1. RENCONTRES DE MORIOND 2005 – QCD and Hadronic Interactions
KLOE results on
f0(980) , a0(980) scalars and
h decays
S. Giovannella (LNFINFN)
for the KLOE Collaboration
Study of the f0(980) , a0(980) mesons
Dynamics of h→ppp decay
BR of h→p0gg
RENCONTRES DE MORIOND 2005 – QCD and Hadronic Interactions
La Thuile, 12–19 March 2005

2. DAFNE: the Frascati f-factory
e+e– s 1020 MeV  M
peak  3 mb
2004 best: Lpeak = 1.2 × 1032 cm-1s-1
F(1020) g
a0(980)
f0(980)
h(550)
h'(960)
r(770) KK p 0- 1- 0+ p0
BR = 83%
BR = 15%
BR = 1.3%
 L(pb-1)
Lint(01/02) = 450 pb-1
Lint (2004) = 700 pb-1
2005: data taking in progress
1.5 ×109 f in 2001/2002
2.3 ×109 f in 2004
S. Giovannella - MORIOND QCD March 2005

3. The KLOE experiment
The KLOE design was driven by the measurement of direct CP violation through
the double ratio: R = G(KL p+p) G(KS p0p0) / G(KS p+p) G(KLp0p0)
DRIFT CHAMBER:
Helium based gas mixture
dpt / pt < 0.4%
xy  150 m ; z  2 mm
E.M. CALORIMETER:
sE / E = 5.4% / (E(GeV))
st = 55 ps / (E(GeV))  40 ps
98% solid angle coverage
PID capabilities
Magnetic field:
B = 0.52 T
S. Giovannella - MORIOND QCD March 2005

4. Light scalar mesons at KLOE
g+f0(980)
(I=0)
g+a0(980)
(I=1)
p+p-
p0p0
h p0 gg
p+p- p0
e+e- →  ─
f0(980) and a0(980) scalar mesons not easily interpreted as qq states
Other interpretations suggested:
 qqqq states [Jaffe 1977]
 KK molecule [Weinstein, Isgur 1990]
Both BR(f→Sg) and scalar mass spectra are sensitive to their nature
[Achasov, Ivanchenko 1989] ─ ─ ─
S. Giovannella - MORIOND QCD March 2005

5. Light scalar mesons at KLOE
Properties of light scalars can be investigated by fitting the Dalitz plot
of the three body PPg final state or the MPP spectra to extract the
relevant couplings.
We use two models:
1) Kaon Loop
[Achasov-Ivanchenko, NPB315 (1989) 465]
2) “No Structure”
[Isidori-Maiani, private comunication] S V g
gVS
gSpp P e+ e- S
gfKK
gSKK
gSPP P K+ K-
S described as a standard BW (fixed S)
+ higher order corrections
S. Giovannella - MORIOND QCD March 2005

6. Fit to the M(hp) spectra
data: → (BR = 39.43%)  104 events
pb →π+π-π0 (BR = 22.6% ) events
Kaon Loop
No Structure
→
→
→
→p+p-p0
→p+p-p0
→p+p-p0
Mπ (MeV)
√s (MeV)
S. Giovannella - MORIOND QCD March 2005

7. The p0p0g final state
450 pb-1 from 2001 – 2002 data taking ~ 400k events
Two main contributions to p0p0g final Mf:
f→Sg→p0p0g
√s = MeV
e+e-→wp0→p0p0g
S. Giovannella - MORIOND QCD March 2005

8. Fit to the p0p0g Dalitz: kaon loop
Fit to the Dalitz plot includes:
1. scalar term [f0(980)]
2. VDM contributions
3. All interference terms
Statistical errors only
r/r
p1(2)
p2(1) g w e+ e─
 Data
 Fit
f/w/w
p1(2)
p2(1) g r e+ e─
VDM well under control
Scalar term not well reproduced
S. Giovannella - MORIOND QCD March 2005

9. Fit to the p0p0g Dalitz: kaon loop
Kaon loop model including s(600) according to PRD 56 (1997) 4084
significantly improves the fit!
 Data
 Fit
Statistical errors only
S. Giovannella - MORIOND QCD March 2005

10. Fit to the p0p0g Dalitz:no structure
Also in this case we fit the p0p0g Dalitz with the scalar term, the
VDM contributions and their interferences
f0 + s (KL)
Scalar (NS)
Statistical errors only
 Data
 Fit
Similar fit quality w.r.t. KL including s(600)
Low mass tail less evident
S. Giovannella - MORIOND QCD March 2005

