Abstract
In this study, the deoxygenation pathway was proposed to eliminate oxygen species from biomass-derived oil, thereby producing a high quality of hydrocarbon chains (green fuel). The catalytic deoxygenation reaction of bio-oil model compound (oleic acid) successfully produced green gasoline (C8–C12) and diesel (C13–C20) via activated hydrotalcite-derived catalysts (i.e. CMgAl, CFeAl, CZnAl and CNiAl). The reaction was performed under inert N2 condition at 300 °C for 3 h, and the liquid products were analysed by GC–MS and GC–FID analyses to determine the hydrocarbon yield and product selectivity. The activity of the catalysts towards the deoxygenation reaction presented the following increasing order: CNiAl > CMgAl > CZnAl > CFeAl. CNiAl produced a hydrocarbon yield of up to 89 %. CNiAl demonstrated the highest selectivity with 83 % diesel production, whereas CMgAl showed the highest gasoline selectivity with 30 %. These results indicated that catalysts with a high acidic profile facilitate C–O cleavage via deoxygenation, producing hydrocarbons (mainly diesel-range hydrocarbons). Meanwhile, highly basic catalysts exhibit significant selectivity towards gasoline-range hydrocarbons via cracking and lead to the occurrence of C–C cleavage. The large surface area of CNiAl (117 m2 g−1) offered high approachability of the reactant with the catalyst’s active sites, thereby promoting high hydrocarbon yield. Consequently, the hydrocarbon yield and selectivity of the deoxygenation products were predominantly influenced by the acid–base properties and structural behaviour (porosity and surface area) of the catalyst.
Introduction
Environmental degradation caused by toxic fumes (i.e. SOx and NOx) originating from the transportation and industrial sectors, which are powered by non-biodegradable petroleum fuel, has become a major issue for decades. Alternative fuels derived from non-food grade triglycerides and fats provide a feasible solution for these problems due to their renewability and clean energy [1], [2]. The most common pathway to improve the properties of triglycerides and fats is via transesterification/esterification reactions, which produce a renewable fuel known as biodiesel or fatty acid methyl ester (FAME) [3], [4]. Although the addition of FAME into fossil diesel results in a high cetane number, it contains high O-containing species that can lead to issues in oxidative instability. Furthermore, utilisation of FAME results in high emissions of oxygenated gases (CO2 and CO) to the atmosphere, causing a detrimental effect to the environment [5], [6], [7], [8]. Hydrodeoxygenation (HDO) is preferred because it can remove the O-containing species from triglycerides and fats, producing a long-chain paraffin within diesel-range fuel [9], [10], [11]. Although the HDO reaction is an ideal process to produce high yields of diesel hydrocarbons, the HDO mechanism is costly because it involves high H2 consumption [12], [13]. Thus, the deoxygenation (DO) pathway is highly promising for the development of diesel-range hydrocarbons via low-cost decarboxylation and decarbonylation routes. Similar to HDO, DO is a process of eliminating O-containing species from triglycerides and fats but without H2. The O-containing species are removed in the form of CO2, CO and water. The DO process does not only produce diesel-range hydrocarbons, which exhibit similar properties to petroleum fuel, but it is also an economical process because it operates without H2 supply.
Numerous solid catalysts have been studied, including noble metal catalysts (Pd and Pt), sulphide and zeolite catalysts [14], [15], [16]. Although noble metal catalysts offer high affinity towards the production of diesel, the high cost of noble metal catalysts makes them unappealing and limits their application for industrial use. The use of sulphide catalysts leads to the formation of undesirable sulphur-containing products and affects the quality of oil, which does not act in accordance with the demand for clean fuel. Highly acidic catalysts, such as mesoporous zeolite-containing catalysts, generally yield high deoxygenised hydrocarbons. Unfortunately, such hydrocarbons result in a wide spectrum of hydrocarbon products. In general, highly acidic catalysts can be deactivated easily by coke formation, diminishing deoxygenation activity. Given the drawbacks of the aforementioned catalysts, the exploration of low-cost, non-sulphated and mild acidic catalysts for high yields of diesel-range hydrocarbon fractions has gained recent popularity. Interestingly, previous studies have discovered that highly basic catalysts do not only act as coke inhibitors but also improve the rate of the decarboxylation reaction mechanism [17].
Hydrotalcite is a well-established heterogeneous basic catalyst. It is represented by the general formula [M(II)1−xM(III)x(OH)2]x+[An−]x/n·mH2O, where M(II) and M(III) are divalent and trivalent metal ions such as Mg2+, Mn2+, Ni2+, Co2+, Zn2+, Cu2+ and Fe2+ and Al3+, respectively, and An− is an intercalated anion (such as Cl−, NO3−, CO32− or other macrocyclic multivalent anion). The structure of hydrotalcite is based on layered double hydroxide (LDH). It can be synthesised with a tuneable number of acid and basic sites and has ion exchange characteristic, which is preferable for DO. The increasing number of studies on hydrotalcites as catalyst and support indicates that hydrotalcite is an active catalyst for DO [18], [19], [20], [21]. This result is supported by findings from Na et al., who proposed that hydrotalcite-containing catalysts (Mg/Al) promote decarboxylation by converting oleic acid to octane, nonane and heptadecene hydrocarbons [22]. Despite its high decarboxylation activity, the high basicity of MgO led to the formation of solid oleic acid (saponification). Therefore, several metals such as Ni, Fe and Zn were considered to substitute Mg as M2+, and catalysts were speculated to show high tolerance towards feedstock with high free fatty acid (FFA) content. Ni has notorious ability in cracking of oil due to its high acidity [23]. Meanwhile, Fe species exhibits redox properties due to its strong oxygen affinity, so it can bind to oxygen from C=O of oleic acid [24]. The moderate basic strength and high selectivity of Zn under mild conditions demonstrate its remarkable potential in DO [25], [26]. Given the high basicity of hydrotalcite-like catalysts, coke formation by the acid sites can be minimised during the DO reaction.
