Favoring the Methane Oxychlorination Reaction over EuOCl by Synergistic Effects with Lanthanum.

The direct conversion of CH4 into fuels and chemicals produces less waste, requires smaller capital investments, and has improved energy efficiency compared to multistep processes. While the methane oxychlorination (MOC) reaction has been given little attention, it offers the potential to achieve high CH4 conversion levels at high selectivities. In a continuing effort to design commercially interesting MOC catalysts, we have improved the catalyst design of EuOCl by the partial replacement of Eu3+ by La3+. A set of catalytic solid solutions of La3+ and Eu3+ (i.e., La x Eu1-x OCl, where x = 0, 0.25, 0.50, 0.75, and 1) were synthesized and tested in the MOC reaction. The La3+-Eu3+ catalysts exhibit an increased CH3Cl selectivity (i.e., 54-66 vs 41-52%), a lower CH2Cl2 selectivity (i.e., 8-24 vs 18-34%), and a comparable CO selectivity (i.e., 11-28 vs 14-28%) compared to EuOCl under the same reaction conditions and varying HCl concentrations in the feed. The La3+-Eu3+ catalysts possessed a higher CH4 conversion rate than when the individual activities of LaOCl and EuOCl are summed with a similar La3+/Eu3+ ratio (i.e., the linear combination). In the solid solution, La3+ is readily chlorinated and acts as a chlorine buffer that can transfer chlorine to the active Eu3+ phase, thereby enhancing the activity. The improved catalyst design enhances the CH3Cl yield and selectivity and reduces the catalyst cost and the separation cost of the unreacted HCl. These results showcase that, by matching intrinsic material properties, catalyst design can be altered to overcome reaction bottlenecks.

Introduction

CH4 is a relatively cheap and widely available natural resource, but it requires multistep processes to produce fuels and chemicals from it.[1] Single-step processes conceptually produce less waste, require smaller capital investments, and have improved energy efficiency.[2,3] However, practical considerations make that none of the direct methane conversion routes have seen industrialization so far.[2] The key challenges with direct conversion routes that need to be addressed, e.g., low conversion levels and/or poor selectivity, all require better catalyst design.[4,5] Of the direct conversion routes, methane oxyhalogenation (MOH) reaction has one of the highest potentials to see industrialization due to the moderate reaction temperatures and high conversion levels of CH4.[6] Moreover, a high selectivity toward the desired mono-halogenated methane CH3X (where X = Cl, Br, or I) can be achieved.[7,8] Being able to produce CH3X selectively in high quantities is of great interest. The chemical analogy between CH3OH and CH3X is remarkable[2,9−11] and makes mono-halogenated methane as valuable as methanol.[5,12,13] However, relatively little research has been performed on the MOH reaction.[6,12,14]

From the perspective of a circular economy approach, methane oxychlorination (MOC) has the additional advantage of being able to utilize HCl, a byproduct of other chlorination reactions.[15,16] However, the corrosive and oxidative environment under which the MOC catalysts must operate pose technological challenges and hinder the industrialization of the process.[6,17,18] A commercially interesting catalyst must be able to operate over prolonged times with high CH3Cl selectivity and CH4 conversion level.[19] Furthermore, the selectivity to CO needs to be minimized to make optimal use of the chemical feedstock and to lower separation costs.[14] These aforementioned requirements are challenging, and very little is known about how to fulfill these catalyst requirements.[20,21] Hence, more work is required to develop suitable MOC catalysts for commercial applications.

A number of catalyst compositions are published in the academic and patent literature, which can be divided into transition metal-based catalysts (e.g., TiO2,[8,22] VPO,[8,22] FePO4,[8] FeCl2/KCl,[23] ZrO2,[24] and Nb2O5[22]), noble metal-based catalysts (e.g., RuO2,[8,22] NM/MO[14] where NM= Ru, Rh, Pd, Ir, Pt, and MO = metal oxide support material), lanthanide-based catalysts (e.g., LaOCl,[25−27] CeO2,[8,12,22,28] and EuOCl[29]) and bimetallic catalysts (e.g., Cu/K/La,[8,30,31] FeO/CeO2,[28,32] LaVO4,[22] Ce/LaOCl,[17] Ni/LaOCl,[17] and Co/LaOCl[17]). None of these groups outperforms any of the other groups by definition, and only a handful of individual solid catalysts were studied in depth. A more fundamental approach to catalyst design needs to be adopted to understand the kinetic and thermodynamic bottlenecks encountered when operating certain catalyst materials.

We recently showed that EuOCl is a promising candidate for the MOC reaction as its performance is stable and, by varying the reaction temperature and feed mixture, also highly tunable.[33] EuOCl is suitable to be studied under working conditions with operando spectroscopy because of the Raman active modes of the material and the photoluminescent properties of Eu3+. Hence, we were able to conclude that the chlorination of the catalyst surface was rate limiting. While EuOCl outperformed the other lanthanides tested in our study, a number of improvements need to be made to the catalyst design to have a potential industrial catalyst: (i) improve CH3Cl selectivity (SCH), preferably at higher CH4 conversion levels (XCH); (ii) reduce catalyst cost by lowering the Eu3+ content in the catalyst; and (iii) lower the HCl concentration in the feed while still maintaining a high degree of surface chlorination. A large excess of HCl and unreacted feed are undesired as they result in high separation costs.

