Catalytic Decomposition of n-C7 Asphaltenes Using Tungsten Oxides–Functionalized SiO2 Nanoparticles in Steam/Air Atmospheres

A wide range of technologies are being developed to increase oil recovery, reserves, and perform in situ upgrading of heavy crude oils. In this study, supported tungsten oxide nanoparticles were synthesized, characterized, and evaluated for adsorption and catalytic performance during wet in situ combustion (6% of steam in the air, in volumetric fraction) of n-C7 asphaltenes. Silica nanoparticles of 30 nm in diameter were synthesized using a sol–gel methodology and functionalized with tungsten oxides, using three different concentrations and calcination temperatures: 1%, 3%, 5% (mass fraction), and 350 °C, 450 °C, and 650 °C, respectively. Equilibrium batch adsorption experiments were carried out at 25 ℃ with model solutions of n-C7 asphaltenes diluted in toluene at concentrations from 100 mg·L−1 to 2000 mg·L−1, and catalytic wet in situ combustion of adsorbed heavy fractions was carried out by thermogravimetric analysis coupled to FT-IR. The results showed improvements of asphaltenes decomposition by the action of the tungsten oxide nanoparticles due to the reduction in the decomposition temperature of the asphaltenes up to 120 °C in comparison with the system in the absence of WOX nanoparticles. Those synthesis parameters, such as temperature and impregnation dosage, play an important role in the adsorptive and catalytic activity of the materials, due to the different WOX–support interactions as were found through XPS. The mixture released during the catalyzed asphaltene decomposition in the wet air atmosphere reveals an increase in light hydrocarbons, methane, and hydrogen content. Hydrogen production was prioritized between 300 and 400 °C where, similarly, the reduction of CO, CH4, and the increase in CO2 content, associated with water–gas shift, and methane reforming reactions occur, respectively. The results show that these catalysts can be used either for in situ upgrading of crude oil, or any application where heavy fractions must be transformed.


Introduction
The growth of the population, as well as the economic and technological developments, imply a continuous increase in the global demand for energy [1]. Consequently, unconventional crude oils, such as heavy crude oil (HO) and extra heavy crude oil (EHO), for adsorption and catalytic decomposition of asphaltenes in wet in situ combustion atmosphere. In this sense, the synthesis was evaluated using incipient wetness technique, using three different concentrations and calcination temperatures: 1%, 3%, and 5% (mass fraction), and 350 • C, 450 • C, and 650 • C, respectively. The nanomaterials were physical, and chemically characterized by surface area (S BET ), dynamic light scattering technique (DLS), X-ray photoelectron spectroscopy (XPS), and H 2 pulse chemisorption. The adsorption was conducted in batch-mode experiments at 25 • C and the wet combustion decomposition was carried out by thermogravimetric analysis coupled to a mass spectrometer.

Asphaltene Isolation and Characterization
Asphaltenes were isolated from a heavy crude oil (6.4 • API) with an approximately contain asphaltenes of 13% (mass fraction), obtained from a reservoir located in southeast Colombia. Asphaltenes were extracted from the heavy crude oil sample by mixing each crude oil in a 40:1 ratio with n-heptane following the ASTM D2892 and ASTM D5236 [57][58][59]. An excess of n-heptane was added to the crude oil in a volume ratio of 1:40 (g·mL −1 ). The mixture was then sonicated for 2 h at 25 • C and further stirred at 300 rpm for 20 h; then it was centrifuged for 30 min at 4500 rpm. The precipitated asphaltenes were filtered and washed with n-heptane until obtaining a shiny black tone. Details of the chemical and physical properties of the asphaltenes can be found in previous work [29].

Nanoparticles Preparation
The synthesis method used for obtaining SiO 2 nanoparticles was developed in previous works [60,61]. Some modifications to the protocol were made with the purpose of obtaining particle sizes that adjust to the IUPAC definition of nanoparticle (1 nm-100 nm) [62]. The procedure consists of the preparation of a silica suspension from an aqueous solution of sodium silicate (24.5% volumetric fraction, silicate/water) by the addition of hydrochloric acid (1 M) through the precipitation method under constant stirring. A sodium silicate solution (SSS) was prepared and heated until adjusting the temperature to 55 • C; using a constant stirring speed of 300 rpm, cetyltrimethylammonium bromide was dissolved until obtaining a concentration of 10% (in volumetric fraction) with respect to the SSS, then, a HCl solution was slowly added. The use of the acid had two purposes: first, to initiate the formation of a gel (pH: 9-9.5); and the second one was to acidify the solution and bring it to a pH of 1-1.5. Between two steps, the solution was stirred for 10 min, with no further addition of HCl, to prevent the formation of a strong silica gel. Then, the gel was aged for 20 h at 55 • C, and the nanoparticles obtained were washed using ethanol. Finally, the silica nanoparticles were dried for 8 h at 120 • C and calcined at 550 • C.

