Highly selective photothermal CO₂ hydrogenation to C₂⁺ hydrocarbons over Mn-modified K–Fe catalysts

Abstract

Efficient and selective CO₂ hydrogenation pathways are pivotal for tackling climate change while enabling sustainable energy solutions. Herein, we report a series of potassium-promoted Fe_xMn_y catalysts for the photothermal hydrogenation of CO₂ to C₂⁺ hydrocarbons under mild reaction conditions. By tuning the Fe/Mn ratio, we identify K–Fe₁₀Mn₁ as the optimal composition, exhibiting a remarkable C₂⁺ selectivity of 79.2% and a C_2-5 hydrocarbon yield of 4.8 mmol g⁻¹ h⁻¹ at 250 °C and 20 bar, under 4 W cm⁻² light irradiation. This catalyst also exhibited the lowest CO selectivity (24.1%) and the highest chain-growth probability (α = 0.64) among the series. In contrast, detailed characterization revealed that excessive Mn leads to over-stabilization of Fe₁Mn_1−xO species, diminishing the availability of active Fe₅C₂ species and reducing C–C coupling efficiency. Mechanistic studies, including in situ DRIFTS, demonstrate that non-thermal effects dominate the reaction pathway, with light irradiation accelerating the generation of key intermediates without altering the intrinsic mechanism. These findings provide fundamental insights into the structure–activity relationship of K–FeMn systems and offer a promising route for solar-driven CO₂-to-hydrocarbon conversion under industrially relevant conditions.

---

Broader context

The sustainable transformation of carbon dioxide into energy-rich hydrocarbons represents a cornerstone challenge in achieving a circular carbon economy. Traditional thermal CO₂ hydrogenation requires high temperatures and pressures, leading to significant energy consumption and limited selectivity toward multi-carbon products. Harnessing solar energy to drive this reaction efficiently under mild conditions would constitute a major step forward in carbon utilization. In this work, we introduce a family of Mn-modified K–Fe catalysts capable of converting CO₂ to C₂⁺ hydrocarbons with unprecedented selectivity under photothermal conditions. By rationally tuning the Fe/Mn ratio, we demonstrate that optimal metal synergy enhances C–C coupling and hydrocarbon chain growth while suppressing unwanted CO formation. Mechanistic and in situ spectroscopic studies reveal that light-induced non-thermal effects accelerate key surface reactions without altering the fundamental Fischer–Tropsch pathway. These findings offer mechanistic insight into how photothermal catalysis can overcome the selectivity and efficiency limitations of conventional CO₂ hydrogenation. Beyond its immediate catalytic implications, this work provides design principles for solar-driven CO₂ conversion processes that could enable cleaner, more energy-efficient production of sustainable fuels and chemicals.

1. Introduction

The global energy crisis has become a significant challenge in addressing the increasing demand for energy and mitigating the environmental impacts associated with fossil fuel consumption.^1,2 Fossil fuels, such as coal, oil, and natural gas, continue to dominate the global energy supply due to their high energy density and established infrastructure.³ However, the uneven distribution of fossil fuel reserves across the world exacerbates energy access disparities, particularly in developing regions.⁴ In addition, the combustion of fossil fuels is a major contributor to CO₂ emissions, which significantly accelerates climate change and leads to adverse environmental consequences, including rising global temperatures, extreme weather events, and sea-level rise.^5,6

With escalating CO₂ emissions and a growing global energy crisis, carbon capture, utilization, and storage (CCUS) stands at the forefront of strategies to curb carbon accumulation and advance energy sustainability.^7,8 In this regard, the conversion of CO₂ into valuable chemicals such as carbon monoxide,^9,10 methane,^11,12 methanol,^13–15 and hydrocarbon,^16,17 offers a promising approach not only to mitigate CO₂ levels but also to generate high value-added fuels and chemicals. In particular, the conversion of CO₂ into hydrocarbons through the CO₂-based Fischer–Tropsch synthesis (CO₂-FTS) presents a pathway to close the carbon loop, addressing both environmental concerns and the growing demand for energy and chemical feedstocks. Consequently, the development of energy-efficient and selective CO₂ conversion technologies has proved essential to achieve a sustainable energy future.

In this context, solar energy is considered a clean and abundant resource with the potential to fundamentally transform CO₂ conversion technologies without relying on fossil fuels.^18,19 In fact, when it comes to heterogeneous catalysis, beyond serving as a renewable heat source, solar irradiation has emerged as a key factor to accelerating catalytic processes, modifying the reaction kinetics, and improving the reaction efficiency through photon-induced electronic excitation.^20–22 By simultaneously harnessing light and heat from solar radiation, photothermal catalysis can integrate thermal activation with photon-induced effects, enabling efficient solar-driven chemical transformations^23–25 with lower energy barriers, particularly in CO₂ hydrogenation.^26,27 Nevertheless, although the photothermal CO₂ hydrogenation toward hydrocarbons has been previously investigated, most studies remain limited to ambient pressure conditions,^28–30 mainly due to constraints in reactor configuration, heat management and/or photon collection upon increasing the operating pressure. Therefore, the combination of thermal and photochemical effects under industrially relevant CO₂-FTS conditions remains largely unexplored.

Fe-based catalysts, comprising magnetite (e.g., Fe₃O₄) and carbides (e.g., Fe₅C₂ and Fe₂C), are widely used in CO₂-FTS, where iron oxides facilitate CO formation via RWGS and carbides enable C–C coupling.^31–34 However, unpromoted Fe catalysts often suffer from deactivation and low hydrocarbon selectivity due to phase instability and methanation.^35–38 The incorporation of promoters has proven to be one of the most effective strategies for enhancing the activity and selectivity of Fe-based catalysts.³⁶ For instance, alkali promoters (e.g., K, Na, Cs) could effectively enhance catalytic activity and hydrocarbon selectivity by modulating the electronic structure of iron phases and suppressing undesired hydrogenation pathways.^39–42

Transition metals, including Co,^43,44 Cu,⁴⁵ Mn,^32,46–48 and Zn,^34,49 significantly influence the electronic properties of Fe-based catalysts in the CO₂-FTS reaction. Among these, Mn has attracted considerable attention due to its cost-effectiveness and its ability to enhance hydrocarbon selectivity.^46,47 Recently, Kondratenko and colleagues reported that a Mn-modulated Fe-based catalyst enhanced hydrocarbon selectivity in the CO₂-FTS reaction.⁴⁷ They found that an in situ MnO-containing layer formed on the surface of iron carbide, which reduced hydrogenation activity, thereby inhibiting methanation and promoting hydrocarbon formation. However, previous studies have shown that Mn could enhance the dispersion of FeO_x catalysts, while strong Fe-Mn interactions stabilize the Fe₁Mn_1−xO phase, suppressing the formation of active FeC_x species.^50,51 Overall, an appropriate Fe/Mn ratio plays a crucial role in governing the CO₂-FTS reaction.

