The high availability of methane (CH₄) from conversional and unconventional reserves calls for the development of efficient methods for its conversion. Among the various oxidation products from CH₄ conversion, ethanol is an important chemical raw material that has numerous applications such as chemical feedstock, sanitizer, additive and liquid fuel [1-5]. Conversion of CH₄ to ethanol is of great significance for clean and sustainable development. In the conventional processes, CH₄ is firstly reformed into syngas and then converted to ethanol by Fischer-Tropsch synthesis [3-7]. Unfortunately, this process of CH₄ reforming requires high temperature (typically > 800 ℃), resulting in high energy consumption [8]. In this regard, many catalysts (MIL-53 [9], Cu-mordenite zeolite [10], Fe-ZSM-5 [11], Ln₂Zr₂O₇ [12]) are developed to lower the reaction temperature for the CH₄ conversion. However, selective conversion of CH₄ at relatively low temperature is still challenging.

From the perspective of thermodynamically and kinetics, CH₄ can be oxidized to oxygenates at low temperature [13]. According to the calculation of Gibbs free energy, the highest theoretical conversion is near 33% when the equilibrium is reached at room temperature. However, the best selectivity was obtained after maximum conversion is near 5% [14]. The major challenge in the direct selective oxidation of CH₄ is resulted from its large bond dissociation energy (435 kJ/mol), which hinders C-H cleavage reactions. In terms of most of the intermediate products, this high activation barrier means the subsequent oxidation of intermediates is favorable over CH₄ oxidation. For example, with the C-H bond energy of 393 kJ/mol, methanol is easier to be oxidized to stable products of over oxidation (i.e., CO or CO₂) than to oxidize CH₄ itself [2, 15, 16]. As a result, a variety of catalysts such as Cu-Fe/ZSM-5 [8], Au-Pd [17], Au/SiO₂ [18] are attempted to converse CH₄ into oxygenates with high selectivity by suppressing the formation of undesirable over-oxidation products. However, only C₁ compounds as main products were obtained in current CH₄ thermodynamic catalytic processes. Therefore, it is of significance to seek advanced heterogeneous catalyst to enable efficient C-H activation and C-C coupling.

Interestingly, methanotrophic bacteria in nature demonstrate that one-step oxidation of methane to high-added-value products is feasible using methane monooxygenase enzymes. The nature of their active sites provides information for the development of synthetic methane to oxygenates oxidation catalysts [19-21].

Inspired by the methanotrophic bacteria which can one-step oxidize CH₄ to high-added-value products using Cu and/or Fe active species in the enzymes [20, 21], we develop a facile approach to load CuFe₂O₄ on the carbon nanotubes (CNTs). The resultant CuFe₂O₄/CNT is highly active to convert CH₄ at low temperature of 150 ℃. Particularly, 82% selectivity of ethanol can be achieved in a synergetic combination of coordination of Cu and Fe species in CuFe₂O₄ and strong interaction to CNT support. This work provides valuable insights for CH₄ conversion into value-added fuels using non-noble catalysts operating at mild conditions.

