity

– Lepton Charge = Decay flavor of B

Flavor Tagging

No charm contamination

Rare Decay Search: D ⁰ J µµ

•Background (1)

–D⁰Jππ with both πJµ fake –N_bg = N(ππ) x prob(fake) ²

– fake prob. Is measured in D⁰JKπ signal

•Background (2)

–Combinatorial background

–Linear extrapolation of the high mass sideband events

Hadronic B signals

•Two track trigger data (65 pb^-1)

•Reconstruct hadronic B decays – B⁰JD⁺π (D⁺JKππ): 413+40 – B⁺JJ/ψK(J/ψJll): 311+25

Normalization mode for

the other decays

Hadronic B _s and Λ _b Decays

•B_s J D_s π

– Golden mode for B_s mixing

•65 pb^-1 of two track trigger data – B_sJD_sπ(D_sJφπ) : 40+10 events – B_sJD_s*π (D_sJφπ) : 65+20 events

•More channels to be added – B_s J D_s πππ

– D_s J K*K, K⁰_sK, πππ

•Further optimization of trigger strategy to obtain more signals

•Estimate the sensitivity for B_s mixing –Flavor tagging, time resolution…

• Λ_b J Λ_cπ (Λ_cJpKπ) – ~ 40 events in 65 pb^-1

•More channels to be added – Λ_b J Λ_cπππ, pD⁰π

– Λ J Λπππ

BJh ⁺ h ^-

•BJh⁺h^- signal in the two track trigger sample

– 301+27 signal events – Good S/N ~ 1

•This signal is combination of four decay channels

– Tree (Br~5x10^-6)

•B⁰ J ππ : B_s J Kπ

– Penguin (Br ~1.5x10^-5)

•B⁰ J Kπ : B_s J KK

•We can separate these decays – Decay kinematics

BÆh ⁺ h ^- from 2-Track Trigger

Experimental challenge:

Disentangle 4 channels:

kinematics + dE/dx

Final resolution expected is B

Æ ππ B

Æ Kπ

B

Æ KK B

Æ Kπ

dE/dx [ns]

BJh ⁺ h ^-

•Kinematical Separation – α = (1 – p₁/p₂) q₁

– M(ππ)

•dE/dx Separation

First results expected soon - Br(B⁰,B_s->KK,Kπ,ππ)

- Direct CP asymmetry in BJKπ M(ππ) is Lorentz invariant

If it’s really BJππ

M(ππ) is not Lorentz Invariant for BJKπ

Simulation

Simulation