1. BBCB-WS, Nov., 2007, S.Uehara, Belle
Study of pp production in
two-photon collisions at Belle
S.Uehara (KEK)
on behalf of The Belle Collaboration
BES-Belle-Cleo-Babar J.W.S, Nov., 2007, Beijing
BBCB-WS, Nov., 2007, S.Uehara, Belle

2. Motivation of two-photon physics
hadrons
Resonance formation or
Hard scattering with a quark line
Hadron production from two quasi-real photons
(|q2|<0.001GeV2, with zero-tag condition)
Exclusive processes with gg c.m. energy
WWgg= GeV at KEKB/Belle
Wide energy region --- Various physics aspects
can be studied simultaneously.
Light-quark resonances, charmonia
Perturbative /Non-perturbative QCD, Hadron production mechanism
BBCB-WS, Nov., 2007, S.Uehara, Belle

3. General physics in gg p+p-/p0p0 final states
quark
Resonance p
soft production
of mesons
soft hadron
exchange
p+p- and p0p0 :
Physics is almost the same, but gives independent information
Electromagnetic production as mesons: s(p0p0)<<s(p+p-)
Resonances (I=0, JPC=even++) decaying via strong interaction:
　　　　　　　　　　　　　　s(p0p0):s(p+p-) =1:2
Test of QCD: s(p0p0)/s(p+p-) : ~0.03 (LO) – 0.5 (“handbag”)
c.m.-energy (W) and angular dependences must also be examined.
BBCB-WS, Nov., 2007, S.Uehara, Belle

4. Resonances to appear in ggpp
(I=0, JPC=even++) mesons --- fJ or ccJ (J=even)
f2(1270) has been observed in p+p- and in p0p0 as a big peak.
f0(980) not significantly seen in the past experiments.
Are any mesons previously reported seen ?
f GeV
f2 1.6 – 2.1GeV
f4(2050) etc.
- The charmonia: cc0 and cc2
s(p0p0):s(p+p-) =1:2
Interference between a resonance and continuum component modifies the ratio.
We expect that continuum contribution in p0p0 is smaller than
in p+p-, in the intermediate or low energies
BBCB-WS, Nov., 2007, S.Uehara, Belle

5. BBCB-WS, Nov., 2007, S.Uehara, Belle
Results of ggp+p-
Belle Collaboration
Phys. Rev. D　75, (R) (2007)
J. Phys. Soc. Jpn. 76, (2007)
Phys. Lett. B 615, 39 (2005)
W = 0.8 – 1.5 GeV
W = 2.4 – 4.1 GeV
BBCB-WS, Nov., 2007, S.Uehara, Belle

6. BBCB-WS, Nov., 2007, S.Uehara, Belle
f0(980) in gg  p+p-
f0(980)
– is an ordinary qq meson?
- an exotic state?
- some special theory needed?
Its two-photon coupling is a
crucial key.
Prediction of Ggg :
uu, dd – 1.8 keV
ss – 0.5 keV
KK molecule – 0.6 keV
85.9fb-1
f2(1270)
f0(980)
Backupに式 P2
F2(1270)の分布説明
Flatte
Energy depoの何の分布？
かたっぽtag?
Pimu separationの精度
p/m separation:
by comparing energy deposit in
the CsI calorimeter between
data and MC
f0(980) appears as a small but
statistically significant peak.
BBCB-WS, Nov., 2007, S.Uehara, Belle

7. BBCB-WS, Nov., 2007, S.Uehara, Belle
Fit to the f0(980) line shape
f0(980) is observed
as a clear peak.
f0(980) p+p- 
Backupに式 P2 
|resonance|2
M =
Gp+p- =
Ggg =
MeV/c2
MeV eV
Interference term
BBCB-WS, Nov., 2007, S.Uehara, Belle

8. New upper limit for Br (h’p+p-)
No h’p+p- enhancement is seen near 958MeV/c2
This is a P and CP mode;
previous upper limit:
Br (h’p+p-) < (HBC (1969))
New upper limit for
Br(h’p+p-) < 2.8 10-3
(interference with the non-resonant process allowed)
< 3.3 10-4 (no interference)
h’ mass
BBCB-WS, Nov., 2007, S.Uehara, Belle

9. BBCB-WS, Nov., 2007, S.Uehara, Belle
p+p- measurements at high energies
PLB 615, 39 (2005)
Two-photon measurements of 　Ggg(ccJ)Br (eV)
GggBr（cc0p+p-)  2.1  GggBr（cc0K+K-)  1.6  2.3
GggBr（cc2p+p-)  0.14  GggBr（cc2K+K-)  0.76  0.14
BBCB-WS, Nov., 2007, S.Uehara, Belle

