Catalytic Conversion of n-C7 Asphaltenes and Resins II into Hydrogen Using CeO2-Based Nanocatalysts

This study focuses on evaluating the volumetric hydrogen content in the gaseous mixture released from the steam catalytic gasification of n-C7 asphaltenes and resins II at low temperatures (<230 °C). For this purpose, four nanocatalysts were selected: CeO2, CeO2 functionalized with Ni-Pd, Fe-Pd, and Co-Pd. The catalytic capacity was measured by non-isothermal (from 100 to 600 °C) and isothermal (220 °C) thermogravimetric analyses. The samples show the main decomposition peak between 200 and 230 °C for bi-elemental nanocatalysts and 300 °C for the CeO2 support, leading to reductions up to 50% in comparison with the samples in the absence of nanoparticles. At 220 °C, the conversion of both fractions increases in the order CeO2 < Fe-Pd < Co-Pd < Ni-Pd. Hydrogen release was quantified for the isothermal tests. The hydrogen production agrees with each material’s catalytic activity for decomposing both fractions at the evaluated conditions. CeNi1Pd1 showed the highest performance among the other three samples and led to the highest hydrogen production in the effluent gas with values of ~44 vol%. When the samples were heated at higher temperatures (i.e., 230 °C), H2 production increased up to 55 vol% during catalyzed n-C7 asphaltene and resin conversion, indicating an increase of up to 70% in comparison with the non-catalyzed systems at the same temperature conditions.


Introduction
Nowadays, global warming and energy supply challenges have attracted increasing attention worldwide [1]. Hydrogen is a renewable and clean energy, and its application only generates water as by-product. However, fossil fuels are currently the most used sources for energy consumption [2]. Generally, there is a desire to generate an energy transition from fossil fuels to a cleaner energy source, which include the use of hydrogen [3]. During the energy transition, co-production of fossil fuels and hydrogen could be an exciting strategy to supply several issues.
As hydrocarbons are hydrogen-rich compounds, they can be used to produce hydrogen by applying different methods such as gasification with air and oxygen, and steam reforming [4]. The first two methods have been extensively studied in coal gasification processes and natural gas reforming [5,6]; however, hydrogen production from crude CO + H 2 O CO 2 + H 2 (1) Due to the potential application of this type of nanocatalysts, it is expected that they can be used in the reservoir for co-production of upgraded crude oils and hydrogen through steam catalytic gasification [2]. To date, there are no works reported in the literature evaluating this technology. Since crude oil is a complex mixture of different fractions, it is important to evaluate the hydrogen production from the steam catalytic gasification of n-C 7 asphaltenes, resins II, and combinations of them. Therefore, this work aims to evaluate for the first time the hydrogen production efficiency of different CeO 2 based-nanocatalysts containing different ratios between three NOT (i.e., Ni, Co, and Fe) and a NON element oxide (Pd) through the steam catalytic gasification of the crude oil heaviest fractions (i.e., resins II and n-C 7 asphaltenes). The adsorption of both fractions and a mixture of them in the nanocatalysts is evaluated to meet this objective. Then, the hydrogen releasing was calculated during isothermal gasification process. Some variables such as temperature, resins:asphaltenes (R:A) ratio, and nanocatalyst nature were analyzed. This study is expected to broaden the use of fossil fuels for producing a hydrogen-rich gas Nanomaterials 2021, 11, 1301 3 of 21 with low content of CO 2 (<5% vol) from the gasification of the most complex fractions of crude oil.

Materials
CeNi1Pd1, CeFe1Pd1, and CeCo1Pd1 were previously synthesized through the incipient wetness technique [41] using commercial cerium oxide nanocatalysts (Nanostructured & Amorphous Materials, Houston, TX, USA) with an average particle size of 21.6 nm. Salt precursors (FeCl 3 ·6H 2 O, NiCl 2 ·6H 2 O, CoCl 2 ·6H 2 O, and Pd(NO 3 ) 2 ·2H 2 O) were supplied by Merck KGaA, Darmstadt, Germany. Each system was named considering the initials of the support, the elements with which the doping was made and the percentage of impregnation. For example, CeNi1Pd1 is a system supported on CeO 2 , with Ni and Pd at 1.0% (w/w). Details of nanoparticle characterization are found in our previous work [38]. Table 1 shows some properties of the employed nanocatalysts. The molar fraction for the couples Ni:Pd, Fe:Pd, and Co:Pd, are 0.4:0.7, 0.9:0.7, and 1.4:0.9, respectively. Considering the metal dispersion results, the nominal molar ratio affects the structure of nanocrystals of transition element (NOT)-Pd, causing a lower possibility of sintering processes for higher Pd/NOE molar value, (1.75 Pd/Ni, 0.78 Pd/Fe, and 0.64 Pd/Co).
X-ray photoelectron spectroscopy analysis was done to characterize the nanocatalysts' surface chemistry following the protocol described in previous work [42]. The atomic content and relationships from O1s, Ce3d, and Pd3d profiles are shown in Table 2. The O1s pattern was deconvoluted in two main peaks at 529.0 eV and 531.5 eV ascribed to lattice oxygen (O latt ) and ascribed to the surface (O ads ) [43]. Between the functionalized nanocatalysts, the amount of O ads follows the increasing order CeCo1Pd1 < CeFe1Pd1 < CeNi1Pd1, therefore, O ads /O latt follows the same order. The peaks for Ce3d 5/2 and 3d 3/2 that appear at 881.6 eV, 885.3 eV, 899.8 eV, and 903.0 eV, which are associated to Ce 3+ species. In the case of Ce 4+ it is found at 882.7 eV, 889.7 eV, 898.6 eV, 901.3 eV, 907.6 eV, and 919.3 eV [44]. The increasing sequence for Ce 3+ is CeCo1Pd1 < CeFe1Pd1 < CeNi1Pd1, which agrees with the sequences of O ads and O ads /O latt . The formation of Ce 3+ is intrinsically related to the oxygen vacancies formation, promoting the activation of surface oxygen.
