The lower olefins, which generally refer to ethylene, propylene and butylene, are basic building blocks in the chemical industry and are traditionally produced from thermal or catalytic cracking of naphtha and oil in the refining process ^[1]. With the rapid depletion of the earth’s limited petroleum reserves, there is an urgent need for processes that can produce lower olefins from alternative feedstocks. Some of the alternative routes to the lower olefins include the dehydrogenation of lower alkanes, syngas-based processes and specific processes for particular target products, such as the production of ethylene by dehydration of ethanol derived from renewable sources. The processes used for the production of lower olefins from syngas can be divided into two main groups: indirect processes ^[2], which require the synthesis of an intermediate such as methanol or dimethyl ether (known as methanol to olefins (MTO) or dimethyl ether to olefins (DMTO)), and direct processes. Identifying a suitable direct process is one of the most challenging topics in the field of C1 chemistry, but such a process is highly desirable owing to its simplified operation and low energy consumption (Fig. 1) ^[3].

Fig. 1

Fig. 1. Various processes for the production of lower olefins from syngas(figure taken from Ref. [3]).

To simplify the industrial MTO process for the direct production of lower olefins from syngas, scientists have focused on developing a bifunctional catalyst system that couples the methanol synthesis and the C-C coupling reaction. Because the MTO reaction has extremely low activity at low temperatures, the methanol synthesis in the bifunctional system needs to be carried out at high temperatures. As a result, high-temperature methanol synthesis catalysts, such as ZnZr or ZnCr, are chosen as one part of the bifunctional catalyst. Recently, Bao’s group ^[4] reported that ZnCrO _x /MSAPO exhibited very high selectivity toward lower olefins under conditions of 400 ℃, 2.5 MPa and a H₂/CO ratio of 1.5. A C_2-4⁼ selectivity of 80% and C_2-4 of 94% (based on CO₂ excluding that in the products) was obtained with a CO conversion of 17%. It was suggested that the partially reduced ZnCrO _x oxide surface activated the CO and H₂ to form a ketene, which was subsequently converted into lower olefins in the zeolite pores. Wang’s group ^[5] also reported that coupling the methanol synthesis and the MTO reaction using a bifunctional catalyst could be used for the direct conversion of syngas to lower olefins with high selectivity. Under conditions of 400 ℃, 1 MPa and a H₂/CO ratio of 2, selectivity for the lower olefins reached 74% with a CO conversion of 11%. However, neither of these processes is yet in commercial use.

The Fischer-Tropsch-to-olefins (FTO) reaction is very similar to the typical Fischer-Tropsch (FT) process, and FT catalysts can be used for the FTO reaction after suitable modification. Fe-based FTO catalysts are often studied, owing to their low cost and the high olefin content in the final product, and some promising results have been achieved ^[6-8]. de Jong’s group ^[6] reported outstanding results using Fe-based catalysts promoted by S and Na and supported on α -alumina or carbon nanofibers; these catalysts exhibited a selectivity for lower olefins of up to 61%. However, these reactions usually require temperatures of 300-350 ℃, which cause catalyst sintering and carbon deposition on the catalyst surface, and thus deactivate the catalyst and shorten its lifetime ^[9]. In addition, the hydrocarbons obtained with the FTO process typically follow the so-called Anderson-Schulz-Flory (ASF) distribution, which is characterized by a maximum C_2-4 hydrocarbon fraction of about 56.7% and an undesired methane fraction of about 29.2% ^{[10, 11]}. To simultaneously achieve high selectivity for lower olefins, low methane selectivity and high stability, it is therefore necessary to develop new FTO catalysts that deviate from the ASF distribution and operate under mild reaction conditions.

It is widely accepted that metallic Co preferentially produces heavy hydrocarbons with high selectivity for C₅₊ paraffins ^{[12, 13]}. The formation of Co₂C is considered to be one of the main reasons for deactivation of the Co-based FT reaction because it possesses very low activity for CO hydrogenation, with methane and CO₂ as the main products (Fig. 2) ^[14-18]. However, Zhong and Sun’s group ^[17] from the CAS Key Laboratory of Low-Carbon Conversion Science and Engineering now report that, under mild reaction conditions (250 ℃, 0.1-0.5 MPa), Co₂C nanoprisms catalyze syngas conversion with high selectivity for the production of lower olefins. The process generated very little methane (about 5.0%), and the ratio of olefin/ paraffin among the C₂-C₄ products was as high as 30. The product distribution deviated markedly from the classical ASF distribution, and showed the highest selectivity for propylene. The catalyst also showed good stability under long-term operation; no obvious deactivation was detected after 600 h.

Fig. 2

Fig. 2. Reaction selectivity in the Fischer-Tropsch process (figure taken from Ref. [18]).

This surprising result, which is quite different from the industrial Co-catalyzed FT reaction using metal Co nanoparticles and Co2C sphere-like nanoparticles, is attributed to the specific structure and morphology of the Co₂C nanoprisms. Based on structural characterization and DFT calculations, it was suggested that there is a strong facet effect for the Co₂C nanoparticles during syngas conversion. Specifically, the (101) and (020) facets of Co₂C promote the production of olefins and inhibit the formation of methane (Fig. 3).

Fig. 3

Fig. 3. Transmission electron microscopy (TEM) images of the CoMn catalysts after reaching a steady state. (a, b) Low-resolution TEM images; (c-e) High-resolution images of Co2C nanoprisms with exposed (101), (−101) and (020) facets. d, distance (length) of the lattice fringes. (f) The Co2C nanoprism has a parallelepiped shape, with four rectangular faces and two rhomboid faces (figure taken from Ref. [17]).

Zhong’s finding ^[17] is particularly surprising because the spherical Co₂C crystallites have little activity towards olefin formation in this reaction, and because our results showed that the formation of spherical Co₂C and the interface with the Co metal are essential for the production of alcohols ^[16]. However, according to density functional theory (DFT) calculations, non-dissociative or dissociative adsorption of CO on the Co2C nanoprisms is regrettably not involved in Zhong’s work ^[17].

This is a groundbreaking contribution that will deepen the understanding of the FT process and can be considered a good example of nanocatalysis with a facet effect. The high activity, selectivity and stability under mild reaction conditions suggest that these catalysts have immense potential industrial applications. As Prof. Michael Claeys ^[18] has said, “their potential impact cannot be overestimated: they might open up pathways for the development of greatly improved systems for producing valuable chemicals from a variety of carbon sources” and “Zhong et al. ^[17] have thrown open the reaction’s treasure chest, and added fresh momentum to research into methods for making olefins from synthesis gas”.