3.1 XRD characterization of catalysts
The low angle XRD patterns of calcined SBA-15 and Ni(5)-HPW(10)/SBA-15 are shown in Fig. 1a. The reflections at 2θ of 0.5°, 1.2° and 1.4° corresponding to (1 0 0), (1 1 0) and (2 0 0), respectively were observed for both the samples. This indicated that the ordered hexagonal mesoporous SBA-15 was formed and the impregnation of Ni and HPW to SBA-15 have not altered its long range order, similar to observation reported for Ni/SBA-15[24]. From the wide angle XRD patterns of the samples (Fig. 1b), the broad peak observed for all the samples in the 2θ range of 15-30o is attributed to amorphous silica of the SBA-15[25]. Three peaks observed at 2θ positions of 37, 43 and 63o were found to be matching with the standard NiO peaks of JCPDS 04-0835 and intensity of these three peaks increased with the increase in Ni loading from 5 to 15%. The reflection at 2θ of 34.1o, characteristic of HPW[26] was not found in the XRD pattern of Ni(5,10,15)HPW(10)/SBA-15 catalysts, due to the thorough dispersion of HPW in SBA-15.
3.2 Nitrogen adsorption isotherms of catalysts
The physico-chemical and textural characteristics of the catalyst are presented in Table 1. All the SBA-15 supported
Table 1
Textural and physico-chemical characteristics of the SBA-15 support and catalysts
Sample
|
aBET
[m2/g]
|
bVp
[cc/g]
|
cdp
[nm]
|
Ni & HPW Loadingd
[wt %]
|
Metal
crystallite
sizee
[nm]
|
Total acidityf
(mmol g − 1)
|
ga0
(nm)
|
hPW
(nm)
|
Ni
|
HPW
|
SBA-15
|
871.782
|
0.81
|
3.4
|
–
|
–
|
-
|
0.01
|
14.7
|
11.3
|
Ni(10)/SBA-15
|
768.002
|
0.73
|
3.4
|
9.90
|
–
|
4.82
|
-
|
11.9
|
8.5
|
HPW(10)/SBA-15
|
720.116
|
0.62
|
3.2
|
–
|
9.87
|
-
|
0.09
|
12.4
|
9.2
|
Ni(5)-HPW(10)/SBA-15
|
680.433
|
0.58
|
3.0
|
4.80
|
9.91
|
3.44
|
-
|
10.5
|
7.5
|
Ni(10)-HPW(10)/SBA-15
|
650.536
|
0.54
|
2.7
|
9.85
|
9.70
|
2.49
|
0.14
|
10.3
|
7.6
|
Ni(15)-HPW(10)/SBA-15
|
612.417
|
0.51
|
2.5
|
14.70
|
9.89
|
3.43
|
-
|
9.8
|
7.3
|
aBET surface area calculated from the N2 adsorption isotherm.
btotal pore volume calculated from the N2 adsorption branch using the BJH method
caverage pore diameter calculated from the N2 adsorption branch using the BJH method
dactual Ni and HPW loading from ICP-OES analysis
ecrystallite size of Nickel calculated from the wide angle XRD using Debey- Scherrer equation
fvalue obtained from NH3-TPD results
gunit cell parameter obtained from XRD results
hwall thickness of the sample estimated by the expression: Pw= a0− dp where (a0 = 2d100/√3)
catalysts were found to have BET surface area, pore volume and pore diameter less than the support SBA-15, suggested that incorporation of either Ni or HPW or both partially blocked the mesopores of SBA-15. Further, with the increase in Ni loading of HPW(10)/SBA-15 from 0 to 15 Wt%, BET surface area, pore volume and pore diameter decreased. From the N2 adsorption isotherms (Fig. 2a), all the samples showed type IV(a) adsorption isotherm with hysteresis loop evidencing the characteristic of cage-like (Im3m) pore structure[27]. Absence of the sharp inflection at the relative pressure (p/p0) range 0.22 to 0.26 indicated the absence of micropores. This was evidenced further from the pore size distribution, much of which were ranging from 2 to 8 nm (Fig. 2b), falling in the range of mesopores, characteristic of SBA-15.
