Delignification of corncob via combined hydrodynamic cavitation and enzymatic pretreatment: process optimization by response surface methodology

Optimization of HCE method

Optimization of HCE pretreatment of delignification was performed using orifice plate 1 (OP₁) and orifice plate 2 (OP₂) at an optimized inlet pressure of 50 and 100 kPa, respectively. To determine the optimum values of variables for HCE pretreatment, response surface methodology (RSM) was employed with all the input variables set in the range, with output responses set at maximum levels. For process optimization, three factors were evaluated for three different responses such as maximum lignin reduction (%), hemicellulose reduction (%) and cellulose increase (%) using Box–Behnken design in RSM, which includes biomass loading (A), enzyme loading (B), and time (C). These variables were statistically optimized using 3³ factorial design.

The optimized conditions for HCE pretreatment with orifice plate 1 (HCE-OP₁), are biomass loading at 5%, enzyme loading at 6.5 U g⁻¹ of biomass, and a process time of 60 min. The result showed a reduction in lignin (47.4%) and hemicellulose content (3.2%) and an increase in cellulose content (25.3%). Furthermore, it was observed that the biomass loading and duration have resulted in the optimum response with the higher values. However, the enzyme loading resulted in the mid value. The quadratic model of percentage of lignin reduction had a predicted and adjusted R² values of 0.99 and 0.98, respectively. The predicted R² value was in agreement with the adjusted R² value, since the difference between two values was less than 0.02, indicating that the model developed will be able to give a reasonably good estimate of the response of the system [20].

Adequate precision measures the signal to noise ratio and compares the range of the predicted values at the design points to the average prediction error. This ratio was greater than 4, which is desirable and indicates adequate model discrimination [25]. In the present case, the adequate precision for a percentage of lignin reduction has resulted in a ratio of 103.04, indicating an adequate signal, implying greater predicted response relative associated error.

The response surface plots were generated with one variable kept at its optimum level and other variables varying within the experimental range. The interactions between the model terms were expressed by a three-dimensional surface. A check for interactions was necessary to determine the significance of the model equation [22, 23]. The response surface plots for a percentage of lignin reduction of the OP₁ were shown in Fig. 1. An empirical relationship between the response and the variables was expressed by the following quadratic second-order polynomial equation:

According to the p values (p < 0.05), the linear model terms (A and C) are significant at 95% confidence level. The CV value of 13.43 indicated that the model had reliability of the experimental values. An adequate precision ratio of 130.45 represents an adequate signal. The response surface plots are shown in Fig. 2. The empirical equation fitting the quadratic model is given below:

$$\text{Lignin reduction} = \text{1}{.818182 \times A } + 0{.318588 \times B} + 0{.598915 \times C}$$

where A is biomass loading in %, B is enzyme loading in U g⁻¹ of biomass, and C is duration in min.

Thermal behavior of raw and pretreated biomass

Thermogravimetric analysis of lignocellulosic biomass was used to study the thermal degradation profiles of hemicellulose (250–300 °C), cellulose (300–350 °C), and lignin (300–500 °C) [31, 32]. The appearance of the derivative thermogravimetry (DTG) peak at 200–400 °C was pertaining to hemicellulose and cellulose component [31]. This information enables comparison of the changes in composition of biomass due to pretreatment method [33, 34]. Furthermore, it was clear that the two peaks observed between 200 and 400 °C in the DTG curves was associated with the degradation of hemicellulose and cellulose of the sample. Perhaps the ranges of temperature for decomposition of lignin would have overlapped partially with that of hemicelluloses and cellulose. It is obvious that the temperature associated with two DTG peaks of pretreated corncob samples was higher than raw sample and this could be due to pretreatment effect (Fig. 3 and Table 4).

Type of biomass	Catalyst used	Temperature,  °C (maximum loss rate,  %/ °C)
		Peak 1	Peak 2
Raw corncob	–	290.7 (0.5753%/ °C)	333.7 (0.7088%/ °C)
Pretreated biomass (HCE-OP1)	6.5 U of enzyme g−1 of dry biomass	310.9 (0.7122%/ °C)	349.0 (0.9737%/ °C)
Pretreated biomass (HCE-OP2)	6.5 U of enzyme g−1of dry biomass	310.4 (0.7092%/ °C)	352.0 (0.9820%/ °C)

Apart from temperature change, the peaks also differed in position and height indicating changes in proximate composition namely reduction in lignin and hemicellulose and increase in cellulose content, that might be happened during pretreatment process.

FT-IR analysis

The performance of combined HCE pretreatment indicated that there were no major changes in hemicellulose. However, both the orifice plates behaved differently with regard to lignin removal.

A reduction in peak intensity was observed under OP₁ than OP₂ compared to the raw biomass, at wave numbers 1320, 1424, 1445, and 1509 cm⁻¹ (Fig. 4). This signifies that lignin removal was higher in OP₁, and OP₂ was less efficient. Probably this difference between the plates could be attributed to the less number of radicals formed in the cavitational zone by the OP₂. Moreover, other functional groups also showed similar changes in both OP₁ and OP₂.

