Green Synthesis, Characterization, and Catalytic Activity of Amine-multiwalled Carbon Nanotube for Biodiesel Production

An amine-functionalized multiwalled carbon nanotube (MWCNT) was prepared for use as a basic heterogeneous catalyst for the conversion of Cocos nucifera (coconut) oil and Hibiscus cannabinus (kenaf) oil to biodiesel. The 3aminopropyltrimethoxysilane (3-APTMS) was chosen to form an amine-reactive surface to bind with hydroxyl (−OH) and carboxyl (−COOH) groups of oxidized MWCNT. Silanization took place using a green surface modification method in which supercritical carbon dioxide fluid was utilized under the following conditions: 55 °C, 9 MPa, and 1 h. The synthesized catalyst was characterized using Thermogravimetric analysis (TGA), Fourier transform infrared (FTIR), Field emission scanning electron microscopy–energy dispersive x-ray (FESEM-EDX), Time-offlight secondary ion mass spectrometry (TOF-SIMS), X-ray powder diffraction (XRD), and Brunauer–EmmettTeller (BET). Transesterification of coconut oil using 10 wt% NH2-MWCNT catalyst (3 wt% APTMS), 12:1 molar ratio of methanol and oil at 63 °C for 1 h resulted in a >95% conversion. On the other hand, the same catalyst was used in the transesterification of kenaf oil, and formation of ammonium carboxylated salt was observed. The effects of temperature, pressure, and silane concentration on surface modification of MWCNT were evaluated in terms of the catalyst’s basic site density and fatty acid methyl ester conversion. The results indicate that reaction temperature and silane concentration had the most significant effects. Copyright © 2022 by Authors, Published by BCREC Group. This is an open access article under the CC BY-SA License (https://creativecommons.org/licenses/by-sa/4.0).


Introduction
The global energy demand has grown exponentially, and the continued use of non-petrodiesel. Over the past decades, it became well-known for its biodegradability, renewability, lower exhaust emissions, higher flashpoint for easier handling, and lubricity. Consequently, numerous studies are carried out to improve its quality, consistency, and long-term viability [2][3][4].
Transesterification is an equilibrium reaction describing the alcoholysis of carboxylic esters carried out in the presence of an acid or base catalyst to produce fatty acid methyl esters [5]. A heterogeneous base catalyst is often used in the transesterification of low free fatty acid oil and offers a more environmentally desirable option for producing biodiesel [6]. These solid base catalysts are reusable, regenerative, and less corrosive, leading to safer and more environmentally friendly operations. Some of the recently discovered solid base catalysts for biodiesel production are hydrotalcite, oxides of metal (e.g., CaO, MgO, or SrO), oxides of alkalidoped metal oxides (e.g., MgO/Al2O3, CaO/Al2O3, Li/CaO), oxides of mixed metal (e.g., Ca/Mg, Ca/Zn), alkali metal oxides (e.g., Na/NaOH/γ-Al2O3, K2CO3/Al2O3 reinforced on Al2O3) and magnetic composites [7][8][9].
In addition to many other novel base catalysts produced, an amine catalyst has also gained acceptance in homogeneous and heterogeneous transesterification of oils [10,11]. Amine, classified as a weak-base, has a lone pair of electrons on its nitrogen atom that can accept protons, such as H + and function as a Bronsted-Lowry base in a chemical reaction. Despite being classified as a weak base, it has shown a unique basicity strength which can be determined by measuring its logarithmic base dissociation constant (pKb) value. Primary amines have the lowest pKb values indicating a stronger base property than secondary and tertiary amines.
Amino-organosilane is one of the most commonly used primary amines, and 3a m i n o p r o p y l t r i e t h o x y s i l a n e a n d 3aminopropyltrimethoxysilane are examples of this chemical compound. Organosilane, an inexpensive coupling agent, is generally used as a surface modifier, primer, or adhesive [12]. Some of its other applications are in CO2 adsorption [13,14], heavy metal detection [15], interface reagent adhesion [16], synthesis of ethyl acetate [17], manufacturing of epoxy coatings for corrosion protection [18], anti-fouling thinfilm composite [19], and base catalyst for biodiesel production [20].
Studies that used aminosilane catalysts in biodiesel production are also found in the references [21,22]. Although significant results have been recorded, environmental harmful effects may occur during the catalyst production as a result of using toxic solvents, longer reaction time, and inclusion of more chemicals. To minimize environmental impacts from these activities, the use of fluids (e.g., water or CO2) in their supercritical states is examined and utilized in the field of green chemical processes. Presently, there are only a few discussions on the catalytic activity of a heterogeneous base catalyst synthesized with aminosilane and MWCNT under supercritical CO2 conditions for biodiesel production. Supercritical carbon dioxide (Tc = 31.10 °C; Pc = 7.39 MPa) is an inert, inexpensive, clean fluid, and a promising alternative to toxic volatile organic compounds such as toluene [23]. It also has zero surface tension, which effectively improves nanomaterial surface wetting, enabling chemical reaction, and surface modification [24]. Based on our previous research work, oxidized mercaptosilane was successfully grafted on MWCNT using scCO2, and a high free fatty acid oil was successfully catalyzed [25].
This study focuses on scCO2 grafting of a primary amine on MWCNT and its catalytic activity in converting two types of oil with substantially different acid values. The effects of temperature, pressure and silane concentration on the amount of amine grafted on the surface and its influence on biodiesel conversion are also presented. This research is a result of our extensive efforts to develop highperformance green catalysts using organosilane compounds.