3.3 NH3‑TPD and FT-IR spectra of pyridine adsorbed catalysts
The NH3-TPD profiles of SBA-15, HPW(10)/SBA-15, Ni(10)-HPW(10)/SBA-15 are shown in Fig. 3a. Two distinct peaks, first one up to 300oC and second one from 350 to 650oC were observed, indicated the presence of two distinct strength of acid sites of medium and strong respectively. TPD of ammonia increased in the order of SBA-15 < HPW(10)/SBA-15 < Ni(10)-HPW(10)/SBA-15, possibly due to the presence of both Brønsted acid sites of HPW and Lewis acid sites of Ni2+ in Ni(10)-HPW(10)/SBA-15. This was further evidenced from the FTIR spectra of pyridine adsorbed Ni(10)-HPW(10)/SBA-15 catalyst (Fig. 3b) which showed bands at 1590 and 1450 cm-1, attributed to Brønsted and Lewis acid sites respectively[28].
3.4 FT-IR spectra of the support and catalysts
The FT-IR spectra of the support SBA-15 and Ni/SBA-15 and Ni,HPW/SBA-15 catalysts are presented in Fig. 4. A broad band at 3343 cm− 1 is due to the defective Si-OH stretching vibration of SBA-15. The intense broad band observed over 1300 to 1000 cm− 1 is due to the asymmetric and symmetric vibrations of Si–O–Si. Absorption bands at 810 and 436 cm− 1 are due to the bending vibrations of Si–O–Si while the shoulder around 966 cm− 1 is due to the stretching vibration of the defective Si–OH. Bulk HPW with a Keggin phase generally shows four strong peaks at 1084 cm− 1 (P–O), 983 cm− 1 (W–O), 895 cm− 1, and 801 cm− 1 (W–O–W) and a weak band at 524 cm− 1 (W–O–P), were found to be absent for Ni (5,10,15%)-HPW(10)/SBA-15 samples, indicated the absence of Keggin phase thus reinforced the similar conclusion from the XRD results.
3.5 UV-visible diffuse reflectance spectra of the catalysts
The UV-Visible diffuse reflectance spectra (UV-Visible DRS) of the catalysts are shown in Fig. 5. SBA-15 did not show any peak in 200 to 800 nm whereas HPW(10)/SBA-15 showed a broader peak from 200 to 300 nm, attributed to oxygen to metal charge transfer in tungstophosphate anion[29]. Ni(10)/SBA-15 showed a sharp peak around 250 nm, attributed to charge transfer transition of NiO from O2- to Ni2+ ions[30] and absorption increased with increase in Ni loading from 5 to 15% for the bi-functional Ni(5,10,15), HPW(10)/SBA-15 catalysts. Absence of plasma band for the Ni particles proved them to be of Nano size.
3.6 HRTEM characterization of catalysts
The HRTEM image of the catalysts are shown in Fig. 6. The image of SBA − 15 showed the hierarchical mesoporous structure and the same was retained after the incorporation of Ni and HPW. The particle size of Ni in Ni(10)/SBA-15 catalyst varied between 15 to 45 nm and the mean size was found to be 30.57 nm. The particle size distribution of HPW in HPW(10)/SBA-15 was found to be between 10 to 70 nm and the mean size was 32.08 nm. For the bi-functional Ni(10)-HPW(10)/SBA-15 catalyst, particle size distribution was between 10 to 70 nm and the mean particle size of was found to be 31.02, lying between Ni (10)- SBA-15 and HPW (10)-SBA-15, which may be possibly due to the interaction between silica wall of SBA-15, Ni and HPW. The EDAX spectrum (Fig. 6e) confirmed the presence of Ni, W, P, and Si, in the bi-functional catalyst Ni(10)-HPW(15)/SBA-15.
3.7 X-ray photoelectron spectra (XPS) of the catalysts
XPS spectrum of nickel 2p, silicon, tungsten 4f for Ni(10)-HPW(10)/SBA-15 catalyst are given in Fig. 7. The Ni 2p signal is deconvoluted into four peaks in the range of 850–890 eV. The peaks centered at 853.31 and 858.9 eV were attributed to Nio and Ni2+of Ni 2p3/2 respectively, while the peaks at 870.50 and 878.2 eV were attributed to Nio and Ni2+of Ni 2p1/2 respectively, and the proposition of Ni0 and Ni2+ in the sample were similar[31]. The Si 2p has shown a characteristic peak at 104.1 eV, attributed to SiO2 of SBA-15 support[32]. W 4f XPS spectrum showed a doublet at 33.57 and 35.80 eV assigned to W 4f7/2, characteristic of the W6+ ion interacting and non-interacting with the support respectively[33].
