Synthesis and Characterization of 3-DOM IrO2 Electrocatalysts Templated by PMMA for Oxygen Evolution Reaction

Three-dimensional ordered macroporous (3-DOM) IrO2 material was prepared using PMMA as a template and ammonia as a chelator. These 3-DOM IrO2 honeycomb arrays showed a large surface area and ordered macropores (155 nm in diameter) cross-linked by secondary mesopores. Internal structures of 3-DOM IrO2 material were observed microscopically through these secondary pores. According to the X-ray diffraction (XRD) and X-ray photoelectron spectroscopy (XPS) spectra, 3-DOM IrO2 has a rutile crystal structure and is mainly composed of iridium dioxide. In acidic electrolytes, the overpotential of 3-DOM IrO2 material at 0.5 mV cm−2 was only 0.22 V. Accelerated durability tests demonstrated excellent durability of 3-DOM IrO2 as an oxygen evolution reaction (OER) catalyst.


Introduction
Hydrogen, as an environmentally friendly energy carrier with high thermal value, has many advantages, which solves the energy problems and environmental issues associated with other energy resources [1][2][3]. The preparation of hydrogen is mainly based on fossil fuel cracking and water electrolysis. Solid polymer electrolyte (SPE) water electrolysis is considered to be one of the most promising methods for hydrogen production because of its high energy conversion efficiency and power density. Hydrogen production systems are safe and clean. Products of this process are just oxygen and hydrogen, both of which possess high purity [4]. However, some problems associated with SPE water electrolysis still exist, including high anodic overpotential and the high cost of the system [4,5].
Despite numerous research studies in this area, high anodic overpotential remains a great challenge for oxygen evolution reaction (OER). Additionally, the stability of electrocatalysts is another factor restricting widespread SPE applications. The majority of metals and their oxides cannot be stabilized under high anode potential and strong acidic conditions; thus, choices for anode catalysts are very limited. Platinum metals have good catalytic performance because of their excellent adsorption affinity to hydrogen and excellent catalytic properties. Thus, Pt has been widely used in SPE water electrolysis [6,7], fuel cells [8][9][10][11], and hydrogen storage [12,13]. However, Pt-based catalysts are not suitable for OER because of the high overpotential of Pt as it forms platinum oxides, which result in the poor conductivity of Pt [14]. Fortunately, Ir and Ru have good catalytic OER activity. Their oxides also have good conductivity and are stable in acidic environments. Interfaces of Ir and Ru oxides are also very reactive.
To increase the stability of the catalyst at the anode during SPE water electrolysis, a lot of work has been performed especially involving dimensionally stable anodes (DSA). Durable material to load on IrO 2 and RuO 2 were selected, most of which were oxides of transition metals. Although these materials have excellent chemical stability, their conductivity is poor, which makes OER electron transfer dependent on IrO 2 and RuO 2 . The result does not reduce the amount of precious metals, and only when the content of IrO 2 in the IrO 2 -ZrO 2 solid solution reaches 80% does the catalyst demonstrate good performance [15]. Other solid oxides forming catalysts with high levels of IrO 2 and/or RuO 2 are TiO 2 [16][17][18][19], SiC [20], SnO 2 [21][22][23][24], Ta 2 O 5 [25,26], and Nb 2 O 5 [27], etc. Research for a variety of other support materials is still underway.
To increase the catalytic activity during SPE water electrolysis, the activity of precious metal catalysts needs to be improved, for example by studying and analyzing its surface area and micropore structure [28][29][30][31]. Recently, Li et al. [30] synthesized a nanoporous IrO 2 catalyst based on a template-free ammonia complex. Unlike the Adams fusion method, ammonia was used as a chelating agent to replace Cl in H 2 IrCl 4 . This IrO 2 catalyst demonstrated the highest specific surface area ever reported (363.3 m 2 g −1 ) as well as outstanding performance. Ortel et al. [31] synthesized ordered mesoporous IrO 2 film templated by PEO-PB-PEO block copolymers. The films possess nanocrystalline walls and ordered pores, which enhance the durability and activity of OER.
Three-dimensional ordered macroporous (3-DOM) materials are novel molecular sieves with uniform and ordered pores and well-controlled pore structures. The ordered porosity of the 3-DOM material is very conducive to an even distribution of the electrolyte between the electrode, which improves the mass transfer of ions to the electrode surface [32]. This can increase the number of reactive sites and improve the catalytic activity of materials. The interconnected structure of 3-DOM materials can increase conductivity [33], which makes them suitable as electrochemical catalysts. 3-DOM materials were successfully used in a lithium-ion battery and fuel cell [32,34]. However, few 3-DOM oxide materials have been reported for OER electrocatalysts. Hu et al. first synthesized 3-DOM IrO 2 using a SiO 2 microsphere template [35,36], in which 3-DOM IrO 2 exhibited an ordered honeycomb array of macropores. The overpotential of this OER catalyst was only 0.22 V at 0.5 mA cm −2 . However, organic templates have not been reported for the preparation of 3-DOM IrO 2 . Herein, we synthesized 3-DOM IrO 2 using a PMMA template. Compared to the SiO 2 template, PMMA templates can be removed by annealing, which makes the preparation process very simple, thereby reducing its cost. However, the decomposition temperature of the PMMA template is under 300 • C, which is too low for iridium chloride to form iridium oxide. Because the formation of 3-DOM structures is difficult under such conditions, we used ammonia as a chelator, which reacts with iridium chloride to form ammonium chloride and iridium hydroxide. Ammonium chloride is volatile and easily decomposed. Iridium hydroxide is easily dehydrated to produce iridium oxide under 300 • C, which can protect the 3-DOM structure against collapse. 3-DOM IrO 2 was characterized by SEM, Brunauer-Emmett-Teller (BET), X-ray diffraction (XRD), and X-ray photoelectron spectroscopy (XPS). Ordered macropores were clearly seen with a significant decrease in overpotential.

