CSPM EDP Sciences logo
Submit your paper
Search
Menu
All issues Volume 2 / No 2 (2023) Natl Sci Open, 2 2 (2023) 20220044 Full HTML
Issue
Natl Sci Open
Volume 2, Number 2, 2023
Special Topic: Chemistry Boosts Carbon Neutrality
Article Number 20220044
Number of page(s) 13
Section Chemistry
DOI https://doi.org/10.1360/nso/20220044
Published online 17 February 2023

© The Author(s) 2023. Published by China Science Publishing & Media Ltd. and EDP Sciences.

Licence Creative CommonsThis is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

INTRODUCTION

The electrochemical reduction of CO2 (CO2RR) into value-added fuels and chemical feedstock offers a sustainable route to store renewable electricity and to mitigate carbon emissions [1–7]. Currently, copper (Cu) is the most promising material capable of catalyzing C–C coupling to yield more valuable multicarbon (C2+) hydrocarbons and oxygenates [8,9]. However, a bulky flat Cu generally favors CH4 production and generates significant amounts of H2 from competitive hydrogen evolution reaction (HER), leading to unsatisfactory activity and selectivity for C2+ products [10,11]. To surmount this challenge, many strategies have been exploited to improve catalytic performance of Cu, including facet control [12], morphology manipulation [13], surface modification [14], and structural reconstruction [15,16]. Among these emerging strategies, structural reconstruction is of particular interest because it is expected to in-situ introduce structural defects such as vacancies [17], grain boundaries (GBs) [16], and dislocations [18], which contain rich undercoordinated sites and thus are believed to facilitate intermediates adsorption for catalyzing CO2RR efficiently [16,18–21]. However, designing advanced catalysts to achieve high-density defects via reconstruction and ultimately guide the reaction pathway remains challenging. This, to a great extent, is due to the limited investigations on the various catalysts and their reconstruction process [22].

Perovskite-type oxides possess flexible electronic structure and chemical versatility, which thus hold great potential to create highly active sites via structural reconstruction in the heterogeneous catalysis and electrocatalysis field. For example, derived Cu+ sites anchored on LaMn1−xCuxO3 (0.2<x<0.8) matrix exhibited remarkable selectivity for catalyzing CO hydrogenation towards alcohols [23]. SrIrO3 material has demonstrated the superior oxygen evolution reaction activity and stability in acidic electrolyte by in-situ forming the highly active IrOx overlayer via Sr leaching [24]. Although intensive interest has been focused on the production of high-value multicarbon products from CO2RR electrochemistry, the use of perovskite-type oxides-derived Cu as CO2RR catalyst is rare, except for a few reports [25,26]. In 1993, Sammells and co-workers [25] employed perovskite-type La1.8Sr0.2CuO4 to catalyze CO2RR and achieved a ~40% Faradaic efficiency (FE) of alcohols at the total current density of 180 mA cm−2 in 0.5 M KOH using a gas diffusion electrode (GDE). Later, Yellowlees and co-workers [26] gained the 9.4% and 11.4% FEs towards ethylene and methane using the same catalyst. Although interesting, these prior work provided limited information regarding the structural evolution and real active site during electrolysis; moreover, the prospects of such materials for high-rate CO2 conversion towards C2+ products are still unclear.

Herein, we report a La2CuO4 perovskite oxide-derived Cu (POD-Cu) catalyst rich in GBs. The high-density GBs were in-situ created in the structural reconstruction process induced by CO2-assisted La sites leaching. Using this GBs-rich Cu catalyst, we achieved a ~80.3% Faradaic efficiency towards C2+ products with partial current densities over 400 mA cm−2 in neutral environment in a flow-cell electrolyzer, outperforming the conventional CuO-derived Cu (OD-Cu) counterparts. By combining the structural and spectroscopic investigations, we uncovered that the introduction of high-density GBs enables abundant undercoordinated Cu sites and thus promote C–C coupling kinetics for C2+ products generation via enhancing CO intermediate adsorption. Our work highlights the great potential of Cu-based perovskite materials for efficient production of valuable multicarbon compounds via CO2RR electrochemistry.

