1. Evidence of ,  and  meson mass modification in nuclear medium measured in 12 GeV p+A reaction at KEK-PS E325 RIKEN, Kyoto Univ. a, KEK b, CNS Univ. of Tokyo c, ICEPP Univ. of Tokyo d, Tohoku Univ. e Ryotaro Muto, KEK, Japan J. Chiba b, H. En’yo, Y. Fukao a, H. Funahashi a, H. Hamagaki c, M. Ieiri b, M. Ishino d, H. Kanda e, M. Kitaguchi a, S. Mihara d, K. Miwa a, T. Miyashita a, T. Murakami a, T. Nakura a, M. Naruki, K. Ozawa c, F. Sakuma a, O. Sasaki b, M. Sekimoto b, T. Tabaru, K.H. Tanaka b, M. Togawa a, S. Yamada a, S. Yokkaichi, Y. Yoshimura a (KEK-PS E325 Collaboration) Introduction Result of   e + e - analysis Result of   e + e - analysis 

2.    dropping mass Brown-Rho scaling (’91) – m*/m = 0.8 at  =   QCD Sum Rule by Hatsuda & Lee (’92) –m*/m = 1 - 0.16   for  –m*/m = 1 - 0.03  0 for  Lattice Calc. by Muroya, Nakamura & Nonaka(’03) width broadening (at  0 ) Klingl, Kaiser, Weise (’97-8)    ~  for  Rapp & Wambach (’99)  :       ~2 Oset & Ramos (’01) :    22MeV Cabrera & Vicente (’03) :    33MeV Vector meson mass at finite density Hatsuda & Lee PRC46(1992)R34

3. E325 experiment Slowly moving  ( p lab ~2GeV/c ）  Large acceptance spectrometer Primary proton beam ~10 9 ppp Thin targets: 0.2%/0.05% (C/Cu) radiation length: 0.4%/0.5%(C/Cu) Invariant Mass of e + e -, K + K - in 12GeV p + A   + X History ’93 proposed ’96 construction start NIM, A457, 581 (2001) NIM, A516, 390 (2004) ’97 first K + K - data ’98 first e + e - data PRL, 86, 5019 (2001) ’99~’02 x100 statistics in e + e -  PRL 96, 092301 (’06)   ee  PRL 98, 042501 (‘07)  : PRC, 75, 025201 (‘06) x6 statistics in K + K -   KK: PRL 98, 152302 (‘07) beam B

4. Invariant mass spectra of e + e- CCu We examine how well the data are reproduced with known hadronic sources & combinatorial background e+e-e+e- e+e-e+e- e+e-e+e- e+e-e+e- Mass resolution 8 MeV/c 2 for  /  11 MeV/c 2 for 

5. Invariant mass spectra of e + e - C Cu the excess over the known hadronic sources on the low mass side of  peak has been observed. e+e-e+e- e+e-e+e- e+e-e+e- e+e-e+e-  /dof    /dof  

6. 0.0±0.04(stat.)±0.3(sys.) Invariant mass spectra of e + e - (  /  region, background subtracted) N  N  =0.0±0.02(stat.)±0.2(sys.) CCu  the excesses are mainly due to modified  mesons Lifetime of  : c  ~1.3fm  excess can be observed in both C & Cu  /  ratio is consistent with zero

7. Invariant mass spectra of e + e- CCu We examine how well the data are reproduced with known hadronic sources & quadratic background e+e-e+e- e+e-e+e- e+e-e+e- e+e-e+e- Mass resolution 8 MeV/c 2 for  /  11 MeV/c 2 for 

8. Invariant mass spectra of   e + e - CCu  Some hints on the low mass side of the  meson peak in Cu data Slowly moving  has larger probability to decay inside nucleus  We divided the data by  ( = p/m) [GeV/c 2 ] Background : Quadratic curve

9. Invariant mass spectra of   e + e -  <1.25 (Slow)1.25<  <1.75 1.75<  (Fast) Large Nucleus Small Nucleus Rejected at 99% confidence level

10. // m(  )/m(0) (mass dropping) 1 - k 1  /  (  /  0 ) (Hatsuda & Lee) 1 - k 1  (  /  0 ) (Hatsuda & Lee)  (  )/  (  ) (width broadening) 1 1 + k 2 (  /  0 ) generation pointsurfaceuniform     0.710±0.0210.937±0.049 momentum dist.measured density distributionWoods-Saxon, radius: C:2.3fm/Cu:4.1fm Model calc. including mass modification   e+e+ e-e- We attempt to reproduce the observed spectra with model calculation including mass modification

11. Fit results of model calculation for  /  the excesses for both C and Cu are well reproduced by the model including the 9% mass decrease at  0. k 1 = 0.092 (m*/m = 1 - 0.092  /  0 ) CCu [GeV/c 2 ]

12. Fit result of model calculation for   e + e - k 1 =0.034, k 2 =2.6  <1.25 (Slow)1.25<  <1.75 1.75<  (Fast) Large Nucleus Small Nucleus

13. Comparison with the theoretical prediction prediction (Hatsuda,Lee) 0 0.5 1  0 1 0.9 0.8 0.7 fit result  fit result  m(  )/m(0) syst. error is not included m*/m = 1 – k 1  /  0,  */  = 1 + k 2  /  0 The data were well reproduced with the model; m  /  decreases by 9.2%, m  decreases by 3.4% and   increases by 3.6 at  0 Best Fit Values  k1k1 9.2 ± 0.2%3.4 +0.6 -0.7 % k2k2 0 (fixed)2.6 +1.8 -1.2  

14. Summary We have observed significant excesses over the known hadronic sources both in  /   e + e - and   e + e - distributions in 12 GeV p+A reactions The excesses were well reproduced by the model calculation including mass modification, –  /  : 9% mass decrease –  : 3% mass decrease and width broadening by a factor of 3.6 at  0. From   K + K - analyses we have obtained limits on the in- medium decay width broadenings for both   e + e - and   K + K - decay channels  See Poster by Fuminori Sakuma (Tomorrow) We are planning next experiment at J-PARC which will achieve statistics 100 times as large as the present experiment  See Poster by Satoshi Yokkaichi (Tomorrow)