Preparation of PMMA Templates
Conventional emulsifier-free emulsion polymerization was used to prepare polymethyl methacrylate (PMMA) templates [37]. First, 260 mL of deionized water was placed into a 500 mL round-bottomed flask and heated to 70 • C inside a heat-collecting thermostatic magnetic stirrer. After nitrogen was passed through the flask for 15 min, potassium persulfate (KPS, SinopharmChemical Reagent Co., Ltd., Shanghai, China) and methyl methacrylate (MMA, Sinopharm Chemical Reagent Co., Ltd., Shanghai, China) were added in a ratio of 1:118. The whole mixture was stirred for 1 h under nitrogen. After that, the PMMA emulsion was centrifuged and washed three times with deionized water, following which PMMA microspheres were dispersed ultrasonically in deionized water. A certain amount of PMMA emulsion was placed into a 25 mL beaker, which was then covered by tin foil with holes. PMMA microspheres were settled down and dried at room temperature. Bright fluorescence on the surface indicated that the PMMA microspheres were hexagonally close-packed.

Preparation of 3-DOM IrO 2
IrCl 4 ·2H 2 O (Shanxi Kaida Chemical Engineering Co., Ltd., Shanxi, China) was added to H 2 O and isopropyl alcohol (IPA, Sinopharm Chemical Reagent Co., Ltd., Shanghai, China) (H 2 O: IPA = 3:1). Then, 4 mL of IrCl 4 ·2H 2 O 0.03 mol L −1 was added to PMMA templates until it was immersed. After 30 min, the excess solution was removed by a dropper. PMMA templates were put into the oven at 70 • C. After drying, the above steps were repeated three times. Finally, the templates were heated to 450 • C in air at a rate of 5 • C/min in a tube furnace for 4 h. The final product was a black powder (marked as non-ammonia IrO 2 ). Ammonia (Sinopharm Chemical Reagent Co., Ltd., Shanghai, China) (IrCl 4 ·nH 2 O: NH 4 ·H 2 O = 1:4) was then added to the above solution that contains isopropyl alcohol, H 2 O, and IrCl 4 ·2H 2 O. 3-DOM IrO 2 was prepared by repeated preparation of non-ammonia IrO 2 . Colloidal IrO 2 was prepared by the colloidal method [35].

