Test of (N)over-barN potential models: Isospin relations in (p)over-bard annihilations at rest and the search for quasinuclear bound states

We have determined branching ratios for antiproton annihilations at rest on protons or neutrons in liquid deuterium which we compare to frequencies of isospin-related processes in antiproton-proton annihilations. Using the an nihilation rates into pi (0)pi (0) and pi (-)pi (0) where the annihilation took place on the proton or neutron, respectively, we discuss the fraction of S-wave and P-wave annihilation in liquid D-2. The frequencies for pi (-) omega and rho (-)pi (0), and pi (-)eta and pi (-)eta' and the corresponding frequencies for (p) over barp annihilations are used to determine isoscala r and isovector contributions to the protonium wave function. The isospin d ecomposition of the annihilating (p) over barp system in the S-3(1) or S-1( 0) state is consistent with both, pure (p) over barp initial wave function and with the predictions of (N) over barN potential models. For the P-3(0) state of the (p) over barp atom we find consistency with a pure (p) over ba rp system at annihilation while (N) over barN potential models predict larg e (n) over barn contributions. We observe rho-omega interference in (p) ove r barp --> pi (+)pi (-)eta and pi (+)pi (-)pi (0) annihilation which we com pare to rho-omega interference in e(+)e(-) annihilation. The interference p atterns show striking similarities due to similarity of the phase in all 3 data sets demonstrates that isovector and isoscalar parts of the protonium ((p) over barp atomic) wave function are relatively real, again in conflict with (N) over barN potential models. The annihilation rate for (p) over ba rd --> K-K(0)p confirms the dominance of the isovector contribution to (N) over barN --> K (K) over bar annihilations. No complications due to initial state interactions are required by the data. Furthermore, we searched for narrow quasinuclear bound states close to the (N) over barN threshold, also predicted by (N) over barN potential models, but with negative outcome. We conclude that (N) over barN potential models are not suited to provide ins ight into the dynamics of the annihilation process.