In this work, we explore the effect of the partial replacement of Eu3+ by La3+ on the catalytic performance in the MOC and investigate the apparent synergistic effect between La3+ and Eu3+. Operando Raman spectroscopy previously revealed that the chlorination of EuOCl to EuCl3 is a slow process and can be rate limiting during the MOC reaction.[33] Based on thermodynamic calculations and experimental evidence, LaOCl was selected as a chlorine reservoir for Eu3+ as the chlorination from LaOCl to LaCl3 occurs readily at low HCl concentrations. La1–EuOCl (where x = 0, 0.25, 0.50, 0.75, or 1) solid solution catalysts were synthesized and characterized. Incorporation of La3+ into EuOCl crystal lattice was favored, since La3+ has the same oxidation state and a comparable ionic radius to Eu3+. The performance of La1–EuOCl materials in the MOC reaction was tested and compared to the benchmark EuOCl. The addition of La3+ improved the degree of chlorination of the catalyst, thereby improving the CH3Cl yield while preserving the excellent CO selectivity compared to monometallic EuOCl. Furthermore, operando luminescence spectroscopy was applied to provide further insight into the chlorination behavior of La3+–Eu3+ solid solutions. Lastly, physical mixtures of LaOCl and EuOCl were used as catalytic material, showcasing the importance of intimate contact between La3+ and Eu3+ in the MOC reaction. This resulted in the enhancement of the catalytic performance, approaching the performance of the La3+–Eu3+ solid solution. Hence, we showcase that, by matching intrinsic material properties, catalyst design can be altered to overcome reaction bottlenecks.

Experimental Methods

Catalyst Synthesis

The La1–EuOCl (where x = 0, 0.25, 0.5, 0.75, or 1) catalyst materials under study were prepared by dissolving lanthanum(III) chloride hydrate (LaCl3·xH2O, Alfa Aesar, >99.9%) and/or europium(III) chloride hydrate (EuCl3·xH2O, Alfa Aesar, >99.9%) in ethanol (absolute, VWR), followed by precipitation using stoichiometric amounts of ammonium hydroxide (Fisher Scientific, 25% in H2O) at room temperature. After the dropwise addition, the precipitates were stirred for an additional hour and subsequently centrifuged to obtain the gel. Then, the obtained gel was washed with ethanol (absolute, VWR) and dried at 80 °C in air. Lastly, the dried solids were calcined in a static oven at 500 °C for 3 h using a ramp rate of 5 °C/min.

Catalyst Characterization

X-ray diffraction (XRD) patterns were obtained with a Bruker-AXS D8 powder X-ray diffractometer in Bragg–Brentano geometry, using Cu Kα1,2 = 1.54184 Å, operated at 40 kV. The measurements were carried out between 22 and 65° using a step size of 0.02° and a scan speed of 1 s, with a 2 mm slit for the source. N2 adsorption isotherms were measured at −196 °C on a Micromeritics TriStar II Plus instrument. Prior to all measurements, samples were dried at 300 °C in a flow of N2. Specific surface areas were calculated using the multipoint Brunauer–Emmett–Teller (BET) method (0.05 < p/p0 < 0.25). Pore volumes were calculated by the t-plot method; pore size distributions were obtained by the Barrett–Joyner–Halenda (BJH) analysis; Harkins and Jura thickness model was applied for the t-plot and BJH methods.

Inductively coupled plasma-optical emission spectroscopy (ICP-OES) was applied to determine the chemical composition of the catalyst materials, using a SPECTRO CIROSCCD instrument. ICP-OES samples were prepared by destructing the solids in aqua regia.

Operando spectroscopy determination of the qualitative EuOCl/EuCl3 signal ratio by luminescence spectroscopy was performed with an AvaRaman-532 Hero-Evo instrument (λ = 532 nm, laser output 50 mW, spectral resolution of 10 cm–1) equipped with an AvaRaman-PRB-FC-532 probe, capable of withstanding temperatures up to 500 °C. Spectra were collected every minute with the AvaSoft 8 software. The data were subsequently dark corrected. The initial signal was optimized to obtain at least 50% of the saturation value.

Catalyst Testing

All of the catalytic tests and operando spectroscopy characterization experiments were performed in a lab-scale continuous-flow fixed-bed reactor quartz reactor. Details on the experimental setup as well as definitions and calculations are reported elsewhere.[33]

Methane oxychlorination reaction: 500 mg of catalyst material (125–425 μm sieve fraction) was loaded in a quartz reactor and heated to 450 °C under N2 with a 10 °C/min heating rate. The catalyst was activated in 20% HCl/N2 for 2 h prior to catalysis. For the isothermal experiments, the reaction temperature was adjusted to reach XCH = 10% for CH4/HCl/O2/N2/He of 2:2:1:1:14. When a stable conversion was reached, the HCl/He ratio was adjusted so that the HCl concentration was increased to 20, 40, 60, and 80 vol % while keeping a constant flow of 20 mL/min. For the ramp experiments, the reactor was brought to 350 °C and the desired feed mixture (i.e., CH4/HCl/O2/N2/He of 2:2:1:1:14 or 2:16:1:1:0 in mL/min) was fed into the reactor. A stabilization period of 30 min was applied, and then the ramp experiment of 1 °C/min was commenced to 550 °C. For the stability tests, the reactor was brought to 450 °C and CH4/HCl/O2/N2/He of 2:2:1:1:14 was fed into the reactor for 4 h. Subsequently, the HCl concentration was increased to 20, 40, 60, and 80 vol % while keeping a constant flow of 20 mL/min. Every HCl concentration was fed for 2 h. To characterize the spent catalysts, the catalysts were dechlorinated at 550 °C for 5 h under CH4/HCl/O2/N2/He of 2:0:4:1:13. The background of this dechlorination step is provided in SI Section S2. For the determination of the apparent activation energy, 250 mg of catalyst (125–425 μm sieve fraction) was loaded in a quartz reactor to 350 °C under N2 with a 10 °C/min heating rate. The catalyst was subjected to CH4/HCl/O2/N2/He of 2:2:1:1:14 (in mL/min) for 1 h. The temperature was increased to 550 °C with increments of 10 °C with a heating rate of 5 °C/min and kept at every temperature step for 45 min to obtain the steady-state activity. Only the data points where the methane conversion level was below 10% were considered for fitting the apparent activation energy to avoid heat and mass transfer limitations.

HCl oxidation: 500 mg of catalyst material (125–425 μm sieve fraction) was loaded in a quartz reactor and heated to 450 °C under N2 with 10 °C/min. The catalyst was activated in 20% HCl/N2 for 2 h prior to catalysis. Temperature-ramp experiments were performed from 350 to 550 °C at a ramp rate of 1 °C/min under the desired feed mixture (i.e., CH4/HCl/O2/N2/He of 0:2:1:1:16 or 0:16:1:1:2 in mL/min).