Scanning electron microscopy (SEM)

The SEM images of raw sample shows a smooth and undisturbed surface (Fig. 5a). By contrast, after the HCE pretreatment, pores appeared on the surface, and it could be due to removal of lignin. According to Grimaldi et al. [35], the biomass pretreatment removes lignin through destruction of cell wall in two stages namely, loss of cohesion between neighboring cell walls, and degradation inside the cell wall by peeling off and forming holes. The breakdown of structure and pinholes in the surface area of treated sample by cavitation could have enhanced the accessibility for hydrolytic enzymes for increasing the saccharification process [36]. HCE-generated potholes in biomass might have occurred due to collapse of the cavities at the surface of the biomass particles, which could lead to a shear of bimolecular structures, while subsequent laccase treatment led to increase in pothole size (1–6 µm). This indicates effective removal of lignin and deconstruction of biomass in OP₁ than OP₂ (max. 2 µm) (Fig. 5b, c), which could be attributed to the higher intensity of cavities collapse and turbulence generated in OP₁ that resulted in a loosening of the biomass for the effective action of laccase than in plate 2.

X-ray diffraction (XRD)

Raw and pretreated samples were estimated for cellulose crystallinity changes. The value of crystallinity index of pretreated samples (56.0%) for both the orifice plates were higher than the raw sample (52.4%), indicating that HCE pretreatment was more efficient in the removal of amorphous fraction as compared to crystalline fractions. No significant change in the crystallinity of cellulose between treatments and type of plates was observed. This observation testifies that laccase removed only lignin and not changed the cellulose crystallinity. Instead, the HCE pretreatment improved the accessibility of cellulose without modifying the cellulose structure.

Cavitational yield and energy consumption

The optimized conditions for biomass treatment in HCE pretreatment with OP₁ and OP₂ were biomass loading at 5%, enzyme loading at 6.5 U g⁻¹of biomass, and a process time of 60 min. Energy consumption in the HCE pretreatment process for OP₁ (1.35 MJ kg⁻¹) was lower than OP₂ (3.24 MJ kg⁻¹) with cavitational yields of 3.56 × 10⁻⁵ and 2.70 × 10⁻⁵ g J⁻¹, respectively, under optimized conditions for OP₁ and OP₂ for the enzyme pretreatment.

Biomass preparation

The corncob biomass was procured from the farmer fields, Coimbatore, Tamil Nadu, India and dried at ambient temperature (30 °C) for 12 h to reduce the moisture content from 20 to 14%, then the size was reduced in a sequence through shredder (Chudekar Agro Engg Pvt Ltd., Model: 53 H, India), pin mill (Premium Pulman Pvt Ltd., Model: PPM-12, India) and grinder (Aashapura Enterprises, Model: Stylo 750, India) to ≤ 212 µm, which was sieved via ASTM sieve size No. 70.

Enzyme selection

Generally, two families of ligninolytic enzymes are widely considered for enzymatic delignification, which includes phenol oxidase (laccase) and peroxidases (lignin peroxidase, or LiP and manganese peroxidase, or MnP) [37]. Laccase belongs to the copper oxidase enzyme family, similar to other phenol-oxidizing enzymes and it preferably polymerizes lignin by coupling of the phenoxy radicals produced from oxidation of lignin phenolic groups. Laccase enzyme was selected and used as catalyst for this HCE pretreatment [38].

The enzyme laccase used in the study was from microbial source (Trametes versicolor). The laccase was purchased from Sigma-Aldrich, Bangalore, India and used as such. The laccase activity was determined with the reaction mixture contained appropriately diluted enzyme mixed with 1 mM ABTS in sodium phosphate buffer (50 mM, pH 4.5) at 30 °C for 5 min. using 1 mM ABTS by monitoring changes in absorbance at 420 nm (€ max = 3.6 × 104 M⁻¹ cm⁻¹) spectrophotometrically in a Spectramax 360 (Molecular devices, USA). One unit of enzyme activity refers to the amount of enzyme required to oxidize 1 µM min⁻¹ of the ABTS substrate under standard assay conditions.