11. The p+p-g final state e+e-→ +- events with the
photon at large angle (45<<135)
Main contributions:
ISR (radiative return to , )
FSR
Search for the f0 signal as a deviation
on M(+-) spectrum from the
expected ISR + FSR shape
676,000 events selected ( )
Events/1.2 MeV
f0(980) region
M(pp) (MeV)
M(pp) (MeV)
√s (MeV)
Expected interference pattern of svis vs √s
in the signal region Mpp[9001000 MeV]
S. Giovannella - MORIOND QCD March 2005

12. Fit to p+p-g Fit of the full Mpp spectrum with: KL NS Full spectrum
Signal region
ISR+FSR
S. Giovannella - MORIOND QCD March 2005

13. Scalar mesons: coupling to the f
In the “no structure” approach the crucial parameter is the coupling of the scalars to the f (i.e. to the s quark)
Meson
gfMg (GeV-1) p0
0.12 h
0.66 h
0.70 a0
1.3 f0
1.8 – 2.3
Increasing
strange quark
content
Large coupling values!
S. Giovannella - MORIOND QCD March 2005

14. The h→ppp decay in cPT
The amplitude of the h→ppp decay occurs primarily due to the d-u
quark mass difference. From lowest order Chiral Pertubation Theory:
with
and, at L.O. ,
A good understanding of M(s,t,u) can lead to a very accurate determination of Q:
S. Giovannella - MORIOND QCD March 2005

15. The h→ppp KLOE
The h meson is produced via f→ hg (17 Mevts in 2001/2002)
Final state for h → p+p-p0 : +-
Final state for h → p0p0p0 : 7g’s
In both decays the physical background is almost negligible
f→ hg → p0p0p0g
p+p-p0 selection:
2 track vertex + 3g candidates
Kinematic fit
p0p0p0 selection:
7 photon candidates
S. Giovannella - MORIOND QCD March 2005

16. The h→p+p-p0 Dalitz plot
Resolution:
Dalitz shape
“core”
sY = 0.027
“core”
sX = 0.018
X(T+-T-) , YT0
B/S  0.8%
Efficiency  36% , almost flat
S. Giovannella - MORIOND QCD March 2005

17. Fit to the h→p+p-p0 Dalitz plot
|A(X,Y)|2 = 1+aY+bY2+cX+dX2+eXY+fY3
KLOE Preliminary
c and e are C violating
Ndf
P2 (%) a b c d e f
147 60
 0.006
0.117  0.006 
0.047 0.006
0.008
0.13 0.01
150 63
 0.005
---
0.02
 0.004
0.100 0.005
0.12 0.01
 0.003
0.120  0.005
0.043 0.006
S. Giovannella - MORIOND QCD March 2005

18. Q evaluation
Using the preliminary KLOE result ICHEP04 :
[B.V. Martemyanov and V.S. Sopov, PRD 71 (2005) ] :
Q = 22.8 ± 0.4
This result implies a violation of the Dashen theorem (QDT=24.2)
based on DMp(e.m.)=DMK(e.m.) and is in agreement with:
Q = 22.0 ± [J.Donoghue and A.Perez, PRD 55 (1997) 7075]
[J.Bijnens and J.Prades, NPB 490 (1997) 239]
Q = 22.4 ± [J.Kambor., C.Wiesendanger and D.Whyler, NPB 465 (1996) 215]
Q = 22.7 ± [A.V.Anisovich and H.Leutwyler, PLB 375 (1997) 335]
The theoretical error is currently dominated by PDG error on G(h→ppp)
S. Giovannella - MORIOND QCD March 2005

19. The h→p0p0p0 Dalitz plot z  [ 0 , 1 ]
The Dalitz plot density is described with a single parameter: |A|2  a z
with:
Ei : Energy of the i-th pion in the h rest frame
r : Distance from the center of the Dalitz plot
rmax : Maximum value of r
z  [ 0 , 1 ]
S. Giovannella - MORIOND QCD March 2005

20. Fit to the h→p0p0p0 Dalitz plot
350 pb-1 of data, using an high purity sample on the photon pairing in p0 → Purity = 98.5% , e = 4.5%
a = ± 0.005stat  0.004syst
KLOE Preliminary
Current best measurement from Crystal Ball: a = ± 0.004stat+syst
S. Giovannella - MORIOND QCD March 2005