Physical Characterization
Scanning electron microscopy (SEM) was performed using a Hitachi S-4800 microscope operated at 8 kV. X-ray photoelectron spectroscopy (XPS) spectra were recorded by an ESCALAB250Xi spectrometer using Mg Kα as an X-ray source. X-ray diffraction (XRD) spectra were recorded by Bruker D8 Advance using Cu-Kα radiation. Particle sizes were measured by the dynamic light scattering (DLS) technique. Specific surface areas and pore distribution were obtained by implementing Brunauer-Emmett-Teller (BET) and Barrett-Joyner-Halenda (BJH) methods, respectively, using N 2 gas absorption/desorption data collected by Micromeritics ASAP 2020 (BELSORP-max).

Electrochemical Characterization
Electrochemical properties were tested using a three-electrode system at room temperature. The electrolyte was 0.5 mol L −1 H 2 SO 4 solution. The counter electrode was platinum foil, and a saturated calomel electrode (SCE, 0.242 V) was a reference electrode. The working electrode was a glassy carbon (GC) 5 mm in diameter. The catalyst ink was prepared using 5 mg of testing material, 980 µL of isopropanol, and 20 µL of 5% Nafion. Both catalysts had a load of 106 µg cm −2 . Cyclic voltammograms (CV) were measured between 0 V and 1.4 V at a 50 mV s −1 scan rate. Steady-state polarization curves for OER were obtained at a 5 mV s −1 scan rate. Before all tests, the electrode was activated by performing 20 cycles of the CV curve at the scanning rate (500 mV s −1 ). All electrochemical tests were performed using the CHI660 workstation.

