Al2O3 catalytic

ARTICLE IN PRESS

JID: JECHEM

[m5G;January 18, 2016;17:16]

Journal of Energy Chemistry 000 (2016) 1–11

Contents lists available at ScienceDirect

Journal of Energy Chemistry journal homepage: www.elsevier.com/locate/jechem

http://www.journals.elsevier.com/ journal-of-energy-chemistry/

Effect of ethanedioic acid functionalization on Ni/Al2 O3 catalytic hydrodeoxygenation and isomerization of octadec-9-enoic acid into biofuel: kinetics and Arrhenius parameters Olumide Bolarinwa Ayodele a,b,∗, Kallidanthiyil Chellappan Lethesh a, Zahra Gholami c, Yoshimitsu Uemura a a

Centre for Biofuel and Biochemical Research, Universiti Teknologi PETRONAS, 32610 Bandar Seri Iskandar, Perak, Malaysia Department of Chemical Engineering, Faculty of Engineering, University of Malaya, Kuala Lumpur, Malaysia c Centralized Analytical Laboratory, Universiti Teknologi PETRONAS, 32610 Bandar Seri Iskandar, Perak, Malaysia b

a r t i c l e

i n f o

Article history: Received 6 July 2015 Revised 10 August 2015 Accepted 11 August 2015 Available online xxx Keywords: Octadec-9-enoic acid Hydrodeoxygenation Isomerization Nickel oxalate Biofuel

a b s t r a c t The effect of ethanedioic acid (EdA) functionalization on Al2 O3 supported Ni catalyst was studied on the hydrodeoxygenation (HDO), isomerization, kinetics and Arrhenius parameters of octadec-9-enoic acid (OA) into biofuel in this report. This was achieved via synthesis of two catalysts; the first, nickel alumina catalyst (Ni/Al2 O3 ) was via the incorporation of inorganic Ni precursor into Al2 O3 ; the second was via the incorporation nickel oxalate (NiOx ) prepared by functionalization of Ni with EdA into Al2 O3 to obtain organometallic NiOx /Al2 O3 catalyst. Their characterization results showed that Ni species present in Ni/Al2 O3 and NiOx /Al2 O3 were 8.2% and 9.3%, respectively according to the energy dispersive X-ray result. NiOx /Al2 O3 has comparably higher Ni content due to the EdA functionalization which also increases its acidity and guarantees high Ni dispersion with weaker metal-support-interaction leading to highly reducible Ni as seen in the X-ray diffraction, X-ray photoelectron spectroscopy, TPR and Raman spectroscopy results. Their activities tested on the HDO of OA showed that NiOx /Al2 O3 did not only display the best catalytic and reusability abilities, but it also possesses isomerization ability due to its increased acidity. The NiOx /Al2 O3 also has the highest rate constants evaluated using pseudo-first-order kinetics, but the least activation energy of 176 kJ/mol in the biofuel formation step compared to 244 kJ/mol evaluated when using Ni/Al2 O3 . The result is promising for future feasibility studies toward commercialization of catalytic HDO of OA into useful biofuel using organometallic catalysts. © 2016 Science Press and Dalian Institute of Chemical Physics. All rights reserved.

1. Introduction There is no gainsaying in the fact that research into biofuel technologies and continuous improved research toward commercialization has come to stay. This is clearly evident in the proliferation and commercialization of two biofuel production plants by Nestle oil in Porvoo, Finland in 2007 and 2009 having a capacity of 190,000 t/a each [1]. Based on the successful output and economic turnover of the processing plants, two new additional plants were recently commissioned in 2010 and 2011 in Singapore and Rotterdam, respectively each having capacity of 800,000 t/a [1]. All the Nestle oil plants employ hydrodeoxygenation process to remove the oxygen molecule from the feed stocks. The application of HDO by Nestle oil economically validates earlier proposition that HDO

∗

Corresponding author. Tel.: +60 164955453. E-mail address: [email protected] (O.B. Ayodele).

is a more veritable method of oxygen removal compared to decarboxylation and decarbonylation processes (Decarbs) [2,3]. This is because HDO products are usually pure paraffin with constituents and properties matching close to those found in petroleum diesel fuel unlike the mixtures of paraffin and other condensation products such as esters and ketones that are usually observed in Decarbs which adversely reduces the energy density [2–5]. A systematic and keen review of recent literature [2,5–12] revealed that improvement in catalyst synthesis is the most researched area of HDO process. According to some recent reports [4,5], this observation was because HDO process can be doctored to adapt to the existing conventional crude oil refinery facilities and this logic is expected to drastically minimize the capital and production costs of the biofuel. Several attempts have been made to improve on different catalyst synthesis protocols in order to enhance activity, stability and reusability for high HDO product throughput. In this direction, Park et al. [4] studied the catalytic activities of NiMo catalysts using

http://dx.doi.org/10.1016/j.jechem.2015.08.017 2095-4956/© 2016 Science Press and Dalian Institute of Chemical Physics. All rights reserved.

Please cite this article as: O.B. Ayodele et al., Effect of ethanedioic acid functionalization on Ni/Al2 O3 catalytic hydrodeoxygenation and isomerization of octadec-9-enoic acid into biofuel: kinetics and Arrhenius parameters, Journal of Energy Chemistry (2016), http://dx.doi.org/10.1016/j.jechem.2015.08.017

JID: JECHEM 2

ARTICLE IN PRESS

[m5G;January 18, 2016;17:16]

O.B. Ayodele et al. / Journal of Energy Chemistry 000 (2016) 1–11

different supports such as alumina and zeolite. Other reports centered on different catalyst preparation procedures [4,5] and effect of adding additives such as sulfur and phosphorus [8]. Similarly, effect of different metals as catalyst promoters has been studied using Ni and Co to promote Mo supported on alumina [9,10], the authors also studied the effect of sulfur addition on the catalyst and reported that the sulfurized catalyst displayed enhanced HDO activities. In addition, they claimed that the sulfurized catalyst showed isomerization activities. Meanwhile, Kovacs et al. [11] reported that the presence of isomerized product should be considered an advantage because iso-paraffin usually has comparable lower freezing point compared to their straight chain (otherwise known as normal paraffin) counterpart. For example, n-C18 has high freezing point of 28 °C while its isomerized 2-methyl and 5-methyl have freezing points of 8 °C and –20 °C, respectively [11]. Consequently, the presence of about 20% isomerized product is expected to reduce the freezing point considerably to about 18 °C which in turn will improve the biofuel cold flow properties such as cold filter plugging point. However, in view of the potential serious environmental consequences, Kovacs et al. [11] replaces the sulfur employed in the earlier works [9,10] with fluorine and reported exceptional paraffin skeletal isomerization. They ascribed this feat to increased acidity of the synthesized catalyst due to fluorine addition. Recently, it was reported that organometallic catalysts possesses exceptionally high reactivity due to their increased acidity and improved active metal(s) solubility and dispersion in the support which in turns increases active metal octahedral structures thus modifying their morphology and textural properties [12–18]. In view of this, Ayodele et al. [19] recently functionalized supported palladium catalyst with oxalic acid and 18% iso-octadecane and ∼69% n-ocatadecane were obtained from feed stock of oleic acid. However, in view of the high cost of Pd as compared to other HDO catalyst like nickel, for example Pd is about 1000 times more expensive than Ni [19], and then it is imperative to fabricate a cheaper Ni based organometallic supported catalyst that can achieve both the hydrodeoxygenation (HDO) reaction and isomerization in one single processing step to enhance the overall process economics. In view of the above, this study investigates the applicability of an expediently synthesized organometallic catalyst via the functionalization of Ni with ethanedioic acid to develop a highly reactive alumina supported Ni oxalate catalyst (NiOx /Al2 O3 ). The effects of ethanedioic acid functionalization were evaluated on the HDO of octadec-9-enoic acid (OA) by comparing the kinetics and activation energy of the NiOx /Al2 O3 –OA system with another synthesized inorganic Ni supported on alumina catalyst (Ni/Al2 O3 ). Both Ni/Al2 O3 and NiOx /Al2 O3 catalysts were characterized for comparison of their physical and chemical properties. 2. Experimental

