Biocatalytic approach on the treatment of edible oil refinery wastewater

Abstract

The number of edible oil refineries has increased in the last few years, with a corresponding increase in oil production. As a result, edible oil-containing wastewater (EOCW) is being produced in huge quantities. Conventional technologies are inefficient at treating this wastewater due to the highly hydrophobic nature of the lipids. In the present study, we have used lipase immobilized nanoporous activated carbons and surface functionalized nanoporous activated carbons for the treatment of lipid-containing wastewater. The nanoporous activated carbon (NAC) was prepared from rice husks and the NAC was surface functionalized by the addition of ethylenediamine and glutaraldehyde with (FNAC2) and without (FNAC1) the addition of a reducing agent, sodium borohydride. The lipase obtained from marine Pseudomonas otitidis, using cooked waste sunflower oil (CWSO) as the substrate, was then immobilized onto the NAC and the functionalized nanoporous activated carbons (FNAC1 and FNAC2). The maximum immobilization capacities of NAC, FNAC1 and FNAC2 were 3640, 4788 and 4400 U g⁻¹, respectively, at the optimum conditions. The carrier matrices in the free and lipase immobilized form were characterized using scanning electron microscopy, Fourier transform infrared spectrometry and X-ray diffraction. The thermal behavior of the free and immobilized lipases was studied using thermogravimetric analysis. Michaelis–Menten enzyme kinetics, adsorption isotherms and nonlinear kinetic models were evaluated for the immobilization of lipase. The lipase immobilized carrier matrices were employed in the treatment of EOCW under batch and continuous mode operations. At the end of the 50^th cycle, FNAC1-L (89.78%) showed a higher operational stability than FNAC2-L (87.36%) and NAC-L (76.59%). The treatment of EOCW by immobilized lipases followed the pseudo second order rate kinetic model.

---

1. Introduction

Lipid-containing wastewater can cause serious environmental problems due to its hazardous nature. Lipids (fats and oils) are the major organic constituent in municipal and some industrial wastewaters. Edible oil refineries, restaurants, slaughterhouses, wool scouring and the food and dairy industries are the major sources of the discharge of high concentrations of lipids (>100 mg L⁻¹) in wastewater. The conventional biological wastewater treatment system failed to treat such high concentrations of lipids in the wastewater due to the following problem: during the treatment, because of the hydrophobic nature of the lipids, they form a layer on the water surface and adsorb onto the cell wall of the microbes and hinder the diffusion of O₂ into the microbial cell, thereby slowing its metabolism. This may lead to the formation of a filamentous sludge and reduce the overall removal of organic matter. Because of this reason, the treatment of lipid-rich wastewater is still a challenge with conventional technologies.¹ However, high-fat wastewater can be treated with an enzyme (lipase), and this type of enzymatic treatment system has gained much attention due to its high specificity, high catalytic efficiency and bio-degradability.²

For practical and economical reasons, it is advantageous to use lipase in its immobilized form. An important requirement for enzyme immobilization is that the matrix should provide a biocompatible and inert environment, i.e., it should not interfere with the native structure of the enzyme, which could compromise its biological activity.³ To achieve high enzyme loading and catalytic efficiency for large-scale operation and application, supports with a high surface to volume ratio, an inherent characteristic of nanoscale materials, are often desirable. Therefore, in recent years, nano-structured supports such as mesoporous materials (activated carbon), nanoparticles (gold or gamma Fe₂O₃ magnetic nanoparticles), and nanofibers (PANCMPCs with phospholipids) have been widely studied for enzyme immobilization.^4–11 However, there has been no attention on the immobilization of enzymes/lipase onto nanoporous activated carbon (NAC) materials and their stability towards continuous operation. In the case of enzyme immobilization in porous materials, a much reduced mass-transfer resistance is expected for smaller porous particles owing to the shortened diffusion path of the substrates compared with macroporous materials. Overall, NAC provides an ideal remedy for the immobilization of lipase due to its high surface area and high enzyme loading. Therefore, the focal theme of the present study was aimed at immobilization of lipase in NAC from agricultural solid waste (rice husks).

Porous activated carbon has the ability to act as a buffer to the immobilized enzymes, which offers structural stability during immobilization. However, the sloughing off of the immobilized enzyme from the carrier matrices during repeated applications was not prevented. The sustained stability of the enzymes was considered as one of the main requirements for their industrial exploitation.¹² The enrichment of reactive groups on the surface of the matrix by chemical modification may lead to an increase in the stabilization of enzymes, because it may permit a more intense multipoint covalent attachment.^13–15 Chemical modification of the support matrices by ethylenediamine and glutaraldehyde has been followed for the strong or covalent immobilization of enzymes.^14–17 Apart from chemical modification of the carrier matrices with ethylenediamine and glutaraldehyde, Fernandez-Lafuente et al.¹⁸ reported that during the addition of a reducing agent, sodium borohydride, along with the above chemicals, all the enzyme-support bonds will be transformed into very stable secondary amine bonds. Therefore, considering this application, the present study dealt with the functionalization of NAC using two methods: (1) without addition of a reducing agent to the ethylenediamine and glutaraldehyde and (2) with the addition of a reducing agent to the ethylenediamine and glutaraldehyde. Comparative studies on the loading capacity of lipase on the matrices and the biocatalytic ability of the lipase in the immobilized and free form for the treatment of lipid-containing wastewater was the other objective of the present study. The hydrolysis rate kinetics were evaluated for the understanding of the nature of the reaction.

2. Materials and methods

The edible oil-containing wastewater (EOCW) was collected from the local oil refinery, Chennai, India. Ethylenediamine and glutaraldehyde of purity (99.5%) were purchased from Sigma-Aldrich-Fluka Chemical Co., India. The EOCW was characterized for its fatty acid composition using gas chromatography (GC).

