Carbon Nanocatalysts for Aquathermolysis of ... - Wiley Online Library

50 downloads 0 Views 2MB Size Report
Kun Guo,[a, b] Minfen Gu,[c] and Zhixin Yu*[a, b] ..... [11] D. S. Su, J. Zhang, B. Frank, A. Thomas, X. Wang, J. Paraknowitsch, ... meyer, R. L. Mccreery, Chem.
DOI: 10.1002/ente.201600522

Carbon Nanocatalysts for Aquathermolysis of Heavy Crude Oil: Insights into Thiophene Hydrodesulfurization Kun Guo,[a, b] Minfen Gu,[c] and Zhixin Yu*[a, b] Various carbon nanomaterials, including carbon nanotubes (CNTs), ketjenblack (KB) carbon, and graphene nanoplatelets (GNPs), are investigated as catalysts for the hydrodesulfurization (HDS) reaction in the aquathermolysis of heavy crude oil. XRD and Raman spectroscopy are used to characterize the crystallinity and graphitization of these carbon samples. Their morphologies and structures are determined by TEM. Surface functional groups are analyzed by FTIR spectroscopy, and specific surface area and pore volume are studied by nitrogen adsorption–desorption. To evaluate the HDS activity of carbon nanocatalysts, thiophene is employed as a model compound because its ring structure is commonly

present in asphaltene and resin components of heavy crude oil. The reactions are conducted under conditions similar to those of actual reservoir conditions of 120–180 8C and < 3 MPa. Reaction parameters, namely, catalyst dosage, temperature, and reaction time, are studied to optimize the HDS performance. Results show that factors such as degree of graphitization and specific surface area play critical roles in the enhanced HDS activity of GNPs and KB compared with that of CNTs. This work opens up the opportunity for the implementation of carbon nanomaterials as sustainable metal-free catalysts in the in situ upgrading and recovery of heavy crude oil.

Introduction Unconventional oil, including heavy and extra heavy oil, is drawing substantial attention as an important energy resource to supplement conventional fossil fuels and to meet the growing global energy demand. In recent decades, tremendous efforts have been dedicated to the effective exploration and production of heavy oils.[1] However, owing to the high viscosity; low hydrogen to carbon ratio; large hydrocarbon molecules; and considerable amounts of heteroatoms such as sulfur, nitrogen, oxygen, and metals, traditional thermal and chemical recovery methods often end up with high capital investment, low efficiency, and a large environmental footprint. Nevertheless, viscosity reduction by hydrocarbon cracking is widely accepted as an important mechanism to recover heavy oil. Among the reported recovery methods, aquathermolysis has become one of the most promising technologies for the in situ upgrading and recovery of heavy crude oil.[2] The aquathermolysis process refers to chemical reactions between superheated steam and heavy oil inside reservoirs. Once steam is injected, the solubility of nonpolar hydrocarbons in water increases with the water temperature.[3] The improved miscibility between water and oil allows a series of chemical reactions to occur, which include hydrocracking (HCK), hydrodesulfurization (HDS), hydrodenitrogenation (HDN), hydrodeoxygenation (HDO), and hydrodemetallization (HDM). These reactions contribute to the viscosity reduction of heavy crude oils by decomposing heavy hydrocarbons, especially asphaltene and resin components, into light fractions and then ease their flow to the production wells.[4] However, the visbreaking efficiency without additives is too low for large-scale field applications. Furthermore, the reactEnergy Technol. 2017, 5, 1228 – 1234

ed oil could face viscosity regression owing to repolymerization of free radicals generated from the cleavage of C@R (R = S, N, O, etc.) bonds during the reaction processes.[5] In this respect, the concept of introducing an active catalyst to improve the sluggish reaction kinetics has been proposed. Catalytic aquathermolysis can strengthen viscosity reduction and could prevent viscosity regression. In addition, the demand of temperature and pressure for the injected steam is reduced, which is beneficial for reducing the operational costs. So far, numerous studies have reported the utilization of various catalysts, such as minerals,[6] water-[7] or oil-soluble[5b,c] salts, zeolites,[8] and metal nanoparticles.[9] The remarkable catalytic performance achieved by these catalysts highlights the importance of developing cost-effective, active, and robust catalysts for field-scale implementation in aquathermolysis.