subsequently incorporated into 50 g dispersion (deionized water) using simple dissolution method [14,20] and left stirring for 4 h at 50 °C for the deposition of NiOx particles on the alumina support, and the pH was observed to be stabilized at 4.3 ± 0.2 as the stirring progressed. The inorganic nickel supported on alumina catalyst (Ni/Al2 O3 ) was prepared by the same method but without EdA functionalization. Both Ni/Al2 O3 and the nickel oxalate alumina supported (NiOx /Al2 O3 ) catalysts were dried in the oven at 100 °C, grinded and calcined at 400 °C in N2 gas for 4 h in a muffle furnace followed by characterization and testing. 2.3. Catalyst characterization X-ray fluorescence (XRF) analysis of the samples was done using a μXay μEDX 100 Shimadzu, NY and X-ray tube of rhodium anode and scintillation detector operating on a 40 mA current and 40 mV voltage to obtain the samples spectra. Energy dispersive X-ray (EDX) and Scanning electron microscopy (SEM) were performed to determine the samples elemental composition and morphology, respectively using (Model EMJEOL-JSM6301-F) with an Oxford INCA/ENERGY-350 microanalysis system. X-ray photoelectron spectroscopy (XPS) analyses (Thermo-Fischer K-Alpha) were carried out to obtain the chemical nature, surface composition, oxidation state, relative surface compositions and the type of interaction between metal and support. The XPS was equipped with monochromatized Al Kα source and the resulting samples spectra were analyzed using the Avantage software for peak fitting and identification of chemical state. The reduction behavior of the catalysts was studied using a Thermo Finnigan TPD/R/O 1100 equipped with a thermal conductivity detector and a mass spectrometer. Typically, 20 mg catalyst was placed in the U-shaped quartz tube. Catalyst samples were degassed under a flow of nitrogen at 200 °C to remove traces of water and impurities from the catalyst pores. H2 temperature-programmed reduction (TPR) was performed using 5% H2 /N2 with a flow rate of 20 mL/min and heating from 40 to 990 °C at 5 °C/min. Nitrogen adsorption–desorption measurements (BET method) were performed at liquid nitrogen temperature (−196 °C) with an autosorb BET apparatus, Micromeritics ASAP 2020, surface area and porosity analyzer to determine the surface area, pore size and structure, and the pore volume. X-ray diffraction (XRD) patterns of the samples were measured with Philip PW 1820 diffractometer to determine the crystal phase and structure of the metal oxides earlier detected by XRF/EDX analyses. Fourier transformed infrared (FTIR) spectroscopy analyses were performed on the samples to determine the functional groups present in order to understand the chemistry of the synthesized catalyst with respect to the support. The instrument used is Perkin-Elmer Spectrum GX Infrared Spectrometer with resolution of 4 cm−1 operating in the range of 4000–400 cm−1 . The Raman spectra of the samples were obtained with a Spex Triplemate spectrograph coupled to a Tracor Northern 1024 large area intensified diode array detector.

2.1. Materials 2.4. Octadec-9-enoic acid (OA) hydrodeoxygenation experiments Nickel nitrate hexahydrate, anhydrous ethanedioic acid and octadec-9-enoic acid were purchased from Sigma–Aldrich. Synthesis grade alumina (Scharlau AL0830, 80%) and all materials were used without any pretreatment. 2.2. Catalyst development Nickel oxalate catalyst was prepared via functionalization of 18.84 g nickel nitrate with ethanedioic acid (EdA) in stoichiometric ratio (1:1) to yield the polynulcear nickel II oxalate complex (NiOx ) in an aluminum foil wrapped conical flask due to the high photo sensitivity index of metal oxalate ligands [20]. The NiOx was

Hydrodeoxygenation of OA was conducted using a 100 mL high pressure semi-batch reactor. The reaction temperature were varied within 320–360 °C at a reaction pressure of 20 bar and 100 mL gas flow rate (90 vol% N2 and 10 vol% H2 ) based on preliminary and previous studies [11,19]. The flow of carrier gas and reaction pressure inlet and outlet were controlled by a flow (Brooks 58505 S) and a pressure controller (Brooks 5866), respectively. In a typically experiment, 20 mg of catalyst (Ni/Al2 O3 or NiOx /Al2 O3 ) was reduced in situ under flowing H2 at 400 °C for 1 h prior to use after which the reactor was purged with He and 40 g (∼45 mL) of OA being added. The operating temperature was established and


ARTICLE IN PRESS

JID: JECHEM

[m5G;January 18, 2016;17:16]


3

monitored by a type-K Omega thermocouple placed inside the reactor. Before the reaction started, 100 mL/min of 90 vol% N2 and 10 vol% H2 was passed through the reactor to build reactor pressure to 20 bar and the reaction commences by turning on the stirrer at a speed of 2000 rpm. Liquid samples withdrawn from the reactor were dissolved in pyridine and thereafter silylated with (100 wt% excess of) N,O–bis(trimethyl)-trifluoroacetamide, BSTFA in an oven at 60 °C for 1 h prior to GC analysis. The internal standard eicosane, C20 H42 was added for quantitative calculations. The withdrawn samples were analyzed with a gas chromatograph (GC, HP 6890) equipped with DB-5 column (60 m x 0.32 mm x 0.5 mm) and a flame ionization detector. 1 mL sample was injected into the GC with split ratio of 50:1 and helium was used as the carrier gas. The chromatographic pressure program was well-adjusted to achieve satisfactory separation of the desired product and the product identification was validated with a gas chromatographmass spectrometer (GC-MS). Since there was technical limitation and difficulty in the online quantification and analysis of the evolved gases (Mgas ) during the study, they were calculated according to Eq. 1, and the liquid product distribution was evaluated using Eq. 2.

Mgas = [Mb + MH2 − Ma ]

ωi ( % ) =

ni

ni

× 100

(1) (2)