The NAC was prepared from rice husks. The rice husks were pre-carbonized at 400 °C in an airtight crucible and then the pre-carbonized carbon was activated using 40% hydrofluoric acid (∼22 N) at 800 °C in a muffle furnace, to render it porous in nature. The prepared material was washed with distilled water to remove the excess hydrofluoric acid and dried at 110 °C for 6 h to obtain the final product, and it was sieved to 300 μm size. The material was labeled as NAC for further use.

The functionalization of the NAC was carried out by two methods. In order to find out the importance of the reducing agent on the functionalization of the carrier matrices, the NAC was functionalized with reducing agent and without reducing agent as follows.

The amine solution (2.5%, v/v) was prepared by mixing 2.5 ml of ethylenediamine with and without 0.3 g of sodium borohydrate (reducing agent) in 97.5 ml of acetone and it was added to 30 g of the dried (moisture free) NAC and mixed thoroughly for 10 min to obtain the amino-functionalized NAC. To facilitate the strong immobilization of enzyme, the amino-functionalized NAC was activated with an aldehyde group by the addition of glutaraldehyde. The aldehyde solution was prepared by mixing 100 ml of glutaraldehyde (25%, w/v) with and without 0.3 g of sodium borohydrate in 100 ml of acetone and then the mixture was added to the amino-functionalized NAC, and gently stirred for 30 min, using a magnetic stirrer. The above materials were dried under vacuum for the complete evaporation of the solvent and heated at 150 °C for 24 h. The prepared materials were washed with distilled water to remove the unbound chemicals. The washed material was dried at 110 °C for 6 h to obtain the final product, which was then sieved to 300 μm in size. The NAC without and with sodium borohydrate were labeled as functionalized NAC 1 (FNAC1)¹⁵ and NAC 2 (FNAC2), respectively, for further studies.

The carbon, hydrogen and nitrogen contents of the NAC, FNAC1 and FNAC2 were determined using a 1108 model Carlo Erba elemental analyzer.

The marine bacterial strain P. otitidis was isolated from marine sediment as reported by Ramani et al.,¹⁹ and grown in Zobell Marine Broth (HiMedia) with 1% CWSO as the substrate. The substrate was emulsified in broth, using an ultrasonicator (BANDELIN, Germany) at 23 kHz. The culture conditions followed were: 300 ml of the medium in one litre Erlenmeyer flasks, rotation at 100 rpm at 37 °C, and at pH 7.5. Biomass in the broth was removed by centrifugation and the supernatant was subjected to lipase activity determination.¹⁹ The enzyme from the cell-free supernatant was purified as explained by Ramani et al.¹⁹ The protein content of the lipase was determined using the Lowry method.²⁰ The lipase activity was measured by titrimetric assay according to an olive oil emulsion as described in our work. A unit of lipase activity was defined as the amount of enzyme that released 1 μmole of fatty acid per minute under assay conditions.

The optimum conditions for the immobilization of lipase in NAC, FNAC1 and FNAC2 were carried out by varying the parameters such as time (30 to 240 min), pH (3 to 9), temperature (20 to 50 °C) and initial concentration of lipase (88 × 10³, 176 × 10³, 264 × 10³, 352 × 10³, 440 × 10³ and 528 × 10³ U L⁻¹ in phosphate buffer (pH 7.0)). The mass of NAC, FNAC1 and FNAC2 used for the immobilization of lipase was 1 g per 15 ml of buffered lipase solution. Lipase immobilization was investigated with hydrolysis activity as the objective function. The residual lipase activity was determined using a lipase assay.¹⁹ The immobilized lipase activity was calculated by subtracting the final activity of lipase solution from the initial activity. The lipase-immobilized NAC, FNAC1 and FNAC2 were named as NAC-L, FNAC1-L and FNAC2-L, respectively, and stored for further studies.

Lipase activity measurements were made under different substrate (olive oil) concentrations (0.0403 to 0.403 mM) to determine the kinetic parameters, maximum reaction rate (V_max) and Michaelis–Menten constant (K_m) for free (L) and immobilized lipases (NAC-L, FNAC1-L and FNAC2-L). Olive oil was used as the substrate and the resulting product (liberated fatty acid) concentration was determined according to the method followed by Ramani et al.¹⁹ It is assumed that polyvinyl alcohol (PVA) mixed with the substrate in the lipase assay produces an emulsion of olive oil, and acts only as a stabilizer of olive oil,²¹ but it does not take part in the hydrolysis reaction. The kinetic parameters were estimated from the Lineweaver–Burk plot equation (eqn (1)).

where [S] is substrate concentration (mM) and v is reaction rate (mM min⁻¹). The surface morphology of the nanoporous carrier matrices (NAC, FNAC1 and FNAC2) and lipase immobilized nanoporous carrier matrices (NAC-L, FNAC1-L and FNAC2-L) was determined using a Leo-Jeol scanning electron microscope at a magnification of 2500–10 000×. The carbon samples were coated with gold by a gold sputtering device for clear visibility of the surface morphology.

A PerkinElmer infrared spectrophotometer was used for the investigation of the surface functional groups. Samples with KBr (spectroscopic grade) pellets were prepared in the size range of about 10–13 mm in diameter and 1 mm in thickness. The free lipase (L), nanoporous carrier matrices (NAC, FNAC1 and FNAC2) and lipase immobilized nanoporous carrier matrices (NAC-L, FNAC1-L and FNAC2-L) were scanned in the spectral range of 4000–400 cm⁻¹.