[a] K. Guo, Prof. Z. Yu Department of Petroleum Engineering Universitetet i Stavanger 4036 Stavanger (Norway) E-mail: [email protected] [b] K. Guo, Prof. Z. Yu The National IOR Centre of Norway Universitetet i Stavanger 4036 Stavanger (Norway) [c] Prof. M. Gu Center for Analysis and Testing Nanjing Normal University Nanjing 210046 (P.R. China) Supporting Information for this article can be found under http://dx.doi.org/10.1002/ente.201600522.

T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

1228

Among the reported catalysts, most of them involve transition metals or ions as the catalytic active species. The coordination effect between metal atoms and C@R (R = S, N, O, etc.) bonds can weaken the C@R bonds, which facilitates bond-cleavage processes. However, catalysts containing transition metals are normally expensive because of their scarcity. The catalyst manufacture process could be complex to prepare well-defined transition-metal nanoparticles. Furthermore, metal-based catalysts are troublesome to separate and recycle, particularly in processes involved in heavy oil upgrading, and can potentially become pollutants in the produced oil and result in detrimental environmental impact. These negative aspects, therefore, have hindered the widespread deployment of transition-metal-based catalysts in in situ heavy oil upgrading. One possible solution to address these issues is to develop efficient metal-free catalysts. Carbon materials have been used as catalysts or catalyst supports in heterogeneous catalysis for decades.[10] The multiple bonding approaches between carbon atoms allow the formation of linear, planar, and tetrahedral carbon structures. Physicochemical properties, such as surface area, porosity, electrical conductivity, and surface functionality, can be delicately tuned, which makes carbon materials versatile for various applications.[11] For example, carbon nanomaterials, such as carbon nanotubes (CNTs) and activated carbon, have been widely used as catalysts or supports in the catalytic oxidative desulfurization of commercial diesel fuels.[12] Surface quinone-type oxygen functional groups have been proposed as the active sites in oxidative dehydrogenation reactions.[13] Recently, carbon black and graphite materials have been reported with remarkable activity in the cracking of heavy crude oil.[14] A high viscosity reduction ratio of 96 % and reduced content of asphaltene and resin components are achieved by carbon nanocatalysts with a particle size of 21 nm. The application of such carbon catalysts will make the upgrading process more environmentally friendly and significantly reduce the operational costs. Because C@S bonds have the lowest bond-dissociation energies among heavy oil molecules, the critical mechanism for successful upgrading has been identified to be the cleavage of C@S bonds in several studies.[15] It is therefore important to develop commercially viable catalysts and processes for the catalytic decomposition of C@S bonds. Although desulfurization technology has been commercially available in petroleum refineries, reaction conditions such as temperatures of 320–380 8C and H2 pressures of 4–8 MPa are normally applied, which are virtually impossible to implement inside the actual reservoirs. To this end, this work focuses on the study of the HDS reaction of the aquathermolysis process by employing several widely used carbon nanomaterials as catalysts, including CNTs, ketjenblack (KB) carbon, and graphene nanoplatelets (GNPs), at reservoir-relevant conditions of temperatures of 120–180 8C and pressures of < 3 MPa. These carbon nanocatalysts can be utilized without a preactivation process, which would make them suitable for reservoir applications. Their catalytic activities towards the HDS reaction are evaluated with thiophene as a model compound, Energy Technol. 2017, 5, 1228 – 1234

which represents the typical sulfur-containing ring structure in heavy hydrocarbon molecules. Parameters such as catalyst dosage, temperature, and reaction time are also investigated to improve the catalytic performance. The results presented herein can contribute to the future advancement of metalfree, carbon-based nanocatalysts in the implementation of in situ upgrading and recovery of heavy crude oil. It is worth mentioning that in situ upgrading of heavy oil will also result in a less demanding HDS unit in downstream oil refineries. The application of metal-free carbon catalysts will therefore contribute to sustainable chemistry for the production and refining of crude oils.