Material
A pristine MWCNT (p-MWCNT) with a length of 10-30 mm and diameter of 20-30 nm was purchased from Chengdu Organic Chemicals Co., Ltd (China). The chemicals used in oxidation of p-MWCNT, such as hydrochloric acid (HCl, 36%) and nitric acid (HNO3, 70%), were procured from Wako Pure Chemical Industries (Japan). The chemicals used in surface m o d i f i c a t i o n , s p e c i f i c a l l y 3aminopropyltrimethoxysilane (3-APTMS, 98%) and carbon dioxide (CO2, 99.9% purity), were supplied by Wako Pure Chemical Industries (Japan) and Uchimura Sanso Co., Ltd (Japan), respectively. Inorganic chemicals such as boron trifluoride in methanol complex solution (49%-53% BF3, Tokyo Chemical Industry Co., Ltd), sodium chloride (NaCl, Wako Pure Chemical gen phthalate (KHP, Nacalai Tesque Inc.), and sodium hydroxide (NaOH, Wako Pure Chemical Industries) were used without further purification. Other solvents were also procured from Wako Pure Chemical Industries (Japan), such as n-hexane (C6H14), ethanol (C2H5OH), methanol (CH3OH), and hydrogen peroxide (H2O2, 30%). The coconut oil was purchased from Fujifilm Wako Pure Chemical Corporation, while kenaf oil was extracted from fresh ground seeds using ultrasound-assisted chemical solvent extraction method.

Oxidation of p-MWCNT
The p-MWCNT was first pre-treated using a concentrated acid mixture of 1:1 HNO3 and HCl using a modified chemical oxidation procedure [26]. Oxidation produced active sites and moieties on the surface of MWCNT, such as carboxylic (−COOH), carbonyl (−C=O), and hydroxyl (−OH) groups [27]. The acid-treated p-MWCNT was referred to as o-MWCNT in this study.

Surface Modification of o-MWCNT using 3-APTMS
The surface modification of o-MWCNT in scCO2 condition was conducted in a laboratory set-up shown schematically in Figure 1. A 100 mL beaker was initially loaded with 0.3 g of o-MWCNT and 25 mL of 3-APTMS (3 wt%). The mixture was ultrasonicated for 15 min at room temperature. It was subsequently transferred to a 100 mL stainless reactor in which it was heated at 55 °C and pressurized at 9 MPa for 1 h. Gradual depressurization of the system was performed after functionalization, and excess organosilane was collected in vials. The aminefunctionalized o-MWCNT, represented by NH2-MWCNT, was dried at 80 °C for 5 h. It was then washed with 100 mL C2H5OH, filtered through a funnel, and oven-dried at 80 °C for 2 h.