The XRD patterns of free lipase (L), NAC, FNAC1, FNAC2, NAC-L, FNAC1-L and FNAC2-L were determined for 2θ values from 10 to 80° using Cu K_α radiation at a wavelength of λ = 1.514 Å (40 kV, 30 mA) with silicon as the reference using a high resolution GUINER powder X-ray diffractometer (SEIFERT 3003 TT, Germany).

The required quantity (8–10 mg) of the free lipase, NAC, FNAC1, FNAC2, NAC-L, FNAC1-L and FNAC2-L samples was loaded onto a platinum TGA pan and gravimetric analysis was made under a pure nitrogen atmosphere, from 0 °C to 800 °C using a temperature gradient of 10 °C min⁻¹. Scans were routinely recorded in duplicate using TGA Universal V4.4A TA instruments.

For DSC analysis, the required quantity of the lipase, biosurfactant NAC, FNAC1, FNAC2, NAC-L, FNAC1-L and FNAC2-L samples (8–10 mg) was loaded in an aluminum DSC pan and gravimetric analysis was made under a reduced nitrogen atmosphere, from 0 °C to 200 °C using a temperature gradient of 10 °C min⁻¹. Scans were routinely recorded in duplicate using a DSC Q200 (V23.10 Build 79).

The hydrolysis study was carried out by using the immobilized lipases loaded on a continuous packed bed glass column reactor. The continuous packed bed column reactor with the dimensions of 1.5 cm diameter; 20 cm height and at 10 cm height from the bottom was fabricated. The hydrolysis experiment was carried out by varying time (30 to 240 min), pH (4 to 8), flow rate (0.5, 1.0, 1.5 and 2.0 ml min⁻¹) and volume of EOCW (20, 25, and 30 ml L⁻¹). The glass column was placed in a thermostable water bath, maintained at constant temperature. The samples were withdrawn at regular time intervals and analyzed for the residual lipid content. The % conversion of CWSO was calculated using eqn (2).

% Conversion = (1 − (final lipid content/initial lipid content)) × 100 (2)

The EOCW (20 ml) was passed through the glass column reactor packed with lipase immobilized matrices at a flow rate of 1 ml min⁻¹. The samples were collected and the residual lipid content was measured. The cumulative volume of EOCW passing through the column was recorded.

In order to investigate the kinetic rate constants for the hydrolysis of oil containing wastewater, using the purified lipase from P. otitidis, non-linear kinetic models were applied. The pseudo first order²² and pseudo second order²³ kinetic models were employed, following the equations (eqn 3 and 4), respectively.

where r_e and r_t are the amount of lipid hydrolyzed (mg mg⁻¹ of lipase) at equilibrium and at time (t), k₁ and k₂ are the first and second order rate constants.

3. Results and discussion

The composition of the EOCW was linoleic acid (56%), oleic acid (24%), palmitic acid (11%), and stearic acid (9%). The preparation of mesoporous activated carbon (MAC) from the agricultural solid waste, rice husks, for the immobilization of lipase has been reported elsewhere.²⁴ The MAC was porous in nature with an amorphous structure due to the presence of silica. In the presented study, hydrofluoric acid was used to remove the silica present in the rice husks with the intention to create a high yield of nanopores and a crystalline structure (Fig. 1).

Fig. 1 Schematic diagram of the preparation of NAC, functionalization of NAC with and without reducing agent and immobilization of lipase.

The functionalization of the NAC was carried out by two methods: 1) without the addition of reducing agent (FNAC1) and 2) with the addition of reducing agent, sodium borohydride, (FNAC2) (Fig. 1). The mechanism behind the binding of amine and aldehyde groups (without reducing agent) with the NAC (FNAC1) may be explained as follows. The ketone group present in the NAC was bound with ethylenediamine to form aminated NAC. This aminated NAC was further treated with glutaraldehyde to introduce the aldehyde groups on the matrix surface. It may be proposed that the amino groups extending from the aminated NAC surface were condensed with glutaraldehyde, having one (monomer) or two molecules (dimer) of glutaraldehyde per primary amino group.¹⁵

In order to compare the efficiency of the reducing agent on the bonding of the amino–aldehyde group (Am–Al) with NAC and on the strong immobilization of lipase, the NAC was functionalized with amine (Am) and aldehyde (Al) groups in the presence of sodium borohydride (FNAC2). The elemental composition of NAC, FNAC1, and FNAC2 is presented in Table 1.

Table 1 Characterization of surface functionalized nanoporous activated carbon catalysts

S. no.	Parameters	NAC	FNAC1	FNAC2
1.	Carbon (%)	36.4	42.5	43.8
2.	Hydrogen (%)	0.95	1.69	2.96
3.	Nitrogen (%)	2.89	12.5	19.33
4.	Moisture content (%)	6.7	5.6	5.4
5.	Ash content (%)	12.7	14.1	13.1
6.	Bulk density (g ml^−1)	0.59	0.26	0.34
7.	Apparent density (g ml^−1)	0.73	0.52	0.49
8.	Phenol number (mg g^−1)	18	27.8	23.6
9.	Decolorizing capacity (mg g^−1)	32.2	41.2	39.6
10.	Point of zero charge (PZC)	7	7.6	7.5

---

The effect of time on the immobilization of lipase was carried out in order to determine the equilibrium points. The activity of lipase was measured at different time intervals (30 to 240 min) during immobilization (Fig. 2a). The immobilization was performed at different time periods (30 to 240 min) at pH 7 and at 35 °C. It was found that in all the experiments, the immobilization was rapid up to 90 min and then there was a slight increase in immobilization until the equilibrium (150 min) was attained. Initially, the number of adsorption sites available is higher and the driving force for the mass transfer is greater. As the immobilization time was increased, the number of bare active sites became less and the lipase molecules might become clustered inside the carbon particles, thus impairing the diffusion of lipase. The immobilization capacity of FNAC1, FNAC2 and NAC, respectively, was 4080, 3340 and 2720 U g⁻¹ of the matrix.

