Catalyst Used in Fluid Catalytic Cracking (FCC) Unit As A Support of NiMoP Catalyst For Light Cycle Oil Hydroprocessing
Catalyst Used in Fluid Catalytic Cracking (FCC) Unit As A Support of NiMoP Catalyst For Light Cycle Oil Hydroprocessing
Fuel
journal homepage: www.elsevier.com/locate/fuel
G RA P H I C A L AB S T R A C T
A R T I C L E I N F O A B S T R A C T
Keywords: The behavior of the NiMoP catalysts prepared by means of impregnation of different supports (MCM-41, SBA-15,
LCO catalyst used in FCC unit) has been studied, using them in the hydroprocessing of Light Cycle Oil (LCO) with the
Hydroprocessing aim of maximizing the sulfur removal and reduce polyaromatics content; so the hydroprocessed LCO can be
Hydrodesulfurization hydrocracked in another unit with metal noble catalysts.
Hydrodearomatization
Catalysts have been characterized through several techniques: N2 adsorption–desorption, elemental analysis
FCC catalyst
Deactivation
(ICP-AES), X-ray diffraction (XRD), acidity measurement by temperature programmed desorption (TPD) of tert-
butylamine, temperature programmed reduction (TPR), X-ray photoelectron spectroscopy (XPS), UV–Vis spec-
troscopy, transmission electron microscopy (TEM), and temperature programmed oxidation (TPO) (spent cat-
alysts). The hydroprocessing runs have been carried out in a fixed bed reactor working in trickle bed regime at
320–400 °C; 80 bar; H2:LCO ratio, 1000 mLNH2 mL−1 LCO; space time, 0.2 h, and time on stream (TOS), 8 h.
The catalyst NiMoP/HY-Al2O3 (spent FCC catalyst) stands out due to its high hydrodesulfurization activity
attributable to the ease of impregnation of metals over FCC catalyst and the accessibility of the components of
the LCO to the hierarchical porous structure. Its activity for reducing polyaromatics compounds of LCO is also
remarkable, without diminishing the diesel fraction content by overcracking. Besides, the deposition of coke
over this catalyst is scarce, since the retention of the high molecular weight molecules of LCO in its porous
structure is minimized.
1. Introduction the increasingly strict policies about fuel composition [1]. In this sce-
nario, refineries have to steadily adapt their units for the processing, in
The petroleum industry has to meet the growing demand of fuels the medium and long term, of the non-conventional crudes (heavy-oil,
solving the inherent difficulties of the lower quality of the crude and of extra-heavy oil, oil sands). Besides, the incorporation of new feeds
⁎
Corresponding author.
E-mail address: [email protected] (A. Gutiérrez).
https://doi.org/10.1016/j.fuel.2017.11.148
Received 26 June 2017; Received in revised form 27 October 2017; Accepted 29 November 2017
Available online 08 December 2017
0016-2361/ © 2017 Elsevier Ltd. All rights reserved.
R. Palos et al. Fuel 216 (2018) 142–152
derived from biomass (bio-refinery) [2–4] and residua from consumer 2. Experimental
society, such as plastics and tires (waste-refinery) [5–7] is being studied
with the aim of obtaining fuels and raw materials for the petrochemical 2.1. Catalysts preparation
industry (such as olefins and BTX aromatics).
Currently, a technological effort in industrial development is being For LCO hydroprocessing three NiMoP catalysts have been used
done to intensify the valorization of petroleum, by means of the with the same nominal content of Ni, Mo and P, of 3, 8 and 1 wt%,
adaptation (revamping) of the units to co-feed refinery secondary respectively. Three different acidic materials, with different porous
streams. Due to its high capacity and versatility, the catalytic cracking structure, have been used as support. Discarded FCC catalyst (referred
units (FCC) are receiving great attention for the co-feeding of non-usual as HY-Al2O3 from now on), being its composition: 53.79 wt% alumina,
streams [2,4,8,9]. Nevertheless, within the scheme of a conventional 13.8 wt% zeolite, 4 wt% metals (Cu, Ni, V, Fe, Na, REO) and binder, has
refinery, best prospects are focused in the usage of hydroprocessing been removed from the catalytic cracking unit of the Petronor S.A.