Thus, the present work aimed to produce green diesel from oleic acid by using hydrotalcite-like catalysts (Fe, Zn, Ni and Mg). The effects of the physicochemical properties of the synthesised hydrotalcite-like catalysts (basicity–acidity, surface area, crystallite sizes and surface morphology) on DO activity were investigated. The reusability of the hydrotalcite-like catalysts was evaluated by batch experiments, and the reasons for the deactivation of the catalysts were discussed.
Experimental
All metal salts for hydrotalcite synthesis, namely, Al(NO3)3·9H2O (98.5%), Zn(NO3)2·6H2O (98%), Ni(NO3)2·6H2O (97%), Fe(NO3)2·9H2O (98%) and Mg(NO3)2·6H2O (99%), and Na2CO3 (99.5%) were obtained from R&M Chemicals Sdn. Bhd. NaOH (99%) was purchased from Merck. The internal standard 1-CH3Br (>98%) and standard solution for alkane and alkene (C8–C20) were purchased form Sigma–Aldrich. Hexane (GC grade, >98%) was also obtained from Merck for dilution purpose. Oleic acid (99.5%) was acquired from QRec. The physicochemical properties of oleic acid are tabulated in Table 1 [27].
Physicochemical properties of oleic acid [27].
| Properties | Description | Method |
|---|---|---|
| Oleic acid | ||
| Molecular formula | C18H34O2 | |
| Decomposition temperature (°C) | 239.37 | TGA analysis |
| Density (g cm−3) | 0.89 | ASTM D1298 |
| Viscosity at 40°C (cSt) | 4.5 | ASTM D445-15a |
| Moisture content (wt.%) | ≤0.2 | AOCS Ja 2b-87 |
| Acid value (mg KOH g−1) | 196–204 | AOCS Te 1a-64 |
| FFA value (%) | 98.5–102.5 | AOCS Te 1a-64 |
| Composition of oil (%) | GCMS analysis | |
| Oleic acid | 99.54 | |
| Hydroperoxide | 0.07 | |
| n-Decanoic acid | 0.03 | |
| Dodecanoic acid | 0.06 | |
| Heptadecane | 0.04 | |
| n-Hexadecanoic acid | 0.10 | |
| 9-Hexadecen-1-ol | 0.15 | |
Preparation of catalyst
Hydrotalcite-like catalysts M6Al2CO3(OH)16·4H2O (M=Mg, Zn, Fe, Ni) were synthesised via co-precipitation. The synthesis involved a solution containing M salt and Al(NO3)3·9H2O mixtures in a molar ratio of 4:1. The salt mixture was precipitated with 2 M NaOH and Na2CO3 in a molar ratio of 1:1 at a rate of 1 mL min−1 at room temperature under continuous vigorous stirring. The final pH was 11. The gelatinous slurry formed was aged at 70°C for 18 h, centrifuged and washed by deionised water until the pH was maintained at 7. The precipitate was dried at 100°C and thermally activated at 500°C for 3 h. The hydrotalcites formed were denoted as ZnAl, NiAl, FeAl and MgAl, whereas the calcined hydrotalcites were denoted as CZnAl, CNiAl, CFeAl and CMgAl. Synthesised alumina was prepared to compare reactivity [28].
Catalyst characterisation
The thermal decomposition profiles of the hydrotalcites were studied using TGA model Q500 V20.13 analysers. Samples in the crucible were heated to 900°C at a rate of 10°C min−1 under inert condition (N2). XRD analysis was performed by using a Shimadzu XRD-6000 diffractometer at a scanning rate of 2° min−1 (2θ=20°–80°). The catalyst was scanned by Cu Kα radiation, which was generated by Philip glass diffraction X-ray tube (broad focus 2.7 kW type). The pattern obtained was compared with the Joint Committee on Powder Diffraction Standards (JCPDS) files based on the peaks’ positions and intensities. The specific surface area and pore sizes of the catalyst were measured by using a Thermo Finnigan Scorptomatic 1900 model. The adsorption and desorption of N2 on the degassed catalyst surfaces were carried out at the temperature of liquid nitrogen (−196°C). The specific surface area was based on Brunauer–Emmet–Teller (BET) theory, while pore sizes distribution was based on Barrett, Joyner, and Halenda (BJH) method. A Philips FEI Field Emission Scanning Electron Microscope model Quanta FEG 450 was used to capture the surface morphologies of the catalysts. Solid samples were coated with gold by BIO-RAS Sputter prior to analysis. The images were then captured at scales of 1 and 5 μm. The TPD-CO2 and TPD-NH3 analyses were carried out by using Thermo Finnigan TPD/R/O 1100 instrument equipped with a thermal conductivity detector (TCD). Pretreated catalysts were exposed to CO2 and NH3 gases for 1 h at ambient temperature to allow the adsorption of CO2 and NH3 onto the surfaces. During analysis, the temperature was increased gradually from 50°C to 900°C. The desorption of CO2 and NH3 was determined by TCD, which produced a spectrum.
Catalytic deoxygenation of oleic acid
The deoxygenation of oleic acid was carried out in a simple glass reactor equipped with a temperature detector and stirrer. In a typical experiment, about 10 g of oleic acid and 5 wt.% catalyst were added into the reactor. The DO reaction of oleic acid was performed by heating the reaction medium to 300°C under inert condition (N2 flow=15 cc min−1) for 3 h. The deoxygenated product was condensed and collected in the receiving flask. The reactor was cooled to room temperature using an external water circulation cooling system, and the deoxygenated liquid products were further analysed by gas chromatography flame ionisation detection (GC–FID) and gas chromatography–mass spectrometry (GC–MS).
