Promotion Effect of H₂S at High Concentrations on Catalytic Dry Reforming of Methane in Sour Natural Gas

Abstract

The effect of high concentrations of H₂S in sour natural gas on the catalytic dry reforming of methane (DRM) process has seldom been studied previously in the literature. Herein, several types of catalysts, including MgO, NiO/MgO, and LaNiO₃ in different states, were prepared for conducting DRM at 800 °C and 0.1 MPa in a feed of 20 vol% CO₂ and 20 vol% CH₄, and their catalytic performance under conditions of the absence and presence of H₂S was compared. A promotion effect of increasing H₂S concentration on both the conversions of CO₂ and CH₄ and the molar yields of CO and H₂ was observed on all the catalysts and was particularly remarkable on the MgO and the pristine NiO/MgO. For NiO/MgO, the addition of 15 vol% H₂S increased the conversion of CH₄ from 6.92% to 26.86% and CO₂ from 9.15% to 42.10%. While there was a significant decline in the catalytic activity of the reduced NiO/MgO and LaNiO₃ catalysts after adding H₂S, moderate reactant conversions were still sustained. The results of process analysis and catalyst structure characterization suggest that H₂S participation can contribute to the increment in CO₂ and CH₄ conversion, and active S-adsorbed species may play the key role of catalysis in reactions involving H₂S.

1. Introduction

The energy shortage and climate change are currently two major global issues of concern to people. Natural gas is one of the most wildly used clean energy sources in the world, and a production output of ca. 4 × 10¹² m³ was reached in the year of 2022. Along with the continuous extraction of natural gas, unfortunately, large quantities of sour natural gas containing high percentages of both CO₂ and H₂S come out of the ground. The data extracted from some high-sulfur gas fields given in

Table 1. CO₂ and H₂S concentration data of several gas fields worldwide [1].

Gas Field	CO2 (vol%)	H2S (vol%)	Country
East Crossfield	12.0	36.0	Canada
BEC	40.0	21.0	America
Lacq	9.3	15.6	France
Zechstein	20.0–50.0	15.0–20.0	Germany
Astrakhan	15.0	25.0	Russia
Puguang	8.6	15.2	China

provide the typical cases of CO₂ and H₂S content in natural gas worldwide [1]. Information on the distribution of sour natural gas fields in the world can be found in ref. [2].

Taking advantage of available catalytic dry reforming of methane (DRM) technology, the natural gas industry can not only diminish CO₂ emissions but also produce high value-added organic chemicals via gas synthesis [3,4,5]. However, the coexistence of H₂S in feedstock makes the application of catalytic DRM technology exceedingly challenging, since H₂S is generally a notorious poison of catalysts.

So far, researchers have made tremendous efforts to address the deactivation of catalysts caused by H₂S, with the aim of attaining highly efficient catalysts that can tolerate H₂S. For example, Appari et al. [6] developed a detailed kinetic model for biogas steam reforming in the absence and presence of H₂S having 20–108 ppm over a Ni-based catalyst. As stated, the model predicted a high sulfur coverage near the reactor inlet during the initial stages of poisoning. When the reaction proceeded further, the sulfur coverage increased towards the reactor outlet. Low-temperature operation could render the catalyst completely deactivated, while high-temperature operation could mitigate sulfur adsorption and slow down the catalyst deactivation. Gaillard et al. [7] investigated the effect of trace (50 ppm) H₂S on the DRM process over Mo- and Ni-based catalysts and dissected the mechanism of catalyst deactivation. They found that carbon accumulation was the main cause of the deactivation of Mo-based catalysts, whereas H₂S poisoning was responsible chiefly for the activity loss of Ni-based catalysts. The effect of H₂S on dry reforming of biogas was studied by Chein et al. [8]. Their experimental results showed that the presence of H₂S led to a severe deactivation of Ni-based catalysts. The bimetallic Pt-Ni catalysts outperformed the monometallic Ni-based catalysts, but deactivation still emerged. Akansu et al. [9] explored the effect of Mn doping on the activity and the H₂S tolerance of Ni-based catalysts in dry reforming of biogas. As reported, both the monometallic Ni and the bimetallic Ni-Mn catalysts exhibited stable activity in the absence of H₂S. In the presence of 2 ppm H₂S, the activities of the two types of Ni-based catalysts decreased gradually with reacting time, and Mn doping did not improve the resistance of the catalysts to H₂S. More recently, Gao et al. [10] inquired into the single and synergistic effects of H₂S and NH₃ on Ni-based catalysts in biogas dry reforming. They stated that trace amount of H₂S caused severe catalyst deactivation, in contrast to NH₃, which had a slight inhibitory effect. In the condition of H₂S and NH₃ coexistence, the catalyst activity decreased more rapidly than when a single component was present.

Meanwhile, there are occasionally research reports that indicate the addition of H₂S can promote the catalytic DRM process. Fidalgo et al. [11] investigated the effect of a small quantity (0.5–1.0 vol%) of H₂S present in the feedstock on DRM over activated carbon- and Ni-based catalysts. As disclosed, for the carbon catalyst, the conversions of both CH₄ and CO₂ were noticeably increased by the addition of H₂S to the reactants. For the Ni-based catalysts, on the contrary, significant deactivation was observed after adding H₂S. The influence of trace amounts of H₂S on commercial methane reforming catalysts during hydrogen production at different temperatures (650–850 °C) was plumbed by Chattanathan et al. [12]. They noticed that even with the addition of 0.5 mol% H₂S, the conversions of CH₄ and CO₂ were significantly dropped from 67% and 87% to 19% and 22%, respectively. However, there was a slight increase in the conversions of both CH₄ and CO₂ when the H₂S concentration was raised from 1.0 mol% to 1.5 mol%. Recently, Jin et al. exerted an in situ observation of the promoting effect of H₂S on the formation of efficient MoS₂ catalysts for CH₄/CO₂ reforming and found that introducing H₂S further improved the catalytic reactivity by >10% [13].