3.8 Products and reaction pathways of catalytic HDO of trans-anethole
Products of HDO of trans-anethole were identified by GC-MS. Oxygenated products identified were 4-propyl anisole formed by the hydrogenation of propenyl group and, 4-propyl phenol formed by the hydrogenation of propenyl group and demethylation of methoxy group of the trans-anethole. Propyl benzene, propyl cyclohexane, 1-propenyl benzene, benzene and cyclohexane were the deoxygenated products identified. Two reaction pathways namely direct HDO and hydrogenation were reported for the HDO of lignin derived model compounds[11]. Direct HDO reportedly involved hydrogenolysis of C-O bond, yielding aromatics while hydrogenation involved hydrogenation of benzene ring followed by hydrogenolysis of C-O bond yielding cycloalkanes. Thus formation of propyl benzene is accounted by the direct HDO of C-O bond of the methoxy group and hydrogenation of propenyl group of trans-anethole while 1-propenyl benzene is due to the occurrence of the first step and the absence of second step. Propyl cyclohexane formation is by hydrogenation pathway involving the hydrogenation of propenyl group and benzene ring followed by hydrogenolysis of C-O bond of the methoxy group of trans-anethole. Absence of C9 alkanes or isoalkanes in the product indicated that the hydrogenation did not proceed to the extent of opening up of the cyclohexane ring, unlike reported for HDO of eugenol over Ni supported on HZSM-5 and combined Al-SBA-15 and HZSM-5 supports[34]. This indicated the efficacy of the present catalyst system in producing and preserving the propyl cyclohexane desired to be present in the 100%SAF. Benzene and cyclohexane observed in the product are due to the consecutive dealkylation of propyl benzene and propyl cyclohexane respectively.
3.9 Effect of Time on Stream on the HDO of trans-anethole
The effect of time on stream (TOS) on the conversion of trans-anethole and HDO selectivity (HDO%) for different catalysts are presented in Fig. 8a and 8b respectively. Among all the catalysts studied, Ni(10)-HPW(10)/SBA-15 has shown the highest conversion at TOS of 2 h and remained almost the same except for the TOS of 10 h, while for all other catalysts conversion declined with TOS and the extent of decline varied according to catalyst composition. From the thermogravimetric analysis of spent catalysts (Fig. 9), it was found that weight loss was minimum for Ni(10)-HPW(10)/SBA-15, indicated less coke deposition and hence less deactivation accounted for stable conversion over the TOS of 12 h. On the other hand Ni(5)-HPW(10)/SBA-15 has shown the highest weight loss due to higher coke deposition and thus accounted for its faster deactivation. Selectivity for the hydrodeoxygenated products (HDO%) was found to be the highest for HPW(10)/SBA-15 at TOS of 2 h and declined sharply with the increase in TOS and reached the lowest at 8 h, among the five catalysts studied. Another mono-functional catalyst Ni(10)/SBA-15 has also shown a sharp decline in HDO% from 2 to 4 h of TOS beyond which, fluctuating trend was observed. In contrast, for the bi-functional catalysts viz. Ni(5)-HPW(10)/SBA-15, Ni(10)-HPW(10)/SBA-15, Ni(15)-HPW(10)/SBA-15, HDO% increased with the increase in TOS from 2 to 4 h and further increase in TOS has not resulted any significant change in HDO%, except for Ni(5%)-HPW(10%)/SBA-15. Thus it could be concluded that the bi-functional catalysts have stable HDO selectivity compared to mono-functional catalysts.
Effect of TOS on the selectivity for different products obtained over different catalysts are presented in Fig. 10. HPW (10)/SBA-15 has given exclusively benzene and cyclohexane as HDO products, because of the dealkylation of propyl benzene and propyl cyclohexane. Selectivity for cyclohexane was found to be higher than benzene up to the TOS of 2 h and with further increase up to 12 h, selectivity for benzene became higher than the cyclohexane. This suggested that there may be two distinct active centers of HPW, one facilitated hydrogenation and other direct HDO pathway, of which the former get selectively deactivated with TOS. The only oxygenated product obtained was propyl phenol indicated the strong demethylation activity of HPW. The other mono-functional catalyst Ni(10)/SBA-15 yielded both the propyl anisole and propyl phenol as oxygenated products and the selectivity of former was higher than later, thus Ni based catalyst has less demethylation activity compared to HPW based catalyst. Benzene and cyclohexane alone obtained as HDO products by Ni(10%)/SBA-15 catalyst as well, but the selectivity for cyclohexane was much lower than benzene and did not change much with TOS. This suggested that the hydrogenation mechanistic pathway was less prevalent compared to HPW(10)/SBA-15, but at the same time, the active centers facilitated the hydrogenation were not selectively deactivated with TOS unlike over HPW(10)/SBA-15 catalyst. Exclusive formation of benzene and cyclohexane by both the mono-functional catalysts which are shorter in carbon length to be part of SAF, but valuable petrochemicals indicated that mono-functional catalysts are not useful for SAF production, but useful for the sustainable production of petrochemicals. All the five HDO products and two oxygenated products were obtained over the three bi-functional catalysts Ni(5, 10, 15)-HPW(10)/SBA-15. Selectivity for C9 compounds including propyl cyclohexane, the most desired HDO product because of being naphthenic and in the SAF range, did not change much with TOS for Ni(5)-HPW(10)/SBA-15 and Ni(10)-HPW(10)/SBA-15 whereas significantly changed for Ni(15)-HPW(10)/SBA-15 and the selectivity is the highest for Ni(10)-HPW(10)/SBA-15 at all TOS. Thus Ni(10%)-HPW(10%)/SBA-15 could be concluded as the best catalyst for the HDO of trans-anethole due to its higher and stable conversion, HDO selectivity and SAF components.