Fig. 2 Effect of (a) time, (b) pH, (c) temperature and (d) initial concentration of lipase on the immobilization of lipase on NAC, FNAC1 and FNAC2.

In order to determine the optimum pH for the immobilization of the lipase onto NAC, FNAC1 and FNAC2, the experiments were carried out at varying pH values between pH 3 and 9 at an initial lipase activity of 88 × 10³ U L⁻¹ and a temperature of 35 °C. A bell-shaped curve was obtained with the maxima attained at pH 5 for FNAC1 and FNAC2 and pH 6 for NAC with immobilization capacities of 4592, 3760 and 3420 U g⁻¹, respectively, as shown in Fig. 2b. Outside this pH range, the enzyme may become denatured, lose its tertiary structure and, therefore, lose its ability to function as a catalyst for the reaction.

The optimum temperature for the immobilization of lipase onto NAC, FNAC1 and FNAC2 was carried out by varying the temperature from 20 to 50 °C at an initial lipase activity of 88 × 10³ U L⁻¹ and at their optimum pH. A bell-shaped curve was also obtained with a maxima reached at 30 °C, as shown in Fig. 2c. At this temperature, the maximum lipase loading achieved was 4624, 4320 and 3520 U g⁻¹ for FNAC1, FNAC2 and NAC, respectively. If the temperature of the reaction is raised too high past the optimum temperature, the reaction rate decreases as the enzyme becomes denatured and loses its ability to function as a catalyst for the reaction.

The immobilization capacity of NAC, FNAC1 and FNAC2 was increased while increasing the initial lipase activities. The enzyme load at initial lipase activities of 88 × 10³, 176 × 10³, 264 × 10³, 352 × 10³, 440 × 10³ and 528 × 10³ U L⁻¹ was studied and the optimum initial lipase activity was found to be 440 × 10³ U L⁻¹ with immobilization capacities of 5788 and 3640 U g⁻¹ for FNAC1 and NAC, respectively, and 352 × 10³ U L⁻¹ with an immobilization capacity of 4400 U g⁻¹ for FNAC2 at the optimum conditions of pH and temperature (Fig. 2d). The enzyme load onto NAC increased with an increase in initial lipase activity, however, the percentage immobilization decreased with an increase in lipase activity. The high loading of enzyme in FNAC1 may be due to the higher surface area than the other matrices.

Adsorption data are represented by isotherms and they were used in determining the immobilization capacity of the FNAC1, FNAC2 and NAC. In the present study, the Langmuir and Freundlich isotherm models were used for this purpose. The Langmuir isotherms were developed based upon an assumption of monolayer adsorption onto a surface containing a finite number of adsorption sites of uniform energies of adsorption. It is represented as follows (eqn (5)).²⁵

where q_e (U g⁻¹) and C_e (U L⁻¹) are the amount of lipase adsorbed per unit weight of adsorbent and equilibrium liquid phase concentration (U L⁻¹), respectively. K_L (L g⁻¹) and b (L U⁻¹) are the Langmuir adsorption constants. The essential features of the Langmuir isotherm can be expressed in terms of a dimensionless constant called the separation factor (R_L, also called the equilibrium parameter), defined by eqn (6).where C_o (U L⁻¹) is the initial lipase concentration. The value of R_L whether the shape of the isotherms is either unfavourable (R_L > 1), linear (R_L = 1), favorable (0 < R_L < 1) or irreversible (R_L = 0).

The Freundlich isotherm can be applied to non-ideal adsorption on heterogeneous surfaces as well as multilayer sorption and is expressed by eqn (7).

where K_F (U g⁻¹)(L U⁻¹) is the Freundlich constant related to the immobilization capacity and n is the Freundlich exponent.

The adsorption isotherms presented in Table 2 indicated that FNAC1 and FNAC2 obeyed the Freundlich isotherm and NAC obeyed the Langmuir isotherm based on the regression coefficient (R²) and χ². The value of R_L, the separation factor, fell in the range of less than zero, indicating that the immobilization of lipase onto NAC, FNAC1 and FNAC2 was irreversible. This confirmed the immobilization of lipase by formation of a strong attachment with the carrier matrices NAC, FNAC1 and FNAC2.

Table 2 Isotherm parameters for the immobilization of lipase onto FNAC1, FNAC2 and NAC

	FNAC1	FNAC2	NAC
Langmuir
KL (L g^−1)	4.475	2.827	0.447
b (L U^−1)	0.05	0.02	0.44
RL	2.53 × 10^−6	4.01 × 10^−6	0.254 × 10^−6
R^2	0.98	0.68	0.987
Freundlich
KF (U g^−1) (L U^−1)	15.45	0.00032	0.328
1/n	2.08	0.23	0.78
R^2	0.99	0.97	0.976

---

The validity of the order of immobilization processes was based on the regression coefficients and χ² values. The first order rate constant, k₁, was found to be 0.011 min⁻¹, 0.367 min⁻¹ and 0.013 min⁻¹ for NAC, FNAC1 and FNAC2, respectively. The second order rate constant, k₂, was found to be 3.75 × 10⁻⁶, 2.35 × 10⁻⁵ and 1.444 × 10⁻⁶ U g⁻¹ min⁻¹ for NAC, FNAC1 and FNAC2, respectively. The regression coefficients for the first order kinetic model were 0.93, 0.88 and 0.86, respectively, for NAC, FNAC1 and FNAC2 and the regression coefficients for the second order kinetic model were 0.97, 0.99 and 0.96, respectively, for NAC, FNAC1 and FNAC2.