Although the relevant research was very limited, the findings of a promotion effect of H₂S on the catalytic process of dry reforming is undoubtedly exciting. H₂S is ubiquitous, either as primary impurity or as a hydrogenolysis product of organic sulfurs, in fossil fuels and biomass processing. To avoid affecting catalyst performance, removing H₂S from feedstock as much as possible has long been an arduous task. On the other hand, with the worldwide emphasis on hydrogen energy, it has been realized that H₂S is also a valuable source of H₂. Utilizing H₂S to produce H₂ rather than just acquiring sulfur through Claus processes has been put on the agenda. However, rare attention has been paid to the possible reactions occurring directly among the H₂S, CO₂, and CH₄ of sour natural gas, which can in effect generate beneficial products including H₂.

Herein, we will focus on the catalytic DRM reactions in the presence of H₂S with high concentrations and compare the catalyst performance in the conditions of with and without H₂S. Several typical DRM catalyst materials have been elaborately investigated in reaction systems, including MgO, NiO/MgO, and LaNiO₃ of different chemical states. A promotion effect arising from the reactions of H₂S with CO₂ and CH₄ has been discovered on all the catalysts, which can bring the conversions of CO₂ and CH₄ and yields of CO and H₂ in the presence of H₂S to reach a meaningful level of production in industry. A relatively long-term stability of the reduced NiO/MgO catalyst has been achieved with little carbon deposition and sulfur poisoning. The possible promotion mechanisms of high concentration H₂S have been discussed on the basis of a thermodynamic analysis and the characterization of the catalyst structures before and after the reactions. S-adsorbed species formed by dissociation of H₂S on the surface of catalysts has been suggested to be crucial for initiating the reactions among H₂S, CO₂, and CH₄. Moreover, the promoting effect of H₂S can even occur on the MgO support, manifesting its universality. To the best of our knowledge, there seems to have been no report up to now that has revealed the promotion effect of H₂S in the process of catalytic dry reforming. The results and deductions of the present study may be conducive to guide rational processing and high-value utilization of sour natural gas.

2. Results and Discussion

2.1. Thermodynamic Analysis of the Effect of H₂S Concentration on DRM Reactions

Thermodynamic equilibrium analysis of the reactions involving CO₂, CH₄, and H₂S as feedstock was conducted firstly by Aspen Plus simulation to ascertain the theoretical basis for the participation of H₂S in reactions. Considering that reaction systems involve polar gases reacting at high temperature and low pressure, the NRTL property equation was chosen and a Gibbs isobaric and isothermal reactor was used. The flow chart for Aspen Plus simulation is supplied in Figure S1 of Supplementary Materials. The reaction conditions for the simulation were assumed as 20% CO₂, 20% CH₄, and H₂S of 0–15%, all in volume, balanced with N₂ at the temperature range of 400–1000 °C and pressure of 0.1 MPa. Nine possible products, including H₂, H₂O, C, C₂H₆, CO, COS, CS₂, S₂, and SO₂, all in a gas state, excepting solid C, were supposed to form in the reaction system. The results are shown in Figure 1, where (a) and (b) display the changes in the equilibrium conversions of three reactants and the molar fractions of the products with reaction temperature, respectively. Because the molar fractions of C, C₂H₆, and SO₂ are all below the order of 10⁻⁶, these three products will not be taken into account thereafter.

As can be seen, there is a slight fluctuation for the equilibrium conversion of H₂S, which varies between 4.8–8.7% in the temperature range of 400–1000 °C, although that of either CO₂ or CH₄ increases significantly from ca. 5% to nearly 100% with the temperature rising. Meanwhile, the molar fractions of both CO and H₂ overwhelm other products in the whole temperature range, especially when higher than 800 °C. The byproducts COS and H₂O each account for a small proportion. Their individual maximum fractions do not exceed 1.1% and 1.9%, appearing around 400 °C and 600 °C, respectively. The quantity of CS₂ and S₂ is even more trivial.

The equilibrium conversions of the three reactants and the molar yields of the products changing with H₂S concentration (0–20 vol%) at the temperature of 800 °C are drawn in Figure 2a,b, respectively. There is a marginal growth of the H₂S conversion with the increase of H₂S concentration. Nevertheless, for the conversions of both CO₂ and CH₄, almost no impacts of H₂S concentration can be perceived. On the other hand, the molar yield of CO, one of the dominant products, declines from 95.65% to 93.38%, and the molar yield of H₂, the other dominant product, drops sharply from 92.97% to 63.78%. It should be noted, however, that the yield of H₂ could increase linearly from 92.97% to 95.67%, as presented in Figure 2b, if it is calculated on the basis of feed CH₄ only, just like when there is no H₂S present. This suggests that there would be a potential promotion effect of H₂S concentration on H₂ production. With respect to the other products, the molar yields of COS and H₂O are scarcely affected by the wave of H₂S concentration, with their values maintained separately around 4.25% and 2.33%. The byproduct CS₂ seems to have an obvious increase but, in fact, its maximal molar yield is still very low. For S₂, the effect of H₂S concentration is not worth mentioning either. The results of thermodynamic analysis show that high concentrations of H₂S could not inhibit the conversions of both CO₂ and CH₄; instead, they were able to increase the production of H₂. This situation is a comprehensive consequence of the possible reactions that may occur between the three raw materials and their products, depending on the Gibbs free energy value of each substance and Gibbs free energy change of each reaction, which will be discussed later.