where, Mb is the mass of reactor with the OA and catalyst before reaction, MH2 is the mass of total H2 gas required during the experiment evaluated from the H2 flow rate and its density. Ma is the mass of the reactor with the liquid product and used catalyst after reaction. Similarly, ωi (%) is the mass fraction of the components in the liquid product and n refers to the total number of liquid components. 3. Result and discussion 3.1. Characterization result 3.1.1. X-ray flourescence spectra The XRF spectra results of Al2 O3 , Ni/Al2 O3 and NiOx /Al2 O3 samples are shown in Fig. 1. The spectrum of Al2 O3 showed its characteristics peaks between 1.48 and 21.4 keV. However, upon the incorporation of Ni (inorganic) precursor into the structure of Al2 O3 , a new and relatively distinct peak at Kα value of 7.47 keV appeared in the Ni/Al2 O3 spectra. Similarly, sequel to the incorporation of NiOx into Al2 O3 in the NiOx /Al2 O3 sample all the characteristics peaks of Al2 O3 were dwarfed by the brilliant intensity of a peak at Kα value of 7.47 keV and another subordinate peak at Kβ of 8.26 keV. These peaks according to the standard card of peak identification (EDXRF-EPSILON 3 XL, PANalytical) were ascribed to the presence of Ni species. In the NiOx /Al2 O3 sample, the intense peak at Kα value of 7.47 keV could be ascribed to high dispersion of Ni species on Al2 O3 due to the formation of Ni-oxalate complex during the EdA functionalization. 3.1.2. Elemental dispersive X-ray The EDX result of Al2 O3 , Ni/Al2 O3 and NiOx /Al2 O3 samples are shown in Fig. 2. The composition of Al2 O3 is basically Al and O with some relatively low quantities of Ca and Na as shown in Table 1. Both Fig. 2(b) and (c) corroborates the presence of Ni species in Ni/Al2 O3 and NiOx /Al2 O3 samples. However, as both samples showed Ni peaks at low energy band of 1.9 keV, NiOx /Al2 O3 sample showed additional peak at high energy band of 7.5 keV. Furthermore, the Ni peaks in NiOx /Al2 O3 sample appeared to be more intense compared to the single peak in Ni/Al2 O3 . This observation is not at variance with what was earlier seen in

Fig. 1. XRF spectra of Al2 O3 , Ni/Al2 O3 and NiOx /Al2 O3 samples.

their XRF spectra and it further confirms that there is more than one Ni species in the NiOx /Al2 O3 sample. There was variation in the amount of Ni from the expected 10% to 8.2% and 9.3% in the Ni/Al2 O3 and NiOx /Al2 O3 samples, respectively (obtained from the average of four repeated different points since EDX is more of a point analysis technique). This variation was ascribed to different degrees of hydration at the catalysts synthesis stage [14]. The low variation observed in the NiOx /Al2 O3 was due to the formation of NiOx complex which is highly insoluble in water due to its organometallic ligand compared to NiO. 3.1.3. Scanning electron microscopy (SEM) The morphology of Al2 O3 shown in Fig. 2(a) (inset) revealed lumpy and thick flakes with partial uniform particle size distribution and relatively smooth surfaces and edges. After the catalysts synthesis both of Ni/Al2 O3 and NiOx /Al2 O3 showed formation of irregular flaky morphology with random orientation and loss of crystallinity. This observation was due to certain synthesis protocols such as calcination which has been known to remove framework alumina [19,20], earlier reports [21,22] also showed that alumina and aluminosilicates transforms from crystalline into amorphous under thermal treatment. However, the morphological variation appeared to be more pronounced in NiOx /Al2 O3 due to the acidic effect of the NiOx on the Al2 O3 support thus removing extra-framework alumina [22] and increasing catalytic activity [6]. 3.1.4. X-ray photoelectron spectroscopy (XPS) Fig. 3 shows the results of XPS analysis performed to validate the elemental composition earlier observed in the EDX result and


ARTICLE IN PRESS

JID: JECHEM 4

[m5G;January 18, 2016;17:16]


Fig. 3. (a) XPS survey scans of Al2 O3 , Ni/Al2 O3 and NiOx /Al2 O3 , (b) XPS Ni 2p spectra of Ni/Al2 O3 , (c) XPS Ni 2p spectra of NiOx /Al2 O3 .

Fig. 2. EDX spectra and SEM (inset) morphologies of (a) Al2 O3 , (b) Ni/Al2 O3 and (c) NiOx /Al2 O3 samples.

852.6–877.59 eV which is consistent with literature reports [23,24]. However, the intensity of Ni peak in NiOx /Al2 O3 is more conspicuous due to the formation of Metal Organic Framework (MOF) from the effect of EdA functionalization. In addition there is a distinctive peak at 504.08 eV in NiOx /Al2 O3 which was not seen in Ni/Al2 O3 and it was tentatively assigned to the presence of O in the Ni–oxalate ligands. These phenomena corroborated earlier observations in the EDX and XRF results. Fig. 3(b) and (c) shows XPS spectra of the Ni 2p region for Ni/Al2 O3 and NiOx /Al2 O3 catalysts. For the NiOx /Al2 O3 catalyst, the Ni 2p3/2 and Ni 2p1/2 peaks were

to determine the oxidation state of Ni based on the physicochemical variations in the catalysts due to ethanedioic acid functionalization. The survey scan of the three samples is shown in Fig. 3(a), the Al2 O3 survey scan confirmed the presence of Na, O, Al and C (from calcium carbonate) at binding energy (BE) of 1069.23 eV, 539.06 eV, 82.26 eV and 293.09 eV, respectively. Both Ni/Al2 O3 and NiOx /Al2 O3 reflected the presence of Ni particles at BE of

Table 1. Elemental composition and textural properties of Al2 O3 , Ni/Al2 O3 and NiOx /Al2 O3 . Sample

Al2 O3 Ni/Al2 O3 NiOx /Al2 O3

Elemental composition (wt%)

BET

Al

O

Ca+Na

Ni

Surface area (m2 /g)

Pore volume (cm3 /g)

Pore diameter (nm)

Average particle size (nm)

39.10 36.65 34.16

43.70 46.02 47.72

17.20 9.13 8.90

0.0 8.2 9.3

64 78 120

0.21 0.26 0.38

13.32 13.64 14.22

4687 3605 2225


JID: JECHEM

ARTICLE IN PRESS

[m5G;January 18, 2016;17:16]