From Pd3d pattern, four main peaks are found. The binding energy at 337.2 eV (Pd 3d 5/2 ) and 344.5 eV (Pd 3d 3/2 ) refers to Pd 2+ in bulk PdO [45]. The content of Pd 2+ increases in the same order than Ce 3+ , O ads , and O ads /O lat , demonstrating the correlation between the three species. In the case of Pd 2+ , they can function as a reservoir for oxygen adsorption. The other peak at 335.2 eV and its respective satellite refer to the metallic palladium, which follows the opposite behavior than Pd 2+ .
A Colombian extra-heavy crude oil (EHO) of 6.4 • API, viscosity of 3.1 × 10 6 cP at 25 • C was used for isolation of n-C 7 asphaltenes (A) and resins II (R). The content of asphaltenes and total resins were 12.64 and 51.98 wt%, respectively. n-Heptane (99%, Sigma-Aldrich, St. Louis, MO, USA) was used to isolate the asphaltenes from the EHO, and chromatographic silica (Sigma-Aldrich, St. Louis, MO, USA) was added to the de-asphalted crude oil to separate the resins from the solution. Both fractions were isolated following the protocols described elsewhere [46,47]. Finally, toluene (99%, Sigma-Aldrich, St. Louis, MO, USA) was used as a solvent for the adsorption experiments.

Asphaltenes and Resins Characterization
The characterization of individual n-C 7 asphaltenes and resins II was carried out through elemental analysis (C, H, N, S, and O) following the ASTM D5291 standard method [48] using an elemental analyzer (Perkin-Elmer, Waltham, MA, USA). Average molecular weight (M W ) was obtained by vapor pressure osmometry (VPO) using a Knauer osmometer (Knauer, Berlin/Heidelberg, Germany). 1 H NMR and 13 C NMR were carried out in a Bruker AMX 300 spectrometer (Karlsruhe, Germany). X-ray photoelectron spectrometry analysis was obtained in a PHOIBOS 150 1D-DLD analyzer (SPECS, Berlin, Germany). The procedures are described elsewhere [49]. Based on these results, average representative molecules for both fractions were constructed and optimized by density functional theory applying Becke's three-parameter and Lee-Yang-Parr functions. Material Studio (BIOVIA, San Diego, CA, USA) was used for this purpose.

Adsorption Experiments
Adsorption isotherms were constructed at 25 • C by making stock solutions of 2.2 mmol·L −1 of resins II and/or n-C 7 asphaltenes in toluene and 100 mg of nanocatalysts per 10 mL of volume solution. Initial concentrations of resins II were varied between 0.1 mmol·L −1 and 2.2 mmol·L −1 , and asphaltenes between 0.1 mmol·L −1 and 1.5 mmol·L −1 . The instrument and protocol employed for the adsorption isotherms construction were like those described in the previous studies [23,43,50,51].
In the first stage, the individual components' adsorption was performed in the three bielemental systems (i.e., CeNi1Pd1, CeCo1Pd1, and CeFe1Pd1) and the support CeO 2 . The system with higher adsorption was then selected to evaluate the competitive adsorption in different R:A mass ratios of 8:2, 1:1, and 2:8. The amount adsorbed of each component was obtained by combining softening point (SP), and thermogravimetric (TGA) experiments following the ASTM E28-12 standard methods [50]. Figure S1 shows the calibration curve of SP against n-C 7 asphaltenes percentage where a R 2 value close to the unity was obtained.
Finally, this work uses the Solid-Liquid Equilibrium (SLE) model to describe the adsorption isotherms based on the theory of adsorption and association of molecules on microporous surfaces [51]. The model is described in Section S.1 of the Supplementary Materials.

Steam Catalytic Gasification of Resins II and n-C 7 Asphaltenes
The steam catalytic gasification of virgin and adsorbed compounds on nanocatalysts was performed using a Q50 thermogravimetric analyzer (TA Instruments, Inc., New Castel, DE, USA). Two different processes were executed. The first one, under non-isothermal heating between 100-600 • C, at a heating rate of 20 • C·min −1 . In the second one, the samples were subjected to isothermal heating at three different temperatures. Samples in the absence of nanocatalysts were heated at 370 • C, while the asphaltenes/resinscontaining nanocatalysts at 220 • C. The steam atmosphere was simulated by introducing 100 mL·min −1 of N 2 and 6.30 mL·min −1 of H 2 O (g) using a gas saturator controlled by a thermostatic bath at atmospheric pressure [37].
Nanocatalysts with the highest catalytic activity were selected to evaluate the isothermal conversion of different R:A ratios. Catalytic experiments were developed for an adsorbed asphaltene and/or resin II amount of 0.0002 mmol·m −2 ± 0.00002 mmol·m −2 [52]. The experimental results were used to calculate the activation energy according to the isothermal procedure [53] described in Section S.2 of the supplementary material.