2.2. Effect of H₂S Concentration on DRM over Different Catalysts

2.2.1. MgO Catalyst

Different catalytic materials were tested to judge whether the effects of H₂S concentrations on DRM are distinguishable from one another. MgO, an alkaline earth oxide, generally has a good adsorption ability towards acidic gases such as H₂S and CO₂. Furthermore, MgO can be used as a support to disperse various active phases, which can not only promote the activation of acidic reactants but also prevent coke formation from hydrocarbon cracking. As Hu et al. claimed in their early research [14], Ni/MgO solid solution catalysts can display excellent dry reforming performance without coke formation and metal sintering. The latest work reported by Feng et al. [15] believed that the addition of MgO in Ni/Al₂O₃ catalysts could improve the alkalinity of the support and enhance the metal-support interaction, both of which are beneficial to the reaction of DRM.

In view of the MgO properties, we firstly examined the performance of MgO during DRM reactions in the absence and in the presence of H₂S with different concentrations. Figure 3a shows the conversions of CO₂ and CH₄ and the molar yields of CO, H₂, and H₂O in the temperature range of 400–800 °C in the absence of H₂S. The reactant mixture gas is composed of 20% CO₂ and 20% CH₄ in volume balanced with N₂ at the pressure of 0.1 MPa and the GHSV of 10,000 h⁻¹. It can be inspected that the conversions of CO₂ and CH₄ are merely 4.09% and 0.92%, respectively, even with the temperature as high as 800 °C, indicating the poor catalytic performance of MgO for DRM. The main products are H₂O and CO, despite the fact that their productions are very low. In the meantime, the formation of H₂ can hardly be observed in the whole temperature range.

For the DRM proceeding in the presence of 15 vol% H₂S in addition to 20 vol% CO₂ and 20 vol% CH₄, the variations in the conversions of CO₂, CH₄, and H₂S as well are displayed in Figure 3b. Surprisingly, the conversions of both CO₂ and CH₄ increase significantly in comparison with those obtained at the same temperatures but in the absence of H₂S. At 800 °C, the conversion of CO₂ can reach 36.30% and that of CH₄ is 19.36%, which are about 9 and 21 times the corresponding conversions under the reaction conditions without H₂S, respectively. The comparison between the reactant conversions under the conditions without and with H₂S discloses the apparent promotion effect of H₂S addition. Unlike the monotonic increases in CO₂ and CH₄ conversions, the conversion of H₂S exhibits a volcanic curve-like change with the rise of reaction temperature. The summit of H₂S conversion is 14.30%, appearing at 600 °C.

The effect of H₂S addition can be further evidenced by the boost in product molar yield. As illustrated in Figure 3c, the molar yields of the products such as CO, H₂, and H₂O at 800 °C all grow remarkably as compared with the counterpart results shown in Figure 3a. In particular, the production of H₂ achieves a molar yield of 10.07%, being second only to that of CO; the latter is 24.7%. Additionally, a new product, COS, was observed which has a maximal molar yield of 10.47% at 700 °C. The influence of H₂S concentration on the conversion of each reactant can be ascertained from the curve trend shown in Figure 3d. For CO₂ and CH₄, there are unequivocally positive effects of H₂S concentration on their conversions. Regarding H₂S transformation, there appears to be an optimal concentration of about 10 vol% H₂S to achieve the largest conversion of 10.99%.

2.2.2. NiO/MgO Catalyst in Pristine and Reduced Form

As mentioned earlier, catalysts based on Ni/MgO are frequently used in the processes of DRM due to their outstanding catalytic performances. Herein, we prepared a NiO/MgO catalyst for DRM and used it in either its pristine or reduced form to explore the sensitivity of the chemical state of active metal on the H₂S-induced effect. The catalytic performances of the pristine NiO/MgO catalyst in the absence of H₂S and in the presence of 15 vol% H₂S at different temperatures are depicted in Figure 4a,b. By comparing between the results of NiO/MgO in Figure 4a,b and those of MgO in Figure 3a,b, one can be aware of that the variation trends with rising temperatures of either the reactant conversions or product molar yields are very similar for the two catalyst types, whatever in the absence or in the presence of H₂S. Nonetheless, on the NiO/MgO catalyst, there are perceptible increments of both reactant conversions and product molar yields. This proves the superior behaviors of NiO/MgO to the simple MgO catalyst. Furthermore, at 800 °C, the conversion of CO₂ and the molar yield of CO are 42.10% and 33.10%, respectively, on the NiO/MgO catalyst in the presence of H₂S, which are not only higher than the results recorded in the same conditions on the simple MgO catalyst but also much higher than the corresponding values gained in the absence of H₂S. Obviously, the promotion effect of H₂S has also been confirmed on the pristine NiO/MgO catalyst.

For the reduced NiO/MgO catalyst, the catalytic performances in the reaction conditions without and with 15 vol% H₂S at different temperatures are demonstrated in Figure 4c,d, respectively. It is interesting that the conversions of both CO₂ and CH₄ and the molar yields of both CO and H₂ are improved greatly on the catalyst of NiO/MgO after reduction, as shown in Figure 4c, when compared with the corresponding results on the pristine catalyst in Figure 3a. At 800 °C, the conversion of CO₂ and the molar yield of CO are 94.05% and 85.04%, respectively, which are comparable to the data reported in the literature [14,16]. The results manifest the importance of the chemically reduced state of the active metal phase in the NiO/MgO catalyst for the reaction of DRM. In the condition of 15 vol% H₂S presence, the changes of reactant conversions and product molar yields with the temperature in Figure 4d are virtually identical to those presented for the pristine NiO/MgO catalyst in Figure 4b. However, in comparison with the results gained in the absence of H₂S in Figure 4c, noticeable drops in the conversions of CO₂ and CH₄ and the molar yields of CO and H₂ can be found, suggesting an inhibition effect of H₂S on the catalytic activity of the reduced NiO/MgO catalyst for DRM.

