The Vanillin project was undertaken by AECI as it was believed that with the raw materials available from SASOL i.e. cresols, it would be possible to develop a process that would be globally competitive. A process for preparing both vanillin and ethyl vanillin was developed by AECI’s Research and Development Department using a mixed m- and p-cresol stream as feed. The displacement of bromide by methoxide is very successful when using a dimethyl formamide (DMF)/alcohol solvent system and excellent conversions (>95%) and selectivities (>95%) may be obtained. However, the use of DMF, which poses a serious chronic health risk, is unacceptable in this process since the products are intended for use as food and flavouring chemicals. In view of the above the main objectives of this study were: · To find a suitable alternative solvent system, which could produce comparable results while still being economically viable; · To develop an appropriate experimental protocol in the laboratory based on the alternative solvent system; · To determine the important reaction variables by conducting statistically designed experiments; · To optimise the reaction to produce a reproducible and robust experimental protocol; and · To test the reaction thoroughly at bench-scale level and to obtain experimental data for scale-up to pilot plant The most promising alternative solvent system was a methanol/methyl acetate mixture, which produced satisfactory results in the preliminary assessment (conversion of 98.3% and selectivity of 92.0%). DMA and acetonitrile also produced promising results but were not considered for further investigation because of toxicity and cost issues. A set of statistically designed experiments was carried out on the methanol/methyl acetate solvent system where four variables were tested i.e., substrate concentration, temperature, catalyst loading, and methanol to methyl acetate volume ratio. The experimentally determined response surface model showed that the most important variable was catalyst loading (63.2%) for conversion. With respect to selectivity, the most important variables were catalyst loading (31.9%) and methanol to methyl acetate ratio (33.1%). The optimum reaction conditions were as follows: · Temperature: 120°C · Methanol:methyl acetate: 15:1 vol/vol · Catalyst loading: 8 mol % to substrate · Substrate concentration: 22 %m/m on solvent · Catalyst: Copper(I) bromide · Sodium Equivalents: 2.7 wrt substrate · Time: 3 hours The optimum conditions were tested for reproducibility in a 1 Labmax pressure reactor. Replicated reactions, two at a 10% and two at a 20% substrate concentration gave conversions and selectivities all greater than 90%. Although the reaction mixture was a slurry at these concentrations, the reactions were very fast and virtually complete within the first hour (~95% conversion). Initial scale-up studies were conducted in an 8 Parr reactor where five reactions were carried out using the optimum conditions described above. The conversion of substrate and vanillin selectivity was consistently high and compared favourably to the Labmax reactions. The average conversion was 97.3% (96.3 to 98.5%) at an average selectivity of 98.2% (97.4 to 99.1%). A study of the reaction kinetics confirmed that the reaction was first order with respect to the substrate as a plot of substrate concentration versus reaction rate gave a straight line. The rate constant was calculated as 1.1096 k(h-1). The reaction mechanism proposed for the copper assisted nucleophilic aromatic substitution involves the formation of an adduct between sodium methoxide, methyl acetate and copper(I) bromide. The formation of a transient intermediate with the substrate allows intramolecular delivery of the methoxide ion to the aryl moiety through a CuI – CuIII type cycle.
展开▼
