The Effect of CO Partial Pressure on Important Kinetic Parameters of Methanation Reaction on Co-Based FTS Catalyst Studied by SSITKA-MS and Operando DRIFTS-MS Techniques

Abstract

A 20 wt% Co-0.05 wt% Pt/γ-Al₂O₃ catalyst was investigated to obtain a fundamental understanding of the effect of CO partial pressure (constant H₂ partial pressure) on important kinetic parameters of the methanation reaction (x vol% CO/25 vol% H₂, x = 3, 5 and 7) by performing advanced transient isotopic and operando diffuse reflectance infrared Fourier transform spectroscopy–mass spectrometry (DRIFTS-MS) experiments. Steady State Isotopic Transient Kinetic Analysis (SSITKA) experiments conducted at 1.2 bar, 230 °C after 5 h in CO/H₂ revealed that the surface coverages, θ_CO and θ_CHx and the mean residence times, τ_CO, and τ_CHx (s) of the reversibly adsorbed CO-s and active CH_x-s (C_α) intermediates leading to CH₄, respectively, increased with increasing CO partial pressure. On the contrary, the apparent activity (k_eff, s⁻¹) of CH_x-s intermediates, turnover frequency (TOF, s⁻¹) of methanation reaction, and the CH₄-selectivity (S_CH4, %) were found to decrease. Transient isothermal hydrogenation (TIH) following the SSITKA step-gas switch provided important information regarding the reactivity and concentration of active (C_α) and inactive -C_xH_y (C_β) carbonaceous species formed after 5 h in the CO/H₂ reaction. The latter C_β species were readily hydrogenated at 230 °C in 50%H₂/Ar. The surface coverage of C_β was found to vary only slightly with increasing CO partial pressure. Temperature-programmed hydrogenation (TPH) following SSITKA and TIH revealed that other types of inactive carbonaceous species (C_γ) were formed during Fischer-Tropsch Synthesis (FTS) and hydrogenated at elevated temperatures (250–550 °C). The amount of C_γ was found to significantly increase with increasing CO partial pressure. All carbonaceous species hydrogenated during TIH and TPH revealed large differences in their kinetics of hydrogenation with respect to the CO partial pressure in the CO/H₂ reaction mixture. Operando DRIFTS-MS transient isothermal hydrogenation of adsorbed CO-s formed after 2 h in 5 vol% CO/25 vol% H₂/Ar at 200 °C coupled with kinetic modeling (H-assisted CO hydrogenation) provided information regarding the relative reactivity (k_eff) for CH₄ formation of the two kinds of linear-type adsorbed CO-s on the cobalt surface.

1. Introduction

Low-temperature Fischer-Tropsch Synthesis (FTS) is a non-reversible, highly exothermic, and complex reaction (Equation (1)), which has been industrially applied for many decades for using syngas (CO and H₂), which is mainly derived from natural gas, biogas, and coal, towards the formation of chemicals and fuels [1,2,3].

$$n \left(CO+2H_2\right)\to {\left(-C{H}_2-\right)}_n+n H_{2}O$$

(1)

Great efforts have been made toward understanding the FTS mechanism and kinetically relevant elementary reaction steps in order to design new and improved catalytic materials via different preparation methods with optimum metal loading, particle size, and chemical and structural promoters’ composition. Additionally, reaction conditions (e.g., temperature, pressure, reactor-type, syngas ratio) must be tuned for the optimization of selectivity and yield of specific products including olefins, higher hydrocarbons, and alcohols [4,5,6]. Cobalt supported on different metal oxides (e.g., γ-Al₂O₃, SiO₂, TiO₂) and usually promoted with a small amount of noble metal (e.g., Pt, Re) but also of other metals, such as Mn and La, were mainly investigated for low-T FTS [2,7,8,9,10].

The use of transient and isotopic techniques coupled with the operando methodology in elucidating FTS mechanisms and kinetics, but also the deactivation of Co-based catalysts is rather limited [11,12,13,14,15,16,17,18,19,20,21,22,23]. In particular, Steady State Isotopic Transient Kinetic Analysis (SSITKA) has been recognized as one of the most advanced techniques providing important kinetic information for a heterogeneous catalytic reaction under working reaction conditions, such as the concentration (mol g⁻¹), surface coverage (θ), mean residence time (τ, s), and intrinsic site activity (k) of truly active reaction intermediates [14,24,25,26,27,28,29,30]. In addition, the intrinsic turnover frequency (TOF_ITK, s⁻¹) based on the concentration of active reaction intermediates measured by SSITKA (not on the total exposed metal surface sites measured by selective chemisorption) leads to the given reaction product and can be estimated [11,13,14,15,16,18,19]. Density Functional Theory (DFT) computations and microkinetic modelling alone or coupled with DFT [31,32,33,34,35,36,37,38,39,40,41] have been applied over Co-based catalysts for CO hydrogenation aimed to discriminate rival mechanisms and extract important kinetic parameters of methanation and/or Fischer-Tropsch Synthesis, and identify kinetic reasons for the effects of variation of H₂ and/or CO partial pressure on FTS product distribution. These advanced kinetic studies pointed out that hydrogenation of carbon derived from CO dissociation (e.g., CH_x) appears to be the rate-determining step (RDS), and, by varying the H₂ and CO partial pressures, there are differences in the rates of formation of CH_x-s intermediates, which eventually affect their surface lifetime (τ_CHx, s) on the catalytic surface.

The present study aims to provide important mechanistic and kinetic information on the carbon pathway of methane formation at 230 °C (feed composition: x vol% CO/25 vol% H₂/Ar, P_T = 1.2 bar) over a commercially relevant FTS Co-based catalyst (supported on γ-Al₂O₃ and promoted with 0.05 wt% Pt) [42,43], when the CO partial pressure is varied (36–84 mbar) at a constant hydrogen partial pressure of 0.3 bar. Thus, the effect of P_CO or syngas ratio (H₂/CO) at constant P_H2 on the following kinetic parameters was investigated.

(i)

The amount (N_CO, μmol g⁻¹) and surface coverage (θ_CO) of reversibly adsorbed CO-s;

(ii)

the amount (N_CHx, μmol g⁻¹) and surface coverage (θ_CHx) of active reaction CH_x-s (C_α) intermediates;

(iii)

the mean residence time of CO-s (τ_CO, s) and active CH_x-s (τ_CHx, s) intermediates;

(iv)

the turnover frequency leading to CH₄, TOF_chem, or TOF_ITK (s⁻¹), respectively, estimated based on all Co surface metal atoms or on the active reaction intermediates (μmol g⁻¹);

(v)

the amount (μmol g⁻¹) of inactive carbonaceous species (C_β) formed during FTS (230 °C) readily hydrogenated at 230 °C in 50% H₂/Ar;

(vi)

the amount (μmol g⁻¹) of the refractory carbonaceous species (C_γ) formed during FTS but hydrogenated at higher temperatures (230–550 °C);

(vii)

the relative reactivity (k_eff) of the various forms of linear-type adsorbed CO-s species formed over the Co surface during methanation after using operando transient DRIFTS-MS coupled with kinetic modelling.

2. Results

2.1. Co/γ-Al₂O₃ Catalyst Characterization

The powder X-ray diffraction (XRD) pattern of the 20 wt% Co-0.05 wt% Pt supported on Puralox SCCa γ-alumina carrier after calcination revealed the existence of a Co₃O₄ crystalline phase (35.5–38.5° 2θ range, JCPDS file No. 42-1467) as reported in our previous publication [19]. The mean Co particle size (d_Co, nm) was found to be 10.1 nm (based on Scherrer equation for Co₃O₄ and the derived relationship between d_Co and d_Co3O4 particle size), resulting in a Co dispersion with D_Co (%) of 9.5% and a Co metal surface area of 10.8 m² g⁻¹. This result was in good agreement with the D_Co (%) estimated from H₂ chemisorption/Temperature-programmed desorption (TPD) measurements [19]. TEM studies on the present catalyst [19] showed the formation of agglomerates of individual cobalt particles, which is in agreement with the literature [44]. The textural properties of the present Co/γ-Al₂O₃ catalyst were previously reported [19], where the specific surface area (SSA), pore volume (V_p), and average pore size (d_p) were found to be 88 m² g⁻¹, 0.21 cm³ g⁻¹, and 8.8 nm, respectively.