Product analysis
The liquid DO products were characterised by GC–MS to investigate the composition of each sample. A GC–MS analyser (Shimadzu QP2010) was equipped with a TCD and RTX-5MS column (30.0 m×0.25 μm×0.25 mm) with splitless inlet. Helium gas was used as carrier gas with the flow rate of 0.80 mL min−1. The temperature of the column was initially set to 50.0°C, increased gradually to 300°C at a rate of 3°C min−1 and held for 30 min at the final temperature. The samples dissolved in hexane (Merck GC grade, ≥98%) were injected at 250°C. The GC–MS peaks were analysed and identified using the National Institute of Standards and Testing-11 (NIST11) library. GC–FID analysis was carried out to determine the total hydrocarbon yield n-(C8–C20) and product selectivity. The GC–FID analyser (Agilent technologies 7890 A) was equipped with an HP5 capillary column (30.0 m×0.25 μm×0.32 mm). CH3Br was used as an internal standard, and both saturated and unsaturated hydrocarbon standard solutions were used as standard. An aliquot of 1 μL of sample was injected into the GC column. The injection temperature was 250°C. Nitrogen gas was the carrier gas. The initial temperature of the oven was set at 40°C, held for 6 min and raised to 260°C at the heating rate of 7°C min−1. The total reacted product (X) and yield of hydrocarbon (Y) were determined by Eqs. (1) and (2) [27], [29], [30]:
where at=area of total reacted (product) and unreacted oleic acid after DO, a0=area of total reacted oleic acid (product) after DO, an=area of alkane hydrocarbon and am=area of alkene hydrocarbon. Meanwhile, the selectivity (S) of the hydrocarbon product was determined by Eq. (3)
where no=area of the selected range of carbon number; gasoline: C8–C12, diesel: C13–C20.
Results and discussion
Physicochemical properties of hydrotalcite-derived catalyst
Figure 1a shows the thermal decomposition patterns of uncalcined parent hydrotalcites M6Al2CO3(OH)16·4H2O (M=Mg, Zn, Fe, Ni). The parent hydrotalcites exhibited two major decomposition weight losses at 100°C–200°C and 300°C–450°C, corresponding to ~10 wt.% of water molecules from the surface of hydrotalcite precursor and ~30 wt.% of metal nitrate decomposition [31], [32]. All the uncalcined parent hydrotalcites were fully decomposed at >500°C for the formation of mixed metal oxides [ZnO-Al2O3, FeO-Al2O3 and NiO-Al2O3]. These results were in agreement with the findings of XRD analysis of uncalcined and calcined parent hydrotalcites (Fig. 1b and c). The XRD results of uncalcined parent hydrotalcites (Fig. 1b) demonstrated that Al exhibited peaks at 2θ=22.8°, 35.3°, 38.9°, 42.5°, 48.3°, 55.6° and 66.6°, which corresponded to the presence of diaspore AlO(OH) (JCPDS file no.: 96-100-8751). The incorporation of Mg metals resulted in reflections at 2θ=11.4°, 22.8°, 34.3°, 38.0°, 46.6°, 60.3° and 61.7° (JCPDS file no.: 41-1428), which corresponded to the formation of the rhombohedral crystal system of MgAl hydrotalcites [33]. Similarly, the hydrotalcite phases appeared with the introduction of Ni, Fe and Zn into Al, thereby proving the replacement of Mg2+ cation by Zn2+ (JCPDS file no.: 38-0486), Ni2+ (JCPDS file no.: 15-0087) and Fe2+ cations into the rhombohedral system of hydrotalcites, respectively. Thus, the hydrotalcite structure of MgAl was converted from Mg6Al2CO3(OH)16·4H2O to Zn6Al2CO3(OH)16·4H2O, Fe6Al2CO3(OH)16·4H2O and Ni6Al2CO3(OH)16·4H2O. All uncalcined parent hydrotalcites showed intense peaks at 2θ=29.4° and 31.9°, which corresponded to the high crystallinity of Al2O3·H2O (JCPDS No.: 00-001-447).
(a) TGA and (b) XRD profiles for uncalcined parent hydrotalcites and (c) calcined parent hydrotalcites (hydrotalcite-derived catalysts). *A=AlO(OH), M=Mg6Al2CO3(OH)16.4H2O, Z=Zn6Al2CO3(OH)16.4H2O, F=Fe6Al2CO3(OH)16·4H2O, N=Ni6Al2CO3(OH)16·4H2O, G=γ-Al2O3, P=Periclase – MgO, C=Zincite, ZnO, H=Hematite, Fe2O3, X=NiO and o=Al2O3·H2O.
The calcined Al catalyst at 500°C exhibited the presence of the γ-Al2O3 phase, making it preferable as a catalyst due to its high activity [34]. The hydroxide phase from M6Al2CO3(OH)16·4H2O (M=Mg, Zn, Fe, Ni) was eliminated at 500°C, which was in accordance with the TGA findings. Thus, thermal treatment at 500°C was sufficient to decompose hydroxide and carbonate compounds from the catalyst. The incorporation of Mg to Mg/Al (CMgAl) resulted in the presence of cubic periclase belonging to the MgO phase at 2θ=42.9° and 62.6° (JCPDS file no.: 45-0946). By contrast, ZnAl was involved in the formation of pure zincite with the ZnO phase at 2θ=31.8°, 34.5°, 36.2°, 47.5°, 56.6°, 62.8° and 68.0° (JCPDS file no.: 96-900-4179). Reflections of cubic NiO from the CNiAl catalyst appeared at 2θ=37.3°, 43.4° and 63.2°. The XRD pattern of CFeAl mainly corresponded to the Fe2O3 phases with the peaks present at 2θ=33.4 °, 36.0°, 41.1°, 49.8 °, 54.5°, 62.8° and 64.6° (JCPDS No.: 2-0919). The crystallite size of the calcined parent hydrotalcites was calculated by the Debye–Scherrer equation based on the center of the peak with the highest intensity reflected at 2θ=29.4° (Table 2). The average crystallite size of all materials decreased in the following order: CAl (9.6 nm)>CMgAl (6.0 nm)>CZnAl (3.7 nm)>CFeAl (1.7 nm)>CNiAl (1.5 nm). The addition of acidic active metal (Ni) indicated the smallest crystallite size relative to basic (Mg) and acidic–basic (Zn, Fe) active metals, which showed that acidic active metals displayed high resistance towards the sintering process during calcination. A small crystallite size should provide additional benefits to the deoxygenated catalyst because it can prevent the formation of undesired carbon residue/coke during DO [35].
