Measurement of D →π e ν decays at BaBar and implications ...ific.uv.es › ~oyangur › tempo...

1

Measurement of D0 →π- e+ ν decays at BaBar

and implications for Vub

Arantza Oyanguren (IFIC – Valencia)

On behalf of the BaBar Collaboration

ICHEP 2014, Valencia

• Analysis of D0→π-e+ν events

• Results of the form factor and BR

• Physics interpretation, implications for Vub

• Conclusions

• Motivation

ICHEP 2014, Valencia Arantza Oyanguren 2

Outline

Example:

• Charm semileptonic decays allow to measure form factors in the charm sector: → validate Lattice QCD calculations

→ increase theory precision involving B decays → ensure possible New Physics signs


Motivation

• The D0→π-e+ν decay channel:

Motivation

• Only one form factor: f+ (q2) (me ∼ 0)

• Partially known: contributions from the D* and D*’1 poles

• Related to the B→π form factor at the same Eπ→ access to Vub


m2D*’(2610) m2D*’’ … D*i are JP = 1- states (→Dπ)

∞

d

• From 347.2 fb-1 of continuum events at the Υ(4S) recontruct D*+→D0π+, D0→π-e+ν:

ϒ(4S) rest frame

→ Require tight PID signal pions and veto against kaons

• Based on similar techniques as in other BaBar analyses D0→K-e+ν (PRD 76 (07) 052005), Ds→K+K-e+ν (PRD78 (08) 051101 (RC)) , D+→K-π+e+ν (PRD 83 (11) 072001)

Analysis method

• Hadronic control channels: D0 →K-π+ 5 ICHEP 2014, Valencia Arantza Oyanguren

→ Impose soft π+, π- and e+ in the same hemisphere

→ Compute D direction (- ∑pall tracks ≠ π,e)

→ Fit pD0 = pπ-+pe++pν → Constraints using mD0 and mD*+

→ Compute the missing energy in the e+ hemisphere

→ Compute q2=(pD0 – pπ- )2

• D0→π-e+ν : Cabibbo suppressed (BR~0.3%); large backgrounds from π’s

Mass (GeV)

Cros

s sec

tion

→ Divide the event in two hemispheres

σcc ~1.3 nb

Analysis method

• Use δm=mD*+-mD0 sidebands from on-peak (BB+cc+light) and off-peak (37fb-1;cc+light) data samples to determine the different backgrounds (fit Emiss vs pπ) → Main systematic uncertainty in the analysis

(kept under control using data)

- BB background - Charm non-peaking (π not from D*) - Charm peaking (13 subcategories) - Light quarks

Background sources:


→ Two types of selections: tight (ε: 1.8%, S/B ~ 1.2) (nominal) loose (ε: 5.5%, S/B~0.6)

• Signal events selected in δm=mD*+-mD0 Fbb Fcc ~ 10000 candidates 50 % background in the signal region

• The background is reduced using Fisher discriminant variables against BB events (event shape) and non-signal cc events (additional tracks topology)

δm = m (D0π) – m(D0) < 0.150 GeV

• The q2 = (pD0 - pπ-)2 = (pe+ + pν)2 distribution is measured in 10 bins:

Analysis method

→ Resolution σ(q2)∼ 0.085 GeV2 (50%) and 0.311 GeV2 (50%)


Cross-check: Pseudoscalar, known angular distribution (sin2(θ) )

Signal: 5303 Bkg: 4623

(worst than experiments at the charm threshold, but not a problem since f(q2) is smooth)

• Normalization: relative to the D0→K-π+ decay channel

Measure B(D0→π-e+ν)/B(D0→K-π+) in data and in MC

PDG 2014 : BR(D0→π-e+ν) = (2.89 ± 0.08) x 10-3


Using the world average for BR(D0→K-π+):

Try to have a selection as similar as possible for the D0→π-e+ν and D0→K-π+ channels

Correct by differences between the two channels

Analysis method

• Branching fraction measurement in q2 bins, corrected for resolution and acceptance effects:

Results on the branching fraction


Diagonal: uncertainties (σ∆B x 103) Off-diagonal: bin correlations ρij

Results on the form factor • Fit to different models (used by experiments):

Model prediction: mD* = 2.010 GeV/c2

Model prediction: aI = 0.104 GeV-2

Simple pole

Modified pole (BK)