MATERIALS AND METHODS

For Cu-based perovskite oxides, the common materials are lanthanum cuprates with the general formula Lan+lCunO3n+l, which are well-known models for studying superconductivity phenomena in physics [27]. In this homologous series, La2CuO4 is considered to be the simplest member (n=1) [28], and recently exhibited electrocatalytic activity for some reduction reactions, such as oxygen reduction reaction and nitrate reduction reaction [29,30]. These advances have motivated us to choose La2CuO4 as an archetypal example of Cu-based perovskite to study its structural evolution and catalytic property in CO2RR electrochemistry.

The La2CuO4 material was prepared by a sol-gel method. Briefly, 2.165 g La(NO3)3∙6H2O, 0.582 g Cu(NO3)2∙3H2O and the complexing reagent citric acid monohydrate (1.576 g, equal to total metal ion molar concentration) were sequentially dissolved in deionized water in a glass beaker. The solution was concentrated with the evaporation of water by heating and became the viscous gel. Then the glass beaker containing viscous gel was placed in an oven kept at 200°C. The mixture underwent further dehydration followed by decomposition with swelling and frothing, then ruptured with a flame, resulting in the voluminous and foamy precursor. By annealing the ground precursor powder at an optimized temperature of 700°C for 6 h in air (Figures S1−S3), followed by cooling in air, we gained the reddish brown La2CuO4 power. More details can be found in the Supplementary Information.

RESULTS AND DISCUSSION

We performed X-ray diffraction (XRD) measurement of the obtained powder. As shown in Figure 1A, all of the diffraction peaks match well with orthorhombic-type perovskites structure of La2CuO4, suggesting the absence of impurity in the sample. Figure 1B displays X-ray photoelectron spectroscopy (XPS) result of the La 3d and Cu 2p binding energy region. La 3d region shows the double peak structure of each spin-orbit split component with configurations of 3d94f0L and 3d94f1L (L denotes the oxygen ligand), which is typical of La3+ compounds [31]. Generally, the f0 and f1 separately dominate the low binding energy band and the high binding energy band. The f1-f0 energy separations are characteristic of different La compounds, such as 3.9 eV for La(OH)3 [32], 4.3 eV for LaCoO3 [31]. The value (3.2 eV) of the f1-f0 energy separation here is similar to that reported (3.1 eV) for La2CuO4 [33]. The Cu 2p spectrum exhibits satellite peaks, which are characteristic of Cu2+ compounds, corresponding to the divalent Cu sites in the La2CuO4 perovskite structure. Scanning electron microscopy (SEM) and transmission electron microscopy (TEM) images of La2CuO4 show a porous worm-like nanostructures consisting of interconnected nanoparticles with an average diameter of about 50−200 nm (Figure 1C and Figure S4). The high resolution-transmission electron microscopy (HR-TEM) and selected-area electron diffraction (SAED) pattern uncover that the single nanoparticle is single crystal and is interconnected through distinct grain boundaries (Figure S4). Energy dispersive X-ray (EDX) spectrum elemental mapping in Figure 1D exhibits a uniform spatial distribution of La, Cu, and O elements in the sample. We further studied the detailed atomic structure of the La2CuO4 by aberration corrected high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM). The atomic-resolution Z-contrast images in Figure 1E clearly reveal an adjacent layered structures with different layer spacing of 3.4 and 2.8 Å respectively, where La atoms exhibit higher image intensity compared with Cu atoms. Using image contrast, the La and Cu atoms can be further identified by the line intensity profile (inset of Figure 1E) acquired along the yellow arrow of [001] direction. This result is consistent with the investigation that La2CuO4 is a kind of K2NiF4-type structure consisting of alternate stackings of perovskite and rock salt blocks [34], where CuO2 planes are separated by LaO-LaO rock-salt layers, forming the corner-shared CuO6 octahedra as shown in Figure 1F.

thumbnail Figure 1

Physical characterization of La2CuO4 perovskite oxide. (A) XRD pattern of La2CuO4 powder. (B) XPS high resolution spectrum of the La 3d and Cu 2p region. (C) SEM image of La2CuO4 particle. Scale bar: 500 nm. (D) HAADF-STEM image and the corresponding EDS elemental mapping of La2CuO4 particle. Scale bar: 50 nm. (E) Atomic-resolution HAADF-STEM image of La2CuO4. Scale bar: 1 nm. Inset is the line intensity profile acquired along the yellow arrow of [001] direction. Green and yellow balls represent La and Cu atoms, respectively. (F) Crystal structure of La2CuO4.