Gas monitoring during the experiments was done in a mass spectrometer (Shimadzu GC-MS, Tokyo, Japan) coupled to the TGA. The equipment was adjusted at an ion trap linear scan rate of 0.03 m/z between 0 and 200 m/z and an electron impact mode of 100 eV to obtain detailed information of the sample's gasification. The gases were analyzed following the protocols described elsewhere [24,49,54,55]. The evolved gases during steam catalytic gasification were evaluated at three different temperatures (210 • C, 220 • C, and 230 • C). Each run was repeated, at least twice, to be ensured of the reproducibility of the experiments.

Resins II and n-C 7 Asphaltenes Characterization
The estimated elemental composition of n-C 7 asphaltenes and resins II is shown in Table 3. The resins I data was also reported in Table 3 taken from our previous work for comparative purposes only [56]. The content of heteroatoms differs for each fraction. Nitrogen was lower than 0.5% for n-C 7 asphaltenes and resins II, while sulfur content was similar for resin II and n-C 7 asphaltene. As expected, the most significant difference between the two resins is the oxygen content, which was higher for resins II than that for resins I. Due to the strong aromaticity of n-C 7 asphaltenes and their degree of unsaturation, resins II and I have a higher H/C ratio. The average molecular weight follows the increasing order resin I < n-C 7 asphaltenes < resin II. This result shows that resin II isolated from the EHO employed in this study is heavier than asphaltenes.  Tables 4 and 5 show the XPS and NMR results for both fractions, respectively. The survey spectra for both fractions indicate X-ray induced Auger transitions of oxygen, nitrogen, carbon, and sulfur. Analyzing the high-resolution spectra, the following results were obtained: Oxygen was present as single bonds (C-O-C, C-O) (532.4 eV) or carboxyl (533.5 eV) in both fractions. Carboxyl groups were in a higher percentage in resins II than n-C 7 asphaltenes. The C1s was divided into single bonds (C-C, C-H) (284.6 eV), C=O (287.9 eV), and COO-(289.4 eV) species. As expected, C-C and C-H predominate for the both samples, while C=O was higher for n-C 7 asphaltenes and COO-for resins II. Although nitrogen was present in low concentration for n-C 7 asphaltenes (elemental analysis), it was slightly detected by XPS. The spectra were deconvoluted in three peaks (i.e., pyridinic (398.8 eV), pyrrolic (399.9), and amines (400.7 eV)) for both fractions. The pyridinic content was the highest for both and the lowest for amines. Finally, sulfur highresolution spectra show the presence of thioethers (163.4 eV), thiophenes (164.4 eV), and sulfones (167.7 eV). For asphaltenes, the sulfur form percentage increases in the order sulfones (1.8%) < thioethers < (23.4%) < thiophenes (74.8%). In resins the trend changes as follows: thiophenes (25.2%) < sulfones (25.3%) < thioethers (49.5%). Table 4. Atomic concentrations (%) of (a) oxygen forms, (b) nitrogen forms, (c) carbon forms, and (d) sulfur forms obtained by the fitting results of the O1s, N1s, C1s, and S2p spectra, respectively, for n-C 7 asphaltenes and resins II.

C1s
O1s S2p N1s  Results from NMR shows that hydrogen was found in four different forms: Aromatic hydrogen (H a ), and aliphatic hydrogens α, β, and γ linked to aromatic rings, (H α , H β , and H γ , respectively). H a content was higher for n-C 7 asphaltenes than resins II. For both fractions, the H β species were in higher proportion than H α and H γ . This result indicates the predominant presence of paraffinic/naphthenic β-hydrogens in aliphatic and aromatic structures, respectively. Resins II present a higher content of long aliphatic chains in γposition (H γ ) than n-C 7 asphaltenes [57]. From the 13 C NMR spectra, carbon was found in aromatic (C ar ) and aliphatic (C al ) form. As expected, resins II present a predominant C al content (60.96%) and asphaltenes a higher proportion of C ar (64.48%).
Based on the general properties obtained from elemental analysis, molecular mass, nuclear magnetic resonance, and X-ray photoelectron spectroscopy, average structure molecules were proposed. Figure 1 shows the n-C 7 asphaltene and resin II molecules. The samples show the continental type model widely accepted for both fractions. n-C 7 asphaltene molecular formula proposed is C 62 H 70 S 2 O 2 , the theoretical molecular weight of 911.36 g·mol −1 and deviation of 0.4%. Resin II molecular formula proposed is C 63 H 85 NS 2 O 2 , molecular weight estimated in 952.5 g·mol −1 and deviation of 0.5%. Figure 2 shows the adsorption isotherms of n-C 7 asphaltenes (panel a) and resins II (panel b) onto CeNi1Pd1, CeCo1Pd1, CeFe1Pd1, and CeO 2 nanocatalysts. A high affinity observed between the different nanocatalysts and the n-C 7 asphaltenes is related to the type Ib isotherms following the International Union of Pure and Applied Chemistry (IUPAC) [58].