The differentiation in the H₂S concentrations’ influence on the catalytic performances between the pristine and the reduced NiO/MgO catalysts can be surveyed in Figure 4e,f. On the reduced NiO/MgO catalyst, as displayed in Figure 4e, the trends of the reactant conversions with H₂S concentration are basically the same for the two catalyst forms, excepting that there are sharp declines in the CO₂ and CH₄ conversions when the reaction atmosphere is altered from not containing H₂S to containing H₂S. Further increasing H₂S concentration can benefit the conversions of CO₂ and CH₄ over both of the NiO/MgO catalysts, resembling the circumstance occurring over the MgO catalyst. The changes of the product molar yields with H₂S concentration, shown in Figure 4f, are consistent with the paces of reactant conversions. Additionally, the pristine NiO/MgO catalyst performs to some extent better than the reduced one in the reaction system containing H₂S, inferring that the catalytic performance may be sensitive to subtle variations in the surface of the catalyst.

2.2.3. LaNiO₃-Derived Catalyst

LaNiO₃ is the most widespread Ni-based perovskite oxide used as a precursor for DRM, due to the stable performance of its derived catalyst during reaction, in contrast with conventional Ni-based catalysts [17,18,19]. We also prepared the material of LaNiO₃ and reduced it to form a Ni-based active phase before running the process of DRM. The results of CO₂ and CH₄ conversions and the molar yields of CO and H₂ at different temperatures in the condition of no H₂S addition are shown in Figure 5a. High activity of the LaNiO₃-derived catalyst is achieved as expected. In the presence of H₂S, as shown in Figure 5b, the changes in the conversions of CO₂ and CH₄ as well as H₂S with the H₂S concentration are somewhat analogous to those appearing over the reduced NiO/MgO catalyst. The addition of H₂S suppresses to a certain extent the activity of the LaNiO₃-derived catalyst for CO₂ and CH₄ transformation; however, a partial revival of catalytic performance can be effectuated by increasing the H₂S concentration.

2.3. Discussion

As revealed above, the promotion effect on catalytic performance arising from the addition of a high concentration of H₂S can be observed on all of the materials investigated including the MgO, the NiO/MgO, reduced or not, and the reduced LaNiO₃ catalysts, implying that there must be a similar reason behind each of these results. After comprehensive deliberation of the product composition detected and their possible formation routes through consulting our previous study and other research published [20,21,22,23], five independent reactions and their thermodynamic functions have been figured out, as listed in

Table 2. Independent reactions involved in the DMR process having H₂S addition.

No.	Reaction Equation	ΔH(1073.15 K)/kJ·mol−1	ΔG(1073.15 K)/kJ·mol−1
(1)	CO2 + H2S ↔ H2O + COS	34.4	31.6
(2)	CH4 + 2H2S ↔ 4H2 + CS2	261.0	29.6
(3)	H2S ↔ H2 + 1/2S2	90.3	37.5
(4)	CO2 + CH4 ↔ 2H2 + 2CO	260.4	−45.0
(5)	CO2 + H2 ↔ H2O + CO	34.0	0.7

below.

It is apparent that the pathway through which H₂S participates in the reaction may not only be embodied in Equation (1) but also probably in Equations (2) and (3). As a matter of fact, traces of CS₂ and S₂ were searched out occasionally during reaction tests in the collected liquid samples and on the reactor outlet wall, respectively, affirming the possibility of the occurrence of Equations (2) and (3). According to the thermodynamic analysis results relevant to CS₂ and S₂ formation given in Figure 2b, the yields of CS₂ and S₂ are exactly very low in theory. Therefore, the negligible amount of CS₂ and S₂ gathered is reasonable. In contrast, COS is one of the more critical products formed on all the catalysts, which most likely arises from H₂S transformation with CO₂. Moreover, the molar yield of COS in practice is generally higher than the theoretical value. This consequence hints that the reaction between H₂S and CO₂ may take priority in virtue of selective catalysis over other CO₂ reactions, especially those proceeding following Equation (4), while the latter is thermodynamically more favorable. The fall-off in conversions of CO₂ and CH₄ after adding H₂S to the systems of the reduced NiO/MgO and LaNiO₃ catalysts points to the alteration of catalytic active phases, thus causing a decrease in the activity of the two catalysts for the direct reaction between CO₂ and CH₄. On the other hand, the occurrence of diverse reactions of H₂S with CO₂ and CH₄ can multiply the transformation pathways of both CO₂ and CH₄, and a high H₂S concentration is definitely promotive to the conversion of CO₂ and CH₄. This should be the principal reason for the promotion effect of high concentrations of H₂S in catalytic DRM reaction systems.

The reaction results of the reactant mixture of 20 vol% CO₂, 20 vol% CH₄, and 15 vol% H₂S on different catalysts tested at 800 °C are summarized in

Table 3. Conversions of three reactants and molar yields of main products on various catalysts at 800 °C and 0.1 MPa.

Catalyst Sample	CO2 Conversion (%)	CH4 Conversion (%)	H2S Conversion (%)	CO Molar Yield (%)	H2 Molar Yield (%)
MgO	36.30	19.39	8.52	24.72	10.07
Pristine NiO/MgO	42.10	26.86	8.03	33.10	16.93
Reduced NiO/MgO	34.76	19.03	7.39	25.33	10.85
Reduced LaNiO3	28.86	15.38	9.06	18.88	6.71

. The pristine NiO/MgO catalyst performs prominently among all in terms of reactant conversion and product yield. The conversions of CH₄, CO₂, and H₂S are 26.86%, 42.10%, and 8.05%, respectively, comparable to the recent literature results. In the report of Jin et al., the corresponding conversions of CH₄, CO₂, and H₂S are 23.2%, 54.3%, and 5%, respectively, for a feed flow of CH₄:CO₂:H₂S:N₂ = 25:13:13:49 on a MoS₂/Al₂O₃-F catalyst at the same reaction temperature [13]. It is meaningful that the conversions of CO₂ and CH₄ in the presence of H₂S can be sustained at a level that can be applicable to industry, in spite of being somewhat lower than those obtained over the reduced NiO/MgO and LaNiO₃ catalysts in the absence of H₂S. In particular, the catalysts can at least partially overcome the impact of activity reduction caused by H₂S addition.

