In situ product recovery of butanol from the acetone butanol ethanol fermentation

Abstract

From 1916 the “acetone butanol ethanol”, or “ABE”, fermentation process was the main production method for n-butanol. It was superseded in the 1950s by a more economical petrochemical process, causing the majority of plants to cease operation. In the fermentation, product inhibition led to low productivity and high energy demand in the downstream processing, making the process unable to compete with the petrochemical route. Overcoming these problems could revive the ABE industry and promote a bio-based economy. In situ product recovery (ISPR) can be applied to the fermentation process to counteract the effects of product toxicity. Productivity increases of greater than 300% are theoretically possible. Many ISPR techniques have been applied to the ABE process at laboratory scale, but a direct comparison of the different techniques has been hindered by experimental inconsistencies. Here, a techno-economic analysis was performed to compare the most developed ISPR techniques, with process simulations providing comparative data on the separation efficiency and energy demand. All the techniques were found to be economically viable, with profit increases compared to an equivalent batch plant of 110-175% and payback times of 2.2-4.5 years. In addition to generating the most profit and having the shortest payback time, perstraction was the only technique to lead to a reduction in overall plant energy demand, by ~5%, compared to a traditional ABE process. Thus perstraction warrants further investigation for application to the ABE process. Perstraction is significantly underdeveloped compared to other ISPR techniques. It was originally designed to overcome various problems associated with liquid-liquid extractions, including solvent toxicity. Here, experiments focused on the use of high-distribution toxic extractants with commercially available membranes. Results showed that high-distribution toxic extractants (1-pentanol, 1-hexanol, 1-heptanol, 1-octanol and 2-ethyl-1-hexanol) have a larger mass transfer coefficient than oleyl alcohol (the main non-toxic extractant), although chemical structure differences, such as branching, can have a greater impact on mass transfer than distribution coefficient. Unfortunately, all extractants investigated here were transferred across the membrane to some extent, which would limit perstraction to non-toxic extractants. However, differences in membrane type have a greater impact on mass transfer than the choice of extractant. Porous membranes have a mass transfer coefficient 10 times greater than non-porous membranes, which would see a factor of 10 reduction in ii membrane size and cost. Overall, this work has confirmed that perstraction is technically viable and compared options for process improvements through membrane and extractant selection.