Adsorption Isotherms
Adsorption capacity at any concentration decreases in the order CeNi1Pd1 > CeCo1Pd1 > CeFe1Pd1 > CeO 2 for n-C 7 asphaltenes; whereas, for resins II uptake, the order increases as: CeNi1Pd1 > CeFe1Pd1 > CeCo1Pd1 > CeO 2 . This result indicates an increase in the adsorptive capacity by NOT impregnation following the order Fe < Co < Ni for asphaltene adsorption, since the effective nuclear charge increases in the same order changing Lewis acidity [59]; hence, facilitating the formation of coordinated bonds between heteroatoms (Lewis bases) and metals of heavy hydrocarbons, and the active site of NOTs in the different nanocatalysts [60]. On the other hand, the resins II chemical structure changes regarding asphaltenes increase their selectivity for Fe crystals than Co.  According to the slope in Henry's region for the adsorption isotherms curves of resins II on CeNi1Pd1 and CeFe1Pd1 nanocatalysts, a high affinity is observed. The isotherms obtained follow the IUPAC standard corresponding to type Ib isotherm [58], indicating that resins II have greater selectivity for the Ni and Fe functional groups [61]. Authors explain that aromatic nitrogen and sulfur species' content has a high affinity for iron and nickel nanocatalysts [62]. Resins II adsorption on non-functionalized support and bielemental CeCo1Pd1 nanocatalyst shows an isotherm type II. At low concentrations, adsorption at the active sites accessible for resins II is high, indicating a high slope in Henry's region; however, as the concentration of resins II increases, the formation of aggregates increases, generating adsorption multilayer and taking a profile of type II isotherms [63].
Comparing the adsorptive capacity of different samples, for a fixed concentration of resins, the adsorption amount increases in the following order CeO 2 < CeCo1Pd1 < CeFe1Pd1 <CeNi1Pd1. Unlike the adsorption of n-C 7 asphaltenes, the adsorption of resins II is greater for the Fe-Pd couple than that for the Co-Pd couple. Based on the results, the competitive adsorption in R:A systems was evaluated in CeNi1Pd1 nanocatalysts and Figure S2 shows the obtained results. Figure S2a shows that n-C 7 asphaltene adsorption capacity on CeNi1Pd1 nanocatalysts decreases with resin II content. For 2:8 and 1:1 ratios, the isotherms obtained are classified as type I, while the isotherm for 2:8 system behaves type III according to the IUPAC [58]. The change in the asphaltene adsorption mechanism can be associated with a change in their colloidal structure or competitive adsorption between the two polar fractions. As the amount of resins II increases in the system, there is a reduction of the active surface sites available for n-C 7 asphaltene adsorption. For example, at fixed C E = 0.33 mmol·L −1 , the asphaltene adsorption amount was 0.002 mmol·m −2 , 0.0018 mmol·m −2 , 0.0011 mmol·m −2 , and 0.0008 mmol·m −2 for n-C 7 asphaltene, 2:8; 1:1, and 8:2 systems, respectively.
Structurally, given the similarity between both fractions, it is likely that strong and stable interactions will occur between them. As shown in Table 1, asphaltenes and resins contain similar percentages of N-, S-. and O-heteroatoms. Because asphaltenes could interact with traces of resins containing oxygen and/or nitrogen, the primary interaction is likely of the attractive type. In this sense, resins tend to interact more with asphaltenes, as shown in Figure 3 below. When the amount of resins in the system is high enough, a solvation phenomenon could be generated, which reduces the asphaltene-nanoparticle interactions by blocking important active sites, such as aromatic, naphthenic rings, and heteroatoms. Figure S2b shows the adsorption isotherms for resins II at different R:A ratios. The results show that type I isotherms for 8:2 and 1:1 ratio. However, in 2:8 system, the isotherm obtained is changed to type III. Results suggest that resins II adsorption is affected by n-C 7 asphaltene content. The multilayer behavior of 2:8 ratio show that the resin-asphaltene interactions prevail over resin-resin interactions [64]. The adsorption amount of resins at C E = 0.33 mmol·L −1 were 0.0013 mmol·m −2 , 0.0011 mmol·m −2 , 0.0007 mmol·m −2 , and 0.0005 mmol·m −2 for individual resins, 2:8; 1:1, and 8:2 systems, respectively. Table S1 summarizes H, K, y Q m values of the SLE model for each system evaluated. The experimental data adjustment through the SLE model shows the highest affinity of both resin II and n-C 7 asphaltene fractions for CeNi1Pd1 nanocatalysts according to the low H values. Resins II presented lower K values than n-C 7 asphaltenes, which could result from the lower amount of C ar on their structure. It is reported that the aggregation mechanism of both fractions occurs through π-π stacking interactions between the aromatic cores. Hence, the lower C ar the lower the self-association degree on the nanocatalysts surface. On the contrary, the H values were higher for n-C 7 asphaltenes than those for resins II due to the higher content of carboxyl, thiophene, and pyrrolic functional groups. It is worth mentioning that resins presented a higher value for Fe-Pd couple due to the high content of pyridines, which present a high affinity for this NOT.
Concerning the self-association degree of n-C 7 asphaltenes and resins, each fraction's parameter decreases as the content of the other fraction increases. This result suggests that asphaltene-asphaltene and resins-resins interactions are affected by stronger resin-asphaltene interactions [64,65]. In this case, multilayer adsorption of n-C 7 asphaltenes on CeNi1Pd1 nanocatalysts is favored for higher R:A ratios.