Results and Discussion

Catalyst Properties

The synthesized La1–EuOCl catalyst material was characterized by N2 physisorption, inductively coupled plasma-optical emission spectroscopy (ICP-OES), and X-ray diffraction (XRD) to gain insights into their physicochemical properties (Table ). The applied base precipitation method yielded catalyst materials with specific surface area (SBET) and pore volume (Vpore) of the same order of magnitude. The SBET ranges between 24.4 and 41.5 m2/g, while the Vpore ranges between 0.06 and 0.23 cm3/g. Furthermore, the experimental La3+/Eu3+ molar ratio obtained from ICP-OES after the precipitation of the bimetallic catalysts is in good agreement with the desired theoretical ratio (Table ).

Table 1

Physicochemical Properties of the As-Synthesized LaOCl, La0.75Eu0.25OCl, La0.50Eu0.50OCl, La0.25Eu0.75OCl, and EuOCla

	physisorption results			phase 1 (La3+ rich)			phase 2 (Eu3+ rich)
catalyst material LnOCl where Ln =	SBET (m2/g)	Vpore (cm3/g)	La3+/Eu3+ molar ratio (ICP-OES)	position (deg)	La3+/Eu3+	relative area (%)	position (deg)	La3+/Eu3+	relative area (%)
La	24.4	0.06		30.62
La0.75Eu0.25	39.6	0.22	74:26	30.80	86:14	54	31.02	68.1:31.9 ±1.2	46
La0.50Eu0.50	41.1	0.18	50:50	30.88	79:21	47	31.42	34.5:65.5 ± 1.3	53
La0.25Eu0.75	41.5	0.16	24:76	30.99	70:30	21	31.69	16.0:84.0 ±1.7	79
Eu	37.4	0.23					31.91

Specific surface area (SBET) and pore volume (Vpore) were derived based on N2 physisorption results. The La3+/Eu3+ ratios obtained from inductively coupled plasma-optical emission spectroscopy (ICP-OES) corresponded well with the theoretical values. Positions of the deconvoluted (110) X-ray diffraction (XRD) peak, the corresponding La3+/Eu3+ ratio, and relative area as calculated with Vegard’s Law for as-synthesized La0.75Eu0.25OCl, La0.50Eu0.50OCl, and La0.25Eu0.75OCl are also tabulated.

While ICP-OES provided the elemental ratio of the bulk materials, it did not provide information on the distribution of the two elements throughout the material. XRD was applied to investigate if the desired oxychloride phase was obtained, and if solid solutions of La3+ and Eu3+ were obtained. The XRD patterns of the as-synthesized catalyst materials are given in Figure A. As previously reported, LaOCl and EuOCl are easily synthesized in the oxychloride phase without any noticeable contaminations from other crystalline phases.[33] Since LaOCl and EuOCl have the same space group, P4/nmm, and comparable ionic radii,[34] solid-state ion mixing of the two elements is expected to occur.[35,36] By deconvolution of the (110) XRD peaks of the as-synthesized LnOCl catalysts (Figure B–F) and applying Vegard’s law (see SI Section S1B for more details on the applied procedure), at least two mixed phases were distinguished with varying La3+/Eu3+ ratios, referred to as phase 1 and phase 2 (Table ). Noticeable is that for every bimetallic La3+–Eu3+ catalyst, we obtained one La3+-rich phase (x > 70%, referred to as phase 1) and one phase with a larger distribution in the La3+–Eu3+ ratio (phase 2). We hypothesize that LaOCl is precipitated at a higher rate than EuOCl during the synthesis, thereby always obtaining one La3+-rich phase. The synthesized catalysts, with known molar ratios and comparable SBET and Vpore, enabled us to investigate the role of the La3+/Eu3+ ratio in the MOC reaction.

Figure 1

(A) X-ray diffraction (XRD) patterns of the as-synthesized LnOCl catalyst materials under study, including LaOCl, La0.75Eu0.25OCl, La0.50Eu0.50OCl, La0.25Eu0.75OCl, and EuOCl and LaOCl reference pattern (ICDD 00-00800477). (B–F) Zoom-in of the (110) XRD peaks displays the fitted peaks used for determining the degree of La3+–Eu3+ mixing in Table according to Vegard’s law (see SI Section 1B for the applied procedure).

Catalytic Performances

Temperature-ramp experiments under MOC reaction conditions were performed to study the catalytic activity trends of the bimetallic La3+–Eu3+ catalysts. An overview of the catalytic performance of the La3+–Eu3+ catalysts is given in Figure . Individual activity and selectivity versus reaction temperature plots are given in Figure S1. The catalytic performance of pure LaOCl and EuOCl are described elsewhere,[33] but the plots are given for facile comparison. The reaction temperature at which the catalyst becomes active, referred to as the onset temperature, is determined as the reaction temperature at which the XCH > 2%.

Figure 2

Methane oxychlorination (MOC) experiments for the synthesized La3+–Eu3+ catalysts. (A) Methane conversion (XCH) plotted versus the reaction temperature for LaOCl, La0.75Eu0.25OCl, La0.50Eu0.50OCl, La0.25Eu0.75OCl, and EuOCl at 10% HCl. The derivative of the XCH versus reaction temperature is plotted in (B). Yields of (C) CH3Cl, (D) CH2Cl2, and (E) CO are plotted versus the reaction temperature at 10% HCl. The CH4 conversion rate normalized to the amount of the catalyst is given in (F). Lastly, the rate difference with respect to the linear combination of LaOCl and EuOCl with the same La3+/Eu3+ ratio is given in (G). Reaction conditions: CH4/HCl/O2/N2/He of 2:2:1:1:14 (10% HCl, in mL/min) or 2:16:1:1:0 (80% HCl, in mL/min), 350–550 °C with a ramp rate of 1 °C/min. The temperature-dependent XCH over LaOCl and EuOCl is obtained from ref (33).

