Edges of Layered FePSe₃ Exhibit Increased Electrochemical and Electrocatalytic Activity Compared to Basal Planes

Abstract

Transition metal trichalcogenphosphites (MPX₃), belonging to the class of 2D materials, are potentially viable electrocatalysts for the hydrogen evolution reaction (HER). Many 2D and layered materials exhibit different magnitudes of electrochemical and electrocatalytic activity at their edge and basal sites. To find out whether edges or basal planes are the primary sites for catalytic processes at these compounds, we studied the local electrochemical and electrocatalytic activity of FePSe₃, an MPX₃ representative that was previously found to be catalytically active. Using scanning electrochemical microscopy, we discovered that electrochemical processes and the HER are occurring at an increased rate at edge-like defects of FePSe₃ crystals. We correlate our observations using optical microscopy, confocal laser scanning microscopy, scanning electron microscopy, and electron-dispersive X-ray spectroscopy. These findings have profound implications for the application of these materials for electrochemistry as well as for understanding general rules governing the electrochemical performance of layered compounds.

Introduction

Hydrogen has been identified as a promising medium to store energy and as a green fuel for the transport sector. However, obtaining green hydrogen in an economically feasible manner remains a big challenge. The main green route for producing hydrogen is electrochemical water splitting.¹ The energy consumption of this process can be decreased using suitable electrocatalysts such as platinum, but its rarity and high cost are the main reasons why electrochemical water splitting cannot be performed on an industrial scale yet.²

Consequently, research for alternative catalyst materials has sparked in recent years.² A large group of promising candidates can be found among the so-called 2D materials, which consist of layers of atomic thickness held together by van der Waals forces.³ Starting from graphene,⁴ multiple classes of 2D materials have been identified, including, but not limited to, layered pnictogens,⁵⁻⁸ transition metal dichalcogenides,⁹⁻¹³ transition metal oxides,¹⁴ metal–organic frameworks,^15,16 and MXenes.¹⁷⁻¹⁹ They cover a wide array of remarkable properties such as high electrical conductivity, charge capacity,^20,21 or electrocatalytic activity^{10−13,19,22,23} depending on the individual material. Another group of 2D materials that has emerged recently is the group of transition metal trichalcogenphosphites (MPX₃; M: transition metal, P: phosphor, and X: chalcogenide).²⁴ They were initially studied mainly for their magnetic properties.²⁵ Recent studies however have shown their remarkable electrocatalytic and photocatalytic properties.²⁶⁻³⁰

For the optimization of catalyst materials, it is important to know where the catalytic hotspots are located. It was found that graphite,³¹ graphene,³²⁻³⁴ transition metal dichalcogenides,³⁵⁻⁴⁰ and layered pnictogens^41,42 show increased electrochemical and electrocatalytic activity at the edges of individual layers compared to their basal planes. In the case of MXenes, the opposite was observed in a study by Djire et al.,⁴³ suggesting that the basal planes are more electrocatalytically active. For MPX₃ compounds, a theoretical study of catalytic hotspots for the cases of FePSe₃ and MnPSe₃ was performed.⁴⁴ Based on density functional theory, the authors conclude that edges are more active for the hydrogen evolution reaction (HER) than basal planes. However, these results need to be confirmed by electrochemical experiments. A useful technique to investigate the local electrochemistry of catalysts is scanning electrochemical microscopy (SECM).⁴⁵ It was recently applied for the investigation of the local electrochemical⁴⁰ and electrocatalytic^37,39 activity of transition metal dichalcogenides. Consequently, the aim of our work was to localize the electrochemical and electrocatalytic hotspots of FePSe₃ using SECM. We correlated our observations using scanning electron microscopy (SEM) combined with electron-dispersive X-ray spectroscopy (EDS). In addition, X-ray diffraction (XRD) was employed to confirm the composition of the FePSe₃ sample. For correlating SECM measurements to the sample topography, optical microscopy (OM) and confocal laser scanning microscopy (CLSM) were employed.