units [10], even though this process shows several limitations, such as refinery (Somorrostro, Biscay, Spain) and prior to its usage as support,
the high deactivation of the catalyst [11]. it has been regenerated for about 12 h in oxidizing atmosphere at
Among the refinery streams that potentially can be valorized, the 550 °C, in order to remove the residual coke. MCM-41 mesoporous
light cycle oil (LCO) awakens great interest as its conversion into fuels support has been synthesized from a sodium silicate solution using
has a high economic incidence. The LCO is produced in FCC units with a CnH2n+1N(CH3)3Br as surfactant [26]. The SBA-15 silicate has been
high yield (15–20%) but its composition, with high content of aro- prepared from triblock amphiphilic copolymers (polyethylene and
matics (75–90 wt%), sulfur (0.2–1.5 wt%) and low cetane number (CN polypropylene oxides) in acidic medium, with the aim of producing
15-25), limit its inclusion into refinery diesel pool [12]. The LCO can be highly ordered bidimensional mesophases of hexagonal silica [27].
valorized both via cracking [13,14] and hydroprocessing [15–18], The incorporation of Ni, Mo and P has been carried out by means of
being the latest the most effective and versatile route to achieve the consecutive wet impregnations with aqueous solutions (water solution
sulfur (hydrodesulfurization) and aromatics (hydrodearomatization) ratio, 20:1) of H3PO4, then (NH4)6Mo7O2·4H2O and, finally, Ni
content values demanded for gasoline and diesel pools. Although noble (NO3)·6H2O, being the following stages: (i) stabilization via stirring for
metal catalysts are very active not only for the hydrodesulfurization 24 h; (ii) drying with air at 110 °C during 12 h, and; (iii) calcination at
reactions, but also for the ring opening reactions implied in the LCO 500 °C in N2 atmosphere during 2 h. Afterwards, catalysts have been
hydrocracking, they suffer from a fast deactivation caused by sulfur tableted and sieved (150–300 µm). Phosphorus has been incorporated
poisoning and coke deposition [19,20]. Consequently, the most ap- due to following benefits: (i) metal-support interaction decrease; (ii)
propriate strategy for LCO and another residual streams valorization ease Ni and Mo metals impregnation on the support; (iii) inhibition the
(REVAHS, REsidua VAlorization by means of Hydrogen Surplus) con- formation of inactive NiAl2O4 species; (iv) act as a second promoter;
sists on the hydrotreating of LCO in two stages [21] i) hydroprocessing and (v) improve the formation of type (II) Ni–Mo–S structures [28].
which is generally considered as the process to perform hydrogenation
and hydrogenolysis reactions; i.e. the removal of C=C and C–X bonds 2.2. Catalysts characterization
(where X=C, S, N, metals…), using low cost catalysts based on tran-
sition metals (Co, Ni, Mo, W) [15,16]; and; ii) hydrocracking, using The surface properties of the catalysts have been determined from
more severe conditions and noble metal based catalysts with acidic the N2 adsorption-desorption isotherms obtained with Micromeritics
supports, which are able to achieve high dearomatization yields by ASAP 2010 equipment. The specific surfaces of the samples have been
means of ring opening [22]. These catalysts show a moderate deacti- determined according to BET procedure using the data in the relative
vation when are used in the hydrocracking of the hydroprocessed pressure range of 0.05 to 0.2. Average pore sizes have been calculated
streams obtained in the first stage (low content of sulfur and polyaro- by means of BJH method.
matics) [23]. The acidity has been determined by temperature programmed des-
In this work, the influence of the support in the first stage of the orption (TPD) of tert-butylamine (t-BA) adsorbed at 100 °C, using a
REVAHS process for LCO valorization has been studied. For this pur- thermogravimetric-calorimetric analyzer Setaram DSC coupled in line
pose, three NiMoP catalysts prepared with different acidic supports to a mass spectrometer Balzers Quadstar 422, for the monitoring of the
have been used. MCM-41 and SBA-15 are mesoporous silica, with un- formed butenes as a result of t-BA cracking [29].