The La3+–Eu3+ catalysts showed many resemblances with respect to each other in terms of catalytic performance as the same qualitative trends could be observed. In general, the bimetallic catalysts showed a steady increase in the XCH up to ∼450 °C, after which the XCH curve leveled off (Figure B). With increasing Eu3+ content in the catalyst, the flattening of the XCH curve was not only more pronounced but also started at a higher reaction temperature, and thus a higher overall activity was obtained. Also, in terms of the product yield, the same qualitative trends were observed. YCH reached a maximum at a reaction temperature between 450 and 475 °C, and CH3Cl is the dominant product below 500 °C (Figure C). YCH was overall quite low, with a maximum yield of ∼3% at 480 °C (Figure D). Lastly, YCO increased steadily over the entire reaction temperature range, reaching its maximum value at 550 °C (Figure E). CH3Cl and CCl4 were detected in minor quantities, with selectivities <3% (Figure S2). No CO2 was detected under these reaction conditions.

The bimetallic catalysts showed different catalytic performances compared to their monometallic counterparts. The most striking difference is that the XCH of the bimetallic catalysts levels off above 500 °C, while a large increase in XCH is observed for both LaOCl and EuOCl (Figure A). Furthermore, the observed XCH drop for EuOCl, attributed to the dechlorination of EuCl3 to EuOCl, is not present when a solid solution is formed between La3+ and Eu3+ (Figure B). Interestingly, the highest YCH of all catalysts was obtained for La0.50Eu0.50OCl and La0.25Eu0.75OCl, reaching a maximum value of 11% at 460 °C. This was significantly higher than the 8% YCH of EuOCl at the same reaction temperature. This difference was caused by the lower YCH for the La3+–Eu3+ catalyst compared to EuOCl, as XCH and YCO were similar. One additional advantage of using the bimetallic La3+–Eu3+ catalysts was that no CO2 was detected over the entire tested range, unlike with other catalysts reported in the literature.[8,14,32]

The most balanced performance was observed for La0.50Eu0.50OCl. The observed XCH, YCH, YCH, and YCO were similar to La0.25Eu0.75 and significantly improved compared to La0.75Eu0.25OCl. This is visualized by normalizing the CH4 conversion rate at 480 °C to the amount of catalyst (Figure F). A clear trend between the Eu3+ content in the catalyst material and the obtained conversion rate is apparent when the activity is normalized to the amount of catalyst and SBET. The following activity ranking was obtained: EuOCl > La0.25Eu0.75OCl ∼ La0.50Eu0.50OCl ≫ La0.75Eu0.25OCl ≫ LaOCl. Large increments in conversion rates were observed going from LaOCl to La0.75Eu0.25OCl and to La0.50Eu0.50OCl, while the CH4 conversion rate increments decreased going from La0.50Eu0.50OCl to EuOCl. Conversely, when the observed activity was corrected for the activity of the linear combination of LaOCl and EuOCl, a synergistic effect between La3+ and Eu3+ was observed (Figure G). The addition of La3+ to EuOCl enhanced the activity of Eu3+ as all of the La3+–Eu3+ catalysts possessed a higher conversion rate than when the individual activities of LaOCl and EuOCl are summed with a similar La3+/Eu3+ ratio (i.e., the linear combination). An optimum was found when an equal amount of La3+ and Eu3+ was present, as the observed rate difference was the largest. Since monometallic LaOCl showed little activity at this reaction temperature by itself, we hypothesize that LaOCl acts as a chlorine buffer, supplying chlorine to the active Eu3+ phase. This effect is caused by the facile chlorination of LaOCl, which increases the degree of chlorination of the catalyst material and hence the activity. The role of La3+ and Eu3+ is further discussed in Section . Nevertheless, the observed selectivities for the bimetallic catalysts were not significantly influenced by the catalyst composition (Figure S2). The SCH lied between 53 and 60% for the bimetallic catalysts, which is much better than the SCH of 40% obtained for EuOCl. The SCO in all cases is ∼28% and seems to be governed by the reaction conditions and not by the catalyst composition.

The results presented in Figure show that La3+ had a major influence on the activity and selectivity in the MOC reaction. Previously, we applied higher HCl concentrations, i.e., 10–80% HCl in the feed, to boost the catalytic performance of EuOCl.[33] The catalytic destruction of chloromethanes was circumvented by the high degree of surface chlorination, resulting in improved product selectivity.[9,10,37−41] With the incorporation of La, similar functionality is incorporated into the catalyst design, and the question arises whether an increment in the HCl concentration is still needed to boost the catalytic performance of La3+–Eu3+ solid solution catalysts. To investigate the effect of HCl concentration on the La3+–Eu3+ solid solution catalysts, the reaction temperature was adjusted to obtain XCH = 10% after which the HCl concentration in the feed was increased. The XCH and SCH are plotted versus the HCl concentration in Figure A,B, respectively. The SCH and SCO are plotted versus the HCl concentration in Figure S3A,B, respectively. All Eu-containing catalysts were still positively influenced in terms of XCH by the increment in HCl concentration. A clear trend in the activity profile was observable going from LaOCl to EuOCl. With increasing Eu3+ concentration in the catalyst materials, XCH is also proportionally more influenced by the increase in HCl concentration. The reaction selectivity was not influenced drastically by the change in HCl concentration. In general, very small distinctions in terms of selectivity are found comparing the La3+–Eu3+ catalysts. The La3+–Eu3+ catalysts follow the same qualitative trend as Eu; only the quantitative performance is more suited for commercial application. Compared to EuOCl, the La3+–Eu3+ catalysts have an increased SCH (i.e., 54–66 vs 41–52%), lower SCH (i.e., 8–24 vs 18–34%), and comparable SCO (i.e., 11–28 vs 14–28%).

Figure 3

(A) CH4 conversion (XCH) and (B) selectivity toward CH3Cl (SCH) versus the HCl concentration for LaOCl (T = 520 °C), La0.75Eu0.25OCl (T = 475 °C), La0.50Eu0.50OCl (T = 450 °C), La0.25Eu0.75OCl (T = 450 °C), and EuOCl (T = 450 °C) in the methane oxychlorination (MOC) reaction. The La3+–Eu3+ catalyst materials all show increasing XCH with increasing HCl concentration. SCH is higher compared to LaOCl and EuOCl over the entire HCl concentration range tested. The temperature was adjusted to reach XCH = 10% for CH4/HCl/O2/N2/He of 2:2:1:1:14. When the stable conversion was reached, the HCl/He ratio was adjusted so that the HCl concentration was increased to 20, 40, 60, and 80% while keeping a constant flow of 20 mL/min.