It is worthy of mentioning that there was no detectable decline in catalytic activity of any catalyst with reaction time on stream throughout the entire testing period in the condition of H₂S presence. As a matter of fact, we measured the stability of reduced NiO/MgO catalyst for longer than 60 h in feed gas of 20 vol% CO₂, 20 vol% CH₄ and 15 vol% H₂S at 700 °C. The conversions of the three reactants changing with time on stream are supplied in Figure S2 of Supplementary Materials. The conversion curves are considerably flat, implying that the deactivation typically resulting from coke formation or H₂S poisoning in such kinds of reaction systems may slightly happen to the catalyst.

In order to understand the distinctive role each catalyst played in the reactions, a series of characterizations of the catalysts was implemented by use of BET, XRD, and XPS techniques, etc. The sample labelled as “used” refers to the catalyst having undergone the DRM reactions at 800 °C for 1 h in the presence of 15 vol% H₂S. The BET measurement results of the catalyst samples before and after the reactions are listed in

Table 4. BET results of the catalysts before and after reaction.

Catalyst Sample	Specific Surface Area/m2·g−1	Pore Volume/cm3·g−1	Average Pore Diameter/nm
MgO	16.20	0.055	13.7
Used MgO	15.66	0.047	11.9
Pristine NiO/MgO	40.12	0.30	30.3
Used NiO/MgO	45.67	0.46	40.5
Reduced NiO/MgO	43.28	0.40	35.5
Used reduced NiO/MgO	42.10	0.37	35.0
Pristine LaNiO3	4.21	0.025	24.0
Used reduced LaNiO3	3.28	0.017	20.3

, and the corresponding cryogenic N₂ adsorption–desorption isotherms and pore size distribution profiles are displayed in Figures S5 and S6, respectively. The catalysts of the MgO, the pristine and reduced NiO/MgO, and the reduced LaNiO₃ differ greatly from one another in their specific surface area, pore volume, and average pore diameter. After completing the reactions at 800 °C for 1 h in the presence of H₂S, almost all the catalysts are little altered with respect to their textural features in comparison with their fresh states, reflecting the high tolerance of the catalyst structures to the H₂S-containing reactive atmosphere at high temperatures. For LaNiO₃, there are obvious changes in N₂ adsorption–desorption isotherm and pore size distribution profile, shown in Figures S3 and S4, respectively. This must originate from the transformation of the crystal structure of LaNiO₃ in the process of reaction, as will be revealed in the discussion of the XRD results in the following text. Nevertheless, probably due to its inherently small specific surface area, the surface area of LaNiO₃ becomes more trivial after experiencing reactions. Associating with the catalytic performance of these catalysts in H₂S atmosphere, as supplied in Table 3, the texture properties of a catalyst seem to have an impact on the catalytic reactions to some extent. A high specific surface area of the catalyst might be helpful to the conversion of the reactants.

The XRD spectra of the three types of the catalysts in various states are spread out in Figure 6a–c. As shown in Figure 6a, the two samples of the MgO before and after the reactions involving H₂S are virtually identical in the appearance of the characteristic peaks of crystal phases. Both of them can be assigned to the standard MgO of PDF #45-0946, indicating that there is little influence of H₂S on the crystal structure of MgO. For the pristine NiO/MgO, its characteristic peaks on XRD spectrum in Figure 6b also look to be the same as those of MgO. This situation should be associated with the formation of solid solution of NiO and MgO. As revealed in the literature, a pure phase solid solution of brucite can be easily formed from a mixture of NiO and MgO by isomorphous substitution, resulting in an almost identical XRD pattern of NiO/MgO to that of MgO, due to the fact that bivalent cations of Ni²⁺ (0.62 Å) and Mg²⁺ (0.69 Å) have close ion radius [24,25,26]. The formation of a solid solution can favor the homogeneous Ni distribution in the NiO/MgO catalyst, despite a NiO loading as high as 20 wt%. Incidentally, such a high NiO loading is fairly profitable to achieve better catalytic performance in the dry reforming process, as already evidenced by a few researchers [24,27]. On the spectrum of the reduced NiO/MgO sample, there appears a small protrusion at the foot of the chief peak of MgO at 2θ of 42.916°, the position of which is basically in concordance with the dominant peak at 44.507° of metallic Ni belong to PDF #04-0850. Making a comparison between the pristine and reduced NiO/MgO, one can notice the approximately consistent XRD spectra for the two catalysts, both after the reactions involving H₂S. A few of characteristic peaks of NiS emerge beside those of MgO, manifesting that the sulfidation of Ni and/or NiO to NiS occurred during the reactions. The sulfidation phenomenon also exists on the catalyst of reduced LaNiO₃, as shown in Figure 6c, where the phases of NiS and La₂O₂S were transformed from the reduced LaNiO₃ after the reactions in the presence of H₂S. In consideration of the relatively close reaction results over different catalysts, it can be deduced that the catalytic performance does not depend greatly on the crystal phase of the catalyst, whether it is MgO, NiS, or La₂O₂S.