Characterization of Catalyst
The TGA was carried out using Perkin Elmer STA 6000 instrument. A specified amount of sample was placed in a ceramic crucible and heated to a temperature of up to 900 °C (heating rate of 20 °C/min) under flowing nitrogen of 50 mL/min. The morphological features and elemental composition of the samples were observed using a Dual Beam Helios Nanolab 600i FESEM-EDX analyzer. It was operated under an accelerating voltage of 2.0 kV, beam current of 86 pA, EDS accelerating volt-age of 10.0 kV, and beam current of 0.69 nA. The changes in the crystal structure or orientation of MWCNT samples, before and after surface modification, were examined using Rigaku miniflex 600 analyzer. The surface area (Brunauer-Emmett-Teller), pore-volume, and pore diameter of the samples were determined in Belsorp-mini II (ver.1.2.6) and BELMaster/BELSim (ver.2.3.2). Before measurements, all the samples were outgassed in a vacuum at 433 K and 10 −4 Pa for 6 h. FTIR spectra were recorded using the attenuated total reflectance accessory technique. Measurements were conducted in the wavenumber range of 4000-600 cm −1 and 20 scans. TOF-SIMS (TOF.SIMS 5 by IONTOF) analysis was done to identify the anions and cations present after the addition of 3-APTMS. The following parameters were observed during this analysis: beam of Bi + , energy of 3000 eV, analysis current of 0.8145 pA, raster mode of random, raster size of 128 pixels  128 pixels, and time of analysis of 200 seconds. The basic site density of the samples was determined by acid-base back titration method. A sample of 0.02 g of NH2-MWCNT and 10 mL of 0.02 M of HCl were mixed for 24 h. The catalyst was removed from the mixture by filtration, and 2 mL of filtrate was mixed with 4 mL of 0.02 M NaOH. The liquid solution was neutralized with 0.02 M HCl using phenolphthalein as an indicator.

Transesterification of Oil using NH2-MWCNT as Catalyst
Transesterification reactions were carried out in a 150 mL, three-necked glass-bottom flask equipped with a Graham condenser and a fiber optic thermometer. The flask was loaded with a 12:1 molar ratio of methanol to oil and NH2-MWCNT catalyst (0.1 g, 10% to oil mass). It was placed in a magnetic stirrer oil bath (AS ONE OBS 200 M) at a reaction temperature of 63 °C and a stirring rate of 600 rpm at 1 h. After reaching room temperature, the mixture was transferred into a vial and added with 15 mL of saturated NaCl solution, 5 mL of nhexane, and approximately 1 g of Na2SO4. The n-hexane was utilized to recover methyl esters from the mixture. Sodium chloride improved the separation of the n-hexane from alcohol layer, while Na2SO4 removed H2O traces. It was placed in a vortex mixer for 5 min and laboratory centrifuge (Tony low-speed centrifuge LCX 100) for 30 min at 5000 rpm.

Preparation of Standard Fatty Acid Methyl Ester
The standard coconut methyl ester was prepared according to the Association of Official Analytical Chemists (AOAC) 969.33 for fatty acids in oils and fats method [28]. After the separation of FAME from the mixture, a portion of the upper layer was collected and analyzed using Gas chromatography-Mass spectrometer (GC-MS) and a capillary column 19091S-433 HP-5MS (30 mm×0.25 mm, particle size: 2.25 µm). The GC-MS temperature program in this study is similar to a published work found in the references [25]. The coconut oil conversion, as FAME, was determined by comparing the area peaks of dodecanoic acid methyl esters (lauric acid methyl ester) from sample chromatograms. The calculation of % conversion was determined using Equation (1).
(1) On the other hand, it can be seen that NH2-MWCNT had three stages of weight loss at 30, 140, and 360 °C. At a temperature range of 30-140 °C, NH2-MWCNT had eliminated its moisture and ethanol solvent by up to 8% weight. A slightly moderate weight loss of 4% from 140 °C to 560 °C is associated with the removal of carboxylic and carbonyl functional groups. Finally, the NH2-MWCNT weight reduction from 560 °C to 900 °C is correlated with the continuous decomposition of chemically bonded organosilane compounds.