obtained at 858.72 eV and 875.96 eV, respectively. These BE were shifted toward higher BEs of 860.28 eV and 877.59 eV, respectively Ni/Al2 O3 . In addition, NiOx /Al2 O3 showed two satellite peaks at 864.58 eV and 882.18 eV for Ni 2p3/2 and Ni 2p1/2 , respectively, while Ni/Al2 O3 showed only one distinctive satellite peak at BE of 877.59 eV for Ni 2p1/2 . The Ni 2p intensive shake-up satellites are characteristic of Ni2+ ions [25]. The split in the BE, ࢞ENi , i.e. Ni 2p1/2 –Ni 2p3/2 is 17.31 eV and 17.26 eV for Ni/Al2 O3 and NiOx /Al2 O3 , respectively. The shift from the Ni (metal) main BE peak for both catalyst at Ni 2p3/2 and Ni 2p1/2 indicated that nickel is present in varying oxidized form. The shift of the Ni 2p3/2 peaks to higher BE in the Ni/Al2 O3 was attributed to Ni3+ and tetrahedral Ni2+ ions from bulk NiO which indicated that there is close assembly and strong metal-support interaction resulting in possible formation of homogeneous structure such as nickel aluminate with poor reducibility [24,25], thus an indicator for comparably lower HDO activity. On the other hand, the shift of the Ni 2p3/2 peaks to lower BE values in NiOx /Al2 O3 as well as the absence of the second satellite peak in Ni/Al2 O3 implied that nickel is present more in lower oxidation state (Nio and Ni2+ ) in NiOx /Al2 O3 than in Ni/Al2 O3 . Consequently, there is weaker nickel–alumina interaction leading to high dispersion of Ni nano-particles on the amophorized alumina support with higher reducibility (see TPR result) [24]. This implied that the oxalate ion (C2 O4 2− ) was able to constrain the electrooxidation of Ni during the synthesis stage and guarantee the presence of nickel in lower oxidation which is a signal toward excellent reducibility and higher activity. Previous study [23] have also shown that MOF such as nickel-bis(cyclooctadiene) is a useful catalyst in organo-nickel chemistry exhibited Ni in zero oxidation state. The strong interaction between Ni and Al2 O3 in Ni/Al2 O3 compared to NiOx/Al2 O3 can also be seen in their TPR results and their relative XRD diffractogram shifts compared to the diffractogram of the alumina support. The XPS spectra of O 1s in Al2 O3 , Ni/Al2 O3 and NiOx /Al2 O3 are shown in Fig. 4(a) and (b); for the Al2 O3 sample, O 1s showed two BEs at 539.38 eV and 531.27 eV. The former value is characteristics of metal-oxygen species and adsorbed water while the latter is characteristics of O in Al2 O3 (i.e. lattice oxygen) [23]. The BE at the lower value is almost unchanged in both Ni/Al2 O3 and NiOx /Al2 O3 which suggested that the lattice structure of the support is not distorted due to the catalyst synthesis protocols. The O 1s BE at higher values of 539.38 eV in Al2 O3 shifted to lower BE of 534.78 eV and 534.98 eV in Ni/Al2 O3 and NiOx /Al2 O3 , respectively due to the bonding/interaction with nickel particles. The ࢞EO between the two catalysts showed negligible difference of 0.2 eV. However, their O/Ni ratio measured by the ratio of area under the peak at higher BE of O 1s of the catalysts (534.78 eV for Ni/Al2 O3 and 534.98 eV for NiOx /Al2 O3 ) to area under the peaks at Ni 2p3/2 for Ni/Al2 O3 and NiOx /Al2 O3 showed 51.921 and 8.705, respectively. This confirmed that the oxalate ion was able to constrain the electro-oxidation of Ni in NiOx /Al2 O3 as earlier pointed. 3.1.5. Temperature-programmed reduction (TPR) The TPR patterns of the catalysts are shown in Fig. 5 with Ni/Al2 O3 showing a broad low intensity band with a reduction peak at 583 °C, while NiOx /Al2 O3 showed a major and intense reduction peak at 455 °C. The shifting of the reduction temperature by 68 °C to the lower temperature region in the case of the NiOx /Al2 O3 catalyst indicated that Ni(NO3 )2 is more stable and difficult to reduce than the NiOx complex structure. This is confirmed by the time reduction peaks in Table 2 which showed that reduction process began at least 22 min in NiOx /Al2 O3 before Ni/Al2 O3 . From Table 2, the reduction peak of Ni/Al2 O3 starts at 87.23 min and stops at 99.60 min with maxima at 96.28 min, while that of NiOx /Al2 O3 starts at 64.65 min and stops at 96.98 min with

5

Fig. 4. (a) XPS O 1s spectra of Al2 O3 , (b) XPS O 1s spectra of Ni/Al2 O3 , (c) XPS O 1s spectra of NiOx /Al2 O3 .

Fig. 5. TPR patterns of Al2 O3 , Ni/Al2 O3 and NiOx /Al2 O3 samples.

maxima at 82.37 min. This observation showed that the initiation of reduction process is faster in NiOx /Al2 O3 than in Ni/Al2 O3 due to the chelating effect of EdA functionalization at the NiOx synthesis stage. Previous reports [26] where EDTA was employed as the chelating agent also showed that EDTA reduced the strong interaction between Mo species and the support in the synthesized


ARTICLE IN PRESS

JID: JECHEM 6

[m5G;January 18, 2016;17:16]


Table 2. Temperature-programmed reduction data of Al2 O3 , Ni/Al2 O3 and NiOx /Al2 O3 . Sample

Amount of gas

Amount of

Metal surface

Dispersion

Mean metal

Peaks

adsorbed

metal reacted

of catalyst

degree (%)

diameter (nm)

Time (min)

(μmol/g)

(m²/g)

503.78 999.27

19.72 39.12

(μmol/g) Ni/Al2 O3 NiOx /Al2 O3

251.89 499.63

29.57 58.66

Mo-EDTA/SiO2 -Al2 O3 which led to easy reduction of all Mo species at a low temperature region (450–550 °C) indicated by a major and highly intense reduction peak at 482 °C whereas, the catalyst prepared without EDTA (i.e. Mo/SiO2 -Al2 O3 ) contained a significant amount of Mo species strongly bound to the support that required higher temperatures (> 800 °C) for reduction. The major and highly intense reduction peak observed in the Mo-EDTA/SiO2 -Al2 O3 compared to the broad low intensity band of the Mo/SiO2 -Al2 O3 by the authors is also similar to what is seen in this present study (Fig. 5) as earlier discussed. The faster and easy initiation of reduction process seen in NiOx /Al2 O3 is obviously due to the chelating ligands around nickel particles that guarantee weaker metal-supportinteraction as seen in the XPS result where the Ni 2p3/2 and Ni 2p1/2 peaks seen at BE of 858.72 eV and 875.96 eV, respectively for NiOx /Al2 O3 catalyst were shifted toward higher BEs of 860.28 eV and 877.59 eV, respectively Ni/Al2 O3 . The amount of hydrogen gas adsorbed and the amount of metal that was reduced (NiO + H2 → Niº + H2 O) in Ni/Al2 O3 were approximately doubled in NiOx /Al2 O3 as seen in Table 2 which can be justified by the high O/Ni ratio of 51.921 observed in Ni/Al2 O3 compared to 8.705 in NiOx /Al2 O3 according to the XPS results. This explains why NiOx /Al2 O3 has exceptionally high metal surface per gram of catalyst, higher degree of dispersion and more importantly half of the mean metal (Ni) diameter seen in Ni/Al2 O3 . This is also in agreement with previous work [26] that showed that Mo/SiO2 –Al2 O3 catalyst prepared with EDTA contains well dispersed octahedral active metal oxide species with high reducibility in the temperature region of 400– 500 °C. The doubly increase in the specific surface area and pore volume of NiOx /Al2 O3 catalyst (BET result in Table 1) also gave credence to the observed high metal dispersion. This phenomenon is expected to have significant effect on the hydrodeoxygenation and isomerization process of OA in biofuel in line with the report of Al-Dalama and Stanislaus [26]. 3.1.6. N2 adsorbtion/desorption studies (BET methods) This loss of crystallinity in the catalysts earlier seen in the SEM micrographs were seen to have enhanced their textural properties as seen in the reduction of the average particle size which in turn resulted into increase in the specific surface area (Table 1) as measured by the N2 adsorption/desorption studies (BET methods). This textural enhancement is more and clearly evident in NiOx /Al2 O3 than in Ni/Al2 O3 due to the effect of acidic NiOx precursor. Generally, organometallic catalyst precursors especially of low pH are known to be structural modifiers, enhancing the mesoporosity of synthesized catalysts [6,14]. Fig. SI (shown in Supporting Information) shows the N2 adsorption/desorption isotherms of Al2 O3 , Ni/Al2 O3 and NiOx /Al2 O3 samples. Al2 O3 showed characteristics formation of monolayer isotherm at lower relative pressure as seen in the intermediate flat region which is typical of Type II isotherms and it systematically conformed to the Type IV isotherm typical of mesoporous materials at higher relative pressures. The isotherms of Ni/Al2 O3 sample did not show much difference from the Al2 O3 , however, NiOx /Al2 O3 showed characteristics of Type IV over the range of the isotherms signifying enhanced mesoporosity. This was made more evident by the increase in its H3-like hysteresis loop which is typical of inter-particle voids formed by accumulation of plate-like and nano-sized particles having slit-shaped

3.42 1.72

Temperature (°C)