10. Preliminary results of ggp0p0
Belle Collaboration
arXiv: v1 [hep-ex]　(2007)
BBCB-WS, Nov., 2007, S.Uehara, Belle

11. Selection of p0p0 candidates
Based on data with an integrated luminosity of 95 fb-1
Only 4 photons are visible
- Triggered by Calorimeter triggers
- No reconstructed track with
pt>0.1GeV/c, dr<5cm, |dz|<5cm　
- 2 or more photons OR 1 or more p0s
- Just 2 p0s with E(decay product photon)>70MeV, pt>0.15GeV/c
- Transverse-momentum (pt)-balance of the two p0’s
|Spt|<0.05GeV/c
- Satisfying a rough trigger condition at offline:
N_gamma=4 or Esum_lab>1.25 GeV
in the ECL-trigger region, 17<qlab<128.7deg
BBCB-WS, Nov., 2007, S.Uehara, Belle

12. Raw-event distribution of p0p0
W = M(p0p0)
|cosq*| in gg c.m.s
About 1.26M events (before
the background subtraction)
| cos q*|
|cos q*|<0.4
| cos q*|
W (GeV)
W (GeV)
cc0
|cos q*|<0.8
cc2
Estimated
background
W (GeV)
BBCB-WS, Nov., 2007, S.Uehara, Belle

13. pt-balance distributions and backgrounds
W=0.90GeV, |cosq*|= W=0.66GeV, |cosq*|=0.05
pt-unbalanced backgrounds
are seen in W<1.2 GeV
Histogram –
experimental
Low-energy
beam-background photons
Intermediate or
Higher energies
Non-exclusive process
sources not significantly seen
W=1.18GeV, |cosq*|= W>3.6GeV, |cosq*|<0.4
Dots – experimental
Histo. – Signal MC
|Spt| (GeV/c)
BBCB-WS, Nov., 2007, S.Uehara, Belle

14. BBCB-WS, Nov., 2007, S.Uehara, Belle
Detection Efficiency
Determined by signal MC simulation
and fitted with smooth functions of
(W, |cos q*|)
Trigger efficiency is determined
by the trigger simulator after tunings of the energy thresholds
the total energy trigger – 1.15 GeV
counting number of clusters (>=4) MeV
Close to 100% above W>1.3 GeV
Acceptance
Around 11% at maximum, typically 5 %
Large angular dependence due to geometrical condition of the detector
Overall efficiency = trigger efficiency Acceptance
W (GeV)
BBCB-WS, Nov., 2007, S.Uehara, Belle

15. Unfolding and Energy calibration
Correction for a smearing effect from a finite W resolution
Apply Unfolding to the experimental W distribution (in each angle region)
using an asymmetric Gaussian (1.3% higher side and 1.9 % lower side).
Unfolding is applied in 0.9 GeV < W < 2.4 GeV.
Experimental sample of h’  gg
The energy scale is correct.
Histogram:
Data
Smearing func. peaking at the nominal mass
Before (orange) and
After (blue) the unfolding
|cosq*|=0.225
Yield
W(GeV)
BBCB-WS, Nov., 2007, S.Uehara, Belle

16. Cross sections of gg p0p0
Obtain first
ds/d|cosq*| in each D|cos q*|=0.05 bin
Then, accumulate them over the angle
Preliminary
Note:
After the Unfolding, different data bins and errors are no longer independent of each other.
Data in the charmonium region
are omitted from the plot.
Comparison with Crystal Ball at DORIS II (1990)
They are consistent.
Belle has ~103 times more data.
Statistical errors only.
BBCB-WS, Nov., 2007, S.Uehara, Belle

17. Systematic-error sources
Trigger efficiency % @ W > ~ 1.1 GeV (HiE)
5% - W< ~1.1 GeV (Clst4)
-p0 reconstruction efficiency %
Acceptance (dominated by pt-balance cut) - - 3% - 5%
Background subtraction % of the subtraction component
3% in High W where no subtraction applied.
Luminosity function % - 5%
Total % - 10% @ W<~1.05GeV
~9% @ 1.05GeV<W<~2.7GeV
10% - 11% @ W>~2.7GeV
BBCB-WS, Nov., 2007, S.Uehara, Belle