To deeply explore the active phase or sites on the catalyst surface for the reactions, XPS investigation was carried out. The full-range spectra of various catalysts in different states are exhibited in Figure 7a,b. The characteristic peaks of the elements belonging to the catalyst compositions of MgO and NiO/MgO, such as Mg, Ni, and O, can be easily discovered in Figure 7a. However, the signal of S 2p within 160–170 eV, which usually represents the existence of the S element on the catalyst, can hardly be probed on the surface of the used MgO. The S 2p signal is also very faint for the used NiO/MgO catalyst, regardless of whether the NiO/MgO was reduced or not before the reactions. In contrast, on the used LaNiO₃-reduced catalyst, the peak of S 2p can be distinctly observed in addition to the peaks of La, Ni, and O, etc. Taking into account the greater amount of S-containing crystal species, such as La₂O₂S and NiS, in the used LaNiO₃-reduced catalyst than that in the used NiO/MgO catalyst, which contains NiS only, the higher peak intensity of S 2p of the former catalyst on XPS spectrum is thus explainable. Since the active sites for the DRM reactions over Ni-based catalysts are generally associated with metallic Ni species, the chemical states of Ni involved in related NiO/MgO and LaNiO₃ catalysts were accordingly resolved, respectively, through deconvolution of the peaks of Ni 2p and 3p on the fine-range spectra. The deconvolution results of Ni 2p_1/2 at 854.68 eV and Ni 2p_3/2 at 872.18 eV for the NiO/MgO catalyst in different states are illustrated in Figure 7c, showing that there is a slight variation of the proportions of Ni²⁺ and Ni³⁺ after the catalysts have experienced the reactions. This is not a surprise because the Ni valence in the NiS phase should be approximate to that in NiO. Nevertheless, the deconvolution results of Ni²⁺ at 66.38 eV and Ni³⁺ at 67.88 eV both assigned to Ni 3p of LaNiO₃ catalyst in Figure 7d unveil a significant difference in Ni chemical valance between the catalysts in pristine and in used states. The species of Ni³⁺ was overwhelming in the pristine LaNiO₃, which evolved partially to Ni²⁺ when the LaNiO₃ was subjected to the reactions in the presence of H₂S. The ratio of Ni²⁺ to Ni³⁺ in the used LaNiO₃-reduced catalyst is considerably comparable to that in the used NiO/MgO catalyst, in concert with the common formation of NiS in both catalysts. Therefore, the chemical state of Ni in the two types of catalysts during the reactions may be much the same.

It should be emphasized that there is no any active Ni species on the MgO catalyst. As discussed above, the catalytic performance of the MgO catalyst in the condition of H₂S presence is in no sense inferior to the two types of catalysts containing Ni species. This raises the question of what is the active center on the MgO catalyst for the reactions. In our earlier study, the catalytic activity of the MgO catalyst was also found to be competitive to that of the NiO/MgO and NiO/γ-Al₂O₃ catalysts during the reactions between CO₂ and H₂S at 800 °C, and consequentially free radicals initiated on the catalyst surface were supposed to dominate the reactions [20]. In recent research of Wang et al., the key active sites were identified as sulfur species (S*) that are dynamically bound to metal cations during the reactions between CH₄ and H₂S at high temperatures, which was believed to be common for all the catalysts tested, covering a wide range of oxides such as MgO, γ-Al₂O₃, and IUPAC group 4–6 metal oxide-derived materials [28,29]. Being inspired by these viewpoints, we speculate that the adsorbed S species, which are formed ineluctably through H₂S dissociation on the surface of catalyst regardless of catalyst type [28,29,30], may be the key to initiating the activation of CO₂ and CH₄ in the presence of H₂S. Schematic diagrams of the possible reaction mechanisms that may occur separately on MgO and NiO/MgO catalysts in the presence of H₂S are drawn as example and illustrated in Figure 8a,b. Regardless of whether MgO is loaded with Ni, its cations are supposed to be able to dissociate H₂S, thus promoting the conversion of CO₂ and CH₄.

Compared with the high catalytic activity of metallic Ni active sites for DRM reactions, however, the catalytic capability of S-adsorbed species may fall into a disadvantage. Due to the inferior activity of S-adsorbed species along with the vanishment of metallic Ni active sites, the reduction in CO₂ and CH₄ conversions is inevitable in the presence of H₂S. Despite all this, the addition of H₂S with a high concentration can undoubtedly promote these conversions by means of H₂S participating in the reactions with both CO₂ and CH₄. Moreover, the deviation in catalytic performance among the three types of catalysts, although weak, still could provide us with clues to improve catalyst. Developing a catalyst having a large specific surface area which can persist at high temperature to afford abundant active S-adsorbed species may be beneficial to the DRM reactions in the presence of H₂S.

3. Materials and Methods

3.1. Catalyst Preparation

(1) MgO: The powdery MgO of analytical purity (≥98.5%) purchased from Sinopharm Chemical Reagent Co. Ltd. in Shanghai (China) was put into a Muffle furnace and heated at a rate of 2 °C/min to 800 °C and maintained for 8 h. After the completion of calcination, the sample of MgO was naturally cooled to room temperature. Prior to being tested as a catalyst, the MgO sample needed to be cast in a mold, and the pellets formed were broken and sieved into granules in the size of 60–80 mesh.

(2) NiO/MgO: The sample of NiO/MgO was prepared by the incipient wetness impregnation method, in which NiO loading was set as 20 wt%. The precursor of NiO was Ni(NO₃)₂ (analytical purity, ≥98.0%, Sinopharm Chemical Reagent Co. Ltd. in Shanghai, China) solution and the powdery MgO obtained above was used as the support. Being aged for 4 h, the impregnated sample was dried in an oven at 110 °C for 6 h, and then calcined in a Muffle furnace at a rate of 2 °C/min to 800 °C and maintained for 4 h. After cooling down, the sample was also granulated. The pristine NiO/MgO catalyst having the size of 60–80 mesh was thus obtained to be ready for use. The real NiO loading of NiO/MgO prepared was detected to be 20.67 wt% (16.24 wt% Ni) by means of an inductively coupled plasma optical emission spectrometry (ICP-OES) using a Perkin Elmer Optima 4300 dual view model.