Start

Maximum

Stop

87.23 64.65

96.28 82.37

99.60 96.98

583 455

pores [27]. Similarly, the increase in the amount of N2 adsorbed in NiOx /Al2 O3 confirmed the enhanced porosity observed in the pore volume (Table 1). These modified textural and morphological qualities were due to the incorporation of acidic organometallic NiOx precursor into the Al2 O3 lattice [6,14]. The well-enhanced mesoporous structure of NiOx /Al2 O3 validated the high dispersion of octahedral nickel species on the support as suggested by the XPS and TPR results. 3.1.7. X-ray diffraction of samples The XRD results of Al2 O3 , Ni/Al2 O3 and NiOx /Al2 O3 samples are shown in Fig. 6. The characteristic peaks at 2θ values of 25.5°, 35.8°, 37.6°, 43.6°, 52.7°, 57.6°, 60.6°, 66.4°, 77° and 84.8° typical of γ -Al2 O3 are also observed in Ni/Al2 O3 and NiOx /Al2 O3 and hence it is not easy to differentiate the characteristics peaks of Ni. This observation indicates that nickel species were finely dispersed owing to the effective synthesis protocols [15–17,28]. The synthesized catalysts exhibited comparative reduced peaks due to loss of crystallinity earlier observed in the SEM result and it was ascribed to the effect of calcination [6,14]. The further reduction in the NiOx /Al2 O3 peaks was ascribed to the effect of EdA functionalization at the synthesis stage which also prevented Ni sintering due to its well-developed mesoporous structure and the high metal dispersion previously seen in the BET and the TPR results, respectively. Furthermore, a close examination revealed that there is a shift in the diffraction peaks of the catalyst compared to Al2 O3 , which implied that the Ni species have intimate contact with the γ -Al2 O3 lattice thus expanding it to form a more nickel-saturated surface nickel aluminate phase [[16]–[18]]. This shift in the diffractogram peaks is more pronounced in the NiOx /Al2 O3 than in the Ni/Al2 O3 sample because NiOx complex has comparably larger effective molecular size. 3.1.8. Raman spectroscopy The Raman spectra of Al2 O3 , Ni/Al2 O3 and NiOx /Al2 O3 samples are shown in Fig. 7. Typical Al2 O3 bands are seen at 1174, 1085, 1057, 1009, 916, 828, 745, 637, 600, 518, 423 and 377 cm−1 , while the strong band at 1300 cm−1 as well as those at 1130 and 1060 cm−1 reflects Al(OH)3 due to the adsorbed water molecules [17,18]. Obviously after catalysts synthesis and calcination those bands drastically reduced in intensity due to dehydroxylation (i.e. 2Al(OH)3 →Al2 O3 +3H2 O) in Ni/Al2 O3 and NiOx /Al2 O3 . Raman spectroscopic studies have fundamentally established two types of supported metals on Al2 O3 due to either adsorption or absorption [17,18]. Adsorption is typical of metals with high oxidation state and as such cannot be accommodated into the matrix of Al2 O3 . On the other hand, absorption is characteristics of low oxidation state metals that can be absorbed into the surface of the alumina support as Ni2+ forming a surface spinel-like phase and they usually possess Raman bands between ∼800 cm−1 and 300 cm−1 . From the Ni/Al2 O3 spectra, the bands at 460, 510 and 670 cm−1 are characteristics of crystalline NiO adsorbed on the Al2 O3 surface [17] which implied the presence of higher oxidation state Ni. On the other hand, from the spectra of NiOx/Al2 O3 , there are Raman bands in the range of ∼1200–300 cm−1 which suggests that Ni species are being absorbed and adsorbed probably as Ni2+ from NiOx and NiO, respectively. This also supports earlier observation in the XRF,


JID: JECHEM

ARTICLE IN PRESS

[m5G;January 18, 2016;17:16]


7

Fig. 6. XRD patterns of Al2 O3 , Ni/Al2 O3 and NiOx /Al2 O3 samples.

Fig. 7. Raman spectra of Al2 O3 , Ni/Al2 O3 and NiOx /Al2 O3 samples.

EDX and XPS results. The NiOx is definitely due to the functionalization with EdA, while the NiO might be from the unreacted Ni salt during the catalyst synthesis stage and possible partial thermally decomposed NiOx (i.e. NiC2 O4 + ½O2 → NiO + 2CO2 ) during calcination. However, the prevalence of the NiOx over NiO in NiOx /Al2 O3 sample is justified by the absence of bands at 460, 510 and 670 cm−1 which are characteristics of crystalline NiO as well as the low intensity peaks at 375 and 600 cm−1 which are typical of crystalline NiAl2 O4 [17,18]. These also support the earlier observation in XPS and XRD result that the Ni species present in the NiOx /Al2 O3 are highly dispersed. According to Ghule et al. [18] the observed Raman bands at 1029 and 1109 cm−1 in NiOx /Al2 O3 can be unambiguously assigned to the symmetric stretching mode of short terminal M = O bonds (i.e., ν s [M = O]), whereas Raman bands at 950 and ∼820 cm−1 can be attributed to either the symmetric stretch of (–O–M–O–)n bonds (i.e., ν s [(–O–M–O–)n ]) or the antisymmetric stretch of M–O–M bonds (i.e., ν as [M–O–M]) which are characteristic of absorbed metal oxides (i.e. low oxidation state Ni). Both catalysts showed the presence of adsorbed and absorbed metals which also supported the XPS result that Ni of different oxidation states was present in both catalyst, but lower oxidation state Ni is more prevalent in the NiOx /Al2 O3 , while Ni/Al2 O3 has more of higher oxidation state and bulk NiO. Low oxidation state metal has excellent reducibility and hence higher catalytic activity while the higher oxidation state does not. It can therefore be concluded that the EdA functionalization was effective, and its presence in the NiOx /Al2 O3 is very evident by certain bands of multiple background-like vibrations which are typical of the presence of

organic functional group from nickel oxalate, the background-like vibrations were not present in the inorganic synthesized Ni/Al2 O3 . Ghule et al. [17] reported similar background-like vibrations in a study on nickel acetate.

3.2. Catalytic activities and kinetics studies 3.2.1. Preliminary studies on the HDO of oleic acid Prior to the main studies on the catalytic HDO and kinetic of OA into paraffinic biofuel using Ni/Al2 O3 and NiOx /Al2 O3 , a blank test was carried out at the previously best observed operating conditions of 360 °C, 20 bar and 100 mL/min gas flow [19] without any catalyst dosage. This study became imperative in view of the reactor and/or it accessories having any possible contributory autocatalytic effect such as hydrogenation or thermal cracking of the OA feed stock which may tamper with the accuracy of the intended kinetics and Arrhenius parameter studies. Though the result actually showed a minuscule conversion of OA, but there was no formation of any C18 hydrocarbon except about 3% stearic acid (SA) formation. This suggests that a fraction of OA was marginally hydrogenated to form saturated SA (Eq. 3) and that cracking of the OA was not imminent at that prevailing operating condition. It can therefore be concluded that there would be only marginal hydrogenation contribution for the formation of SA from the reactor and the associated internals made of stainless steel. In our previous study [19] on the HDO of OA into C18 H38 (octadecane) biofuel using acidified zeolite supported fluoropalladium oxalate catalyst we


JID: JECHEM 8

ARTICLE IN PRESS

[m5G;January 18, 2016;17:16]


Scheme 1. Simplified proposed mechanism for the HDO kinetics of oleic acid.

Fig. 8. Product distribution of the HDO of octadec-9-enoic acid under the condition of 360 °C, 20 mg catalyst loading, 20 bar, 10% H2 gas flow and 45 min.

equally observed the formation of SA as an intermediate product (Eqs. 3 and 4) from the FTIR of the evolved species.