In this work, we prepared CuFe₂O₄ catalysts via a co-reduction following by heat treatment method (details see Supporting information) to improve the selectivity of ethanol from CH₄. As shown in Fig. S1 in Supporting information, the diffraction peaks in X-ray diffraction (XRD) pattern can be indexed to copper ferrite (CuFe₂O₄, JCPDS No. 34-0425) [22, 23]. The sample obtained in the absence of Cu and Fe source is identified as Fe₂O₃ and CuO, respectively. To get more detailed information on the crystalline structure of the composite, the transmission electron microscopy (TEM) images of CuFe₂O₄/CNT at different magnifications are displayed in Fig. 1. The low-resolution image (Fig. 1a) confirms that the CuFe₂O₄ particles with diameter of 10–20 nm are well dispersed among the CNT substrate. As displayed in Fig. 1b, two sets of the lattice fringes are observed for the CuFe₂O₄/CNT catalyst. The interplanar distance of 0.25 nm corresponds to the (211) plane of CuFe₂O₄ (JCPDS No. 34-0425), while the fringe spacing of 0.34 nm corresponds to the (002) lattice plane of CNT [24]. The seamless contact of the lattice fringes suggests that CuFe₂O₄ nanoparticles are tightly attached to the CNT to form CuFe₂O₄/CNT composite. The intimate contact usually leads to interaction between metal oxides and support [25-28]. The uniform distributions of Cu, Fe and O elements in the Energy-dispersive detector spectra (Figs. 1c–e) further confirm the formation of CuFe₂O₄ nanoparticles. X-ray photoelectron spectroscopy (XPS) is performed to obtain the information on the surface composition and chemical states of CuFe₂O₄/CNT catalysts. The signals of Cu, Fe, O and C elements are detected in the survey spectra of CuFe₂O₄/CNT (Fig. S2a in Supporting information). For the high resolution Fe 2p XPS spectra of CuFe₂O₄/CNT (Fig. S2b in Supporting information), two peaks at the binding energies of 711.7 eV and 713.5 eV in the realm of Fe 2p_3/2 correspond to tetrahedral Fe³⁺ ions and octahedral Fe³⁺ ions, respectively [29-33]. Concurrently, the peaks referred to the tetrahedral Fe³⁺ ions (724.8 eV) and octahedral Fe³⁺ ions (726.6 eV) are also found in the Fe 2p_1/2 region. Similarly, in the spectra of Cu 2p (Fig. S2c in Supporting information) for CuFe₂O₄/CNT, the peaks at binding energies of 953.02 eV and 933.26 eV are assigned to Cu²⁺ on octahedral sites, while the peaks at binding energies of 955.15 eV and 935.46 eV are assigned to Cu²⁺ on tetrahedral sites [23, 31-34]. Moreover, the spectral profile of CuFe₂O₄/CNT shows 0.3 eV and 0.2 eV shift to the higher binding energies, compared to Fe₂O₃/CNT and CuO/CNT, respectively. The results demonstrate that the chemical states of Fe and Cu in CuFe₂O₄/CNT are totally different from those in Fe₂O₃ and CuO.

The as-prepared catalysts are employed to oxidize CH₄ at low temperature of 150 ℃. As shown in Fig. 2a, ethanol is the main oxidation products for CH₄ oxidation on CuFe₂O₄/CNT catalysts with a major proportion of 82%, accompanying with methanol, acetone and formic acid with selectivity of 7.3%, 7.3% and 3.2%, respectively, which is identified by gas chromatographymass spectrometry (GC-MS) (Fig. S3 in Supporting information). Comparatively, these four oxidation products are also detected for Fe₂O₃/CNT and CuO/CNT (Fig. S4a in Supporting information). In addition, over-oxidation products CO_x converted from CH₄ are also found for Fe₂O₃/CNT. However, the selectivity of ethanol on Fe₂O₃/CNT and CuO/CNT is 13% and 44%, respectively, which is much lower than CuFe₂O₄/CNT (Fig. 2b). Additionally, CNT is inactive for ethanol generation. Thus, the metal species (Fe₂O₃, CuO and CuFe₂O₄) are the active phase for ethanol formation from CH₄. Given that the configuration environment of Cu and Fe species in CuFe₂O₄/CNT is totally different from those in Fe₂O₃/CNT and CuO/CNT, the physical mixtures of Fe₂O₃/CNT and CuO/CNT with the same metal weight percent are also used for CH₄ selective oxidation. Compared to CuFe₂O₄/CNT, the physically mixed catalysts exhibit inferior ethanol selectivity (10%) with more over-oxidation products CO_x (19%) (Fig. 2b and Fig. S4a in Supporting information). The superior catalytic performance of CuFe₂O₄/CNT demonstrate that the coordination of Cu and Fe species in CuFe₂O₄/CNT composite contributes to the highly selective ethanol generation from CH₄ oxidation.