Surface Functionalization
Silica nanoparticles were dried at 120 • C. Sodium tungstate dihydrate Na 2 WO 4 ·2H 2 O and ammonium tungstate (NH 4 ) 10 H 2 (W 2 O 7 ) 6 were used. The precursors were dissolved in a volume of water equal to the pore volume of a determined amount of the catalyst support. To determine volume of liquid required, the selected solid silica is dried at room temperature for several weeks, weighed, and immersed in alcohol for 20 min The bubbling observed on the surface of the silica indicates that trapped air is being released. Once saturated, the silica is placed on an alcohol-saturated paper towel and immediately weighed at room temperature (20 ºC). This process is completed in less than 10 s. The internal volume of the silica is calculated by converting the difference in masses of the alcohol-saturated silica and the dry aggregate to volume using the density of the alcohol.
The impregnation procedure in the support was carried out using incipient wet impregnation, which was carried out by adding, drop by drop, the precursor solution to the support until saturation.
Then, the impregnated materials were then dried at 120 • C for six hours, followed by calcination [28,37]. The calcination temperature was varied at 350 • C, 450 • C, and 650 • C separately for 4 h, obtaining the tungsten oxide nanoparticles. The hybrid or supported hygroscopic salt (called SHS) in this study was named using the chemical symbol of the support (i.e., Si) followed by the symbol of tungsten (i.e., W). If the precursor salt was the sodium tungstate dihydrate, the Na symbol was added after W. Then, it was included in the mass fraction percentage of the nominal tungsten used [28]. The number after the space indicates the calcination temperature of the material. For example, the sample SiWNa1 350 corresponds to nanosilica doped with 1% (in mass fraction) of sodium tungstate dihydrate and calcined at 350 • C. Table 1 lists the characteristics of the tungsten functionalized silica nanoparticles (WSN) used in this study.

Nanoparticles Characterization
The silica nanoparticles were characterized using dynamic light scattering technique (DLS) for hydrodynamic diameter by a nanoplus-3 from Micromeritics, USA. Characterization of functional groups on the nanoparticles and WSN surface was performed by Fourier transform infrared spectroscopy (FTIR) with an IRAffinity-1 FTIR spectrophotometer from Shimadzu, Japan. Determination of tungsten dispersion and the average size of crystal on silica nanoparticles or mean particle size of active phase (MSAP) was performed by H 2 pulse chemisorption with a Chembet 3000 (Quantachrome Instruments, Boynton Beach, FL, USA) with a high-sensitivity thermal conductivity detector. The procedure is detailed in other work [63]. Quantitative hydrogen pulses (10 µL) were injected until the surface was saturated, and W-H species were formed [64][65][66]. The surface area (S BET ) of catalysts were obtained by N 2 adsorption/desorption experiments in an Autosorb-1 from Quantachrome [67,68]. The S BET was estimated based on the Brunauer−Emmet−Teller (BET) method which is detailed elsewhere [69]. X-ray photoelectron spectroscopy tests were performed for selected catalysts on a Specs X-ray photoelectronic spectrometer (NAP-XPS) with a PHOIBOS 150 1D-DLD analyzer, using a monochromatic source of Al-Kα (1486.7 eV, 13 kV, 100 W). Details of the experimental setup are found in a previous work [44].

Equilibrium Adsorption Isotherms
A stock solution of 2000 mg·L −1 was prepared by dissolving the extracted asphaltenes in toluene; later, model solutions were made ranging from 100 mg·L −1 to 2000 mg·L −1 . Batch adsorption experiments were carried out at 25 • C, adding the nanoparticles to the model solutions at a fixed nanoparticles-to-solution ratio of 100 mg to 10 mL [70]. The mixtures were stirred for 12 h at 300 rpm to ensure equilibrium; then, the WSN with the adsorbed asphaltenes were separated from the asphaltenes in solution by centrifugation. The absorbance of the supernatant was measured using a Genesys 10S UV-Vis spectrophotometer, taking the toluene as a blank. A calibration curve was previously constructed for a wavelength of 295 nm, and, in this way, determining the adsorbed amount of asphaltenes per nanoparticle area surface is determined by the mass balance of the analysis. The amount adsorbed q (mg·m −2 ) was determined using Equation (1): where C 0 (mg·L −1 ) is the initial concentration of asphaltenes, C E (mg·L −1 ) is equilibrium concentrations in solution after adsorption, V (L) is volume of the solution, and A (m 2 ·g −1 ) is the dry surface area of the employed material. The adsorption isotherms for the asphaltenes onto the nanoparticles were modeled using the solid-liquid equilibrium (SLE) model [38] (see Equations (S1) to (S3) in the support information).

