Study of the radiative decays f  f0(980)g and f  a0(980)g with the KLOE detector C.Bini Universita’ “La Sapienza” and INFN Roma for the KLOE collaboration Outline Scalar Mesons at a f - factory The KLOE experiment Results: 3.1 p+p-g 3.2 p0p0g 3.3 hp0g 4. Summary and perspectives

How a f-factory can contribute to the understanding of the scalar mesons f(1020) Mass (MeV/c2) 500 1000 Motivations: Extract f to scalar “coupling” Since f  |ss> G(f  g”scalar”)  s-quark content  4-quark vs. 2-quark states Radiative decays: f  f0(980)g f  a0(980)g f  f0(600)g a0(980) I=0 I=1/2 I=1 f0(980) f0(600) “s” K*0(800) “k” 2. Comparison a0(980) – f0(980); 3. Need to introduce f0(600) ?

How to detect these radiative decays - I  f0(980)g  p+p-g (charged mode)  p0p0g (neutral mode)  K+K-g [ 2m(K)~m(f0)~m(f) ]  BR [Achasov, Gubin] ~ 10-6  K0K0g “ “ ~ 10-8 (charged mode): f0(980) signal vs. huge unreducible background: Initial state radiation (ISR), Final state radiation (FSR), f r±p± with r ±  p±g (neutral mode): same as above BUT different backgrounds e+e-  wp0 with w  p0g, f r0p0 with r 0  p0g General Comments:  fit of mass spectra are needed;  the background is not fully known (parameters from data);  interference between signal and background;

How to detect these radiative decays - II  a0(980)g  hp0g  K+K-g BR [Achasov, Gubin] ~ 10-6  K0K0g “ ~ 10-8 2 decay modes considered: h  gg [39.4%]  5 photons h  p+p-p0 [22.6%]  2 tracks and 5 photons Combined analysis General comments:  the unreducible background is lower;  the combined analysis allows systematics checks.  f0(600)g  p+p-g (charged mode)  p0p0g (neutral mode) Search for “structures” in the m(p+p-) and m(p0p0) spectra.

How to detect these radiative decays - III How do we extract the signal ? Electric Dipole Transitions (E1):  G(E1)  Eg3 × |Mif(Eg)|2 2. Distortions due to KK thresholds; 3. Mif(Eg) requires a model. Kaon-loop model (by N.N.Achasov): For each scalar meson S: gSpp, gSKK, MS (2) No-Structure model (in collaboration with G.Isidori and L.Maiani): A modified BW + a polynomial continuum; Parameters: gfSg, gSpp, gSKK, MS + polynomial continuum f0,a0 f Kaon-loop No-structure K+ K- p+ p- g

The KLOE experiment (see S.Miscetti talk in plenary session) The KLOE detector: Drift Chamber E.M. Calorimeter Magnetic field = 0.52 T (solenoid) Data taking period Integrated Luminosity (pb-1) 2000 20 p0p0g : PLB537 (2002) 21 hp0g : PLB536 (2002) 209 2001+2002 450 Results presented here on p0p0g, hp0g, p+p-g 2004+2005 (up to date) 1500 ( 2000 end y2005) “final results” next years N(scalars) = N(f) × BR(fg+scalars)  8 ×109 × 10-4 = 8 ×105

The p+p-g analysis - I I - event selection: 2 tracks with qt>45o; missing momentum qpp>45o (Large Angle); Each track is pion-like; “trackmass” compatible with pion; 1 photon matching the missing momentum (“trackmass”) muons pions (Likelihood: Tof and Shower shape) electrons pions, muons W(rad) (Photon matching) p+p-g p+p-p0 Reducible Background rejection:

The p+p-g analysis - II II – The data sample 6.7 ×105 events / 350 pb-1 @ √s = Mf 2.2 ×104 events / 11 pb-1 “off-peak” m(pp) spectra: (blue) “Small angle” qpp<15o; (red) “Large angle” qpp>45o; This analysis m(pp) (MeV) photon efficiency f0(980) region m(pp) (MeV) “Large angle”: clear f0(980) signal

The p+p-g analysis - III (Kaon-loop fit) f0(980) only [p(c2)=5%]; III - Fit to the m(pp) spectrum (491 bins, 1.2 MeV wide, 420 to 1009 MeV) F=ISR (Kuhn-Santamaria, mr, Gr, a , b) +FSR+rp +scalar (Kaon-Loop, mf0 ,g2fKK /4p ,R) +interference Comments: 1. Adding f0(600)  no improvement; 2. f0(980) is narrow (FWHM = 30 MeV) BUT large destructive interference with FSR 3. f0(980) parameters: mf0 (MeV) 980 ÷ 987 g2fKK /4p (GeV2) 2.0 ÷ 3.2 R= g2fKK /g2fp+p- 2.5 ÷ 2.8 R>1 [as in PLB537(2002) 21]

The p+p-g analysis - IV IV – Alternative fit: No Structure. scalar = NS amplitude. f0(980) + polynomial (2nd order): f0(980)+p2 [p(c2)=4%]; Comments: Adding f0(600) no improvement; Parametrization is still under development; 3. Parameters (preliminary values) mf0 (MeV) 970 ÷ 981 gff0g (GeV-1) 1.2 ÷ 2.0 R= g2fKK /g2fp+p- 2.6 ÷ 4.4 R > 1 in both fits; A value for gffg.