Fourier-transform infrared analysis
The FTIR spectrum of NH2-MWCNT in transmittance mode is shown in Figure 3. At wavelengths between 2625 cm −1 and 2950 cm −1 , the NH2-MWCNT spectrum showed changes in   % Transmittance, which are related with −OH (in carboxylic acids) stretching and −OH (in carboxylic acids) in plane bending, respectively. Also, a prominent peak was observed at 1372 cm −1 signifying the presence of the COO − group. All these observations confirm that MWCNT underwent surface modification by the addition of an acid mixture. The introduction of 3-APTMS, on the other hand, formed new peaks at 1521 cm −1 (NH3 + deformation) and 1063 cm −1 (C−N stretching). Other notable peaks appearing exclusively after organosilane addition in-clude those at 1271 cm −1 and 1000 cm −1 , matching to Si−CH3 symmetric deformation and antisymmetric Si−O−Si stretching, respectively. The identified functional groups have established the successful grafting of organosilane using scCO2 on the surface of NH2-MWCNT.
3.1.3 Field emission scanning electron microscopy -energy dispersive X-ray analysis The surface morphology of NH2-MWCNT at 100,000 magnification was observed using FESEM, as appeared in Figure 4. The effects of   nd -non detected during TOF-SIMS analysis oxidation (Figure 4(b)) and silanization ( Figure  4(c)) on the surface orientation of p-MWCNT (Figure 4(a)) are seen from these micrographs. Oxidation slightly loosened the carbon nanotubes but maintained their long and cylindrical-shaped structure. On the other hand, the addition of 3-APTMS made the MWCNT clustered together in bundles and groups. The organosilane appears to be a uniformly distributed material covering the surfaces and sides of each MWCNT bundle. The surface structure modification on MWCNT, after oxidation and silanization, is illustrated in Figure 5. During the oxidation process, hydroxyl (−OH) and carboxyl (−COOH) groups are introduced, and the −OH from these groups reacts with 3-APTMS during silanization.

Time-of-Flight Secondary Ion Mass Spectrometry Analysis
The lists of positive and negative ions present in p-MWCNT and NH2-MWCNT were summarized in Tables 1 and 2. Generally, more hydrocarbon cations and anions were identified after oxidation and addition of 3-APTMS. This result suggests the formation of intermediates in an organic reaction under scCO2 conditions.

X-ray diffraction analysis
The XRD analysis was performed to identify any modifications made in the crystallographic structure of p-MWCNT. In Figure 6, the XRD patterns of p-MWCNT, o-MWCNT, and NH2-MWCNT have sharp peaks at angle 2θ of 25.5° (C002), 43° (C100), 53° (C004), and 77° (C110), which are associated with the reflection of graphite [29]. This figure shows a decrease in the XRD profile intensity of the MWCNT samples after oxidation and silanization, indicating that new functional groups were successfully introduced and grafted.

Elemental composition
The presence of 3-APTMS was also confirmed using an EDX analyzer. Figure 7 illustrates the FESEM-EDX micrographs of NH2-MWCNT and the two selected areas used in describing the organosilane distribution on the surface. It shows that the organosilane is evenly distributed over the MWCNT, having the same wt% N on spectra 1 and 2 ( Table 3). On the other hand, elements, such as C, O, Si, Al,   Table 3. Elemental composition of MWCNT samples using FESEM-EDX analysis.

Bulletin of Chemical Reaction Engineering & Catalysis, 17 (2), 2022, 294
Copyright © 2022, ISSN 1978-2993 and Ni, and their corresponding wt% were summarized in Table 3. It is possible that Al and Ni-based compounds were employed in the production of p-MWCNT, and they were regarded as impurities of the sample in this study. The absence of these trace metals indicates a successful acid-oxidation treatment and washing of p-MWCNT.

