In a recent article in
Nature Energy, Zhang and He from Soochow University, Ozin from the University of Toronto, and their colleagues reported on a partially substituted perovskite that facilitates the efficient photocatalytic dehydrogenation of ethane [
1]. This material represents another prototypical example of surface frustrated Lewis pairs (SFLPs) in heterogeneous photocatalysis. The concept of SFLPs is based on the original idea of ‘frustrated Lewis pairs’ (FLPs) — sterically hindered Lewis acids and bases that cannot form the classical adducts [
2]. While FLPs are well-established in homogeneous catalysis, their surface counterparts, SFLPs, have found important applications in heterogeneous catalysis using solid-state materials [
3]. SFLP-based catalysts often consist of (bi)metallic oxyhydroxide/oxides/hydroxides, such as In
2O
3–x(OH)
y, CoGeO
2–x(OH)
y, Bi
xIn
2–xO
3, ZnSn(OH)
6, Cu
2O/CeO
2, modified CeO
2, and MOFs [
4‒
7]. More recently, materials like wurtzite-structured GaN, ZnO, and AlP have also been identified as exceptions. SFLPs are known to activate small molecules such as CO
2, H
2, alkenes, dienes, and alkynes, and have gained particular CO
2 reduction [
4‒
9]. However, the application of SFLPs in C−H activation of ethane via photocatalysis has been less explored. Therefore, this work opens new possibilities for utilizing SFLPs in catalysis beyond CO
2 hydrogenation [
10,
11].
Ethylene is an essential commodity chemical and the most-consumed olefin building block. Its production through non-oxidative dehydrogenation of ethane offers greater carbon utilization potential than other methods, but is still limited by high fossil energy consumption. To address this challenge, isomorphic B-site-substituted lanthanum copper manganese oxide perovskites (LaMn
1–xCu
xO
3, or LMCO) were demonstrated as efficient carriers of SFLPs. These materials effectively utilize photogenerated electrons and holes to reduce the activation barrier and minimize energy consumption. In the series of samples with Mn(IV) sites readily substituted by Cu(II), more OH and oxygen vacancies (O
v) were observed compared to pristine LaMnO
3. Specifically, with a 10% Cu substitution (LMCO-10), the number of oxygen vacancies nearly doubled. The substitution also increased the Lewis basicity of the Mn–OH site in the SFLP, with an adjacent Mn atom acting as a Lewis acid, as shown by the DFT simulations. Upon photo-excitation, this SFLP site should promote heterolytic activation of C
2H
6 [
12].
Photocatalytic tests in a Harrick cell reactor confirmed the superior performance of Cu-substituted LMCO samples. LMCO-10 exhibited the best performance, with a C2H4 production rate of 1106.5 μmol/(g·h), and a selectivity of 91.0%, significantly higher than pristine LMCO-0 and comparable to those in the latest reports (Fig.1(a)). In contrast, when Cu was deposited on the surface rather than incorporated in the lattice, the activity was much lower, highlighting the importance of isomorphic substitution. Furthermore, the light enhancement in the C2H4 production rate at various light intensities was more pronounced for LMCO-10 than for pristine LMCO, even though the light-induced heating effect was nearly identical for both (Fig.1(b)). The activation energies derived for both samples under light were significantly lower than those measured under purely thermal conditions, confirming the contribution of the photochemical process (Fig.1(c)). However, while LMCO-10 exhibited a high C2H4 production rate, coke formation led to a decrease in stability over time (Fig.1(d)). To address this issue, the authors applied a simple air oxidation step under light for regeneration, which effectively recovered most of the activity.
In situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) was employed to investigate the reaction pathway and to identify the potential cause of the superior performance of LMCO-10 compared to pristine LMCO-0 (Fig.2(a) and Fig.2(b)). Upon light irradiation, diagnostic peaks for Mn–H at 1845 cm−1, and for –OH and –OH2 vibrations at 3730 and 3700 cm−1, appeared for LMCO-10. In contrast, no such signals were observed for LMCO-0 under the same conditions. These findings suggest the presence of abundant SFLPs on LMCO-10, facilitating the heterolytic dissociation of C2H6. What distinguishes this work from conventional photocatalysis research is the use of solid-state 1H NMR spectroscopy. The authors identified a lengthened Mn−H bond after reaction under light, indicated by a positive chemical shift (Fig.2(c)). This result strongly supports the hypothesis of a photochemically enhanced reaction process.
Theoretical calculations further corroborate these findings, showing that ethane adsorbs more favorably on LMCO-10 than on LMCO-0. This, along with the experimental results, led to the proposed reaction pathway. Upon photo-excitation, photogenerated electrons and holes localize at the Lewis acid Mn and Lewis base Mn–OH sites, respectively, promoting ethane activation. As shown in Fig.2(d), the α-hydrogen bonds to the Lewis base –OH site, forming Mn–OH
2 [
13], while the remaining C
2H
4 bonds to the Lewis acid Mn site. Subsequently, the β-hydrogen is abstracted by an adjacent Mn site, forming the ethylene product. Finally, the remaining proton and hydride combine to generate H
2.
The authors also demonstrated a prototype setup on a rooftop for the photocatalytic conversion of ethane to ethylene. However, solar intermittency remains a key challenge for industrial implementation. To address this, the authors proposed using a 24-7 LED powered by electricity as an alternative light source. A techno-economic analysis (TEA) was conducted comparing the two approaches, revealing that the selling prices of photo-catalytically produced ethylene are higher than the current market price for commercial ethylene (0.7–1.5 $/kg) (Fig.2(e)). With further optimization, the LED approach could significantly reduce the minimum selling price in the long term, bringing it closer to the market price, thus improving the commercial viability of the process.
Overall, this work provides valuable insights into the role of SFLPs in perovskite materials, enhancing the direct dehydrogenation of ethane into ethylene under light. The advantage of this approach lies in its sustainability, as it eliminates the need for external heating and relies on LED or outdoor sunlight as the energy source. The ethylene production rate of 1.1 mmol/(g·h) demonstrates the promise of the process. The TEA suggests that further optimization of both the process and catalyst performance is necessary for market competitiveness. Future studies should focus on improving the optical properties of the reactor and enhancing the photon-to-product efficiency [
14,
15].
Despite the feasibility of the catalyst for regeneration, its rapid decay within hours is far from industrial standard. Additional anti-coking designs for the catalytic material or process are needed to address this issue. Moreover, although structures with SFLPs have demonstrated stable production rates for reactions at relatively low temperatures (e.g., water splitting and CO2 hydrogenation to methanol, typically performed below 300 °C), maintaining their stability at higher temperatures remains a challenge. New strategies are required to either lower the temperature requirements for these reactions or ensure that a sufficient number of SFLPs remain active for sustained reaction rates.
While this paper did not specifically mention the term of “photothermal” in the title, the approach is driven by both the photochemical process and the focused-light-induced heating. The success demonstrated in this paper could help promote the broader use of photothermal approaches for sustainable, low-carbon-footprint energy catalysis [
16]. Furthermore, the insights into SFLPs chemistry could be extended to other perovskites and more versatile oxides, enabling vital chemical conversion processes.
PDF
(1965KB)