imodal and bimodal structures, respectively; and hexagonal and round The amount of coke deposited on the spent catalysts has been de-
porous shapes, respectively; having the MCM-41 support a higher termined by temperature programmed oxidation (TPO), by means of air
acidity. The usage of discarded FCC catalyst as a support is a promising combustion in a TA Instruments TGA Q5000 IR thermogravimetric
strategy due to its porous structure designed for easing the access of apparatus, coupled in line to a mass spectrometer Balzers Quadstar 422
high molecular weight molecules, such as those present in LCO [24], to monitor the CO2 formed in the coke combustion [30]. Previous to the
since the catalyst particles are prepared with a combination of HY combustion, an initial sweeping with N2 was carried out to quantify the
zeolite and a meso and macroporous matrix aluminas (HY-Al2O3) [25]. content of the retained compounds in the porous structure (possible
Furthermore, as the spent FCC catalyst is a residue highly available in coke precursors). In this stage, once the sample was stabilized at 100 °C,
refineries, if it is used in the first stage of the REVAHS process for LCO the temperature was raised up to 450 °C with a heating rate of
valorization, its lifespan will be extended. 5 °C min−1 under a N2 flow of 50 mL min−1. Temperature was held for
The main objective of the first stage of the REVAHS process for LCO 30 min and then the sample was cooled down again to 100 °C. After-
is achieving the maximum conversion of the sulfur compounds (hy- wards, the gas was switched to air (50 mL min−1) and stabilized for
drodesulfurization) with a remarkable removal of the polyaromatics 5 min. Then, to carry out the coke combustion, the sample was heated
(hydrodearomatization), but without significantly diminishing the following a ramp of 5 °C min−1 up to 550 °C, holding that temperature
boiling point range of its components. For this purpose, the spent FCC for 2 h with the aim of obtaining complete coke combustion. Finally,
(HY-Al2O3) catalyst is a suitable support, since it has a limited acidity, the sample was cooled down to room temperature.
as a consequence of the severe dealumination suffered in the con- Coke content (Cc, wt%) is defined from the mass of spent catalyst
secutive reaction-regeneration cycles, providing to it a low activity for (m0) and the mass of remaining catalyst after coke is burned (mcat):
the cracking of the aliphatic chains of LCO.
m 0−m cat
Cc = ·100
m cat (1)
143
R. Palos et al. Fuel 216 (2018) 142–152
Table 1 Table 3
LCO properties. Sulfur compounds content in LCO.
144
R. Palos et al. Fuel 216 (2018) 142–152
145
R. Palos et al. Fuel 216 (2018) 142–152
Fig. 3. Wide (a) and low (b) angle XRD patterns of the fresh catalysts.
Fig. 2. N2 adsorption-desorption isotherms of the fresh catalysts (a), mesopore size dis-
tribution (b), and macropore size distribution of the NiMoP/HY-Al2O3 catalyst (c).
146
R. Palos et al. Fuel 216 (2018) 142–152
species. The presence of Ni becomes evident with a well defined band at 3.2.1. Hydrodesulfurization
390 nm and a broad band in the 600–1000 nm range, with a maximum Fig. 9 shows the results for the hydrodesulfurization conversion and
at around 840 nm. These bands are related to octahedral Ni species and sulfur compounds content in the liquid products, for the three catalysts
to [Ni2+6O2−] species with octahedral coordination, respectively [49]. at studied temperatures. These results correspond to a relatively small
TEM micrographs of fresh catalysts before and after sulfurization space time (0.2 h) with the aim of comparing the catalysts and evaluate
are shown in Fig. 8. Analyzing the images of the fresh catalysts before the effect of the temperature, although under these conditions there is a
sulfurization, the corresponding morphology of every support can be partial sulfur removal. For the first stage of LCO hydroprocessing, as an
a) b)
HY-Al2O3
MCM-41
SBA-15
245 240 235 230 225 875 870 865 860 855 850 845
Binding energy (eV) Binding energy (eV)
147
R. Palos et al. Fuel 216 (2018) 142–152
148
R. Palos et al. Fuel 216 (2018) 142–152
3.3. Catalyst deactivation catalyst (HY-Al2O3) used as support. This macroporous matrix is sui-
table to make easier the entrance of the LCO molecules to the active
In Fig. 12, the N2 sweeping and coke combustion at programmed sites, but also to minimize the retention of the heaviest molecules as it
temperature (TPS and TPO, respectively) are shown for the three used favors their diffusion outside the catalyst particles.
