DD Shape · Tag e ciency ( i) for each mode is determined from MC simulation via i = Nfound i =N...

$: DD Shape · Tag e ciency ( i) for each mode is determined from MC simulation via i = Nfound i =N generated i and then weighted by the branching fraction of mode i (B i) to obtain$
DD̄ Shape

Speaker: Yi FANGfor BESIII Collaboration

The 7th International Workshop on Charm PhysicsMay 18-22, 2015

Detroit, Michigan

Y. Fang (IHEP) DD̄ Shape Charm 2015 1 / 16

Outline

1 Introduction

2 Analysis

3 Summary


Introduction

Introduction

Since the discovery of ψ(3770), it is a long-standing puzzle in understandingof ψ(3770) production and decays

(MeV)(3770)ψM3760 3770 3780 3790

BES-II

Belle

BABAR

KEDR

PDG’2014

hadrons→ -e+e

hadrons→ -e+e

+K0D0 D→B

DDγ → -e+e

KD D→B

hadrons→ -e+e

PLB 652, 238 (2007)

PLB 660, 315 (2008)

PRL 100, 092001 (2008)

PRD 76, 11105(R) (2007)

PRD 77, 011102 (2008)

PLB 711, 292 (2012)

0.3±0.4±3772.4

1.9±3772.0

4.0±5.0±3776.0

0.9±1.9±3778.8

0.5±2.4±3775.5

-1.7-0.7-0.3+1.8+0.5+0.33779.2

0.33±3773.15

Includes interference between resonant

productionDand nonresonat D

(MeV)(3770)ψM3760 3770 3780 3790

Discrepant results of ψ(3770) parameters are observedModel? Interference?

Analyze DD̄ shape around ψ(3770) at BESIII with higher statistic


Introduction

BESIII Experiment

BEPCII Collider

symmetric e+e− collider, double-rings, 2.0 GeV < ECM < 4.6 GeV

BESIII Detector


Introduction

Data Sets

ψ(3770) Scan Data

∼ 70 pb−1, 3.74 < ECM < 3.89 GeVLuminosity (L) is determined using large-angle Bhabha scattering events

(GeV)CME3.75 3.8 3.85

)-1

L (p

b

0

2

4

6

8

Monte Carlo Simulation

1 ψ(3770)→ D0D̄0,D+D−

2 e+e− → qq̄, τ+τ−,γISRJ/ψ,γISRψ(2S)


Analysis

Reconstruction of D Mesons

The cross sections for e+e− → DD̄ are measured usingsingle tag method

Tag modes (charge conjunction is implied):

D0 → K−π+ D+ → K−π+π+ D+ → K 0Sπ

+π0

D0 → K−π+π0 D+ → K−π+π+π0 D+ → K 0Sπ

+

D0 → K−π+π+π− D+ → K+K−π+ D+ → K 0Sπ

+π+π−

Define variables:∆E = Etag − Ebeam

mBC =√

E 2beam − |~ptag|2

Select D candidate with ∆E closest to 0 for each mode


Analysis

Extraction of Signal Yields

Signal yield (Ntag) is extracted from a maximum-likelihood fit to the

2D distribution in ∆E vs. mBC

Signal and background PDFs are formed from MC simulation

Float normalization of signal and background

• Data

— Signal

- - Background

— Fit


Analysis

Changes in mBC Shapes due to ISR Effect

Two peaks seen in higher beam energy mBC distribution

• Data

— Signal

- - Background

— Fit

Left Peak — Events from Born Level contribution

Right Peak — Events from ISR contribution


Analysis

Cross Section

Reconstruction Efficiency

Tag efficiency (εi ) for each mode is determined from MC simulation via

εi = N foundi /Ngenerated

i and then weighted by the branching fraction ofmode i (Bi ) to obtain the average efficiency

ε =∑i

εiBi

ε0tag = (11.3± 0.2)%, ε+tag = (9.8± 0.1)%

(the branching fractions of sub-resonance decays are included)

Calculation of Cross Section

σRCDD̄

(Ei ) =Ntag(Ei )

2εtagL(Ei )


Analysis

Fits to Cross Sections

Fit the measured D0D̄0 and D+D− cross sections simultaneously using the theoret-ical cross sections

σRCDD̄

(W ) =

∫zDD̄(W

√1 − x)σDD̄(W

√1 − x)F(x ,W 2)dx

zDD̄ Factor describing Coulomb interaction

F(x , s) Probability to lose a fraction of s in initial state radiation

σDD̄(W ) =π2α

3W 2β3

D |FD(W )|2, FD(W ) = FRD (W )e iφR + FNR

D (W )

Use Breit-Wigner formula for resonant component

FRD (W ) =

6W√(Γee/α2)(ΓDD̄(W )/β3

D)

M2 −W 2 − iMΓ(W ), ΓDD̄(W ) = Γ(W )× (1 −BnDD̄)

Analyze two models for non-resonant component

Exponential Model - FNRD (W ) = FNR exp(−q2

D/α2NR)

Vector Dominance Model (VDM) - FNRD (W ) = F

ψ(2S)D (W ) + F0


Analysis

Fits to Cross SectionsSingle Breit-Wigner Shape

Single Breit-Wigner formula is unable to describe data


Analysis

Fits to Cross SectionsExponential Model


Analysis

Fits to Cross SectionsVector Dominance Model

VDM provides more physically based resultsY. Fang (IHEP) DD̄ Shape Charm 2015 13 / 16

Analysis

Preliminary Results and Comparisons

Use Γψ(3770)→DD̄ee = Γ

ψ(3770)ee ×B(ψ(3770)→ DD̄)

Remains constant from fit independent of branching fraction

Preliminary results of ψ(3770) parameters (errors are only statistical)

Preliminary results of ψ(3770) parameters are consistent with those

measured at KEDR


Analysis

Systematics

Expect statistics-limited result due to scan data size

Systematics evaluation still in progress

Current main sources (ranked by contribution to total)

1 Meson radii used for ψ(2S) and ψ(3770)

2 Charged tracking

3 Neutral tracking

4 Luminosity

Negligible effect seen when altering B(ψ(3770)→ DD̄)

All parameters remain constant except Γψ(3770)ee

Scales inversely to input branching fraction


Summary

Summary

Cross sections for e+e− → DD̄ in the vicinity of the ψ(3770) are

studied at BESIII

Able to well fit the DD̄ line shape near ψ(3770) with interference based

models

Both exponential model and vector dominance model provide quality

description of data

Upcoming aspects

1 Finalize estimation of systematic uncertainty

2 Compare to alternate models


DD Shape · Tag e ciency ( i) for each mode is determined from MC simulation via i = Nfound i =N...

Documents

Transcript of DD Shape · Tag e ciency ( i) for each mode is determined from MC simulation via i = Nfound i =N...