Motivated by the challenges and opportunities outlined above, we investigate potassium (K)-promoted Fe_xMn_y catalysts with varying Fe/Mn ratios for photothermal CO₂ hydrogenation, targeting the production of C₂⁺ hydrocarbons in a continuous flow reactor operated under industrially relevant conditions (250 °C, 20 bar). Structure–activity relationship optimization reveals that the K-promoted Fe₁₀Mn₁ catalyst achieves the lowest CO selectivity (24.1%) and the highest chain growth probability (α = 0.64) among the tested K–Fe_xMn_y catalysts, indicating an optimal Fe/Mn ratio. To the best of our knowledge, this catalyst delivers a C₂⁺ hydrocarbon selectivity of 79.2% at 20 bar, with a C_2–5 hydrocarbon yield of 4.8 mmol g⁻¹ h⁻¹, representing state-of-the-art performance for photothermal CO₂ hydrogenation. Mechanistic investigations suggest that non-thermal effects serve as the primary driving force for photothermal CO₂ hydrogenation. Complementary in situ DRIFTS measurements further confirm that light irradiation accelerates the formation of key intermediates, thereby enhancing overall catalytic performance.

2. Catalysts preparation

A series of Fe_xMn_y catalysts, with different Fe/Mn atomic ratios, was prepared via the co-precipitation method using iron(III) nitrate nonahydrate (Fe(NO₃)₃·9H₂O) and manganese(II) nitrate tetrahydrate (Mn(NO₃)₂·4H₂O) as the metal precursors. Typically, appropriate amounts of Fe(NO₃)₃·9H₂O and Mn(NO₃)₂·4H₂O were dissolved in deionized water (100 mL) for a total molar concentration of Fe³⁺ and Mn²⁺ of 0.4 mol L⁻¹. Each mass of Fe(NO₃)₃·9H₂O and Mn(NO₃)₂·4H₂O was determined by the atomic ratio of Fe to Mn (10/1, 7/1, 5/1, 4/1, 3/1, and 2/1). The mixture was stirred at room temperature for 1 h. Then, ammonia solution (NH₄OH; Sigma Aldrich, 25% NH₃ in H₂O) was added dropwise into the above solution under stirring until the pH value reached 10. The suspension was aged for 2 h, filtered and washed until a neutral pH value was achieved. The obtained samples were dried at 80 °C overnight. Subsequently, the dried precursor was impregnated with a solution of potassium nitrate (KNO₃) in water by wetness impregnation to obtain 2 wt%. The K-promoted samples were first dried at 100 °C for 1.5 h and then calcined at 450 °C with a heating rate of 3 °C min⁻¹ for 3 h. They were named K–Fe_xMn_y, where x/y stands for the atomic ratio of Fe to Mn. The catalyst without Mn was also prepared using only Fe(NO₃)₃·9H₂O while other preparation conditions were kept the same, named as K–Fe.

Lastly, the calcined sample was reduced in a flow of 6 vol% H₂/Ar at 400 °C for 2 hours, and then passivated in a flow of 5 vol% O₂/N₂ at room temperature for 1 hour. Before the catalytic testing, the reduced samples were kept in a vial with a seal inside a glove box to prevent further oxidation.

3. Photothermal CO₂ hydrogenation reaction

The photo-thermal CO₂ hydrogenation reaction was evaluated using a commercial flow reactor equipped with a quartz window (Harrick, HVC-MRA-5) and a 300 W Xe lamp (PE300BFA, Perfectlight) as the irradiation source. In a typical experiment, 50 mg of pre-reduced catalyst was loaded on the reactor. A thermocouple to monitor the temperature was located ca. 1 mm below the catalyst's surface. The reaction gas mixture (H₂/CO₂ ratio of 4 : 1) was introduced through the reactor with a total flow rate of 10 mL min⁻¹ (GSHV = 12 000 mL h⁻¹ g⁻¹). Nitrogen was used as an internal standard. The evolved gases were analyzed by a gas compact chromatograph (CompactGC, Thermo Fisher) directly connected to the reactor outlet (Fig. S11).

CO₂ conversion was calculated by the following formula (eqn (1)):

The CO selectivity (S_CO) was given as formula (eqn (2)):

The selectivity towards C_i hydrocarbons (S_{Ci hydrocarbon}) was given as formula (eqn (3)):

4. Results and discussion

Potassium-promoted iron-based catalysts with a range of Fe/Mn atomic ratios (denoted as K–Fe_xMn_y; Table S1) were prepared using a co-precipitation method. Powder X-ray diffraction (PXRD) patterns of the calcined samples (Fig. S1) demonstrated that α-Fe₂O₃ (hematite) was the only crystalline phase observed in catalysts with Fe/Mn ratios between 10 : 1 (K–Fe₁₀Mn₁) and 5 : 1 (K–Fe₅Mn₁). No additional peaks corresponding to potassium are observed in the PXRD patterns, indicating its uniform dispersion within the K–Fe_xMn_y catalysts. Notably, increasing the Mn content led to a pronounced decrease in both crystallinity and crystal size of Fe₂O₃ (Table S2). When the Fe/Mn ratio dropped below 5 : 1, Fe species became predominantly X-ray amorphous. This behavior highlights that Mn as a structural promoter could effectively hinder crystallite growth and promote Fe dispersion.⁴⁶ Such structural modification led to a pronounced enhancement in specific surface area. Moreover, the UV-Vis-NIR absorption spectra of K–Fe_xMn_y oxides show that the monometallic K–Fe catalyst exhibits two distinct absorption bands at 345 and 530 nm, characteristic of α-Fe₂O₃. With increasing Mn content, both bands progressively weaken, further indicating a decrease in crystallinity (Fig. S2). As shown in Table S3 and Fig. S3, the BET surface area increased significantly from 26 m² g⁻¹ for K–Fe to 115 m² g⁻¹ for K–Fe₁₀Mn₁, and further rose to 287 m² g⁻¹ at a Fe/Mn ratio of 7 : 1. However, a subsequent increase in Mn content (Fe/Mn = 2 : 1) resulted in a reduction of surface area to 172 m² g⁻¹.