To truly compare the catalytic performance of the catalyst material under study, the nonisothermal conversion–selectivity relation was given plotted toward CH3Cl and CO (Figure S4). In general, LaEu1–OCl catalyst materials performed significantly better compared to EuOCl at 10% HCl concentrations as SCH (Figure S4A) and SCO (Figure S4B) were drastically improved at the same conversion level. For example, at XCH = 10%, the SCH and SCO of EuOCl were 54 and 25% while for La0.50Eu0.50OCl, values of 74 and 17% were obtained. Only at high conversion levels (XCH > 20%), the EuOCl catalyst performed better than the LaEu1–OCl catalyst materials, with the important caveat that SCH became too low for practical applications. In the extreme case where the HCl concentration was increased to 80%, the performance of the LaEu1–OCl catalyst materials was still superior to the performance of EuOCl in terms of SCH (Figure S4C), while the SCO (Figure S4D) were fairly comparable. Here, the La0.25Eu0.75OCl catalyst performed slightly better than the other LaEu1–OCl catalyst materials with an SCH and SCO of 74 and 8% at XCH = 10%. At the same conversion level, the SCH and SCO of EuOCl were 56 and 6%, respectively. The main difference in product selectivity at 80% HCl concentration is that CH3Cl is not further chlorinated to higher chloromethanes for the LaEu1–OCl catalyst.

The catalytic performance of La0.50Eu0.50OCl was put in perspective to showcase its excellent performance compared to the catalytic systems reported in literature. For the benchmark catalysts reported in literature, SCH was plotted versus the T at which the XCH reached 10%, and the reaction rate is also provided (Figure S5). The exact values of the performance of the catalytic systems are tabulated in Table S1. While many catalytic systems show an SCH above 70% at XCH = 10% (Figure S5A), a large portion of these catalytic systems are not stable or were not tested for their stability. To comply with the stability criterium, only the catalysts reported as stable in terms of chemical, structural, and catalytic stability are considered (Figure S5B). Now, only a few catalytic systems show SCH above 70% at XCH = 10%, making La0.50Eu0.50OCl a benchmark catalyst. Lastly, the activity was normalized to the volume of the catalyst bed (Figure S5C), evidencing that the La0.50Eu0.50OCl catalyst is more reactive per unit volume than other catalyst materials reported in the literature.

Lastly, the change in the chemical composition of the catalyst material may alter the reaction mechanism that is responsible for the chlorination of CH4. Gas-phase chlorination via tandem reactions, HCl oxidation, and free radical chlorination is in competition with the surface-driven MOC reaction. To investigate the contribution of the gas-phase chlorination to the observed activity, the HCl oxidation performance of La0.50Eu0.50OCl was tested. The oxygen conversion (XO) of the HCl oxidation was compared to the XO of the MOC reaction under 10 and 80% HCl in the feed in Figure A,B, respectively. For facile comparison, the same plots are given for EuOCl obtained from ref (33) in Figure C,D, respectively. At 10% HCl, XO for La0.50Eu0.50OCl increased to a reaction temperature of 500 °C, after which it stabilized at the final XO value of ∼20%. This was significantly less than the XO during the MOC reaction, which gradually increased to a final XO value of ∼62%. A discrepancy between the XO of the HCl oxidation and MOC was already observed from 405 °C onwards, evidencing that the surface-driven CH4 chlorination is the dominant pathway during MOC at 10% HCl. When the HCl concentration was increased to 80% HCl, thereby also increasing the activity of the catalyst material in the MOC, a steeper increase in the XO was observed for the HCl oxidation, which gradually increased up to a final XO value of ∼53% at 550 °C. The XO was significantly higher when the HCl concentration was increased, and the thermal chlorination had a larger contribution to the overall activity. These trends in both HCl oxidation and MOC match well with the trends observed for monometallic EuOCl. The addition of La3+ does not influence the HCl oxidation capability of EuOCl qualitatively.

Figure 4

Temperature-ramp experiments where the oxygen conversion (XO) is plotted versus the reaction temperature for the HCl oxidation reaction (filled squares) and methane oxychlorination (MOC) reaction (open circles) for (A) La0.5Eu0.5OCl 10% HCl, (B) La0.5Eu0.5OCl 80% HCl, (C) EuOCl 10% HCl, and (D) EuOCl 80% HCl in the feed. Reaction conditions for the HCl oxidation: CH4/HCl/O2/N2/He of 0:2:1:1:16 (10% HCl, in mL/min) or 0:16:1:1:2 (80% HCl, in mL/min), 350–550 °C with a ramp rate of 1 °C/min. Reaction conditions for the oxychlorination: CH4/HCl/O2/N2/He of 2:2:1:1:14 (10% HCl, in mL/min) or 2:16:1:1:0 (80% HCl, in mL/min), 350–550 °C with a ramp rate of 1 °C/min. The temperature-dependent XO over EuOCl was obtained from ref (33).

Understanding the Working Mechanism

The catalytic performance of La3+–Eu3+ solid catalysts showed clear synergetic behavior when compared to either LaOCl or EuOCl. The premise of making La3+–Eu3+ solid solutions was to improve the chlorination rate of EuOCl, as this chlorination step was found to be rate limiting.[33] High HCl concentrations in the feed were needed to boost the activity of EuOCl, which is unfavorable in terms of product separation and size of recycle streams. The chlorination and dechlorination behavior of La3+ was studied, and we observed that La3+ was readily chlorinated to LaCl3. Thermodynamic calculations are consistent with this observation, as the chlorination of LnOCl (Ln = lanthanide) to LnCl3 is the most facile for LaOCl (Figure S6). Thus, LaOCl most probably functions as a chlorine acceptor/capacitator for the active EuOCl. However, the harsh reaction conditions under which these solid catalysts operate cause many changes in the physicochemical properties over time, and the intimate contact between La3+ and Eu3+ could be lost. The loss of intimate contact between La3+ and Eu3+ implies that the exchange of ions between La3+ and Eu3+ is made more difficult, thereby losing the synergistic effect. Hence, catalyst stability could pose an issue.

