Acid catalysis for conversion of carbohydrates to sulphide derivatives

The chemical synthesis of methionine at industrial scale was attained in Germany in 1948 through a process involving acrolein, methanethiol and hydrogen cyanide, among others, to yield the racemic mixture of D- and L-methionine (D,L-Met). Since then, this process and variations thereof, among them those that synthesize methionine hydroxy analogues (MHAs), have been used to manufacture supplements for animal feeds that enable to produce animal protein for human consumption while minimizing livestock waste. Due to their importance to mitigate human hunger, close to 1.6 million tonnes of D,L-Met and MHAs are produced every year by said chemical methods. However, feedstocks such as acrolein and hydrogen cyanide are (i) highly toxic and (ii) obtained from non-renewable petroleum derivatives, making the production of D,L-Met and MHAs less sustainable than desired.

Among the different synthetic approaches proposed to mitigate the drawbacks, the chemical conversion of carbohydrates to MHAs appears as a holistic approach that can solve the problems while using renewable carbohydrates. Unfortunately, the chemical conversion of carbohydrates to MHAs has been scarcely studied. The best results of the state of the art for batch reactions report the conversion of erythrulose to MHAs in yields of 29% (selectivity 36%) and 24% (selectivity 31%) in four hour reactions at 373 K, when catalyzed by post-synthetic Sn-MFI and Sn-BEA, respectively. However, there is no understanding of the fundamental reasons limiting reaction’s selectivity. The reaction pathways of the transformation of carbohydrates to MHAs are unknown as well as the effects of the presence of thiols in the catalyst’s reactivity and stability.

This doctoral thesis investigated the acid catalyzed conversion of tetroses in the presence of butanethiol as a model thiol molecule. This study elucidated the main reaction pathways that tetroses undergo in the presence of a thiol and homogeneous acid catalysts. A screening of homogeneous catalysts established that Brønsted acids were very selective to catalyze thioacetalization side-reactions while Lewis acids were more selective towards the 1,2-hydride shift reaction required to produce MHAs from tetroses, especially tin, tungsten and molybdenum chlorides. A kinetic profile and in-situ NMR studies allowed establishing that the 1,2-hydride shift was the rate determining step of the reaction cascade. Therefore, increasing the reaction’s selectivity towards MHAs demanded to establish strategies to either (i) inhibit the thioacetalization reactions or (ii) minimize the formation of thioacetals by selecting the reaction parameters so that thioacetals formation was thermodynamically less favorable. Based on the reaction pathways established for the transformation, we rationalized and proved that the addition of either potassium hydroxide, water or methanol to the reaction mixture were effective to minimize the formation of thioacetals and increase the selectivity for MHAs.

The previous knowledge was extended to the conversion of tetroses to industrially relevant MHAs in the presence of methanethiol (MeSH) and a Sn (IV) Lewis acid catalyst. It was found that MeSH altered at some extend the reaction pathways established for the reaction with butanethiol. The lower nucleophilicity of MeSH prevented the formation of detrimental thioacetals of vinyl glyoxal resulting in an increase of selectivity towards MHAs. The pressure of inert gas also impacted the reaction’s selectivity, with high nitrogen pressures leading to higher yields of sulphur containing products. When the effect of high nitrogen pressure was combined with an adequate balance between MeSH and methanol, and the addition of small amounts of water to the reaction, the yield and selectivity towards MHAs increased to 38% in one hour reactions at 413 K.

This doctoral research shows that the reaction cascade converting tetroses to MHAs presents crucial steps that are better catalyzed by Brønsted acidity while others by Lewis acidity. Therefore, Sn-BEA zeolites with tailored Lewis and Brønsted acid properties were synthesized and tested in the reaction. Our best Sn-BEA zeolite attained a yield and a selectivity towards MHAs of 39% and 42%, respectively. A kinetic profile and post-reaction analysis of catalysts through TGA, N2 physisorption, FT-IR spectroscopy of absorbed pyridine and CD3CN and PXRD demonstrated the main problems limiting the synthesis of MHAs. We also established that the addition of other components to the reaction mixture contributed to preserve the stability of the acids sites in the catalyst leading to a further increase in yield and selectivity towards MHAs.

Finally, the thesis provides some future perspectives of the remaining challenges to bring the chemical conversion of carbohydrates to methionine hydroxy analogues to an industrial reality.