ISGW2

Taylor expansion

t ≡ q2 |z|

BaBar (2014)

• Fit to different models:


• Normalization: considering

→ data can be fitted by all models if parameters ≠ predictions → no physics meaning

Results on the form factor

• Going further in the understanding of the form factor:


Burdman and Kambor [PRD55 (1997) 2817]

being H* = D*,D*’1, D*’2,… (or B*, B*’1, B*’2, …)

→ The form factor cannot be explained by the D* and D*’1

fH*, gH* are the decay contant and coupling

• For D0→π+e-ν: - fD* , fD*’1 determined by Lattice - gH*, gH*’1 from D*, D*’1 widths measured at BaBar

→ fixed D* and D*’1 contributions

• BaBar data D* contribution D* and D*’1

D0→π-e+ν


∞


Three poles model (multipole) New!: Becirevic et al (arXiv: xxxxx)

ci given by the residua of the poles (relative to D*) in terms of decay constants and couplings

(more info in my talk tomorrow)

Fitting a 3rd pole effective with the condition (superconvergence):

→ larger than the predicted third JP=1- state by quark models ~3.1GeV, (as expected since it is effective)

Data is well described by this model


π

π

mMqmM

wDB

DBDB

,

22,

2

, 2−+

=Using wB,D instead of q2

At wB=wD :

Experimentally, the common range in wB,D

→ Aim to extract Vub with a reliable theoretical uncertainty

( )( )

22

//

π

π

νπνπ

→+

→+

=

→Γ→Γ

D

B

D

B

cd

ub

D

B

ff

MM

VV

dwDddwBd

→ Kinematic factors cancel out (same Eπ)

wB,D ∈ [1,6.7] : q2D ∈ [0; 2.975]GeV2 q2B ∈ [18; 26.4] GeV2

~17% of B→ πlν evts dΓ

/dw

0 100% of

D evts

B → πν

D→ πν

• Having measured dΓD→πeν /dq2 we can extract Vub from the relation between the D→πν and B→πν channels:

(shape from BK)

wB,D

A physics interpretation of the charm form factor may allow to use it outside the D physical region

Physics interpretation


1) From Lattice

2) From a phenome- nological model

• 1) Vub extraction (from Lattice):

→ extrapolated to the unphysical region

Statistics + pole constraints Form factor ratio = 1.8 ± 0.2

→ good fit for several considered ranges in w: data are compatible with a constant fB(wB)/fD(wD)

It can be improved by LQCD, providing values for this ratio with better accuracy and for several values of q2.

BaBar B and D sl data



Lattice QCD (Preliminary)

f +D(q2) HPQCD (preliminary) f +B(q2) HPQCD

→ the three poles form factor fitted on D0→π-e+ν

→ assuming a constant ratio of f+B(wB)/f+D(wD)

→ Using BaBar D0→π+e- and B0→π-e+ν data

Three pole form factor

[PRD86(12)092004]

• 2) Vub extraction (from the three poles model)*:

Result on the third pole (effective)

Statistics + pole constraints gb/gc ∧ ∧

B→ πeν


ICHEP 2014, Valencia 16

→ Having tested the three-pole model in D0→π-e+ν → We can use it for fitting only B0→π-e+ν data → Constraints from the first two poles (B*, B*’1) and fitting the third one

Arantza Oyanguren

couplings, entering in the residua

It can can be improved by Lattice QCD

* more info in my talk tomorrow [Becirevic et al, arXiv: XXXXX]

[PRD86(12)092004]

BaBar

Conclusions


• Measurement of the D0→π-e+ν form factor and branching fraction at BaBar, competitive and in agreement with CLEO-c and BELLE. • Physics interpretation of the form factor: - The form factor cannot be explained by the D* and D*’1 contributions. - Description in terms of a three poles model agrees well with data. • Vub has been extracted using the information of charm sl data: →Using a constant form factor ratio from Lattice. →Using the three poles model.