To analyze whether further phase segregation occurs over time, La0.50Eu0.50OCl was subjected to MOC conditions for 1, 2, 4, 8, and 16 h, and the postcharacterization results of the chemical composition and structure are presented in Figure . Additional transmission electron microscopy (TEM) images of the time series are given in Figure S7 to visualize the morphological changes. Aggregation of particles is visible with increasing time on stream (TOS); however, the dechlorinated catalyst might be altered morphologically, see SI Section S2. The as-synthesized La0.50Eu0.50OCl displayed two XRD peaks in the region where the (110) lies (Figure A), both consisted of La3+ and Eu3+ (Figure B). Over time, the La3+-rich phase starts to move to lower angles, indicating the further enrichment of this phase with La3+. The Eu3+-rich phase, however, does not change in chemical composition (±2% over the entire duration). Simultaneous to the segregation is the change in relative peak area where the La3+-rich phase gained in relative peak area. The largest differences were observed in the first 8 h, where the La3+/Eu3+ ratio of the La3+-rich phase changed from 61:39 to 80:20. After 16 h TOS, the La3+/Eu3+ ratio reached 83:17 for the La3+-rich phase.

Figure 5

Time series of La0.50Eu0.50OCl exposed to methane oxychlorination (MOC) conditions to study the phase segregation behavior of La3+–Eu3+ solid solutions. La0.50Eu0.50OCl was tested for 1, 2, 4, 8, and 16 h time on stream (TOS) at 450 °C, and the catalyst material was characterized with X-ray diffraction (XRD). The fresh catalyst was loaded into the reactor for every measurement. (A) Zoom-in of the (110) XRD peaks displays phase segregation over time. The obtained (B) La3+/Eu3+ ratio and (C) relative area of Fit 1 and Fit 2 indicate that the phase segregation predominantly occurs within the first 8 h of reaction. A schematic representation of the phase segregation is depicted in (D), where the La3+-rich phase starts to increase in La3+ concentration and relative amount.

The observed phase segregation suggests that total phase segregation could occur over prolonged reaction times or harsher reaction conditions, thereby losing the intimate contact between La3+ and Eu3+. It is unclear if the segregation of these two phases would result in the loss of the synergistic effect between La3+ and Eu3+. Therefore, to investigate whether this synergistic effect between La3+ and Eu3+ also exists when the two phases are completely segregated, two physical mixtures of LaOCl and EuOCl were prepared and tested under the same reaction conditions as La0.50Eu0.50OCl. Physical mixture 1 (PM1) was prepared by sonicating a mixture of LaOCl and EuOCl nanopowders in ethanol, after which the solvent was evaporated and the powder mixture was sieved (125–425 μm size fraction). Intimate mixing of the powders was achieved, but no solid solution was formed. Physical mixture 2 (PM2) was prepared by mixing sieved LaOCl and EuOCl particles (125–425 μm size fraction); hence, no intimate contact is expected. PM1 and PM2 were tested by performing temperature-ramp experiments under 10% HCl and post characterized with XRD. The XCH, YCH, and the (110) XRD peak of PM1 are presented in Figure A–C, respectively, and compared to La0.50Eu0.50OCl. The same plots as for PM1 were made for PM2 and presented in Figure D–F, respectively. A comparison between PM2 and the linear combination of LaOCl and EuOCl is made.

Figure 6

Catalytic performance of PM1 compared to La0.50Eu0.50OCl and PM2 compared to the linear combination of LaOCl and EuOCl. The (A) XCH, (B) YCH, and (C) analysis of the (110) X-ray diffraction (XRD) peak of PM1 indicate that the performance of PM1 is very comparable to La0.50Eu0.50OCl if the intimate contact between LaOCl and EuOCl is established. The (D) XCH, (E) YCH, and (F) analysis of the (110) XRD peak of PM2 reveal that similar performance to the linear combination of LaOCl and EuOCl is obtained when no intimate contact is established. Reaction conditions: CH4/HCl/O2/N2/He of 2:2:1:1:14 (10% HCl, in mL/min), 350–550 °C with a ramp rate of 1 °C/min.

A clear distinction between the observed performance of PM1 and PM2 was apparent. When an intimate contact was achieved, thus in the case of PM1, XCH and YCH much resemble the same trend as observed for La0.50Eu0.50OCl. Even though some quantitative differences exist, and the overall performance is slightly lower, an enhancement of the activity compared to the linear combination was present (Figure S8). The drop in activity, unique for EuOCl, was not observed, indicating that an intimate contact is established between La3+ and Eu3+. Surprisingly, mixing of Eu3+ in the La3+-rich phase occurred, indicated by the shift to higher angles for the La3+-rich phase. The La3+/Eu3+ ratio changed from 100:0 to 88:12. No La3+ was incorporated in the EuOCl crystal structure, but migration of Eu3+ into LaOCl occurred, possibly because of the higher thermodynamic stability of such phase. The enhancement of activity and mixing of phases did not occur in PM2, when no intimate contact between La3+ and Eu3+ was present. The activity profile and selectivity of PM2 much resembled a linear combination of the activity of monometallic LaOCl and EuOCl. The drop in activity does occur for this catalyst, which is characteristic of monometallic EuOCl. Furthermore, XRD patterns reveal that no mixing of Eu3+ and La3+ occurred at these reaction conditions and reaction times. The premise of mixing La3+ and Eu3+ was to accelerate the chlorination rate of the catalyst material, and hence the activity of Eu, by incorporating a chlorine accepting element in the material. At this point, we observed a synergistic effect between La3+ and Eu3+ and established the fact that the intimate contact between La3+ and Eu3+ responsible for this synergistic effect will be preserved. However, it is yet unclear what the mechanism behind this synergistic effect is. Furthermore, during the reaction, a La3+-rich oxychloride phase with minor amounts of Eu3+ and a (almost) pure EuOCl phase was obtained. To unravel the active phase, we looked at the chlorination behavior of Eu3+ in different Eu-containing catalysts.

