Floating behaviour of molten copper and matte droplets on slag surface has beeninvestigated using X-ray radio photograph technique in order to clarify the mechanism of the mechanical copper loss. It was experimently found that copper matte droplets of which density is higher than that of slag stably floated on surface of the slag by a surporting force of surface tension of the slag. About 1×10-3kg of matte droplets floated on an fayalite slag and the profile of it was not spherical but flat because of low interfacial tension between the matte and slag.The quantitative analysis of floating behaviour was carried out. The profile and amount of floating droplets were calculated using interfacial properties such as surface tensions of the slag and matte and interfacial tensions between the slag and matte. The floating model in this study could explain the floating behaviour of copper matte droplet on an fayalite slag and copper droplet on a sodium silicate slag.
The He-He and He-Li+ potentials are investigated by means of extensive STO SCF and configuration interaction calculations. The calculated molecular constants r m and εm are 5·61 a.u. and 9·81 K for He-He which are very close to experiment. The calculated potential curve for He-Li+ is also considered to be reliable; r m=3·63 a.u. and εm=0·072 eV.
To determine the effects of the vacuum environment on fatigue crack propagations, K-decreasing tests with compact tension (CT) specimens were conducted in air and vacuum environments. The test data clarified the fatigue crack propagation rate (da/dN) and threshold stress intensity factor range (ΔK_
) for both environments. The da/dN becomes lower and ΔK_
becomes larger with decreasing vacuum pressure. However, this tendency was not explained only by the crack closure. Based on fracture surface observations, the crack propagation properties in vacuum were compared with those in air. A few micrometer size granular region was observed on the fracture surface only in the high vacuum (〜10^<-6>Pa) and ultra high vacuum (〜10^<-7>Pa), but not in the air and medium vacuum (〜10^<-1>Pa). The high vacuum environment is one of the necessary conditions for the formation of the granular region, and the fraction of surface coverage of fracture surfaces probably relates to the phenomenon. The formation of the granular region represents the difference of the crack propagation mechanism between in vacuum and in air environments, and the tendency observed in da/dN and ΔK_
The effect of anode bismuth level and of concentrations of dissolved oxygen and bismuth ions in copper electrolyte were investigated to understand the formation mechanism of anode slime layer and its influence toward passivation behaviour. The metallic granular forms of bismuth were found in the anode at the lower bismuth content. However, the higher bismuth content of anode exhibited a hairy or meshes-like shape. The bismuth dissolved preferentially during electrolysis and the anode surface after electrolysis became excessively uneven. The slimes were classified into the adhering and the falling types. The amount of falling slimes was about 90% of total slime formed and it mainly consisted of copper powder mechanically separated from anode. On the other hand, the adhering slimes consisted of copper powder, Bi2O3 and Bi2 (SO4)3. The copper powder in the adhering slimes was the fine particles formed by the disproportionation reaction of Cu+ ions. Bi2(SO4)3 in the slimes were precipitated by the supersaturation of dissolved Bi3+ ions and observed in the inside of grain boundary.Under the electrolytic condition of high current density and in the case of the addition of Bi3+ ions into electrolyte, layer of needle-like precipitates of Bi2 (SO4)3 was observed on the adhering slime. This slime layer caused the anode to passivate sensitively.When Bi, As and Sb impurities were coexisted in the copper anode, the composition of the resulting anode slime was Bi2O3, BiAsO4 and SbAsO4.
This paper deals with the electronic structure of the Zn2 and Zn3 clusters. Independent ab initio SCF calculations are performed for the ground and excited states of Zn2, and the SCF MOs are determined for respective states. By using these MOs, CI calculations taking account of the correlation effects among 4s- and 4p-like electrons are carried out. This is an improvement over previous CI calculations, which used the occupied and virtual (or improved virtual) MOs computed for the ground state, and thus disregarded the reorganization effect. The calculated excitation energies for Zn2 are 0.8–1.3 eV smaller than experiment. The inclusion of the correlation between the inner (3d-like) and valence (4s- and 4p-like) shells ameliorates the result only slightly (by 0.03–0.05 eV). The remaining discrepancy is attributed to relatively poor description of the ground state as compared with that of the excited states. Similar calculations are performed for the electronic structure of Zn3. Calculated excitation energies are compared with the absorption spectra of Zn aggregates (size unknown) in an argon matrix.
