Single-Crystalline Mesoporous Palladium and Palladium-Copper Nanocubes for Highly Efficient Electrochemical CO ₂ Reduction

Abstract

Open AccessCCS ChemistryRESEARCH ARTICLE1 Apr 2022Single-Crystalline Mesoporous Palladium and Palladium-Copper Nanocubes for Highly Efficient Electrochemical CO2 Reduction Hao Lv†, Fang Lv†, Huaiyu Qin, Xiaowen Min, Lizhi Sun, Na Han, Dongdong Xu, Yanguang Li and Ben Liu Hao Lv† College of Chemistry, Sichuan University, Chengdu 610064 , Fang Lv† Institute of Functional Nano and Soft Materials (FUNSOM), Jiangsu Key Laboratory for Carbon-Based Functional Materials and Devices, Soochow University, Suzhou 215123 , Huaiyu Qin College of Chemistry, Sichuan University, Chengdu 610064 , Xiaowen Min College of Chemistry, Sichuan University, Chengdu 610064 , Lizhi Sun Jiangsu Key Laboratory of New Power Batteries, Jiangsu Collaborative Innovation Center of Biomedical Functional Materials, School of Chemistry and Materials Science, Nanjing Normal University, Nanjing 210023 , Na Han Institute of Functional Nano and Soft Materials (FUNSOM), Jiangsu Key Laboratory for Carbon-Based Functional Materials and Devices, Soochow University, Suzhou 215123 , Dongdong Xu Jiangsu Key Laboratory of New Power Batteries, Jiangsu Collaborative Innovation Center of Biomedical Functional Materials, School of Chemistry and Materials Science, Nanjing Normal University, Nanjing 210023 , Yanguang Li *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] Institute of Functional Nano and Soft Materials (FUNSOM), Jiangsu Key Laboratory for Carbon-Based Functional Materials and Devices, Soochow University, Suzhou 215123 and Ben Liu *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected]da.edu.cn College of Chemistry, Sichuan University, Chengdu 610064 Jiangsu Key Laboratory of New Power Batteries, Jiangsu Collaborative Innovation Center of Biomedical Functional Materials, School of Chemistry and Materials Science, Nanjing Normal University, Nanjing 210023 https://doi.org/10.31635/ccschem.021.202100958 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail Mesoporous single crystals have unique potential in catalysis, but remain unexplored owing to the enormous synthetic challenge that they pose. Herein, we report a facile soft-template method to prepare palladium (Pd) and Pd alloy nanocubes with single-crystallinity and abundant mesoporosity. The successful formation of these exotic nanostructures essentially relies on the cointroduction of cetyltrimethylammonium chloride as the surfactant template and extra Cl− ions as the facet-selective capping agent under well controlled experimental conditions. Thanks to their large surface areas and penetrating mesoporous channels, our products exhibit a great performance for electrochemical CO2 reduction. The best sample from alloying palladium with copper enables the efficient formate production with high selectivity (90∼100%) over a broad potential range, and great stability even under the working potential as cathodic as −0.5 V versus a reversible hydrogen electrode. These performance metrics are far superior to previous Pd-based materials, and underscore the structural advantages of our products. Download figure Download PowerPoint Introduction Single-crystalline mesoporous metals constitute a unique class of materials for catalytic applications. Their ordered mesopores expose large accessible surface areas and active sites, and facilitate the rapid diffusion of reactants and products,1–6 while their single-crystallinity ensures long-range structural coherence and enhances electron transport.7–10 Unfortunately, the combination of ordered mesoporosity and single-crystallinity appears inherently unfavorable. Complete crystallization of mesoporous walls significantly compromises structural stability and integrity, leading to the distortion or even collapse of ordered mesopores and consequently the loss of accessible surface areas and active sites.11–14 Indeed, most mesoporous materials (such as silica, carbon, and metal oxides) are amorphous or poorly crystalline in nature. There are only scarce reports available about single-crystalline mesoporous oxides or nitrides in the literature.14–23 The preparation of single-crystalline mesoporous metals is even more challenging since metals generally have larger surface energy than oxides, and have a greater propensity to