The other task was even more critical. In the 1970s hitherto unknown toxicity of vinyl chloride at extremely low concentrations was discovered and the last traces of this gas had to be removed from the finished PVC polymer particles, well below the limits set by the health authorities. Since the gas was dissolved in the polymer and removal was diffusion-controlled, elevated temperatures were required, yet PVC tends to degrade at high temperatures. The problem was solved by an elegant continuous process of high speed stripping gas from the solids suspended in water cascading down a counter-currently operating column, which reduced the gas content to fractions of parts per million without thermal degradation of the polymer. A full scale plant unit was installed (Fig. 30) and the process was subsequently adopted by overseas companies. The sudden discovery of vinyl chloride toxicity was a world-wide problem; everywhere industry had to face the task of controlling emissions to fractions of parts per million. Throughout this world-wide safety campaign of improved process control the key analytical tool was the ICI/McWilliam flame ionisation detector; by then automation of the instrument with computer control had progressed to the stage that well over one thousand analyses were done each day on the plant to ensure absolutely safe working conditions.

30 Continuous vinyl chloride stripping form PVC. Full scale plant at Botany.

Ethylene oxide and derivatives

While the oxychlorination process was primarily an achievement of Australian research and development, and polythene in essence was technology transfer, ethylene oxide and its derivatives were a mixture of the two. In the UK, ICI and in Australia, Timbrol had long made ethylene-glycol, an important chemical intermediate best known to the wider public for only one of its uses, as a car radiator fluid. The ICI process based on epichlorhydrin did not suit ICI Australia's constellation; chlorine prices were rising and the process appeared obsolescent. Research and development teams, therefore, evaluated alternative processes and, eventually, secured an early licence from an emergent American company, R. Landau's Scientific Design. This was a matter of prudent buying, as the process was still at an early development stage, when entry terms were favourable but entailed an element of risk, and hence confidence in one's own evaluation.

Australian Academy of Technological Sciences and Engineering