To understand the adsorption mechanism of n-C 7 asphaltenes in resins II presence on nanoparticles, the adsorbed and desorbed amount of asphaltenes was predicted. If the colloidal state of n-C 7 asphaltenes is affected by resin II, the predicted and the experimental adsorbed amount will differ. The predicted adsorbed amount in R:A systems is calculated considering as a fraction of the individual adsorbed amount of each component. For example, in 8:2 system the predicted adsorbed amount of n-C 7 asphaltenes are assumed as 20.0% of the individual mass fraction adsorbed. A linear graph q predicted against q experimental with a slope m = 1, intersection b = 0, and R 2 = 1.0 is associated with a good prediction. Table S2 shows the results for the predicted adsorbed amount of n-C 7 asphaltenes on CeNi1Pd1 nanoparticles in R:A systems. The results show a good fit between the predicted and experimental values. The values of m were close to 1 in all cases, and b were close to the origin. These results give a clear idea about the influence of resins II on the phenomenon evaluated, indicating that the adsorption of n-C 7 asphaltenes are mainly controlled by their concentration in the systems and by the affinity of the nanoparticle for resins II. In this way, the colloidal state of the asphaltenes is not affected by the presence of resins II. These results are in a good agreement with the reported results in previous works [18,66]. Figure 4 shows the rate for mass loss for the steam catalytic gasification of n-C 7 asphaltenes and resins II, respectively. Results reveal that pure resins II and n-C 7 asphaltenes decompose around 420 • C and 450 • C, respectively (rate for mass loss peak). However, with nanocatalysts, the decomposition temperature reduces, and decomposition begins at approximately 200 • C in all systems.

Thermogravimetric Experiments.
Bielemental and support systems reduce the n-C 7 asphaltenes decomposition temperature from 450 • C, to 220 • C, 230 • C, 250 • C, and 370 • C for CeNi1Pd1, CeCo1Pd1, CeCo1Pd1, and CeO 2, respectively. On the other hand, the three bielemental and CeO 2 nanocatalysts decompose resins II at 220 • C and 300 • C, respectively. The conversion of both fractions continues at high temperatures around 350 • C, as a result of the distribution of high, medium, and low molecular weight hydrocarbons [29].
The decomposition temperature peaks of resins II suggest a high content of lower molecular weight hydrocarbons (aliphatic structures) than those present in the asphaltenes' molecular structure. This agrees with 13 C-NMR results. In both cases, nanocatalysts' catalytic activity was improved with the addition of NOT in the order of Fe < Co < Ni. The highest performance for the couple Ni-Pd is due to the support's ability to interact with Ni and Pd nanocrystals to promote cracking and isomerization reactions [45,[67][68][69]. The strong metal-support interaction between Ce 3+ ions and Ni nanocrystals may promote water gas shift reactions at low temperatures [67]. During this process, the support conducts a redox cycle, which generates oxygen anion vacancies (V 2• o ). These are active sites for the adsorption of steam molecules and, therefore, greater interaction with the adsorbed heavy oil fractions [45,68]. Figure 5 shows the isothermal conversion at 220 • C for n-C 7 asphaltenes and resins II adsorbed on the bielemental systems and the support. It is appreciated that the conversion degree increases in the order of CeO 2 < CeFe1Pd1 < CeCo1Pd1 < CeNi1Pd1 in both fractions.
Between the support and the best bielemental nanoparticle (Ni-Pd), to convert 20% of resins II and n-C 7 asphaltenes, the time decreases approximately 70 min and 155 min, respectively. Among many factors, this is due to the role played by the Ni-Pd couple. NiO nanocrystals can inhibit PdO nanocrystals growth [38]; therefore, they acquire a higher dispersion than Co-Pd and Fe-Pd bielemental systems, generating many HA-PdO interactions. This implies an increase in hydrogen and light gas production capable of stabilizing the cracked heavy oil fractions' free radicals [7].   Figure 6 shows the isothermal conversions at 220 • C for R:A ratios of 8:2, 1:1, and 2:8 on CeNi1Pd1 nanocatalysts. The conversion degree increases in the order A < 2:8 < 1:1 < 8:2 < R at any time. It was reported that resins are composed for a higher degree of aliphatic carbon structures. This composition influences the number of interactions generated in R:A systems between asphaltenes and resins, and therefore on catalysis [56]. Decomposing structures by R-R bonds is less complicated than breaking R-A structures [56]. Additionally, a higher asphaltene content, the presence of thiophenes and pyrroles prevails; therefore, more time is required for their cracking [49]. In addition, the first cracked products can polymerize and form coke. The effective activation energy was used to estimate the energy requirements needed to carry out the chemical reaction. Figure 7a shows the results of effective activation energy (E a ) for n-C 7 asphaltenes and resins II with and without bielemental nanocatalysts and support. E a decreases considerably by the addition of nanocatalysts, in the order of CeO 2 < CeFe1Pd1 < CeCo1Pd1 < CeNi1Pd1. NOTs impregnation modifies the decomposition mechanism of the heavy oil fractions, indicating that the catalyst's chemical nature influences gasification reactions. On the other hand, the energy required for thermo-gasification and thermo-catalytic gasification is lower for resins II than that for asphaltenes. In addition, the estimated activation energies values for the R:A systems above CeNi1Pd1 nanocatalysts are shown in panel b. It is observed that E a increases in the order of 8:2 < 1:

Effect of Nanocatalyst Chemical Nature
The effect of nanocatalyst chemical nature on evolved gases during the steam gasification of n-C 7 asphaltenes and resins II was analyzed under isothermal heating at 220 • C and the results are shown in Figure 8 as a function of time. It is observed that the release of CO 2 , CH 4 , CO, a mixture of light hydrocarbons (LHC) and H 2 . Some traces of other gases were evidenced in different time intervals, but their contribution was less than 0.1%; thus, they were neglected. It is worth mentioning that non-catalyzed systems were evaluated at 370 • C due to their complex structure. The gasification results for the non-catalyzed systems show considerable CO and CO 2 production during the conversion of the samples. The higher production of CO is associated with the reverse Boudouard reaction. As the reaction is endothermic, the production of CO becomes meaningful over time. On the other hand, the LHC mixture was composed mainly of C 2 H 2 , C 2 H 4 , and C 6 H 6 . Its production was markedly at the beginning of the warm-up. This is because the aliphatic chains break down at low temperatures and promote the production of these gases. However, over time, both asphaltenes and resins produce a higher proportion of CO 2 and CO and do not promote intermediate reactions that give rise to LHC.