minimize their surface areas and form disordered structures.20,24,25 This synthetic obstacle largely precludes their potential application at the current stage. One of the emerging catalytic processes is the electrochemical CO2 reduction reaction (CO2RR). It converts CO2 to value-added industrial chemicals or fuels, and is an essential step to close the artificial carbon cycle. Most CO2RR electrocatalysts are composed of metals.26–29 Depending on the selection of catalysts and experimental conditions, different reduction products can be attained from CO2 reduction, of which formic acid or formate is recommended as the most economically viable.30,31 Pd is so far the only known candidate capable of selectively reducing CO2 to formate at close-to-zero overpotential in aqueous solution.32–35 It unfortunately suffers from severe CO poisoning and is subject to rapid selectivity and stability loss with increasing overpotential.36–38 To this end, efforts have been made through nanostructural engineering and/or compositional regulation to enhance the operational stability of Pd-based materials. The performance gain, however, remains limited. In this contribution, we develop a facile aqueous method to prepare single-crystalline mesoporous Pd (s-mesoPd) and Pd alloy nanocubes in the copresence of a suitable quaternary ammonium surfactant and extra Cl− ions. The product exhibits uniform nanocubic morphology with the {100} enclosure, and features abundant mesoporous channels and single-crystallinity. When investigated for CO2 reduction in an aqueous solution, the best catalyst achieved unprecedented selectivity and stability for the formate production at the cathodic potential that was previously believed to cause quick catalyst deactivation. Experimental Methods Synthesis of s-mesoPd nanocubes and other nanocrystals Synthesis of s-mesoPd nanocubes In a typical synthesis of s-mesoPd nanocubes, 0.325 mg of cetyltrimethylammonium chloride (CTAC) was first dissolved in 5.0 mL of deionized H2O, followed by the addition of 0.24 mL of 80 mM KCl and 0.25 mL of 10 mM H2PdCl4. After incubation at 50 °C for 30 min, 0.50 mL of 0.30 M freshly prepared l-ascorbic acid (AA) was rapidly injected into the above solution with gentle shaking. The color of the solution gradually changed from orange to dark brown. After another 30 min, the product was collected by centrifugation and washed several times with ethanol and H2O. The surfactant type and/or KCl concentration were varied under otherwise identical conditions to investigate their effects on the product morphology and crystallinity. The size of s-mesoPd nanocubes was adjusted by varying the added amount of H2PdCl4 under otherwise identical conditions. Synthesis of p-mesoPd nanoparticles In a typical synthesis of p-mesoPd nanoparticles, 0.325 mg of CTAC was first dissolved in 5.0 mL of deionized H2O, followed by the addition of 0.25 mL of 10 mM H2PdCl4. After incubation at 50 °C for 30 min, 0.50 mL of 0.30 M freshly prepared AA was rapidly injected into the above solution with gentle shaking. After another 30 min, the product was collected by centrifugation and washed several times with ethanol and H2O. Synthesis of s-mesoPdCu nanocubes In a typical synthesis, 0.325 mg of CTAC was first dissolved in 5.0 mL of deionized H2O, followed by the addition of 0.24 mL of 80 mM KCl, 0.25 mL of 10 mM H2PdCl4, and 0.125 mL of 10 mM Cu(NO3)2. After incubation at 50 °C for 30 min, 0.50 mL of 0.30 M freshly prepared AA was rapidly injected into the above solution with gentle shaking. After another 30 min, the product was collected by centrifugation and washed several times with ethanol and H2O. Synthesis of s-mesoPdRh nanocubes In a typical synthesis, 0.325 mg of CTAC was first dissolved in 5.0 mL of deionized H2O, followed by the addition of 0.24 mL of 80 mM KCl, 0.25 mL of 10 mM H2PdCl4, and 0.125 mL of 10 mM (NH4)3RhCl6. After incubation at 50 °C for 30 min, 0.50 mL of 0.30 M freshly prepared AA was rapidly injected into the above solution with gentle shaking. After another 30 min, the product was collected by centrifugation and washed several times with ethanol and H2O. Synthesis of Pd nanoparticles In a typical synthesis, 13.0 mg of cetylpyridinium bromide was dissolved in 5.0 mL of H2O, followed by the addition of 0.25 mL of 10 mM H2PdCl4. After incubation at 50 °C for 30 min, 0.50 mL of 0.30 M