The data indicates that the immobilization of lipase onto NAC, FNAC1 and FNAC2 follows pseudo second order rate kinetics. The active sites of the enzymes are the determining factors in the carrier matrices.

The experimental q_e value was found to be 4080, 3340 and 2720 U g⁻¹ for FNAC1, FNAC2 and NAC, respectively. The inferred results showed that the immobilization of lipase onto NAC and FNAC1 follows the pseudo second order kinetic model and on FNAC2 it follows the pseudo first order kinetic model.

Kinetic parameters, K_m and V_max for free and immobilized lipase were determined by a Lineweaver–Burk plot (Fig. 3). The K_m and V_max values of purified lipase were found to be 9.50 mM and 1.03 mM min⁻¹, respectively. The K_m value represents a higher affinity between enzymes and substrates while V_max represents the higher catalytic efficiency of lipase.²⁶ The K_m value is low for the NAC-L indicating a higher affinity towards lipase. The reaction rate (V_max) was higher for FNAC2-L showing a higher catalytic efficiency of lipase in the immobilized form as shown in Table 3.

Fig. 3 Lineweaver–Burk plot of purified lipase and immobilized lipases using olive oil as the substrate.

Table 3 K_m and V_max values from the LB plot of the immobilized lipases

Immobilized lipases	Km^a (mM)	Vmax^a (mM min^−1)
a Experiments were done in triplicate.
NAC-L	3.57 ± 0.08	0.62 ± 0.03
FNAC1-L	4.34 ± 0.03	0.63 ± 0.01
FNAC2-L	4.76 ± 0.06	0.70 ± 0.13

---

The surface morphology of NAC, FNAC1 and FNAC2 is shown in Fig. 4 (a, c and e). The micrograph reveals that the NAC was highly porous in nature with the average pore diameter of 250 nm. The porous nature may be due to the chemical activation, e.g., following hydrofluoric acid treatment. The SEM images of FNAC1 and FNAC2 (Fig. 4c and e) reveal that the NAC was covered with an epitaxially grown deposit on the surface of the outer pores and the chemical deposit was uniform in nature with regards to the opacity. This may be regarded as an amino–aldehyde (Am–Al) chemical deposit on the NAC surface. The surface morphology images of the NAC-L, FNAC1-L and FNAC2-L are shown in Fig. 4b, d and f. The micrographs indicate that the enzyme molecules were well bound on the carrier matrices upon immobilization.

Fig. 4 SEM images of (a) NAC, (b) NAC-L, (c) FNAC1, (d) FNAC1-L, (e) FNAC2 and (f) FNAC2-L.

The FT-IR spectrum of NAC (Fig. 5a) has a wide band at 3383.53 cm⁻¹, due to the O–H stretching mode of aromatic rings and adsorbed water. The peak observed at 2358.3 cm⁻¹ is attributed to asymmetric stretching of the CH₃ group.

Fig. 5 FT-IR spectra of (a) NAC, (b) FNAC1, (c) FNAC2, (d) lipase, (e) NAC-L, (f) FNAC1-L and (g) FNAC2-L.

The FT-IR spectra of FNAC1 and FNAC2 (Fig. 5b and c) show the peak corresponding to the N–H stretching vibration of a secondary amine at 3397.14 and 3401.94 cm⁻¹, respectively. The significant increase in intensity of the bands at 1700 and 1704.76 cm⁻¹ in FNAC1 and FNAC2, respectively, correspond to C–O stretching vibrations of carboxyl or aldehyde groups. This may be due to the addition of ethylenediamine and glutaraldehyde. This confirms the stabilization of ethylenediamine and glutaraldehyde groups with NAC.

The FT-IR spectrum of lipase in Fig. 5d shows major protein bands (due to vibration of the peptide linkage) occurring in the spectral region of 1200–1700 cm⁻¹. The band at 1652.38 cm⁻¹ is due to the C=O stretching vibrations of amide I. The bands at 1233.33 and 1461.52 cm⁻¹ can be attributed to NH bending and C–N stretching vibrations.

The force constant of the stretching frequency is directly proportional to the strength of the bond. The IR stretching frequencies of C–O, N–H and C–N are 1100, 3300 and 1220 cm⁻¹. The force constants of C–O, N–H and C–N stretching are 489, 599 and 567 N m⁻¹.

The FT-IR spectrum of NAC-L (Fig. 5e) shows a band at 1652.38 cm⁻¹ that may be attributed to amide I of the immobilized enzyme. The peak corresponding to the C–O stretching vibration (1704.76 cm⁻¹) of the ketone and aldehyde groups was masked because of the binding of the NH₂ group of the enzyme with the aldehyde group of NAC. The wide peak observed at 3401.50 cm⁻¹ may be assigned to the N–H stretching vibration of a secondary amine. The force constants of N–H and C–O stretching are 636 and 1174 N m⁻¹. This increase in force constant indicates the strength of the strong binding between the enzyme and the matrices.

The FT-IR spectra of FNAC1-L and FNAC2-L (Fig. 5f and g) show a band at 1628.6 cm⁻¹ that may be attributed to amide I of the immobilized enzyme. The peaks corresponding to the C–O stretching vibration in FNAC1-L (1700 cm⁻¹) and FNAC2-L (1704.76 cm⁻¹) of the ketone and aldehyde groups were masked because of the strong binding of the NH₂ group of the enzyme with the aldehyde group of FNAC. The wide peak observed at 3395.96 cm⁻¹ in FNAC1-L and 3399.60 cm⁻¹ in FNAC2-L were due to the N–H stretching vibration of a secondary amine. The force constants of NH and CO stretching in FNAC1-L are 634 and 1167 N m⁻¹ and those in FNAC2-L are 635 and 1174 N m⁻¹. This increase in force constant indicates the strength of the strong binding between the enzyme and the matrices.