The reducibility of the K–Fe_xMn_y catalysts was investigated using H₂ temperature-programmed reduction (H₂-TPR, Fig. S4). For K–Fe, two major hydrogen consumption peaks were observed: The first at 422 °C, corresponding to the reduction of Fe₂O₃ to Fe₃O₄, and the second at around 595 °C, attributed to the further reduction to FeO/Fe⁰.^32,46 In contrast, the K–Fe_xMn_y catalysts displayed more complex reduction behavior. A low-temperature peak near 270 °C was assigned to the reduction of MnO₂ to Mn₂O₃,⁵² while a broader peak between 350–370 °C was associated with overlapping reduction events of Fe₂O₃ → Fe₃O₄ and Mn₂O₃ → Mn₃O₄.^50,53 A subsequent peak at approximately 500 °C was attributed to the reduction of Mn₃O₄ to MnO. Interestingly, the incorporation of Mn lowered the temperature required for Fe₃O₄ formation, suggesting the formation of a Fe₁Mn_1−xO solid solution.^47,54 However, the subsequent reduction of Fe₃O₄ to metallic Fe was significantly hindered in Mn-containing samples, with reduction temperatures exceeding 600 °C. This shift implies that the Fe₃O₄/Fe₁Mn_1−xO phase becomes increasingly resistant to reduction as Mn content increases, potentially stabilizing intermediate mixed oxide phases.⁴⁷

Based on H₂-TPR analysis, all catalysts were pre-reduced at 400 °C under a flow of 6 vol% H₂/Ar for 2 h and subsequently passivated under 5 vol% O₂/N₂ at room temperature for 1 h. The phase composition of the reduced catalysts was analyzed by PXRD (Fig. S5). The reduced monometallic K–Fe sample exhibited a mixture of Fe₃O₄ and metallic Fe. In contrast, the reduced K–Fe_xMn_y catalysts showed the emergence of a new diffraction peak at 41.6°, alongside reflections attributed to Fe₁Mn_1−xO. This peak progressively shifted to lower 2θ angles with increasing Mn content. Phase identification revealed that the peak corresponds to Fe_0.899Mn_0.101O (ICSD-60684) in K–Fe₁₀Mn₁ and evolves to Fe_0.497Mn_0.503O (ICSD-60687) in K–Fe₂Mn₁. Moreover, the intensity of metallic Fe diminished with increasing Mn content and disappeared entirely in the K–Fe₂Mn₁ sample. Quantitative XRD analysis confirmed a decrease in metallic Fe and a concurrent increase in Fe₁Mn_1−xO with higher Mn loading (Fig. S6 and Table S4). These findings are consistent with H₂-TPR results, reinforcing that the incorporation of Mn could stabilize Fe-containing mixed oxides against full reduction to metallic Fe. Therefore, a pre-reduction temperature of 400 °C was selected to predominantly form Fe₃O₄ and Fe-Mn mixed oxides, which are active for the RWGS reaction and can subsequently undergo in situ carburization to generate iron carbide species under reaction conditions. There carbide phases are widely recognized as the active phase for C–C coupling and C₂⁺ hydrocarbon formation.^33,55 Moreover, as shown in Fig. S7, the UV-Vis-NIR absorption spectra of the reduced K–Fe_xMn_y catalysts show the complete disappearance of the α-Fe₂O₃ absorption bands compared to the oxidized samples. All reduced catalysts exhibit broad, continuous absorption across the visible to NIR region, indicative of excellent light-harvesting capability and suggesting strong potential for enhanced photothermal and photochemical activity.

Transmission electron microscopy (TEM) analysis of the reduced K–Fe₁₀Mn₁ catalyst (Fig. S8) revealed well-defined lattice fringes with spacings of 0.249, 0.296, and 0.203 nm, corresponding to the (111) plane of Fe₁Mn_1−xO, the (200) plane of Fe₃O₄, and the (110) plane of metallic Fe, respectively, which is consistent with the PXRD results. In addition, elemental mapping (Fig. S9 and S10) confirmed a homogeneous distribution of Mn and K throughout the reduced sample, indicating uniform incorporation of promoter elements within the catalyst matrix.

The catalytic performance and hydrocarbon product distribution for CO₂ hydrogenation over a series of K–Fe_xMn_y catalysts were evaluated under photo-thermal conditions, using a 300 W xenon lamp (light intensity: 4 W cm⁻²) with external heating at 250 °C, a total pressure of 20 bar, and a space velocity of 12 000 mL g⁻¹ h⁻¹ (Fig. S11). The CO₂ conversion and product selectivity reached a steady state within around 15 hours, which can be attributed to the gradual formation of active sites (Fig. S13). The CO₂ conversion and product selectivity were monitored over 35 h of time-on-stream (TOS) for the various K–Fe_xMn_y catalysts (Fig. 1a and Fig. S14 and Table S6). As shown in Fig. 1a and Fig. S14, all catalysts exhibited comparable CO₂ conversions in the range of 44–50%, highlighting the high activity of K-promoted FeMn systems for photo-thermal CO₂ hydrogenation. Notably, the CO selectivity decreased from 37.3% to 24.1% as the Mn content decreased from Fe/Mn = 2/1 to Fe/Mn = 10/1, while the C₂⁺ hydrocarbon selectivity correspondingly rose from 46.8% to 59.6%. This inverse trend in CO and C₂⁺ selectivity correlates with XRD analysis of the reduced catalysts, which reveals that a higher Mn content stabilizes Fe-based oxides and promotes the formation of Fe₁Mn_1−xO solid solutions. Such structural stabilization, arising from the low reducibility of Fe-Mn solid solutions, could reduce the amount and surface accessibility of Fe species available for carburization. Consequently, the formation of iron carbide phases under reaction conditions is suppressed, leading to an attenuation of the Fischer–Tropsch synthesis (FTS) pathway essential for hydrocarbon production. Moreover, CO₂-temperature-programmed desorption (CO₂-TPD) analysis (Fig. S15) revealed that an optimal Mn content acts as an effective structural promoter, enhancing CO₂ adsorption capacity and strengthening the interaction between CO₂ molecules and the catalyst surface.