C17 H33 COOH + H2 →C17 H35 COOH

(3)

C17 H35 COOH + 3H2 →C18 H38 + 2H2 O

(4)

3.2.2. Catalytic activities of Ni/Al2 O3 and NiOx /Al2 O3 on HDO of oleic acid The catalytic activities of Ni/Al2 O3 and NiOx /Al2 O3 on the HDO of OA at 360 °C, 20 bar, 100 mL/min H2 gas flow and 45 min using 20 mg loading of each catalyst showed that NiOx /Al2 O3 has superior activity producing 67.6% n-C18 H38 and 26.4% i-C18 H38 compared to 51.2% n-C18 H38 and 1.2% i-C18 H38 produced when using Ni/Al2 O3 . However, since the amount and intensity of Ni species in NiOx /Al2 O3 is comparably higher than in the Ni/Al2 O3 from Table 1 and Fig. 1, respectively, it is relatively difficult to ascribe the superiority of NiOx /Al2 O3 over Ni/Al2 O3 directly to either the effect of EdA functionalization or the amount of total Ni species present. Consequently, a repeated HDO of OA was conducted at the same experimental condition but using 22.68 mg of Ni/Al2 O3 , an equivalent of 9.3% Ni of 20 mg NiOx /Al2 O3 . The observed result in Fig. 8 showed that there is a marginal increase of about 4.3% in n-C18 H38 production and almost negligible effect on i-C18 H38 production. Thus, this observation confirmed that the superior activity of NiOx /Al2 O3 over Ni/Al2 O3 can be ascribed to the effect of EdA functionalization which enhances higher Ni dispersion (as seen in the TPR plot) and also ensures the formation of subsurface absorbed octahedral Ni species with increased reducibility as shown by the XPS result [6,24,26]. In addition, the EdA functionalization also engrained in NiOx /Al2 O3 the ability to withstand high degree of hydration thereby resulting in comparably higher Ni content and strong Ni spectra as seen in Table 1 and Fig. 1, respectively. Generally, organometallic catalysts with metal-oxalate ligand functionalities have been reported to be acidic and highly reactive as well as having propensity to minimize nefarious side reactions [6,29,30]. The presence of i-C18 H38 was due to the increase in the NiOx /Al2 O3 acidity owing to the EdA functionalization. Previous reports [11,13] have shown that the presence of about 20% isomerized components in biofuel can reduce its freezing point by about 12 °C thus enhancing the biofuel cold flow property. There are also instances of decarboxylation/decarbonylation reaction as evident by the presence of C17 H36 in the products of Ni/Al2 O3 which were not observed in the NiOx /Al2 O3 and this was ascribed to the EdA functionalization at the NiOx /Al2 O3 synthesis stage. About 29% of non-hydrocarbons comprising unreacted OA, unconverted SA and traces of esters and alcohol functional groups were also conspicuous in Ni/Al2 O3 product distribution but were insignificant in NiOx /Al2 O3 . The almost-total HDO of OA into C18 H38 when using NiOx /Al2 O3 was similar to the results obtained by Sousa et al.

[12] on the hydrotreatment of sunflower oil using β -Mo2 C/Al2 O3 where all the C18 acids in the sunflower oil were converted to C18 biofuel. Similarly, just as no C17 H36 was observed when using NiOx /Al2 O3 , Sousa et al. [12] also did not observe decarboxylation/decarbonylation, and they attributed the efficiency and ability of the β -Mo2 C/Al2 O3 catalyst to the expedient organometallic synthesis protocol based on the temperature-programmed carburization (TPC) methodology. 3.2.3. Kinetics studies of hydrodeoxygenation of oleic acid In order to study the effect of EdA functionalization on alumina supported Ni catalyst on the HDO of OA, the development of a proposed HDO mechanism becomes important for the kinetic study. In previous section and our past study [19], hydrodeoxygenation of OA proceeded via sequential hydrogenation to form saturated SA followed by oxygen molecule extraction to produce the final biofuel (Eqs. 3 and 4). This was also supported by other previous works [5,11] and it was due to the presence of double bond in the OA structure. In the HDO of OA, SA was observed to be the main intermediate product, although some other species may be present in varying minute quantity. In order to model a mechanism that can be modestly solved, a lumped kinetic model was adopted as shown in Scheme 1, where SA represents all the intermediate compounds since its concentration excessively outweighs that of any other possible intermediate products [14,31]. Similarly, since the concentration of OA is far in excess of H2 gas, a pseudo-first-order kinetics was assumed with respect to OA to determine the rate constants [14,32], ki (i = 1, 2, 3) especially since few preliminary experiments on effect of variation of OA loading at constant catalyst dosage showed that its disappearance conformed to pseudofirst-order kinetics. The kinetic study is also to establish whether the SA formation step or its consumption step is the limiting step. The following sets of differential Eqs. (Eqs. 5–12) were obtained according to the reaction pathway in Scheme 1. Notation: C17 H33 COOH → OA; C17 H35 COOH → SA; C18 H38 → C18

−dcOA = (K1 + K2 )COA dt

(5)

dcSA = k2COA − k3CSA dt

(6)

From Eq. 5,

COA = COAo · e−(k1 +k2 )t

(7)

As earlier noted that HDO of OA proceeded via sequential hydrogenation to form SA, therefore k1 representing the direct formation of the C18 biofuel in Scheme 1 can be set to zero, because the rate of direct formation, if at all takes place, will be far too low to be reckoned with in this study. Therefore Eq. (7) can be reduced to:

COA = COAo · e−(k2 )t

(8)

Substituting Eq. (8) into Eq. (6) gives Eq. (9), which upon integration with boundary limits: when t = 0, COA = COAo and


ARTICLE IN PRESS

JID: JECHEM

[m5G;January 18, 2016;17:16]


9

Fig. 9. Experimental data fitting of the HDO of octadec-9-enoic acid at 20 mg catalyst loading, 10% H2 gas flow and 20 bar.

CSA = CSAo = 0 gives rise to Eq. (10).

dcSA = k2COAo · e−(k2 )t − k3CSA dt CSA =

(9)

k2COA −(k2 )t e − e−(k3 )t (k3 − k2 )

(10)

Consequently, the formation of C18 biofuel can be obtained by substituting Eqs. (8) and (10) into Eq. (11). It is important to note that both n-C18 H38 and i-C18 H38 formed are summed together as C18 since their molecular formula and weight are same.

i.e.,

C18 = COAo − COA − CSA

C18 = COAo − COAo · e−(k2 )t −

(11)

k2 cOA

(k3 − k2 )

e−(k2 )t − e−(k3 )t

(11b)

Rearranging Eq. (11b) gives Eq. (12) which gives the ratio of C18 biofuel production as a function of time with respect to the initial OA concentration.