Crystallite sizes, textural properties and acidity and basicity profiles of the synthesised catalysts.
| Catalyst | XRD Crystallite size (nm)a |
BET Surface area (m2 g−1)b |
BJH |
TPD-CO2 |
TPD-NH3 |
|||
|---|---|---|---|---|---|---|---|---|
| Pore diameter (nm)c | Pore volume (cc g−1)c | CO2 desorption temperature (oC)d | Amount of CO2 desorbed (μmol g−1)d | NH3 desorption temperature (°C)e | Amount of NH3 desorbed (μmol g−1)e | |||
| CAl | 9.6 | 122 | 4 | 0.24 | 98 | 40.30 | 261 | 481.04 |
| CMgAl | 6.0 | 106 | 34 | 1.10 | 704 | 9002.92 | 422 | 1388.62 |
| CZnAl | 3.7 | 98 | 37 | 0.26 | 686 | 8638.51 | 653 | 5943.78 |
| CFeAl | 1.7 | 95 | 10 | 0.38 | 698 | 2336.79 | 523 | 1396.56 |
| CNiAl | 1.5 | 117 | 5 | 0.21 | 647 | 2045.01 | 635 | 6050.22 |
aDetermined by using Debye–Scherrer equation.
bDetermined by BET analysis.
cDetermined by BJH analysis.
dDetermined by TPD-CO2 analysis.
eDetermined by TPD-NH3 analysis.
The BET surface area and BJH porosity profiles of the catalysts are tabulated in Table 2. CAl had the highest surface area (122 m2 g−1) compared with the M-Al (M=Mg, Ni, Zn, Fe) catalysts, and the surface areas decreased in the order of CAl>CNiAl>CMgAl>CZnAl>CFeAl. The reduction in surface area of the hydrotalcite catalysts was attributed to the incorporation of Ni, Mg, Zn and Fe metal oxides with Al2O3. The metals (M=Ni, Mg, Zn and Fe) incorporation with the mesoporous Al2O3 lead to the formation of hydrotalcite-like catalysts with the reduction of surface area, as compared to the unreacted Al2O3 due to the collapsing of the HT structure during the calcination (sintering process), which is consistent with the XRD characterization result [36]. CNiAl produced the highest surface area (117 m2 g−1) amongst the hydrotalcite catalysts, which indicated that Ni presented the high accessible surface for DO activity. Surface areas showed no correlation with crystallite sizes of the catalysts (refer to XRD results). In terms of porosity, all the calcined M-Al catalysts resulted in larger pore diameters (10–37 nm) than CAl (5 nm), proving that the incorporation of Mg, Ni, Fe and Zn into Al led the pore walls of the catalysts to collapse. Furthermore, the occurrence of internal pore strain between the metal oxides (MO-Al2O3) also contributed to the increment of the pore sizes of catalysts. CZnAl exhibited the largest pore diameter (37 nm), which was strongly due to the dispersion of ZnO onto Al2O3, resulting in the disruption of the pore lattice and destruction of pore walls during the sintering process [36]. In summary, the pore volumes decreased in the order of CMgAl>CFeAl>CZnAl>CAl>CNiAl. The large pore volume of the CMgAl catalyst corresponded to the aggregation of plate-like particles, which were also attributed to the large pore diameter [37]. On the basis of the BET results, CNiAl was predicted to exhibit high catalytic activity towards DO, thereby producing a high hydrocarbon yield due to its high surface area corresponding to the catalyst’s high surface accessibility.
The variation in surface morphology for the series of modified hydrotalcites was determined by FESEM analysis (Fig. 2). CAl showed a compact surface of γ-Al2O3 particles with a notable rectangular shape (Fig. 2a). However, the FESEM morphology for CM-Al (M=Mg, Ni, Zn, Fe) catalysts varied significantly, suggesting that the surface of the catalyst was dependent entirely on the nature of active metal used. A curly, multi-layer and plate-like surface with a visible pore structure was observed on the CMgAl catalyst (Fig. 2b), which was consistent with its large pore characteristics (refer to BJH results). The CZnAl surface presented rough surface morphology decorated with a uniform size (50–200 nm) of highly visible crystal structure of hexagonal symmetry (Fig. 2c), which corresponded to the hexagonal phase structure of ZnO. This result was in agreement with the XRD findings. By contrast, a fine fibrous-like structure was perceived by the CFeAl catalyst (Fig. 2d). CNiAl exhibited disproportionate surfaces without a visible pore structure (Fig. 2e), which was related to the pore characteristics of the catalyst (refer to BJH results). Therefore, hydrotalcites have rough and irregular morphological surfaces depending on the metal incorporated (M=Mg, Ni, Zn, Fe).
FESEM image of (a) CAl, (b) CMgAl, (c) CZnAl, (d) CFeAl and (e) CNiAl.
The effects of the incorporation of Mg-, Zn-, Fe- and Ni-active metals with Al on the catalysts’ basic strengths and amount of basicity were analysed by TPD-CO2 (Table 2 and Fig. 3a). All activated catalysts (CAl, CMgAl, CZnAl, CFeAl and CNiAl) resulted in CO2 desorption peaks >500°C, so they were classified as strongly basic catalysts [30]. The basic strengths of the catalysts followed the increasing sequence of CMgAl>CFeAl>CZnAl>CNiAl>CAl, whereas the amount of basicity followed the order of CMgAl>CZnAl>CFeAl>CNiAl>CAl. A strong basic strength with large basic density was exhibited over the CMgAl catalyst, corresponding to Mg2+-O2- pairs from hexagonal MgO surfaces as presented in XRD analysis [38]. CNiAl showed the lowest basic characteristic amongst the modified hydrotalcites due to the surface lattice oxygen belonging to Ni2+-O2− pairs attributed by Lewis acid sites. Meanwhile, Mg2+-O2−, Fe3+-O2− and Zn2+-O2− were ascribed by their Bronsted base sites. On the basis of the TPD–CO2 profile, binary interactions between the metallic ions of MgO, Fe2O3, ZnO and NiO with Al species in hydrotalcite-derived catalysts produced a synergistic effect that further enhanced the basicity and basic strength, compared with synthesised CAl [39].
