Optimium conditions for the production of sub-micron cobalt power

Abstract

Cobalt powder is a grey metallic powder that is produced by the thermal decomposition and reduction of a cobalt compound. The challenge faced by Shu Powders Africa was that sub-micron cobalt powder had never been produced in a two-step furnace by any manufacturer in the cobalt powder industry. Hence there was no prior information to guide this type of processing. Therefore this research set out to investigate the production of sub-micron cobalt powder through a two-step furnace to determine the optimum parameters for this process. For the company to remain competitive, it was imperative to begin producing sub-micron cobalt powder. Sub-micron cobalt powder is much more valuable and profitable to produce. The second production line would be operational due to the production of sub-micron cobalt powder hence creating job opportunities for the local community. Sub-micron cobalt powder shares the same chemical composition and physical characteristics as cobalt powder. The only differences are particle size (0.60 - 0.90 µm), oxygen content (0.30 - 0.80%) and the microscopic structure which is the particle size distribution d90 (7 - 10 µm). The approach taken was to understand the variables that had a large effect on the powder. The effects needed to be established by determining how it impacted on the quality of the powder which is pertinent to making sub-micron cobalt powder. Due to the experience in producing cobalt powder, variables that had a large effect on normal cobalt powder production were assumed to be the same variables that would impact the production of sub-micron cobalt powder. Some of these effects were also confirmed by literature. A strategy of statistical design of experiments was used to evaluate the conditions for sub-micron cobalt powder production. Design of experiments assisted in planning the experimental design matrices for both experiments. For the furnace experimentation a 24 factor design was selected. For the jet mill experimentation a 23 factor design was selected. Response surface methodology was used to determine optimum ranges of the variables at various process conditions. The central composite rotatable design laid out the design in which the variables interacted with one another at different process conditions. Evaluation of results was based on the generated model. Models such as the 3D surface model, cubic model and the contour model were generated to graphically illustrate the effects that the variables have on the response. Analysis of furnace data indicated that the optimal response was achieved at a temperature range (445 - 460)°C, hydrogen gas range (225 - 250) Nm3/h, belt speed (80 - 90) mm/min, and carbon dioxide gas range (80 - 90) Nm3/h. Analysis of the jet mill experimental data indicated that the optimal response particle size distribution, was achieved at a classifier speed range of (5500 - 6000) rpm, AFG grinding bin range (30 - 35) kgs and grinding gas pressure of (4.0 - 4.5) bar. The study confirms the efficiency of a two-step furnace to produce sub-micron cobalt powder at high volumes. The advantage of the two-step furnace was the increased throughput of 2.3-2.7 tons/day whilst in industry furnace throughputs are 1.3-1.6 tons/day. This represented a 60% increase in productivity over conventional furnaces. The response surface methodology also proved to be a suitable technique for process optimization.