Structural information, combined with the observed activity in the MOC reaction, provides crucial insight into the working mechanism of these MOC catalyst materials. According to our understanding, the oxychlorination reaction consists of two noncatalytic reactions combined to form a catalytic cycle: the chlorination of lanthanide oxychloride (eq ) and the dechlorination of lanthanide chloride (eq )Many more reactions occur in the complex methane oxychlorination reaction, as, e.g., the dechlorination can also occur via the reaction with H2O.[40,41] For simplicity reasons, the two reaction equations that make up the standard oxychlorination reaction to methyl chloride are given as the main point is the concept of catalyst chlorination and dechlorination. From eqs and 2, it becomes apparent that the state of the catalyst, or the degree of catalyst chlorination, is controlled by . By altering the feed composition, either k1 or k2 is directly influenced, which is represented by a change in catalytic performance.

The structural information was obtained with operando luminescence spectroscopy. The area of the Eu3+ luminescence signal was used as a measure for the degree of Eu3+ chlorination in previous research.[33] Since EuCl3 shows no luminescence, the decrease in luminescence intensity can be correlated with the degree of chlorination. The Eu3+ luminescence spectra of La0.50Eu0.50OCl and PM1 showed the same emissions as Eu3+ in EuOCl and responded in the same manner to a change in degree of chlorination (Figure A). Thus, the same analysis can be performed to show the qualitative trends in the degree of chlorination of Eu3+ in La3+–Eu3+ catalyst materials.

Figure 7

(A) Photoluminescence spectra of La0.50Eu0.50OCl, PM1, and EuOCl corresponding to the runtimes in (B) show the same behavior to the response in degree of chlorination as observed for EuOCl. The only change appeared in the spectral intensity and not in the shape of the spectrum. The applied integrated spectral area is graphically depicted for EuOCl by the blue area. (B) Relative spectral area of the Eu3+ luminescence signal observed during methane oxychlorination (MOC) reaction under varying reaction conditions at 450 °C and (C) corresponding XCH plotted versus time on stream (TOS). The incorporation of La3+ caused a faster chlorination of Eu3+. Reaction conditions: CH4/HCl/O2/N2/He of 2:2:1:1:14 (10% HCl, in mL/min), at 450 °C. Subsequently, the HCl/He ratio was altered to obtain 20, 40, 60, and 80 vol % HCl while keeping a constant flow of 20 mL/min.

When considering EuOCl, very high HCl concentrations and prolonged reaction times were needed to convert EuOCl into EuCl3. The relative spectral area of the Eu3+ luminescence signal (Figure B) and the XCH (Figure C) are plotted versus the time on stream (TOS), where the HCl concentration in the feed is gradually increased. Here, the first signs of catalyst chlorination started after 10 h and reached their final state after 12 h. The XCH gradually increased up to 60% HCl, and a steady downward trend in the XCH of EuOCl was visible when the final HCl concentration of 80% was fed, which coincides with previously reported observations that full chlorination deactivates the catalyst material. For EuOCl, only at these very high HCl concentrations, the > 1, combined with the fact that the activity correlated with the HCl concentration, indicated that the chlorination of the EuOCl surface is the rate-determining step (RDS). Any chlorine present on the surface had reacted before it could diffuse to the bulk; hence, no phase change was observed. If the surface chlorination would not be rate limiting, increasing the HCl concentration would not result in an increase in the activity.

We applied the same principle for La3+–Eu3+ catalysts to show that La3+ addition heavily affects the rate of EuOCl chlorination and thus the rate-determining step. When La3+ was in close proximity to Eu3+, more facile catalyst chlorination was observed. The highest chlorination rate was observed for La0.50Eu0.50OCl, as the integrated spectral area already shows a decreasing trend with 10% HCl in MOC reaction conditions. Right from the start, > 1. This is remarkable, as EuOCl was proven to be difficult to chlorinate under these conditions. The chlorination continued with an increasing rate when the HCl concentration was further increased up to 8 h, where it reached its final state. Complete chlorination was achieved, as no emissions from EuOCl could be detected anymore. Interestingly, up to 8 h, XCH increased from 9 to 15%, after which it decreased back to 9% after reaching full chlorination. Qualitatively, the same trend was observed for PM1, but chlorination of the catalyst material occurred at a slower rate. The catalyst material was fully chlorinated after 10 h.

A crucial observation is that a fast chlorination of Eu3+ was expected for La0.50Eu0.50OCl but not for PM1. PM1 showed no incorporation of La3+ into the EuOCl phase (Figure ), and therefore the same trend as for pure EuOCl would be expected. However, the excellent particle mixing of LaOCl and EuOCl heavily influenced the rate of chlorination of the pure EuOCl. This showcases that the ions in these materials are very mobile, and that facile exchange of ions occurs when the two phases are within close proximity. The apparent activation energy (Eapp) of La0.50Eu0.50OCl (126 kJ/mol) was very comparable to the Eapp of EuOCl (120 kJ/mol), suggesting that the energy needed for the reaction was not altered (Figure S9). A hypothesis on the process of ion exchange is schematically depicted in Scheme , responsible for the observed synergistic effect in catalysis. In the case where only EuOCl is present (Scheme A), the rate-determining step (RDS) is eq . The dechlorination of the catalyst surface is rapid, and therefore the bulk stays in the dechlorinated state. In the case where both Eu3+ and La3+ are present (Scheme B), ion exchange through the bulk occurs. LaOCl, acting as a Cl– acceptor/capacitator, is rapidly chlorinated by the reaction with HCl. Subsequently, the mobile excess Cl– is transferred to the Cl-deficient EuOCl, where an exchange with O2– occurs. The Cl– is reacted with CH4 and O2 on the EuOCl catalyst surface, replenishing the O2– group. While LaOCl and EuOCl individually are active in the MOC, both capable of surface chlorination and CH4 activation, the process of ion exchange is accelerated. Hence, PM1 also exhibited synergistic effects when tested for its MOC performance.