For the preparation of the reduced NiO/MgO catalyst, a sample of pristine NiO/MgO granules was put into a fixed-bed reactor and heated in a flow of 10 vol% H₂/N₂ mixture at a rate of 2 °C/min to 600 °C and kept for 4 h. The gas hourly space velocity (GHSV) during reduction was set around 10,000 h⁻¹. After that, the sample was cooled down and purged using N₂ for 2 h before being fetched out from the reactor.

(3) LaNiO₃: A citrate sol-gel method was adopted for the synthesis of LaNiO₃ according to the procedure descried in reference [31]. Briefly, nickel nitrate (Ni(NO₃)₂), lanthanum nitrate (La(NO₃)₃, analytical purity, La₂O₃ ≥ 44.0%, Sinopharm Chemical Reagent Co. Ltd. in Shanghai, China), and citric acid were used as the precursors and their molar ratio was fitted to be 1:1:3. The solutions of Ni(NO₃)₂ and La(NO₃)₃ formed separately were mixed firstly in a beaker, which was placed in a water bath at 80 °C. Along with continuously stirring the mixture, the citric acid solution was added into the beaker drop by drop. A green transparent colloid was formed afterwards, which gradually became a sticky gel under the constant stirring. The gel was dried in an oven at 120 °C for 24 h, and the dried gel after cooling was then crushed into pieces. For calcining the dried gel in the Muffle furnace, the temperature was programed to be 500 °C at a rate of 3 °C/min for 3 h, and then to 800 °C at a rate of 5 °C/min for 4 h. After cooling, the sample of LaNiO₃ was also molded and sieved. The reduction of LaNiO₃ was fulfilled following the same steps for NiO/MgO reduction recorded above.

3.2. Catalytic DRM Test and Reaction Gas Composition Analysis

The catalytic performance testing of a variety of catalysts used for the DRM process was executed in a fixed-bed tubular quartz reactor under atmospheric pressure in the condition of either the absence or presence of H₂S. The reactor had an inner diameter of 8 mm and a length of 300 mm, which was heated using an electrical furnace when carrying out the reactions. The flow chart of the apparatus for the tests is provided in Figure S5. In a typical procedure for reaction testing, 0.6 mL catalyst granules (60–80 mesh) were filled in the middle of the reactor. A gas flow of N₂ with high purity was purged through the reactor until the temperature of the reactor reached a set value. After that, a mixture gas composed of 20 vol% CO₂ and 20 vol% CH₄ balanced with N₂ at a total flow 100 mL/min was introduced to attain an assigned GHSV of 10,000 h⁻¹. The flow of each gas was adjusted by a Sevenstar mass flow controller (type D07, Beijing Sevenstar Electronic Technology Co. Ltd., Beijing, China).

The compositions of the mixture gas fore-and-aft the reactor were monitored constantly, including CO₂, CH₄, H₂S, CO, COS, CS₂, H₂O, H₂, and N₂. Two gas chromatographic (GC) instruments, both with thermal conductive detectors (Clause 680, PerkinElmer Instruments, Waltham, MA, USA), were used for routine composition analysis. GC-1 was mounted in series with PQ and 5A columns using H₂ as carrier gas to separate, in succession, CO₂, H₂O, H₂S, COS, N₂, CH₄, CO, and CS₂, the in case of all existing. A schematic diagram of different components flowing through two GC columns separately and the corresponding switching time of the six-way sampling valve is illustrated in Figure S6 for the convenience of understanding this unique GC analysis method. The GC peak order of all the components is displayed in Figure S7 for reference. Another GC (GC-2) was equipped with a 5A column and adopted N₂ as carrier gas for H₂ detection. An additional GC with flame photometric detector was ready for SO₂ detection, which was fixed with a PQs column using H₂ as the carrier gas. The elemental sulfur, when formed, could be captured downstream of the reactor onto the wall of a glass vessel immersed in an ice-bath. Liquid water could be also removed by this glass vessel. The tail gas was discharged at the end after passing through secondary ethanolamine washing bottles to absorb the toxic and harmful substances.

It needs to be pointed out that N₂ as an inert gas did not take part in any reactions and its flow rate entering and leaving the reactor was constant. What is more, the composition of N₂ could be determined through GC analysis. Therefore, N₂ was used as the standard substance for quantifying the concentration and flow rate of the other components contained in the reaction system. The content of various substances relative to N₂ was further utilized for calculating the mass balance of C, H, O, and S before and after the reaction. In the present study, good mass balances for C and S, both with less than 5% deviation, were acquired because of few elemental carbon and sulfur formations during reaction and, meanwhile, the smooth discharge of other carbon- and sulfur-contained products from the reactor outlet. For the H and O, however, probably due to the condensation of H₂O in the pipeline, there is a mass loss gap between before and after the reaction. Nevertheless, this does not hinder the quantification and calculation of changes in important components other than H₂O for catalytic performance evaluation.