3.10 Effect of Ni loading into HPW(10)/SBA-15 on the HDO of trans-anethole
Effect of Ni loading into HPW(10)/SBA-15 on the average conversion% and HDO% at 400o C are presented in Fig. 11a. Increase in Ni% from 0 to 10% increased the conversion and then decreased with further increase to 15%. HDO% did not change much from 0 to 5% of Ni, but increased sharply with an increase to 10% and then decreased with further increase to 15% and hence concluded that 10% is the optimum Ni loading for HDO. HPW(10)/SBA-15, a purely acidic catalyst having 0% Ni, yielded benzene and cyclohexane exclusively as HDO products and propyl phenol as the only oxygenated product (Fig. 11b). With the addition of 5% Ni i.e. Ni(5)-HPW(10)/SBA-15, a bi-functional catalyst yielded C9 compounds of SAF range viz. propenyl benzene, propyl benzene and propyl cyclohexane as HDO products and both propyl anisole and propyl phenol as oxygenated products. Increase of Ni from 5 to 10% increased the selectivity for propyl cyclohexane from 1 to 11%, selectivity for C9 compounds from 11 to 35% and further increase to 15% decreased the selectivity for propyl cyclohexane and C9 compounds to 3 and 21% respectively. Thus it may be concluded that Ni(10)-HPW(10)/SBA-15 is the best catalyst among the five SBA-15 supported catalysts studied.
3.5 Effect of carbon support on the HDO of trans anethole
As Ni(10)-HPW(10)/SBA-15 has shown the highest conversion, selectivity and stability for the HDO of trans- anethole, effect of different carbon support viz. activated carbon, biomass derived activated carbon, and graphitic carbon sphere on the HDO of trans-anethole were studied in lieu of SBA-15. For the various catalysts, specific reaction rate (SRR), turnover frequency (TOF), conversion% and selectivity for HDO (HDO%) are presented in Table 2.
Table 2
Catalytic activity for HDO of trans anethole over different catalysts at 400°C
Catalysts
|
SRR (mol g-1 s-1)
|
TOF (s-1)
|
Conversion%
|
HDO%
|
Ni(5)-HPW(10)/SBA-15
|
0.951x10− 5
|
0.117x10− 1
|
51.45
|
35
|
Ni(10)-HPW(10)/SBA-15
|
1.64x10− 5
|
8.90x10− 4
|
81.86
|
79.5
|
Ni(15)-HPW(10)/SBA-15
|
1.347x10− 5
|
5.27x10− 4
|
72.7
|
58.5
|
Ni(10)/SBA-15
|
0.95x10− 5
|
5.59x10− 3
|
51.39
|
37.7
|
HPW(10)/SBA-15
|
0.592x10− 5
|
0.17x106
|
32.14
|
38.7
|
Ni(10%)-HPW(10%)/AC
|
14.5x10− 6
|
8.52x10− 3
|
78.76
|
6.57
|
Ni(10%)-HPW(10%)/ BDAC
|
17.9x10− 6
|
10.52x10− 3
|
97.94
|
15.88
|
Ni(10%)-HPW(10%)/ GCS
|
12.4x10− 6
|
7.29x10− 3
|
67.46
|
12.17
|
Among all the catalysts studied Ni(10)-HPW(10)/BDAC gave the highest conversion of 97.94% and SRR of 1.79 x 10− 5 mol g− 1 s− 1, but its HDO% was only 15.88%. Ni(10)-HPW(10)/AC has shown the HDO% of 6.57%, the lowest among all the catalysts studied and for Ni(10)-HPW(10)/GCS, HDO% was 12.17%. On the other hand Ni(10)-HPW(10)/SBA-15 has shown the highest HDO% of 79.5 among all the catalysts studied with comparable SRR of 1.64 x 10− 5, suggested that SBA-15 is a better support for HDO of trans-anethole compared to all the three carbon supports. This reinforces the conclusion of HRTEM that there is a synergetic interaction between silica wall of SBA-15, Ni and HPW which may be the reason for higher HDO%.