Results and Discussion Crystallinity or graphitic order of the carbon nanomaterials are examined by XRD and the diffraction patterns are shown in Figure 1. All diffraction peaks of the four samples can be indexed to standard carbon materials. The sharp characteristic peaks at 26.48 of CNT1, CNT2, and GNPs, corresponding to the {002} facets, indicate good crystallinity of these carbon samples. The crystallinity can be attributed to the ordered stacking of the multilayer graphitic-like sheets. On the contrary, KB carbon displays a broad and weak diffraction peak at 26.48, which suggests the amorphous nature of KB. This result agrees with the traditional structure of carbon black, which is composed of stacks of small and disordered graphitic-like sheets.

Figure 1. XRD patterns of the carbon nanomaterials: a) CNT1 and CNT2, b) KB carbon, and c) GNPs. Commercial CNTs with outer diameters of 10–20 and 40–60 nm are denoted as CNT1 and CNT2, respectively.

T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

1229

Furthermore, Raman spectroscopy is utilized to analyze these carbon allotropes. Figure 2 shows the Raman spectra of the four carbon samples and Table 1 lists the positions of D and G bands, as well as the calculated intensity ratio of the D and G bands (ID/IG). The two principle bands appearing at n˜ & 1350 and 1580 cm@1 in the spectra are the D and G bands, respectively. The D band originates from a hybridized vibrational mode associated with carbon atoms close to the edges and defects, and it indicates the presence of disorder in the sample structure. The G band represents the in-plane vibrational mode related to the sp2-hybridized carbon atoms in the graphene layer. The band intensity ratio is often used as a measure of the degree of graphitization of carbon materials.[16] KB shows the highest ID/IG ratio of 1.87, which suggests the highest disorder. This agrees with the fact that carbon black is normally composed of stacks of small and disordered graphitic-like sheets. In addition, the ID/IG ratio of CNT1 (0.84) is larger than that of CNT2 (0.44), which implies that CNT2 possesses a higher degree of graphitization than CNT1. Considering that CNTs are essentially rolled-up

graphene sheets, the larger diameter of CNT2 with more sheet layers is expected to exhibit better graphitic order than that of CNT1. With an ID/IG ratio of 0.28, GNPs give the best degree of graphitization among the four carbon samples; this is attributed to the nanoplatelet structure composed of multiple and ordered graphitic-like sheets. The 2 D band appearing at n˜ & 2700 cm@1 with an asymmetric shape and a low intensity further confirms the nature of multiple layers. The results observed from XRD and Raman analysis are in accordance with each other. TEM was further performed to characterize the morphology of all carbon nanomaterials. As shown in Figure 3, CNT1 (Figure 3 a and b) presents smaller outer diameters than those of CNT2 (Figure 3 c and d), as expected from the mate-

Figure 2. Raman spectra of the carbon nanomaterials.

Table 1. Summary of the Raman spectroscopy results of the carbon nanomaterials. Carbon nanomaterial

D

CNT1 CNT2 KB carbon GNPs

1348 1350 1346 1351

Band position [cm@1] G

ID/IG

1577 1579 1600 1579

0.84 0.44 1.87 0.28

Energy Technol. 2017, 5, 1228 – 1234

Figure 3. High- and low-magnification TEM images of CNT1 (a, b), CNT2 (c, d), KB carbon (e, f), and GNP (g, h).