freshly prepared AA was rapidly injected into the above solution with gentle shaking. After another 30 min, the product was collected by centrifugation and washed several times with ethanol/H2O. Electrochemical measurements Electrochemical CO2RR was carried out in a gas-tight H-cell controlled by the standard three-electrode system as reported in our previous work.38 To prepare the working electrodes, 1.00 mg of the catalyst powders under investigation and 0.50 mg of Ketjenblack carbon were added to 6.0 μL of 5 wt % Nafion solution and 250 μL of ethanol, and sonicated for >30 min to form a uniform dispersion. This catalyst ink was then dropped onto a 1 × 1 cm2 glassy carbon plate to achieve a catalyst loading of 1 mg cm−2 and dried under ambient conditions. For CO2RR measurements, the working electrode and a saturated calomel reference electrode (SCE) were placed in the cathodic compartment. A graphitic rod counter electrode was placed in the anodic compartment. These two compartments were separated by a Nafion-117 membrane, and each filled with 30 mL 0.10 M KHCO3 electrolyte presaturated with CO2 (pH = 6.8). All the potential readings in our study were measured against SCE and converted with respect to a reversible hydrogen electrode (RHE) with ∼90% iR compensation. Only geometric current densities were reported. Polarization curves were recorded from the cathodic sweeping of the working electrode at the scan rate of 10 mV s−1. Chronoamperometric analysis was carried out at a few selected potential points for the selectivity and stability assessment. At the end of the chronoamperometric study, formate accumulated in the catholyte was analyzed using ion chromatography (Dionex ICS-600; Thermo Scientific, USA) by comparing with the calibration curve from a series of standard formate solutions. Characterizations Scanning electron microscopy (SEM) images were collected using a JSM-7600F field emission scanning electron microscope (JEOL, Japan). SEM samples were prepared by dropcasting a suspension of the sample powder onto a silicon wafer. Transmission electron microscopy (TEM) studies were carried out using a JEM-F200 field emission transmission electron microscope (JEOL, Japan) with an accelerating voltage of 200 kV. TEM samples were prepared by dropcasting a diluted suspension of the sample powder onto a carbon coated copper grid (300 mesh). Scanning TEM (STEM) images were collected on a Talos F200X scanning/transmission electron microscope (Thermo Scientific, USA) operating at an accelerating voltage of 200 kV and equipped with an energy-dispersive X-ray spectroscopy (EDS) detector for elemental mapping analysis. Small-angle X-ray scattering (SAXS) patterns were measured using an Anton Paar SAXSess mc2 instrument (Austria). Powder X-ray diffraction (XRD) patterns were recorded on powder samples using a D/max 2500 VL/PC diffractometer (Japan) equipped with graphite-monochromatized Cu Kα radiation in 2θ ranging from 30° to 90°. The working voltage and current were 40 kV and 100 mA, respectively. X-ray photoelectron spectroscopy (XPS) was performed on a Thermo ESCALAB 250Xi X-ray photoelectron spectrometer (Thermo Scientific, USA) using Al Kα radiation. The binding energy of the C 1s peak (284.8 eV) was employed as a standard to calibrate the binding energies of other elements (Pd and Cu). Results and Discussion s-mesoPd nanocubes were prepared via a facile one-step soft-template method by reducing PdCl42− with AA in the presence of CTAC as the surfactant template and KCl as the facet-selective capping agent (Figure 1a). The cointroduction of CTAC and extra Cl− ions holds the decisive key to the successful formation of s-mesoPd nanocubes. The SAXS pattern of the final product shows a well-defined peak at 1.04 nm−1, suggesting mesoporous structure with an average periodicity of 6.04 nm (Figure 1b). The powder XRD pattern of s-mesoPd displays a set of diffraction peaks in the 2θ range of 30–90° assignable to face-centered cubic (fcc) Pd (JCPDS: 05-0681), evidencing that it is crystalline at the atomic scale (Figure 1c). Under SEM, the product is observed to have a cubic morphology with rounded edges and corners and an average size of ∼100 nm (Figure 1d). Figure 1e depicts an individual nanocube viewed perpendicular to its face, edge, or corner. Careful examination reveals that the nanocube surface is riddled with