On the other hand, methane and hydrogen were produced in low quantities all the time evaluated. It should be noted that these fractions decompose less than 20% of their mass at 370 • C during the time analyzed. It is noted that asphaltenes produce a higher amount of CO 2 and a lower amount of H 2 than resins II in all the time range evaluated. This result agrees with the higher content of aromatic carbon in asphaltenes. During thermal heating, the aromatic core is susceptible to the formation of ketones and finally releasing as CO 2 . The higher the content of hydrogens α and γ linked to aromatic rings, the higher the hydrogen release. Figures 9 and 10 show the selectivity distribution of light gases for n-C 7 asphaltenes and resins adsorbed on different nanocatalysts, respectively. The results reveal an increase in LHC, CH 4 , and H 2 in all the times at 220 • C for both fractions. The asphaltenes begin their decomposition with a CO 2 production between 20% and 30% (vol) for all nanocatalysts. Unlike the virgin fractions, there is a higher production of H 2 and LHC from the first 10 min of heating. Subsequently, CO 2 production falls to a greater degree for asphaltenes adsorbed on CeNi1Pd1 than in the rest of the systems. In turn, the production of CO, CH 4 and H 2 increases. When reacting with CO 2 , carbon produces CO, while CH 2 is produced by breaking short aliphatic chains (methyl, methylene, etc.). Hydrogen in the first instance can be produced by the reaction of C in asphaltenes and H 2 O (g) molecules.
During the thermal gasification there is a drop in the CH 4 production in all cases. For CeO 2 it occurs at 50 min, for CeFe1Pd1 at 40 min, and for CeCo1Pd1 and CeNi1Pd1 at 30 min. In turn, the production of H 2 increases. This may suggest the development of steam reforming reactions assisted by different nanocatalysts [69]. The difference in the time in which occurs, is a response to each system's catalytic activity. Therefore, the most catalytic system (CeNi1Pd1) generates this change in less time and with more outstanding vol% production of H 2 . Likewise, the drop in CO and a slight increase in CO 2 and H 2 indicate the WGS reaction development [69]. These changes in the selective distribution of gaseous products were first observed in the system (CeNi1Pd1) between 40 and 50 min. Moreover, the increase in CO 2 was minimal. The system that took the longest was CeO 2 (90-100 min).
Regarding several studies on the catalytic gasification of carbon, all the materials in this study produced a lower content of CO 2 in the entire time interval analyzed [5,[70][71][72]. products were first observed in the system (CeNi1Pd1) between 40 and 50 min. Moreover, the increase in CO2 was minimal. The system that took the longest was CeO2 (90-100 min).
Regarding several studies on the catalytic gasification of carbon, all the materials in this study produced a lower content of CO2 in the entire time interval analyzed [5,[70][71][72].
Finally, the mixture of light hydrocarbons was mainly composed of C2H2, obtaining a total gas mixture with an average calorific value of 82.45 kcal·kg −1 for CeO2, 102.51 kcal·kg −1 for CeFe1Pd, 124.23 kcal·kg −1 for CeCo1Pd1, and 139.14 kcal·kg −1 for CeNi1Pd1 on time evaluated. It is observed that the dispersion and average crystallite size of Pd in functionalized systems follows the same order as the increase in H2 release and the average calorific values of the release gas mixture. As higher the dispersion and lower the average crystal size, the higher the calorific values and the content of H2 produced.   of multiple metals as active phases creates defects in the structure and oxygen vacancies in which pure, highly dispersed metal catalysts are located [62,67]. This allows the adsorption of reactants and facilitates the WGS and methane reforming reactions themselves. The hydrogen production trend is also agreed with each material's catalytic activity for decomposing both fractions at the evaluated conditions. CeNi1Pd1 showed a higher performance than the other three samples and led to a higher hydrogen vol% in the effluent gas. Cerium nanoparticles as active support facilitate the production of low molecular weight compounds due to (a) the positive effect of the NOT and NON elements and ceria interactions, and (b) the activity of redox couple Ce 3+ /Ce 4+ . There is evidence that oxygen atoms migrate from the support (CeO2) to transition element particles at higher temperatures during spillover phenomena. This oxygen can be used to oxidize CO to CO2, and subsequently, the partially reduced ceria can be re-oxidized by water, completing the catalytic cycle. Moreover, the gases can be used to produce H2 during the WGS reaction. In this sense, the surface of the support will be composed of Ce 3+ and Ce 4+ ions. Ce 3+ ions Finally, the mixture of light hydrocarbons was mainly composed of C 2 H 2 , obtaining a total gas mixture with an average calorific value of 82.45 kcal·kg −1 for CeO 2 , 102.51 kcal·kg −1 for CeFe1Pd, 124.23 kcal·kg −1 for CeCo1Pd1, and 139.14 kcal·kg −1 for CeNi1Pd1 on time evaluated. It is observed that the dispersion and average crystallite size of Pd in functionalized systems follows the same order as the increase in H 2 release and the average calorific values of the release gas mixture. As higher the dispersion and lower the average crystal size, the higher the calorific values and the content of H 2 produced. Figure 10 depicts the results for resins II. The resins followed the same trend as n-C 7 asphaltenes. The main differences lie in how the nanocatalysts generate significant changes in the selective distribution of gases and their volumetric proportion.