2.2. SSITKA-MS after 5 h in FTS

Figure 1A shows transient normalized concentration (Z) response curves of Kr (tracer gas), ¹³CO, and ¹³CH₄ obtained during the SSITKA step-gas switch 3 vol% ¹²CO/25 vol% H₂/Ar (5 h) → 3 vol% ¹³CO/25% H₂/1% Kr/Ar (t) performed at 230 °C over the Co/γ-Al₂O₃ catalyst. It is seen that the Z(¹³CO) transient response curve lags behind that of Kr due to a measurable concentration (see Equation (5), Section 4.2)) of molecularly and reversibly chemisorbed CO formed after 5 h in FTS. The Z(¹³CH₄) response of gas phase ¹³CH₄ clearly lags behind that of Z(¹³CO), and this is attributed to the ¹³C-containing active reaction intermediates formed (following the ¹³CO-s pool), which are sequentially hydrogenated to ¹³CH₄ (g), and the concentration of which is estimated via Equation (6) in Section 4.2.

Figure 1A,B, and

Table 1. Important kinetic parameters obtained after SSITKA following 5 h in CO/H₂ reaction at 230 °C over the 20 wt% Co/γ-Al₂O₃ catalyst with various feed gas compositions (x vol% CO/25 vol% H₂/Ar; x = 3, 5, and 7).

CO	NCO	θCO	NCHx	θCHx	yCH4, s.s (ppm)	TOFchem	TOFITK	τCHx (s)	keff	SCH4 (%)
vol%	(μmol g−1)		(μmol g−1)			(× 103, s−1)	(× 103, s−1)		(× 102, s−1)
3	10.0	0.029	15.8	0.047	3870	9.7	127.7	4.8	20.8	60
5	10.8	0.032	28.8	0.085	2690	7.4	61.5	12.0	8.3	38.5
7	16.5	0.048	39.7	0.117	4380	7.3	44.2	16.0	6.3	35

show that, by increasing the CO feed gas concentration from 3 to 7 vol% (36 to 84 mbar), the concentration (N_CO, μmol g⁻¹) and surface coverage (θ_CO) of reversibly adsorbed CO-s are increased (see Equations (5) and (7), Section 4.2), which are in agreement with the SSITKA work of Chen et al. [18]. Figure 1C presents the effect of CO partial pressure (P_CO = 36–84 mbar) and, thus, of the H₂/CO ratio (3.6–8.3, x vol% ¹²CO/25 vol% H₂/Ar) on the transient dimensionless concentration of ¹³CH₄ formed during the SSITKA step-gas switch. A clear increasing delay in the appearance of ¹³CH₄ with respect to the Z(Kr) of the tracer is observed. This behavior relates to both the reactivity of CH_X-s (named C_α) active intermediates, which lead to methane and their surface coverage. The amount and surface coverage of the active CH_x-s intermediates, N_CHx (μmol g⁻¹) and θ_CHx, respectively, along with other important kinetic parameters are reported in Table 1. It is seen that N_CHx and θ_CHx are increased to a significant extent with increasing P_CO. An increase by a factor of 2.5 is observed after increasing P_CO from 36 to 84 mbar.

Figure 2 presents transient rates (μmol g⁻¹ s⁻¹) of exchange, $R_{CO}^{ex},$ of adsorbed ¹²CO-s with ¹³CO(g) (Figure 2A) and ¹²CH_x-s with ¹³CH_x-s ($R_{CHx}^{ex},$Figure 2B) estimated from the SSITKA transient response curves recorded after 5 h in FTS at 230 °C (Figure 1, x vol% ¹²CO/25 vol% H₂/Ar → x vol% ¹³CO/25 vol% H₂/Kr/Ar, x = 3, 5, and 7). Estimation of these transient exchange rates was made after using Equations (3) and (4) (Section 4.2). The $R_{CO}^{ex}$ appears very fast, where, within the first ~8 s after the switch from the non-isotopic ¹²CO/H₂ to the equivalent isotopic ¹³CO/H₂ gas mixture, all reversibly adsorbed CO-s was exchanged. It is also illustrated that this $R_{CO}^{ex}$ rate of exchange increases with partial pressure of CO since it reflects the rate of adsorption of ¹³CO on the surface after a Co-s surface site is freed by the desorption of ¹²CO-s. The rate of the latter process is proportional to the surface coverage of ¹²CO-s. If the ¹²CO-s at steady-state is in equilibrium with gas-phase ¹²CO, then the rate of desorption of ¹²CO-s must be equal to the rate of adsorption or exchange for ¹³CO(g) during the SSITKA step-gas switch. Considering the shape profile of $R_{CO}^{ex}$(Figure 2A), similar kinetics of ¹²CO-s/¹³CO(g) exchange occurs independently of P_CO (36–84 mbar). The surface coverage of CO-s (θ_CO) in equilibrium with P_CO increases with partial pressure of CO (area under the R_CO vs. time response curve), as previously mentioned.

The R_CHx (μmol g⁻¹ s⁻¹) transient rate response curve depicted in Figure 2B show a similar shape with increasing P_CO, where the rate maximum is shifted to higher reaction times and presents a longer tail until completion of CH_x-s exchange. By increasing the CO partial pressure (x = 3, 5, and 7 vol% CO) in the feed gas stream, the initial R_CHx rate (during the very first few seconds of the transient) increases from 0.95 μmol g⁻¹ s⁻¹ to 1.23 and 1.35 μmol g⁻¹ s⁻¹, respectively, with the time required for complete exchange to occur at 50, 120 and 170 s, respectively. These features support the view of two kinds of CH_x-s species formed in the CO/H₂ reaction (230 °C, 5 h). One is hydrogenated to CH₄ with a faster rate than a second one of a lower rate, where the latter is responsible for the appearance of a tail (Figure 2B). These features will be further discussed below.

The mean lifetimes τ_CO and τ_CHx (s) of the adsorbed CO-s and active CH_x-s intermediates, respectively, were estimated based on the experimental results shown in Figure 1 after using Equations (8) and (9) (Section 4.2). The obtained results are given in Table 1 and Figure 3A as a function of the partial pressure of CO. The mean residence time τ_CO (s) is practically independent of P_CO (0.35–0.5 s), while that of τ_CHx (s) largely increases with increasing P_CO (Figure 3A and Table 1). An increase by ~3.3 times is seen after increasing the P_CO from 36 to 84 mbar (4.8 vs. 16 s).

The concentration (μmol g⁻¹) and surface coverage of adsorbed CO, θ_CO, were found to increase by 10% to 65% with increasing CO partial pressure, ca. θ_CO = 0.029, 0.032, and 0.048 for 36, 60, and 84 mbar P_CO (Figure 3B and Table 1). A significantly larger increase is obtained for the concentration (μmol g⁻¹) and surface coverage of CH_x-s (Figure 3B), where θ_CHx increases from 0.047 to 0.085 and 0.117 for 36 to 60 and 84 mbar CO partial pressure, respectively. An increase by a factor of ~2.5 is obtained after increasing the P_CO from 36 to 84 mbar. The latter values are in the range of those previously published over supported Co catalysts [8,39]. Carvalho et al. [45] reported that, for the methanation reaction at 250 °C over 20 wt% Co/SiO₂ and 20 wt% Co-0.1 wt% Pt/SiO₂ catalysts, the mean lifetime of CO-s and CH_x-s active intermediates decreased with an increasing H₂/CO gas ratio due to the greater surface coverage of hydrogen, θ_H, and the lower surface coverage of CO-s, θ_CO. They reported that these changes with H₂/CO gas ratio were responsible for the increased hydrogenation rate of CH_x-s. The decrease of θ_CO with an increasing H₂/CO gas ratio was reported [46] to have a positive effect on the ratio of θ_H/θ_*H, where θ_*H represents the surface coverage of empty sites on the Co surface. This θ_H/θ_*H parameter plays a crucial role in the FTS reaction discussed below.

The k_eff (s⁻¹) value of the methanation reaction was estimated based on Equation (12) (Section 4.2). This is plotted against the partial pressure of CO as depicted in Figure 3C. It is noted that k_eff = k_CHx θ_H. A three-fold decrease in k_eff (s⁻¹) is observed with increasing CO partial pressure, namely, k_eff = 20.8, 8.31, and 6.26 × 10⁻² s⁻¹, for 36, 60, and 84 mbar CO, respectively (Figure 3C).