Altogether, multiple characterizations above show that high-quality La2CuO4 perovskite was successfully synthesized by a simple sol-gel synthetic strategy. Additionally, this approach is scalable, which enables the production of high-yield La2CuO4 powders (>15 g one batch) with good fidelity for potential large-scale use (Figure S5). For a fair catalytic performance comparison, we also synthesized high-purity CuO powders by the same synthetic protocol without the addition of La. The obtained samples are mainly composed of irregular gravel-like particles with an average diameter of less than 200 nm (Figures S6 and S7).

To gain OD-Cu and POD-Cu electrocatalysts, the CuO and La2CuO4-coated GDEs (with a catalyst loading of ~1.0 mg cm−2) were firstly activated through in-situ electrochemical reduction at a constant potential of −1.8 V (versus normal hydrogen electrode (NHE)) for 15 min in a three-compartment flow-cell setup (Figure S8). Interestingly, unlike CuO electrode that can be easily reduced to metallic Cu regardless of the atmosphere (Figure S9), the structural evolution of La2CuO4 perovskite during electrochemical process exhibits a strong atmosphere-dependent behavior. In Ar-saturated KCl, no obvious changes in the structure and morphology of La2CuO4 were observed after electrochemical activation (Figure 2A and Figure S10), indicating that La2CuO4 perovskite is hardly reducible under this condition, consistent with previous reports [30,35]. By contrast, when the electrolyte was aerated with CO2, we found that La element was progressively leached into the electrolyte during the activation process, confirmed by inductively coupled plasma atomic emission spectroscopy (ICP-AES, Figure S11). Consequently, the electrode was evolved from La2CuO4 perovskite structure into cubic Cu phase, as suggested by XPD pattern in Figure 2A.

thumbnail Figure 2

Physical characterization of POD-Cu. (A) XRD patterns of La2CuO4 electrodes after activation in Ar and CO2-feed catholyte. The diffraction peaks marked by black rhombus originated from GDL substrate. The weak Cu diffraction in XRD pattern was the result of the very low content of Cu on the GDL (about 0.16 mg cm−1). (B) Schematic diagram of the structural evolution of La2CuO4. (C) HAADF-STEM image and the corresponding EDX elemental mapping of POD-Cu. Scale bar: 100 nm. (D), (E) HR-TEM images of POD-Cu. Inset in (D) is the corresponding FFT pattern. Scale bars: 10 nm in (D) and 5 nm in (E). (F) Pb UPD profile of POD-Cu.

The above results reveal that CO2 plays a unique and important role in the evolution of La2CuO4 perovskite during electrochemical process, which can be reasonably explained by the following reasons. Under negative bias, a small amount of Cu sites on Cu-terminated surface of La2CuO4 are firstly reduced with the concomitant formation of La oxides regardless of in the Ar or CO2-feed catholyte [35]. In Ar-feed electrolyte, due to the extremely negative reduction potential of La oxides (<−2.9 V versus standard hydrogen electrode (SHE) [36]), the internal Cu sites of perovskite are blocked by La-O cages, thus preventing from being exsoluted [37]. As a result, O vacancies may be created on the perovskite surface, and no significant restructuration occurs [35]. When introducing CO2 into the system, however, La oxides can react with CO2 and H2O to form La2(CO3)3, which would be dissolved into the electrolyte by forming complex with carbonates generated in HER or CO2RR process [38]. Accordingly, the leaching of La sites further leads to deep exsolution and reorganization of internal Cu sites, resulting in the collapse of perovskite structure with the concomitant formation of the metallic Cu phase, as illustrated in Figure 2B.

To gain the detailed structural information, we characterized the POD-Cu catalyst by multiple techniques. EDX elemental mapping measurements of randomly selected regions permit a quantitative analysis of composition at a relatively large scale. As displayed in Figure 2C and Figures S12 and S13, the atomic percentages of lanthanum are generally less than 0.15%, a value comparable to the detection limit of the instrument (about 0.1%), which suggest a complete leaching of La element during in-situ electrochemical activation process. Additionally, EDX elemental mapping and XPS spectrum (Figure S14) exhibited a thin oxide layer on the POD-Cu surface, originating from natural oxidation of nanostructured Cu exposed in the air. By combining XRD and EDX results, we determined a thoroughly structural transition from layer perovskite oxide to metallic Cu without residual La after the activation process. Correspondingly, the worm-like nanostructures of La2CuO4 reorganize into rougher multiparticles (Figure S15).