The maximum production of CO 2 in all cases is observed at 10 min, which follows the incremental order of CeNi1Pd1 < CeFe1Pd1 < CeCo1Pd1 < CeO 2 . The order in which the LHCs were produced in this interval follows the opposite order. This indicates that in the presence of the CeNi1Pd1 catalyst, the resins decompose a high content of their aliphatic structure. The production of CO 2 tends to decrease, being more noticeable in the catalyst CeNi1Pd1. In other words, the catalyst promotes C-CO 2 interactions, increasing the release of CO [69]. Methane reforming reactions appear to occur between 30 and 40 min for functionalized catalysts and between 60 and 70 min for CeO 2 support. On the other hand, the production of H 2 intensified with the reduction of CO; that is, there is a participation of this gas in WGS reactions. This reaction occurs at higher temperatures, between 40 and 60 min for CeNi1Pd1, 60 and 80 min for CeCo1Pd1 and CeFe1Pd1, and 90 and 100 min for CeO 2 .
The future trend in hydrogen production is increasing toward Fe < Co < Ni for both fractions. Many researchers cite the oxygen storage/reduction of cerium-based materials and nanocrystals' catalytic activity on the supports' surface as a key component of the WGS and methane reforming activity of catalysts [67,69,[73][74][75].
There have also been studies showing that the addition of dopants or a combination of multiple metals as active phases creates defects in the structure and oxygen vacancies in which pure, highly dispersed metal catalysts are located [62,67]. This allows the adsorption of reactants and facilitates the WGS and methane reforming reactions themselves. The hydrogen production trend is also agreed with each material's catalytic activity for decomposing both fractions at the evaluated conditions. CeNi1Pd1 showed a higher performance than the other three samples and led to a higher hydrogen vol% in the effluent gas.
Cerium nanoparticles as active support facilitate the production of low molecular weight compounds due to (a) the positive effect of the NOT and NON elements and ceria interactions, and (b) the activity of redox couple Ce 3+ /Ce 4+ . There is evidence that oxygen atoms migrate from the support (CeO 2 ) to transition element particles at higher temperatures during spillover phenomena. This oxygen can be used to oxidize CO to CO 2 , and subsequently, the partially reduced ceria can be re-oxidized by water, completing the catalytic cycle. Moreover, the gases can be used to produce H 2 during the WGS reaction. In this sense, the surface of the support will be composed of Ce 3+ and Ce 4+ ions. Ce 3+ ions formed after reducing Pd-based catalysts play an essential role in activating CO and CO 2 molecules at the ceria-transition element interface.
The heterogeneous surface of the material promotes different types of reactions. For example, the Ni/CeO 2 active sites first generate the total combustion of CH 4 followed by H 2 O and CO reforming. For Pd/CeO 2 , the reaction follows a direct mechanism with H 2 , CO, CO 2 , and H 2 O's simultaneous formation. These differences are explained due to the transfer of electrons between the element's particles and the support. In a steam atmosphere, the CO oxidation is enhanced by OH· and O 2 species' reaction to the support's basic sites, following the gasification mechanisms, increasing CO 2 and H 2 production. Methane activation occurs through the active surface of oxygen. The high electronic affinity with the coordinated surface of unsaturated oxygen (cus) generates a strong interaction with methane and light hydrocarbon molecules, allowing the activation of C-H bonds at low temperatures and their subsequent breakdown, finally desorbing as CO 2 and H 2 O.
The surface chemistry of the nanocatalysts was correlated with the results of selective distribution in hydrogen releasing from asphaltene and resin decomposition. The results are shown in Figure 11. As described, hydrogen release increases in the order CeFe1Pd1 < CeCo1Pd1 < CeNi1Pd1, which agrees with the tendencies for Pd 2+ , Ce 3+ and O ads /O lat in the surface chemistry of the nanocatalysts. It is well recognized the redox property of CeO 2 influences catalytic activity. The relative abundance of Ce 3+ and O ads /O lat suggests their participation inf hydrogen formation. First, the redox cycle allows the oxygen adsorption under oxidation conditions and its release under reduced conditions, as is shown in Equations (3) and (4): Furthermore, hydroxyl groups (OH) formed by the reaction between H 2 O and partially reduced ceria oxygen vacancies react with CO to form methanoate species of bridge [76]. Above 170 • C, these species are transformed into bidentate ormats to decompose into final products of CO 2 and H 2 [76]. Similarly, the higher the content of Pd 2+ species, the higher the selective distribution for hydrogen production in catalytic decomposition of both fractions. These ions could substitute Ce 4+ in the ceria lattice to increase the defect concentration. Hence, the catalysts improved their oxygen vacancy density and oxygen mobility as increases Pd 2+ in their surface. In complete agreement, it is reported that the strong interactions between Ni and Pd species with Ce support lead to more thermal stable metal species [38], which contributed to the synergistic effect of both phases for crude oil fractions decomposition and hydrogen production.