(a) TPD-CO2 and (b) TPD-NH3 profiles for the hydrotalcite-derived catalysts.
The acidic properties of the catalysts are summarised in Table 2 and Fig. 3b. Similarly, the acidity of CAl was the lowest. The acid strength and amount of basicity of CAl improved significantly after the incorporation of Mg, Fe, Zn and Ni. The acid strengths increased in the order of CZnAl~CNiAl>CFeAl>CMgAl>CAl, whereas the amount of acidity followed the order of CNiAl>CZnAl>CFeAl~CMgAl>CAl. Thus, CMgAl exhibited the least acid amount (1396.56 μmol g−1) with moderate acid strength (NH3 desorption at 200°C–450°C), whereas CNiAl presented the strongest acid strength (NH3 desorption at 635°C) with the largest total acid amount (6050.22 μmol g−1). These findings indicated that the presence of alkaline metal (Mg2+) ions detrimentally affected the acid sites of the catalyst and led to the formation of medium acidity as presented by CMgAl, whereas the improvement in acidity over CNiAl could be explained by the interaction between Ni2+-O2− and Al23+-O2− Lewis acid pairs [40]. The above results strongly suggested that the type of active metals used plays a key role in determining the acid–base properties of the catalyst.
Catalytic deoxygenation activity
The catalytic activity of hydrotalcite-based catalysts was investigated via the DO of oleic acid at the reaction conditions of 300°C, 3 h and 5 wt.% catalyst loading under inert N2 flow condition, and the results are tabulated in Table 3. The hydrocarbon yield increased with following order of the catalysts: CAl<CFeAl<CZnAl<CMgAl<CNiAl, whereas the total reaction products followed the trend CAl<CFeAl<CMgAl<CZnAl<CNiAl. The results in Fig. 4 and Table 3 revealed that CAl only resulted in 17% of hydrocarbon yield with 20% of total reaction product. Thus, the catalytic activity of CAl towards the DO of oleic acid was very low. Similarly, GCMS detected a high carboxylic acid content (83%), which was mainly attributed to unreacted oleic acid.
Deoxygenation products: determination of total reaction product, hydrocarbon yield and selectivity of gasoline and diesel.
| Catalysts | Total reaction product (%) | Hydrocarbon yield (%) | Selectivity (%) |
|
|---|---|---|---|---|
| Gasoline (C8–C12) | Diesel (C13–C20) | |||
| CAl | 20 | 17 | 0 | 100 |
| CMgAl | 90 | 81 | 30 | 71 |
| CNiAl | 98 | 89 | 17 | 83 |
| CFeAl | 89 | 75 | 23 | 77 |
| CZnAl | 97 | 77 | 25 | 75 |
(a) Total reaction product and hydrocarbon yield (b) product selectivity of liquid deoxygenated product of hydrotalcite-catalysed DO reaction.
CNiAl produced 98% of the total reaction products and hydrocarbon yield of 89% with high selectivity towards diesel-range hydrocarbon, which was mainly composed of C17: n-heptadecane and n-heptadecene (60%; Fig. 4, Table 3). Ni functioned as an activity promoter of the catalyst in the DO of oleic acid. A high acid amount of CNiAl catalyst (6050.22 μmol g−1) (refer to the TPD-NH3 results for further details) accelerated the occurrence of C–O cleavage, thereby maximizing the DO of oleic acid towards hydrocarbon products, which were dominated by diesel-range hydrocarbon fuels [41], [42]. The CFeAl catalyst resulted in the lowest total reaction product (89%) and hydrocarbon yield (75%) amongst the hydrotalcite catalysts due to the insufficient amount of acid sites (1396.56 μmol g−1), thereby reducing the efficiency of CFeAl in the subsequent DO step. Notably, the usage of base active metals (Mg, Zn and Fe) provided a significant amount of light hydrocarbon fractions (gasoline-range fuels) (>23%), implying that the basic promoter predominantly facilitated C–C cleavage via cracking reaction. Evidently, CMgAl with the highest basic site distribution (9002.92 μmol g−1) exhibited the highest gasoline-range fractions (30%). Similarly, the CZnAl and CFeAl catalysts produced 25% and 23% of gasoline-range fuels with basicity of 8638.51 and 2336.76 μmol g−1, respectively. The GCMS results (Fig. 5) demonstrated that the ultimate products of DO were hydrocarbons (>75%). Oxygenates were the reaction intermediates, which comprised fatty acids (e.g. oleic acid and n-decanoic acid), aldehydes (e.g. 9-hexadecenal and nonanal) and alcohol (e.g. 9-hexadecen-1-ol). CNiAl exhibited the highest C8–C20 hydrocarbon fractions (87%), followed by CZnAl, CFeAl, CMgAl and CAl. The pronounced formation of oxygenates by CMgAl and CZnAl (>20%) confirmed that basic metals Mg and Zn were prone to produce more intermediates and by-products. A previous study showed that the formation of oxygenates is related to the use of basic metal [43]. In the present study, the overall DO activity of hydrotalcite catalysts was also attributed to the high surface area of the catalysts. This finding was in accordance with the BET results (Table 2) showing the high surface area of CNiAl (117 m2 g−1), which resulted in the highest DO activity. In addition, small crystallite sizes of catalyst showed high deoxygenation activity due to its larger external surface. The presence of small crystallite of catalyst particles offers more opening channels, thus provides high accessibility of contact between reactant and active sites for deoxygenation reaction [44], [45]. Thus, fatty acids were converted directly into hydrocarbons.
Product distribution of deoxygenated products using synthesised catalysts.
All synthesised hydrotalcite catalysts successfully catalysed the DO reaction to produce >70% of hydrocarbon yield (C8–C20). Catalysts with high acidity present high catalytic activity for DO due to their cracking ability. By contrast, catalysts with high basicity produce more gasoline products with significant amounts of aldehyde intermediates and carboxylic acids than those with low basicity. Therefore, catalysts with high acid characteristic exhibited strong ability to restrain the distillation of unreacted oleic acid while simultaneously reducing the formation of intermediates and optimizing the hydrocarbon yield.
