Ni and Zn doping of Cu sites in superconducting Nd2-zCezCuO4, La2-βSrβCuO4, Bi2Sr2CaCu2O8, Bi1·8Pb0·2Sr2Ca2Cu3O10, YBa2Cu3O7, La0·6Ca0·4Ba1·35La0·65Cu3Ox and YBa2Cu4O8

A unified Cooper pair-breaking picture is presented for the observed degradation of the superconducting critical temperatures by Cu-site Zn and Ni impurities in Nd2-zCezCuO4, La2-βSrβCuO4, Bi2Sr2CaCu2O8, Bi1·8Pb0·2Sr2Ca2-Cu3O10, Lao.0·6Ca0·4.4Ba1·35La0.65Cu3O x (with x ≈ 7), YBa2Cu3O7 and YBa2Cu4O8. Independent of any specific model, the data alone require physics outside the cuprate planes, and hence no two-dimensional cuprate-plane model can consistently explain all the Zn and Ni pair-breaking data. In Nd2-zCezCuO4, Ni and Zn behave as in a BCS superconductor. Although the Ni- and Zn-doping data for YBa2Cu3O7 have been cited as providing the so-called ‘smoking gun’ conclusive proof of the spin-fluctuation model, we conclude that the underlying assumptions of that data-analysis, that Ni and Zn occupy the same sites and that Ni is a weaker pair-breaker than Zn, are invalid: we argue that the exchange scattering of Cooper-pairs in YBa2Cu3Ox is inoperative. Thus, except for Nd2-zCezCuO4 whose charge-reservoirs are adjacent to its cuprate-planes and hence are uniquely within the range of exchange scattering by cuprate-plane Cu-site Ni ions, the exchange scattering caused by the difference between cuprate-plane Ni and Zn impurities is inoperative in these high-temperature superconductors—a fact difficult and perhaps impossible to reconcile with any cuprate-plane model of superconductivity that also features a Meissner effect and Cooper pairing. We interpret the inoperative exchange scattering as evidence that the primary superconducting condensate is outside the range of the exchange interaction-and hence in the charge reservoirs, rather than in the cuprate-planes. The trends in observed pair-breaking critical compositions uc define approximately exponential functions of the distance d between the impurity sites and the charge-reservoir dopant-oxygen-also indicative of charge reservoir superconductivity. The critical tempertures Tc increase (approximately linearly) with d (again, consistent with charge-reservoir superconductivity), suggesting that higher critical temperatures can be achieved in materials with increased separation between the cuprate planes and the charge reservoirs. The cuprate planes appear to be mechanically favourable, but electronically unfavourable to superconductivity. A consistent explanation of all the Ni- and Zn-doping data is obtained (i) if the superconductivity originates in the charge-resevoir or dopant-oxygen regions of each unit cell (not in the cuprate planes), (ii) if the short-ranged exchange scattering by magnetic Ni readily breaks Cooper-pairs whose holes are located on nearest-neighbour oxygen ions only, (iii) if a longer-ranged interaction is responsible for the degradation of T c by Zn and by Ni distant from the superconducting condensate, and (iv) if polarization fluctuations are responsible for the Cooper pairing. The pair-breaking data contradict not only the spin-fluctuation d-wave pairing model, but also all models with the superconductivity originating in cuprate planes.