CLEOIII Upsilon results In principle includes CLEOIII dipion

CLEOIII Upsilon results • In principle, includes: • CLEO-III dipion transitions between vectors – Complements CLEO 05 results on transitions between L=1 P-states • High-precision measurement of dielectronic width of Y(1 S), (2 S) (3 S) • Many radiative results: – Observation of exclusives already presented: , KK, – Upper limits on and ’ modes – UL on multibody modes (>=4 charged tracks) – Comparison of inclusive quark/gluon production in radiative decays of Y vs. qq+photon (ISR)

CLEOIII CLEAN signals, angular analysis underway.

Dipion transitions • Renewed interest in `double-bump’ structure in (3 S) (1 S) following Ba. Bar observation of 4 S (n. S) Goal: spin/parity analysis across invariant mass to determine whether low-mass bump is sigma 0 – if not, what is it?

Exclusives: Multibody modes • Exclusive radiative events ‘bumps’ in the inclusive (scaled to Ebeam) photon spectrum (assume narrow recoil object) • We perform a series of fits to the inclusive photon spectra as a function of E in order to set an E -dependent upper limit on these radiative events. • Nota bene: ‘bumps’ in the inclusive photon spectra can also be caused by continuum threshold effects (ccbar, e. g. )

*→ + , → 4 MC An example, albeit exaggerated, of signal. . . (10 -2)

Method (Fitting Spectrum) • We fit each step to a Gaussian+Chebyshev polynomial • Step along the photon spectra with the Gaussian mean • Fix Gaussian sigma at each step to be the detector resolution (~1% @ 5 Ge. V) • Looking for narrow resonances so the measured photon energy dist. should be Gaussian with Gaussian width E.

Efficiencies ( *→ + , →? ) 4 4 K 4 p 2 p 2 K 2 2 K 4 0 4 K 0 4 p 0 2 p 2 0 59 2% 50 2% 67 2% 62 3% 56 2% 53 3% 60 2% 48 2% 65 2% 54 5% 2 p 2 K 0 2 2 K 0 4 2 0 4 K 2 0 4 p 2 0 2 p 2 K 2 0 2 2 K 2 0 4 4 0 4 6 0 50 5% 53 2% 59 1% 49 2% 63 5% 57 2% 54 3% 57 2% 60 2% 4 8 0 6 6 K 6 p 60 2% 74 3% 68 4% 52 4% Worst Phase Space High Mult.

All limits on the order of 10 -4

In/Out and Sensitivity Check • Embed signals at a given level into data. • We then apply our procedure to the resulting spectra • We construct all signals above our upper limit floor (~10 -4) in our accessible recoil mass range

A(M )+1. 645* A(M )

d. N/d(A/ A)(< (1 S)) Check of pulls: Continuum data A/ A

Results • Our sensitivity is of order 10 -4 across all accessible values of M • Above threshold for any known B( (1 S)→ +pseudoscalar, pseudoscalar h+h-+neutrals) • We measure for all M : B( (1 S)→ + , 4 charged tracks) < B( (2 S)→ + , 4 charged tracks) < B( (3 S)→ + , 4 charged tracks) < 1. 05 x 10 -3 1. 65 x 10 -3 5. 70 x 10 -3

Results (2) • Restricting M to 1. 5 Ge. V < M < 5. 0 Ge. V we measure: 1. 82 x 10 -4 1. 69 x 10 -4 3. 00 x 10 -4 • We report these upper limits as a function of recoiling mass M (see conf. Paper) • B. R. ’s are all ~10 -4. • N. B. Not in conflict with any observed two -body radiative decays to-date (due to 4 charged track requirement here) B( (1 S)→ + , 4 charged tracks) < B( (2 S)→ + , 4 charged tracks) < B( (3 S)→ + , 4 charged tracks) <

Many modes! Dedicated search for 1 S gh and 1 S gh’; Observed in J/psi decay at 10 -4 and 4. 7 x 10 -4 level

Only upper limits quoted at this time… Suggests dedicated search for (1 S) c?

Quarks v. Gluons • 1981 (CESR): e+e- collisions (ECM ~ 10 Ge. V) produce ; ggg allows high-statistics study of gluon fragmentation • Isolate gluons: ggg decay of Isolate quarks: fragmentation • 1984 Find: more baryons/event in ggg decay than • Weakness: 3 partons (ggg) vs. 2 partons ( ) 3 strings (ggg) vs. 1 string ( ) • Solution: decay of vs. decay of continuum

Y(1 S) 3 gluons, but also 2 -gluon source: • e+e- (CLEO) • e+e- (1 S) (CLEO) • e+e- Z 0 (LEP) Z 0

Data Sets Data Set 1 S 2 S 3 S 4 S Below 4 S Luminosity (1/fb) 1. 19 1. 07 1. 42 5. 52 2. 10 ECM (Ge. V) 9. 46 10. 02 10. 36 10. 58 10. 55 Note that for 2 S and 3 S have not corrected for cascades: • (2 S) (1 S) + X • (3 S) (2 S) + X (3 S) (1 S) + X Are included as consistency checks, but have subtractions and corrections that have not been included.