Physical Characterization
Since PMMA microspheres serve as a synthesis template, their particle size distribution is an important physical property. Analysis of the particle size distribution of PMMA by DLS demonstrates that the average diameter of PMMA microspheres is~220 nm (Figure 1a). Polydispersity of PMMA microspheres is 1.7%, which is indicative of good uniformity. A single particle size of PMMA microspheres is an important reason for forming a regular colloidal crystal template. Also, sizes of PMMA microspheres are controlled by the quantity of the initiator and MMA. The PMMA template consists of hexagonally close-packed PMMA microspheres (Figure 1b), which have a uniform distribution of particle size (the diameter of the particles is 220 nm). An obvious layered structure can be clearly observed in Figure 1b. All layers have the same close-packed hexagonal arrangement of the microspheres. Several point defects are observed as well as some microsphere misalignments (markings in Figure 1b). These defects have no impact on the template ordered macroporous material formed because of this ordered regular template. The PMMA template is covered by the precursor, as observed in Figure 1d. After immersion, the PMMA template retains the same close-packed hexagonal arrangement of the microspheres. The PMMA colloidal crystal template exhibits fluorescence (Figure 1c) under natural sunlight, which is a typical Bragg behavior of ordered microspheres. 3-DOM IrO 2 exhibits a three-dimensional ordered honeycomb macroporous structure ( Figure 2). Most of the macropores are 155 nm in diameter, which is~30% smaller than the diameters of PMMA microspheres in the template, probably because of the carbonization and dehydration of the organic templates during annealing. PMMA contains many hydrogen and oxygen elements, and dehydration can greatly reduce the microsphere size as well as carbonize the PMMA template. The honeycomb pore structure can be seen in the high-magnification SEM image. In addition, small pores in the walls of the large pores, which can connect to the surrounding large pore structures, can also be clearly observed (Figure 2b). These pore structures are formed because when accumulated hexagonally close-packed PMMA microspheres are in contact with each other, the precursor solution cannot pass through the points of contact of the PMMA spheres during impregnation. After the removal of the PMMA template, these contact points form small pores. The solid structure of the 3-DOM IrO 2 can be seen through these small pores. These pores are favorable for the even distribution of the electrolyte between the electrodes, which will improve the mass transfer of ions to the surface of the electrodes. This can also increase reactive sites and improve the catalytic activity of the material. Additionally, such interconnected structures can increase the overall conductivity. A few pores can be observed in non-ammonia IrO 2 (Figure 2c). These pores are broken and disordered probably because the decomposition temperature of IrCl 4 is higher than that of the organic template. Before IrCl 4 decomposes, the decomposition of the organic template causes the pore structure to collapse (Figure 2d). The specific surface area of the OER catalyst is an important physical property. Because of its microporous structure, it has a large specific surface area. Figure 3a exhibits N 2 adsorption/desorption isotherms for 3-DOM IrO 2 and non-ammonia IrO 2 . According to the International Union of Pure and Applied Chemistry (IUPAC) classification [38], the curve of 3-DOM IrO 2 belongs to the type IV with H3-type hysteresis, which is usually associated with highly consistent mesopore structures. Non-ammonia IrO 2 also shows the type IV with H3-type hysteresis. However, the hysteresis loop is smaller than that in 3-DOM IrO 2 , showing that 3-DOM IrO 2 has more mesopores. According to the BJH pore-size distribution analysis (Figure 3b), the pore sizes of both materials are concentrated below 20 nm. Diameters of 3-DOM IrO 2 mesopores are~1.2 nm, while diameters of non-ammonia IrO 2 mesopores are~2.1 nm. The BJH surface areas of the mesopores of 3-DOM IrO 2 and non-ammonia IrO 2 are 34.5 and 23.8 m 2 g −1 , respectively, indicating that 3-DOM IrO 2 has more mesopores. The BET surface areas of 3-DOM IrO 2 and non-ammonia IrO 2 are 67.3 and 37.1 m 2 g −1 , respectively. The BET surface area of 3-DOM IrO 2 is close to the mesoporous IrO 2 structure reported by Ortel et al. [31].   Non-ammonia IrO 2 pattern does not show rutile IrO 2 peaks. However, characteristic peaks of the elemental Ir at 2θ of~40 • , 47 • , and 69 • (which correspond to (111), (200), and (220) planes) could be clearly seen. This indicates that 3-DOM IrO 2 contained metallic Ir formed because of template carbonization; after PMMA carbonization, internal Ir 4+ had no access to air. Thus, a small amount of IrO 2 was reduced to Ir. However, Ir 4+ in the precursor that has no ammonia is IrCl 4 , and since Cl is less electronegative than O, Ir 4+ was reduced almost entirely. XPS analysis of 3-DOM suggests that IrO 2 is present in the catalyst ( Figure 5). No other impurities are observed on the surface (Figure 5a). Characteristic peaks of Ir are clearly seen as well. High-resolution spectra of the Ir4f demonstrates characteristic peaks at 61.5 eV (Figure 5b). This binding energy value agrees with values obtained from nanoporous IrO 2 synthesized by the template-free ammonia complex [30].