Basic site density
The basic site density of NH2-MWCNT was measured using the standard acid-base back titration method. As shown in Figure 8, the basic site density increased from 1.165 mmol OH − /g to 4.498 OH − /g upon the addition of 3-APTMS. The increase in basic site density is due to the presence of amine groups (−NH2) from aminosilane that covalently bonded on the hydroxyl groups (−OH) of o-MWCNT. As estimated, the basic site density of o-MWCNT is much lower than p-MWCNT, owing to the presence of acidic functional groups on its surface. This result conforms with elemental analysis, where a higher %N concentration was observed after silanization.   Figure 9 shows the N2 adsorption-desorption isotherms of p-MWCNT, o-MWCNT, and NH2-MWCNT. All samples exhibited similar isotherms (Type IV) with evident hysteresis loops in the relative pressure about P/Po > 0.8. The presence of a hysteresis loop, over a high range of p/po values, indicates capillary condensation taking place at mesoporous particles (pore size = 2-50 nm). Among the three MWCNT samples, p-MWCNT had the highest adsorption capacity due to its highly porous surface, whereas NH2-MWCNT had the lowest adsorption capacity due to the added amine group on its surface [30]. This was confirmed by the textural and structural properties of MWCNT samples, summarized in Table 4. The oxidation of p-MWCNT caused a reduction in values of its surface area, pore-volume, and pore diameter. Oxidation created defects that cause interference among the nanotubes leading to a smaller surface area and pore diameter [31][32][33]. In addition, the carboxylic and amine functional groups had slightly taken up the unfilled volume on the pores of MWCNT, thus reducing the pore volumes of o-MWCNT and NH2-MWCNT. In comparing the o-MWCNT surface properties to NH2-MWCNT, the latter's surface area was reduced due to the presence of 3-APTMS, which clusters the nanoparticles in larger particle sizes.

Brunauer-Emmett-Teller analysis
In general, the surface area of catalyst decreased with increased addition of reactive functional groups, such as in other published literature where g-methacryloxypropyltrimethoxysilane was added to silica [34] and metal Sn to silicon dioxide [35]. Furthermore, the addition of organosilane filled up the smaller mesopores of MWCNT, causing an increase in the average pore diameter from 25.90 nm to 39.36 nm and a reduction in pore volume from 1.05 cm 3 /g to 0.06 cm 3 /g.

Transesterification of coconut oil
The catalytic activity of MWCNT samples was investigated by transesterification of coconut oil (no detected acid value) at 63 °C, 12:1 methanol to oil molar ratio, 10 wt% catalyst (3 wt% APTMS), and 1 h reaction time. Figure 10 shows the transesterification result, in which p-MWCNT and NH2-MWCNT recorded 23.6% (lowest) and 95.8% (highest) conversion, respectively. The use of o-MWCNT as catalyst resulted in a slightly higher conversion of 35.2% compared to p-MWCNT. The acidic functional groups of o-MWCNT catalyzed the transesterification of coconut oil. On the other hand, the highest % conversion is attributed to the presence of high basic sites in NH2-MWCNT. The 95.8% conversion is higher than most of the values published in the literature, summarized in Table 5. All solid catalysts from this table were synthesized using toluene as its solvent at a reaction time of approximately 24-48 hours.
The proposed four-step reaction mechanism for amine-catalyzed transesterification of coconut oil is shown in Figure 11. The first step is the reaction of the base catalyst and alcohol, forming protonated NH2-MWCNT and alkoxide. In this proton reaction transfer, the lone pair electrons on nitrogen form a new covalent bond with the hydrogen atom and produce the conjugate base of alcohol. This process is followed by the reaction of alkoxide and triglycer-

Bulletin of Chemical Reaction Engineering & Catalysis, 17 (2), 2022, 297
Copyright © 2022, ISSN 1978-2993 ides, forming a tetrahedral intermediate. Then, these tetrahedral intermediates produced the initial methyl ester and diglyceride anion. Finally, the base catalyst was recovered, and another catalytic cycle took place. The methyl esters produced using the NH2-MWCNT were identified and quantified using GC-MS. The chromatograms and % composition of coconut methyl esters are presented in Figure 12 and Table 6, respectively.
The scCO2 method of functionalization is proven effective in grafting 3-APTMS on o-MWCNT, even in a slightly low-pH environment. This low-pH condition is caused by a reversible reaction of alcohol and CO2 that produces carbonic acid and a lower pH environment [36]. In this study, the use of scCO2 has reduced the reaction time, solvent waste generation and eliminated the multi-step method of surface modification. Figure 11. Proposed reaction mechanism for transesterification of coconut oil.