Experimental Section

Materials and Chemicals

Crystalline FePSe₃ was bought from XFNANO, China. A scheme of the surface morphology of the bulk crystal in given in Figure 1A, with the atomic structure given in Figure 1B. The crystal was prepared according to a previous study⁴⁰ to obtain a flat FePSe₃ electrode using the following materials. The crystal was embedded in a matrix consisting of a 1:1 (m/m) mixture of two-component epoxy resin (Struers Aps, Denmark) and graphite powder (<20 μm, synthetic, Sigma-Aldrich). Carbon SEM stubs obtained from Micro to Nano, Netherlands, were used as conductive support for the sample. Polydimethylsiloxane (SYLGARD 184) used during the electrode fabrication process was bought from Dow Inc., Michigan, USA. Conductive copper tape was used to establish electrical contact with the sample. All measurements in this work were executed with this sample. Feedback mode and substrate generation/tip collection (SG/TC) mode SECM images were recorded in a solution containing 1.5 mM ferrocene methanol (FcMeOH, 99%, ABCR GmbH, Germany) and 0.2 M potassium nitrate (KNO₃, analytical grade, Merck KGaA, Germany). For studying the local differences in HER, 0.5 M sulfuric acid (H₂SO₄, 96%, analytical grade, Penta, Czech Republic) was used. Deionized water with a resistivity >18.2 MΩ cm (Milli-Q Advantage A10 system, Merck Millipore, Germany) was used to prepare solutions.

Figure 1.

(A) Schematic representation of the surface morphology of a FePSe₃ crystal. (B) Atomic structure of the FePSe₃ crystal layers.

Instrumentation

Localized electrochemical and electrocatalytic activity studies were carried out using a commercially available scanning electrochemical microscope (SECM, Sensolytics, Germany) with a bipotentiostat (PGSTAT302N, Autolab, Netherlands). A 25 μm diameter Pt disk ultramicroelectrode (UME, RG = 11, Sensolytics, Germany) was used for SECM experiments. As counter and reference electrodes for electrochemical measurements, a graphite rod and a Ag/AgCl (3 M KCl) reference electrode were employed. The potentials stated herein refer to this reference system.

Scanning electron micrographs were recorded using a MIRA 3 SEM (Tescan, Czech Republic). For EDS maps, this setup was expanded with a Bruker XFlash 5010 EDS. Accelerating voltages of 5 or 20 kV for SEM and EDS, respectively. Gwyddion 2.55 and Origin 2020 software were used for analyzing and visualizing SECM experiments.

Before recording SECM images, the SECM probe was positioned close to the sample by performing a probe approach curve (PAC) toward the sample surface. The PAC was executed with a probe velocity of 1 μm s^–1 and a probe potential of E_probe = 0.5 V. The PAC was stopped when the probe current reached 50% of the current measured in bulk solution. At that distance, SECM images were recorded with a pixel size of 10 μm, a scan rate of 100 μm s^–1, and a waiting time of 10 ms. Feedback and SG/TC mode images were recorded in a solution of 1.5 mM FcMeOH and 0.2 M KNO₃. Feedback mode images were obtained with E_probe = 0.5 V and the substrate at open circuit potential. The SG/TC mode images in FcMeOH were recorded with E_probe = −0.2 and E_substrate = 0.5 V, respectively. HER experiments were performed in a solution of 0.5 M H₂SO₄. Among these experiments, a linear scan voltammogram was recorded at the FePSe₃ sample, decreasing the potential from 0.0 to −0.8 V at a scan rate of 5 mV s^–1. To obtain images of the local catalytic activity of the sample toward the HER, SECM images of the local hydrogen evolution were obtained using potentials of E_probe = 0.2 V and E_substrate = −0.5 V. XRD using a diffractometer (SmartLab 3 kW, Rigaku) with a Bragg–Brentano geometry was employed for the structural analysis of the sample. The optical images and height profile measurements of the sample surface were carried out using CLSM (Olympus LEXT OLS4100) with a laser light source (λ = 405 nm).