System description and operating conditions

The main hurdles associated with commercialization of higher biomass loading pretreatment reactors are complexity of design, reactor geometry and upscale, poor mixing characteristics of reactants, and it is an energy-intensive process. For these reasons, low biomass loading rate (< 20%) is largely preferred in most of the biomass pretreatments [39]. The main criteria to be considered for reactor design is rheological properties of different slurries collected from different unit operations in the fermentation process [40]. Mostly, water with catalyst is used as the working fluid in the recent HC biomass pretreatment studies; whereas, in the present study corncob biomass slurry (buffer + biocatalyst + powdered biomass) was used as the working liquid. The process design of this pretreatment is to supply the biomass slurry via holes in the orifice plate to cavitation zone. The rheological study involving different corncob biomass slurries (2.50, 3.75, 5.00, 6.75, 7.50, 8.75 and 10.00%) showed increased viscosity and yield stress, along with increased biomass solid loading and these slurries exhibit pseudo-plastic or shear-thinning behavior [41]. It is also observed that high biomass loading rate (> 6.75%) of corncob slurry have blocked the holes in orifice plate. Based on this result, low biomass loading rates (2.50, 3.75 and 5.00%) were selected for HCE pretreatment. Hydrodynamic cavitation reactor (HCR) consists of circulation tank (6 L capacity), orifice plate, flanges for orifice plate, centrifugal pump, electrical motor, gate valves for priming and bypass, pressure gauges, and pipe accessories (Fig. 6a). Two pressure gauges (P₁ and P₂) were fixed on both the downstream and upstream sides of the orifice plate to measure the pressure drop. For same area opening in the orifice plate, higher diameter of the hole is recommended for intensive cavitation applications and vice versa. In this reactor, total hole area openings made in the plate were kept as constant (28.26 mm²) and for constant area openings, two configurations of orifice plates were used, viz., orifice plate 1 (OP₁:9 holes and 2 mm Ø) and orifice plate 2 (OP₂:4 holes and 3 mm Ø) (Fig. 6c). The suction pipe of the pump was connected at the base of the circulation tank. The main pipeline was divided into three sub-pipelines to serve for three purposes namely priming, bypass line, sub mainline to accommodate flanges and orifice plate, and it was connected to the delivery pipe of the centrifugal pump. To make a closed loop circulation, the working fluid was supplied from the circulation tank to the orifice plate with the help of a centrifugal pump and then sent back to circulation tank. The purpose of the bypass valve was to regulate the flow rate/pressure of the working fluid (biomass slurry) by passing through the orifice plate of the reactor.

For HCE biomass pretreatment, the biomass slurry was prepared by mixing appropriate quantities of biomass powder and a biocatalyst (laccase enzyme) in an acetate buffer pH 4.5. For example, 200 g of biomass and 6.5 U g⁻¹ of enzyme were added to 4000 mL of acetate buffer to make biomass slurry for 5% biomass loading and 6.5 U g⁻¹ of enzyme. The prepared slurry was added to circulation tank and made to circulate via selected orifice plate continuously, which enabled the biomass to get exposed to cavitation action. Cavitation involves production and aggressive collapse of micro bubbles to generate more hotspots, having higher temperature and pressure. This is sufficient to make the chemical and physical transformations in the lignocellulosic biomass. During the process, decomposition of water molecule causes free radical formation, such as hydroxyl radicals leading to turbulence action of working fluid in the cavitation zone, which eventually helps in rupturing the lignin barrier. Laccase enzyme can oxidize a variety of phenolic subunits of lignin and other aromatic compounds via radical-catalyzed mechanism to yield oxygen-centred free radical and quinine for subsequent reduction in the polymerization reaction [42–44]. Hence, HCE-based biomass treatment helps in the generation of highly reactive free radicals in HCR, which can improve the mass transport process rates and enhance the lignin degradation. Based on these results, three biomass loadings (2.50, 3.75 and 5.00%) were selected for HCE pretreatment.

Compositional analysis

Experimental design

Response surface methodology (RSM) is a statistical and mathematical technique widely used for optimization of process parameters and their interactions on output response(s). The added advantages of optimization via RSM approach are (i) only a minimum number of trials are sufficient to find optimum, (ii) to find a correlation between independent inputs and output responses and (iii) reduction in time, materials and cost because of the less number of trials are needed [46]. The optimization of biomass pretreatment process involves studying the influence of operational parameters and their interactions on lignin removal from raw biomass. For enzymatic biomass pretreatment, process parameters such as biomass loading, enzyme loading and reaction time are crucial factors in deconstruction of the lignin structure as well as lead to release of reducing sugars due to solubilization of biomass.

Response surface methodology (RSM) was employed to determine the optimal conditions for HCE pretreatment to attain the maximum percentage of lignin reduction in the corncob biomass samples. The response was assumed to be influenced by three independent variables, catalyst concentration (A), biomass to liquid ratio (B) and reaction time (C). The range of three selected independent variables used for pretreatment process are: biomass loading of 2.5–5.0%, enzyme loading of 3–10 U g⁻¹ of dried biomass, and reaction time of 5–60 min. Based on the results of the preliminary trials, the above range of levels of the three independent variables were fixed. A total of 17 experimental trials of the three variables were designed by Box–Behnken design via Design-Expert software 10.0 (Stat-Ease, Inc., USA) [47].

Structural composition of biomass

Effect of cavitation on temperature of working liquid and residual enzyme activity

During the experimental trials, the temperature of the working liquid was measured at an interval of 5 min by digital thermometer (Multi-thermometer, India). Similarly, the residual enzyme activity during the reaction period was also determined [50].

Cavitational yield

The energy required per kg of biomass in an HCE pretreatment process