T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

1230

rials data sheet. The CNTs are of good quality because the diameters of the CNTs are uniform. The lengths of both CNT samples are on the micron scale and the multiwalled tubular structure can be confirmed. KB carbon (Figure 3 e and f) displays the appearance of aggregated nanopellets with sizes of tens of nanometers. In addition, these pellets are amorphous, which is in agreement with the XRD and Raman results. In Figure 3 g and h, the stacked platelet shape of GNPs on the micron scale is observed. The poor image contrast of these platelets indicates a small thickness, which agrees with the thickness of 10–15 nm in the materials data sheet. The TEM observations also confirm the good quality and purity of these carbon nanomaterials. Figure 4 displays the thermogravimetric analysis (TGA) curves of the four carbon nanomaterials in a 50 mL min@1 flow of air, and Table 2 summarizes the weight loss observed during TGA experiments. TGA analysis further confirms the graphitic order of the carbon nanomaterial, as well as their high purity, because all carbon nanomaterials lose weight mainly over a very narrow temperature range.

of the four carbon samples are listed in Table 3. KB carbon possesses an ultrahigh specific surface area (SSA) of 1339.25 m2 g@1 and a large pore volume of 1.71 cm3 g@1, which are much higher than those of CNTs and GNP. It should be noted that KB, as one activated carbon, is full of defects in its structural framework, which could potentially result in superior reactivity.

Table 3. Measured SSA and BJH pore volume of the carbon nanomaterials. Carbon nanomaterial

SSA [m2g@1]

BJH pore volume [cm3g@1]

CNT1 CNT2 KB carbon GNPs

113.75 57.71 1339.25 62.51

0.21 0.09 1.71 0.14

FTIR spectroscopy was performed to characterize the surface functional groups. The resulting FTIR spectra of the four carbon samples are shown in Figure 5. All samples present four major bands at similar locations, which implies that they have similar surface functional groups. This characteristic is important for a fair comparison of their catalytic activity in the subsequent HDS reaction. The most intense band at n˜ & 3424 cm@1 can be assigned to O@H stretching vibrations in alcohols, phenols, and @COOH. The bands at n˜ & 1635, 1557, and 1103 cm@1 correspond to C=O, C=C, and C@O stretching vibrations, respectively.[17] It could be speculated that carboxylic, hydroxyl, and unsaturated bond groups are present on the surface of these carbon nanomaterials.

Figure 4. TGA curves of the four carbon nanomaterials in a 50 mL min@1 flow of air.

Table 2. Weight loss of the four carbon nanomaterials during TGA analyses. Carbon nanomaterial

Weight loss [%]

CNT1 CNT2 KB carbon GNP

98.0 97.2 99.5 99.3

Figure S1 in the Supporting Information shows the nitrogen adsorption–desorption isotherms of four carbon nanomaterials, which exhibit type IV isotherms, according to the IUPAC classification, indicating the porous structure. The hysteresis loops of both CNT samples are mainly located in a higher relative pressure region than those of KB and GNP, which suggests a high proportion of large mesopores in CNTs. The measured Brunauer–Emmett–Teller (BET) surface area and Barrett–Joyner–Halenda (BJH) pore volume Energy Technol. 2017, 5, 1228 – 1234

Figure 5. FTIR spectra of the carbon nanomaterials.

The carbon nanomaterials were utilized as catalysts in the HDS reaction with thiophene as the sulfur-containing model compound. To optimize the catalytic performance, a parametric study in terms of catalyst dosage, temperature, and reac-

T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

1231

tion time was conducted. Figure 6 shows the thiophene conversion rate with different amounts of KB catalysts together with the blank experiment for comparison. When no catalyst is applied, thiophene is barely converted, which indicates the chemical stability of this aromatic structure. Increasing the dosage from 20 to 50 mg delivers a significant increment of thiophene conversion rate from 2.68 to 7.41 %, which could be attributed to the increased number of catalytically active sites with more KB catalyst. However, further doubling the amount does not give a big conversion improvement; this suggests that the contact between thiophene and KB catalyst is approaching saturation. Therefore, in all subsequent HDS reactions, the catalyst dosage was maintained at 50 mg.