Results and Discussion

Before electrochemical investigation, an SEM image and EDS elemental maps of the sample were recorded to evaluate the purity and location of the crystal (Figure 2). The SEM image in Figure 2A shows a black area corresponding to the carbon/epoxy matrix surrounding the crystal, which is visible as a gray structure. The EDS elemental maps of Fe, P, and Se (Figure 2B–D, respectively) show a mostly homogeneous distribution within the crystal, with lower amounts detected at large cracks in the sample. Furthermore, local carbon impurities are visible in the crystal (Figure 2E). Nevertheless, most of the crystal surface appears clean. From the EDS spectrum in Figure 3A, atomic percentages for Fe, P, and Se of 21.0%, 19.7%, and 59.3% were derived, indicating high purity of the crystal. The purity of the sample was confirmed by XRD (Figure 3B). The sample gave characteristic crystalline peaks at their respective 2θ value. The values were found to be similar to the ones in the spectrum provided by the crystal manufacturer.⁴⁶ The application of hydrogen evolution to the sample resulted in changes in the relative peak intensity in the FePSe₃ sample. Figure 3C shows the XRD spectrum of the FePSe₃ sample after HER. Because the crystal was embedded into a conductive carbon epoxy matrix, the peaks stemming from this matrix are visible as well. They are marked by a *-symbol. XRD spectra of the carbon epoxy matrix and the crystal embedded in said matrix after HER application are given in Figure 3D. Comparing Figure 3B,C, it can be seen that all XRD peaks of the pristine FePSe₃ crystal are present post-HER as well, and no additional peaks aside the matrix peaks appeared. Thus, it can be assumed that the crystal edges did not undergo reorganization during the HER.

Figure 2.

SEM image of the FePSe₃ sample (A) with EDS elemental maps of Fe (B), P (C), Se (D), and C (E).

Figure 3.

EDS (A) and XRD spectra of FePSe₃. (B) XRD of the FePSe₃ crystal prior to sample preparation and HER. (C) XRD spectrum of the crystal embedded in the carbon matrix after HER. (D): Top: XRD of the conductive carbon matrix the FePSe₃ crystal was embedded into. Bottom: FePSe₃ sample after HER, embedded in the carbon matrix (same spectrum as in C).

Prior to spatially resolved electrochemical analysis of the FePSe₃ crystal, an LSV in 0.5 M H₂SO₄ was recorded (Figure 4A). Upon scanning toward more negative potentials, the measured current starts decreasing at a potential of −0.5 V due to the beginning of the HER. Decreasing the potential further led to a lower and noisier cathodic current resulting from H₂ bubble formation at the sample surface. Because a continuous and steady hydrogen evolution is required for SECM imaging, a potential of −0.5 V was applied to the crystal sample during the SECM experiments for imaging the HER.

Figure 4.

(A) Linear scan voltammogram recorded with the FePSe₃ sample in 0.5 M H₂SO₄ at a scan rate of 5 mV s^–1. (B) PAC toward the graphite/epoxy region of the FePSe₃ crystal sample. Parameters: max speed: 1.0 μm s^–1, step width: 1.0 μm, E_probe = 0.5 V.

Before SECM imaging, the probe was brought close to the sample by performing a PAC toward the carbon matrix around the crystal (Figure 4B). As the probe approached the surface, the measured current started to decrease significantly after the probe traveled 60 μm. The approach was stopped when the current decreased by 50% with respect to the initially measured value.