Thermogravimetric Analysis
To evaluate the catalytic behavior of the materials, the thermal stability of the asphaltenes (adsorbate) in the presence and absence of the adsorbent was studied. Thermogravimetric tests were performed on asphaltenes, and nanoparticles obtained after the batch adsorption tests (previously dried) on a TGA Q50 from TA Instruments, Inc, New Castle, DE. The samples were dried for 12 h, and 5 mg of each one was taken for thermogravimetric analysis kept low to circumvent the diffusion limitations [71]. The samples were analyzed on wet combustion atmosphere from 30 • C to 600 • C at three different rates (5 • C·min −1 , 10 • C·min −1 , and 20 • C·min −1 ) using a wet airflow (5%, volumetric fraction) rate constant 100 mL·min −1 throughout the experiment. Steam is generated in an evaporator (filled with deionized water) with a temperature controller, which fixes the vapor pressure based on the temperature set point. The wet airflow is obtained when the airflow stream is injected through the evaporator, reaching its saturation at the vapor pressure required [72]. For the adsorbents selected as the best catalysts, their capacity to produce in situ hydrogen and other calorific gases was analyzed via mass spectrometry using a mass spectrometer (Shimadzu GC-MS, Tokyo, Japan). Details of the setup are found in other works [28,73]. The components targeted for the analysis by single-ion monitoring in mass spectrometry were H 2 (2 g·mol −1 ), CO (28 g·mol −1 ), CO 2 (44 g·mol −1 ), and LHC (C 2 H 4 , C 2 H 6 , CH 4 ). Reproducibility was ensured, realizing the experiments by triplicate. In addition, the effective activation energies were estimated by the method of Ozawa-Flynn-Wall (OFW). This methodology is extensively found in the literature and reported in previous reports [28,73], with good reproducibility (see Equations (S4)-(S7)).

Nanoparticle Characterization
The results obtained for the hydrodynamic diameter of synthesized SiO 2 nanoparticles are in the nanometer range with a mean value of 12.5 nm ± 2 nm. This size is suitable for application in oilfields because its size is smaller than pore throats of heavy oil reservoirs [74]. Table 1 shows the properties of the nanoparticles, BET surface area, tungsten dispersion, and average size of the tungsten nanoparticles in the supported catalyst. Each type of nanoparticle exhibited a different chemical nature.
Pulsed H 2 chemisorption determines the exposed metallic atoms at the surface, using H 2 as a molecule probe, which does not interact significantly with the support material (SiO 2 ), but forms a strong chemical bond to the surface metallic atoms causing the H 2 spillover and consequently its consumption. The W surface area and dispersion were calculated, assuming dissociative chemisorption of hydrogen with a stoichiometry of W-H of 1 [65,75]. Results of the tungsten mean particle size under the measurement of reduced particles for the studied samples are also shown in Table 1 and confirm the formation of tungsten nanoparticles (tungsten oxides) over the support surface. It can be observed that by increasing the calcination temperature, the obtained nanoparticles are bigger in size due to the thermally promoted particle growth. Meanwhile, the particle dispersion has an inverse trend: as the calcination temperature is increased, particle dispersion decreases. The tendency obtained for the variation of impregnation dosage used is similar; a greater quantity implies a crystal with larger size, and, in turn, lower dispersion. Hence, the formation of the tungsten oxide nanoparticles over the SiO 2 surface is strongly influenced by synthesis conditions.
The obtained S BET values for the nanomaterials are also listed in Table 1. As expected, the sample prepared at 350 • C has the highest surface area of 122.3 m 2 ·g −1 . The values of the measured S BET followed the order 350 • C > 450 • C > 650 • C and 1% > 3% > 5%, according to calcination temperature and impregnation dosage, respectively. As the temperature of calcination is raised, the surface area decreases [76]. It is well known that the temperature favors the atomic movements and, therefore, an increase of the crystallite size. Thus, a bigger size of the crystalline domain implies an increase of the particle dimension and a reduction of its surface area. In porous materials, it is also reported that the increase of the temperature can induce a collapse of the porous structure and a densification of the material. The behavior described in the reduction of the surface area is in agreement with the reported in other works [28,33].