Figure 6. Thiophene conversion with different amounts of KB carbon catalysts and without catalyst at 160 8C for 24 h.

Figure 7 presents the thiophene conversion of the four carbon nanocatalysts at temperatures of 120, 160, and 180 8C. Along with elevated temperature, the conversion rate is slightly enhanced for all catalysts, which suggests that a relatively low temperature of 120 8C is thermodynamically possible to overcome the activation energy of the thiophene HDS reaction. At each specific temperature, the HDS activity is generally in the order of KB carbon > GNP > CNT1 > CNT2. KB as a type of carbon black is composed of essentially

graphitic-like sheets stacked by random cross-linking, which results in high porosity with numerous edges and cracks. These defects can serve as catalytically active sites. Meanwhile, the ultrahigh SSA suggests its high capability of adsorbing reactants and making thiophene molecules interact with the catalytically active defects. Thus, KB carbon is expected to deliver superior activity. However, GNPs, with a SSA approximately half that of CNT1 and the same as that of CNT2, also show surprisingly high activity. It is worth mentioning that the catalytic activity of metal-free carbon nanomaterials is normally correlated with parameters such as SSA, pore size and volume, and surface functional groups and defects. Herein, it turns out that activity correlation determined by these factors might be inadequate to explain the high activity of GNPs. Other physicochemical parameters, including electrical and thermal conductivity, which are correlated to the degree of graphitization, should also be taken into consideration. For instance, the group of Li found that the CNT catalyst with a higher degree of graphitization presented higher activity in the oxidation of dibenzothiophene because higher electrical conductivity contributed to electron transfer involved in oxidation–reduction reactions.[12a] With regard to graphene, highly mobile p electrons result in remarkable electrical conductivity, which could also promote the HDS of thiophene. Furthermore, the interaction between thiophene molecules and the hexagonal lattices of graphene is also stronger, leading to a higher adsorption capacity of thiophene compared with that of other carbon materials.[18] Therefore, GNPs could exhibit a large adsorption capacity, although it has a small SSA. It can be proposed that the excellent electrical conductivity, combined with their remarkable adsorption capacity, have contributed to their high catalytic activity for thiophene HDS. Herein, as concluded from the FTIR spectra, the four carbon nanocatalysts present similar surface functional groups with similar intensity. Neither the SSA nor the pore volume are directly correlated with the activity sequence. Therefore, it could be speculated that both the SSA and degree of graphitization are critical factors correlated to the catalytic activity of carbon nanocatalysts in thiophene desulfurization. Because GNP has a relatively low SSA and pore volume, while exhibiting surprisingly high activity, we further speculate that the high degree of graphitization is of paramount importance in the carbon-based HDS reaction. To investigate the influence of reaction time on the catalytic activity, HDS reactions with carbon nanocatalysts were conducted for 12, 24, and 48 h at 160 8C. As shown in Figure 8, the HDS activity of carbon nanocatalysts is still in the same order of KB carbon > GNP > CNT1 > CNT2. Nevertheless, thiophene conversion does not show a major increment for all catalysts when extending the reaction time from 12 h to 24 or 48 h, which suggests that a period of 12 h is sufficient for the HDS of thiophene under moderate reaction conditions. This result is in agreement with the previous report by Liu et al.[19]

Figure 7. Thiophene conversion with four carbon nanocatalysts at different temperatures of 120, 160, and 180 8C for 24 h.

Energy Technol. 2017, 5, 1228 – 1234

T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

1232

plied by Shenzhen Nanotech Port Co., Ltd. Ketjenblack EC600 JD was ordered from AkzoNobel. GNPs were bought from American Elements. The detailed specifications from the suppliers of carbon nanomaterials are listed in Table 4. Thiophene (+ 99 %), 1,2,3,4-tetrahydronaphthalene (tetralin, 99 %), and toluene (+ 99.5 %) were purchased from Sigma–Aldrich and used as received without further treatment.