Fig. 1 Photothermal CO₂ hydrogenation reaction. (a) CO₂ conversion and product selectivity over K–Fe₂Mn₁, K–Fe₅Mn₁, and K–Fe₁₀Mn₁ catalysts at TOS of 35 hours. (b-c) Hydrocarbon distribution and ASF plot over K–Fe₁₀Mn₁ and K–Fe catalyst at TOS of 35 hours. (d) Stability test of K–Fe₁₀Mn₁ catalyst. Reaction conditions: GSHV = 12 000 mL g⁻¹ h⁻¹, H₂/CO₂ = 4, P =20 bar, 4 W cm⁻² light irradiation with external heating at 250 °C. (e) Comparison of the hydrocarbon selectivity of C₂⁺ with reported state-of-the-art catalysts in the photo-thermal CO₂-to-hydrocarbon reaction.

The Anderson–Schulz–Flory (ASF) product distributions for the K-promoted Fe₁₀Mn₁ and monometallic K–Fe catalysts are shown in Fig. 1b and c. Compared to K–Fe, the K–Fe₁₀Mn₁ catalyst exhibited a higher chain growth probability, indicative of more efficient C–C coupling. However, further increasing the Mn content led to a decline in chain growth probability, suggesting that an optimal Fe/Mn ratio is essential for effective CO₂-FTS (Fig. S16 and S17). A balanced Fe/Mn composition not only enhances overall hydrocarbon productivity but also promotes the formation of longer-chain hydrocarbons.

Catalytic stability is a key point for industrial applications. As illustrated in Fig. 1d, the K–Fe₁₀Mn₁ catalyst exhibited remarkable stability over 80 h of time-on-stream (TOS) under photo-thermal conditions, maintaining consistent CO₂ conversion, CO selectivity, and hydrocarbon selectivity without observable degradation. Notably, steady-state performance was achieved after approximately 15 h, likely due to the gradual formation and evolution of active catalytic sites. To mitigate potential surface temperature drops caused by hydrocarbon accumulation, the feed was switched to N₂ after 35 h of reaction for a 3 h purge under identical conditions, effectively removing adsorbed hydrocarbon species from the catalyst surface. These results underscore the structural robustness of the catalyst and its strong potential for long-term solar-driven CO₂ hydrogenation reaction.

In comparison with previously developed photothermal catalysts in the CO₂ hydrogenation reaction, the K-promoted Fe₁₀Mn₁ exhibited exceptional performance, achieving a C₂⁺ hydrocarbon selectivity of 79.2% (CO free) at 20 bar, surpassing previously reported benchmarks for photo-thermal CO₂-to-hydrocarbon conversion (Fig. 1e). Under these conditions, the C_2–5 hydrocarbon yield reached 4.8 mmol g⁻¹ h⁻¹, representing the highest value reported to date for photo-thermal CO₂ hydrogenation. (Table S10) Furthermore, relative to purely thermal catalysis, the K–Fe₁₀Mn₁ catalyst demonstrated outstanding photothermal CO₂ hydrogenation activity under mild reaction temperatures (Table S11).

To gain deeper insight into phase transformations and the role of Mn in Fe catalyst during the photo-thermal CO₂ hydrogenation reaction, a series of characterizations was conducted on the spent K–Fe_xMn_y/K–Fe catalysts after 35 h of time-on-stream. As shown in Fig. 2a and Fig. S18, PXRD analysis revealed distinct diffraction peaks corresponding to Fe₅C₂ (ICSD-423887) and Fe₃O₄ (ICSD-249047) across all K–Fe_xMn_y and K–Fe catalysts. This suggests that Fe₅C₂ is formed in situ from Fe₃O₄ and metallic Fe during the reaction and is responsible for C–C coupling in the FTS pathway.^46,56 In contrast to the monometallic K–Fe catalyst, additional reflections assigned to MnCO₃ (ICSD-80867) and Fe₁Mn_1−xO were observed in the K–Fe_xMn_y samples, with increasing intensity upon higher Mn loading. TEM images of the K–Fe₁₀Mn₁ catalyst after the photothermal reaction (Fig. 2g) confirmed the formation of χ-Fe₅C₂ and MnCO₃ phases, with lattice spacings of 0.205 nm and 0.287 nm corresponding to the (510) and (104) planes, respectively. In addition, elemental mapping (Fig. 2h) revealed a homogeneous distribution of K and Mn throughout the spent catalyst, indicating the structural stability and uniform dispersion of promoter elements under reaction conditions.

Fig. 2 Spent catalyst characterizations. (a) PXRD patterns for spent K–Fe_xMn_y/K–Fe catalysts after photo-thermal CO₂ hydrogenation reaction after TOS = 35 hours at 20 bar under 4 W cm⁻² irradiation with 250 °C external heating. Simulated XRD patterns for Fe₅C₂, Fe₃O₄, Fe₁Mn_1−xO, and MnCO₃ are included for reference. (b) Phase composition of the spent catalysts based on the Rietveld quantitative phase analysis of their PXRD patterns. (c) CO-TPR for spent K–Fe_xMn_y/K–Fe catalysts. (d)–(f) Mössbauer spectra of the spent K–Fe₁₀Mn₁, spent K–Fe, and spent K–Fe₂Mn₁ catalysts. (g) TEM images and (h) STEM images with elemental mapping by EDX of spent K–Fe₁₀Mn₁ catalyst.