HR-TEM image (Figure 2D) shows that POD-Cu catalyst appeared as irregular aggregates, which are compactly composed of multiple nanocrystallites interconnected by distinct GBs (marked by white arrows), corresponding to the polycrystalline diffraction rings in fast Fourier transform (FFT) pattern. Such single-to-polycrystalline transformation revealed that the initial La2CuO4 single crystals are fragmented into nanosized Cu grains connected each other during the activation process, which are more noticeable in the enlarged HR-TEM images (Figure 2E and Figure S16), thus resulting in high-density GBs. We highlight that these GBs can kinetically trap and stabilize lattice dislocations and deformations in polycrystalline materials, providing a way to create high-energy surfaces with abundant undercoordinated sites, i.e. GBs effects, as previously demonstrated by Kanan group [39,40]. Unlike POD-Cu catalyst, we identified that OD-Cu sample mainly presents continuous and regular lattice fringes in the microscopic analysis of multiple regions (Figures S17 and S18), indicating much fewer GBs and defective sites.

To obtain more insights into surface properties of the POD-Cu electrode, we employed highly sensitive Pb underpotential deposition (UPD) technique to probe the distribution of crystallographic domains on the catalyst surface (Figure S19). The UPD profile of POD-Cu in Figure 2F displays three overlapped voltammetric peaks in the potential window of −0.27 to −0.36 V versus saturated calomel electrode (SCE). According to previous reports on the Cu disk [41], these peaks are characteristic to Cu (111) facet, Cu (100) facet and defective sites respectively. The surface charge value of each peak was then obtained by integrating the deconvoluted peak after subtracting the base line that corresponds to the capacitive charge. We observe that the <111> and <100> domains are the predominant orientation on the POD-Cu surface, respectively, which contribute to 47% and 32% of the total surface charge (Table S1). Notably, the defective sites contribute to the total surface charge up to 20.9%, far exceeding that of 3.2% for OD-Cu. Based on the microscopic analysis and surface-sensitive probe measurement, we unveil that compared to OD-Cu, POD-Cu possesses abundant structural defects mainly trapped in high-density GBs. We reasonably attribute the defect-site-rich surface of POD-Cu to the dramatic collapse of perovskite structure that induced by the La sites leaching, as well as the exsolution and rearrangement of Cu sites.

We evaluated the CO2RR catalytic properties of POD-Cu and OD-Cu electrodes using the same three-compartment flow-cell setup. We used neutral KCl (1 M, pH ~6.5) electrolyte as catholyte to mitigate (bi)carbonate issues [42]. The low buffering capability of the KCl electrolyte allows surface pH of the electrode to increase to a weakly basic range, which benefits C2+ production via positively shifting the C2+ onset potential and suppressing HER, as described by previous studies [43,44]. Considering that the electrode surface pH is substantially higher than the bulk pH during high-rate electrolysis in the non-buffering catholyte, all potentials are converted into the NHE rather than RHE scale to facilitate performance comparison on an absolute scale [43,45,46]. Gas-phase and liquid-phase products were quantified by gas chromatography (GC) and nuclear magnetic resonance (NMR) spectroscopy, respectively. Detailed information regarding the CO2RR measurement is provided in the Supplementary Information.

Multiple products were detected from the CO2RR on two Cu electrodes, including C1 compounds (CO, formate and CH4) and C2+ compounds (ethylene, ethanol, acetate and n-propanol). The FEs of different products yielded on POD-Cu and OD-Cu are summarized in Figure 3A and Tables S2 and S3. Initially, POD-Cu electrode exhibits a C2+ FE of ~46% at −1.67 V. As the potential shifts negatively, this FE value rapidly reaches ~70% at −2.12 V. Correspondingly, the FE of C1 products, especially CO, significantly drops, suggesting the promoted C−C coupling kinetics for the formation of C2+ products. Intriguingly, the POD-Cu catalyst holds the high C2+ FE of over 70% in a broad potential range of ≤−2.12 V. A maximal C2+ FE of 80.3% is achieved at −2.34 V, corresponding to 38.3% ethylene, 30.1% ethanol, 10.1% acetate, and 1.8% n-propanol, which is among the best CO2RR performance in neutral electrolytes (Table S4). By contrast, the OD-Cu catalyst bears a lower C2+ FE less than 70% (Figure 3A) in the similar potential range, which agrees with previous reports on the OD-Cu catalysts [43,47,48]. In addition, compared with OD-Cu, POD-Cu exhibits lower H2 selectivity (FE≤12%, Figure 3A) and generation rate in the whole potential range (Figure S20), demonstrating that the competitive HER reaction was suppressed over the POD-Cu surface.

