Inclusive b-quark and upsilon production in DØ Horst D. Wahl Florida State University DIS 2005, Madison

Outline l Tevatron and DØ detector Bottomonium ϒ(1S) production l High p t μ-tagged jet production l conclusion

Tevatron – data taking l peak luminosity in 2005 above cm -2 s -1 l DØ collected > 690 pb -1 l Results shown use pb -1

Leading order Flavor creation Next to leading order Flavor excitation Gluon splitting

Recent developments l Beyond NLO: resummation of log(p t /m) terms  FONLL l Changes in extraction of fragmentation function from LEP data l New PDFs l Improved treatment of experimental inputs (use b-jets and b-hadrons instead of b-quarks)

Open Heavy Flavor Production l Long-standing “discrepancies” between predicted and measured cross sections now resolved; e.g. Cacciari, Frixione, Mangano, Nason, Ridolfi,JHEP 0407 (2004) 033 l combined effects of better calculations (Fixed order (NLO)+ NLL= FONLL) relationship/difference between e+e−and hadron colliders different moments of FF relevant better estimate of theory errors (upward!) new appreciation of issues with“quark-level” measurements l Total Cross sections (from CDF): –inclusive b cross section: |y<1| 29.4±0.6±6.2 µb hep-ex/ (submitted to PRD) inclusive c cross section: ~50x higherPRL 91, (2003)

Comments: Tevatron as HF Factory l The cross sections given on the previous slide imply central (|y|<1) b’s at 3kHz at current luminosities ~2x this if you look out to |y|<2 central charm at 150kHz –~2x1010 b’s already seen by CDF and DØ in Run II –~1x1012 charm hadrons already produced(!) l ⇒ Near infinite statistics for some measurements l ⇒ If you can trigger... Rely heavily on muon triggers  J/ψ decays are golden  semi-leptonic decays-  rare decays with leptons Tracks with significant b/σb Missing neutrals troublesome  forget  o identification  all-charged decay modes

Muon Toroid Calorimeter Solenoid, Tracking System (CFT, SMT) The DØ Detector

Forward Preshower detector Silicon TrackerFiber Tracker Solenoid Central Preshower detector 125 cm 50cm 20cm   DØ tracking system

DØ - Muon detectors l Toroid magnet (1.9 T central, 2.0 T forward) l Scintillation counters l PDTs (central) l MDTs (forward) A-  Scint Forward Tracker (MDTs) Shielding Bottom B/C Scint PDT’s Forward Trigger Scint

Bottomonium production l Theory modeling of production Quarkonium production is window on boundary region between perturbative and non-perturbative QCD factorized QCD calculations to O(α 3 ) (currently employed by Pythia) color-singlet, color-evaporation, color- octet models Recent calculations by Berger et al. combining separate perturbative approaches for low and high-p t regions  Predict shape of pt distribution  Absolute cross section not predicted l ϒ (1S) Tevatron: 50% produced promptly, i.e. at primary vertex 50% from decay of higher mass states (e.g. χb →ϒ(1S)  ) l Event selection - luminosity: ± 10.3 pb-1 - di-muon: pT>3, tight Tracking and Calorimeter isolation cut - invariant mass in 7 – 13 GeV

Why measure ϒ (1S) production at DØ Because we can: The ϒ (1S) cross-section had been measured at the Tevatron (Run I measurement by CDF) up to a rapidity of 0.4. DØ has now measured this cross-section up to a rapidity of 1.8 at √s = 1.96 TeV Measuring the ϒ (1S) production cross-section provides an ideal testing ground for our understanding of the production mechanisms of heavy quarks. There is considerable interest from theorists in these kinds of measurements: E.L. Berger, J.Qiu, Y.Wang, Phys Rev D (2005) and hep- ph/ ; V.A. Khoze, A.D. Martin, M.G. Ryskin, W.J. Stirling, hep-ph/

The Analysis l Goal: Measuring the ϒ (1S) cross-section in the channel ϒ (1S) → μ + μ - as a function of p t in three rapidity ranges: 0 < | y ϒ | < 0.6, 0.6 < | y ϒ | < 1.2 and 1.2 < | y ϒ | < 1.8 l Sample selection Opposite sign muons Muon have hits in all three layers of the muon system Muons are matched to a track in the central tracking system  p t (μ) > 3 GeV and |η (μ)| < 2.2 At least one isolated μ Track from central tracking system must have at least one hit in the Silicon Tracker

d 2 σ(  (1S)) dp t × dy N(  ) L × Δp t × Δy × ε acc × ε trig × k dimu × k trk × k qual = L Luminosity k dimu local muon reconstruction y rapidity k trk tracking ε acc Acceptance k qual track quality cuts ε trig Trigger 0.0 < y < < y < < y < 1.8 ε acc – ε trig k dimu k trk k qual Efficiencies,… l Cross section:

MC Data * p t (μ) in GeV η(μ) φ(μ) 0.6 < | y ϒ | < 1.2 * 9.0 GeV < m(μμ) < 9.8 GeV. Data vs Monte Carlo l To determine our efficiencies, we only need an agreement between Monte Carlo and data within a given p T ( ϒ ) and y( ϒ ) bin and not an agreement over the whole p T ( ϒ ) and y( ϒ ) range at once

Fitting the Signal l Signal: 3 states ( ϒ (1S), ϒ (2S), ϒ (3S)), described by Gaussians with masses m i, widths (resolution) σ i, weights c i,(i=1,2,3) Masses m i = m 1 +  m i1 (PDG), widths σ i = σ 1 (m i /m 1 ), for i=2,3 free parameters in signal fit: m 1, σ 1, c 1, c 2, c 3 l Background: 3rd order polynomial All plots: 3 GeV < p t (  < 4 GeV m(  ) = ± GeV m(  ) = 9.415± GeVm(  ) = ± GeV 0 < |y  | < < |y  | < < |y  | < 1.8 PDG: m(  (1S)) = 9.46 GeV

0.0 < y ϒ < ± 19 (stat) ± 73 (syst) ± 48 (lum) pb 0.6 < y ϒ < ± 20 (stat) ± 76 (syst) ± 50 (lum) pb 1.2 < y ϒ < ± 19 (stat) ± 56 (syst) ± 39 (lum) pb 0.0 < y ϒ < ± 14 (stat) ± 68 (syst) ± 45 (lum) pb CDF Run I: 0.0 < y ϒ < ± 15 (stat) ± 18 (syst) ± 26 (lum) pb Results: dσ( ϒ (1S))/dy × B( ϒ (1S) → µ + µ - ) Fro central y bin, expect  factor 1.11 increase in cross section from 1.8TeV to 1.96 TeV (Pythia)

σ(1.2 < y ϒ < 1.8)/σ(0.0 < y ϒ < 0.6) Pythia Comparison with previous results

Effects of polarization l CDF measured ϒ (1S) polarization for |y ϒ | < 0.4. How can we be sure that our forward ϒ (1S) are not significantly polarized ? l So far there is no indication for ϒ (1S) polarization. CDF measured α = ± 0.22 for p T ( ϒ )> 8 GeV α = 1 (-1) ⇔ 100% transverse (longitudinal) polarization The vast majority of our ϒ (1S) has p T ( ϒ ) < 8 GeV Theory predicts that if there is polarization it will be at large p T. l No evidence for polarization in our signal (|y| < 1.8); not enough data for a fit in the forward region alone. l estimated the effect of ϒ (1S) polarization on our cross-section: Even at α = ± 0.3 the cross-section changes by 15% or less in all p T bins. same effect in all rapidity regions.

Question: Why is CDF's systematic error so much smaller than ours ? l Better tracking resolution --- CDF can separate the three ϒ resonances: → Variations in the fit contribute considerable both to our statistical and systematic error. → We believe we have achieved the best resolution currently feasible without killing the signal l Poor understanding of our Monte-Carlo and the resulting large number of correction factors. l Signal is right on the trigger turn-on curve.

Conclusions ϒ (1S) cross-section Presented measurement of ϒ (1S) cross section BR(→μμ) for 3 different rapidity bins out to y( ϒ ) = 1.8, as a function of p t ( ϒ) First measurement of ϒ (1S) cross-section at √s = 1.96 TeV. Shapes of dσ/dp t show very little dependence on rapidity. Normalized dσ/dp t is in good agreement with published results (CDF at 1.8TeV) l μ-tagged jet cross section: Measured dσ/dp t in central rapidity region |y|<0.5 for μ-tagged jets originating from heavy flavor

Motivation Under hypothesis of compositeness, deviation from point-like behavior would likely manifest in third generation. Conclusion: g  bb may exhibit desired deviant behavior. Explore b quark dijet mass as a possible signature. l Problem ~100:1 QCD:bb Solutions   tagging  2 nd VTX tagging  Impact parameter Fit to CDF qQCD calculation CDF: PRL 82 (1999) 2038 Fact: The multi-generational structure of the quark doublets requires explanation and could herald compositeness.

