Effect of oxygen on the surface tension of liquid copper
- PDF / 843,480 Bytes
- 9 Pages / 603.28 x 783.28 pts Page_size
- 73 Downloads / 221 Views
THEeffects of surface-active species on the surface tensions of liquid metals are important in numerous industrial applications. Partially due to the difficulty of their determination, there is a real dearth offeliable experimental data. The sessile drop technique was selected to investigate the effect of oxygen on the surface tension of liquid copper at 1108 ~ A similar study was conducted by Bernard and Lupis I on the Ag-O system at the same temperature (1108 ~ Their theoretical conclusions did not agree with experimental results in the literature on the Cu-O system. 2~ The present contribution resolves this discrepancy. APPARATUS A sketch of the apparatus is shown in Fig. 1. The equipment was a modified version of that previously used by Bernard? The furnace was of the globar type with a reaction chamber constituted by a McDanel tube of 99.8 pct A1203. Inside the chamber, a half-cylindrical tray of 99.8 pet A1203, cut from a tube of 3.5 cm diam, supported the Lucalox substrates used in this investigation. The temperature of the drop was measured by a Pt/Pt-10 pet Rh thermocouple located under the drop's substrate. It was repeatedly calibrated against the melting points of copper, gold and silver and exhibited an accuracy of 0.5 ~ Commercial purity gases were used and purified by standard techniques 6 to eliminate reported contaminants (essentially H2, H20 , CO, CO 2 and O2). Oxygen potentials ranging from 5 • 10 -5 atm to 10 -~6 atm, at 1108 ~ were obtained by means of CO-CO2 mixtures. All gases were metered through capillary flowmeters before entering the reaction chamber. The precision was of the order of 1 pct, even for flowrates as small as 10 cm3/min. Two oxygen probes of calcia-stabilized zirconia were
used. The first probe was placed immediately after the reaction chamber to continuously monitor the oxygen potential during an experiment. The second was mounted directly at the outlet of the gas trains. Its purpose was to verify the composition of the gas flow prior to its introduction in the reaction chamber. Both probes were calibrated against argon-oxygen mixtures of 0.01 pet and 1 pet 02. In this composition range, the agreement between measured and calculated oxygen potentials was within 1 pct. The oxygen potential corresponding to a CO-CO2 mixture could also be measured through knowledge of the flowrates and of the concentrations of CO and CO 2. Agreement between the oxygen potentials resulting from the two methods could only be obtained at very low flowrates. This is probably due to the fact that thermal and thermodynamic equilibria could not be achieved within the small internal volumes of the probes. As a consequence, the following procedure was adopted. The gas mixture was allowed to flow in the probe for a few minutes, the flow was then interrupted and the oxygen potential measured. The uncertainty on log Po2 was estimated at 5 pct. The sessile drop was viewed through two optically flat windows of fused quartz (surfaces parallel to 30 s of arc). They were mounted on the female joints of the reacti
Data Loading...