XPS was used to investigate the oxidation states and chemical environment of W species on the surface of the as-synthesized catalysts; the results are shown in Figure 1. In an XPS survey spectra (see Figure S1) of selected catalysts, binding energies of the O 1s and Si 2s levels for all the catalysts agreed with the values obtained for silica (O 1s = 532.9 eV and Si 1s = 155.3 eV) [77]. The catalysts present significant signals at 531.3 eV and 285 eV, possibly due to the presence of hydroxyl groups and carbonate species or C-C bonds in carbon impurities (adventitious carbon) [78,79].
The W 4f HR-XPS core-level spectrum ( Figure 1) reveals that the W 4f doublet (5/2 and 7/2) presents at least four different signals corresponding to different W species at the surface. From approximately 33.8 eV to 36.2 eV and from 36.2 eV to 40 eV represent the spin orbit splitting W 4f 7/2 and W 4f 5/2 , respectively. Then, each of these regions was decomposed into two different signals, corresponding to different oxidation states or chemical environments of the tungsten superficial atoms. In general, the signals are displaced to higher energy with the increasing calcination temperature, indicating that the tungsten is surrounded by a high electronegative environment or is more oxidized when the temperature increases. W 4f electrons from WO 3 have binding energies at around 35.9 eV and 38.0 eV for 7/2 and 5/2 orbits, respectively [80][81][82]. As is shown in Figure 1a, another signal displaced to less binding energy is found in all W4f HR-XPS. These signals vary between 34.8 eV to 35.5 eV and 36.6 eV to 37.3 eV for 7/2 and 5/2 orbits, respectively. Generally, the spectrum of supported WO 3 is broader than massif WO 3 due to the interaction between the tungsten oxides and the support [83]. On another side, the broadening of the spectrum toward the high and low binding energy flack suggests that the supported tungsten is not fully oxidized, and possibly the surface contains small amounts of W 5+ on the supported catalysts. The analysis of the high-resolution XPS for W 4f suggests that the surface of the tungsten nanoparticles contains a large quantity of defects which diminish with the increasing of the temperature. These defects cause a nonstoichiometry oxide by the apparition of oxygen vacancies V 2 O at the surface and in the bulk as well. A nonstoichiometry oxide can be written as shown in Equation (2), where the missing charges can be balanced by a reduction of the W cations. In the Supplementary Information is shown the whole process (in Kröger-Vink notation) of the vacancy formation and reduction of the cations inside of a tungsten oxide (Equation (S8)-(S10)). Thus, the changes in BE shown in W 4f high-resolution XPS ( Figure 1) are associated with the oxygen vacancy formation and the cation reduction from 6+ to 5+, in accord with the BE values. Different reports show that the tungsten oxides can produce oxygen vacancies, and that its quantity can be modulated though the heating temperature and the atmosphere as well [84][85][86][87][88]. These defects play an important role in the physicochemical properties, such as catalytic [89] and photocatalytic activity [85], quantum confinement [86], and ionic conductivity [85], among others.
The atomic ratio of surface elements is shown in Figure 1c; all the catalysts showed small changes of Si/O ratio whereas W/Si and W/O ratios were different between catalysts. Considering the ratio of W/O and (W)/Si, which indicated the surface dispersion and concentration of tungsten on a silica support, a direct relation is seen to calcination temperature, relatively greater than the impregnation dosage used on each one [77,89]

Asphaltene Adsorption Isotherms Effect of Synthesis Conditions
Figure 2a-f shows the obtained adsorption isotherms of asphaltenes by SiWNa and SiW samples at 25 • C with the SLE model. For every sample, the sorption isotherms revealed an increase in the adsorbed asphaltene content as the asphaltene concentration increased. It should be noted that SHS adsorbed more asphaltenes than SiO 2 nanoparticles for all asphaltene concentrations. The difference was more noticeable at higher concentrations (C E > 200 mg·L −1 ), which agrees with previous results [28,36,[46][47][48][49][50]. These results agree with literature reports for the adsorption of different asphaltene types onto different solid surfaces [36,37,42,43,73]. The adsorption isotherms follow a type Ib behavior according to the IUPAC classification [87], indicating that there is a strong affinity between the n-C 7 asphaltenes and the synthesized materials [31,90].  (2)).