Effect of R:A Ratio
The CeNi1Pd1 was selected to evaluate the effect of R:A ratio on hydrogen release from its isothermal steam catalytic gasification at 220 • C. The results are shown as a function of the conversion degree (α) in Figure 12. It is found that the R:A ratio influences the H 2 content on the effluent gas. At low conversions (α < 0.2) the H 2 content is very similar in all systems. Between 0.4 and 0.7 the trend is evident. For the same conversion, the production of H 2 increases in the order of R < 8:2 < 1:1 < 2:8 < A. As shown in the previous section, different reactions contribute to the generation of H 2 in the presence of the CeNi1Pd1 catalyst. However, depending on the chemical nature of the adsorbed fraction, H 2 release can vary. High content of short aliphatic chains promotes the production of methane in greater quantity, and as previously shown, asphaltenes have a higher content of H α species. In this sense, methane reforming may have a greater contribution as the content of asphaltenes increases in R:A mixtures. On the other hand, the systems show a production peak generated at lower conversions when the asphaltene content is lower. These data coincide with the slight increase in the production of CO 2 in samples A and R, an indicator of WGS reactions. For conversions greater than 0.8, H 2 production is fluctuating. However, it does not fall below 40.0 vol%. These results indicate the nanocatalyst's potential to produce a hydrogen-rich gas from the steam catalytic gasification of asphaltenes and resins adsorbed both individually and in mixtures. All the results demonstrate the potential of the technology to release a hydrogen-rich gaseous mixture.

Effect of Temperature
Operation temperature was varied between 210 and 230 • C to analyze its effect on H 2 generation from the steam catalytic gasification of n-C 7 asphaltenes. Figure 13 shows the results obtained. Higher H 2 concentration was observed as the temperature increased. This indicates that at 230 • C the most favorable conditions for steam reforming and WGS reactions are received. This trend is following the degree of conversion of each fraction. In the same figure, it is observed that asphaltenes convert greater mass at 230 • C. Therefore, it produces a greater quantity of H 2 catalyzed by CeNi1Pd1. The higher the temperature, the higher the H 2 O (g) and CO 2 gasification reaction rates, obtaining a higher release of H 2 at 230 • C than the other two temperatures evaluated.

Conclusions
For the first time, the present study shows the effect of resin II in the adsorption, and catalytic steam gasification of n-C 7 asphaltene on hydrogen production using functionalized CeO 2 nanocatalysts with different pairs of transition element oxides. Individual adsorption of resin II follows the order of Co < Fe < Ni. For the competitive adsorption, the simultaneous presence of two fractions influences the individual adsorption of n-C 7 asphaltene and resin II correspondingly because when the amount of resin II increases on the system, the adsorption of n-C 7 asphaltenes decreases. Additionally, resin II decomposition temperature decreased from 420 to 220 • C for bielemental nanocatalysts and 300 • C for the support of CeO 2 ; whereas n-C 7 asphaltenes decompose about 220 • C for bielemental nanocatalysts and 350 • C when the support catalyzes them. For the R:A systems, the conversion increases in the order of (R:A) 8:2 < 1:1 < 2:8, because asphaltenes have a greater content of heavy molecular weight components that require higher time for their decomposition. The effective activation energy is decreased by the presence of CeNi1Pd1 nanocatalysts in all R:A and individual systems, evidencing the positive thermal effect against the steam injection processes.
Hydrogen release was successfully quantified. The results depict an increasing H 2 production order of CeO 2 < CeFe1Pd1 < CeCo1Pd1 < CeNi1Pd1 under isothermal heating at 220 • C during asphaltene and resin conversion. The CeNi1Pd1 catalyst produces a gas effluent composed of more than 40 vol% of H 2 , decomposing both fractions. When the samples are heated at 230 • C, H 2 production was increased close to 55 vol% during n-C 7 asphaltene conversion.
It was noted that asphaltenes produce a higher amount of CO 2 and a lower amount of H 2 than resins II in all the time range evaluated. This result agrees with the higher content of aromatic carbon in asphaltenes. During thermal heating, the aromatic core is susceptible to the formation of ketones and finally releasing as CO 2 . The higher the content of hydrogens α and γ linked to aromatic rings, the higher the hydrogen release. In the presence of CeNi1Pd1 a high efficiency to produce H 2 was obtained regardless of the nature of the fraction and the R:A systems. That is, the decomposition of the R:A mixtures also generated gas contents similar to the individual fractions.
Likewise, CO 2 production was minimal in all the cases evaluated, highlighting the use of the nanocatalysts studied. An gas average mixture with a content of less than 4% vol in the presence of CeNi1Pd1 was produced, a rather low result and similar to that reported for the gasification of biomass and natural gas.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/nano11051301/s1, Figure S1. Softening point calibration curve. Figure S2. Adsorption isotherms constructed for (a) n-C 7 asphaltenes and (b) resins II on CeNi1Pd1 nanoparticles for different R:A ratios. The dotted lines are from the SLE model and the symbols are experimental data. Table S1. Estimated SLE model parameters for adsorption isotherms of n-C 7 asphaltenes and/or resins II on different nanocatalysts and different Resins-Asphaltene (R:A ratios). H represents the Henry's law constant, K gives an idea about the self-association degree of asphaltenes over nanoparticle surface, and q m is the maximum amount adsorbed. Table S2. Estimated slope and intercept of the linear plots of q predicted as a function of q experimental for the adsorption and desorption of n-C 7 asphaltenes in absence and presence of resins II in the systems at different R:A ratios. Section S.1-Solid-Liquid Equilibrium model. Section S.2-Activation energy estimation Funding: The author thanks Universidad Nacional de Colombia for the financial and logistical support.