Based on the surface coverages of θ_CO and θ_CHx, the τ_CO and τ_CHx values, the TOF_chem and TOF_ITK (s⁻¹) of CH₄ formation were estimated (see Equations (10) and (11), Section 4.2) and results are shown in Figure 3D. The TOF_chem (s⁻¹) (estimated based on the total Co surface metal atoms) was found to slightly decrease with increasing P_CO, namely: 9.74, 7.40, and 7.34 × 10⁻³ s⁻¹, respectively, for 36, 60, and 84 mbar CO. However, in the case of TOF_ITK (s⁻¹), which is estimated based on the total concentration of active CH_x-s and reversibly adsorbed CO-s, this was found to decrease by 2.1 and 2.9 times, respectively, in the case of 60 and 84 mbar CO compared to the case of 36 mbar CO in the feed gas stream (TOF_ITK (s⁻¹) = 127.7, 61.5, and 44.2 × 10⁻³ s⁻¹, respectively, Figure 3D). These differences in TOF_ITK (s⁻¹) are attributed to the different concentration of active sites accommodating CO-s and CH_x-s that truly participate in the carbon-path of methanation reaction, which were both influenced by the partial pressure of CO (Figure 3B). The latter results are in good agreement with the work of Carvalho et al. [45] over the Co/SiO₂ and CoPt/SiO₂ catalysts, where, at higher hydrogen partial pressures (higher H₂/CO), an increased θ_H favoured the rate of CO hydrogenation to methane. The TOF_ITK (s⁻¹) was found to increase with an increasing H₂/CO gas ratio. In fact, an increase by a factor of ~1.6 was reported for the TOF_ITK after an increase of H₂/CO from 2 to 5 to be compared to the increase of ~1.4 in the present work (CoPt/γ-Al₂O₃, after increasing the H₂/CO gas ratio from ~3.6 to 5.0 (Table 1, y^f_CO = 5 and 7 vol%).

Kinetic methane selectivity values S_CH4(%) measured at 230 °C and for the applied partial pressures of H₂ and CO are also reported in Table 1. It is clearly seen that S_CH4(%) decreases substantially with increasing P_CO. In particular, the increase of P_CO from 36 mbar (3 vol%) to 60 (5 vol%) and 84 mbar (7 vol%) results in the decrease of CH₄-selectivity by a factor of 1.55 and 1.71, respectively. Ma et al. [47] reported a very comprehensive kinetic study of methanation reaction over the 25 wt% Co/γ-Al₂O₃ catalyst at 220 °C and for a wide range of partial pressures of H₂ and CO. The authors derived a CH₄-selectivity relationship as a function of P_H2, P_CO, and P_H2O (see Reference [47], Equation (9)). We have applied the latter relationship to the present kinetic experimental data considering that P_H2O can be given by: P_H2O = y_H2O P_T = (F_{CO, f} X_CO/F_T) P_T, where P_T = 1.2 bar (total pressure), F_{CO, f} = CO molar feed flow rate (mols CO/s), and F_T = total feed molar flow rate (mols/s). It was found that S_CH4(%) at the lowest P_CO (36 mbar, 3 vol% CO) decreases by a factor of 1.4 and 1.78 with increasing P_CO to 60 mbar (5 vol%) and 84 mbar (7 vol%), respectively. These values agree very well with the measured experimental values given in Table 1, considering the slightly different reaction T (230 vs. 220 °C) and Co mean particle size used in these studies.

2.3. Transient Isothermal and Temperature-Programmed Hydrogenation (TIH/TPH)

Figure 4 presents the transient rates (μmol g⁻¹ s⁻¹) of ¹²CH₄ and ¹³CH₄ obtained during hydrogenation of carbonaceous species formed after 5 h in the CO/H₂ reaction at 230 °C, according to the following sequence of step-gas switches: x% ¹²CO/25% H₂/Ar (5 h, 230 °C) → x% ¹³CO/25% H₂/1% Kr/Ar (230 °C, 7 min, SSITKA) → Ar (3 min) → 50% H₂/Ar (t), TIH (7 min, 230 °C) (see Section 4.3). The transient response curves of ¹²CH₄ depicted in Figure 4A are due to the hydrogenation of inactive (spectator) ¹²C-containing species (named C_β) not participating in the formation of CH₄ under steady-state FTS reaction conditions. It is seen that P_CO in the feed gas stream influences both the shape of the transient rate and the amount of C_β. The former reflects the kinetics of hydrogenation of C_β, which is expected to depend on its reactivity (k) and the rate-determining step (RDS) in the reaction path of hydrogenation [48,49]. Of interest is the long tail out for more than 100 s in the hydrogen stream, irrespective of the CO partial pressure used in the feed stream. This result is similar to the TIH of carbon formed after the CO/He reaction at 250 °C on a Rh/MgO catalyst [49].

The transient rate of ¹³CH₄ (Figure 4B) reflects the hydrogenation of ¹³C-containing reaction intermediates exchanged at 230 °C after the SSITKA switch. It decays much faster than that of ¹²CH₄ (Figure 4A), and its maximum is also higher than the corresponding one of ¹²CH₄ (Figure 4A). These features show that the ¹²C-containing and ¹³C-containing species are different in reactivity towards hydrogenation to methane. The very sharp transient response of ¹³CH₄ at t < 5 s is associated with the hydrogenation of the small amount of active CH_x-s species (C_α) formed in CO/H₂ (5 h) and exchanged (labelled ¹³CH_x-s) under the SSITKA switch. It can be clearly seen that there is a small decrease in the maximum rate of ¹³CH₄ formation by increasing the CO concentration in the feed gas stream from 3 to 5 vol% (4.4 vs. 4.0 μmol g⁻¹ s⁻¹, respectively), while a further increase to 7 vol% leads to an increase in the rate maximum by ~25% (5.5 μmol g⁻¹ s⁻¹). However, this appears practically at the same time, ca. 4–6 s. Part of the ¹³CH₄ transient response (Figure 4B) is due to the hydrogenation of CO-s formed in CO/H₂ (5 h) and exchanged under the SSITKA switch (Figure 1A,B, and Figure 3B). The hydrogenation of ¹³CO-s lasts for ~20 s in the case when 3 vol% CO was used in the feed. However, in the case of 5 and 7 vol% CO, the hydrogenation takes longer (~40 s). The amounts (μmol/g) and surface coverages (based on the exposed Co surface metal atoms) of the corresponding C_β (¹²CH₄ transient) and those of C_α and CO-s (¹³CH₄ transient) are provided in

Table 2. Amounts (μmol g⁻¹) and surface coverages (θ) of carbonaceous species formed at 230 °C after 5 h in FTS reaction (x%CO/25%H₂/Ar), and which are hydrogenated at 230 °C (TIH) and during TPH (240–600 °C).

vol% CO	ΤΙH	ΤΙH	ΤΙH	ΤΙH	ΤPH	ΤPH
	N (12CH4) or Cβ	θ (12CH4)	N (13CH4)	θ (13CH4)	N (12CH4) or Cγ	θ (12CH4)
	(μmol g−1)		(μmol g−1)		(μmol g−1)
3	34.3	0.10	27.6	0.08	128	0.38
5	27.6	0.08	40.2	0.12	228	0.67
7	30.4	0.09	56.3	0.17	252	0.74

. The θ(¹³CH₄) was found to be 0.08, 0.12, and 0.17 for the 3, 5, and 7 vol% CO used in the feed gas stream, respectively. It is pointed out that the sum of the surface coverages of CO-s and C_α estimated during SSITKA is in very good agreement with the TIH results (θ(¹³CH₄), Figure 4B). The amount (μmol g⁻¹) of hydrogenated C_β species given in Table 2 is found to be approximately the same irrespective of the partial pressure of CO in the feed gas stream, namely: 34.3 (θ = 0.10), 27.6 (θ = 0.08), and 30.4 (θ = 0.09) μmol g⁻¹ for the 3, 5, and 7 vol% CO in the feed.

The ¹²CH₄ transient response curves obtained during temperature-programmed hydrogenation (TPH), following 5 h in CO/H₂ reaction and 7 min TIH at 230 °C, are reported in Figure 5 for the three CO concentrations used in the feed gas stream (3, 5, and 7 vol% CO). The several hydrogenation peaks and shoulders that appeared in these TPH traces are attributed to inactive carbonaceous species (named C_γ) formed during the FTS reaction, and which could not be hydrogenated at 230 °C, as opposed to the C_β hydrogenated to methane at 230 °C (TIH, Figure 4A). As shown in Figure 5A, the amount of C_γ (Table 2) is influenced by the CO feed gas concentration, as opposed to the shape and T_max (°C) of the ¹²CH₄ traces. Furthermore, the fraction of the various types of C_γ carbonaceous species was found not to be influenced by the CO concentration in the feed, according to the deconvolution of the TPH traces performed (Figure 5B), and these results are reported in Table S1 (ESI). By increasing the CO concentration in the feed gas stream from 3 to 5 and 7 vol%, the amount of CH₄ formed was found to increase, ca. 128, 228, and 252 μmol g⁻¹, respectively. Additionally, the main peak maximum is ca. 290, 420, and 520 ppm, respectively (Figure 5A).