On another side, considering the acidic character of tungsten oxide, this behavior suggests the basicity of the asphaltenes in addition the presence of heteroatoms such as nitrogen and oxygen in the asphaltene molecules, which favor the interaction with the nanoparticle surface [91,92]. It could be observed in Figure 2a-f that there is a trend on the adsorptive capacity highly influenced by synthesis conditions of WSN following the general orders SiWNa1 > SiWNa3 > SiWNa5 > SiO 2 and SiWNa350 > SiWNa 450 > SiWNa 650 > SiO 2 . The SLE model parameters are enlisted in Table 2. Additionally, the adsorption affinity, represented by the reciprocal of H value, increases for lower tungsten impregnation dosage, except for the virgin support. The same trend can be observed by the degree of asphaltene self-association on active sites, which is represented by K. This parameter decreases with the increase in impregnation dosage, and the lowest value is obtained for the virgin support. Conversely, the inverse trend can be observed for the q m parameter. From values of q m and the H parameter, it can be observed that the functionalized nanoparticles show more affinity and adsorption capacity than the support. In addition, lower values of the K parameter suggest that the inclusion of tungsten oxide allows the inhibition of the adsorbate self-association. Van der Waals forces, polar, and acid-base (Lewis and/or Brønsted) interactions seem to be the dominant forces that contribute to the adsorption of the asphaltenes onto tungsten oxide and/or salt surfaces [93]. Previous studies concluded that the adsorption of asphaltenes over silica is due to the chemisorption of asphaltenes on the silica silanol groups [94]. On the other hand, acid sites are responsible for adsorption of basic molecules that contain heteroatoms [91,92]. Due to the low acidity of silica, the presence of tungsten oxide generates the formation of Brønsted acid sites and the increase of Lewis acid sites [95], increasing interactions with the basic sites of the asphaltenes, or heteroatoms N, O, and S, which makes the interaction stronger and consequently increases the number of asphaltenes adsorbed. Figure 2g shows the obtained adsorption isotherms of asphaltenes by SiW at 25 • C with the SLE model. The asphaltene uptake is enhanced after functionalizing the SiO 2 with the trend SiW1 > SiW3 > SiW5 > Si and SiW 350 > SiW 450 > SiW 650 > SiO 2 . Figure S2 compares the obtained adsorption isotherms of asphaltenes by SiW1 350 and SiWNa1 350 at 25 • C with the SLE model. In this case, the amount adsorbed of asphaltenes is higher than SiWNa nanoparticles; in addition, SiW nanoparticles showed a similar adsorption affinity than SiWNa. These differences could be due to electronic effects and different interactions between the active phase and the support (as seen in XPS analysis) that influence the selectivity for different functional groups or heteroatoms in n-C 7 asphaltene molecules [43]. Figure 3 presents the trends of the H and K parameters of the SLE model. Figure 3a shows the trends for functionalized nanoparticles from sodium tungstate at different dosage calcination temperatures. Here, in the parameter H, as the calcination temperature increases, the value increases, which suggests that the higher the calcination temperature, the lower the affinity. On the other hand, parameter K shows that as the calcination temperature increases for the same dosage, the self-association of the asphaltenes increases.
During XPS analysis, a direct relationship between calcination temperature and W/O and W/Si ratios was demonstrated, which indicated that the surface dispersion and concentration of tungsten on a silica support depends on the calcination temperature used. In addition, the tungsten nanoparticles contain a large quantity of defects that are diminishing with the increasing calcination temperature. All these properties affect the adsorptive behavior of nanoparticles and their ability to inhibit the self-association degree of asphaltenes on their surfaces.  Figure 4b shows that, likewise for nanoparticles functionalized from ammonium tungstate, the affinity value increases as the calcination temperature increases, which suggests a lower affinity. Regarding the parameter K, there is a slight increase in the value, which suggests a greater self-association of the asphaltenes as the calcination temperature increases.
Asphaltene adsorption over the surface of the nanoparticles depends on factors such as asphaltene concentration on the medium, affinity interactions, and availability of active sites [38,61]. However, at medium and high uptake [38,43], due to their self-associated tendency asphaltenes, molecules will form aggregates around the high-energy sites until saturation [61,96]. On another side, the asphaltenes adsorption is subject to the adsorbent surface selectivity to the aromatic part of the asphaltenes or heteroatoms present in their structure, generating different adsorption orientations (perpendicular, parallel, or flatways) [61]. Following this order of ideas, with the functionalization of the nanoparticles, selectivity is induced, and the conditions used during material synthesis could influence the amount of active sites available for asphaltene adsorption; at the same time, the obtained results suggest that these conditions affect the characteristics of supported metal oxide, such as dispersion and active phase size, influence the adsorptive capacity of support, and enhance its affinity for asphaltenes [97]. It is important to remark that the adsorption capacity seems to be enhanced after functionalization at a lower dosage.

