Porous Iron Single Crystals at 2 cm Scale Delivering Enhanced Electrocatalysis Performance①

CHENG Fang-Yuan MYUNG Jae-Ha XIE Kui ②

a (College of Chemistry and Materials Science, Fujian Normal University, Fuzhou 350007, China)

b (Key Laboratory of Design & Assembly of Functional Nanostructures, Fujian Institute of Research on the Structure of Matter, Chinese Academy of Sciences, Fuzhou 350002, China)

ABSTRACT Porous single crystals would significantly enhance their catalysis functionalities owing to the combination of structural coherence and porous microstructures. Porous single crystals have wormhole microstructures and then we define them as wormcrystals. The twisted surfaces in porous microstructures would produce surface lattice distortions that give rise to high-energy active surfaces. Here we grow porous iron single crystals at an unprecedented 2 cm scale with a lattice reconstruction strategy and create high-energy surfaces through the control of lattice distortions within a thickness region of 1～2 nm. The porous iron crystal therefore boosts electrochemical reduction of nitrobenzene to aminobenzene with ～100% conversion and > 95% selectivity. The exceptionally high current densities with porous iron crystals represent the first level electrocatalysis performance. The current work would open a new pathway not only to the creation of high energy surfaces but also to the growth of porous single crystals at large scales in wealth of other materials.

Keywords: porous single crystals, twisted surfaces, lattice distortions, electrocatalysis;

1 INTRODUCTION

Altering the catalysis properties of a material would require the identification of the structural features that could deliver catalytically active surfaces. Active surfaces usually originate from material’s high-energy surfaces with specific atom coordination structures and chemical compositions[1-3]. Nano- crystals with high surface areas therefore receive a great deal of attention because of their fascinating physical and chemical properties mainly dominated by the synergistic effect of shape, size and composition. The crystal facets, vertices and edges mainly govern the chemical interaction of catalysts with reactants in catalysis reactions. In essence, the surface atom arrangement of materials determine their physical and che- mical properties while the synergistic control of coordination structure and chemical composition at surface would be one of the effective approaches to tailor the catalysis activity.

For a given crystalline material, the surface atom arrange- ments would be confined by the long-range ordering of lattice with specific orientations while the surface free energies are mainly dominated by the atom coordination structure and chemical composition[4-6]. Engineering high-energy surfaces not only relate to controlling the size and shape of nanocrystal but also involving chemical doping with impurity atoms. Porous single crystals, combining the advantages of long- range ordering and porosity, would create high-energy surfaces through the creation of twisted surfaces leading to lattice distortions within a specific thickness region. This may thus provide an alternative way to create high-energy surfaces that are kinetically trapped. The lattice distortions at twisted surfaces induced by curvature produce electrochemically active surfaces while the structural coherence provides electronic connectivity, thus reducing energy losses at grain boundaries and keep high stability and activity.

Iron metal with the advantages of ready availability, low price and environmental benign has made itself an ideal alternative to precious metals in catalysis[7-13]. Iron catalysts have demonstrated promising activity and stability in many catalysis reactions and the activity normally originates from the dispersed nanoparticles, clusters and even single atoms. The active sites are either accommodated in hollow structures like porous carbon or embedded on oxide substrates to prohibit their agglomeration that leads to catalysis per- formance degradation. Iron nanocrystals with faceted surfaces have also confirmed to be effective to tailor high-energy surfaces. The catalysis activity normally comes from the unique coordination structure and even from interaction with substrates[14,15]. Porous iron single crystals would signifi- cantly enhance the catalysis functionalities owing to struc- tural coherence and porous microstructures. The twisted surfaces would be confined in high-energy states that give rise to catalytically active surfaces through the control of lattice distortion within a specific thickness region.

Aminobenzene, as a key intermediate, is produced by selective hydrogenation of nitroaromatic hydrocarbons, which involves many noble metal catalysts and even higher reaction temperature (35～125 ℃) and higher hydrogen pressure (0.6～0.9 MPa). The existing catalysts often fail to meet the dual requirements of activity and selectivity while non- selective hydrogenation of intrinsically active catalysts (such as Pd/C) results in by-products that are difficult to purify[14]. Here we demonstrate a novel electrochemical reduction of nitrobenzene to aminobenzene with non-noble porous iron single crystals at ambient conditions. At porous iron cathode, the protons are in situ used to electrochemically reduce nitrobenzene while the highly active surface in porous microstructures will promote the electrochemical hydro- genation process with efficient conversion to aminobenzene under external bias.

Here we grow porous iron single crystals at an unprece- dented 2 cm scale through a lattice reconstruction strategy and tailor the average curvatures in different porous microstruc- tures. We discuss the lattice distortions at twisted surfaces and closely correlate them to electrocatalysis activity. We show the efficient electrochemical reduction of nitrobenzene to aminobenzene with the porous iron single crystals.