To further quantify the phase composition, Rietveld analysis of the XRD patterns was performed (Fig. 2b and Fig. S19 and Table S5). As per these analyses, there is 91% Fe₅C₂ and 9% Fe₃O₄ in the K–Fe catalyst after reaction, consistent with the classical “tandem mechanism” where Fe₃O₄ facilitates the RWGS reaction and Fe₅C₂ drives subsequent hydrocarbon formation via FTS.^35,57 Upon Mn incorporation, however, the fraction of Fe₅C₂ decreased, while Fe₁Mn_1−xO and MnCO₃ increased with decreasing Fe/Mn ratio. For instance, the spent K–Fe₂Mn₁ catalyst contained only 43.0% Fe₅C₂, along with 10.6% Fe₁Mn_1−xO and 41.4% MnCO₃. To further elucidate the evolution of iron species during the photothermal CO₂-FTS reaction, Mössbauer spectroscopy was performed (Fig. 2d–f and Fig. S20 and Table S7). As shown in Fig. 2d–f, the Mössbauer results of spent catalysts are fully consistent with the XRD analysis, confirming that iron-containing oxides are responsible for RWGS activity, whereas Fe₅C₂ is the exclusive carbide phase associated with hydrocarbon formation. Quantitative comparison reveals good agreement in phase composition between Mössbauer and XRD for the spent K–Fe and K–Fe₁₀Mn₁ catalysts. In contrast, Mössbauer spectroscopy indicates a higher fraction of Fe₅C₂ in spent K–Fe₂Mn₁ (79.8% of iron species) than that estimated from XRD (73.5% of iron species), likely due to the presence of amorphous carbide species that are not detectable by XRD (Table S8).³⁵ Overall, the evolution of phase composition with varying Fe/Mn ratios correlates with the observed CO selectivity. Specifically, the diminished fraction of Fe₅C₂ together with the increased abundance of Fe-containing oxide phases at higher Mn contents, which are active in the RWGS reaction, preferentially promotes CO formation while simultaneously suppressing C–C coupling.^47,50,51

To assess the impact of Mn on carburization behavior, CO-temperature-programmed reduction (CO-TPR) measurements were conducted on reduced K–Fe_xMn_y/K–Fe catalysts (Fig. 2c). The main carburization temperatures increased from 351 °C and 485 °C for K–Fe to 370 °C and 575 °C for K–Fe₂Mn₁, indicating that Mn incorporation raises the carburization barrier. This behavior is associated with lower metallic Fe availability in the reduced samples, as confirmed by H₂-TPR results. Together, CO-TPR and H₂-TPR analyses reveal that the Fe₁Mn_1−xO phase enhances the reduction resistance and suppresses Fe carburization. Overall, these findings demonstrate that Mn plays a critical role in modulating the structural evolution and catalytic function of Fe-based catalysts under photothermal CO₂ hydrogenation conditions. An optimal Fe/Mn ratio is thus essential not only for tuning product selectivity but also for promoting C–C coupling and hydrocarbon chain growth via balanced phase composition and redox behavior.

5. Mechanistic study

In photothermal catalysis, the interplay between photo-chemical and thermal effects derived from sunlight plays an important role in determining catalytic activity. Distinguishing these two contributions is therefore critical for elucidating the reaction pathway. To this end, Liu and Everitt introduced a strategy that isolates the non-thermal effects by depositing a thin layer (≤ 1 mm) of Ti₂O₃ (a material with strong light-to-heat conversion capabilities but no intrinsic catalytic activity for CO₂ hydrogenation) on top of the catalyst bed (Fig. S21).⁵⁸ This configuration creates an equivalent thermal gradient under both direct and indirect illumination, enabling a controlled comparison of catalytic performance with and without photon effect access to the catalyst surface.

Using this approach, we evaluated the photothermal behavior of the K–Fe₁₀Mn₁ catalyst under varying light intensities. As shown in Fig. 3d, direct and indirect illumination resulted in comparable surface temperatures, confirming that the Ti₂O₃ overlayer provides similar thermal input while effectively blocking photon-induced excitation. The corresponding CO₂ conversion and product selectivity are shown in Fig. 3a and b. Notably, CO selectivity increased with decreasing light intensity, consistent with lower surface temperatures. Under direct illumination, CO₂ conversion exhibited a linear dependence on light intensity in the range of 2.8–4.0 W cm⁻², suggesting a dominant contribution from non-thermal effects, with a minor thermal component (Fig. 3a and Fig. S22).^23,59 In contrast, under indirect illumination, where only thermal contributions remain, the CO₂ conversion showed a non-linear trend, highlighting the purely thermal-driven effect in this configuration (Fig. 3b and Fig. S23). Moreover, the substantial difference in CO₂ conversion between direct and indirect conditions further underscores the pivotal role of photo-chemical pathways. As detailed in Table S9, this conversion difference decreased from 36.94% to 20.96% as light intensity increased from 2.8 to 4.8 W cm⁻², indicating that the relative contribution of hot-carrier effects diminishes at higher irradiances. This decline is likely from elevated reaction temperatures that enhance charge recombination, thereby suppressing the non-thermal effect. Overall, these observations indicate that the non-thermal contribution plays a predominant role in the reaction pathway. By harnessing this effect, the thermal requirement could be substantially reduced while maintaining high catalytic efficiency. In particular, with an external temperature of only 250 °C and a light intensity of 4.0 W cm⁻², the K–Fe₁₀Mn₁ catalyst achieved high CO₂ conversion compared to other conventional catalysts in thermal CO₂-FTS reaction (Table S11); underscoring the synergistic effect of light and heat in photothermal catalysis, which could compensate for reduced heat input and thus enable energy-efficient operation.

Fig. 3 Direct and indirect illumination experiments. (a) CO₂ conversion and product selectivity over K–Fe₁₀Mn₁ catalyst under direct illumination of different light intensities with 250 °C external heating. (b) CO₂ conversion and product selectivity over K–Fe₁₀Mn₁ catalyst under indirect illumination (cover Ti₂O₃ layer) with different light intensities at 250 °C external heating. (c) Study of the effect of light intensity on the CO₂ conversion over K–Fe₁₀Mn₁ catalyst under direct and indirect illumination. (d) Surface temperature of direct and indirect light illumination under different light intensities over K–Fe₁₀Mn₁. Reaction condition: GSHV = 12 000 mL g⁻¹ h⁻¹, 20 bar, H₂/CO₂ = 4.

Phase analysis of the spent catalysts under both direct and indirect illumination conditions is presented in Fig. S24. Notably, identical phase compositions were observed regardless of illumination, indicating that light exposure does not induce detectable structural changes in the catalyst.