2.4. Operando DRIFTS—Transient Isothermal Hydrogenation (TIH) of CO

Operando DRIFTS-Mass spectrometry experiments were conducted in the DRIFTS cell (PFR performance) in which the outlet was connected to a mass spectrometer [19]. After 2 h of methanation reaction at 200 °C (5 vol% CO/25 vol% H₂/Ar), the infra-red (IR) cell was purged with Ar for 3 min, which was followed by a step-gas switch to 50 vol% H₂/Ar in order to study the transient kinetics of hydrogenation of each kind of adsorbed CO-s. Figure 6A shows the DRIFTS spectrum in the CO region (1800–2250 cm⁻¹) and its deconvolution into five IR bands, following the deconvolution methodology previously described by us [19]. The IR bands marked (1) and (2) are due to the R and Q branches of gas-phase CO. The IR bands marked as (3) and (4) centered at 2015 and 1980 cm⁻¹, respectively, are due to the presence of two linear-type reversibly chemisorbed CO-s. The third IR band centered at 1905 cm⁻¹ is due to a bridged-type reversibly adsorbed CO-s [50,51,52,53,54]. At least one of these CO-s should be considered a precursor for methanation and higher HCs formation on the Co surface. The DRIFTS spectrum of Figure 6A is in agreement with our previous results (5 vol% CO, H₂/CO = 2) and others reported for different H₂/CO gas ratios [19,50,51,52,53,54].

The DRIFTS spectrum recorded after the 3-min Ar purge is shown in Figure 6B (Ar, t = 0 s) and its deconvolution into two kinds of linear CO-s (L₁, L₂) and one bridged-type CO-s (B) is depicted in Figure 6C. Upon the step-gas switch Ar → 50 vol% H₂/Ar (t), the surface coverage of adsorbed CO-s (sum of L₁, L₂, and B) progressively decreases with time in hydrogen stream, and the maximum of the IR band shifts slightly to lower wavenumbers (Figure 6B). The latter is attributed to the effect of surface coverage of CO-s on its binding strength with the cobalt surface (lateral repulsive interactions between adjacent adsorbed CO-s). Thus, the bond energy between C and O within the adsorbed CO-s decreases, and, thereby, lowers the vibrational frequency as observed.

DRIFTS-TIH deconvoluted spectra with time of hydrogenation of the two linear-type (L₁, L₂) adsorbed CO-s species were further analysed by applying the microkinetic modelling described in detail [19]. Due to the small integral band area of the third bridged-type (B) adsorbed CO-s, no further analysis was attempted. Figure 7A,B show DRIFTS spectra of the evolution of the IR bands of L₁ and L₂ linear-type adsorbed CO-s with time in hydrogen gas stream, after following the same deconvolution procedure shown in Figure 6C (deconvolution in Ar gas stream). Both L₁ and L₂ adsorbed CO-s formed after 2 h in the CO/H₂ reaction at 200 °C were hydrogenated towards methane within the first 3 min in 50 vol% H₂/Ar. Figure 7C plots the integral band intensity (Abs cm⁻¹) of the L₁ and L₂ as a function of time in 50%H₂/Ar gas treatment. Assuming the same extinction coefficient for the two linear-type CO-s species, it is seen that the relative population of the two CO-s is L₁:L₂ ~65:35%. The integral bands of L₁ and L₂ follow a very similar decay with time in hydrogenation, which suggests similar kinetics of hydrogenation. This is confirmed by the following kinetic analysis.

The transient kinetics of hydrogenation of CO-s was modelled via the H-assisted CO dissociation mechanism [55,56,57], where the formation of hydroxymethylene species (HCOH-s) from the sequential hydrogenation of CO-s was considered as the rate-determined step (RDS) [19]. Based on this kinetic analysis, an apparent rate constant (k_eff, s⁻¹) associated with this RDS was derived using Equation (2), where k_eff is the product of the intrinsic rate constant k times θ_H (k_eff = k θ_H).

$$\ln \left[a\right]=-k_{eff}t ;a=\frac{A_{CO}\left(t\right)}{A_{CO}\left(t=0\right)}$$

(2)

In Equation (2), α is equal to the ratio of the integral IR band intensity for each individual linear-type adsorbed CO-s at a given time, t, in H₂/Ar to the integral IR band intensity recorded at t = 0 (after the 3-min Ar purge before the H₂/Ar step-gas switch was made).

Figure 7D presents linear plots of ln[α] versus time in H₂/Ar for the two (L₁, L₂) linear adsorbed CO-s species. Based on Equation (2), the slope of the linear ln[α] vs. time plot provides the apparent rate constant (k_eff) of each of the two linear adsorbed CO-s. It is shown that the L₁ and L₂ adsorbed CO-s exhibit very similar reactivities (k_{eff_L1}/k_{eff_L2} ~1, k_{eff_L1} = 0.016 s⁻¹) toward hydrogenation to CH₄. Since k_eff = k θ_H, the L₁ and L₂ linear-type CO-s have very similar intrinsic reactivities (k_L1 ~k_L2). The k_eff value estimated is smaller than the one reported by us on the same Co/γ-Al₂O₃ catalyst at 230 °C and for H₂/CO = 2 (5% CO, 10% H₂). More precisely, k_{eff_L1} = 0.021 and k_{eff_L2} = 0.027 s⁻¹ with k_L1/k_L2 to be 0.78 (230 °C). Assuming that the pre-exponential factors are the same for the two rate constants, then the k_L1/k_L2 ratio takes the value of 0.76 at 200 °C. Furthermore, the difference in the activation energies (E₁-E₂, kcal/mol) of hydrogenation of these two linear-type CO species was estimated to be ~0.3 kcal mol⁻¹. The increase in the hydrogen partial pressure from 0.12 to 0.3 bar (10 to 25 vol% H₂, P=1.2 bar) has a very small influence on the intrinsic rate constant of CO-s hydrogenation via the H-assisted CO dissociation mechanism (RDS: HCO-s + H-s → HCOH-s + s).

The transient evolution of CH₄ formation during the DRIFTS-TIH experiment was recorded by online mass spectroscopy (operando methodology). This is depicted in Figure 8. At the same time, the gas-phase CO response was measured and compared to that of the Kr tracer, as shown in Figure 8. Based on the Kr and CO response curves, it is illustrated that desorption of CO during the 50%H₂/Ar gas treatment of the catalyst sample was negligible. The CH₄ transient response (Figure 8) presents two rate peak maxima. The first one, which appeared at ~8 s, is largely due to the CO-s and active CH_x-s formed after 2 h of CO/H₂ reaction at 200 °C, whereas the second peak centered at ~40 s is largely due to the C_β species presented and discussed above (Section 2.3, Figure 4A). Integration of this transient response of CH₄ provides the total amount (μmol/g or θ) of CO-s, and active CH_x-s (C_α) and C_β species formed after 2 h in CO/H₂ reaction. This amount was found to be 31.1 μmol/g (θ = 0.09), as opposed to the value of θ = 0.20 obtained at 230 °C after 5 h in CO/H₂ (Table 2). It is reasonable to suggest that, since θ_CO decreases with increasing rection T, and θ_CHx decreases with time-on-stream [11,21], then inactive C_β appears to grow with an increasing reaction temperature from 200 to 230 °C.