Stability of hydrotalcite catalyst
The reusability of the hydrotalcite catalysts for DO under optimum conditions (reaction temperature of 300°C, 5 wt.% catalyst loading and 3 h reaction time under inert N2 flow) was further investigated, and the results are displayed in Fig. 6a. Upon the completion of each cycle, the catalyst was reactivated by simple washing treatment with hexane and directly reused for the next cycle. All DO reactions catalysed by hydotalcite catalysts exhibited >65% yield of hydrocarbon throughout the cycles. Moreover, the catalytic activity for all the catalysts slightly decreased as the number of cycles increased. On the basis of the results of the stability investigation for all hydrotalcite catalysts shown in Fig. 6a, the yield of hydrocarbon products declined with increasing number of cycles as follows: 1st>2nd>3rd>4th. The reusability profile indicated that CMgAl, CZnAl and CFeAl failed to reuse after two cycles of reaction. The deoxygenated products do not have any hydrocarbon fraction from GCMS analysis. CMgAl and CZnAl were limited to two uses due to the formation of gel-like reactants, causing difficulties in catalyst recovery. The formation of gel-like reactants was derived from the reaction of highly basic catalysts with oleic aicd, leading to saponification [3]. By contrast, no yield was obtained after the second cycle of the reaction by using CFeAl as the catalysts not able to be regenerated due dissolvement of the catalyst into the remaining reactant under high temperature and resulted in coking phenomena. Overall, the C17 hydrocarbon selectivity of CNiAl decreased from the first to fourth cycle in less than 10% (Fig. 6b), which demonstrated that the deactivation of the CNiAl catalyst was mild. The decrement in C17 selectivity corresponded to the deactivation of catalyst by coke deposition on the catalyst’s active sites.
(a) Reusability of hydrotalcite-derived catalysts; reaction conditions: 3 h, 5 wt.% catalyst, 300°C, (b) C17 selectivity of products for CNiAl-catalysed DO reaction.
Therefore, the degree of coke formation on CNiAl catalyst was investigated by TGA (Fig. 7). Fresh CNiAl catalyst was observed to have a slight weight decrement stage of 3.96%±0.31% at 50°C–150°C, corresponding to the decomposition of water. By contrast, the fourth cycle spent catalyst resulted in two weight loss stages. The first stage occurred at 50°C–150°C due to the loss of water (1.41%±0.28%). The second stage ensued within the temperature range of 250°C–430°C with weight decrement of 11.09%±0.25%, which was attributed to the desorption of coke as CO or CO2. Coke; also recognised as soft coke due to its temperature range of desorption; is commonly formed as carbon residue of side products of any hydrocarbon reactions [46]. The elemental composition of the spent CNiAl catalyst was determined by using EDX analysis (Fig. 8 and Table 4). The EDX spectrum indicated existence of carbon elements with average major percentage <30% which corroborates with TGA result. Evidently, the reduction in hydrocarbon yield (87%–76%) and C17 selectivity (57%–49%) was attributed to the deactivation of catalyst by the deposition of coke on the catalyst surface.
TGA profiles for fresh CNiAl and spent CNiAl after the fourth cycle.
EDX analysis of spent CNiAl catalyst.
Elemental composition of spent CNiAl catalyst.
| Elements | Ni | Al | O | C |
|---|---|---|---|---|
| Composition (%) | 50.05 | 5.33 | 14.86 | 29.76 |
The catalytic performance of CNiAl catalyst was compared with the findings on hydrotalcite catalyzed deoxygenation reaction reported by previous studies (Table 5). Based on the comparison, it was found that CNiAl exhibited the highest hydrocarbon yield under mild reaction conditions; 3 h of reaction at 300°C, without presence of H2. Furthermore, CNiAl catalyst shows a stable performance for being able to reuse for up to 4 times.
Comparison on catalytic performance between CNiAl catalyst with previous studies.
| Catalyst | Type of feedstock | Reaction conditions | Hydrocarbon yield/selectivity (%) | Reusability | Ref. |
|---|---|---|---|---|---|
| Mg-Al-Co | Jatropha oil | T=400°C, Catalyst=2 wt.%, Time=2 h |
59.0 | – | [47] |
| Pd/Mg(Al)O | Hexadecyl palmitate | H2 (P)=0.9 MPa, T=300°C, Catalyst=1 wt.%, Time=2 h |
90.2 | – | [48] |
| N2 (P)=0.9 MPa, T=300°C, Catalyst=1 wt.%, Time=5 h |
80.4 | ||||
| MG70 | Jatropha oil | T=400°C, Catalyst=3 wt.%, Time=6 h |
81.1 | – | [49] |
| MG70 | Oleic acid | T=400°C, Catalyst=5 wt.%, Time=3 h |
85.2 | – | [22] |
| Ni/Hydrotalcite | Calophyllum inophyllum oil | H2 (P)=20 bar, T=350°C, Catalyst=2.5 wt.%, Time=2 h |
54.2 | – | [50] |
| CNiAl | Oleic acid | T=300°C, Catalyst=5 wt.%, Time=3 h |
89.0 | 4 times | Present study |
Conclusions
In the present study, CMgAl, CZnAl, CFeAl and CNiAl hydrotalcite-derived catalysts show the presence of mixed metal oxide phases with a surface area range of 95–117 m2 g−1 and mesoporous structure with pore sizes of 10–37 nm. The synergistic effects between respective metal oxides (MgO, ZnO, Fe2O3 and NiO) with alumina present a significant influence on the basic and acidic profiles of the catalysts, which increase in the following order: CMgAl>CZnAl>CFeAl>CNiAl and CNiAl> CZnAl>CFeAl>CMgAl, respectively. The deoxygenation activity of oleic acid was greatly influenced by the basicity and acidity of the catalysts. CNiAl shows highest hydrocarbon yield of 89% and diesel selectivity of 83%, attributes by its high acid characteristic and textural properties: high surface area. Meanwhile, CMgAl renders high gasoline selectivity of 30%, due to its very high amount of basicity. Therefore, the hydrocarbon yield and selectivity depend critically on the acid–base properties of the catalyst with the aid of the active catalyst’s large surface area. CNiAl catalyst also possesses high stability due to its acidic properties without having a high significant loss in its catalytic activity.