mesopores, which become more evident under STEM and TEM imaging (Figures 1f and 1g and Supporting Information Figure S1). They extend from the center and form cylindrical mesoporous channels radially penetrating the entire nanocube. The meopore size and framework thickness are measured to be 2.7 and 3.2 nm, respectively, consistent with the average periodicity derived from the above SAXS analysis. Surprisingly, we find that each nanocube is a single crystal despite its abundant mesoporosity. The selected area electron diffraction (SAED) pattern of the nanocube shown in Figure 1h only exhibits a single set of bright spots along the [100] zone direction of fcc Pd (Figure 1i). High-resolution TEM imaging of four randomly selected regions of the nanocube discloses the same lattice orientation and an identical d-spacing of 0.196 nm from the (200) plane (Figure 1j). These results unambiguously confirm that s-mesoPd nanocubes are single-crystalline and enclosed with six {100} facets. Figure 1 | Synthesis and characterizations of s-mesoPd nanocubes. (a) Schematic synthetic procedure. (b) SAXS analysis. (c) XRD pattern. (d and e) SEM images, inset in (d) illustrates the nanocube size distribution. (f) STEM image, inset shows the {100} enclosure of a nanocube. (g) TEM image. (h) TEM image of a nanocube and corresponding (i) SAED pattern and (j) high-resolution TEM images from different regions in (h). Download figure Download PowerPoint Here, the combination of the overall nanocubic morphology, abundant mesoporosity and single-crystallinity is highly unusual, and not available in previous literature to the best of our knowledge. We believe that CTAC and Cl− ions play important roles in precisely regulating the nanocrystal growth. In what follows, a series of control experiments are carried out to elucidate their effects. We first explored the effect of the surfactant template on the product structure and morphology (Figure 2a). In the absence of any surfactant, the product consists of concave nanocubes free of any mesoporous channels ( Supporting Information Figures S2a–S2c). When shorter-chained octyltrimethylammonium chloride (C8TAC) or dodecyltrimethylammonium chloride (C12TAC) are used instead of CTAC under otherwise identical conditions, resultant nanocubes become smaller in size and have less ordered mesoporous channels ( Supporting Information Figures S2d–S2i). When longer-chained behenyltrimethylammonium chloride (C22TAC) is used, two-dimensional (2D) Pd nanosheets are yielded, presumably from templating the lamellar mesophase of C22TAC in solution ( Supporting Information Figures S2j–S2l and S3).39 Figure 2 | Exploration of different experimental parameters. (a) TEM images of products prepared with no surfactant or in the presence of quaternary ammonium surfactants with different chain lengths under otherwise identical conditions. (b) TEM images of products prepared with no extra Cl− or different Cl− concentrations as indicated under otherwise identical conditions. (c) TEM images showing the structural evolution of s-mesoPd nanocubes with the reaction time. Download figure Download PowerPoint We then investigated the effect of Cl− ions during the synthesis of s-mesoPd nanocubes (Figure 2b). In the absence of extra Cl− ions from KCl, the product consists of spherical polycrystalline mesoporous Pd (p-mesoPd) nanoparticles ( Supporting Information Figures S4a–S4c). With the extra addition of Cl− ions, the overall product morphology gradually evolves from spheres to cubes. The optimal Cl− concentration is found to be 4.0 mM, giving rise to s-mesoPd nanocubes with the {100} enclosure as characterized in detail above. EDS and XPS analyses reveal the presence of Cl residue on the product surface after the synthesis ( Supporting Information Figure S5). Adding too many Cl− ions (16 mM) affords smaller polycrystalline nanocubes albeit with the same {100} enclosure ( Supporting Information Figures S4d–S4f). It should be noted that K+ cations do not play a role in the facet selection as replacing KCl with an equimolar amount of NaCl, CaCl2 or NH4Cl results in similar s-mesoPd nanocubes ( Supporting Information Figure S6). However, replacing Cl− ions with Br− ions or adding OH− compromises the nanocubic morphology ( Supporting Information Figure S7). Based on the above experimental observations, we propose that s-mesoPd