Transesterification of Kenaf Oil
The NH2-MWCNT was also used as a catalyst in the conversion of kenaf oil (acid value = 116.20 mg KOH/g of oil) to biodiesel. After an hour of transesterification, carboxylate salts were formed, and no FAME was found when the sample was injected into GC-MS. The free fatty acids of kenaf oil reacted with amine and resulted in salts formation at 63 °C. Ammonium carboxylate salts are formed when there is an internal acid-base reaction of the amine and carboxylic units [38]. The proposed reaction mechanism of the three highest free fatty acid concentrations in kenaf oil is shown in Figure  13.

ScCO2 Functionalization Factors and their Effect on Coconut oil Transesterification
The operational parameters affecting the surface modification of NH2-MWCNT using scCO2, temperature, pressure, and silane concentration, were evaluated using one factor at a time experimentation method. The synthesized catalysts were used in the transesterification of coconut oil, and % conversion was calculated.

Temperature
Three different reaction temperatures of 40 °C, 50 °C, and 60 °C were selected for surface modification of NH2-MWCNT using 12 MPa and 1.5 wt% silane concentration. As shown in Figure 14, the increasing reaction temperature promotes the grafting of 3-APMTS on the MWCNT surface, resulting in a much higher basic site density. At higher reaction temperatures, the reaction rate and chances of a higher energy collision of the reactants increase. ScCO2 diffusivity may also have increased, resulting in a higher rate of mass transfer between the interfaces of the reactants and attachment of more aminosilane groups on MWCNT surface. This result was also observed in studies involving the use of silane in silicafilled tire tread compounds [39] and crosslinking of Ti-alloy [16], in which the degree of silanization is affected by the reaction temperature. On the same figure, the highest conversion of 69.6% was achieved at 50 °C, which corresponds to an NH2-MWCNT catalyst having a basic site density of 2.768 mmol OH − /g. The catalyst having the most number of attached amine groups may not have produced the highest % conversion, a validation that basic site density is only one of several factors affecting a catalyst performance.

Pressure
The effect of pressure on the amount of 3-APTMS grafted on MWCNT is presented in Figure 15. The increase in basic site density was observed from 10 MPa to 12 MPa and slightly decreased as it shifted to 14 MPa. The addition of pressure increased the solvent den- Figure 14. Effect of temperature on basic site density and NH2-MWCNT catalytic activity. Supercritical CO2 functionalization conditions: temperature -40, 50, 60 °C, pressure -12 MPa, wt% silane -1.5 wt%, time -1 h. Transesterification conditions: feedstock -coconut oil, temperature -63 °C, wt% catalyst -10 wt% (1.5 wt% APTMS), methanol to oil ratio -12:1, time -1 h. Figure 15. Effect of pressure on basic site density and NH2-MWCNT catalytic activity. Supercritical CO2 functionalization conditions: pressure -10, 12, 14 MPa, temperature -50 °C, wt% silane -1.5 wt%, time -1 h. Transesterification conditions: feedstock -coconut oil, temperature -63 °C, wt% catalyst -10 wt% (1.5 wt% APTMS), methanol to oil ratio -12:1, time -1 h. sity and solvent power of scCO2 [40]. This condition allows a higher mass transfer rate of 3-APTMS in ethanolic solution and MWCNT. However, at a pressure range of 12 MPa to 14 MPa, the basic site density decreased from 2.951 mmol OH − /g to 2.725 mmol OH − /g. The increase in pressure resulted in CO2 and alcohol interaction forming a carbonic acid that lowers the pH of a mixture [41,42]. The slight change in the pH may have affected the continuous grafting of aminosilane on MWCNT surface, which naturally occurs in an alkaline mixture. On the same figure, the highest conversion of 65.39 % was achieved at 12 MPa, which corresponds to an NH2-MWCNT catalyst having a basic site density of 2.951 mmol OH − /g.