3. Discussion

3.1. Influence of CO Partial Pressure on Intrinsic Kinetic Parameters of Methanation Reaction

The increase of CO partial pressure from 36 to 84 mbar (3–7 vol% CO, P = 1.2 bar), while keeping the H₂ partial pressure constant at 300 mbar (25 vol%), led to significant variations in several intrinsic kinetic parameters of the methanation reaction conducted at 230 °C and after 5 h over the 20 wt% Co-0.05 wt% Pt/γ-Al₂O₃ commercially relevant catalyst. The θ_CO was found to be small, ca. 0.029, 0.032, and 0.049 for 36, 60, and 84 mbar, respectively, but with an increasing trend as the CO partial pressure increases and the H₂/CO gas ratio decreases. An increase of θ_CO by a factor of ~1.68 was obtained as the CO partial pressure increases from 36 to 84 mbar. The latter result agrees with the characteristic features of Langmuir isotherm and the fast ¹²CO-s/¹³CO(g) exchange, which was revealed under SSITKA (see Figure 1A,B, and Figure 2A). The increase of θ_CO with P_CO finds agreement with the SSITKA work conducted by Chen et al. [18] over the 17.1 wt% Co-0.04 wt% Pt/SiO₂ catalyst. However, the reaction conditions of T = 260 °C and P_H2 = 0.45 bar (P_CO in the 15–90 mbar range) are different than in the present work (230 °C, P_H2 = 0.3 bar). The authors reported that θ_CO increased from ~0.18 to ~0.30 in the examined P_CO range. The differences in θ_CO between that work [18] and the present one must be seen as the result of the influence of several reaction parameters on θ_CO, which were not the same. For example, the reaction T, the time-on-stream (5 h in the present work vs. 16 h in Reference [18]), the H₂ partial pressure, and the Co mean particle size (10 nm in the present work vs. 15 nm in Reference [18]).

The effect of La-promotion of Co/γ-Al₂O₃ catalysts toward methanation was investigated by Vada et al. [58] using the SSITKA technique at similar FTS reaction conditions (P_CO = 0.036 bar, H₂/CO = 10, T = 220 °C) to those used in this work. They found similar surface coverages of θ_CO = 0.11 and θ_CHx = 0.05 as in the present work (Figure 3B). Yang et al. [59] in their SSITKA work on Co/γ-Al₂O₃ using a CO partial pressure of P_CO = 55 mbar, H₂/CO = 10, T = 210 °C and P_T = 1.85 bar, and, after 5 h on FTS reaction stream, they reported a similar θ_CHx value but larger θ_CO ~0.3. Thus, a larger P_H2 (0.55 vs. 0.3 bar in the present work) and lower T (210 vs. 230 °C in the present work) led to larger θ_CO values. In the case of TOF_chem, a value of ~12.5 × 10⁻³ (s⁻¹) was reported [59] and compared to the present value of 7.4 × 10⁻³ (s⁻¹) for very similar P_CO (55–60 mbar) but larger P_H2 (0.55 vs. 0.3 bar in the present work). A k_eff (CH_x) value of ~0.1 s⁻¹ was reported [59] compared to the value of 0.08 s⁻¹ (Table 1) for the previously mentioned comparative experimental conditions.

Carvalho et al. [11,45] in their SSITKA works conducted at 250 °C over Co-based catalysts reported that θ_CO decreases and θ_CHx increases with an increasing H₂/CO gas ratio. The authors claimed that higher θ_CHx could be linked to larger hydrogenation rates for higher hydrocarbons, which is in agreement with the decrease of S_CH4 and increase of S_C5+ [45]. These results are similar to those reported by Pena et al. [60] and those in the present work (Table 1), where, after increasing the CO partial pressure, an increase of θ_CHx and a concomitant decrease in S_CH4 are obtained. In addition, Keyvanloo et al. [33] reported that, by increasing the CO partial pressure, while keeping the H₂ partial pressure constant, the S_CH4 decreases. This result is similar to our findings, where the S_CH4 (%) behaviour with P_CO and P_H2 was presented and discussed in Section 2.2 in relation to kinetic modelling studies [47] conducted on the Co/γ-Al₂O₃. In the present work, it was found that, after increasing the H₂/CO gas ratio (decrease of CO partial pressure), the ratio of θ_CO/θ_CHx grows toward unity. This result agrees with that found by Carvalho et al. [45], where, for H₂/CO = 2, θ_CO was higher than θ_CHx, whereas, for H₂/CO = 5, θ_CHx was higher than θ_CO. This trend of θ_CO/θ_CHx with the H₂/CO gas ratio is attributed to the competitive H₂ and CO chemisorption where, after increasing the hydrogen partial pressure in a given range, θ_H increases against θ_CO, which results in increasing rates for -CH_x coupling and further hydrogenation to higher hydrocarbons (lowering the S_CH4).

The mean residence time of active -CH_x species (τ_CHx, s) leading to CH₄, was found to increase ~2.5 times by increasing the CO feed gas concentration from 3 to 5 vol%, whereas a further increase to 7 vol% causes only a slight increase (~30%). This result agrees very well with previous works. More specifically, τ_CHx is in the range of 5–15 s for the present catalytic system and the FTS reaction conditions applied, which is similar to other previous SSITKA works over Co-based catalysts supported on SiO₂ or Al₂O₃, where τ_CHx ranged from 8 to 25 s [8,11,15,16,59]. Vada et al. [58] for Co supported on alumina carrier found that, by decreasing the H₂/CO gas ratio, the τ_CHx increases (7, 13, and 29 s for H₂/CO = 15, 10, and 5, respectively, CO = 2 vol%). This result is in harmony with those of the present work. However, it should be noted that, in that work [58], the effect of H₂/CO gas ratio was investigated by varying the P_H2 as opposed to the present work where the P_CO varied and P_H2 was kept constant (Figure 3A).

According to the results shown in Table 1, the TOF_chem (s⁻¹) for CH₄ formation at 230 °C decreases by ~25% when the P_CO increases from 36 to 84 mbar. A similar behaviour was reported for the methanation reaction at 260 °C on Co/SiO₂ (P_H2 = 450 mbar) [18]. The drop in TOF of CH₄ formation is attributed to the deterioration of site activity (k_eff) of the -CH_x species and not of a possible decrease of the surface coverage of -CH_x, as illustrated in Table 2. Given the fact that k_eff = k θ_H (see Section 4.2), it is reasonable to suggest that, as θ_H decreases with increasing P_CO, then deterioration of intrinsic activity (k) of -CH_x should be considered less than 25% in that range of P_CO. Weststrate and Niemantsverdriet [46] pointed out the importance of θ_CHx and θ_V (coverage of empty catalytic sites) for hydrogen adsorption related to the hydrogenation of CO toward methane formation. The surface coverage of CO-s defines the amount of hydrogen able to chemisorb next to adsorbed CO-s for a given P_CO, and, as a consequence, the τ_CHx strongly depends on the H₂/CO gas ratio [46,58].

The TOF_ITK (s⁻¹) of CH₄ formation estimated based on the active CO-s and CH_x-s intermediates measured by SSITKA shows large differences when compared to TOF_chem for a given P_CO, according to the results reported in Table 1. The influence of P_CO on TOF_ITK is largely different than in the case of TOF_chem. By increasing the CO partial pressure from 36 to 84 mbar, the TOF_ITK decreases by a factor of 2.9 (127.7 vs. 44.2 × 10⁻³ s⁻¹) to be compared to the value of 1.3 for the TOF_chem, which indicates the more sensitive nature of TOF_ITK on P_CO. It is interesting to mention the variation of TOF_ITK/TOF_chem with P_CO, where a decrease from 13 to 6 is obtained by increasing the P_CO from 36 to 84 mbar.

The exchange rates of CO-s and CH_x-s in the SSITKA experiment were estimated (see Section 4.2, Equations (3) and (4)) and reported (Figure 2) for the first time to the best of our knowledge. It is clearly illustrated the significantly larger adsorption/desorption rate of CO, $R_{CO}^{ex}$, compared to that of hydrogenation of -CH_x ($R_{CHx}^{ex}$) under working FTS reaction conditions (5–7 vol% CO in the feed), but the similar rates for the lowest feed concentration of 3 vol% CO. Furthermore, the shape profile of $R_{CHx}^{ex}$ suggests the influence of P_CO on the formation of different kinds of CH_x-s in the methanation reaction path with different reactivity toward hydrogenation to methane. Bell and co-workers [61] were the first to describe an advanced mathematical analysis of the ¹³CH₄-SSITKA response curve for the methanation reaction, which reveals the distribution of reactivity of -CH_x-s (f(k)—reactivity distribution function). Thus, the heterogeneity of sites exists for this particular active reaction intermediate. Weststrate and Niemantsverdriet [46] pointed out that the long CH_x-s residence time (e.g., 10 s) observed on the Co(0001) surface, which is similar to the τ_CHx values reported in Table 1, cannot be due to the existence of low values of θ_H or θ_H/θ_*H, where θ_*H is the surface coverage of empty sites for hydrogen chemisorption. The authors proposed, based on DFT calculations and the θ_CHx values observed (0.1–0.2 ML), that CH_x are formed on step edge sites, which then migrate to the terrace sites. The latter was considered the rate-determining step (RDS) that gives rise to residence times of ~10 s. Given the fact that more than one kind of step edge sites can be formed on the Co surface with a distribution that depends on the Co particle size, hydrogenation of such CH_x-s species is more sensitive to the P_CO, which is shown in the present work (Figure 2B). An increase of CO feed gas concentration from 3 to 7 vol% causes the increase of τ_CHx by ~3.5 times (Table 1).