nanocubes are formed through a concurrent surfactant-templating and facet-selective growth mechanism. CTAC is directly responsible for the formation of abundant mesoporous channels. During the reaction, amphiphilic CTAC molecules interact with PdCl42− through the Coulombic attraction, self-assemble to one-dimensional (1D) cylindrical micelles that further organize into a three-dimensional (3D) mesophase. The reduction of PdCl42− by AA gives rise to the templated growth of metallic Pd nanocrystals on the surfactant micelles, creating ordered mesoporous channels upon the surfactant removal. In the meantime, Cl− ions strongly bind to Pd(100) and selectively stabilize this facet.40,41 Such a facet-selective capping effect ultimately results in overall nanocubic morphology of our product. It is the copresence of CTAC and extra Cl− ions that collectively makes possible the successful growth of our s-mesoPd nanocubes in solution. To lend further support to the proposed mechanism, we track the time-dependent nanocube growth under TEM (Figure 2c). Upon the injection of AA, single-crystalline dendritic nanoparticles of 10–25 nm are formed immediately ( Supporting Information Figures S8a–S8c). They grow bigger and gradually evolve into nanocubes in the next 60 s ( Supporting Information Figures S8d–S8i). The nanocubic morphology is maintained throughout the rest of the reaction. Finally, s-mesoPd nanocubes were obtained in 10 min. Our solution synthetic strategy is versatile. For example, by increasing the amount of the PdCl42− precursor under otherwise identical conditions, the size of s-mesoPd nanocubes can be systematically varied from 31 to 132 nm, all of which are single-crystalline and mesoporous (Figures 3a–3d and Supporting Information Figure S9). More interestingly, this method can be extended to the preparation of single-crystalline mesoporous nanocubes of bimetallic alloys (s-mesoPdM nanocubes) such as s-mesoPdCu (Figures 3e–3i) and s-mesoPdRh ( Supporting Information Figure S10). Taking s-mesoPdCu nanocubes (with a nominal Pd/Cu molar ratio of 2) as an example, we find that the coreduction of PdCl42− and Cu2+ does not noticeably alter the templating effect of CTAC and facet-selective capping effect of Cl− ions. Resultant nanocubes are observed to have a similar morphology as that of s-mesoPd nanocubes, while the EDS elemental mapping of Pd and Cu clearly evidences their uniform spatial distribution and attests to the alloy formation instead of Their SAXS and XRD measurements support the formation of the bimetallic alloy with ordered mesopores ( Supporting Information Figure XPS analysis evidences the electron from Cu to Pd in the alloy ( Supporting Information Figure The compositional in nanocubes an to their and their catalytic Figure | and control of s-mesoPd and nanocubes. TEM images of s-mesoPd nanocubes prepared with different CTAC STEM image of s-mesoPdCu nanocubes. (f) TEM image of a s-mesoPdCu nanocube and corresponding (g) SAED pattern and (h) high-resolution TEM image. (i) STEM image of a s-mesoPdCu nanocube and corresponding elemental Download figure Download PowerPoint The unique combination of mesoporosity and single-crystallinity our materials for many applications. a of we investigated the of s-mesoPd and s-mesoPdCu nanocubes for CO2RR to and with p-mesoPd nanoparticles and Pd nanoparticles (Pd Pd is the only known that can the CO2 reduction to formate at close-to-zero but unfortunately suffers from selectivity and stability under increasing overpotential 200 to CO We that this challenge be by of the large electrochemical surface areas and of single-crystalline mesoporous nanocubes as well as the effect from alloying Pd and Figure depicts the curves of s-mesoPdCu nanocubes in or M These two curves at which the CO2RR The cathodic current the potential in the presence of and cm−2 at which is over times larger than that measured in Polarization curves of other samples are shown in Supporting Information Figures Figure | CO2RR (a) Polarization curves of s-mesoPdCu in and M (b) Chronoamperometric curves of s-mesoPdCu nanocubes at different under (c) formate selectivity and of (d) of the formate selectivity of and Pd chronoamperometric stability of and Pd at Download figure Download PowerPoint For the product analysis and we performed chronoamperometric analysis at a few selected working and −0.5 V