Catalyst characterization

Figure 8. Thiophene conversion with four carbon nanocatalysts at 160 8C for different reaction periods of 12, 24, and 48 h.

Conclusions Different carbon nanomaterials, including carbon nanotubes (CNTs), ketjenblack (KB) carbon, and graphene nanoplatelets (GNPs), were utilized as catalysts in the hydrodesulfurization (HDS) reaction with thiophene as a sulfur-containing model compound to investigate the potential application of metal-free carbon nanocatalysts for the in situ upgrading and recovery of heavy crude oil. Experimental results showed that KB carbon with an ultrahigh specific surface area (SSA) and GNPs with superior degree of graphitization exhibited similar HDS activities that were much better than that of CNTs. It has been proposed that the degree of graphitization of carbon-based nanocatalysts is of great importance in HDS reactions. Further investigation is necessary to clarify the specific roles of critical factors, such as SSA and degree of graphitization, as well as the number of surface functional groups, on the catalytic activity of carbon nanocatalysts. Nevertheless, this work demonstrates an interesting topic on the application of carbon nanomaterials as metal-free catalysts in the in situ upgrading and recovery of heavy crude oil, which will contribute to more sustainable chemistry in terms of the production and refining of heavy crude oils.

Experimental Section

XRD was conducted to characterize the crystal phase and structure of the samples, and the patterns were recorded on a BrukerAXS Microdiffractometer (D8 ADVANCE) by using a CuKa radiation source (l = 1.54 c). Scanning angles for all samples were set in the 2q range of 10–908 with a step interval of 2.258 min@1. Peaks were indexed according to the database established by the Joint Committee on Powder Diffraction Standards (JCPDS). The Raman spectra were obtained by using a JOBIN YVON HR800 laser confocal micro-Raman spectrometer equipped with an optical microscope, a charge-coupled device (CCD) camera, and an argon-ion laser source. The laser provided 50 mW power at l = 514.5 nm for the exciting line. Raman spectra were fitted with a Gaussian function to calculate the D and G band areas, which were used to represent the band intensities, ID and IG, respectively. TEM was performed to observe the microstructures and morphologies of the carbon nanomaterials on the JEOL JEM-2100F microscope with an accelerating voltage of 200 kV. TEM specimens were prepared by dispersing samples in ethanol for 15 min by using an ultrasonic bath. One droplet of the suspension was dropped onto a copper grid (400 mesh, TAAB) and dried in air. TGA was conducted on PerkinElmer Pyris 1 TGA apparatus at a heating rate of 10 8C min@1 from room temperature to 950 8C under a 50 mL min@1 flow of air. Nitrogen adsorption–desorption measurements were performed at liquid nitrogen temperature of 77 K on a Micromeritics TriStar II surface area and porosity analyzer. SSA was calculated by using the BET method and the pore volume was analyzed by using the BJH method. FTIR spectra were recorded on a Nicolet NEXUS 670 spectrometer by using a deuterated triglycine sulfate detector. All spectra were recorded with a resolution of 4 cm@1 for 32 scans in the spectral range between n˜ = 400 and 4000 cm@1. The background spectrum of air was measured as a single beam and used as reference.

Materials All carbon nanomaterials were purchased and used as received without further treatment. Multiwalled CNT samples were sup-

Table 4. Specifications of the carbon nanomaterials. Carbon nanomaterial

SSA [m2 g@1]

Morphology

CNT1 CNT2 KB carbon GNPs

100–160 40–70 1400 50–80

outer diameter: 10–20 nm; length: > 5 mm outer diameter: 40–60 nm; length: > 5 mm pellets nanoplatelets; thickness: 10–15 nm; average diameter: 15 mm