thumbnail Figure 3

CO2RR performance. (A) Faradaic efficiencies, (B) geometric current densities and (C) mass activities of C2+ products on POD-Cu and OD-Cu catalyst at various potentials. Error bars are based on the standard deviation of at least three independent measurements, and points are average values.

We also compared geometric current densities for C2+ production of two studied Cu catalysts at various applied potentials (Figure 3B). For POD-Cu electrode, there was a near-exponential increase of the C2+ current density in the whole potential range, indicating excellent C2+ generation rate with roughly stable selectivity even under high overpotentials. Impressively, the POD-Cu possesses a high C2+ current density of 511 mA cm−2 at −2.45 V, holding promise to operate at industrial electrolyzer-relevant current densities. By contrast, the C2+ partial current density on OD-Cu increases exponentially only at low overpotentials and then turns to a near linear increase at high overpotentials. The limited C2+ products formation rate on OD-Cu is mainly caused by substantial generation of competitive H2 and CH4 under high overpotentials (Figure 3A). As a result, OD-Cu electrode only offered a C2+ partial current density of 332 mA cm−2 even applying a more negative potential of −2.73 V. We further normalized the C2+ partial current density by electrochemically active surface area (ECSA) gained from Pb UPD profile. ECSA-normalized C2+ partial current densities of POD-Cu electrode largely exceed that of OD-Cu at various studied potentials (Figure S21), indicating its superior intrinsic activity for C2+ production.

We highlight that the content of Cu element in La2CuO4 is much lower than that in CuO (0.16 mg cm−2 versus 0.80 mg cm−2) under the same loading on GDE (typically 1 mg cm−2), suggesting that the catalytic activity with respect to catalyst loading, i.e. the Cu mass activity, of the POD-Cu catalyst is higher. Indeed, Figure 3C shows that the C2+ mass activity of POD-Cu overwhelmingly outperforms OD-Cu at all potentials examined. In particular, POD-Cu possesses an outstanding mass activity of ~3.26 A mgCu−1 at −2.45 V, which represents the best mass activities observed on Cu-based catalysts for C2+ compounds (Table S4). The ultra-high mass activity of POD-Cu can be ascribed to small grain size, fully exposed surface and superior intrinsic activity.

Now, we turn to assess the long-term stability of our POD-Cu catalyst. To this end, we operated a continuous CO2 electrolysis under a constant current density of 300 mA cm−2 in the flow-cell electrolyzer. A stable C2+ FE of 70%−75% was recorded during the 4-h continuous operation. The slight decrease in C2+ Faradaic efficiency later was predominantly due to the electrolyte flooding caused by the reduced hydrophobicity of the GDL (inset in Figure S22), thus leading to the CO2 diffusion toward the catalyst layer to be blocked [49–52]. Post-mortem characterizations reveal that the abundant GBs are largely maintained (Figure S23), confirming the good operational stability of the defective catalyst.

For the CO2RR on the Cu surface, adsorbed CO (COad) is an essential intermediate towards the formation of multicarbon products. Prior studies have showed that the C≡O stretch mode is highly sensitive to the applied potential, surface morphology and reaction site of catalysts [53]. Therefore, COad becomes a powerful and broadly applicable spectroscopic probe of the electrochemical interface. We thus performed in-situ surface enhanced Raman spectroscopy (SERS, Figure S24) to exploit the utility of COad probe to study the electrochemical interface of POD-Cu electrode.