The shift in frequency of the amide I and secondary amine bands in FNAC1-L and FNAC2-L towards the lower frequency region and thus, a decrease in force constant is a clear indication of the delocalization of electrons in the N–H stretching of FNAC1-L and FNAC2-L. The delocalization of electrons increases the stability of the enzyme and thus the activity of FNAC1-L and FNAC2-L is higher than the NAC-L, however, it is less than the lipase in the free state.

XRD patterns of the NAC, FNAC1, FNAC2, free lipase, lipase immobilized NAC, lipase immobilized FNAC1 and lipase immobilized FNAC2 are shown in Fig. 6a–g, respectively. The XRD pattern of NAC shows the crystalline nature which was derived from the activation by hydrofluoric acid, followed by the removal of silica from the rice husks (Fig. 6a). On the other hand, the XRD pattern of MAC prepared in our earlier study from rice husks using phosphoric acid activation showed that the MAC was amorphous in nature and that the silica peak was observed at 22°. Fig. 6d shows the crystalline structure of lipase at 32° and 35°, which is also observed in the NAC-L sample (Fig. 6e), confirming the immobilization of lipase in NAC. The immobilization of lipase in FNAC1 and FNAC2 (Fig. 6f and g) was further confirmed from the XRD patterns, which showed peaks at 11.3, 12.9, 15.8, 18.0, 20.1, 25.05, 25.6, 26.6 and 30.9° due to surface functionalization, and peaks at 2θ values of 32° and 35°confirmed the presence of lipase.

Fig. 6 XRD patterns of (a) NAC, (b) FNAC1, (c) FNAC2, (d) lipase, (e) NAC-L, (f) FNAC1-L and (g) FNAC2-L.

The TGA of the free lipase (Fig. 7a) showed a major weight loss (47.73%) occurred from 107.98 to 481.45 °C due to decomposition of the major components of the lipase. At the end of the scan (800 °C), 41.43% of the sample remained, indicating the thermal resistance of the constituents of lipase.

Fig. 7 TGA and DTA of (a) lipase, (b) NAC, (c) NAC-L, (d) FNAC1, (e) FNAC1-L, (f) FNAC2 and (g) FNAC2-L.

TGA of NAC (Fig. 7b) showed that at 800 °C, there was only a 14.08% weight loss observed and 85.92% was left as residue. This suggests that the NAC was a very stable carrier matrix.

TGA of NAC-L (Fig. 7c) showed 11.53% and 14.77% of weight loss occurred at 278.60 °C and 359.17 °C, respectively. After this decomposition, there was a drastic decrease in weight loss (13.29%) and 71.94% was observed as residue at the end of the scan (800 °C). Moreover, the TGA of NAC-L showed about 13.98% weight loss compared to NAC, which could be due to the decomposition of the components of the enzyme molecules. The immobilized lipase was stable compared to free lipase, confirmed by 30.51% stability over the temperature ranges.

TGA of FNAC1 (Fig. 7d) showed a major weight loss (23.02%) occurred from 222.67 to 580.03 °C. At the end of the scan, i.e., at 800 °C, 63.69% was left as residue, indicating that the sample was less stable than NAC.

TGA of FNAC1-L (Fig. 7e) showed 10.93% and 33.2% of weight loss occurred at 195.18 °C and 525.05 °C, respectively. At the end of the scan, i.e., at 800 °C, 58.62% was left as residue. Moreover, the TGA of FNAC1-L showed about 5.07% weight loss when compared with the FNAC1, which may be attributed to the decomposition of the components of the enzyme molecules. The immobilized lipase was stable compared to free lipase, which was confirmed by the 17.19% stability over the temperature ranges. The FNAC1-L had 13.32% weight loss compared with the NAC-L.

TGA of FNAC2 (Fig. 7f) showed a major weight loss (27.36%) occurred from 194.24 to 533.58 °C.

TGA of FNAC2-L (Fig. 7g) showed 16.06% and 40.41% of weight loss occurred at 172.43 °C and 544.01 °C, respectively. Moreover, the TGA of FNAC2-L showed 6.49% of weight loss when compared with the FNAC2, which is attributed to the decomposition of the components of the enzyme molecules. The immobilized lipase was stable compared to free lipase, which was confirmed by the 14.13% stability over the temperature ranges. The FNAC2-L had a weight loss of 16.38% and of 3.06% compared with NAC-L (Fig. 7c) and FNAC1-L (Fig. 7e).

The DSC of lipase (Fig. 8a) showed a sharp thermal transition at 96.77 °C with an enthalpy of transition of 496.8 J g⁻¹ due to the weight loss of components present in the lipase. The NAC sample (Fig. 8b) gave rise to a thermal transition at 123.39 °C with an enthalpy of transition of 125.7 J g⁻¹. The DSC of NAC-L (Fig. 8c) showed a sharp thermal transition at 108.72 °C with an enthalpy of transition of 489.8 J g⁻¹.

Fig. 8 DSC of (a) lipase, (b) NAC, (c) NAC-L, (d) FNAC1, (e) FNAC1-L, (f) FNAC2 and (g) FNAC2-L.