For correlating the local activity of the FePSe₃ crystal, an OM image and an SEM image were recorded (Figure 5A+B). Because the topography of samples can impact the current measured in SECM, a CLSM image of the sample was recorded as well. The CLSM image in Figure 5C shows the height profile of the investigated area. The feedback mode image (Figure 5D), indicating the local conductivity of the sample, shows that the electron transport at the crystal surface is nonuniform. While the central piece of the crystal appears uniformly conductive, the left part shows local differences in conductivity (highlighted by a blue box, see Figure 5G for better visibility). In that region, the areas where higher currents were recorded are located where cracks (and thus crystal edges) are visible in the image in Figure 5A. Moreover, in the bottom left of the optical image, both low and high currents were measured by SECM. This high contrast in the SECM image indicates a low substrate-to-tip distance and that this piece of crystal protrudes from the sample surface. The CLSM image in Figure 5C proves that the crystal is protruding from the surface compared to the rest of the imaged area. The SG/TC SECM image in Figure 5E gives an insight into the local electrochemical activity of the FePSe₃ crystal surface. The image shows well-pronounced local differences over the entirety of the investigated area. Especially in the left highlighted area (extracted in Figure 5H), the regions where a high current was recorded correlate to cracks visible in the optical image in Figure 5A. Because high currents in the SG/TC mode image indicate a high electrochemical activity, we can conclude that the edges of the crystal tend to be more electrochemically active than the basal planes. To investigate whether the electrocatalytic activity for the HER follows the same trend, another SG/TC mode image in 0.5 M H₂SO₄ was recorded. This resulted in an image showing local differences in the HER (Figure 5F). Relatively large current differences were located at the top and bottom borders of the image. Consequently, clearly distinguishing edge versus basal plane activity in these regions is not possible. In the left highlighted part (see Figure 5I for an extract of that area), however, more clear local differences in current were measured. Here, lines of high current, and thus high electrocatalytic activity, can be localized. They correspond to both electrochemically active regions (Figure 5D) and cracks within the sample (Figure 5A). Furthermore, the current pattern recorded in SECM images does not match the topography of the sample shown in Figure 5C, and consequently, these patterns of high currents are not caused by topographic effects. Thus, these results show that edges of FePSe₃ are both more electrochemically and electrocatalytically active than the basal planes.

Figure 5.

OM (A), SEM (B), CLSM (C), and SECM images of the FePSe₃ sample (D–F). (D) Feedback SECM image recorded in 1.5 mM FcMeOH. E_probe = 0.5 V. (E) SG/TC SECM image recorded in 1.5 mM FcMeOH. E_probe = −0.2 V, E_substrate = 0.5 V. (F) SG/TC SECM image recorded in 0.5 M H₂SO₄. E_probe = 0.2 V, E_substrate = −0.5 V. Images (G–I) are extracted from the SECM images (D–F), highlighting the area included by the blue boxes. (G) Feedback SECM image. (H) SG/TC SECM image recorded in FcMeOH. (I) SG/TC SECM image recorded in H₂SO₄.

Conclusions

In this work, we investigated whether the theoretical prediction that edges of MPX₃ materials are more electroactive than their basal planes is valid. The local electrochemical investigation of an MPX₃ representative, namely, FePSe₃, via SECM has shown that edges exhibit both increased electrochemical and electrocatalytic activity toward the HER. Thus, our measurements indicate that the ideal MPX₃-based electrocatalyst is rich in edge planes. This knowledge is very important for future catalyst design. It also adds MPX₃ compounds to the family of materials where edges and defects are more active than the basal planes of 2D materials. Our work took a qualitative approach for basal plane and edge activity characterization. Thus, this study opens the door toward quantitative analysis of the activity of the basal planes and edges of MPX₃ compounds. Another important question is whether the edges and basal planes of single layers of FePSe₃ show similar characteristics as the bulk material analyzed in this study.

Glossary

Abbreviations

EDS

energy-dispersive X-ray spectroscopy

HER

hydrogen evolution reaction

LSV

linear scan voltammogram

OM

optical microscopy

PAC

probe approach curve

PDMS

polydimethylsiloxane

SECM

scanning electrochemical microscopy

SEM

scanning electron microscopy

SG/TC

substrate generation/tip collection

Author Contributions

S.W. performed SECM, OM, and EDS measurements and wrote the paper. C.I. was involved in the conceptualization and discussion. A.K.K.P. performed the XRD and CLSM measurements. M.P. supervised and conceptualized this work. All authors have given approval to the final version of the paper.

The authors declare no competing financial interest.