Charm physics at BaBar

1. Charm physics at BaBar • Mat Charles • (The University of Iowa)

2. Outline • Overview of what charm physics do we do • The detector & data sample • Some physics analyses • Summary & thoughts for LHCb

3. so data sample contains ~ 700M events The BaBar experiment Operating energy: √s ~10.6 GeV (... plus data at Y(3S), Y(2S) -- not discussed here.) 531.4/fb recorded

4. Charm physics at BaBar • D0 mixing and time-dependent (indirect) CP violation • Time-integrated CP violation in D0, D+ • Rare charm decays • Spectroscopy • Charm mesons (Ds, D0, D+) • Charm baryons (Λc, Σc, Ξc, Ωc and searches for Ξcc) • Dalitz plot analysis of charm decays • (D Dbar) production -- e.g. from charmonium-like states • Charm production in B decays • Semi-leptonic charm decays

5. D0 mixing & indirect CPV • Flagship analysis area for BaBar • Several different ways to measure D0 mixing in an inclusive environment like BaBar (or LHCb) • Details of methods & results later in the talk • Mixing parameters are unlikely to establish New Physics, but are important to measure to constrain models. • Indirect CPV is tiny in SM & would be a clear signal of NP • CPV search is nearly free -- just need to remeasure mixing parameters with D0 and D0bar separately. Normalization drops out.

6. Time-integrated CPV • Direct CPV depends on final state • In SM: Expected rate typically < 10−3 • Much smaller for some final states • With NP, rate can naturally be much larger • Depends on model & final state • Singly-Cabibbo-suppressed modes especially sensitive -- O(10−3), perhaps O(10−2) • Experimentally tricky: • Rate measurements depend on knowing production & detector asymmetries. (Not easy at per-mille level!) • Can also look for distribution asymmetries in multi-body decays(e.g. helicity moments, Dalitz plot asymmetries) • Examples from BaBar coming up Grossman, Kagan & Nir, PRD 75, 036008 (2007)

7. Rare charm decays • Searches for decay modes that are very suppressed in SM but can be enhanced by NP. • Classic modes: D0 → l+l− (very clean) • BaBar also looks for suite of 3-body modes(e.g. D → l+l− π, Λc → l+l− p) • Expands search horizon a lot... • ... but need to watch out for SM contributions like D → ϕπ, ϕ → l+l−

8. Spectroscopy • Searches for new states & measurements of quantum numbers, properties, cross-sections of existing states • Several discoveries & first measurements made at BaBar • Some expected -- e.g. first observation of Ωc*, spin of Ω− from Ξc decays • Some very unexpected -- e.g. Ds0(2317) and Ds1(2460) • Why should LHCb care about charm spectroscopy? • Better understanding of final state • Better understanding of QCD & its limitations • LHCb will be an interesting place for spectroscopy too • Huge production rates, if the background doesn’t kill you • You can see things we didn’t have the energy/luminosity/environment for before, e.g. Ξcc/Ωcc/Ωccc, new b-baryons (maybe even pentaquarks!) • Exclusive B → charm is a great place to pin down quantum numbers when you have huge statistics

9. Charm Dalitz plot analysis • Long-standing interest in Dalitz plot structure, partial waves • Good tool to probe properties of broad light-quark resonances (a0, f0, etc) • Theory interest for finding/excluding candidates for glueballs, hybrid mesons, etc. • Don’t mention the κ. • Recent attention for use in ADS method (γ measurements) and in charm mixing measurements.

10. D0 Mixing • Formalism & implications • New results from BaBar, Belle, CDF

11. Standard mixing formalism Mixing occurs for neutral mesons M0 = K0, D0, B0, Bs0 General time evolution:

12. Cartoon of mixing For convenience, define: and

13. Mixing in charmed mesons Charm mixing small compared to other mesons in SM: Mixing via box diagram (short-range) Mixing via hadronic intermediate states (long-range) K+K− π+π−π0 π+π− K+π− etc Contributes mainly to x Non-perturbative; hard to predict SM contribution. Most predictions give x,y ~ (0.001–0.01) and |x|<|y| Intermediate b: CKM-suppressed Intermediate d,s: GIM-suppressed Recent calculation: |x|≤0.01, |y|≤0.01 – less tiny! PRD 69,114021 (Falk, Grossman, Ligeti, Nir & Petrov) Tiny!