18. BBCB-WS, Nov., 2007, S.Uehara, Belle
Differential cross section (Angular dependence)
gg p0p0
Preliminary
Low W  High-W
Large angle enhancement
appears  disappears
Small angle enhancement
disappears  appears
ds/d|cos q*| (nb)
Change  ~ 2.0 GeV
ds/d|cos q*| (nb)
BBCB-WS, Nov., 2007, S.Uehara, Belle

19. BBCB-WS, Nov., 2007, S.Uehara, Belle
Resonant structures
gg  Resonance  p0p0
Clearly seen: f0(980) and f2(1270)
New broad peaks ~1.65 GeV and ~1.95 GeV
Hints with lower significance : ~1.45 GeV, cc0 (and cc2)
Sizes of the charmonia
Ggg(cc0)B(cc0p0p0):
8.4  2.2  0.8 eV (4.5s)
Ggg(cc2)B(cc2p0p0):
0.29  0.23  0.03 eV (1.4s)
Upper limit < 0.75 eV (90%CL)
Preliminary
The central values are consistent with
the expectations from isospin invariance
between the p+p-/p0p0 decays
|cos q*|<0.4
BBCB-WS, Nov., 2007, S.Uehara, Belle

20. Cross section ratio s(p0p0)/s(p+p-)
|cosq*|<0.6
Preliminary
Error bars are statistical only.
Systematic errors:
About 10% at each measurement point,
except p+p- above the charmonium masses (~25%).
Preliminary
Charmonium region is
omitted from the plot.
The ratio is 0.3 – 0.4 at W>2.7 GeV
BBCB-WS, Nov., 2007, S.Uehara, Belle

21. Conclusion of p+p- measurement
- We measure the process gg p+p- with 85.9 fb-1 data.
The differential cross sections are obtained in
0.8 GeV < W < 1.5 GeV and |cos q*|<0.6 in the gg c.m. frame.
(# We separately published before for the region 2.4 GeV<W<4.1GeV)
- We confirm small but statistically significant peak from f0(980).
New upper limit for P and CP violating process h’p+p- is obtained.
BBCB-WS, Nov., 2007, S.Uehara, Belle

22. Summary for p0p0 measurement
- We measure the process gg p0p0 with 95fb-1 data.
Preliminary results of the differential cross sections are obtained
in 0.6 GeV < W < 4.0 GeV and |cos q*|<0.8 in the gg c.m. frame.
- Several resonant structures are seen, including a statistically significant peak from f0(980).
Behavior of the angular dependence (forward- and/or large-angle enhancements) changes drastically at around W= 2.0 GeV.
The cross section ratio s(p0p0)/s(p+p-) in 3 GeV region is obtained.
BBCB-WS, Nov., 2007, S.Uehara, Belle

23. BBCB-WS, Nov., 2007, S.Uehara, Belle
BACKUP slides
BBCB-WS, Nov., 2007, S.Uehara, Belle

24. KEKB Accelerator and Belle Detector
Asymmetric e- e+ collider
8 GeV e- (HER) x 3.5 GeV e+ (LER) s=10.58 GeV  (4S)
Beam crossing angle: 22mrad
Continuous injection
Luminosity
Lmax=1.71x1034 cm-2s-1
∫ Ldt  740 fb (Nov.2007)　　
High momentum/energy resolutions
CDC+Solenoid, CsI
Vertex measurement – Si strips
Particle identification
TOF, Si-aerogel, CDC-dE/dx,
RPC for KL/muon
BBCB-WS, Nov., 2007, S.Uehara, Belle

25. Measurements of two-photon processes at Belle
ggcc2  gJ/y PLB 540, 33(2002)
gg(f2'(1525)etc.  )K+K- ( GeV) EPJC 32, 323 (2003)
gg (cc0,cc2)p+p-,K+K- ( GeV) PLB 615, 39 (2005)
gg (hc )pp ( GeV) PLB 621, 41 (2005)
gg  cc2'  DD PRL 96, (2006)
gg (f0(980)etc. )p+p- ( GeV) PRD 75, (R) (2007),
Phys. Soc. Jpn. 76, (2007)
gg (cc0,cc2 )K+K- ( GeV) PLB 651, 15 (2007)
gg (hc,cc0,cc2 )p+p-p+p-,K+K-p+p-,K+K-K+K-
arXiv: [hep-ex], to appear in EPJC
gg  a2's  p+p-p0 ( GeV) hep-ex/
gg (hc )LL,S0S0 ( GeV) hep-ex/
gg p0p arXiv: [hep-ex]　
BBCB-WS, Nov., 2007, S.Uehara, Belle