Mass Loss Analysis
The tests were carried out in a wet air atmosphere for a specific asphaltene loading of 0.20 mg·m −2 . Figure 4 shows the rate of mass loss for the virgin n-C 7 asphaltenes in the absence and presence of SiWNa. Figure 4a-c compares the presence of nanoparticles using the same impregnation dosage, and Figure 4e-f compare the presence of nanoparticles using the same calcination temperature during nanoparticles synthesis. As seen for virgin n-C 7 asphaltenes, the reaction has an onset approximately at 325 • C, with a maximum peak at 465 • C and an offset near to 574 • C. In this case, the presence of silica nanoparticles did not prove to be effective in reducing the decomposition temperature of asphaltenes. Figure 5 shows that, generally, the curve for asphaltenes rate of mass loss in the presence of WSN shifts to the left, confirming that decomposition occurs at a lower temperature. For the best scenarios evaluated with the WSN (SiWNa1 350, SiWNa450, and SiWNa650), the maximum peak of mass loss rate was reduced from 465 • C to 325 • C. In previous reports, the support (SiO 2 ) was evaluated without any functionalization, showing that it reduces the temperature of transformation of asphaltenes a bit under different atmospheres, including inert or thermal cracking [73], steam gasification, and air (or combustion [28]). In this case, the presence and absence of water into the stream was also evaluated, showing that the decomposition profile remains unchanged. However, a change in the produced gases was observed and more gasification products were found (H 2 , CO, and CH 4 ). These products will be analyzed in the following section. The conversion of n-C 7 asphaltenes to gaseous products can be associated with their partial oxidation, pyrolysis, and gasification [16]. Steam presence in the decomposition of asphaltenes as carbonaceous source implies different reactions, such as water-gas shift reaction and methanation [43]. The catalytic decomposition of asphaltenes in the presence of WSN can be divided into two reaction regions, LTR (low-temperature reaction, before 400 • C) and HTR (high-temperature reaction, between 400 • C and 600 • C). Initially, reactions on the breakdown of alkyl chains take place and the dissociation of weak bonds (S-C and N-C) [93,98], followed by the opening of polycyclic aromatic hydrocarbons and radical or addition reactions that can form heavier compounds (these compounds could be larger from the starting species turning into coke) [51,99]. With the increase in the temperature, there are still some reactions attributed to the gasification and oxidation of resultant compounds produced by free radicals that were not stabilized and due to addition reactions [42]. In the catalytic decomposition process, irrespective of the atmosphere (wet air), the first step includes the oxidation and low-temperature gasification of the asphaltenes. Hence, the volatile content is released, and the remaining condensed aromatic sheets subsequently react. Later, the oxygen present reacts with condensed aromatic sheets oxidizing first outer layers and, at relatively higher temperatures, the internal layers [93].
The inclusion of WSN reduces the temperature of asphaltene consumption. In all cases, two main peaks are observed, one on LTR and another on HTR. It is inferred that the first peak corresponds to the breaking of weakest bonds of asphaltenes and the second, to the decomposition of more stable structures, such as aromatic condensed rings. The presence of H 2 O can enhance the process of asphaltenes decomposition, due to the reaction of gaseous H 2 or O 2 previously adsorbed such as water over the nanoparticle matrix and active sites with adsorbed molecules that would be cracked, and free radicals formed which are highly reactive and can aid to decompose other molecules, i.e., raw asphaltenes. According to the results obtained, the surface properties and the catalytic activity of the nanoparticles affect how asphaltenes and water are adsorbed over the active sites, leading to a highly efficient process in energetic terms [43]. In the studied cases, the first peak temperature is considerably reduced by nanoparticles at approximately 100 • C and for the WSN that shows the best catalytic activity, SiWNa1 350. Changing the precursor does not show significant catalytic activity (Figure 4g), a factor that could be attributed to poor dispersion, acidity, affinity, or the big size of the active phase. It is considered that acid sites influence the asphaltenes conversion [100]. Isomerization reactions occur on the Brønsted acid sites of the WO 3 /SiO 2 catalysts, and studies have concluded that by poisoning the surface acid sites with small amounts of alkali elements, such as Na, K, Rb, and Cs, branched metathesis products were decreased, that is, isomerization was reduced, obtaining a greater degree of asphaltenes consumption [101][102][103].