Figure 4A exhibits the time-dependent SERS spectra in the range of C≡O stretch mode during La2CuO4 activation process under the bias of −1.8 V. At 2 min, a band centered at ~1820 cm−1 and additional broad band in 1900−2100 cm−1 emerge, which are assigned as bridge-bound CO (CObridge) and atop-bound CO (COatop), respectively. The SERS results here imply that the Cu sites in perovskite underwent exsolution and reorganized into the metallic Cu phase (Figure S25). We found that the CObridge band is dominantly present in the first 4 min, then the COatop band develops into a prominent feature in the following 10 min. The conversion of COad configuration suggests that two distinct active sites were formed during the reconstruction process [54]. Generally, CObridge behaves relatively inert without participating in further C–C coupling step for yielding multicarbon products [55]. Instead, it is more likely an on-pathway intermediate in favor of CH4 generation [56]. Indeed, product measurement during the in-situ electrochemical activation of La2CuO4 displayed a high CH4 FE of ~35% in the first 10 min, but the value rapidly dropped to ~7% and C2H4 FE sharply increased from 6% to 37% in the following 10 min (Figure S26). Combining spectroscopic observation and experimental results, we conclude that at a negative bias, La2CuO4 perovskite first evolved into dispersed Cu particles supported on perovskite matrix, which prefer CObridge and generate CH4 as the dominant product [57]. With deeper exsolution and reorganization of internal Cu sites, the perovskite oxide eventually transformed into Cu aggregates that favors COatop and promotes the formation of multicarbon products.

thumbnail Figure 4

In-situ SERS spectra studies. (A) Time-dependent SERS spectra collected during the perovskite activation process at −1.8 V. (B) Potential-dependent SERS spectra of POD-Cu catalyst during CO2RR. (C) Diagram of structure-activity relationship on POD-Cu surface. The green, blue balls represent Cu atoms in grain 1 and 2, and the black, red, gray balls represent C, O and H atoms, respectively.

Potential-dependent SERS spectra of POD-Cu catalyst were also collected, as shown in Figure 4B. Clearly, relatively inert CObridge band is very weak that can be ignored. By contrast, the reactive COatop is exclusively present at studied potentials, consistent with experimental results demonstrated in Figure 3A. Overall, at a lower overpotential range, the COatop band intensity raises with the increased CO coverage. At higher overpotencials, however, the fast consumption of COad populations via deeper reduction leads to the declining of COatop band intensity. Moreover, COad band typically consists of a low frequency band (LFB) centered at ~1960 cm−1 and a high frequency band (HFB) centered at ~2040 cm−1 (Figure S27).

Previous studies have well established that COad binds on defective Cu sites more strongly, resulting in a blue-shift of the C≡O vibrational frequency of COad [58]. This is because the electronic charge distribution at undercoordinated defects is significantly different from that at flat terraces (so-called Smoluchowski effect [59]). The reduced charge density at defects sites could lead to a reduction in repulsion between COad molecules, thus giving rise to an increase of the CO binding intensity. Therefore, the LFB and HFE here are associated with COatop on terrace and defective sites, respectively. Due to dynamic dipole coupling effect of COad, the HFB is amplified by intensity transfer from the LFB mode to the HFB mode, failing to quantify the ratio of these two sites [53]. Nonetheless, the apparent HFB on POD-Cu reveals the rich defective sites on the surface, mainly from GBs-trapped lattice dislocations and deformations [39,40], consistent with HR-TEM and Pb UPD analysis demonstrated above. In contrast to POD-Cu, OD-Cu surface mainly exhibits LFB (Figure S28), suggesting limited defective sites on its surface and thus weaker COatop binding strength.

For the formation of multicarbon products, suitably high CO binding energy on the Cu surface is necessary so that the COad can participate in further C–C coupling step before desorption. Compared with terrace sites, defective sites typically bind CO more strongly by about 50–100 meV [60,61]. Accordingly, they are predicted to give orders of magnitude higher reaction rates in CO2RR than terrace sites because of the more favorable CO adsorption [62]. On the basis of the above analysis, we therefore attribute the excellent catalytic activity of POD-Cu to abundant defective sites, which are trapped and stabilized by the high-density GBs created from structural reconstruction. Compared with the terrace sites on grain surface (GS), these defective sites can bind the reactive COatop intermediates stronger, which not only promoted the possibility of dimerization of neighboring COatop species but also largely suppressed H2 evolution by limiting the access of *H intermediate to active sites.