2 EXPERIMENTAL

2. 1 Crystal growth

Parent FeCO3crystals with dimensions of 20mm × 10mm × 0.5mm are utilized to convert into porous iron single crystals. The FeCO3crystal is washed and dried with absolute ethanol for several times and placed in an alumina tube vacuum system. A constant flow rate of H2/Ar gas (50～300 sccm) is introduced and the pressure is maintained at 20～700 mbar. The heating rate is 10 ℃·min-1to 300～700 ℃ for 20 hours.

2. 2 Microstructure characterization

XRD is used to examine the phase formation (Cu-Kα, Mniflex 600). The surface morphology of porous crystal was studied using a field emission scanning electron microscope (FE-SEM, SU8010). Crystal structure is analyzed using transmission electron microscopy (TEM) combined with focused ion beam (FIB) (Zeiss Auriga). We perform selected area electron diffraction (SAED) at 200 KV (Tecnai F30) and Cs-STEM at 300 KV (FEI Titan3 G2 60-300).

2. 3 Nitrobenzene electroreduction

We test the reduction of nitrobenzene with porous iron single crystals on an electrochemical workstation (IM6, Zahner) in a three-electrode mode. Pt foil is the counter electrode and a saturated calomel electrode serves as the reference electrode. The cell has two-compartments separated by an anion exchange membrane (Nafion® 212). For this electrochemical reduction experiment, 200 mg nitrobenzene was dissolved in 1 M Na2SO4solution (pH = 7.3). And the external bias is set versus RHE. The conversion of nitrobenzene is analyzed by gas chromatography and mass spectrometry (Agilent Technologies 7820A).

2. 4 Theoretical calculation

Density functional theory (DFT) calculations are performed using the Vienna ab-initio simulation package (VASP)[16]. Within the projector augmented wave (PAW) framework, the plane wave cut-off is set to 500 eV. The generalized gradient approach (GGA) was used including Perdew-Burke-Ernzerhof (PBE) functional[17]. We calculate 4 different primitive cells: a standard cell and 3 cells with different lattice parameters to simulate the lattice distortions. An 8 × 8 × 8 k-point grid is used for all cells. A p (3×3) superstructure of the [110] surface is used for these slab models with four layers. A 22 Å vacuum region is inserted to avoid the interactions.

3 RESULTS AND DISCUSSION

We use parent crystal FeCO3to grow porous iron single crystals. We treat the substrates (10 × 20 × 0.5 mm3) with <110> orientation in vacuumed H2/Ar atmospheres (20～700 mbar) at 300～700 ℃. Fig.1a shows the X-ray diffraction (XRD) patterns of parent crystal FeCO3, which confirms the exposed [110] facet[18]. The porous iron single crystal shows a porosity of 74% with the optical photo shown in the inset. Fig.1b and 1c depict the microstructures of porous iron single crystals with pore sizes ranging from 200 to 850 nm grown along the <110> direction of parent crystals. The porosities are similar for the three porous crystals even though the pore sizes are different.

Fig.1. (a) XRD of iron crystals grown on the [110] FeCO3 substrates. The digital images of porous crystals have dimensions of 20 × 10 × 0.5 mm3. (b), (c) and (d) Microstructures of porous single crystals with different pore sizes

Fig.2 shows the cross-sectional view of the porous iron single crystal grown along the <110> direction of parent crystals, which demonstrates the homogeneous distribution of interconnected pores with the diameter of 50～100 nm. We further use selected area electron diffraction (SAED) to validate the single-crystalline nature at different locations on the porous skeleton. We observe identical facet orientations well consistent with the XRD diffraction even at different locations, which further confirms the long-range ordering of single crystalline nature. The [110] facet is the low-index facet with the lowest surface free energy which therefore dominates the growth of porous iron crystals with the growth mechanism (Fig.3). Fig.3a～3c show the crystal structure of the parent crystal FeCO3viewed from the <110> direction which demonstrates the vertical and periodical lattice channels distributed closely along the Fe–O octahedra. The removal of C and O atoms in lattice would proceed accompanied with the polyhedron collapsing while the channels would facilitate atom diffusion leaving the open iron skeleton which simultaneously transforms into porous iron single crystals to maintain a low-energy state in vacuumed atmosphere. The unique structural feature with lattice channels would also facilitate the growth of porous iron single crystals during the lattice reconstruction process.