catalysts. The contents of coke precursors and coke (displayed in Comparing the results of the coke combustion for the three used
Table 6) have been determined from the area below the curves. It can catalyst, it can be seen that coke content (Table 6), as well as, TPO
be seen the following trend for coke precursors content: NiMoP/SBA- profiles (Fig. 12) are different. The less coke content for the NiMoP/HY-
15 > NiMoP/MCM-41 > NiMoP/HY-Al2O3. This order corresponds to Al2O3 catalyst can be related to the properties of the FCC (HY-Al2O3)
the different capacity of the supports to retain the components of the catalyst (weak acidity and presence of a macroporous matrix). The role
reaction media in their porous structure. The lower value (16.03 wt%) of these properties have been well established in the catalytic cracking
corresponds to NiMoP/HY-Al2O3 catalyst, which also requires a lower and are crucial to mitigate the coke formation in the hydroprocessing
temperature for the desorption of retained compounds. This result can [55,56]. Wang et al. [57] have studied the mechanism of coke forma-
be attributed to the hierarchical porous structure of the discarded FCC tion in the deep hydroprocessing of vacuum residue, obtaining that
149
R. Palos et al. Fuel 216 (2018) 142–152
6000 90 80
Sulfur concentration (ppm)
80 70
Composition (wt%)
60
70
4000
50
320 ºC 60
3000
40
50
2000 30
40
20
1000 360 ºC
320 °C
360 °C
400 °C
400 ºC
30 10
0
HY-Al2O3 MCM-41 SBA-15 0
LCO HY-Al2O3 MCM-41 SBA-15
Fig. 9. Comparison for the three catalysts of the distribution of the sulfur compounds and
hydrodesulfurization conversion, for different reaction temperatures. Catalyst
Fig. 11. Comparison for the three catalysts of the concentration of the liquid product
P+iP N A1 A2 A3 lumps, for different reaction temperatures.
100
90
80
70
Composition (wt%)
60
50
40
30
20
320 °C
360 °C
400 °C
10
Fig. 12. TPS and TPO profiles of the used catalysts.
0
LCO HY-Al2O3 MCM-41 SBA-15 Table 6
Catalyst Coke precursors and coke contents of spent catalysts.
Fig. 10. Comparison for the three catalysts of the composition of the liquid product, for Catalyst HY-Al2O3 MCM-41 SBA-15
different reaction temperatures.
a
Coke precursors (wt%) 16.03 31.43 39.81
Coke (wt%)b 2.42 6.82 4.62
coke is mainly formed by fast condensation of intermediate polyaro-
matics obtained after a hydrogenation and latter cracking of phenan- Data obtained by TPSa and TPOb analyses.
threne and pyrene molecules, to form heavier compounds.
Comparing the TPO profiles plotted in Fig. 12, it can be seen that these polyaromatics towards structures that burn at higher tempera-
coke combustion over NiMoP/SBA-15 and NiMoP/MCM-41 catalysts tures due to its low H/C ratio [58].
starts at lower temperature and leads to a well defined peak, with the Another remarkable fact is that combustion temperature and coke
maximum combustion rate at around 400 °C. However, the TPO profile content (2.42 wt%) of the used NiMoP/HY-Al2O3 catalyst are inside the
for NiMoP/HY-Al2O3 catalyst corresponds to the combustion of a het- usual range in the FCC unit, so it is expected that this catalyst can re-
erogeneous coke, which shows only a shoulder at 400 °C and a max- cover its activity by means of regeneration via coke combustion in
imum in the 500–520 °C range. The greater difficulty in the combustion oxidizing atmosphere. The character of equilibrated material of the
of the NiMoP/HY-Al2O3 catalyst can also be related with the porous support guarantees that it maintains its properties (porous structure and
structure of this catalyst that allows the entrance of big molecules from acidity) along reaction-regeneration cycles, which is an advantage for
LCO. The macroporous of the matrix ease the condensation reactions of the industrial use of the FCC catalyst as a support of LCO
150
R. Palos et al. Fuel 216 (2018) 142–152
4. Conclusions
NiMoP catalysts are suitable for a first stage of the REVAHS process
for LCO (hydroprocessing of LCO) because they lead to high hydro-
desulfurization conversion at 400 °C with a remarkable reduction of
polyaromatics and without an important reduction of diesel fraction. As
a consequence, with enough space time, an adequate composition of
treated LCO can be obtained for its subsequent valorization by means of
hydrocracking with noble metal based catalysts.
Discarded FCC (HY-Al2O3) catalyst (residua from industrial unit)
has turned out to be an appropriate support for the metallic function
(NiMoP), because it allows the impregnation of these metals with high
dispersion. This property and the hierarchical porous structure of the
catalyst particle are appropriate for favoring the hydroprocessing of
LCO components with less diffusional limitations than those meso-
porous supports usually employed in the synthesis of transition metal
catalysts. The low acidity of equilibrated FCC catalyst is also suitable
for mitigating the cracking of linear chains of diesel fraction in LCO.