Furthermore, a comparison of product selectivity at similar CO₂ conversion levels, achieved under direct illumination at 3.4 W cm⁻² and indirect illumination at 4.0 W cm⁻², revealed no significant differences in product distribution. These findings suggest that light irradiation does not alter the hydrocarbon selectivity, thereby implying that the reaction follows the same mechanistic pathway under both photothermal and purely thermal conditions.

To gain mechanistic insights into the light-induced CO₂ hydrogenation over the K-promoted Fe₁₀Mn₁ catalyst, we performed in situ diffuse reflectance infrared Fourier transform spectroscopy (in situ DRIFTS) under both dark and illuminated conditions at 10 bar. The experimental procedure is illustrated in Fig. S25. Fig. 4a presents the IR spectra collected under light irradiation in a H₂/CO₂ (4 : 1) mixture, with temperature ramping from ambient to 250 °C, capturing the activation and progression of the photothermal CO₂ hydrogenation reaction. At low temperatures, characteristic bands assigned to surface bicarbonate (HCO₃⁻, 1624 cm⁻¹) and bidentate carbonate (b-CO₃²⁻, 1285 cm⁻¹) were observed, indicating that CO₂ adsorption on the catalyst surface is the initial step.^60,61 As the temperature increased, the intensities of these bicarbonate/carbonate species diminished, accompanied by the emergence of vibrational bands corresponding to formate intermediates (HCOO*, 1559 and 1369 cm⁻¹)^62,63 and adsorbed CO species (*CO, 2076 cm⁻¹),^64,65 signifying the onset of the RWGS reaction. Further temperature elevation led to an increase in the intensity of formate and *CO signals, alongside the appearance of gaseous CO (2174 and 2112 cm⁻¹),⁶⁶ reflecting enhanced RWGS activity. Simultaneously, hydrocarbon-related bands were detected at around 220 °C, including *C=CH (1650, 1624, and 842 cm⁻¹),^46,67,68 and *C–H (896 cm⁻¹),⁶⁹ pointing to the formation of hydrocarbon species via a Fischer–Tropsch synthesis (FTS) pathway. Additionally, thermally stable polydentate carbonates (p-CO₃²⁻) could still be observed until 250 °C, evidenced by bands at 1473, 1456, and 1074 cm⁻¹.^26,60 Overall, those findings confirmed the reaction pathway of photothermal CO₂ hydrogenation.

Fig. 4 In situ DRIFTS experiments. (a) In situ DRIFTS spectra of K–Fe₁₀Mn₁ catalyst under H₂ and CO₂ (4 : 1) at 10 bar with increased temperature from room temperature to 250 °C under light illumination. (b) IR peak intensities of *CH/*CH₃, CO*, and *C=CH species as a function of time under light and dark conditions at 250 °C. (c) In situ DRIFTS spectra of K–Fe₁₀Mn₁ catalyst under H₂ and CO₂ (4 : 1) at 10 bar and 250 °C upon light and dark illumination.

To further understand the reaction pathway of CO₂ hydrogenation under both light and dark conditions, in situ IR spectra were collected at 250 °C and 10 bar for each condition over a 2-hour period. As shown in Fig. 4c, the intensities of key surface species increased steadily under continuous light irradiation, indicating the progressive accumulation of active intermediates. In contrast, upon switching off the light, a marked decrease in all signals was observed, reflecting a rapid decline of these intermediates. Fig. 4b presents the normalized intensities of representative species (*C=CH, *CO, and *CH) as a function of time, clearly demonstrating an abrupt drop in signal intensity when illumination stopped. Importantly, the identical intermediates were detected under both illuminated and dark conditions, indicating that the CO₂ hydrogenation reaction pathway remains unchanged. These results suggest that light irradiation primarily enhances the reaction performance by accelerating the formation of reactive intermediates, rather than altering the reaction mechanism. Photocurrent measurements for the reduced K–Fe₁₀Mn₁ catalyst further confirmed the photochemical properties. As shown in Fig. S26, the photocurrent intensity increases progressively with higher light intensity, indicating enhanced generation and separation of photogenerated charge carriers. To confirm the origin of the photoactivity and the role of Mn, we conduct additional photocurrent measurements of the K–Fe₁₀Mn₁ catalyst and a Mn-free sample (K–Fe). As it can be seen in Fig. S27, the photocurrent intensity decreases upon the introduction of Mn. This observation could be explained by (i) a lower proportion of the photoactive phase (Fe₃O₄) and/or (ii) the transfer of photogenerated charges from Fe₃O₄ to Mn. To confirm this hypothesis, we performed steady-state photoluminescence (PL) measurements. The results confirmed that, upon Mn incorporation, the PL intensity of K–Fe₁₀Mn₁ is markedly reduced compared to monometallic K–Fe, indicating enhanced charge separation and transfer between Fe and Mn. (Fig. S28) Overall, those findings are consistent with observations from the direct/indirect illumination experiments, supporting the idea that light enhances catalytic activity through a predominant non-thermal contribution, in combination with a thermal effect.

6. Conclusions

In summary, we developed a series of K–Fe_xMn_y catalysts for the photothermal CO₂ hydrogenation to hydrocarbons, systematically elucidating the role of Mn incorporation into Fe-based systems. An optimal Mn content could enhance both hydrocarbon productivity and chain-growth probability, whereas excessive Mn reduced reduction ability and carburization process, suppressing C–C coupling and increasing CO selectivity. The K–Fe₁₀Mn₁ catalyst, with a finely tuned Fe/Mn ratio, achieved a C₂⁺ hydrocarbon selectivity of 79.2% and a C_2–5 yield of 4.8 mmol g⁻¹ h⁻¹ under 4 W cm⁻² light irradiation with external heating at 250 °C, 20 bar, and 12 000 mL g⁻¹ h⁻¹. Mechanistic investigations revealed that non-thermal effects dominate the reaction pathway, accompanied by a minor thermal contribution, enabling exceptional performance under mild conditions. In situ DRIFTS further demonstrated that light irradiation could accelerate the formation of key reaction intermediates without altering the fundamental pathway, thereby enhancing overall catalytic efficiency.

Conflicts of interest

There are no conflicts to declare.

Data availability

All data supporting the findings of this study are available within the article and its supplementary information (SI). Supplementary information is available. See DOI: https://doi.org/10.1039/d6ey00013d.