We use focused-ion-beam nano-tomography to disclose the microstructure of porous iron crystals grown along the <110> direction of parent crystals. The average pore radius and throats length calculated from discrete frequency are about 200 and 1000 nm in Fig.4a, respectively. The 3D-inter- connected pores therefore create twisted surfaces throughout the porous iron single crystals. We simulate the curvature in microstructures and give an average value of 10.1 for #1 iron crystal in Fig.4d. We then grow another two porous single crystals with larger pore sizes. Similarly, the #2 iron crystal in Fig.4b gives the average pore radius of 490 nm and throats length of ～1000 nm in the ～100% 3D-interconnected porous microstructures. The average curvature value reaches 8.1 in the porous microstructure. The #3 iron crystal in Fig.4c gives the average pore radius of 840 nm and throats length of ～1000 nm while the average curvature value finally reaches 6.4. The specific surface areas reach 9～10 m2·g-1while the pore sizes range from 200 to 900 nm well consistent with the SEM results (Fig.5). The average curvatures in different microstructures represent the twisting extent of the exposed surfaces in the porous microstructure, which therefore leads to significant lattice distortions that give rise to high-energy surfaces.

Fig.2. (a～d) and (f～i) SAED pattern at different locations on the cross-section of the porous iron single crystals. (e) Cross-sectional view of the porous [110] iron single crystals with different locations for SAED patterns labeled

Fig.3. (a～c) Schematic diagram of the structural transformation from the [110] FeCO3 precursor to the [110] plane porous Fe single crystal

Fig.4. (a～c) show the distribution of pore and throat lengths for samples # 1, # 2, and # 3 with different pore sizes. (d～f) represent the curvature of the former microstructure, respectively

Fig.5. Mean pore diameter (a) and specific surface area (b) of the #1, #2, and #3 samples of iron single crystal. The error bar represents the standard deviation of repeated measurements

We use a spherical aberration corrected scanning transmis- sion electron microscope (Cs-corrected STEM) coupled with focused ion beam to investigate the lattice distortions at twisted surfaces in porous microstructures. Fig.6a shows the Cs-corrected STEM of the porous iron single crystal viewed along the <001> direction with the iron atoms periodically stacked in the bulk, while the lattice spacing of 0.201 and 0.143 nm is assigned to [110] and [200] fringe as confirmed by the SEAD pattern in the inset[19]. The lattice distortions are clearly present at twisted surfaces within the thickness of 1～2 nm (Fig.6b～6d). The lattice spacings at twisted surfaces are increased by 16～26% for the porous iron crystals and the large twisting extent of exposed surfaces facilitates the lattice distortions. Larger curvatures enhance the twisting extents of exposed surface, which accordingly promotes the lattice distortions at twisted surface in the porous microstructures.

Fig.6. (a) Cs-corrected STEM image of the bulk of porous iron crystals. Inset corresponds to SAED pattern. (b), (c) and (d) respectively represent lattice distortions at twisted surfaces for #3, #2 and #1 porous single crystals with different curvatures

Fig.7a shows the electrochemical reduction of nitroben- zene to aminobenzene with porous iron single crystals. Crystals with porosity possesses both structural and com- positional coherences to accommodate the lattice distortions at twisted surfaces. The lattice distortions lead to significantly enhanced electrolysis activity with current density reaching ～400 mA·cm-2even only at 0.5 V for porous iron crystal. This represents the first-level current densities in most electrocatalysis performances, which further validates the enhanced activity with twisted surfaces through the control of lattice distortions within a specific thickness region. We operate the electrocatalysis reduction for a duration of 5 hours, which indicates a pretty stable electrode activity as shown in Fig.8. The optimum performance is observed with ～100% of conversion of nitrobenzene and > 95% of the aminoben- zene selectivity with applied bias up to –1.0 V in Fig.7b and 7c. We simulate the lattice distortion as observed and validate the changes of free energies of the twisted surfaces with different extents of lattice distortions. Fig.7d shows that the system energy is significantly rising even by 45～50% in the presence of lattice distortion at twisted surfaces, which further validates the high-energy surfaces with lattice distortions. And the rising of lattice distortion at a reasonable extent would further increase the surface energy. The combination of structural coherence and porosity therefore gives rise to high-energy twisted surface that are kinetically trapped, which thus leads to the electrochemically active surfaces in the porous microstructures.

4 CONCLUSION

In this work, we show enhanced electrocatalysis at twisted surfaces with lattice distortion in 2 cm porous iron single crystals. The combination of structural coherence and porosity in porous single crystals provides a unique advantage to create high-energy twisted surfaces that leads to the electroche- mically active surfaces. The presence of lattice distortions within the thickness of 1～2 nm at twisted surfaces therefore enhances the electrocatalysis activity. The highest perfor- mance is observed with ～100% of conversion of nitroben- zene and >95% of aminobenzene selectivity. The current work would open a new pathway not only to grow porous single crystals at large scales but also create high energy surfaces in wealth of other materials.

Fig.7. (a) I-V curves of electrochemical reduction of nitrobenzene with porous iron single crystals. (b) Conversion of nitrobenzene to aminobenzene with different porous iron single crystals. (c) Aminobenzene selectivity at different potentials. (d) System energy of porous iron single crystals with different lattice expansions

Fig.8. (a), (b) and (c) are the durability tests of porous iron single crystal cathode samples #1, #2 and #3 in 1 M Na2SO4 electrolyte, respectively