Scheme 1

Schematic Representation of the Role of (A) EuOCl and (B) Combination of LaOCl and EuOCl Exhibiting a Synergistic Effect in the Methane Oxychlorination (MOC) Reaction

For EuOCl, the rate-determining step (RDS) is the chlorination of the catalyst surface. When La3+-rich and Eu3+-rich phases are in close proximity to each other, the exchange of ions can occur. The rate-determining step, the chlorination of EuOCl, is accelerated by the presence of LaOCl. The oxygen on the LaOCl surface is replaced with Cl by the reaction with HCl. Subsequently, the excess Cl is transferred to the Cl-deficient EuOCl, after which it is transferred to the surface of the EuOCl phase. The Cl is reacted with CH4 and O2 on the catalyst surface, leaving an O2– group. Conversely, O2– travels the reverse path.

Lastly, the stability of La0.50Eu0.50OCl under MOC conditions was tested for 48 h at 450 °C under varying HCl concentrations in the feed. Every 10 h, the HCl concentration was increased to find the upper limit under which the catalyst material still exhibits stable performance. Simultaneously, the photoluminescent properties of Eu3+ were again used to monitor the degree of EuOCl chlorination. The activity/selectivity in the MOC reaction and the corresponding spectral data are plotted versus the time on stream (TOS) in Figure A,B, respectively. La0.50Eu0.50OCl exhibited very stable XCH under 10 and 20% HCl in the MOC reaction, with values of 12 and 16%, respectively. At 40% HCl, a slight downward trend in XCH was observable, going from 21 to 19%. The decline was accelerated when the HCl concentration was further increased to 60%. A final XCH of 16% was achieved after 48 h. The selectivity in the MOC reaction showed the same stability as observed for XCH. At 10 and 20% HCl in the feed, an SCH of ∼64% was achieved. When XCH showed a decreasing trend, from 60% HCl onwards till the end of the experiment, SCH slightly increased from 59 to 64% in favor of SCH and SCHCl. SCO remained unaltered under these reaction conditions at ∼13%. In line with the trends observable for XCH were the observed changes in the spectral intensity. After an initial stabilization period of ∼8 h in which the catalyst is slowly chlorinated, a steady-state composition of the catalyst was achieved as the spectral area did not change until the HCl concentration was further increased to 20%. Again, a stabilization period was observed, which now took roughly 3 h whereafter a steady state was achieved. At 40%, where the XCH slowly decreased over time, the integrated spectral area also showed a slightly decreasing slope. From 60% HCl onwards, the catalyst was gradually chlorinated almost to completion (Figure S10). These results suggest that La0.50Eu0.50OCl is stable in the MOC reaction under the condition that EuOCl is not fully chlorinated to EuCl3. This was further evidenced by performing a 100 h during stability test under the same conditions (Figure S11). No sign of deactivation was observed for La0.50Eu0.50OCl under 10% HCl at 450 °C. Furthermore, the catalytic benefits arising from the synergistic effect between La3+ and Eu3+, i.e., increased SCH, lower SCH, and similar SCO and XCH were preserved.

Figure 8

Stability test of La0.50Eu0.50OCl at 450 °C while varying the HCl concentration in the feed every 10 h. (A) XCH and SCH, SCH, SCHCl, SCCl, SCO, and SCO are plotted versus time on stream (TOS). (B) Operando luminescence spectroscopy of Eu3+ where the spectra are plotted as a heat map versus the time on stream. Furthermore, the integrated spectral area is plotted versus the time on stream as a measure for the degree of catalyst chlorination. With increasing HCl concentration up to 60%, the XCH increased while the SCO and SCH decreased. When 60% HCl was fed in the reactor, the XCH sloped down, while simultaneously the catalyst fully chlorinated. Reaction conditions: CH4/HCl/O2/N2/He of 2:2:1:1:14 (10% HCl, in mL/min), at 450 °C. Subsequently, the HCl/He ratio was altered to obtain 20, 40, 60, and 80 vol % HCl while keeping a constant flow of 20 mL/min.

Conclusions

In this work, a set of LaEu1–OCl (where x = 0, 0.25, 0.50, 0.75, and 1) solid solutions with comparable physicochemical properties were synthesized. An intimate contact between La3+ and Eu3+ was achieved, as La3+ and Eu3+ were incorporated into the same crystal structure. However, methane oxychlorination (MOC) conditions caused phase segregation into two phases: a La3+-rich phase and a Eu3+-rich phase. These phases were still in close contact with one another, exhibiting synergistic effects in the MOC reaction. LaOCl, which readily chlorinates, acts as a chlorine buffer in the EuOCl catalyst and accelerates the catalyst chlorination rate. Transport of chlorides from the La3+-rich phase to the active EuOCl is suspected to take place, facilitating the difficult EuOCl chlorination step. This synergistic effect resulted in the fact that all La3+–Eu3+ solid solution catalysts possessed enhanced activity as compared to the linear combination of LaOCl and EuOCl. Even in absolute terms, the activity of, e.g., La0.50Eu0.50OCl approached the activity of EuOCl, even though the material contains 50% less of the active Eu3+. Furthermore, mixing La3+ and Eu3+ also significantly improved the observed selectivity. Compared to EuOCl, the La3+–Eu3+ catalysts have an increased SCH (i.e., 54–66 vs 41–52%), lower SCH (i.e., 8–24 vs 18–34%), and comparable SCO (i.e., 11–28 vs 14–28%) under the same reaction conditions and varying HCl concentrations in the feed. Finally, the synergistic effect between La3+ and Eu3+ can be assured over extended reaction times as the same synergistic effect can be reached by physically mixing LaOCl and EuOCl. This physical mixture showed qualitatively the same trends as La0.50Eu0.50OCl, and after reaction, incorporation of Eu3+ in the LaOCl crystal structure was found. The improved catalyst design by the partial replacement of Eu3+ by La3+ makes Eu-based catalysts even more attractive for commercial applications as better CH3Cl yield and selectivity could be achieved while also reducing the raw material cost of the MOC catalyst.