Article note
A collection of invited papers based on presentations at the 8th IUPAC International Conference on Green Chemistry (ICGC-8), Bangkok, Thailand, 9–14 September 2018.
Acknowledgements
The authors acknowledge the financial support from University of Malaya (GC001B-14AET, ST012-2018, ST021-2019 and RU007C-2017D), the Postgraduate Research Grant Scheme PPP (PG062-2015A) and Ministry of Education Malaysia for MyPhD scholarship.
References
[1] S. N. Naik, V. V Goud, P. K. Rout, A. K. Dalai. Renew. Sustain. Energy Rev. 14, 578 (2010).10.1016/j.rser.2009.10.003Search in Google Scholar
[2] A. Kleinová, I. Vailing, J. Lábaj, J. Mikulec, J. Cvengroš. Fuel Process. Technol. 92, 1980 (2011).10.1016/j.fuproc.2011.05.018Search in Google Scholar
[3] Z. A. Shajaratun Nur, Y. H. Taufiq-Yap, M. F. Rabiah Nizah, S. H. Teo, O. N. Syazwani, A. Islam. Energy Convers. Manag. 78, 738 (2014).10.1016/j.enconman.2013.11.012Search in Google Scholar
[4] Y.-T. Wang, Z. Fang, F. Zhang, B.-J. Xue. Bioresour. Technol. 193, 84 (2015).10.1016/j.biortech.2015.06.059Search in Google Scholar
[5] S. K. Tanneru, P. H. Steele. Renew. Energy. 80, 251 (2015).10.1016/j.renene.2015.01.062Search in Google Scholar
[6] G. Knothe. Prog. Energy Combust. Sci. 36, 364 (2010).10.1016/j.pecs.2009.11.004Search in Google Scholar
[7] Z. Pan, R. Wang, M. Li, Y. Chu, J. Chen. J. Energy Chem. 24, 77 (2015).10.1016/S2095-4956(15)60287-XSearch in Google Scholar
[8] E. Sendzikiene, V. Makareviciene, P. Janulis. Renew. Energy. 31, 2505 (2006).10.1016/j.renene.2005.11.010Search in Google Scholar
[9] Z. He, X. Wang. Catal. Sustain. Energy. 1, 28 (2012).Search in Google Scholar
[10] T. V Choudhary, C. B. Phillips. Appl. Catal. A Gen. 1, 397 (2011).Search in Google Scholar
[11] E. Furimsky. Appl. Catal. A Gen. 199, 147 (2000).10.1016/S0926-860X(99)00555-4Search in Google Scholar
[12] Y. Yang, Q. Wang, X. Zhang, L. Wang, G. Li. Fuel Process. Technol. 116, 165 (2013).10.1016/j.fuproc.2013.05.008Search in Google Scholar
[13] K. C. Kwon, H. Mayfield, T. Marolla, B. Nichols, M. Mashburn. Renew. Energy. 36, 907 (2011).10.1016/j.renene.2010.09.004Search in Google Scholar
[14] A. T. Madsen, B. Rozmysłowicz, I. L. Simakova, T. Kilpiö, A.-R. Leino, K. Kordás, K. Eränen, P. Mäki-Arvela, D. Y. Murzin. Ind. Eng. Chem. Res. 50, 11049 (2011).10.1021/ie201273nSearch in Google Scholar
[15] T. Morgan, D. Grubb, E. Santillan-Jimenez, M. Crocker. Top. Catal. 53, 820 (2010).10.1007/s11244-010-9456-1Search in Google Scholar
[16] O. B. Ayodele, H. F. Abbas, W. M. A. W. Daud. Energy Convers. Manag. 88, 1111 (2014).10.1016/j.enconman.2014.02.014Search in Google Scholar
[17] B. Puértolas, T. C. Keller, S. Mitchell, J. Pérez-Ramírez. Appl. Catal. B Environ. 184, 77 (2016).10.1016/j.apcatb.2015.11.017Search in Google Scholar
[18] K. Kaneda, T. Mitsudome, T. Mizugaki, K. Jitsukawa. Mol. 15, 8988 (2010).10.3390/molecules15128988Search in Google Scholar PubMed PubMed Central
[19] T. Mitsudome, A. Noujima, Y. Mikami, T. Mizugaki, K. Jitsukawa, K. Kaneda. Chem. – A Eur. J. 16, 11818 (2010).10.1002/chem.201001387Search in Google Scholar PubMed
[20] T. Morgan, E. Santillan-Jimenez, A. E. Harman-Ware, Y. Ji, D. Grubb, M. Crocker. Chem. Eng. J. 189, 34 (2012).Search in Google Scholar
[21] M. Romero, A. Pizzi, G. Toscano, A. A. Casazza, G. Busca, B. Bosio, E. Arato. Fuel Process. Technol. 137, 31 (2015).10.1016/j.fuproc.2015.04.002Search in Google Scholar
[22] J.-G. Na, B. E. Yi, J. N. Kim, K. B. Yi, S.-Y. Park, J.-H. Park, J.-N. Kim, C. H. Ko. Catal. Today. 156, 44 (2010).10.1016/j.cattod.2009.11.008Search in Google Scholar
[23] E. Santillan-Jimenez, T. Morgan, J. Lacny, S. Mohapatra, M. Crocker. Fuel. 103, 10102013.Search in Google Scholar
[24] X. Yu, J. Chen, T. Ren. RSC Adv. 4, 46427 (2014).10.1039/C4RA07932ASearch in Google Scholar
[25] L. Chen, F. Zhang, G. Li, X. Li. Appl. Catal. A Gen. 529, 175 (2017).10.1016/j.apcata.2016.11.012Search in Google Scholar