- Systematics of different origin, hope to be reduced in future by Lattice calculations: May shed light into the differences between the exclusive and inclusive Vub determination…

Thank you!

source variation δ(mp)/ mp (%) δ(α)/ α (%) δ(a1)/ a1(%) S1 charm background tuning 30% of total corrections -0,66 13,85 17,87 S2 D*+ production tuning 30% of total corrections 0,08 -2,31 1,69 S3 BB production tuning full effect -0,19 0,77 4,93 S4 backg. control using sidebands statistics from sidebands 0,82 14,62 -21,10 S5 Non-peaking bkg Corrections (On-Off) -0,02 0,38 0,66 S6 Peaking bkg (D0→ρνe) Br=(0.22→ 0.19 ± 0.04)% -0,12 1,92 3,75 S7 Peaking bkg (K misID) 50% of total corrections -0,07 0,38 -0,05 S8 Peaking bkg (D0→no s.l.) Increase fraction by 30% -0,01 0,38 -1,10 S9 D sl. decays (FF, Keν) 30 MeV on pole mass 0,03 -0,77 -0,59 S10 Ds sl. Decays (FF, Peν) 100 MeV on pole mass -0,17 2,69 6,32 S11 D sl. Decays (FF, Veν) r2 (0.80 → 0.80±0.03) 0,04 -0,77 -1,40 S12 D sl. Decays (FF, Veν) rV (1.50 →1.463±0.035) -0,15 2,31 5,74 S13 D sl. Decays (FF, Veν) mA (2.5 →2.63 ± 0.16 GeV) 0,11 -1,92 -3,60 S14 D sl. Decays (FF, Veν) mV (2.1 →1.9 GeV) -0,08 0,77 6,62 S15 D0 energy and direction, and missing Eν Smearing corrections (On-Off) 0,04 -0,77 -0,51 S16 Electron ID Tweaking (On-Off) -0,13 2,69 0,29 S17 Radiative corrections Increase radiative events by 30% -0,18 3,08 5,07 S18 Tight Pion ID 20% of total corrections -0,12 2,31 1,25 S19 Resolution on q2 1% slope in q2 eff. (D0→ Kππ0 ) -0,34 6,92 3,24 S20 Kaon loose PID Correction at low momentum -0,01 0,00 0,59

Total 1,19 22,69 -31,3

• Systematic uncertainties: form factor model parameter

Emiss (GeV) in the D0 hemisphere

D0→Kπ data MC

D0→Kπ data MC

ϑD,true - ϑD,event (rad) (true: from K and π tracks) (event: from the rest of the event)

Backup • Determination of D momentum:

Backup • Sidebands :

Backup

• Loose selection:

Backup • Unfolded distribution:

• Form factor model: Based on the Burdman and Kambor model: [PRD55 (1997) 2817]

Pole contribution (H*= D* or B*…)

Continuum from threshold (Hπ) to first excited state Λ∼ mH*’1

Contribution from radially excited states JP=1- ≡ H*’i

m2 D*’(2610)

with the residues of the vector resonances:

Coupling to the Hπ state

Decay constant

SUPERCONVERGENCE CONDITION (SC): (from ff behaviour at large q2)

Backup

Three poles model (Orsay)

First experimental determination of the D* decay constant:

(fixed masses for the poles)

→ The sum of the residues = 1.32 ± 0.36 ± 0.27 ≠ 0 (SC) by ~3σ. → Using SC, contraining the 1st and 2nd residues and fitting the 3rd pole as an effective pole:

mD*’2eff = 3.6 ± 0.3 GeV (χ2/NDF = 4.8/9)

~ same order as the 2nd pole

(constrained -1.1±0.4 GeV2)

Backup • Form factor model:

• Vub extraction (from the three pole model): Using the three pole form factor model in B→π data: (with SC: )

Fit the residues of the two first poles (constrained), the effective mass of the 3rd pole and Vub

(measured for D→πνe)

Backup

Número de diapositiva 1Número de diapositiva 2Número de diapositiva 3Número de diapositiva 4Número de diapositiva 5Número de diapositiva 6Número de diapositiva 7Número de diapositiva 8Número de diapositiva 9Número de diapositiva 10Número de diapositiva 11Número de diapositiva 12Número de diapositiva 13Número de diapositiva 14Número de diapositiva 15Número de diapositiva 16Número de diapositiva 17Número de diapositiva 18Número de diapositiva 19Número de diapositiva 20Número de diapositiva 21Número de diapositiva 22Número de diapositiva 23Número de diapositiva 24Número de diapositiva 25Número de diapositiva 26

Measurement of D →π e ν decays at BaBar and implications ...ific.uv.es › ~oyangur › tempo...

Documents

Transcript of Measurement of D →π e ν decays at BaBar and implications ...ific.uv.es › ~oyangur › tempo...