3.2. The Influence of H₂ Partial Pressure on the Reactivity of CO-s in the Methanation Reaction

Transient Isothermal Hydrogenation (TIH) of adsorbed CO-s coupled with DRIFTS performed at 200 °C after 2 h in 5 vol% CO/25 vol% H₂/Ar along with microkinetic analysis, which was based on a hydrogen-assisted CO hydrogenation mechanism, revealed practically the same reactivity (k) for each of the two L₁ and L₂ linear-type adsorbed CO-s species populated on the present Co/γ-Al₂O₃ catalyst (Figure 7C,D). According to the results depicted in Figure 6C and Figure 7C, the initial surface coverage of L₁ and L₂ adsorbed CO-s is largely different. The θ_CO apparently had no influence on the hydrogenation activity of each of these two linear-type CO-s, thus, on the RDS (HCO-s + H-s → HCOH-s + s), in agreement with the recent work of Zijlstra et al. [34]. The authors based on Density Functional Theory (DFT) calculations reported that the C-O bond dissociation energy on step-edge sites is hardly affected by CO co-adsorbates. The comparison between the CO-s hydrogenation rate constant (k) found in the present work (Figure 7, Section 2.4) for a H₂/CO gas ratio of 5 and that from previous results [19] for a H₂/CO gas ratio of 2 (same 5% CO concentration was used in the FTS reaction) over the same catalyst, suggest that the H₂/CO gas ratio hardly affects the k of RDS (H-assisted CO hydrogenation mechanism).

The results of the operando DRIFTS-Mass spectrometry methodology applied in the isothermal hydrogenation of adsorbed CO-s and in the transient mode, which are depicted in Figure 6, Figure 7 and Figure 8, combined with the SSITKA-MS results depicted in Figure 3B, allowed to learn about the nature of the various kinds of adsorbed CO-s, their distribution (individual θ_CO (L_i), i = 1, 2), and their reactivity (k_i, s⁻¹) toward hydrogenation to CH₄. On the other hand, the SSITKA-MS alone and the TIH-Mass spectrometry alone (Figure 4B) allowed to estimate only the total surface coverage of all types of CO-s and an effective reactivity, k_eff (s⁻¹) for methanation associated with the active CH_x-s formed from all these related adsorbed CO-s precursor species. Thus, the powerful combination of SSITKA with an operando methodology becomes clear from the results of the present example of methanation reaction on Co/γ-Al₂O₃, as demonstrated also in the WGS reaction and others reported therein [24].

3.3. Effect of CO Partial Pressure on the Formation of Inactive C_β and C_γ Carbonaceous Species

Transient isothermal (TIH) and temperature programmed hydrogenation (TPH) conducted following the SSITKA step-gas switch (¹²CO/H₂ → ¹³CO/H₂) over alumina-supported cobalt for the first time was described in detail in our previous publication [19]. In the present work, the influence of CO partial pressure in the 36–84 mbar range (constant H₂ partial pressure of 300 mbar) on the rate of formation of inactive carbonaceous species (based on the measured amount for a given time-on-stream, TOS) was investigated. The transient rate of ¹²CH₄ formation estimated during TIH at 230 °C (Figure 4A) corresponds to the hydrogenation of inactive C_β, C_xH_y-s (y ≠ 0), the structure of which was revealed via in situ DRIFTS [19], which is in agreement with other studies on Co-based FTS catalysts at similar reaction temperatures [46,62].

The amount of C_β (Table 2) was found to be 1.2 and 1.1 times higher in the case where 3 vol% CO compared to 5 and 7 vol% CO, respectively, were used (34.3, 27.6, and 30.4 μmol C g⁻¹, respectively). The shape of ¹²CH₄ response (Figure 4A) revealed two kinds of inactive C_β species (C_β,I—small peak at ~5 s, and C_β,II—broad peak with maximum at ~30 s) in the case where 3 vol% CO was used in the feed gas stream, as opposed to the case of 5 and 7 vol% CO, where only single and very similar in shape peaks were observed, which likely correspond to adsorbed carbonaceous species of the same chemical structure. Based on our previous DRIFTS work [19] and the results of Figure 4A (peak shapes), the C_β,I—small peak could be attributed to the presence of a small concentration of HCOO-s (formate species) formed on the Co-alumina interface able to be hydrogenated at 230 °C. This assignment finds support from the TIH experiments reported by Paredes et al. [23].

The rate of hydrogenation of CO-s and active CH_x-s (Figure 4B) is significantly larger than that of inactive C_β (¹²CH₄ response, Figure 4A), and this explains the accumulation of C_β in CO/H₂ reaction conditions, where θ_H is lower than that established in 50%H₂/Ar gas treatment. In the former case, CO and H₂ competes for Co active sites, where the former presents larger adsorption energy. Thus, θ_CO will depend largely on the partial pressure of CO only, as opposed to θ_H, which will also depend on the concentration of Co surface vacant sites [18]. As mentioned previously (see Section 3.1), by reducing θ_H, the concentration of active CH_x-s formed will be reduced. As shown in Table 2 and Figure 4, with decreasing CO partial pressure in the feed gas stream (ca. 3 vol%), θ_CO and θ_CHx become smaller (Figure 3B), which leads to less blockage of Co active sites for H-s formation. Thus, more C_β can be formed. The latter finds support from the literature [17], where -CH_x monomers polymerization and termination (C_β formation) proceed with faster rates than the H-assisted CO dissociation step at FTS reaction conditions. Based on the amounts of C_β (Table 2), the effect of P_CO is not that large. Approximately, a 15% decrease occurs by increasing the P_CO from 36 to 84 mbar at 230 °C on the present Co/γ-Al₂O₃ commercially relevant catalyst.

The TPH results shown in Figure 5A,B, Table 2, and Table S1 (ESI) reveal the likely existence of four kinds of carbonaceous species of different distribution but of similar individual activity (note the similar T_M values, Table S1), independent of CO partial pressure. The surface coverage of each of these C_γ carbonaceous species largely depends on P_CO (Table 2). By increasing the CO feed concentration from 3 to 7 vol%, the total amount of C_γ was nearly increased by a factor of two (Table 2). The chemical structure of C_γ was reported [31,32,34] to include: (i) oligomeric strongly adsorbed hydrocarbon species (peaks 1, 2, and 3, T_M = 320–390 °C) and (ii) polymeric carbon (Peak 4, T_M = 400–450 °C). The latter species were not eliminated during DRIFTS-TIH at 230 °C as reported in our previous publication [19] (IR band centred at 2915 cm⁻¹), but at higher temperatures (230 < T < 400 °C). It was reported [63] that larger amounts of polymeric carbon are formed with decreasing hydrogen partial pressure at constant P_CO over Co-based FTS catalysts. The possibility that formates could be considered as part of C_γ should be excluded, as these were hydrogenated under DRIFTS-TPH at T < 300 °C [19].

Pena et al. [60] in support of the present findings showed that, by increasing the H₂/CO gas ratio, higher amounts of polymeric carbon reacted with hydrogen under TPH. Keyvanloo et al. [33] showed via TPH following FTS reaction under industrial conditions that, by increasing the H₂/CO ratio at a constant hydrogen partial pressure, more hydrogenated polymeric carbon was formed, which is similar to the present findings, where C_γ is larger in the case of 7 vol% CO with most of carbonaceous species to be hydrogenated at T > 350 °C but lower than 550 °C. Thus, this cannot be attributed to atomic (low Ts) or graphitic carbon (high Ts). It was reported [33,64] that, by increasing P_CO at constant P_H2, the deactivation rate increases, which might be due to the increase of hydrogen-deficient polymeric inactive carbonaceous species. Additionally, Pena and co-workers [60,65] showed via TPH and Scanning Transmission Electron Microscopy–Electron Energy Loss Spectroscopy (STEM-EELS) that, under low hydrogen partial pressures in FTS reaction conditions, more stable carbon species are formed with polymeric carbon and strongly adsorbed hydrocarbon-like species to be responsible for the deactivation of the catalyst at a lower H₂/CO gas ratio. This result agrees with the present work where, at a lower H₂/CO gas ratio, larger amounts of C_γ are formed.