Additional datasets and raw characterization files are available from the corresponding authors upon reasonable request.

Acknowledgements

We gratefully acknowledge the financial support from King Abdullah University of Science and Technology (KAUST).

References

1. P. Achakulwisut, P. Erickson, C. Guivarch, R. Schaeffer, E. Brutschin and S. Pye, Nat. Commun., 2023, 14, 5425.
2. D. Zhang, C. Rong, T. Ahmad, H. Xie, H. Zhu, X. Li and T. Wu, Eng. Rep., 2023, 5, e12584.
3. Q. Hassan, P. Viktor, T. J. Al-Musawi, B. M. Ali, S. Algburi, H. M. Alzoubi, A. K. Al-Jiboory, A. Z. Sameen, H. M. Salman and M. Jaszczur, Renewable Energy Focus, 2024, 48, 100545.
4. I. Overland, J. Juraev and R. Vakulchuk, Renewable Energy, 2022, 200, 379–386.
5. M. Filonchyk, M. P. Peterson, L. Zhang, V. Hurynovich and Y. He, Sci. Total Environ., 2024, 935, 173359.
6. D. Nong, P. Simshauser and D. B. Nguyen, Appl. Energy, 2021, 298, 117223.
7. N. Mac Dowell, P. S. Fennell, N. Shah and G. C. Maitland, Nat. Clim. Change, 2017, 7, 243–249.
8. P. Gabrielli, M. Gazzani and M. Mazzotti, Ind. Eng. Chem. Res., 2020, 59, 7033–7045.
9. M. Ahmadi Khoshooei, X. Wang, G. Vitale, F. Formalik, K. O. Kirlikovali, R. Q. Snurr, P. Pereira-Almao and O. K. Farha, Science, 2024, 384, 540–546.
10. X. Wang, E. Toshcheva, A. Rendón-Patiño, C. Martín, U. H. Bhatti, D. Mateo, R. Ahmad, M. H. Alabsi, L. Cavallo and J. M. R. Gallo, Chem. Eng. J., 2025, 507, 160326.
11. D. Schmider, L. Maier and O. Deutschmann, Ind. Eng. Chem. Res., 2021, 60, 5792–5805.
12. C. Molinet-Chinaglia, S. Shafiq and P. Serp, ChemCatChem, 2024, 16, e202401213.
13. O. Martin, C. Mondelli, D. Curulla-Ferré, C. Drouilly, R. Hauert and J. Pérez-Ramírez, ACS Catal., 2015, 5, 5607–5616.
14. O. Martin, A. J. Martín, C. Mondelli, S. Mitchell, T. F. Segawa, R. Hauert, C. Drouilly, D. Curulla-Ferré and J. Pérez-Ramírez, Angew. Chem., 2016, 128, 6369–6373.
15. T. Pinheiro Araújo, J. Morales-Vidal, T. Zou, M. Agrachev, S. Verstraeten, P. O. Willi, R. N. Grass, G. Jeschke, S. Mitchell and N. López, Adv. Energy Mater., 2023, 13, 2204122.
16. F. Jiang, B. Liu, S. Geng, Y. Xu and X. Liu, Catal. Sci. Technol., 2018, 8, 4097–4107.
17. A. Ramirez, L. Gevers, A. Bavykina, S. Ould-Chikh and J. Gascon, ACS Catal., 2018, 8, 9174–9182.
18. K. Ukoba, K. O. Yoro, O. Eterigho-Ikelegbe, C. Ibegbulam and T.-C. Jen, Heliyon, 2024, 10, e28009.
19. A. O. Maka and J. M. Alabid, Clean Energy, 2022, 6, 476–483.
20. X. Wu, S. Zhang, S. Ning, C. Yang, L. Li, L. Tang, J. Wang, R. Liu, X. Yin and Y. Zhu, Chem. Sci., 2025, 16, 4568–4594.
21. S. Luo, X. Ren, H. Lin, H. Song and J. Ye, Chem. Sci., 2021, 12, 5701–5719.
22. D. Mateo, E. V. Ramos-Fernandez and J. Gascon, Chem. Catal., 2025, 5, 101569.
23. D. Mateo, J. L. Cerrillo, S. Durini and J. Gascon, Chem. Soc. Rev., 2021, 50, 2173–2210.
24. C. Song, Z. Wang, Z. Yin, D. Xiao and D. Ma, Chem. Catal., 2022, 2, 52–83.
25. Y. Li, X. Pei, Z.-J. Wang, L. Shi, H. Song and J. Ye, ACS Sustainable Chem. Eng., 2024, 12, 17069–17097.
26. B. Deng, H. Song, Q. Wang, J. Hong, S. Song, Y. Zhang, K. Peng, H. Zhang, T. Kako and J. Ye, Appl. Catal., B, 2023, 327, 122471.
27. X. Wang, A. Rendón-Patiño, J. M. R. Gallo, D. Mateo and J. Gascon, J. Mater. Chem. A, 2024, 12, 23541–23550.
28. S. Ning, J. Wang, X. Wu, L. Li, S. Zhang, S. Chen, X. Ren, L. Gao, Y. Hao and C. Lv, Adv. Funct. Mater., 2024, 2400798.
29. S. Ning, X. Wu, H. Song, X. Ma, S. Yue, S. Zhang, L. Tang, R. Liu, X. Yin and S. Ouyang, J. Am. Chem. Soc., 2025, 148, 1728–1740.
30. J. Zhao, R. Shi, G. I. Waterhouse and T. Zhang, Nano Energy, 2022, 102, 107650.
31. L. Krausser, Q. Yang and E. V. Kondratenko, ChemCatChem, 2024, 16, e202301716.
32. B. Yao, T. Xiao, O. A. Makgae, X. Jie, S. Gonzalez-Cortes, S. Guan, A. I. Kirkland, J. R. Dilworth, H. A. Al-Megren and S. M. Alshihri, Nat. Commun., 2020, 11, 6395.
