Limited oil resources and the devastating effects of climate change from the burning of fossil fuels make it necessary to identify sustainable alternative sources of energy for the future . Many alternative, potentially sustainable sources of energy exist, however there are limited choices for the replacement of liquid fossil fuels. One of these possibilities is the production of fuels from lignocellulosic biomass. The first step in one of the more promising conversion pathways includes a biochemical conversion of the lignocellulosic material whereby the structural sugars present in lignocellulosic biomass are depolymerized into their monomeric constituents via an enzymatic hydrolysis step, providing a fermentable sugar stream rich in glucose .
Recent commercial cellulase preparations have been shown to be effective at hydrolyzing cellulose under industrially relevant conditions, however the high cost of enzymes remains a significant barrier to the economical production of ethanol from lignocellulosic biomass [3, 4]. It is therefore necessary to reduce the amount of enzyme required for the enzymatic hydrolysis step. Current enzyme loadings for cellulose hydrolysis remain high compared to enzymatic starch hydrolysis. Hence, reducing the amount of enzyme needed or increasing the enzyme productivity in the process is a promising approach. This has been investigated by looking for methods to improve hydrolysis yields while lowering enzyme doses. A variety of methods have been suggested to achieve increased hydrolysis yields, including via surfactant addition , gradual substrate loading  or advanced reactor configurations coupled with product removal to avoid inhibition [7–9]. However, these methods have yet to be shown to be cost effective. One method that may reduce the amount of enzyme used and increase enzymatic productivity is to recycle the enzymes [10, 11]. Conceptually, the assumption is that by recovering the active enzymes from the output of the enzymatic hydrolysis step, it is possible to decrease the amount of new enzyme which must be added to the hydrolysis, and therefore reduce the overall enzyme cost in the process. This method also lends itself to continuous processes, a necessity for industrial application. Recycling can also increase the enzyme substrate interaction time, which can lead to an overall increase in enzyme conversion efficiency. Due to mass transfer limitations of the insoluble substrate, immobilization, the most commonly applied method of enzyme recycle, is not an option for cellulose hydrolysis, so other options for enzyme recycling must be developed.
The cellulase enzymes currently employed in the hydrolysis of lignocellulosic biomass readily bind to cellulose, and those enzymes which are active on the cellulose polymer, specifically cellobiohydro-lyases (EC 188.8.131.52) and endoglucanases (EC 184.108.40.206) remain adsorbed to the cellulose polymer during hydrolysis [12–14]. Because the substrate is present as a solid, the cellulases stay attached to the insoluble solids fraction. For cellulose biomass substrates that have been de-lignified during pretreatment, a significant amount of cellulase enzymes have been found to desorb from the solid substrate during the hydrolysis [13, 15]. β-glucosidases (EC 220.127.116.11) have a soluble substrate, cellobiose, and a majority of the enzyme activity has been found to remain in the soluble (liquid) fraction during hydrolysis . Cellulases also adsorb to lignin, and therefore a significant fraction can be found adsorbed to the lignin throughout the reaction, as lignin concentration remains constant . Thus the enzymes are primarily associated with the insoluble solids fraction, though a significant amount of activity, especially β-glucosidase, can still be found in the liquid fraction. To successfully recover the enzymes, they must either be separated and collected from their associated fractions, or the enzyme containing fractions must be recycled into the subsequent hydrolyses.
A significant amount of work has already been carried out investigating the recycle of cellulases [10, 11, 14, 17–21]. The majority of cellulase recycling methods that have been reported involved either separating the enzymes from the solid or liquid phases, or recycling of the solid and/or liquid phase directly. Approaches have been demonstrated where free enzymes were recovered from the liquid fraction by membrane filtration , or where fresh substrate was introduced to the liquor fraction and the enzymes were allowed to adsorb to the substrate before a further separation and hydrolysis step [18, 19]. Similarly, the enzymes associated with the solids have been recovered by washing with excess volumes of buffer, sometimes with surfactants, to desorb the enzymes, which were then concentrated and added to the fresh substrate . These methods have shown varying levels of success under controlled laboratory conditions and with specially prepared feedstocks. The recycle methods which rely on enzyme isolation and recovery have yet to be demonstrated under process relevant conditions, and the effective scale up of the separation processes used have not been shown. A more straightforward approach to enzyme recycle is the direct recycle of the residual solid fraction into subsequent hydrolyses. The recycle of the solids fraction after solid liquid separation has been demonstrated on a number of substrates and in combination with other recycle methods [15, 19, 22]. This approach has also been coupled with methods for recovering enzymes from the liquor fraction, primarily by the method of a limited exposure time of the liquor to fresh substrate at low temperatures [18, 19]. Previous attempts have shown that a large portion of the enzymatic activity could be recovered using a combination of these methods. However, these studies relied on high enzyme loadings, which often resulted in complete hydrolysis of the cellulose, low total solids concentrations (from 2-5% TS), extended reaction times, and supplementation of each recycle round with β-glucosidase [14, 15, 18, 19, 22]. As well, the best results were demonstrated on pretreated materials with very low lignin content, which were produced using pretreatments specifically designed to remove lignin [18, 19, 22]. Recycle performance decreased significantly when applied to lignin containing substrates created with more standard pretreatment . While these studies have shown that enzyme recycle is technically possible under defined conditions and with ideal enzyme extraction methods, no work has yet demonstrated cellulase enzyme recycle under process relevant conditions and by relying on unit operations which could be economically feasible for industrial production. Little consideration has been made into the process implications of enzyme recycle on an industrial scale, and how various methods of enzyme separation or re-adsorption could be applied. Most recycle studies have attempted to maximize the amount of activity recycled, irrespective of process intensity, with the idea that recovery must approach 100% of initial activity to be industrially interesting. Compared to the current cellulose hydrolysis processing regimes for lignocellulosic biomass, which requires fresh enzyme addition with each new batch of substrate, a significant fraction of enzyme activity could be recycled with minimal sample processing, and large decreases in enzyme cost could be achieved.
The objective of this study was to determine if, by recycling the insoluble solids fraction, a significant amount of enzyme activity could be reused, and therefore increase overall product yields or decrease the amount of required enzyme needed to reach a given level of conversion. It was also desired to determine which process variables, such as solids washing and fraction of solids recycled, had significant effects on process yields, and how the manipulation of these variables would impact the hydrolysis reaction conditions. In this study, the recycle of cellulase enzymes by recycling the insoluble solids residue present after hydrolysis was investigated and evaluated at conditions which were closer to industrial processing conditions than many previous studies. The efficacy of enzyme recycle by the proposed method was evaluated for a number of successive recycle scenarios while modifying a number of recycle conditions and measuring the changes in product formation. The data from these experiments were then used to develop a computational model to predict process parameters after a large number of recycles, thus giving an idea of steady state conditions. The influence of the lignin rich residue was also directly evaluated for its effect on enzyme hydrolysis.