Additionally, the hierarchical porous structure of the FCC (HY-
Al2O3) catalyst is suitable (the same as in catalytic cracking) to mini-
mize the polyaromatics retention in the NiMoP/HY-Al2O3 catalyst. In
this way, the amount of coke deposited is remarkably lower than that
for the other catalysts that attenuates the blocking of active sites.
The coke content and combustion characteristics are similar to those
Fig. 13. N2 adsorption-desorption isotherms (a) and pore size distribution (b) of the used of FCC unit, so the regeneration of NiMoP/HY-Al2O3 catalyst will be
catalysts. viable under similar operating conditions, without affecting the support
properties.
Table 7
Properties of the spent catalysts after N2 sweeping.
Acknowledgements
Catalyst HY-Al2O3 MCM-41 SBA-15 This work has been carried out with financial support of the
Ministry of Economy and Competitiveness (MINECO) of the Spanish
SBET (m2 g−1) 42 114 149
VP/P0=0,2Ads (mL g−1) 11.7 32.3 42
Government (CTQ2013-46172-P, CTQ2015-67425R), the ERDF funds
VP/P0=0,5Ads (mL g−1) 12.7 49.7 61 of the European Union, the Basque Government (IT748-13) and the
VP/P0=0,5Des (mL g−1) 12.9 55.7 63 University of the Basque Country UPV/EHU (UFI 11/39).
Mesopore volume (mL g−1) 0.045* 0.12 0.28 The authors thank for technical and human support provided by
Average pore diameter (nm) 16.0* 6.2 11.0
SGIker of UPV/EHU and European funding (ERDF and ESF).
Data obtained by N2 adsorption-desorption.
* Without considering the macropores of the matrix. Appendix A. Supplementary data
151
R. Palos et al. Fuel 216 (2018) 142–152
[10] Al-Sabawi M, Chen J. Energy Fuels 2012;26:5373–99. [35] Ravikovitch PI, Domhnail SCÓ, Neimark AV, Schüth F, Unger KK. Langmuir
[11] Thybaut JW, Marin GB. Adv. Catal. 2016;59:109–238. 1995;11:4765–72.
[12] Yun GN, Lee YK. Appl. Catal. B Environ. 2014;150–151:647–55. [36] Huh S, Wiench JW, Yoo J, Pruski M, Lin VSY. Chem. Mater. 2003;15:4247–56.
[13] Gilbert WR, Morgado E, De Abreu MAS, De La Puente G, Passamonti F, Sedran U. [37] Ravikovitch PI, Neimark AV. Colloids Surfaces A Physicochem. Eng. Asp.
Fuel Process. Technol. 2011;92:2235–40. 2001;188:11–21.
[14] Del Rio D, Bastos R, Sedran U. Catal. Today 2013;213:206–10. [38] Nava R, Pawelec B, Castaño P, Álvarez-Galván MC, Loricera CV, Fierro JLG. Appl.
[15] Azizi N, Ali SA, Alhooshani K, Kim T, Lee Y, Il Park J, et al. Fuel Process. Technol. Catal. B Environ. 2009;92:154–67.
2013;109:172–8. [39] Boufaden N, Akkari R, Pawelec B, Fierro JLG, Zina MS, Ghorbel A. Appl. Catal. A
[16] Betancourt P, Marrero S, Pinto-Castilla S. Fuel Process. Technol. 2013;114:21–5. Gen. 2015;502:329–39.
[17] Peng C, Huang X, Duan X, Cheng Z, Zeng R, Guo R, et al. Catal. Today [40] Klimova T, Calderón M, Ramírez J. Appl. Catal. A Gen. 2003;240:29–40.
2016;271:149–53. [41] Gutiérrez A, Arandes JM, Castaño P, Olazar M, Bilbao J. Fuel 2012;94:504–15.
[18] Peng C, Fang X, Zeng R, Guo R, Hao W. Catal. Today 2016;276:11–8. [42] Selvam P, Bhatia SK, Sonwane CG. Ind. Eng. Chem. Res. 2001;40:3237–61.
[19] Gutiérrez A, Arandes JM, Castaño P, Aguayo AT, Bilbao J. Energy Fuels [43] Ortega-Domínguez RA, Vargas-Villagrán H, Peñaloza-Orta C, Saavedra-Rubio K,
2011;25:3389–99. Bokhimi X, Klimova TE. Fuel 2017;198:110–22.