21. h→p0gg: a window on p6 cPT… 0 8
L2 contributions
at tree level:
= 0 (Q = 0) p2
Coupling proportional to the
charges, zero also for
tree level. 
+,K+
-,K- 0 8
1-loop contributions from L2 vertices, suppressed by G parity conservation and kaon mass suppression: p4
Gp-loops = 0.84 ×10-3 eV
GK-loops = 2.45 ×10-3 eV h
r0, w, f p0 g
L6 coefficients determined by the meson propagator. Th. expectations:
G6VMD = ( 0.42 ± 0.20 ) eV
G6ENJL = ( 0.58 ± ) eV
G6Ch unit= ( 0.47 ± 0.10 ) eV h8 a0 p0 g p6
S. Giovannella - MORIOND QCD March 2005

22. … and an experimental challenge!
p- + p ® h + n
(CERN, Brookhaven, GAMS, Crystal Ball)
p+ + d ® p + p + h (67)
p+ + p ® p+ + p +h (67,69)
K- + p ® L + h (70 AGS)
p+ + n ® h + p (71)
p- + n ® p- + n + h (80)
f ®hg (SND 01)
S. Giovannella - MORIOND QCD March 2005

23. Analysis of h→p0gg @ KLOE}
f → h g
p0gg
5g final state
s ~ 8 pb (GAMS BR measurement)
Background candidates:
Process
s (pb)
e+e-→wp0 , w→p0g , p0→gg
450
f→f0g , f0→ p0p0 , p0→gg
300
f→a0g , a0→ hp0 , h/p0→gg
260
f→hg , h→gg
17000
f→hg , h→ p0 p0 p0
13000
Process
s (pb)
e+e-→wp0 , w→p0g , p0→gg
450
f→f0g , f0→ p0p0 , p0→gg
300
f→a0g , a0→ hp0 , h/p0→gg
260
f→hg , h→gg
17000
f→hg , h→ p0 p0 p0
13000
Process
s (pb)
e+e-→wp0 , w→p0g , p0→gg
450
f→f0g , f0→ p0p0 , p0→gg
300
f→a0g , a0→ hp0 , h/p0→gg
260
f→hg , h→gg
17000
f→hg , h→ p0 p0 p0
13000
Rejected with veto on additional w-p0-h
Rejected with energy momentum conservation and likelihood technique to identify merged clusters
S. Giovannella - MORIOND QCD March 2005

24. BR(h→p0gg) = ( 8.4 ± 2.7stat ± 1.4syst ) × 10-5
h→p0gg: BR evaluation
The shape of signal + background after fit
well reproduces the data
NDATA = 735
Nbkg = 667  36
Nsig = 68  23
N(3p0) = 2,288,882
BR(h→p0gg) = ( 8.4 ± 2.7stat ± 1.4syst ) × 10-5
KLOE Preliminary
S. Giovannella - MORIOND QCD March 2005

25. BRh→p0gg: comparison with models
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
GAMS (1984)
[1] J. N. Ng et al.(1992)
[2] J.N. Ng et al (1993)
[3-4] L. Ametller et al. (1992)
[5] S. Bellucci et al. (1995)
[6] E. Oset et al (2003)
[7] J. Bijnens et al. (1996)
[8] A. A. Belkov et al.(1996)
68%
90%
95%
99%
Statistical significance
Crystal Ball
(2004)
KLOE
KLOE 1s
KLOE 2s
KLOE 3s
P6 calculations
S. Giovannella - MORIOND QCD March 2005

26. Conclusions/Outlook
The f0(980), a0(980) scalar resonances are clearly observed in different
final states, indicating large values for the couplings with f
We are analyzing an unprecedented statistics of h→3p decays, determining
accurate values of the Dalitz plot slopes for p+p- p0 /p0 p0 p0 final states
We measure a value of BR(h→p0gg) smaller than previous measurement
and in agreement with ChPT p6 calculations
A factor 5 improvement in statistics expected at the end of current run:
Complection/extensions of the on-going analysis (including out of peak data)
Search for S→KK
Measure of all main h BRs + rare/forbidden h decays (Nh ~ 90 Mevts )
Limit on gluonium content in h [BR(f→hg)/BR(f→hg)]
S. Giovannella - MORIOND QCD March 2005