1 c18 = k (1 − e−(k2 )t ) − k2 (1 − e−(k3 )t ) . cOAo (k3 − k2 ) 3

(12)

The experimental data of C18 formation for both Ni/Al2 O3 and NiOx /Al2 O3 at temperature range of 320–360 °C were fitted into c Eq. (12) using MathCAD v13 and plotted in (Fig. 7), where C SA is OAo

denoted by Y(t) and represents the developed model, while C18i is the experimental value. The plot fitting on Fig. 9 showed a high degree of perfect fitting with observed R2 ranges between 0.92 and 0.98. The fitted plots showed that the C18 production is enhanced as the HDO temperature is increased from 320 to 360 °C. For example, when using Ni/Al2 O3 the C18 production increased from about 70% to 85% when the temperature was increased from 320 to 360 °C. Similarly, in the same temperature range the C18 production of NiOx /Al2 O3 increased from 85% to 96%. This clearly indicates that the HDO process of OA irrespective of the catalyst used conforms to the

Arrhenius theory of temperature dependent of reaction rates. This implied that as the temperature was increased the molecules of the OA gained more kinetic energy in excess of activation energy to vigorously interact with the H2 gas at the catalysts active sites. Furthermore, since the viscosity of OA has an inverse relation with temperature, for example according to Aspen Hysys process simulator v7.2, a simulation of the effect of temperature on OA showed that at 320, 340 and 360 °C, the viscosity of OA was 0.4021, 0.3441, 0.2963 cp, respectively. Consequently, the reduction in OA viscosity with increased temperature is also considered to have enhanced both the H2 mass transfer into the bulk of OA and its propensity for solubility in the reaction mixture. A careful look at the fitted plots (Fig. 9) also revealed that irrespective of the temperature the C18 production of the Ni/Al2 O3 is clearly inferior to NiOx /Al2 O3 . For example, at 360 °C the C18 production of Ni/Al2 O3 is 85% while NiOx /Al2 O3 has 96%. This feat was due to the functionalization of the Ni catalyst with EdA, which did not only improves the solubility of Ni salt [6,26] but also enhanced its dispersion and potential for reducibility as earlier observed in the XPS, XRD and Raman spectroscopy results. In addition, according to the EDX result, EdA functionalization increases the ability of the NiOx /Al2 O3 to withstand the degree of hydration at the catalyst synthesis stage due to its organometallic functionalities. Consequently, NiOx /Al2 O3 has comparably higher Ni species of 9.3% compared to 8.2% observed in Ni/Al2 O3 , all these contributed to its high catalytic activities. According to the studies of Li et al. [6] on the synthesis of NiMo/γ -Al2 O3 catalyst, they reported that the increase in the catalyst acidity due to functionalization with organic tetraphenylporphyrin increased the metals dispersion, enhanced the morphological and textural qualities which in turn resulted in the formation of octahedral Ni and Mo known to be highly reactive species. Their report is also in agreement with the report of Al-Dalama and Stanislaus [26] which showed that the effect of EDTA on Mo–EDTA/SiO2 –Al2 O3 resulted in high metal dispersion having weaker interaction with the support leading to highly reducible active metal with comparably higher


ARTICLE IN PRESS

JID: JECHEM 10

[m5G;January 18, 2016;17:16]


Table 3. Evaluated kinetic parameter of the HDO of octadec-9-enoic acid using Ni/Al2 O3 and NiOx /Al2 O3 . Ni/Al2 O3

NiOx /Al2 O3

Temperature (°C)

k2 (s−1 )

k3 (s−1 )

k2 (s−1 )

k3 (s−1 )

360 340 320

4.10 × 10− 5 1.29 × 10− 5 5.03 × 10− 6

3.12 × 10−12 7.26 × 10−13 1.26 × 10−13

3.74 × 10− 2 1.64 × 10−2 8.88 × 10−3

2.22 × 10− 7 7.59 × 10−8 2.09 × 10−8

Fig. 10. Arrhenius parameter fitting plot (a) stearic acid formation step, k2 ; (b) C18 formation step, k3 .

hydrotreating activity compared to Mo/SiO2 -Al2 O3 . These findings [6,26] and another US patent [30] supported the high Ni dispersion, enhanced morphological and textural qualities of NiOx /Al2 O3 compared to Ni/Al2 O3 as earlier seen in their characterization results, hence resulted in its superior HDO and isomerization activity seen in Figs. 8 and 9. 3.2.4. Rate constants and Arrhenius parameters The rate constants obtained from Fig. 9 are shown in Table 3. As expected the rate constants when using NiOx /Al2 O3 are clearly higher than Ni/Al2 O3 , which is a confirmation that the former is highly active owing to the EdA functionalization as earlier explained. Furthermore, a closer observation showed that k2 > k3 in all cases which implied that sequential hydrogenation of OA into SA is more easily achieved and faster than the oxygen extraction stages. This supports and explains why minuscule conversion (about 3%) of OA into SA was the only observation during the preliminary study in the absence of catalysts. The kinetic data in Table 3 were fitted into the Arrhenius equation (Eq. 13) and plotted as shown in Fig. 10 to evaluate the pre-exponential factor, Ao (s−1 ) and the activation energy, Ea (kJ/mol/K).

ln k = ln Ao − Ea /RT

Table 4. Evaluated Arrhenius parameter of the HDO of octadec-9-enoic acid using Ni/Al2 O3 and NiOx /Al2 O3 .

(13)

From Table 4, the values of Ao (which is the total number of collisions either leading to a reaction or not per second) observed

Ao (s−1 ) Ea (kJ/mol)

Stearic acid formation

C18 biofuel formation

Ni/Al2 O3

NiOx /Al2 O3

Ni/Al2 O3

NiOx /Al2 O3

1.19 × 109 163

6.31 × 107 112

4.88 × 108 244

5.54 × 107 176

for the process showed that SA formation step (k2 ) has higher values compared to the C18 formation step (k3 ). Similarly, comparing the two catalysts, NiOx /Al2 O3 has comparably lower value. According to collision theory [33], the higher values observed when Ni/Al2 O3 was used could be due to the catalyst’s inability to initiate faster reaction rates to use up the continuous inflow of H2 gas, hence the collision rate increases since the H2 flow rate was kept constant and the temperature was elevated. On the other hand, the low Ao observed in NiOx /Al2 O3 could imply that the H2 gas was quickly used up since NiOx /Al2 O3 has comparable higher reactivity. The lowest Ea of 112 kJ/mol was observed for the SA formation step when using NiOx /Al2 O3 while the highest value of 244 kJ/mol was observed for the C18 formation using Ni/Al2 O3 and in both cases the Ea of the SA formation steps is the lowest. This results showed that it is much more easier to overcome the energy barrier in the SA formation step than in the C18 formation step and this corroborate the earlier results on the kinetic study and especially the preliminary analysis that showed only marginal conversion of OA into SA without catalyst loadings. In addition, EdA functionalization was able to adequately reduce the activation energy barrier in both the SA formation step and the C18 formation step. A careful analysis of the activation energy revealed that 51 kJ/mol (i.e. 163–112, SA formation step) and 68 kJ/mol (i.e. 244–176, C18 formation step) would be saved if NiOx /Al2 O3 is used instead of Ni/Al2 O3 . The kinetic data adequately supports the earlier observation on the comparison between the catalytic activities of Ni/Al2 O3 and NiOx /Al2 O3 on the HDO and isomerization of OA in Fig. 8 and it is considered a good credit for the process economics. 3.3. Ni/Al2 O3 and NiOx /Al2 O3 reusability studies The reusability studies of Ni/Al2 O3 and NiOx /Al2 O3 on the HDO of OA was carried out at 360 °C, 20 bar, 20 mg catalyst, 100 mL/min gas flow and 45 min. The results when using NiOx /Al2 O3 showed consistency with the result in Fig. 8 for both n-C18 H38 and iso- C18 H38 in three consecutive runs after which only iso-C18 showed 4% drop after the 4th run. On the other hand, Ni/Al2 O3 showed a marginal drop of about 2% in the n-C18 H38 production in the second run, but a drastic reduction of 8% in the third use. The major reason for the deactivation of Ni/Al2 O3 is the carbon deposition due to the decarboxylation/decarbonylation (i.e. removal of O2 as CO2 or CO) as evident by the presence of C17 H36 in its products composition. The high reusability of NiOx /Al2 O3 compared to Ni/Al2 O3 was undoubtedly due to the EdA functionalization which intercalated in it the exceptional high reusability quality of organometallic catalysts [22]. According to previous reports [34,35] organometallic catalysts with metal oxalate ligand complex have been reported to be highly reactive and reaction specific thus minimizing reaction time and prolonging the catalyst life span. Similarly, the presence of the strong M+ -oxalate ligand in the NiOx /Al2 O3 increases the active metal resistance to leaching out of the support which also minimizes the tendencies of multiple side reactions [22,30]. The reduction in the iso-C18 according to Kovacs et al. [11] was ascribed to partial loss of the NiOx /Al2 O3 acidity.