4. Materials and Methods

4.1. Co/γ-Al₂O₃ Catalyst—Synthesis and Characterization

The 20 wt% Co supported on Sasol Germany’s Puralox SCCa γ-Al₂O₃ catalyst was prepared by SASOL S.A., as described elsewhere [19,42]. More precisely, 20 wt% Co and 0.05 wt% Pt were deposited on the alumina support via a slurry-phase impregnation procedure with aqueous metal precursor solutions described in References [42,43]. The resulting slurry was dried under vacuum and calcined in static air at 250 °C. Sequential impregnation and calcination steps were applied in order to achieve 20 wt% Co nominal loading. Prior to any catalytic and transient kinetic measurements, the catalyst was in situ reduced in pure H₂ (1 bar, 50 NmL min⁻¹) at 425 °C for 10 h (from room T to 425 °C, the temperature of the solid was increased at β = 1 °C min⁻¹), which was followed by an He-purge and cooling to 230 °C for FTS. A 5 vol% O₂/He (2 h) was applied at 30 °C, following catalytic measurements, for catalyst passivation. The latter sample was further used in DRIFTS studies (Section 4.4). The as-received commercially relevant catalyst in powder form was sieved to less than a 106-μm particle size for avoidance of internal mass transport resistances before catalytic and transient kinetic measurements. Textural (BET/BJH), structural (PXRD, HAADF-TEM), and H₂-TPR/H₂-TPD studies over the examined catalyst were described and presented in detail in our previous publication [19].

4.2. SSITKA-MS Following 5 h in FTS

Steady State Isotopic Transient Kinetic Analysis (SSITKA) experiments with the use of ¹³CO were conducted in order to investigate the effects of the variation of CO partial pressure (constant H₂ partial pressure), or the H₂/CO gas ratio, on the carbon pathway of methane formation at 230 °C. After 5 h in CO/H₂ (x vol% ¹²CO/25 vol% H₂/Ar, GHSV = 30,000 h⁻¹), the SSITKA step-gas switch to x vol% ¹³CO/25 vol% H₂/Ar (t) was made. At this point, it is worth mentioning that internal and external mass transport resistances in the catalytic bed placed in the quartz micro-reactor used were absent, as previously reported [19,66]. During the SSITKA step-gas switch, the mass numbers (m/z) 15, 17, 18, 28, 29, and 84, which correspond to ¹²CH₄, ¹³CH₄, H₂O, ¹²CO, ¹³CO, and Kr, respectively, were continuously monitored by MS. The contribution of H₂O (m/z = 18) to the m/z = 17 was considered for the analysis of ¹³CH₄ signal to concentration (ppm or mol%) in the case a different signal at m/z = 18 in the two ¹²CO/H₂ and ¹³CO/H₂ gas mixtures would be recorded. An ideally designed and conducted ¹²CO/¹³CO-SSITKA experiment for FTS should provide the same signals at m/z = 18 [19,21].

The transient rates (μmol g⁻¹s⁻¹) of CO and active CH_x-s intermediates which are exchanged (¹²CO-s is replaced for ¹³CO-s and ¹²CH_x-s for ¹³CH_x-s) during SSITKA were estimated based on Equations (3) and (4), respectively.

$$R_{CO}^{ex} \left(mol{g}^{-1}{s}^{-1}\right)=\frac{F_T{y}_{CO}^f}{W}\left[A-\left(X_{CO}^{s.s}\left(S_{C{H}_4}{Z}_{{}_{ }{}^{13}C{H}_4}\left(t\right)+S_{C2}{Z}_{C2}\left(t\right)+S_{C3}{Z}_{C3}\left(t\right)+\dots \right)\right)\right]$$

(3)

$$R_{CHx}^{ex}\left(mol{g}^{-1}{s}^{-1}\right)=\frac{F_T{y}_{{}_{ }{}^{13}C{H}_4}}{W}\left[Z_{{}_{ }{}^{13}CO}\left(t\right)-Z_{{}_{ }{}^{13}C{H}_4}\left(t\right)\right]$$

(4)

where F_T is the total molar flow rate (mol s⁻¹), y^f_CO is the mole fraction of CO in the feed, $X_{CO}^{s.s}$ is the CO conversion (%) at a steady state, W is the mass of the catalyst (g), y_13CH4 is the mole fraction of ¹³CH₄ at the effluent stream from the micro-reactor in the new steady-state obtained in ¹³CO/H₂, Z is the dimensionless concentration of a given gas-phase species, and A(t) = $\left(Z_{Kr}\left(t\right)-Z_{{}_{ }{}^{13}CO}\left(t\right)\right)$.

Equation (3) is based on the fact that the rate of ¹³CO fed in the reactor (mols ¹³CO/s) during the step-gas SSITKA switch (until the new steady-state is reached) must be equal to the sum of the rates of: (i) exchange of ¹³CO with ¹²CO-s ($R_{CO}^{ex}$), (ii) conversion of ¹³CO into ¹³CH₄, (iii) conversion of ¹³CO into ¹³C₂₊-hydrocarbons, and (iv) the rate of non-reacted ¹³CO(g) leaving the reactor (rate of outlet flow of ¹³CO(g) from the micro-reactor). The Z_C2, Z_C3, and those of higher hydrocarbons (Z_Cn) that appear in Equation (3) represent the fractional change in ¹³C for the C₂-, C₃-, and C_n-hydrocarbons, respectively, and can only be measured after using the SSITKA-GS-MS methodology [17,67]. In the present work, only the Z_C1 (Z(¹³CH₄)) was measured by mass spectrometry. However, according to the experimental work of Rebmann et al. [67] on a 13 wt% Co/γ-Al₂O₃ catalyst, the Z_C2 and Z_C3 transient curves were very close (almost identical) to that of Z_C1. Given the fact that, in the present work, the total carbon selectivity to C₁-C₃ is more than 85%, the Z_Cn (n > 3) terms in Equation (3) were neglected, and it was considered that Z_C1 ~Z_C2 ~Z_C3. In Equations (3) and (4), the accumulation of ¹³CO(g) in the CSTR micro-reactor used [66] was found to be very small when compared to the other terms. Thus, it was also neglected.

The concentration (μmol g⁻¹) of the reversibly adsorbed CO-s and active CH_x-s intermediates leading to CH₄ are given by the following material balance Equations (5) and (6), respectively [19].

$$N_{CO}\left(mol{g}^{-1}\right)=\frac{F_T{y}_{CO}^f\left(1-X_{CO}\right)}{W}\underset{0}{\overset{t_{s.s}}{{\displaystyle \int }}}\left[Z_{Kr}\left(t\right)-Z_{{}_{ }{}^{13}CO}\left(t\right)\right]dt$$

(5)

$$N_{CHx}\left(mol{g}^{-1}\right)=\frac{F_T{y}_{{}_{ }{}^{13}C{H}_4}}{W}\underset{0}{\overset{t_{s.s}}{{\displaystyle \int }}}\left[Z_{{}_{ }{}^{13}CO}\left(t\right)-Z_{{}_{ }{}^{13}C{H}_4}\left(t\right)\right]dt$$

(6)

The surface coverage, θ, and the mean residence time, τ (s), of the reversibly adsorbed CO-s and active CH_x-s intermediates, can be estimated based on the following Equations (7)–(9).

$${\theta }_i=\frac{N_i}{N_{Co,surf}}\left(i=CO, C{H}_x\right)$$

(7)

$${\tau }_{CO}=\underset{0}{\overset{t_{s.s}}{{\displaystyle \int }}}\left[Z_{Kr}\left(t\right)-Z_{{}_{ }{}^{13}CO}\left(t\right)\right]dt$$

(8)

$${\tau }_{CHx}=\underset{0}{\overset{t_{s.s}}{{\displaystyle \int }}}\left[Z_{{}_{ }{}^{13}CO}\left(t\right)-Z_{{}_{ }{}^{13}C{H}_4}\left(t\right)\right]dt$$

(9)

where t_s.s. is the time at which the new steady-state is obtained under the ¹³CO/H₂ gas mixture.