The conversions of CO₂, CH₄, and H₂S, and the molar yields of CO, H₂, COS, H₂O, and CS₂ are defined, respectively, as follows:

$${\mathrm{CO}}_2 \mathrm{conversion}=\frac{F_{{\mathrm{CO}}_2,\mathrm{in}}-F_{{\mathrm{CO}}_2,\mathrm{out}}}{F_{{\mathrm{CO}}_2,\mathrm{in}}}\times 100\%$$

(6)

$${\mathrm{CH}}_4 \mathrm{conversion}=\frac{F_{{\mathrm{CH}}_4,\mathrm{in}}-F_{{\mathrm{CH}}_4,\mathrm{out}}}{F_{{\mathrm{CH}}_4,\mathrm{in}}}\times 100\%$$

(7)

$${\mathrm{H}}_2\mathrm{S} \mathrm{conversion}=\frac{F_{{\mathrm{H}}_2\mathrm{S},\mathrm{in}}-F_{{\mathrm{H}}_2\mathrm{S},\mathrm{out}}}{F_{{\mathrm{H}}_2\mathrm{S},\mathrm{in}}}\times 100\%$$

(8)

$$\mathrm{CO} \mathrm{molar} \mathrm{yield}=\frac{F_{\mathrm{CO},\mathrm{out}}}{F_{{\mathrm{CO}}_2,\mathrm{in}}+F_{{\mathrm{CH}}_4,\mathrm{in}}}\times 100\%$$

(9)

$${\mathrm{H}}_2 \mathrm{molar} \mathrm{yield}=\frac{F_{{\mathrm{H}}_2,\mathrm{out}}}{F_{{\mathrm{CH}}_4,\mathrm{in}}\times 2+F_{{\mathrm{H}}_2\mathrm{S},\mathrm{in}}}\times 100\%$$

(10)

$$\mathrm{COS} \mathrm{molar} \mathrm{yield}=\frac{F_{\mathrm{COS},\mathrm{out}}}{F_{{\mathrm{H}}_2\mathrm{S},\mathrm{in}}}\times 100\%$$

(11)

$${\mathrm{H}}_2\mathrm{O} \mathrm{molar} \mathrm{yield}=\frac{F_{{\mathrm{H}}_2\mathrm{O},\mathrm{out}}}{F_{{\mathrm{CO}}_2,\mathrm{in}}\times 2}\times 100\%$$

(12)

$${\mathrm{CS}}_2 \mathrm{molar} \mathrm{yield}=\frac{F_{{\mathrm{CS}}_2,\mathrm{out}}\times 2}{F_{{\mathrm{H}}_2\mathrm{S},\mathrm{in}}}\times 100\%$$

(13)

where F_i,in and F_i,out refer to the molar flow rate of the component of i at inlet and outlet, respectively, of the DRM reactor. Because the product of COS should come from the equimolar conversion of H₂S and CO₂, and as the molar flow rate of H₂S is smaller than that of CO₂ in the feedstock throughout, the COS molar yield is reduced on the basis of the H₂S feed flow rate, as is the CS₂ molar yield. Since the collection and quantification of H₂O generated with reaction time on stream is unachievable, the molar flow rate of H₂O is derived from the mass balance of the oxygen element between the inlet and outlet flow.

3.3. Catalyst Characterization

(1) N₂ cryo-physisorption measurement (BET): The textural measurement of the sample which had in advance been degassed in vacuum at 200 °C for 6 h was performed at −196 °C on a Micromeretics ASAP 2020 apparatus (Norcross, GA, USA) using N₂ as the adsorbate. Specific surface area of the sample was calculated with the Brunauer–Emmet–Teller (BET) equation, and pore volume between 1.7 and 300.0 nm in diameter was determined from the N₂ desorption isotherm using the Barrett–Joyner–Halenda (BJH) method.

(2) X-ray diffraction (XRD): The crystal phases of the catalysts were detected by means of XRD using a Rigaku Smartlab 9 X-ray diffractometer (Akishima, Japan) with an operating voltage of 40 kV and an operating current of 40 mA. The X-ray light source was a Cu target with a scanning range of 10–90° and a scanning speed of 10°/min. The diffraction peaks were identified by comparing the results of corresponding materials with the standard JCPDF cards.

(3) X-ray photoelectron spectroscopy (XPS): XPS was used to analyze the elemental valence information on the surface of the catalysts, which was carried out using an Escalab 250Xi photoelectron spectrometer from Thermo Scientific, Waltham, MA, USA. The main parameters were monochromatic, micro-focused Al-K-Alpha radiant light source (1486.6 eV), and analyzer flux energy of 20 eV. During the measurement, a full-spectrum scan of the sample was performed first to collect information, followed by a fine-spectrum scan for the target element. The obtained data were analyzed by Thermo Avantage v5.9921 software, the binding energy was corrected using the C 1s peak at 284.8 eV, and the background was subtracted by Shirley’s method.

4. Conclusions

The thermodynamic simulation results of the reaction mixture of 20 vol% CO₂, 20 vol% CH₄, and (0–20 vol%) H₂S in the temperature range of 400–1000 °C show that the DRM reaction with CO₂ has a significant thermodynamic advantage over the reactions involving H₂S. Meanwhile, the equilibrium molar yield of H₂ based on CH₄ changes little when increasing the H₂S concentration from 0 to 20 vol% at 800 °C.

For the catalysts of MgO and the pristine NiO/MgO, both having low catalytic activity for DRM at 800 °C, the addition of H₂S can significantly increase the conversions of CO₂ and CH₄ and the yields of CO and H₂. On NiO/MgO, adding 15 vol% H₂S can increase the conversion of CH₄ from 6.92% to 26.86% and CO₂ from 9.15% to 42.10%. To the contrary, for the reduced NiO/MgO and LaNiO₃ catalysts with high catalytic activity for DRM, adding a small amount of H₂S severely inhibits their catalytic performance. However, by further increasing the H₂S concentration to 15 vol%, the conversions of CO₂ and CH₄ and the yields of CO and H₂ can be partially recovered. The byproduct COS emerges in all the catalytic systems involving H₂S.

The results of process analysis and catalyst structural characterization reveal that a high H₂S concentration is promotive to the transformation of either CO₂ or CH₄ through the reactions of H₂S with CO₂ and CH₄. The active S-adsorbed species on the surface of catalysts, rather than the traditional active metallic Ni sites, have been speculated to be responsible for the moderate conversion of CO₂ and CH₄ in the presence of H₂S at high temperatures. The obtained results are of great significance for the direct catalytic conversion of sour natural gas to produce valuable chemicals.