JID: JECHEM

ARTICLE IN PRESS O.B. Ayodele et al. / Journal of Energy Chemistry 000 (2016) 1–11

4. Conclusions Two hydrodeoxygenation (HDO) catalysts were synthesized; the first (Ni/Al2 O3 ) was via incorporation of inorganic Ni precursor into Al2 O3 , and the second (NiOx /Al2 O3 ), an organometallic catalyst was via the incorporation of nickel oxalate (NiOx ) prepared by functionalization of Ni with ethanedioic acid (EdA) into Al2 O3 . The characterization result of both catalysts showed that Ni species present in Ni/Al2 O3 was 8.2% while 9.3% was observed for NiOx /Al2 O3 according to the energy dispersive X-ray (EDX) result. The Ni contents of both samples were below the expected value of 10% due to the degree of hydration at their synthesis stage. However, the comparably higher value of 9.3% in NiOx /Al2 O3 was ascribed to the EdA functionalization because NiOx is highly insoluble. Furthermore, temperature-programmed reduction, X-ray fluorescence, Xray photoelectron spectroscopy, X-ray diffraction, and Raman spectroscopy results showed that Ni species in NiOx /Al2 O3 was highly dispersed due to the EdA functionalization which also modified the textural and morphological properties as seen in the N2 adsorption/desorption and scanning electron microscopy results, respectively. The catalysts activities test on the HDO of octadec-9-enoic acid showed that NiOx /Al2 O3 did not only display the best catalytic and reusability abilities, but it also possesses isomerization ability as a result of the EdA functionalization which increased its acidity. In the kinetic and Arrhenius parameter study, NiOx /Al2 O3 also has the highest rate constants evaluated using pseudo-first-order kinetics, but the lowest activation energies. Acknowledgments The authors sincerely acknowledge the financial support from Higher Impact Research-Ministry of Higher Education project no D000011-16001 of the Faculty of Engineering, University of Malaya, Malaysia and the Mitsubishi Corporation Education Trust Fund, University Teknologi PETRONAS, Malaysia. Supplementary materials Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.jechem.2015.08.017.

[m5G;January 18, 2016;17:16] 11

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35]

Neste Oil Corporation http://wwwnesteoilcom:accessed December, (2013). P.T. Do, M. Chiappero, L.L. Lobban, D.E. Resasco, Catal Lett 130 (2009) 9. P.S. Nigam, A. Singh, Prog Energy Combust Sci 37 (2011) 52. J. Park, J-K. Lee, J. Miyawaki, Y-K. Kim, S-H. Yoon, I. Mochida, Fuel 90 (2011) 182. P. Priecel, D. Kubicka, L. Capeka, Z. Bast, P. Rysanek, Appl Catal A 397 (2011) 127. J. Li, Z. Xia, W. Lai, J. Zheng, B. Chen, X. Yi, W. Fang, Fuel 97 (2012) 504. Sotelo-Boyás R., Liu Y., Minowa T. Ind Eng Chem Res, 2011, 50: 2791. Y. Yang, C. Ochoa-Hernández, V.A.P. O’Shea, J.M. Coronado, D.P. Serrano, ACS Catal 2 (2012) 592. P. Simácek, D. Kubicka, Fuel 89 (2010) 1508. A. Guzman, J.E. Torres, L.P. Prada, M.L. Nunez, Catal Today 156 (2010) 38. S. Kovacs, T. Kasza, A. Thernesz, I.W. Horvath, J. Hancsok, Chem Eng J 176-177 (2011) 237. L.A. Sousa, J.L. Zotin, V. Teixeira da Silva, Appl Catal A 449 (2012) 105. O.B. Ayodele, H.U. Farouk, J. Mohammed, Y. Uemura, W.M.A.W. Daud, J Mol Catal A 400 (2015) 179. O.B. Ayodele, O.S. Togunwa, Appl Catal A 470 (2014) 285. Y. Bang, S.J. Han, J. Yoo, J.H. Choi, K.H. Kang, J.H. Song, J.G. Seo, J.C. Jung, I.K. Song, Int J Hydrogen Energ 38 (2013) 8751. Y. Bang, S.J. Han, J.G. Seo, M.H. Youn, J.H. Song, I.K. Song, Int J Hydrogen Energ 37 (2012) 17967. A.V. Ghule, K. Ghule, T. Punde, J. Liu, S. Tzing, J. Chang, H. Chang, Y. Ling, Mater Chem Phy 119 (2010) 86. A.V. Ghule, K. Ghule, S. Tzing, T. Punde, J. Liu, H. Chang, Y. Ling, J Solid State Chem 182 (2009) 3406. O.B. Ayodele, H.F. Abbas, W.M.A.W. Daud, Ind Eng Chem Res 53 (2014) 650. O.B. Ayodele, B.H. Hameed, J Ind Eng Chem 19 (3) (2012) 966. O.B. Ayodele, W.M.A.W. Daud, J Taiwan Inst Chem Eng 47 (2015) 113. L. Xiaoling, W. Yan, W. Xujin, Z. Yafei, G. Yanjun, X. Qinghu, X. Jun, D. Feng, D. Tao, Chin J Catal 33 (2012) 1889. I.G. Casella, M. Contursi, J Electroanal Chem 692 (2013) 80. T.Y. Ma, S. Dai, M. Jaroniec, S.Z. Qiao, J Am Chem Soc 136 (2014) 13925. N.V. Kosova, E.T. Devyatkina, V.V. Kaichev, J Power Sour. 174 (2007) 965. K. Al-Dalama, A. Stanislaus, Thermochimica Acta 520 (2011) 67. T. Xue, Y.M. Wang, M. He, Micropor Mesopor Mater 156 (2012) 29. O.B. Ayodele, B.H. Hameed, Appl Clay Sci 83–84 (2013) 171. O.B. Ayodele, Appl Clay Sci 72 (2013) 74. Tanev P.T., Lange De Oliveira A. US 2012/0123176 A1. Zhang H., Lin H., Wang W., Zheng Y., Hu P. Appl Catal B, http://dxdoiorg/doi: 101016/japcatb201312006. O.B. Ayodele, H.S. Auta, N. Md Nor, Ind Eng Chem Res 51 (2012) 16311. R.W. Missen, C.A. Mims, B.A. Saville, Introduction to Chemical Reaction Engineering and Kinetics, John Wiley & Sons Inc, New York, 1999, p. 128. O.B. Ayodele, H.F. Abbas, W.M.A.W. Daud, Energy and Fuels 28 (2014) 5872. O.B. Ayodele, H.U. Farouk, J. Mohammed, Y. Uemura, W.M.A.W. Daud, J Taiwan Inst Chem Eng 50 (2015) 142.