The DSC of FNAC1 is shown in Fig. 8d. It shows there was a thermal transition at 118.09 °C with an enthalpy of transition of 203.9 J g⁻¹. The DSC of FNAC1-L (Fig. 8e) showed a sharp thermal transition at 124.72 °C with an enthalpy of transition of 298.32 J g⁻¹. The FNAC2 sample (Fig. 8f) gave rise to a thermal transition at 135.17 °C with an enthalpy of transition of 209.4 J g⁻¹. The DSC of FNAC2-L (Fig. 8g) showed a sharp thermal transition at 126.46 °C with an enthalpy of transition of 345.2 J g⁻¹.

These results suggested that the thermal stability of the lipase was enhanced due to the immobilization. The results also suggested that there was not much change in their enthalpy of transition, indicating that the nature of lipase was not being altered upon immobilization. When 50% of the lipase is unfolded at the higher thermal transition midpoint, the more stable the biomolecule. The order of thermal stability of lipase is FNAC2-L > FNAC1-L > NAC-L > lipase.

The various parameters such as time, temperature and pH for the hydrolysis process by free and immobilized lipases were optimized in the batch conditions. The optimum conditions for the hydrolysis of EOCW by free lipase was a time of 4 h, a pH of 7.0, and a temperature of 30 °C. At the optimum conditions, 92.3% of EOCW hydrolysis was achieved by free lipase. The optimum conditions for the hydrolysis of EOCW by the lipase immobilized matrices were a time of 3.5 h, a pH of 7.0, and a temperature of 30 °C. The % hydrolysis values of EOCW by FNAC1-L, FNAC2-L and NAC-L were found to be 98.43, 97.34 and 93.04%, respectively.

The hydrolysis of EOCW under continuous mode was carried out using immobilized lipases in a continuous packed column reactor. The study was carried out to determine the efficiency of immobilized lipases for large scale operation. Various parameters like time, pH, flow rate and volume of EOCW for the hydrolysis process by immobilized lipases were optimized under continuous mode. The optimum conditions for the efficient hydrolysis of oil-containing wastewater under continuous mode were observed to be a time of 4 h, a pH of 7.0, a flow rate of 1 ml min⁻¹ and a volume of EOCW of 20 ml (Fig. 9) at a fixed bed height of 5.5 cm, containing 15 g of carbon matrix. The results revealed that FNAC1-L, FNAC2-L and NAC-L hydrolyze oil-containing wastewater by 98.52%, 96.38% and 92.24%. The results from the kinetic models confirmed that the hydrolysis of EOCW by the immobilized lipase under continuous mode obeyed the second order rate kinetic model as greater R² values were observed (Table 4).

Fig. 9 Effect of (a) time, (b) pH, (c) flow rate and (d) volume of wastewater on the hydrolysis of EOCW under continuous mode.

Table 4 Kinetic rate constants for the hydrolysis of EOCW using the immobilized lipases in continuous mode

Carrier matrices	Pseudo first order		Pseudo second order
	k1	R^2	k2	R^2
NAC-L	0.0248	0.995	3.05 × 10^−4	0.993
FNAC1-L	0.0290	0.948	3.379 × 10^−4	0.987
FNAC2-L	0.0293	0.992	3.798 × 10^−4	0.997

---

The reusability of immobilized enzymes is very important for their application, especially from a commercial point of view. The reusability of the catalysts (NAC-L, FNAC1-L and FNAC2-L) was investigated to determine the stability of the immobilized lipases. FNAC1-L, FNAC2-L and NAC-L possessed 100% hydrolytic efficiency until the 15^th, 15^th and 10^th cycles, respectively. The immobilized matrices were washed with buffer solution (pH 7.0) at the end of each cycle. Also, it was found that there was no leakage of enzyme from the matrices. At the end of the 25^th cycle, FNAC1-L, FNAC2-L and NAC-L showed hydrolysis efficiency values of 98.86, 98.46 and 94.48%, respectively. FNAC1-L (89.78%) showed a higher operational stability than the FNAC2-L (87.36%) and NAC-L (76.59%) at the end of the 50^th cycle (Fig. 10). And also the lipase immobilized NAC and lipase immobilized surface functionalized NAC showed much higher operational stability than the results reported by Ramani et al.^15,24 The higher operational stability of lipase immobilized NAC may be due to the higher loading of lipase (5788 U g⁻¹) on NAC than on MAC⁴ and thus higher efficiency for the hydrolysis of EOCW and operational stability was achieved. The reduction in hydrolytic efficiency following repeated application cycles may be explained as follows. The immobilization of lipase on a carrier matrix is through strong interactions and also through weak intermolecular bonding, such as hydrogen bonding. The lipase immobilized through hydrogen bonding would have been easily detached as this bonding is weak and unstable, leading to poor hydrolytic efficiency. This was overcome by using the lipase immobilized surface functionalized NAC matrices (FNAC1-L and FNAC2-L). In this case, the operational stability was greatly enhanced without affecting their native active sites when compared to the NAC-L and lipase immobilized surface functionalized MAC.¹⁵ This could be due to the formation of a strong interaction between the lipase and the functional groups (amino–aldehyde) of FNAC1 (3395.96 cm⁻¹) and FNAC2 (3399.60 cm⁻¹) (figure not shown). The enzymes were not dislodged from the FNAC1 and FNAC2 following repeated application of the hydrolysis of EOCW. The strong interaction of lipase with the FNAC1 and FNAC2 matrices was also confirmed using FT-IR studies.

Fig. 10 Operational stability of FNAC1-L, FNAC2-L and NAC-L.

The study also proves that there is not much difference in the operational stability when using either FNAC1-L and FNAC2-L. This suggests that the strong interaction of lipase with the functional groups (Am–Al) of the carrier matrices is possible with or without the addition of sodium borohydride.