for different catalyst samples under It is that s-mesoPdCu exhibits chronoamperometric over the potential (Figure products and are is to be the only product from CO2 reduction. is analyzed to be at which to and and −0.5 V (Figure formate current is to from cm−2 at V to cm−2 at In we find that the cathodic current to the working potential is V for Pd and and V for s-mesoPd owing to the catalyst poisoning by CO from CO2RR ( Supporting Information Figures Figure the formate of four different electrocatalysts under study their potential all of exhibit more or less selectivity as shown in Supporting Information Figure in the overpotential our s-mesoPdCu clearly out for its unique to the selectivity and stability under working potential as cathodic as −0.5 the for CO2RR on Pd-based materials and far superior to any close reported in the literature ( Supporting Information S1). To the performance of stability was at V for as shown in Figure The cathodic current of Pd from 5 to cm−2 the first in with the previous The of mesoporosity significantly the CO and working to the catalyst surface areas that CO At the end of s-mesoPd and p-mesoPd the current of and respectively. Their average formate is measured to be and respectively. The stability and selectivity of s-mesoPd over p-mesoPd is to its single-crystalline that the during Finally, alloying Pd with Cu further enhances the s-mesoPdCu cm−2 even after s with average formate We believe that in addition to its the effect from alloying Pd and Cu is responsible for the observed The of a metal with such as Cu is to the Pd center with the electron from Cu to Pd as from consequently the CO binding on the alloy catalyst and its CO and catalytic After the stability s-mesoPdCu is observed to the single-crystalline mesoporous structure as well as the bimetallic ( Supporting Information Figure The Pd/Cu molar ratio is measured to only to Here, we a facile solution method for the preparation of s-mesoPd and nanocubes. The products abundant mesoporous channels single-crystallinity throughout each nanocube. We proposed that this unique was formed as the of the templating effect of CTAC and the facet-selective capping effect of Cl− ions as by a range of control The large surface areas and penetrating mesoporous channels of our products their catalytic applications. In M the best electrochemical CO2 reduction to formate with great selectivity and stability even under the working potential as cathodic as −0.5 which was previously believed to be too for Pd-based The effect from alloying Pd and Cu was found to play an role in the Our study the great potential and for single-crystalline and mesoporous metallic Supporting Information Supporting Information is available and of different STEM images, TEM images, SAED EDS SAXS XRD and XPS of different chronoamperometric and formate selectivity of different and of the formate of our best sample with literature of There is no of to Information the of Jiangsu the Open of Key Laboratory of and Materials and the of Jiangsu Innovation for the Collaborative Innovation Center of Suzhou Nano and for their the of Jiangsu for its the of for its The the Center of Experimental School of Sichuan for the TEM and STEM and and Their in Li for Mesoporous [email protected] with of or from CO2 on a Cu Liu in Liu Li of in through a with for and on of in Sun of to Liu Li of CO on in Liu Synthesis of Mesoporous and with Liu Li Liu Mesoporous with and and Synthesis in Reduction for Mesoporous Mesoporous Synthesis and the to Mesoporous Liu Sun Li Mesoporous with for Li in Mesoporous to Li in to and Mesoporous for Electrochemical and Xu Li in the Synthesis of Mesoporous Materials for and Based on and Xu Liu via Liu and of Electrochemical CO2 Reduction on in and Efficient for CO2 and the for Electrochemical Reduction of CO2 to and on Li Materials for Electrochemical Liu Li Liu as the for Electrochemical CO2 of CO2 to with for Electrochemical CO2 Reduction of Efficient Palladium for of to Han Liu Li Li for Electrochemical CO2 Reduction to Min of to at and of the Li Palladium as CO for the Electrochemical Reduction of CO2 to Han Sun Xu Li Reduction to Xu Liu Liu Han Liu Palladium with the Synthesis of of in Synthesis of Pd and Their of over Pd for CO2 CO2 via of Pd Information CO2 the Center of Experimental School of Sichuan for the TEM and STEM times

Topics & Concepts