Electrocatalytic Characterization
The electrocatalytic performances of two materials are shown in Figure 6. The peaks observed in the CV curve of colloidal IrO 2 (Figure 6a) clearly indicate Ir transformation (Ir III /Ir IV and Ir IV /Ir IV ), accompanied by protonation/deprotonation. The CV curve of 3-DOM IrO 2 demonstrated a distinct redox peak at a low potential, which was related to the elemental Ir in 3-DOM IrO 2 . Ir, similar to Pt and Ru, can absorb hydrogen in the hydrogen underpotential deposition region [39,40]. 3-DOM IrO 2 showed significantly larger current density in comparison to colloidal IrO 2 . Current responses consist of two effects, the redox pseudo-capacitance and the double-layer capacitance, both of which are related to the electroactive surface areas (ESAs). The linear sweep voltammetry (LSV) of two catalysts demonstrating their OER activities are shown in Figure 6b. 3-DOM IrO 2 demonstrates superior OER activity compared to colloidal IrO 2 . Compared to the thermodynamic minimum potential of OER, the overpotential of 3-DOM IrO 2 is 0.22 V at 0.5 mA cm −2 , which outperforms colloidal IrO 2 (0.25 V) and is the same as the 3-DOM IrO 2 material synthesized by the silica colloidal crystal template [35] and the mesoporous IrO 2 templated by PEO-PB-PEO copolymers [31]. However, the preparation of PMMA microspheres was easy as 3-DOM IrO 2 prepared by the PMMA template did not require acid solution treatment, thereby reducing the cost of preparation. Figure 6b suggests that 3-DOM IrO 2 (106 µg cm −2 ) has the same catalytic activity as colloidal IrO 2 (265 µg cm −2 ), while the OER activity of 3-DOM IrO 2 is outstanding compared to that of colloidal IrO 2 , because of their different ESAs (Figure 6a). The rutile IrO 2 shows lower O 2 -evolving overpotential than the metal Ir [41]; therefore, the metal Ir is not the main reason for the increase of catalytic activity. Catalytic data were normalized relative to the corresponding CV charges (inset of Figure 6c). The curves represent intrinsic activities of the catalysts, which are almost identical (Figure 6c). Thus, intrinsic OER activities are identical, and the difference in porosity does not affect catalytic performance. However, 3-DOM IrO 2 shows superior catalytic performance under the same mass condition. The slope of the Tafel curve for 3-DOM IrO 2 is 57.9 mV dec −1 (Figure 6d), which is lower than that of colloidal IrO 2 (which is equal to 62.2 mV dec −1 ). The Tafel slope is similar to that of nanoporous IrO 2 and within the normal range of conventional IrO 2 [30]. This also confirms the robust catalytic activity of 3-DOM IrO 2 , indicating a faster rate for the OER. Figure 7 shows current change as a function of time. From these curves, the OER stabilities of both catalysts in acidic environments can be obtained. Current density dropped initially for both catalysts. After 3000 s, the decline of the current density slowed down. However, 3-DOM IrO 2 still showed higher current density than that of colloidal IrO 2 . Slopes of the fitting curves in the range of 15,000-20,000 s (insert in Figure 7) are −2.09× 10 −4 and −2.37× 10 −4 , respectively. Thus, the current change of colloidal IrO 2 deceases faster and 3-DOM IrO 2 exhibits superior stability than that of colloidal IrO 2 .

Conclusions
Synthesis of 3-DOM IrO 2 using a PMMA template and ammonia sol has been performed. The issue of the organic template having difficulty in maintaining the 3-DOM IrO 2 structure has been solved by the ammonia sol. The sol consists of iridium chloride, isopropanol, and ammonia. IrCl 4 and ammonia form a stable complex solution, in which NH 4 + and Cl − form ammonium chloride, and ammonium chloride can easily be removed. The prepared 3-DOM IrO 2 material has a honeycomb array macroporous structure with an average pore diameter of 155 nm and a large surface area. Many mesoscale pores are observed on the wall of the macropores. 3-DOM IrO 2 demonstrates superior OER activity compared to colloidal IrO 2 , and the overpotential of 3-DOM IrO 2 is lower than that of colloidal IrO 2 by about 30 mV at 0.5 mA cm −2 . In the same catalytic activity, the loading of 3-DOM IrO 2 is much lower than that of colloidal IrO 2 . Our novel method to synthesize the 3-DOM IrO 2 catalyst can reduce the anode overpotential of water electrolysis. The results suggest that the 3-DOM IrO 2 catalyst is a promising material for OER.