[20] Castaño P, Gutiérrez A, Hita I, Arandes JM, Aguayo AT, Bilbao J. Energy Fuels [44] Hita I, Gutiérrez A, Olazar M, Bilbao J, Arandes JM, Castaño P. Fuel
2012;26:1509–19. 2015;145:158–69.
[21] Hita I, Aguayo AT, Olazar M, Azkoiti MJ, Bilbao J, Arandes JM, et al. Energy Fuels [45] Angeles MJ, Leyva C, Ancheyta J, Ramírez J. Catal. Today 2014;220–222:274–94.
2015;29:7542–53. [46] Morales-Ortuño JC, Ortega-Domínguez RA, Hernández-Hipólito P, Bokhimi X,
[22] Galadima A, Muraza O. Fuel 2016;181:618–29. Klimova TE. Catal. Today 2016;271:127–39.
[23] Hita I, Rodríguez E, Olazar M, Bilbao J, Arandes JM, Castaño P. Energy Fuels [47] Grosvenor AP, Biesinger MC, Smart RSC, McIntyre NS. Surf. Sci. 2006;600:1771–9.
2015;29:5458–66. [48] Wang X, Zhao Z, Chen Z, Li J, Duan A, Xu C, et al. Fuel Process. Technol.
[24] Corma A, Martínez C, Sauvanaud L. Catal. Today 2007;127:3–16. 2017;161:52–61.
[25] Al-Sabawi M, Atias JA, De Lasa H. Ind. Eng. Chem. Res. 2006;45:1583–93. [49] Silva-Rodrigo R, Castillo Jimenez H, Guevara-Lara A, Melo-Banda JA, Olivas
[26] Tanev PT, Pinnavaia TJ. Chem. Mater. 1996;8:2068–79. Sarabia A, Reyes de la Torre AI, Morteo Flores F, Castillo Mares A. Catal. Today
[27] Zhao D, Feng J, Huo Q, Melosh N, Fredrickson G, Chmelka B, et al. Science 2015;250:2–11.
1998;279:548–52. [50] Ding L, Zheng Y, Yang H, Parviz R. Appl. Catal. A Gen. 2009;353:17–23.
[28] Huirache-Acuña R, Pawelec B, Rivera-Muñoz EM, Guil-López R, Fierro JLG. Fuel [51] Ancheyta J, Speight JG. Hydroprocessing Heavy Oils Residua. CRC Press; 2007. p.
2017;198:145–58. 259–79.
[29] Epelde E, Gayubo AG, Olazar M, Bilbao J, Aguayo AT. Chem. Eng. J. [52] Trejo F, Rana MS, Ancheyta J, Chávez S. Fuel 2014;138:104–10.
2014;251:80–91. [53] Hita I, Cordero-Lanzac T, Gallardo A, Arandes JM, Rodríguez-Mirasol J, Bilbao J,
[30] Cordero-Lanzac T, Hita I, Veloso A, Arandes JM, Rodríguez-Mirasol J, Bilbao J, et al. Catal. Commun. 2016;78:48–51.
et al. Chem. Eng. J. 2017;327:454–64. [54] Nakano K, Ali SA, Kim HJ, Kim T, Alhooshani K, Il Park J, et al. Fuel Process.
[31] Gutiérrez A, Arandes JM, Castaño P, Olazar M, Barona A, Bilbao J. Chem. Eng. Technol. 2013;116:44–51.
Technol. 2012;35:653–60. [55] Cerqueira HS, Caeiro G, Costa L, Ramôa Ribeiro F. J. Mol. Catal. A Chem.
[32] Hita I, Palos R, Arandes JM, Hill JM, Castaño P. Fuel Process. Technol. 2008;292:1–13.
2016;144:239–47. [56] Guisnet M, Costa L, Ramôa Ribeiro F. J. Mol. Catal. A Chem. 2009;305:69–83.
[33] Van Herk D, Castaño P, Quaglia M, Kreutzer MT, Makkee M, Moulijn JA. Appl. [57] Wang W, Cai X, Hou H, Dong M, Li Z, Liu F, et al. Energy Fuels 2016;30:8171–6.
Catal. A Gen. 2009;365:110–21. [58] Ibarra Á, Veloso A, Bilbao J, Arandes JM, Castaño P. Appl. Catal. B Environ.
[34] Tuel A, Hubert-Pfalzgraf LG. J. Catal. 2003;217:343–53. 2016;182:336–46.
152