33. J. Wei, Q. Ge, R. Yao, Z. Wen, C. Fang, L. Guo, H. Xu and J. Sun, Nat. Commun., 2017, 8, 15174.
34. K. Zhu, X. Liu, H. Tang, S. Tian, J. Xie, L. Zeng, T. Wang, H. Li, M. Wang and D. Ma, ACS Catal., 2024, 15, 800–808.
35. J. Zhu, P. Wang, X. Zhang, G. Zhang, R. Li, W. Li, T. P. Senftle, W. Liu, J. Wang and Y. Wang, Sci. Adv., 2022, 8, eabm3629.
36. J. Wei, R. Yao, Y. Han, Q. Ge and J. Sun, Chem. Soc. Rev., 2021, 50, 10764–10805.
37. R.-P. Ye, J. Ding, W. Gong, M. D. Argyle, Q. Zhong, Y. Wang, C. K. Russell, Z. Xu, A. G. Russell and Q. Li, Nat. Commun., 2019, 10, 5698.
38. Y. Zhang, C. Cao, C. Zhang, Z. Zhang, X. Liu, Z. Yang, M. Zhu, B. Meng, J. Xu and Y.-F. Han, J. Catal., 2019, 378, 51–62.
39. W. Yoon, H. Jo, S. Ahmed, M. K. Khan, M. Irshad, J. Lee, S. S. Bibi and J. Kim, Chem. Eng. J., 2024, 495, 153617.
40. Y. Han, C. Fang, X. Ji, J. Wei, Q. Ge and J. Sun, ACS Catal., 2020, 10, 12098–12108.
41. A. S. Skrypnik, S. A. Petrov, V. A. Kondratenko, Q. Yang, A. A. Matvienko and E. V. Kondratenko, J. Catal., 2023, 425, 286–295.
42. Q. Yang, V. A. Kondratenko, S. A. Petrov, D. E. Doronkin, E. Saraçi, H. Lund, A. Arinchtein, R. Kraehnert, A. S. Skrypnik and A. A. Matvienko, Angew. Chem., Int. Ed., 2022, 61, e202116517.
43. N. Liu, Q. Fan, J. Wei, G. Zhang, J. Sun, W. Li, C. Song and X. Guo, ACS Catal., 2024, 15, 179–192.
44. L. Zhang, Y. Dang, X. Zhou, P. Gao, A. P. van Bavel, H. Wang, S. Li, L. Shi, Y. Yang and E. I. Vovk, Innovation, 2021, 2, 100170.
45. P. Sai Prasad, J. W. Bae, K.-W. Jun and K.-W. Lee, Catal. Surv. Asia, 2008, 12, 170–183.
46. H. Ren, H. Yang, J. Xin, C. Wu, H. Wang, J. Zhang, X. Bu, G. Yang, J. Li and Y. Sun, Appl. Catal. B, 2024, 358, 124440.
47. Q. Yang, E. A. Fedorova, D.-B. Cao, E. Saraçi, V. A. Kondratenko, C. R. Kreyenschulte, H. Lund, S. Bartling, J. Weiß and D. E. Doronkin, Nat. Catal., 2025, 1–12.
48. K. Wang, Z. Li, T. Liu, W. Gao, T. Yang, K. Liu, X. Gao, Q. Ma, J. Zhang and T. Zhao, ACS Catal., 2024, 14, 17469–17479.
49. L. Guo, J. Li, Y. Cui, R. Kosol, Y. Zeng, G. Liu, J. Wu, T. Zhao, G. Yang and L. Shao, Chem. Commun., 2020, 56, 9372–9375.
50. Z. Yang, Z. Zhang, Y. Liu, X. Ding, J. Zhang, J. Xu and Y. Han, Appl. Catal., B, 2021, 285, 119815.
51. B. Shi, Z. Zhang, Y. Liu, J. Su, X. Liu, X. Li, J. Wang, M. Zhu, Z. Yang and J. Xu, J. Catal., 2020, 381, 150–162.
52. B. Chen, B. Wu, L. Yu, M. Crocker and C. Shi, ACS Catal., 2020, 10, 6176–6187.
53. X. Ding, M. Zhu, Y.-F. Han and Z. Yang, J. Catal., 2023, 417, 213–225.
54. G. Maiti, R. Malessa, U. Löchner, H. Papp and M. Baerns, Appl. Catal., 1985, 16, 215–225.
55. R. W. Dorner, D. R. Hardy, F. W. Williams and H. D. Willauer, Catal. Commun., 2011, 15, 88–92.
56. L. Tang, L. He, Y. Wang, B. Chen, W. Xu, X. Duan and A.-H. Lu, Appl. Catal., B, 2021, 284, 119753.
57. D. Ye, W. Tang, T. Zhang, L. Lv, Z. Zou, R. K. Gupta and S. Tang, Colloids Surf., A, 2022, 654, 130145.
58. X. Li, H. O. Everitt and J. Liu, Nano Res., 2019, 12, 1906–1911.
59. M. J. Landry, A. Gellé, B. Y. Meng, C. J. Barrett and A. Moores, ACS Catal., 2017, 7, 6128–6133.
60. K. Pokrovski, K. T. Jung and A. T. Bell, Langmuir, 2001, 17, 4297–4303.
61. E.-M. Köck, M. Kogler, T. Bielz, B. Klötzer and S. Penner, J. Phys. Chem. C, 2013, 117, 17666.
62. T. Tagawa, N. Nomura, M. Shimakage and S. Goto, Res. Chem. Intermed., 1995, 21, 193–202.
63. M.-Y. He, J. White and J. Ekerdt, J. Mol. Catal., 1985, 30, 415–430.
64. B. Yan, B. Zhao, S. Kattel, Q. Wu, S. Yao, D. Su and J. G. Chen, J. Catal., 2019, 374, 60–71.
65. G. Blyholder, J. Phys. Chem., 1964, 68, 2772–2777.
66. F. C. Meunier, Catal. Today, 2023, 423, 113863.
67. S. Ning, J. Wang, X. Wu, L. Li, S. Zhang, S. Chen, X. Ren, L. Gao, Y. Hao and C. Lv, Adv. Funct. Mater., 2024, 34, 2400798.
68. M. Gao, H. Li, J. Yu, M. Ye and Z. Liu, AIChE J., 2023, 69, e17881.
69. A. Lagutschenkov, A. Springer, U. J. Lorenz, P. Maitre and O. Dopfer, J. Phys. Chem. A, 2010, 114, 2073–2079.

---