14. Example NP contributions to mixing: Supersymmetry Leptoquarks Extended Higgs New physics? • Theoretical uncertainty in SM mixing rate => can’t really observe NP by looking at mixing. • (... though x≫y would be a hint...) • Future theory input might change things. • In the meantime, can bound NP from above • e.g. Golowich, Hewett, Pakvasa & Petrov (PRD76:095009,2007) • CPV has more potential to provide a “smoking gun” -- indirect CPV expected to be very small (10−5 ×10−3) in SM but can be larger with NP. Grossman, Kagan & Nir, PRD 75, 036008 (2007)

15. D0 mixing results • Broadly, four types of measurement: • Lifetime difference between states of different CP • Time-dependence of wrong-sign hadronic decays • Wrong-sign semi-leptonic decays, e.g. D0 → K+ l− νl • Coherent D0D0 production at psi(3770) -- CLEO-c (not covered in this talk)

16. Define Lifetime ratios: Introduction D0 → K− π+: Mixture of CP states D0 → K− K+: CP-even eigenstate (also D0 → π− π+) yCP related to y and CP parameters by: Falk et al, PRD65,054034 AM≠0: CPV in mixing (asymmetry in RM between D0 and D0) cosφ≠1: CPV in interference between mixing and decay CP observables (AΓ or ΔY) defined as: Non-zero value of yCP implies mixing. If no CP violation, yCP = y.

17. Belle: D0/D0 → K−K+ Lifetime ratios: Method • Belle & BaBar methods very similar: • Require D*+ → D0 π+ tag • ID flavour of D0 at production for CPV measurement • Suppresses background • Modest improvement to lifetime resolution • Use D0 → K−K+, π−π+ as signal and K−π+ as control • Many measurement systematics cancel in the ratio • Get as clean a sample as possible • Background model systematics don’tcancel well between modes BaBar: D0 → K−K+ Diagram shows another D0 decay mode

18. Sample Sample yCP yCP ΔY AΓ K+ K− K+ K− (1.25 ± 0.39 ± 0.28)% (1.60 ± 0.46 ± 0.17)% (0.15 ± 0.34 ± 0.16)% (−0.40 ± 0.44 ± 0.12)% π+ π− π+ π− (1.44 ± 0.57 ± 0.42)% (0.46 ± 0.65 ± 0.25)% (0.05 ± 0.64 ± 0.32)% (−0.28 ± 0.52 ± 0.30)% Combined Combined (1.31 ± 0.32 ± 0.25)% (1.24 ± 0.39 ± 0.13)% (−0.26 ± 0.36 ± 0.08)% (0.01 ± 0.30 ± 0.15)% Lifetime ratios: Results BaBar preliminary Belle arXiv:0712.2249 PRL 98:211803,2007 Yellow band: D0 → K− π+ control mode No evidence for CP violation 3.0σ evidence for mixing 3.2σ evidence for mixing BaBar result can be combined with statistically independentuntagged sample (PRL 91, 162001(2002),91 fb−1) to obtainyCP = (1.03 ± 0.33 ± 0.19)%

19. DCS MIX K+π− CF Wrong-sign hadronic decays Look for wrong-sign decays, e.g. D*+ → D0 π+, D0 → K+ π− Two main contributions: Doubly-Cabibbo-suppressed (DCS) decay Mixing & Cabibbo-favoured (CF) decay Distinguish them by their time dependence: [Limit of |x| ≪ 1, |y| ≪ 1, and no CPV.] Why x′ and y′ instead of x and y? x′ = x cosδ + y sinδ y′ = y cosδ − x sinδ where δ is the phase difference between DCS and CF decays and depends on the final state. Note: (x′2 + y′2)/2 = (x2 + y2)/2 ≡ RM

20. x′,y′ highly correlated Wrong-sign D0 → K+ π− Belle: 400 fb−1 PRL 96,151801 (2006) CDF: 1.5 fb−1 arXiv:0712.1567 (preliminary) BaBar: 384 fb−1 PRL 98,211802 (2007) 3.8σ evidence for mixing 3.9σ evidence for mixing 2.0σ evidence for mixing BaBar: AD = (−2.1 ± 5.2 ± 1.5)% Belle: AD = (2.3 ± 4.7)% Belle: AM = 0.67 ± 1.2 Clear evidence for mixing! But no evidence for CP violation found.