Energy Technol. 2017, 5, 1228 – 1234

HDS of thiophene The HDS reaction of thiophene was performed in a 50 mL Teflon-lined stainless-steel autoclave reactor. Initially, thiophene, as the sulfur-containing model compound, and tetralin, as the hydrogen donor, in a total volume of 20 mL together with the carbon nanocatalysts were added to the reactor. The Teflon container was then sonicated for 10 min to disperse the carbon nanocatalysts. The mixture was sealed tightly and the reactor was placed inside an electric oven preheated to 120–180 8C for a duration of 12–48 h. After the reaction, the reactor was cooled naturally. The thiophene content after the reaction was analyzed to evaluate the HDS activity of different catalysts; this was performed on an Agilent 7820 gas chromatograph equipped with an

T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

1233

Agilent J&W HP-5 column (length 30 m, diameter 0.25 mm, thickness 1.0 mm) and a flame-ionization detector (FID). For each analysis, liquid (80 mL) from the reactor was mixed with toluene (1.6 mL) as a solvent, and this mixture was utilized to measure the thiophene concentration. Each sample test was repeated three times and the average value was taken as the thiophene concentration.

[7]

[8]

Acknowledgements [9]

We gratefully acknowledge financial support from the National IOR Centre of Norway at the University of Stavanger. We thank Vidar Folke Hansen and Priscille Cuvillier for help with TEM characterization and assistance from Jorunn Hamre Vr,lstad in GC analysis.

[10]

[11]

Keywords: aquathermolysis · carbon · nanostructures · sulfur · surface analysis [1] a) O. Muraza, A. Galadima, Fuel 2015, 157, 219 – 231; b) O. Muraza, J. Anal. Appl. Pyrolysis 2015, 114, 1 – 10; c) R. Hashemi, N. N. Nassar, P. P. Almao, Appl. Energy 2014, 133, 374 – 387; d) A. Mai, J. Bryan, N. Goodarzi, A. Kantzas, J. Can. Pet. Technol. 2009, 48, 27 – 35; e) P. P. Almao, Can. J. Chem. Eng. 2012, 90, 320 – 329; f) A. Shah, R. Fishwick, J. Wood, G. Leeke, S. Rigby, M. Greaves, Energy Environ. Sci. 2010, 3, 700 – 714; g) S. K. Maity, J. Ancheyta, G. Marroqu&n, Energy Fuels 2010, 24, 2809 – 2816. [2] Aquathermolysis of Heavy Oils: J. B. Hyne, J. W. Greidanus, J. D. Tyrer, D. Verona, C. Rizek, P. D. Clark, R. A. Clarke, J. Koo, The Second International Conference on Heavy Crude and Tar Sands (Caracas, Venezuela), 1982, pp. 25 – 30. [3] J. Li, Y. Chen, H. Liu, P. Wang, F. Liu, Energy Fuels 2013, 27, 2555 – 2562. [4] L. Dong, Y. J. Liu, K. M. Xu, F. J. Zhao, W. W. Liu, X. W. Kong, Adv. Mater. Res. 2013, 773, 298 – 303. [5] a) Y. Yi, S. Li, F. Ding, H. Yu, Pet. Sci. 2009, 6, 194 – 200; b) J. J. Li, X. Wang, X. D. Tang, F. Wang, D. Y. Qing, Energy Fuels 2015, 29, 7771 – 7780; c) A. V. Galukhin, A. A. Erokhin, Y. N. Osin, D. K. Nurgaliev, Energy Fuels 2015, 29, 4768 – 4773. [6] a) The Effects of Minerals on Heavy Oil and Bitumen Chemistry when Recovered Using Steam-Assisted Methods: W. Montgomery, M. Sephton, J. Watson, H. Zeng, SPE Heavy Oil Conference-Canada

Energy Technol. 2017, 5, 1228 – 1234

[12]

[13]

[14] [15]

[16]

[17] [18] [19]