26. Reactions at KEKB/Belle
The cross sections observable there
e+e-U(4S)BB nb
e+e-qq (uds) nb
e+e-cc nb
e+e-t+t nb
gg  hadrons (Wgg>0.8GeV) --- ~1 nb
(within the acceptance)
Ee+e-=10.58GeV
Q2max=1GeV2
Luminosity function (1/GeV)
Luminosity Function
for quasi-real gg collisions
ds(in ee) = s(in gg) Lgg(W) dW
Calculated with equivalent-photon approximation
BBCB-WS, Nov., 2007, S.Uehara, Belle

27. Two-meson exclusive final states
C = + resonances: Peak structure in the invariant mass distribution
Test of QCD: c.m.-energy (W) dependence and
angular dependence in high energies
General Analysis strategy
Invariant-mass of hadron pair = W (c.m. energy of gg)
Quasi-real two-photon collisions with final state fully reconstructed
--- strict transverse momentum balance is required
(| Spt| < MeV/c )
This gives efficient background rejection
BBCB-WS, Nov., 2007, S.Uehara, Belle

28. Selection of p+p- candidates
Based on data with an integrated luminosity of 85.9fb-1
- Two charged tracks (oppositely charged)
with pt>0.3 GeV/c, <cos qlab <+0.82
|Spt*| (in e+e- c.m.s) < 0.1 GeV/c
- p/K, p/p separations
using TOF, Aerogel-Cherenkov and dE/dx
Muon background subtraction
with fitting the energy-deposit distribution on ECL
KLM (muon detector with iron filter) cannot be
available in these low energies
BBCB-WS, Nov., 2007, S.Uehara, Belle

29. Differential cross section
ds/d|cos q*| (nb)
BBCB-WS, Nov., 2007, S.Uehara, Belle

30. Angular dependences at some energy points
Red dashed lines ---
Size of point-to-point systematic error
Systematic structure near |cos q*|=
could be due to a measurement bias
BBCB-WS, Nov., 2007, S.Uehara, Belle

31. Fitting formula for f0(980)
BBCB-WS, Nov., 2007, S.Uehara, Belle

32. Angular dependence in f2(1270) region
Close to |Y2|2 2
Consistent with
pure J=2 and l=2
But, mathematically
the partial waves cannot be fully separated taking into account the interference between J=0 and 2.
BBCB-WS, Nov., 2007, S.Uehara, Belle

33. BBCB-WS, Nov., 2007, S.Uehara, Belle
Fit to f2(1270) line shape
Components:
f2 (1270) resonance
Smooth continuum:
Interfering component (l=2)
No-interfering component (I=0)
resonance
data
and fit
continuum (sum)
This is for consistency check:
Known mass and width of
f2(1270) (PDG2006) are used.
The continuum and relative phase parameters etc. were floated.
continuum (l=2)
Interference +continuum
BBCB-WS, Nov., 2007, S.Uehara, Belle

34. Trigger efficiency study
Correlation between the two triggers:
HiE (Etot_trg>1.15GeV), Clst4 (>=4 clusters each with >110MeV)
(The thresholds have been tuned.)
* -- data, dashed histogram – trigger simulator (tsim)
HiE in Clst4
large p0 c.m. angle
Clst4 in HiE with Nclst=4
Etrg
(GeV)
Minimum cluster energy (GeV)
Clst4 in HiE
small p0 c.m. angle
Clst4 in HiE with Nclst=4
Minimum cluster energy (GeV)
Etrg
(GeV)
BBCB-WS, Nov., 2007, S.Uehara, Belle

35. Trigger efficiency from the simulation
Trigger efficiency is determined
by the trigger simulator, tsim
after tunings of the energy thresholds
for
the total energy trigger (HiE) – 1.15 GeV
and for
counting number of clusters (Clst4)
- 110 MeV
In each of 16 c.m. angular bins
0.0 < |cos q*| < 0.8,
Fitted by a two-dimensional
function of (W, |cos q*| )
BBCB-WS, Nov., 2007, S.Uehara, Belle

36. pt-balance resolution and correction to efficiency
the pt-balance cut
Efficiency is affected by
pt-balance resolution.
W =1.96 – 2.00 GeV
|cos q*| = 0.5 – 0.6
Dots -- experimental
Curve – fit to data
(considering backgrounds)
(peak: 16 MeV/c)
Histogram – signal MC
(peak: 14 MeV/c)
The resolution in the experimental
data is by ~15% worse than the MC.
Correction factor: 5% - 10%,
depending on W and |cos q*|
BBCB-WS, Nov., 2007, S.Uehara, Belle