The p+p-g analysis - V V - The charge asymmetry: A = (N(q+>90o) – N(q+<90o)) / sum p+p- system: A(ISR) C-odd A(FSR) & A(scalar) C-even Cross-section: |A(tot)|2 = |A(ISR)|2 + |A(FSR)|2 + |A(scalar)|2 + 2Re[A(ISR) A(FSR)] + 2Re[A(ISR) A(scalar)] + 2Re[A(FSR) A(scalar)] (Red) = p+ (Blue) = p- Pion polar angle distributions Pion momentum distributions

The p+p-g analysis - VI VI – Effect of the scalar amplitude on the charge asymmetry: Plot of A in slices of m(pp); Comparison with simulation with and without the scalar amplitude. Circles = data points Triangles = predictions ISR+FSR Squares = predictions +f0(Kaon-Loop fit parameters) Qualitative description of: f0(980) region behaviour; Low mass behaviour. Remarkable result: not a fit but an absolute prediction

The p+p-g analysis - VII VII – Cross section dependence on √s: Absolute prediction based on Kaon-Loop fit parameters Red and Blue = data points Green = KL predictions In 2006 we will increase our “off-peak” statistics to study better the scalar contribution in the p+p-g data.

The p0p0g analysis - I I - event selection: 5 photons with qg>21o ; no tracks; Kinematic fit  energy-momentum conservation; Kinematic fit  p0 masses: choice of the pairing. KLOE PLB537 (2002) 21 New analysis scheme: Removed the cut of events with m(p0g) ~ m(w) wp0 with w  p0g in the sample [ s ~ 0.5 nb ~ 2 s(“signal”) ] 2.Bi-dimensional analysis [ Dalitz-plot m(p0p0) – m(p0g) ] New treatment of systematics [ pairing problem...] Improved VDM parametrization of wp0

The p0p0g analysis - II Dalitz plot components II – The data sample: 1. e+e-  wp0 with w  p0g (√s dependence checked agreement with SND results) √s (MeV) II – The data sample: 400 kevents out of 450 pb-1. Dalitz plot [M(pp) vs. M(pg)] 2. “signal”:f  “scalar”+g  p0p0g √s (MeV)

The p0p0g analysis - III III – The fit of the Dalitz plot (still preliminary results) Residuals vs. DP position Data- fit comparison (on projections) Kaon-loop fit: 1. VDM part still not perfect (see residuals); 2. Scalar part ok BUT f0(600) is still needed [p(c2) ~ 10-4  30% !]; 3. f0(980) parameters agree with p+p-g analysis again R > 1 (gfKK > gfp+p-).

The hp0g analysis - I I – The data samples: out of 400 pb-1 : Statistics of PLB536 (2002) 209 × 20 (h  gg) Improved reducible background subtraction: 2.2 ×104 events [ ½ are signal] Red = signal Other colors= bck M(hp) (MeV) (h  p+p-p0) almost “background Free”  4100 events [ bck < 3%] Full points = “20 pb-1” data Empty points = “400 pb-1” data (norm. to the same luminosity)

The hp0g analysis - II II – The combined fit: scalar amplitude with kaon-loop. Red points = data Blue hist = fit M(hp) (MeV) M(hp) (MeV) Results (preliminary statistical errors only): a good combined fit is obtained [BR(gg)/BR(p+p-p0) fixed to PDG value] the spectrum is entirely due to the scalar amplitude; M(a0) = 987.0 ± 0.4 MeV g2aKK/4p = 0.434 ± 0.007 GeV2 R = g2aKK/g2hp = 0.787 ± 0.006

Summarizing: The KLOE scalar analysis is not yet completed. However: (1) The Kaon-Loop frame describes our entire data-set. Emerging picture: f0(980) strongly coupled to kaons g2fKK ~ 2 ÷ 3 GeV2; R = g2fKK/g2fp+p- ~ 2 ÷ 4. a0(980) is less strongly coupled to kaons g2aKK ~ 0.4 GeV2 R = g2aKK/g2ahp ~ 0.8 f0(600) required in the p0p0 channel (2) No Structure analysis is promising: still theoretical effort required BUT first results “confirm” the kaon-loop picture: f0 and a0 have large |ss> contents.

KLOE perspectives on scalar mesons Conclude analysis on 2001-2002 data sample (~ 400 pb-1) for f0(980) (neutral and charge final states) and a0(980) (KL and NS fits).  With 2000 pb-1: improvement expected for f0 → p+p- Better precision on the couplings and on the asymmetry. Combined fit p+p- AND p0p0 ?  Study of the √s-dependence of the cross-section.  Search for f0, a0  KK ?

Spare Slides Hadron 2005

Slightly lower BR obtained (preliminary): (hgg) BR(fhp0g)=(7.49±0.10)×10-5 [it was: (8.5  0.5stat  0.6syst)x10-5] (hp+p-p0) BR(fhp0g)=(7.45±0.19)×10-5 [it was: (8.0  0.6stat  0.5syst)x10-5]

Theory: KL (KL) by N.N.Achasov; where: g(m) = kaon-loop function d(m) = phase shift (based on pp scattering data) Df(m) = f0 propagator (finite width corrections) If only one meson (no s included): 3 free parameters: mf, gfKK, gfpp

Theory: NS (NS) after several discussions with G.Isidori, L.Maiani and (recently) S.Pacetti; where the propagator (Flatte’ revised) is: Nota che 3 dei 7 sono anche in KL quindi questo e’ piu’ generale e molto piu’ libero. with couplings 7 free parameters: mf, gffg, gfKK, gfpp,a0, a1, b1