The increasing concerns for using phthalates have led to the discovery of the safer alternative Grindsted® SOFT-N-SAFE (SNS) by Dansico A/S (now known as DuPont Nutrition Biosciences Aps). The main component of SNS is based on acetylated glycerol monostearate, originating from hydrogenated castor oil. SNS is a 1-1 replacement alternative to the phthalate di(2-ethylhexyl)phthalate (DEHP). This alternative is, however, too expensive to be an actual alternative to phthalates, because of the expensive starting material castor oil and the use of acetic anhydride as acetylation reagent. Besides this, the starting material is not very accessible, meaning that even if the production price of SNS was comparable to the production price of phthalates, SNS could not be produced in large enough quantities to replace the use DEHP. An alternative to SNS, the SNS-analogue (SNS-A), was suggested from glycerol monooleate originating from sunflower oil. Sunflower oil is less expensive and more accessible compared to castor oil and the SNS-A has been tested to have the same plasticizing effect and non-toxic effects as SNS. However, a sustainable and cheap way of synthesizing SNS-A has not been developed. The aim of this project was to find an alternative, sustainable and cheap synthetic pathway for the SNS-A in collaboration with DuPont Nutrition Biosciences Aps, Danish National Food institute (DTU-Food), Department of Engineering at Aarhus University (AU) and other group members from the Centre for Catalysis and Sustainable Chemistry at Department of Chemistry at Department of Chemistry at Technical University of Denmark (DTU-Chemistry). A three step synthetic pathway consisting of epoxidation, hydrogenation and acetylation of glycerol monooleate was early in the project suggested as the best procedure, and the different parts were divided between the parties in the project. This PhD project had three parts. The first part was to find an ionic liquid (IL) which could separate out the product from the reaction mixture and that could function as reaction media for one or more steps in the process. The second step was developing an epoxidation procedure and the third part was developing a hydrogenation method. The idea with using an IL as a reaction media and for product separation was also to a keep any water from the glycerol backbone of the material to avoid hydrolysis. This could be done by choosing an IL non-miscible with water, like the ones having a bis(trifluoromethylsulfonyl)amide anion. From the ILs synthesized or purchased it was not possible to find an IL which could separate out SNS. Several of the tested ILs could, however, be used as separation media after epoxidation or hydrogenation. Only one IL was miscible with the starting material and the epoxide but not the product from the hydrogenation, and all ILs where separation was possible after epoxidation step was non-miscible with the starting material. It was therefore not possible, from the ILs described here, to have a homogeneous reaction media for epoxidation with separation possible after this step. The second part of this study, the work with developing an epoxidation procedure, can be divided into two parts. The first epoxidation method was catalytic epoxidation in ILs where product separation was possible using a heterepoly acid based catalyst and hydrogen peroxide as oxidant. The IL giving the best result was N-butyl-N-methyl imidazolium bis(trifluoromethyl-sulfonyl)amide (abbreviated [BMIm][Tf2N]) and the epoxidation reaction was further investigated and optimized using this IL. Applying different reaction conditions resulted twice in a good epoxidation results, with conversions of 67 % and 70 % respectively. This, however, was not a sufficiently good result for up scaling of the process and reproducibility proved difficult. Accordingly, this method was abandoned. The epoxidation procedure was also tested with peracetic acid with acetic acid as both solvent and by-product from the reaction. Both substrate and product are soluble in the reaction media, but as acetic acid was to be added for the final acetylation step, this did not require a solvent removal step. Already in the initial experiments this system performed very promising, and the reaction was optimized with respect to the temperature, addition time and rate of peracetic acid and substrate concentration i.e. amount of solvent. Furthermore, the effect of water concentration was also investigated. After optimization, the result was satisfying as no large excess of peracetic acid needed to be used. The conversion rate was high at low temperatures, giving a short reaction time, and the amount of by-product low. The reproducibility of this reaction was high and it was tested many times in up to 2 L scale with the same satisfying result. Two methods for hydrogenation of the epoxide to the mono-hydroxy compound examined, a catalytic transfer hydrogenation (CTH) using a hydrogen donor and the use of molecular hydrogen gas. CTH was conducted with selected substrates or with the epoxide of interest in this project. Several solvents, many of them ILs, were tested with three different hydrogen donors and three different catalysts, known to be efficient in CTH reactions. However, no satisfying results were obtained using CTH as a hydrogenation method. Using molecular hydrogen gas for hydrogenation also resulted in problems. The hydrogenation of the epoxide obtained from epoxidation with peracetic acid gave mainly the saturated monoglyceride instead of the mono-hydroxy compound as intended. Several heterogeneous metal catalysts and reaction conditions were tested, but it was not possible to find a suitable method for the hydrogenation. Overall, this PhD study has established a potentially industrially viable epoxidation protocol as part of a new reaction pathway for the synthesis of SNS-A.
Main Research Area:
Wiebe, Lars, Riisager, Anders
Technical University of Denmark, Department of Chemical Engineering, 2013