Regarding the SLE parameters, it can also be noticed a direct relation between the catalytic effect of the synthesized materials with the affinity for asphaltene adsorption and the degree of self-association. With lower values of H and K parameters of the SLE model, the catalytic behavior of the surface is higher. The asphaltene auto-association degree would impact the catalytic activity of the nanoparticle blocking this one, due to that the formation of large aggregates could block some active sites [61]. The highest tungsten oxide dispersion and the lowest active phase showed a higher reduction of decomposition temperature.

Estimation of the Effective Activation Energies
Another way to validate the catalytic activity of the nanoparticles consists of the estimation of the effective activation energies through the isoconversional OFW methodology. For this case, the nanoparticle that demonstrated the highest efficiency in hydrogen production and reduction of asphaltene decomposition temperature was selected to analyze the effective activation energy behavior (that is, SiWNa1 350). For comparative purposes, the nanoparticle was selected with the same metal dosage, calcined at the same temperature, but with different precursors, that is, SiW1 350. Figure 5 shows the estimated effective activation energies as a function of the degree of conversion. For virgin n-C 7 asphaltenes, the value of Ea decreases as the conversion increases, which agrees well with previous reports under different reaction atmospheres [28,72,95]. Conversely, for the decomposition n-C 7 asphaltenes in the presence of SiWNa1 350 and SiW1 350 nanoparticles, the opposite trend is observed. It is worth remarking that, in all cases for conversion < 80%, the effective activation energy is shown to be lower compared with the virgin n-C 7 asphaltene; this again confirms the catalytic activity of the nanoparticles. For the catalytic asphaltene decomposition, the activation energy increases with the degree of conversion, suggesting that the catalytic decomposition consists of a complex mechanism that involves multiple steps [42,104]. SiWNa1 350 displayed the lowest activation energy throughout the process, suggesting that this sample is more catalytically active toward SiW1 350. The reduction of the effective activation energy is of primary importance to enhance the wet in-situ combustion, as a more efficient heat consumption can be ensured and could lead to higher oil recovery with upgraded characteristics.
As discussed above, the temperature of calcination affects the size of particles and the surface chemical environment. For lower temperatures, a small particle, high dispersion, and more quantity of defects are found. Though the XPS analysis, lower binding energies were found when the calcination temperature was set at 350°C, showing that the W cations were in the lowest oxidation state (on average). Therefore, for these catalysts (350°C), it is expected to find the largest quantity of surface defects, such as oxygen vacancies and reduced tungsten cations. These couples of punctual defects V 2 O · · · W x W constitute a redox or basic-acid site that promotes the adsorption and activation of the different reactants, such as water, oxygen, and hydrocarbons, favoring the reaction kinetics (see Figure S3).

Analysis of the Gaseous Products Evolved during Catalytic Decomposition Process
During the asphaltenes decomposition though wet air gasification, different gases such as CO, CH 4 , H 2 , light hydrocarbons (LHC), and CO 2 can be released. In this study, the releases gases during virgin asphaltene and catalytic steam gasification with SiWNa1 350 were analyzed using mass spectrometry. The SiWNa1 350 sample was selected based on catalytic results reported in previous sections. The experimental nonisothermal heating at 10 • C·min −1 was considered, and the results of the gaseous products are shown in Figure 6. As expected, the gaseous mixture was composed of CO 2 , CH 4 , CO, C 2 H 2 , C 2 H 4 , C 6 H 6 , and H 2 . The gases C 2 H 2 , C 2 H 4 , and C 6 H 6 will be referred to as light hydrocarbons (LHC). Some traces of other gases were evidenced in different temperature intervals, but they were neglected (selective distribution < 1.0% vol). Panel a shows the evolution of the different gases with temperature change for virgin asphaltene decomposition. The gas mixture was composed mainly for CO and CO 2 . The CO evolved increased with temperature because of the reverse Boudouard reaction. The CO 2 is produced by different mechanisms, including the transformation of ketones and other oxygenated groups in the asphaltene chemical structure. The produced LHC was evidenced during all the temperature range evaluated. However, at lower temperatures (<200 • C), the higher release of them was noted because of the breaking and decomposition of aliphatic chains and low-molecular-weight hydrocarbons. Finally, H 2 release was poor in the noncatalyzed sample because these heavy fractions follow other reaction mechanisms that prioritize the production of other gases, such as CO 2 and CO [29].
