Z. Phys. C - Particles and Fields 36, 455-460 (1987) Zeitschnft P a r t i d e s for Physik C and Fields 9 Springer-Verlag 1987 Strange quarks in the nucleon and neutron electric dipole moment V.M. Khatsymovsky, I.B. Khriplovich, A.R. Zhitnitsky Institute of Nuclear Physics, SU-630090 Novosibirsk, USSR Received 29 April 1987 Abstract. Arguments in favour of a large value of the matrix element (N[ gs IN) based on semiphenomeno- logical data and low-energy theorems are given. As an application of these considerations, we study the contribution of strange quarks to the neutron electric dipole moment. 1 Introduction It is widely held that the admixture of the pairs of strange quarks gs in nucleons is small. Deep inelastic scattering data seem to support this view. It should be noted, however, that this argument is valid for s quarks in the vector channel only. For the scalar channel, there is no compelling evidence that (gl gs IN) is much more smaller than corresponding tTu and dd matrix elements. Recently [1], arguments were given indicating that, indeed, the matrix element (NI gs IN) is not small. This conclusion seems to us to be very interesting in itself. In this work we present additional arguments based on low-energy theorems and some semipheno- menological data in favour of a large nucleon expecta- tion value for gs. A similar approach applied to calcu- lating baryon matrix elements of (pseudo-)scalar op- erators allows important conclusions to be drawn re- garding interaction between Higgs particles and nu- cleons and also on CP-odd effects in the non-strange baryon system. In particular, prediction of the Weinberg model of CP violation for the nucleon electric dipole moment is definitely contradicted by experimental data. 2 Quark scalar expectation value Let us start by calculating quark scalar matrix ele- ments over the nucleon, assuming an octet nature of SU(3) symmetry breaking. This calculation differs from that in [1, Sect. 2], in technical details only. However, we present it here for completeness. Averaging the results of various fits to the data on ~zN scattering presented in [2] leads to the follow- ing answer for the so-called a term m,+md <PI Ou+dd [P) =63__ 12 MeV. 2 0) (Here and below we omit kinematical structures like 15 p in expressions for matrix elements.) Taking the values of quark masses to be m, =4 MeV, md=7 MeV, from (1) (PI t7u IP) + (P[ dd IP) = 11.5 • 2. (2) Further, assuming octet-type SU(3) breaking to be responsible for mass splittings in the baryon octet, we find (P[ ~7u-dd [P) = mz-rns =0.9, (3) ms (Pl~u+dd-2gs]P)=3mz-mA 4.3. (4) ms Here mz, mz, ma are the masses of ~, S, and A hyper- ons respectively; the s-quark mass value is taken to be 140 MeV. The values (3), (4) are quite reasonable: the former is close to the difference of the number of u and d quarks in a proton, and the latter is close to the total number of valence quarks u and d in a nucleon (specified below). Using (2)-(4) one obtains (P[ tiu lP) =6.2, (5) (P[ dd ]P) = 5.3, (6) (P[gs IP)= 3.6. (7)