CONCLUSIONS

In summary, the work reported here showcases a La2CuO4 perovskite oxide-derived Cu catalyst for electrochemical conversion of CO2 into C2+ products in neutral environment with excellent selectivity and activity. A maximum C2+ FE of 80.3% and with C2+ partial current density beyond 400 mA cm−2 was achieved, far outperforming the CuO-derived Cu counterparts. Structural characterizations and in-situ spectroscopic investigations revealed that the perovskite oxide-derived Cu possesses high-density grain boundaries containing abundant defective sites, which facilitate the C–C coupling step for C2+ products generation via enhancing the adsorption of key CO intermediates. We anticipate that our finding will find immediate use in the design of next generation CO2RR catalysts based on flexible and efficient perovskite materials, which should aid the large-scale production of valuable C2+ compounds via CO2RR electrochemistry.

Data availability

The original data are available from corresponding authors upon reasonable request.

Acknowledgments

We are grateful to Dr. Ming Zuo for his help in collecting the HRTEM data.

Funding

This work was supported by the National Basic Research Program of China (2018YFA0702001), the National Natural Science Foundation of China (21975237 and 51702312), Anhui Provincial Research and Development Program (202004a05020073), the USTC Research Funds of the Double First-Class Initiative (YD2340002007), the Fundamental Research Funds for the Central Universities (WK2340000101), the Technical Talent Promotion Plan (TS2021002), and the Recruitment Program of Global Youth Experts.

Author contributions

M.R.G. conceived and supervised the project. Z.Z.N. and L.P.C. performed the experiments, collected and analyzed the data. P.P.Y., Z.Z.W., and M.H.F helped with material characterization and analysis. M.R.G. and Z.Z.N. co-wrote the manuscript. All authors discussed the results and commented on the manuscript.

Conflict of interest

The authors declare no conflict of interest.

Supplementary information

The supporting information is available online at https://doi.org/10.1360/nso/20220044. The supporting materials are published as submitted, without typesetting or editing. The responsibility for scientific accuracy and content remains entirely with the authors.

References

All Figures

thumbnail Figure 1

Physical characterization of La2CuO4 perovskite oxide. (A) XRD pattern of La2CuO4 powder. (B) XPS high resolution spectrum of the La 3d and Cu 2p region. (C) SEM image of La2CuO4 particle. Scale bar: 500 nm. (D) HAADF-STEM image and the corresponding EDS elemental mapping of La2CuO4 particle. Scale bar: 50 nm. (E) Atomic-resolution HAADF-STEM image of La2CuO4. Scale bar: 1 nm. Inset is the line intensity profile acquired along the yellow arrow of [001] direction. Green and yellow balls represent La and Cu atoms, respectively. (F) Crystal structure of La2CuO4.
In the text
thumbnail Figure 2

Physical characterization of POD-Cu. (A) XRD patterns of La2CuO4 electrodes after activation in Ar and CO2-feed catholyte. The diffraction peaks marked by black rhombus originated from GDL substrate. The weak Cu diffraction in XRD pattern was the result of the very low content of Cu on the GDL (about 0.16 mg cm−1). (B) Schematic diagram of the structural evolution of La2CuO4. (C) HAADF-STEM image and the corresponding EDX elemental mapping of POD-Cu. Scale bar: 100 nm. (D), (E) HR-TEM images of POD-Cu. Inset in (D) is the corresponding FFT pattern. Scale bars: 10 nm in (D) and 5 nm in (E). (F) Pb UPD profile of POD-Cu.
In the text
thumbnail Figure 3

CO2RR performance. (A) Faradaic efficiencies, (B) geometric current densities and (C) mass activities of C2+ products on POD-Cu and OD-Cu catalyst at various potentials. Error bars are based on the standard deviation of at least three independent measurements, and points are average values.
In the text
thumbnail Figure 4

In-situ SERS spectra studies. (A) Time-dependent SERS spectra collected during the perovskite activation process at −1.8 V. (B) Potential-dependent SERS spectra of POD-Cu catalyst during CO2RR. (C) Diagram of structure-activity relationship on POD-Cu surface. The green, blue balls represent Cu atoms in grain 1 and 2, and the black, red, gray balls represent C, O and H atoms, respectively.

In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.