4. Conclusions

The main objective of the study was to treat the oil-containing wastewater from oil refining using lipase immobilized nanoporous and surface functionalized nanoporous activated carbons. The NAC was prepared from agricultural solid waste, rice husks, by precarbonization and chemical activation. The FNACs were prepared using ethylenediamine and glutaraldehyde with and without a reducing agent, sodium borohydride. The NAC, FNAC1 and FNAC2 were used for the immobilization of lipases and the maximum immobilization was achieved by using FNAC1 and FNAC2. The SEM images showed that the lipase was immobilized onto the carrier matrices. The immobilization of lipase was confirmed using FT-IR and XRD pattern analysis. The Michaelis–Menten enzyme kinetics confirmed that a higher affinity of the substrate with the immobilized lipases than the free lipase. The lipase immobilized NAC obeyed the Langmuir isotherm, confirming monolayer coverage of lipase onto NAC. The lipase immobilization in both FNAC1 and FNAC2 obeyed the Freundlich isotherm, confirming the multilayer formation of lipase onto FNAC1 and FNAC2. The immobilization of lipase onto both carrier matrices followed pseudo second order rate kinetic models. The immobilized matrices showed higher efficiency for the hydrolysis of EOCW under batch and continuous mode operations. The lipase immobilized carrier matrices had high operational stability up to 50 cycles of repeated applications of the treatment of EOCW. The study concluded that FNAC1-L and FNAC2-L are potential biocatalysts for the treatment of lipid-containing wastewater from the industrial sectors.

Acknowledgements

P. Saranya is thankful to the Council of Scientific and Industrial Research (CSIR) and the Central Leather Research Institute (CLRI), India, for awarding a senior research fellowship and providing all the facilities needed to carry out this work.

References

1. D. G. Crine, X. Paloumet, L. Bjornsson, M. M. Alves and B. Mattiasson, Renewable Energy, 2007, 32, 965–975.
2. M. P. Prasad and K. Manjunath, Indian J. Biotechnol., 2011, 10, 121–124.
3. D. T. Mitchell, S. B. Lee, L. Trofin, N. Li, T. K. Nevanen, H. Soderlund and C. R. Martin, J. Am. Chem. Soc., 2002, 124, 11864–11865.
4. U. Drechsler, N. O. Fischer, B. L. Frankamp and V. M. Rotello, Adv. Mater., 2004, 16, 271–274.
5. A. Dyal, K. Loos, M. Noto, S. W. Chang, C. Spagnoli, K. Shafi, A. Ulman, M. Cowman and R. A. Gross, J. Am. Chem. Soc., 2003, 125, 1684–1685.
6. X. J. Huang, Z. K. Xu, L. S. Wan, C. Innocent and P. Seta, Macromol. Rapid Commun., 2006, 27, 1341–1345.
7. Z. X. Tang, J. Q. Qian and L. E. Shi, Mater. Lett., 2007, 61, 37–40.
8. L. Wang, C. M. Li, A. J. Ryan and S. P. Armes, Adv. Mater., 2006, 18, 1566–1570.
9. Z. G. Wang, Z. K. Xu, L. S. Wan, J. Wu, C. Innocent and P. Seta, Macromol. Rapid Commun., 2006, 27, 516–521.
10. P. Ye, Z. K. Xu, J. Wu, C. Innocent and P. Seta, Macromolecules, 2006, 39, 1041–1045.
11. P. Ye, Z. K. Xu, J. Wu, C. Innocent and P. Seta, Biomaterials, 2006, 27, 4169–4176.
12. F. Lopez-Gallego, T. Montes, M. Fuentes, N. Alonso, V. Grazu, L. Betancor, J. M. Guisán and R. Fernández-Lafuente, J. Biotechnol., 2005, 116, 1–10.
13. P. K. Jal, S. Patel and B. K. Mishra, Talanta, 2004, 62, 1005–1028.
14. A. A. Mendes, R. C. Giordano, L. C. Giordano Rde and H. F. de Castro, J. Mol. Catal. B: Enzym., 2011, 68, 109–115.
15. K. Ramani, S. Karthikeyan, R. Boopathy, L. John Kennedy, A. B. Mandal and G. Sekaran, Process Biochem., 2012, 47, 435–445.
16. D. Samanta and A. Sarkar, Chem. Soc. Rev., 2011, 40, 2567–2592.
17. G. Bayramoglu, B. Altintas and M. Y. Arica, Appl. Microbiol. Biotechnol., 2013, 97, 1149–1159.
18. R. Fernandez-Lafuente, C. Rosell and L. Caanan- Haden, Enzyme Microb. Technol., 1999, 24, 96–103.
19. K. Ramani, P. Saranya, S. Chandan jain and G. Sekaran, Bioprocess Biosyst. Eng., 2013, 36, 301–315.
20. O. H. Lowry, N. J. Rosebrought, A. Farr and R. J. Randall, J. Biol. Chem., 1951, 139, 265–274.
21. K. Hayakawa, M. Kawaguchi and T. Kato, Langmuir, 1997, 13, 6069–6073.
22. S. Lagergren and B. K. Svenska, Veternskapsakad Handlingar, 1898, 24, 1–39.
23. Y. S. Ho and G. Mckay, Process Saf. Environ. Prot., 1998, 76, 183–191.
24. K. Ramani, R. Boopathy, C. Vidya, L. John Kennedy, M. Velan and G. Sekaran, Process Biochem., 2010, 45, 986–992.
25. E. Bulut and M. Ozacar, Microporous Mesoporous Mater., 2008, 115, 234–246.
26. R. Sharma, Y. Chisti and U. C. Banerjee, Biotechnol. Adv., 2001, 19, 627–662.

---