Model-Based Experimental Design to Estimate Kinetic Parameters of the Enzymatic Hydrolysis of Lignocellulose

Abstract

A model-based nonlinear optimum experimental design technique has been implemented to estimate the kinetic parameters of the lignocellulose enzymatic hydrolysis process, mainly focused on the calculation of reaction rate constants and activation energy parameters. Analysis of the reaction was based on the mechanism of simultaneous consecutive enzymatic reactions of cellulose and hemicellulose to produce sugar-rich syrups. A mathematical model was developed as a set of ordinary differential equations (ODEs). To enhance the parameter estimation process, a control variable that directly affects the response of the system (temperature) was included to predict the state variable profiles over a broad experimental range. To determine the optimal profiles for the temperature gradient, optimal control problems were set up and solved numerically. A comparison between model prediction profiles and experimental data was performed to adjust kinetic parameter estimation to real process features. The results show that nonlinear optimum experiments enhance the quality of the estimates because of the optimality criterion included in the objective function and because the embedded nonlinear behavior is explicitly addressed.

Kinetic modeling for enzymatic hydrolysis of pretreated sugarcane straw

“…According to the parameters values for the original model, cellobiose was the strongest inhibitor of cellobiose formation (r 1 ) to all compared studies except for Câmara [23]. Researchers agree that direct glucose formation from cellulose (r 2 ) was slightly inhibited by cellobiose, most of them found glucose to be the main reaction inhibitor [9,35], however our estimated values for original model suggest that xylose is also a strong inhibitor of cellulase in r 2 , as was found by Prunescu and Sin [36]. Even considering xylose inhibition effect and agreeing about the hierarchical importance of sugar inhibitions in r 2 , the relative inhibition values in Table 3 show important differences of magnitude order between studies (e.g., K 2IX /K 2IG is more than 1000 times higher for Flórez-Sánchez et al [35] that Kadam et al [9]).…”

“…Parameters were estimated via WLS function using EMSO software [32]; results are listed in columns 3-6 of Table 3. Simulation results of enzymatic hydrolysis of pretreated SCS obtained from fitted model for reference condition are shown in Fig 2A. Recent works have estimated the kinetic parameters of Kadam et al [9] model from experimental data of enzymatic hydrolysis for different pretreated biomass, including: corn stover [9], sugarcane bagasse [23], corncob stock [35], and wheat straw [36]. Kadam et al [9] were the first incorporating xylose inhibition from pretreatment; Câmara [23] considered xylose formation from hemicellulose during saccharification; Flórez-Sánchez et al [35] extended the model to xylo-oligosaccharides and arabinose formation and inhibition; and Prunescu and Sin [36] included acid acetic formation and furfural inhibition.…”

Modeling and Kinetic Parameter Estimation of the Enzymatic Hydrolysis Process of Lignocellulosic Materials for Glucose Production

“…Gan 17 proposed a competitive inhibition model considering the crystalline and amorphous fractions of cellulose. Angarita 27 and Flores 24 proposed the formation of xylose from hemicellulose in its hydrolysis mechanism. Following the inhibitory products obtained, such as furan derivatives, organic acids, and phenols, Prunescu 19 modified Kadam's model 18 by incorporating the inhibition generated by furfural, xylose, and acetic acid.…”

“…The modeling quality depends significantly on the availability of a reliable enzymatic hydrolysis model that can reflect the real process, around the main reaction rates and activities of the reagents and products in the process. 24,25 The modeling is complicated because of the heterogeneous nature of the substrate, enzymatic activities, inhibitory effects, and operating conditions, such as pH and temperature. 1,7,18 Zhang 26 classified enzymatic hydrolysis models based on their numbers of solubilization activities and substrate state variables, such as nonmechanical, semimechanical, functional, and structural models.…”

Model‐based analysis of chemical‐looping combustion experiments. Part II: Optimal design of CH4‐NiO reduction experiments

“…10) is the most popular criterion, with several applications in chemical processes. [48][49][50][51][52][53] A potential disadvantage of the D-optimal design is that the calculated design vector is catered to the most sensitive model parameters, resulting in low parameter confidence for the remaining parameters. 54 It is advantageous in that it minimizes the correlation between model parameters, which is of interest here because of the inherent similarity of the reduction and catalytic reactions in CLC.…”