Silane Concentration
To investigate the effect of silane concentration on surface modification of NH2-MWCNT, three different concentrations (1 wt%, 1.5 wt%, and 2 wt%) were used during scCO2 functionalization of o-MWCNT at 50 °C and 12 MPa. The prepared silane concentration is lower than in the previous experiments to observe closely its effect in the transesterification of oil. As illustrated in Figure 16, increasing the silane concentration from 1% to 2% increases the basic site density of NH2-MWCNT. The highest recorded basic site density is 3.099 mmol OH -/g upon the addition of 2 wt% 3-APTMS. Moreover, the highest conversion of 70.54% was achieved at the same silane concentration.
The relation between two dependent variables-basic site density and % conversion, is shown in Figures 14-16. Based on the experiments conducted, a catalyst with the highest basic site density does not always result in a high biodiesel conversion. The high catalytic performance of NH2-MWCNT depends on the number of reactive amine groups, surface area, and pore diameter. The catalyst's pore structure is a critical requirement in biodiesel production since a standard triglyceride molecule has a pore diameter of about 5.8 nm. Moreover, a large surface area increases the reaction rate due to a greater chance of reactant particle collision [43].

Reusability of NH2-MWCNT
The base catalyst was reused for three cycles in transesterification of coconut oil under the following conditions: 10 wt% catalyst loading (3 wt% of 3-APTMS), 63 °C reaction temperature, 12:1 methanol to oil molar ratio, and 1 h reaction time. Solvents such as methanol and n-hexane are alternately used to wash NH2-MWCNT and remove unreacted oils, glycerol, and other impurities. It was then dried at 80 °C for 300 min and allowed to cool at room temperature before further use. The summary of the reusability test results of NH2-MWCNT can be seen in Figure 17.
It shows in the figure the decrease in NH2-MWCNT catalytic activity after the second (13.5%) and third (20.6%) transesterification reactions. The reduction of % conversion could be attributed to the possible leaching of amine functional groups and blocking of active sites Figure 16. Effect of silane concentration on basic site density and NH2-MWCNT catalytic activity. Supercritical CO2 functionalization conditions: wt% silane -1, 1.5, 2 wt%, temperature -50 °C, pressure -12 MPa, time -1 h. Transesterification conditions: feedstock -coconut oil, temperature -63 °C, wt% catalyst-10 wt%, methanol to oil ratio -12:1, time -1 h. Figure 17. Reusability of NH2-MWCNT. Transesterification conditions: feedstock -coconut oil, temperature -63 °C, wt% catalyst = 10% wt (3 wt% APTMS), methanol to oil ratio -12:1, time -1 h. by adsorbed intermediate products, methyl esters, and glycerol on MWCNT. The leaching of amine functional groups can be addressed by increasing the silanization temperature and prolonging the silanization time during the catalyst preparation. It has been proven in other silane-related studies the positive effect of increasing reaction temperature and reaction time [39]. On the other hand, optimizing the transesterification process parameters (e.g., reaction temperature, reaction time, methanol to oil molar ratio, catalyst loading, and mixing speed) will increase catalytic activity at a shorter reaction time, preventing the adsorption of intermediate products into the surfaceactive sites of NH2-MWCNT.

Conclusions
This study focuses on a heterogeneous base catalyst prepared using supercritical CO2, and effectively used in the production of coconut methyl esters (>95% conversion). The application of scCO2 resulted in a milder method of catalyst preparation due to the elimination of toxic solvents (e.g. toluene), multi-step process, and longer reaction time. This study also reaffirmed that the basic site density of NH2-MWCNT is just one of the several factors affecting its catalytic activity. The surface area, pore-volume, and pore diameter at different process parameters during scCO2-silanization must be further examined to optimize its catalytic performance. It was proven that the supercritical CO2 method in grafting amine groups through silanization on multiwalled carbon nanotubes produces an effective catalyst for transesterification of low fatty acid oil.