(Calgary, Alberta, Canada), 2014, DOI: 10.2118/0115-015-JCPT; b) H. Fan, J. Can. Pet. Technol. 2003, 42, 11 – 14. a) P. D. Clark, M. J. Kirk, Energy Fuels 1994, 8, 380 – 387; b) Experimental Evaluation of Transition-Metal Salt Solutions as Additives in Steam Recovery Processes: O. R. Rivas, R. E. Campos, L. G. Borges in SPE Annual Technical Conference and Exhibition (Houston, Texas, USA), 1988, DOI: 10.2118/18076-MS. a) A. S. M. Junaid, M. M. Rahman, G. Rocha, W. Wang, T. Kuznicki, W. C. McCaffrey, S. M. Kuznicki, Energy Fuels 2014, 28, 3367 – 3376; b) A. S. M. Junaid, C. Street, W. Wang, M. M. Rahman, W. An, W. C. McCaffrey, S. M. Kuznicki, Fuel 2012, 94, 457 – 464. a) A. Al-Marshed, A. Hart, G. Leeke, M. Greaves, J. Wood, Energy Fuels 2015, 29, 6306 – 6316; b) R. Hashemi, N. N. Nassar, P. P. Almao, Energy Fuels 2014, 28, 1351 – 1361; c) R. Hashemi, N. N. Nassar, P. P. Almao, Energy Fuels 2014, 28, 1338 – 1350; d) S. Alkhaldi, M. M. Husein, Energy Fuels 2014, 28, 643 – 649. a) P. Trogadas, T. F. Fuller, P. Strasser, Carbon 2014, 75, 5 – 42; b) C. J. Shearer, A. Cherevan, D. Eder, Adv. Mater. 2014, 26, 2295 – 2318; c) D. S. Su, S. Perathoner, G. Centi, Chem. Rev. 2013, 113, 5782 – 5816. D. S. Su, J. Zhang, B. Frank, A. Thomas, X. Wang, J. Paraknowitsch, R. Schlogl, ChemSusChem 2010, 3, 169 – 180. a) W. Zhang, H. Zhang, J. Xiao, Z. X. Zhao, M. X. Yu, Z. Li, Green Chem. 2014, 16, 211 – 220; b) K. G. Haw, W. A. W. Abu Bakar, R. Ali, J. F. Chong, A. A. A. Kadir, Fuel Process. Technol. 2010, 91, 1105 – 1112. J. Zhang, D. Su, A. Zhang, D. Wang, R. Schlçgl, C. Hebert, Angew. Chem. Int. Ed. 2007, 46, 7319 – 7323; Angew. Chem. 2007, 119, 7460 – 7464. K. Li, B. Hou, L. Wang, Y. Cui, Nano Lett. 2014, 14, 3002 – 3008. a) P. D. Clark, R. A. Clarke, J. B. Hyne, K. L. Lesage, AOSTRA J. Res. 1990, 6, 29 – 39; b) P. D. Clark, N. I. Dowling, K. L. Lesage, J. B. Hyne, Fuel 1987, 66, 1699 – 1702; c) P. D. Clark, N. I. Dowling, J. B. Hyne, K. L. Lesage, Fuel 1987, 66, 1353 – 1357; d) P. D. Clark, J. B. Hyne, Fuel 1984, 63, 1649 – 1654; e) P. D. Clark, K. L. Lesage, G. T. Tsang, J. B. Hyne, Energy Fuels 1988, 2, 578 – 581. a) A. Cuesta, P. Dhamelincourt, J. Laureyns, A. Martinezalonso, J. M. D. Tascon, Carbon 1994, 32, 1523 – 1532; b) Y. Wang, D. C. Alsmeyer, R. L. Mccreery, Chem. Mater. 1990, 2, 557 – 563. J. M. OQReilly, R. A. Mosher, Carbon 1983, 21, 47 – 51. L. F. Wang, R. T. Yang, C. L. Sun, AIChE J. 2013, 59, 29 – 32. M. Khalil, R. L. Lee, N. Liu, Fuel 2015, 145, 214 – 220.

Manuscript received: August 24, 2016 Revised manuscript received: September 12, 2016 Accepted manuscript online: September 26, 2016 Version of record online: February 7, 2017

T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

1234