The results of the catalyzed sample are shown in panel b. The results show an increase in LHC, CH 4 , and H 2 during the wide range of temperature. At lower temperatures (<250 • C), the LHC production is slightly higher than the uncatalyzed system because of the loss of aliphatic chains and low-molecular-weight hydrocarbons. In the catalyzed scenario, the LHC mixture is composed mainly of C 2 H 4 (>60%vol), a highly calorific gas; while in the noncatalyzed scenario, this gas was produced to a lesser extent (<30%vol).
On the other hand, CO 2 production decreased between 200-300 • C, whereas CO and CH 4 increased in the same range. CO 2 reacts with carbon atoms to produce CO while CH 4 is released by the breaking of methyl, methylene, and other short aliphatic chains. Then, between 300 and 400 • C, a drop in CO and a slight increase in CO 2 and H 2 is observed. This behavior can be attributed to the occurrence of water-gas-shift reaction [105]. In addition, methane-reforming reactions appear to occur in the same range of temperature as a reduction of the CH 4 content in the gas mixture. It is worth mentioning that hydrogen released was higher than the noncatalyzed system in all the temperature range. It can be produced by the reaction of C in asphaltenes and H 2 O (g) molecules. The higher hydrogen content was 20%vol between 300 and 400 • C.
In the thermic process, reactions of total or partial combustion are common (formation of CO); in the presence of steam, the methanation (formation of CH 4 ) and steam reforming are common [106]. The increase of CH 4 production in the presence of nanoparticles with respect to virgin asphaltenes suggests an increment in methanation reactions promoted by the action of nanoparticles; at the same time, in addition to the combustion reaction by the presence of O 2 , part of the formed CH 4 and the reaction in water present in steam form CO. These reactions are aligned with the upgrading process in the reservoir as lighter compounds would lead to viscosity reduction and API gravity increase.

Conclusions
Asphaltene adsorption and subsequent catalytic decomposition on functionalized silica nanoparticles with tungsten oxides were studied using model solutions on batch adsorption experiments and TGA/MS system for evaluating thermocatalytic properties in a steam/air atmosphere. Dispersion and medium active phase size of supported oxides were shown to be essential factors in the catalytic and adsorptive behavior of nanoparticles. Sodium tungstate and tungsten oxide supported nanocatalysts were compared, and the results showed a similar adsorptive capacity but a marked difference in catalytic behavior.
XPS analysis demonstrated a direct relationship between calcination temperature and W/O and W/Si ratios, which indicated the surface dispersion and concentration of tungsten on a silica support, depending on the calcination temperature used. In addition, the tungsten nanoparticles contain a large quantity of defects that are diminishing with the increasing calcination temperature. All these properties affect the adsorptive behavior of nanoparticles and their ability to inhibit the self-association degree of asphaltenes on their surfaces.
The catalytic results show that synthesized nanoparticles reduce asphaltene decomposition temperature. In addition, synthesis parameters, such as temperature and impregnation dosage, play an important role in the catalytic activity of the materials due to the different WOX-support interactions. The mixture released during the catalyzed asphaltene decomposition in the wet air atmosphere reveals an increase in light hydrocarbons, methane, and hydrogen content. Hydrogen production was prioritized between 300 and 400 • C, where the reduction of CO, CH 4 and the increase in CO 2 content similarly occurs, associated with water-gas-shift and methane-reforming reactions, respectively.
The results obtained in this research project suggest that the SiO 2 nanoparticles functionalized with tungsten oxides (WSN) can be used as catalysts for the in situ upgrading of heavy crude oil during wet in situ combustion Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/pr10020349/s1, Solid-liquid equilibrium model; Section S2. OFW model; Figure S1. XPS survey spectrum and high-resolution spectra for SiWNa5 650 sample.; Section S3. Vacancy formation and compensation charge into a tungsten oxide using Kröger-Vink notation.; Figure S2. Adsorption isotherms of asphaltenes onto supported nanoparticles using the same impregnation dosage and calcination temperature but different metal oxide precursors. Nanoparticle dose, 10 g·L −1 ; agitation speed, 200 rpm; T, 25 • C. The symbols are experimental data, and the solid lines are from the SLE model.; Figure S3. Chemical species around an oxygen vacancy. Model for adsorption of reactants. Funding: This research was funded by Fondo Nacional de Financiamiento para la Ciencia, la Tecnología y la Innovación "FRANCISCO JOSÉ DE CALDAS", Agencia Nacional de Hidrocarburos (ANH), COLCIENCIAS and Universidad Nacional de Colombia for their support provided in Agreement 272 of 2017.