Method: vs. • Bin according to particle momentum • Count N(Baryon) per bin and normalize to hadronic event count • Enhancement is: Continuum-subtracted Resonance Yield Continuum Yield Enhancement = 1. 0 Particle is produced as often on resonance as on continuum

Method: vs. • Bin particle yield recoiling against high-E photon according to tagged photon momentum • Count N(Baryon) per bin and normalize to photon count in that bin • Enhancement is: Continuum-subtracted Resonance Yield Continuum Yield Enhancement = 1. 0 Particle is produced as often on resonance as on continuum

Detector and Generator Level: ggg manageable bias; use correction factor where appropriate; discrepancy in/out used for systematics Λ p p φ

Proton L f f 2 results • Successfully reproduce CLEO 84 indications of baryon enhancement in 1 S (ggg) vs. CO ( ) fragmentation • Comparison of baryon production in 1 S ggγ vs. e+e- (comparing two gluon to two quark fragmentation) -1 S gg baryons shows much reduced enhancement relative to baryons -Effect not reproduced in JETSET MC Ggg/qqbar Ggγ/qqbargamma Ratio p 1. 30 ± 0. 01 1. 10 ± 0. 02 ~ 1. 2 Antip 1. 33 ± 0. 01 1. 19 ± 0. 03 ~1. 1 Λ 2. 56 ± 0. 02 1. 97 ± 0. 03 ~1. 3 φ 0. 85 ± 0. 03 1. 1 ± 0. 3 ~0. 8 f 2 0. 66 ± 0. 04 1. 4 ± 0. 9 ~0. 5

Deuteron Production (Preliminary) B(1 S (ggg+gg )) d+X= 2. 86(0. 30)x 10 -5 Per event enhancement of deuteron production in gluons vs. quarks ~12. 0(2. 0). Also: note 1 S psi >> continuum psi

Summary • Radiative decays (in general) continue to be more elusive than for J/psi • Baryon coupling to 3 -gluons confirmed (even larger for deuterons!); enhancement in 2 -gluons mitigated. • Ramping down these efforts (CLEO-III CLEOc) • Future improvements/results hopefully to emerge from B-factories with dedicated Upsilon running • Thanks to everyone who did the work!

Overview • Reproducing CLEO 84 indications of baryon enhancement in 1 S (ggg) vs. CO ( ) fragmentation • New comparison of baryon production in 1 S ggγ vs. e+e- comparing two gluon to two quark fragmentation -First time such a comparison has been made • Essential results: -1 S gg baryons shows much reduced enhancement relative to baryons -Effect not reproduced in JETSET MC • Additional cross-checks (2 S, 3 S, comparison with mesons) included

Data Results: ggg p and p: 2 S/3 S data corrected

Data Results: ggγ Λ: 2 S corrected

Method (Extracting Limit) • Plot the gaussian area A(x ) from fits to inclusive photon spectra • Convert into an upper limit contour with height=A(x )+ 1. 645* A(x ) • A(x ) is the Gaussian fit sigma • Negative points → 1. 645* A(x )

The M -Dependent Upper Limits • Divide on-resonance fits by efficiency corrected number of (1 S), (2 S) and (3 S) events (-1 events) • Divide off-resonance fits by luminosity of off-resonance running and derive xsct UL’s • Note: +f 2(1270) will not show up in this analysis since B (f 2 4 tracks) is approximately 3% • B ( (1 S) + , + - 0) << 10 -4

CHECK OF PULL DISTRIBUTIONS

Fragmentation Models • Simplistically there are two models: Parton vs. String • Parton: g or q radiates a new particle • String: g and q are connected by a string (gluon). Particles move apart; string stretches and breaks; forms new particles e+ q e- • String model is what is in Jet. Set MC (CLEO: Jetset 7. 4 PYTHIA) Parameters tuned to √s = 90 Ge. V LEP Data

Data Results • Show data and detector level MC enhancements for both ggg and ggγ • “Corrected” data and generator level MC enhancements for those with a low CL fit. • Systematic errors have been introduced based on the correction factor.

Data Results: Momentum-Integrated 1 1 Λ p p φ f 2 Λ p φ f 2