The kinetic rate TOF (s⁻¹) of CH₄ formation is estimated based on the Co dispersion, Equation (10) (TOF_chem, s⁻¹), or the concentration of active reaction intermediates (Equations (5) and (6)) found in the reaction path of CH₄ formation, TOF_ITK (s⁻¹), via Equation (11) [19]. In addition, the effective rate constant, k_eff (s⁻¹), for the methanation reaction is estimated based on Equation (12), which holds when one of the hydrogenation steps of -CH_x/CH_xO intermediates is the RDS of the methanation reaction (hydrogenation and not CO dissociation of CO-s is the RDS).

$$TO{F}_{Chem}=\frac{F_T{y}_{C{H}_4}}{W{N}_{Co,surf}}=\frac{{\theta }_{CHx}}{{\tau }_{CHx}}$$

(10)

$$TO{F}_{ITK}=\frac{F_T{y}_{C{H}_4}}{W\left(N_{CO}+N_{CHx}\right)}$$

(11)

$$k_{eff}=\frac{1}{{\tau }_{CHx}}=k {\theta }_H$$

(12)

N_Co,surf is the total number of surface cobalt per gram of catalyst (μmol g⁻¹) estimated on the basis of the dispersion and loading values of the Co/γ-Al₂O₃ catalyst.

4.3. Transient Isothermal (TIH) and Temperature-Programmed Hydrogenation (TPH) Experiments

Transient hydrogenation experiments (Isothermal and Temperature-Programmed) were performed following the SSITKA step-gas switch: x% ¹³CO/25% H₂/Kr/Ar (230 °C, 7 min) → Ar (3 min) → 50% H₂/Ar, TIH (7 min, 230 °C) → TPH up to 600 °C (β = 10 °C min⁻¹). During the TIH and TPH runs, the mass numbers (m/z) 15, 17, and 18 for ¹²CH₄, ¹³CH₄, and H₂O were continuously monitored by online mass spectrometry. Under TIH at 230 °C, the ¹²CH₄ response reflects the inactive carbon-containing species (named C_β) formed during the non-isotopic ¹²CO/H₂ gas mixture (5 h) and which cannot be exchanged with ¹³CO during the ¹³CO/H₂-SSITKA switch. On the other hand, the ¹³CH₄ response reflects both the adsorbed ¹³CO-s and active ¹³CH_x-s species that are exchanged during the SSITKA switch. Under TPH, the response of ¹²CH₄ is due to adsorbed refractory carbonaceous inactive species (C_xH_y-s, named C_γ) that were not hydrogenated at 230 °C. The transient concentration response curves of ¹²CH₄(t) and ¹³CH₄(t) recorded during TIH/TPH were converted into transient rates (mol g⁻¹ s⁻¹) using Equation (13) (the accumulation term in the Continuous Stirred Tank Reactor (CSTR, ~1.5 mL volume micro-reactor) was found negligible when compared to the outlet rate of CH₄ from the micro-reactor) and after calibration of the MS signal with standard 1.04 mol% ¹²CH₄/Ar and 2.1 mol% ¹³CH₄/Ar gas mixtures.

$$R_{C{H}_4}\left(mol{g}^{-1}{s}^{-1}\right)=\frac{F_{T }{y}_{C{H}_4}}{W}$$

(13)

4.4. Operando DRIFTS—Mass Spectrometry—Transient Isothermal Hydrogenation (TIH) of CO

The reactivity toward hydrogen of the various types of chemisorbed CO-s formed after 2 h of CO/H₂ reaction at 200 °C was determined by operando DRIFTS kinetic experiments. The temperature of 200 °C instead of 230 °C was selected since this allowed the recording of more spectra (lower hydrogenation rate of CO) during this transient experiment. DRIFTS spectra were collected by using a Perkin-Elmer Frontier FT-IR spectrometer equipped with an MCT detector and a high-temperature/high pressure temperature controllable DRIFTS reactor cell (Harrick Scientific, Praying Mantis) with CaF₂ windows. An average spectrum was recorded after 256 scans (resolution of 4 cm⁻¹, scanning speed rate of 2 cm s⁻¹). The sieved and passivated catalyst sample (as described in Section 4.1) was placed in the DRIFTS reactor cell (~90 mg) and its temperature was increased to 425 °C and kept for 2 h in 50 vol% H₂/Ar (50 NmL min⁻¹), which was followed by cooling in H₂/Ar gas flow to the methanation reaction temperature, T = 200 °C, for background spectrum of solid catalyst acquisition. The DRIFTS cell was subsequently purged in Ar gas flow for 20 min and then exposed to the CO/H₂ feed gas stream (5 vol% CO/25 vol% H₂/Ar) for 2 h. At the end of the reaction, an averaged DRIFTS spectrum was collected. The transient kinetics of hydrogenation of individual adsorbed CO-s species formed during the reaction step was investigated by transient isothermal hydrogenation (TIH) in 50 vol% H₂/Ar gas flow. The recorded DRIFTS spectra in the CO region under CO/H₂ and TIH reaction conditions were deconvoluted following the procedure previously described in detail [19]. In the case of a TIH reaction run, where no gas phase IR CO bands were observed, the following criteria for deconvolution were used: (i) FWHM < 100 cm⁻¹, (ii) R² > 0.99, and (iii) IR band center of the CO-s species was allowed for a small shift (< 20 cm⁻¹) to lower wavenumbers (surface coverage effect). By considering fulfillment of the above three criteria, three IR bands of adsorbed CO-s were obtained.

During TIH, the gas-phase responses of CO and CH₄ from the outlet of the DRIFTS reactor cell were continuously monitored by the connected mass spectrometer (MS), which were used for performing the SSITKA-MS and other transient kinetic experiments that followed (TIH/TPH).

5. Conclusions

The present study aimed at providing deeper fundamental understanding on how the CO partial pressure affects important kinetic parameters of the methanation reaction path at 230 °C and 1.2 bar total pressure over a commercially relevant 20 wt% Co-0.05 wt% Pt/γ-Al₂O₃ catalyst. Towards this goal, advanced SSITKA combined with transient isothermal and temperature-programmed hydrogenation experiments as well as transient operando DRIFTS-MS CO hydrogenation to CH₄ coupled with a microkinetic modelling were performed.

For the first time, to the best of our knowledge, the exchange rates of CO-s and CH_x-s related to the SSITKA experiment were estimated via strict application of ¹³C-material balances. The adsorption/desorption rate of CO-s, $R_{CO}^{ex}$, was found to be significantly larger in comparison to the hydrogenation of CH_x-s ($R_{CHx}^{ex}$), and its dependence on the partial pressure of CO was probed. The large influence of CO partial pressure on the formation of different kinds of active CH_x-s during the methanation reaction was suggested by the shape of the transient rate profile of the CH_x-s exchange.

By increasing the CO partial pressure, an increase of θ_CO was observed and this agrees with the characteristic features of the Langmuir isotherm and the fast CO exchange observed. Additionally, by increasing the CO partial pressure, an increase of θ_CHx and a concomitant decrease in S_CH4 were observed, which was linked to the increase of θ_CO/θ_CHx by decreasing the CO partial pressure (competitive chemisorption of H₂ and CO reactants) and the growing rate of -CH_x coupling (increase of θ_CHx).

The amount (μmol/g) and reactivity in hydrogen of the inactive carbonaceous species formed after 5 h of FTS at 230 °C over the Co/γ-Al₂O₃ catalyst was estimated via TIH (C_β) and TPH (C_γ) following SSITKA. The amounts of C_β and C_γ were found to be influenced by the CO partial pressure (36–84 mbar, P_H2 = 300 mbar). In particular, larger amounts of adsorbed active CH_x-s and CO-s were found to react with hydrogen at 230 °C toward CH₄ by increasing the CO partial pressure, as opposed to the amount of inactive carbonaceous species (-C_xH_y, named C_β), which slightly decreased by increasing the CO concentration in the feed. The latter was related to the fact that θ_CO and θ_CHx become larger, which likely reduces the surface coverage of H-s, and, thus, the rate of formation of C_β. The amount of inactive carbonaceous species C_γ was found to significantly increase with increasing CO partial pressure. The C_γ could be due to hydrogen-deficient polymeric carbon and strongly adsorbed hydrocarbon-like species, but not formate (HCOO), atomic, or graphitic carbon.

Based on microkinetic modelling performed on the operando DRIFTS Transient Isothermal Hydrogenation conducted at 200 °C following the methanation reaction (5 vol% CO/25 vol% H₂/Ar, 2 h), the presence of two linear-type adsorbed CO-s species (L₁ and L₂) of a 65:35 (mol/mol) ratio but of very similar activity (k) was revealed. It was concluded that θ_CO had no influence on the hydrogenation activity of linear-type CO-s and on the RDS (HCO-s + H-s → HCOH-s + s) of the considered H-assisted CO hydrogenation mechanism.