[26] S. Enthaler, S. Krackl, E. Irran, S. Inoue. Catal. Letters. 142, 1003 (2012).10.1007/s10562-012-0862-9Search in Google Scholar
[27] S. N. Zdainal Abidin, H. V. Lee, J. C. Juan, N. A. Rahman, Y. H. Taufiq-Yap. RSC Adv. 9, 1642 (2019).10.1039/C8RA07818ASearch in Google Scholar PubMed PubMed Central
[28] J. F. P. Gomes, J. F. B. Puna, L. M. Gonçalves, J. C. M. Bordado. Energy. 36, 6770 (2011).10.1016/j.energy.2011.10.024Search in Google Scholar
[29] N. Asikin-Mijan, H. V. Lee, J. C. Juan, A. R. Noorsaadah, G. Abdulkareem-Alsultan, M. Arumugam, Y. H. Taufiq-Yap. J. Anal. Appl. Pyrolysis. 120, 110 (2016).10.1016/j.jaap.2016.04.015Search in Google Scholar
[30] N. Asikin-Mijan, H. V. Lee, Y. H. Taufiq-Yap. Chem. Eng. Res. Des. 102, 368 (2015).10.1016/j.cherd.2015.07.002Search in Google Scholar
[31] J. Tantirungrotechai, P. Chotmongkolsap, M. Pohmakotr. Microporous Mesoporous Mater. 128, 41 (2010).10.1016/j.micromeso.2009.08.001Search in Google Scholar
[32] S. V Cherepanova, N. N. Leont’eva, A. B. Arbuzov, V. A. Drozdov, O. B. Belskaya, N. V. Antonicheva. J. Solid State Chem. 225, 417 (2015).10.1016/j.jssc.2015.01.022Search in Google Scholar
[33] J. Kong, L. Jiang, Z. Huo, X. Xu, D. G. Evans, J. Song, M. He, Z. Li, Q. Wang, L. Yan. Catal. Commun. 40, 59 (2013).10.1016/j.catcom.2013.05.026Search in Google Scholar
[34] M. J. Pérez Zurita, M. Bartolini, T. Righi, G. Vitale, P. Pereira Almao. Fuel. 154, 71 (2015).10.1016/j.fuel.2015.03.002Search in Google Scholar
[35] D. Carriazo, M. del Arco, E. García-López, G. Marcì, C. Martín, L. Palmisano, V. Rives. J. Mol. Catal. A Chem. 342, 83 (2011).10.1016/j.molcata.2011.04.015Search in Google Scholar
[36] J. Mei-lin, Z. Yan-ping, B. Yong-sheng, W. Jiang, X. Ai-jv. Green Chem. Lett. Rev. 11, 230 (2018).10.1080/17518253.2018.1470681Search in Google Scholar
[37] M. Thommes, K. Kaneko, A. V. Neimark, J. P. Olivier, F. Rodriguez-Reinoso, J. Rouquerol, K. S. W. Sing. Pure Appl. Chem. 87, 1051 (2015).10.1515/pac-2014-1117Search in Google Scholar
[38] C. García-Sancho, R. Moreno-Tost, J. M. Mérida-Robles, J. Santamaría-González, A. Jiménez-López, P. M. Torres. Catal. Today. 167, 84 (2011).10.1016/j.cattod.2010.11.062Search in Google Scholar
[39] I. Istadi, N. A. S. Amin. J. Mol. Catal. A Chem. 259, 61 (2006).10.1016/j.molcata.2006.06.003Search in Google Scholar
[40] V. V. Kumar, G. Naresh, M. Sudhakar, C. Anjaneyulu, S. K. Bhargava, J. Tardio, V. K. Reddy, A. H. Padmasri, A. Venugopal. RSC Adv. 6, 9872 (2016).10.1039/C5RA24199ESearch in Google Scholar
[41] K. Hengst, M. Arend, R. Pfützenreuter, W. F. Hoelderich. Appl. Catal. B Environ. 174, 383 (2015).10.1016/j.apcatb.2015.03.009Search in Google Scholar
[42] T. B. Čelič, M. Grilc, B. Likozar, N. N. Tušar. ChemSusChem. 8, 1703 (2015).10.1002/cssc.201403300Search in Google Scholar PubMed
[43] J. Zhang, Y. S. Choi, B. H. Shanks. Catal. Sci. Technol. 6, 7468 (2016).10.1039/C6CY01307DSearch in Google Scholar
[44] M. Thommes, K. Kaneko, A. V. Neimark, J. P. Olivier, F. Rodriguez-Reinoso, J. Rouquerol, K. S. W. Sing. Pure Appl. Chem. 87, 1051 (2015).10.1515/pac-2014-1117Search in Google Scholar
[45] P. Shu-bin, I. Tomoyuki. Zeolites. 17, 334 (1996).10.1016/0144-2449(96)00083-8Search in Google Scholar
[46] N. Asikin-Mijan, H. V Lee, Y. H. Taufiq-Yap, G. Abdulkrem-Alsultan, M. S. Mastuli, H.C. Ong. Energy Convers. Manag. 141, 325 (2017).10.1016/j.enconman.2016.09.041Search in Google Scholar
[47] H. K. D. Nguyen, H. V. Nguyen, D. S. Dao, L. L. Hoang. J. Porous Mater. 24, 731 (2017).10.1007/s10934-016-0310-0Search in Google Scholar
[48] Y. Liu, M. Inaba, K. Matsuoka. Catalysts. 7, 333 (2017).10.3390/catal7110333Search in Google Scholar
[49] M. J. A. Romero Rivas, A. Pizzi, G. Toscano, B. Bosio, E. Arato. Chem. Eng. Trans. 37, 883 (2014).Search in Google Scholar
[50] H. Hafshah, D. H. Prajitno, A. Roesyadi. Bull. Chem. React. Eng. 12, 273 (2017).10.9767/bcrec.12.2.776.273-280Search in Google Scholar
© 2020 IUPAC & De Gruyter, Berlin/Boston