21. Time-integrated WS data Time-integrated RS data Wrong-sign D0 → K+ π− π0 • DCS and CF components each have a Dalitz plot. • Get CF Dalitz model from time-independent fit to RS data • Get DCS Dalitz model & mixing params from time-dependent fit to WS data Contours: 68.3%, 95%, 99%, 99.9% Consistent with no mixing at 0.8% Mixing results: BaBar preliminary384 fb-1 [x′′, y′′ since phase is in general different from D0 → K+ π−]

22. D0 → KS π+ π− • Another time-dependent Dalitz plotanalysis -- but this time have: • CF contributions (e.g. D0 → K*− π+) • DCS contributions (e.g. D0 → K*+ π−) • CP-even contribution (e.g. D0 → KS ρ0) • CP-odd contribution (e.g. D0 → KS f0) • ... all in the same Dalitz plot, interfering. • ⇒ Can measure relative phases -- and hence x, y -- directly! Belle Mixing results assuming no CPV: Belle No evidence for CPV found: 95% CL contours Belle, 540 fb-1 PRL 99, 131803 (2007)

23. Look for D0 → K(*)+ l− νl Pro: No DCS contribution! Theoretically clean Con: Missing ν makes reconstruction/selection harder Semi-leptonic decays • Belle • arXiv:0802.2952, 492 fb-1 (preliminary) • Electron & muon modes • Kinematic constraints (Ecm, mD, mν) improve ν reconstruction • D0 proper lifetime cut • BaBar • PRD 76:014018 (2007), 344 fb-1 • Electron mode only • Use double-tag to suppress background Data (passing all cuts) Data (recoil cut sideband) No evidence for mixing. 90% CL: RM in (−13, +12) ×10−4 No evidence for mixing. 90% CL: RM < 6.1 ×10−4

24. HFAG Combined results arXiv:0803.0082 Mixing Indirect CP violation No-mixing point excluded at 6.7σ No-CPV point still allowed at 1σ World average: World average:

25. Time-integrated CPV results • Recent results: • D0 → K+K−, π+π− • D0 → K+K−π0, π+π−π0 • Older result (not covered here): • D0 → K+K−π+ -- PRD 71, 091101 (2005)

26. CPV in D0 → K+K−/π+π− CP asymmetry: Experimentally tricky to measure with per-mille systematics: • Tagging efficiency asymmetry for soft pion in D*+ → D0 π+ studied with control sample of D0 → K−π+ events. • Crucial to get this from data, not MC! • Control sample corrected for K+/K− and π+/π− efficiency asymmetry as function of polar angle and momentum. MC simulation • Forward-backward production asymmetry • From Z/γ interference & higher-order QED diagrams • These effects are odd in cos(θ*) • CP asymmetry is even in cos(θ*) • ... so measure aCP in bins of |cos(θ*)| & odd terms vanish 385/fb, PRL 100,061803 (2008)

27. CPV in D0 → K+K−/π+π− Plotting CP asymmetry in bins of |cos(θ*)|: Systematics Last bin excluded (due to acceptance) Results are consistent with zero CP asymmetry: 385/fb, PRL 100,061803 (2008)

28. CPV in D0 → K+K−π0/π+π−π0 • Move to three-body mode -- we now have more tools: • Look for rate asymmetry in bins of |cos(θ*)| as before • Look for asymmetry in distribution. • Second point is crucial -- CP asymmetry may pop up in one corner of phase space or in one intermediate resonance. • Remember: Direct CPV is not universal. • Localized asymmetry may be washed out -- or even cancelled -- when looking at integral over whole phase space. • Several ways used to check for distribution asymmetry: • Bin-by-bin difference in normalized Dalitz plot (model-independent) • Difference in angular moments (model-independent) • Differences in amplitudes & phases of components in Dalitz plot fit

29. CPV in D0 → K+K−π0/π+π−π0 Angular distribution asymmetry(first three Legendre polynomial moments only shown here): Look for distribution asymmetry in normalized Dalitz plots: Efficiency-corrected Dalitz plots P(χ2) = 32.8% P(χ2) = 16.6% 385/fb, arXiv:0802.4035 accepted by PRD-RC Normalized residuals No evidence of CP violation found No evidence of CP violation found

30. c.f. Belle: [arXiv:0801.2439, 532/fb] CPV in D0 → K+K−π0/π+π−π0 Asymmetries in phase-space-integrated rates? Asymmetries in Dalitz plot fits? 385/fb, arXiv:0802.4035 accepted by PRD-RC D0 → π+π−π0 D0 → K+K−π0 No evidence of CP violation found Thus, no evidence for CP violation found in any of the four tests. No evidence of CP violation found

31. Summary • BaBar has done a lot of interesting charm physics • ... and so can you! • Hopefully both results & methods will be helpful for LHCb. • D0 mixing now established (world avg: 6.7σ level) • Still large uncertainties on parameters -- more work to do • Observed mixing rate consistent with SM prediction... • ... within large theory uncertainty... • ... and at upper end of expected range. • No sign of CP violation (direct or indirect) in charm yet • Limits still well above SM expectations -- room for NP. • Lots of other BaBar charm results I glossed over...