 -tagged Jet Cross-section l Given the simplicity of the calculation, there are few likely sources of the excess seen in p13. These could conceivably be N  JES (central value) Resolution (i.e. smearing)  T Trigger Eff  PV Primary Vertex Eff  j Jet Eff    Eff f b  Frac b   (Pt > 4 GeV) f B  Frac B   (Pt > 4 GeV) L Luminosity  p t Pt bin width  HF HF cross-section  bg background cross-section Jet +  (Pt > 5 GeV) Correlated p13

p14 Analysis Summary Inclusive  -tagged jet l corrJCCB (0.5 cone jets) Standard Jet quality cuts, Standard JET Triggers l Jet tagged with MEDIUM muon (more on this later)  R( , jet) < 0.5 l |y jet | < 0.5 l JES 5.3 l Long term goal was b-jet xsec. Difficult due to no data-driven determination of b-fraction.

p14 Skimming l Start with CSG QCD skim l Turn into TMBTrees (40M events…on disk) l Skim on Trees Remove bad runs (CAL, MET, SMT, CFT, JET, Muon) Remove events w/o 2 jets Use Ariel d0root_ based package SKIM 1:  1 leading jet has ~ MEDIUM  (P  (  ) > 4 GeV) SKIM 2:  1 leading jet has ~ loose SVT

p14 All Data: CSG Skims Bad runs & lumblk removed in luminosity. Only bad run removed in event counts for skims. Up until Run (07-JUN), V12.37.

Trigger Turn On Jet Trigger Collinear muon |y jet | < 0.5 Luminosity weighted Statistics uncorrelated poisson –(wrong, of course) JES corrected (5.3)

Efficiency Detail Value TT Trigger Eff  PV Primary Vertex: |z| < 50cm, ≥ 5tracks 0.84 ±   Eff (geom, μ det., tracking, match) 0.37 ± 0.05 jj Jet Eff (jet quality cuts) 0.99 ± 0.01 f bg  Frac background   (Pt > 4 GeV) Pt dependent f HF  Frac heavy flavor   (Pt > 4 GeV) Pt dependent Efficiencies…. [0.37 ± 0.05]

 JES Definitions l Required identically 2 jets Pt(jet 3, uncor) < 8 OR third jet doesn’t pass jet QC l One jet contains muon, the other doesn’t. |  | > 2.84 l Imbalance variable: l Independent variable:

Jet energy scale for μ-tagged jets l μ-tagged jets also have neutrinos ⇒ offset -- correction needed l Imbalance in events with 2 jets (one with, one without μ) – find 3.8% offset, not strongly p t dependent for p t in (75, 250GeV) l Scale energies of μ-tagged jets by factor l Order-randomized imbalance used to get resolution STD JES 5.3 gives a 3.8% offset for  -tagged jets. It is independent of Pt ( GeV). Maybe higher above that. Need to rebin and revisit the idea that the muon Pt may be mis- measured. Same plot when scaling the  -tagged jet energies by 3.8%.

Energy Resolution

resolution Neutrinos in μ-tagged jet  resolution worse than for jets without μ take rms of order randomized imbalance l Parameterize, Fit (fig. (a)) Subtract (in quadrature) resolution for jets without μ  obtain resolution for μ-tagged jets (fig. (b) l Fit: N = 7.7  4.1 S = 1.9  0.1 C = 0.0  0.1 l Resolution parameterization used in “unsmearing”

Fitting Functions VariableValueError N1N k1k N2N k2k VariableValueError N9.56 × × 10 6  

Extraction of Correction Factors exponential “normal”

Point by Point Unsmearing Factors Exponential “normal” Unsmearing Error small ~5% for Pt > 100 GeV

HF fraction of μ-tagged jet sample l Sample of jets with μ-tagged jets contains jets with μ from non-HF sources (e.g. , K decays…) l Use Pythia with standard DØ detector simulation to find HF fraction of jets tagged with muons vs (true) p t l Fit with A + B e -Pt/C A = “plateau”, B = “zero” C = “turnon”

l Pythia using standard DØ MC. l NLO uses NLO++ (CTEQ6L) From Pythia, find fraction of jets tagged with muons (HF only). Multiply NLO cross-section by Pythia muon-fraction. This is effectively the NLO k factor.

Conc l DØ Note and Conference note to EB025 l Residual small bug in code (should have only a few percent effect). l JES error must be reduced to use this before setting limits on new physics.