Fig. 3a shows the H₂-temperature programmed reduction (TPR) profiles of CuFe₂O₄/CNT, CuFe₂O₄, Fe₂O₃/CNT and CuO/CNT. There are two peaks at around 174 and 288 ℃ for CuO/CNT, which are assigned to the reduction of the Cu²⁺ to Cu⁺, Cu⁺ to Cu⁰, respectively [35, 36]. There are three peaks at around 323, 499 and 568 ℃ for Fe₂O₃/CNT. The peak at around 323 ℃ is the reduction of Fe₂O₃ into Fe₃O₄, while the broad peaks at around 499 and 568 ℃ are due to the subsequent multiple reduction of Fe₃O₄ to FeO and Fe [30, 37, 38]. Comparatively, there are four peaks at around 188, 248, 395 and 501 ℃ for CuFe₂O₄/CNT. The peak at around 188 ℃ can be ascribed to the reduction of CuFe₂O₄ to Cu⁰ and Fe₂O₃ phases [37], while the peaks at around 248, 395 and 501 ℃ are due to the further reduction of Fe₂O₃ to Fe₃O₄, Fe₃O₄ to FeO and Fe, respectively [38]. Compared to Fe₂O₃/CNT, the Fe species in the CuFe₂O₄/CNT composite exhibit lower reduction temperature, demonstrating a synergistic effect between Cu and Fe species within the CuFe₂O₄/CNT composite, in which Cu promoted the reduction of Fe at a lower temperature [35-39]. The TPR profile changes induced by the interaction between Cu and Fe species indicate that the enhanced oxygen transfer properties and a better electron acceptor for CuFe₂O₄/CNT [40, 41]. Moreover, the spectra profile of CuFe₂O₄/CNT is also different from that of CuFe₂O₄, demonstrating the interaction between CuFe₂O₄ particles and CNT supports [42, 43], which is consistent with TEM results. Thus, the enhanced redox property of CuFe₂O₄/CNT composite would accelerate the activation of CH₄ and further formation of ethanol. From the above analysis, the Cu and Fe species in CuFe₂O₄/CNT composite remarkably affect the ethanol selectivity. Accordingly, the yield of ethanol over the CuFe₂O₄/CNT is calculated as 2.02%. The CH₄ conversion efficiency increases with the higher concentration of Fe species, so the Fe is reasoned to be the main active center for CH₄ oxidation. There are over-oxidation products of CO and CO₂ using Fe₂O₃/CNT as catalysts. The addition of Cu species can decrease the concentration of generated hydroxyl radicals, which are of strong oxidative ability to oxidize the carbon-containing intermediates. This is consistent with the electron paramagnetic resonance (EPR) radical trapping studies on Fe/ZSM-5 and Cu/ZSM-5 [8]. The intermediates during the CH₄ oxidation process is probed by in situ infrared (IR) spectroscopy. As shown in Fig. 3b, no stable surface species are observed over CNT (Fig. S5 in Supporting information), confirming that the CNT alone is inert for CH₄ conversion. Two bands at around 2927 and 2857 cm⁻¹ are observed for the CuFe₂O₄ catalyst, which is attributed to asymmetric and symmetric CH₂ stretching modes, respectively [44-47]. Comparatively, one more peak at around 2963 cm⁻¹ corresponding to CH₃ stretching modes is found for CuFe₂O₄/CNT [47, 48]. This confirms that the strong interaction between CuFe₂O₄ and CNTs facilitates the CH₄ activation.

Accordingly, the possible reaction paths for the formation of ethanol from CH₄ over the CuFe₂O₄/CNT composite catalysts is proposed in Fig. 4. The hydroxyl radicals are usually believed to precipitate in CH₄ oxidation process [8]. On the one hand, CH₄ is firstly activated mainly by Fe species in CuFe₂O₄. The surface CH₃ and CH₂ species are formed and then coupled to form longer alkyl and alkoxy chains speedily, further leading to the formation of oxygenates. The Cu species in CuFe₂O₄ drastically inhibit the further over-oxidation process in the presence of hydroxyl radicals and yield ethanol as the major reaction products. On the other hand, the strong interaction between CuFe₂O₄ and CNT would promote the electrons transfer from CH₄ to catalysts, and oxygen is transferred from catalysts to CH₄ during the oxidation process [8, 41-43]. Therefore, the coordination of Cu and Fe species in the CuFe₂O₄ and the strong interaction to CNT substrate result in highly selective conversion of CH₄ to ethanol on CuFe₂O₄/CNT composites.

In summary, CuFe₂O₄/CNT composite catalysts are synthesized via a co-reduction process. Ethanol with high selectivity of 82% can be directly converted from CH₄ over the CuFe₂O₄/CNT catalysts at low temperature of 150 ℃. The unique synergistic effects of the coordination of Cu and Fe species in the CuFe₂O₄ as well as strong interaction to CNT substrate in CuFe₂O₄/CNT result in supper high selectivity for ethanol formation. This work elucidates that non-noble metals loaded on CNT can pave the way for efficient CH₄ conversion and highly valuable C₂₊ products generation.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

This work was financially supported by the National Natural Science Foundation of China (No. 21975163), Bureau of Industry and Information Technology of Shenzhen (No. 201901171518) and Shenzhen Science and Technology Program (No. KQTD20190929173914967). We also gratefully acknowledge the support provided by Instrumental Analysis Center of Shenzhen University (Xili Campus).

Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2021.07.019