Though there are no studies reported on HDO of trans anethole using fixed bed plug flow reactor, HDO of lignin derived phenolic compounds in octane solvent was studied in a batch reactor over MoO3 catalyst at 340o C and 3.5 MPa (3 MPa N2 + 0.5 MPa H2) for 6 h and reported 93.6% conversion for trans anethole with selectivity of 75.8, 8.3 and 1.2% for the propenyl benzene, propyl benzene, and propyl cyclohexane respectively while 12.9 and 1.7% for propenyl phenol and propyl phenol respectively[14]. Zhang et al[35] reported the HDO of mixture of lignin derived phenolic compounds in dodecane solvent over Ni/SiO2–ZrO2 catalysts at 300oC and H2 pressure of 5 MPa in a batch reactor for 8 h and got 100% conversion of trans anethole with 98.9% selectivity for propyl cyclohexane. Whereas in the present study Ni(10)-HPW(10)/SBA-15 catalyst which has given 81.86% conversion with 79.5% selectivity for HDO at 400oC, 0.1 MPa of H2 and residence time of just 0.1012 h and hence it may be concluded that Ni(10)-HPW(10)/SBA-15 is a much better catalyst for HDO of trans anethole.
3.6 Effect of Temperature on the HDO of trans anethole
Effect of temperature on the average conversion of trans-anethole and selectivity for HDO (HDO%) for the five SBA-15 supported catalysts are presented in the Fig. 12. For the three bi-functional catalysts viz. Ni(5)-HPW(10)/SBA-15, Ni(10)-HPW(10)/SBA-15, Ni(15)-HPW(10)/SBA-15, average conversion% increased with increase in temperature from 350 to 400 and then decreased with further increase to 450oC. While for the mono-functional catalyst HPW(10)/SBA-15 average conversion% increased continuously with increase in temperature, but the lowest among all the catalysts. For the other mono-functional catalyst Ni(10)/SBA-15, average conversion% was the highest among all the catalysts at 350oC, decreased with an increase in temperature to 400oC and then increased with further increase in temperature to 450oC. The selectivity for HDO (HDO%) increased from 350 to 400oC and decreased with further increase to 450oC for all the catalysts, except Ni(5)-HPW(10)/SBA-15 whose average HDO% was not much affected by temperature.
Increase in temperature markedly affected the selectivity of both the oxygenated and HDO products which varied with the nature of catalysts (Fig. 13). For Ni(10)/SBA-15, propyl anisole was the major oxygenated product at all temperatures though the selectivity for propyl phenol increased with increase in temperature from 350 to 450oC. At 350 and 400oC propenyl benzene and benzene were obtained as HDO products and only at 450oC additionally propyl benzene and cyclohexane were obtained indicated that the activation energy for hydrogenation is higher for mono-functional Ni(10)/SBA-15 catalyst. Other mono-functional catalyst HPW(10)/SBA-15 has shown a complimentary behavior. It has shown exclusive selectivity for propyl phenol among the oxygenated product at 350 and 400oC and only at 450oC propyl anisole was also found as the oxygenated product, indicated its stronger demethylation ability. Benzene and cyclohexane are the exclusive HDO products at 350 and 400oC and the selectivity for the cyclohexane increased with temperature indicated that surprisingly it has better hydrogenation ability than Ni(10)/SBA-15. Among the bi-functional catalysts, for Ni(5)-HPW(10)/SBA-15, average selectivity for cyclohexane decreased with increase in temperature while the selectivity for propenyl and propyl benzene increased from 350 to 400 and then decreased with further increase to 450oC. For Ni(10)-HPW(10)/SBA-15, average selectivity for all the HDO products including SAF components increased to the highest with increase in temperature from 350 to 400oC, and then decreased with further increase to 450oC. Similar trend was shown by the Ni(15)-HPW(10)/SBA-15 catalyst and hence it could be concluded that 400oC was the optimum temperature.