Study and fabricated electrocatalysts on the iro2 basis for oxygen evolution reaction in proton exchange membrane water electrolyzer

reaction and hydrogen evolution reaction on the catalytic materials based on IrO2 in PEMWE. - Introduction of development history of catalytic materials used in PEMWE and research & development of catalytic materials at the anode and cathode electrodes of PEMWE. 4 CHAPTER 2. EXPERIMENTAL AND RESEARCH METHOD 2.3. Fabrication of electrocatalyst materials on the IrO2 basis Three methods applied to catalyst synthesis are hydrolysis, Adams and Adams modified method. - Hydrolysis method: At first, metal precursors were dissolved in deionization water with the exactly calculated metal precursors. The aqueous solution was then heated (100°C) under air atmosphere and magnetically stirred for 1 hour. Afterward, sodium hydroxide (1 M) was added to the solution in order to obtain the precursor- hydroxide. This mixture was maintained under stirring and heat (100°C) for 45 minutes. The given precipitate was then filtered and washed with deionization water. After washing the precursor-hydroxide was dried for 5 hours at 80°C. Finally, the dryed paste was calcined in air at 450°C for 1 hour with a heating ramp of 5°C.min-1. - Adams method: metal precursors were added to isopropanol to obtain a total metal concentration of 0.01 M, this solution was ultrasonic for 30 minutes and magnetically stirred for 1-2 hours to ensure complete dissolution of the precursors, followed by the addition of 10 gram of finely ground NaNO3. The mixture was heated at 70 o C in air until completely dry. Using isopropanol as the precursor solvent appeared to give a more homogenous mixture than if water was used. The dry salt mixture was then placed into a preheated furnace. The fused salt oxide mixture was cooled slowly to room temperature then washed in deionised water to remove the remaining salts, filtered and dried in an oven at 80 o C. - Modified Adams method: the steps are the same as the Adams method differs only at the furnace stage: before heating at 500°C for 1 hour the 5 salt mixture need pre-heated at 325°C for 30 minutes at a rate of 5°C.min-1. 2.4. Physical and electrochemical cheracterization of electrocatalyst 2.4.1. Physical cheracterization The mechanisms of the thermal decomposition process of metal precursors to form oxide powders were studied by means of thermal gravity analysis (TGA). The physical phase and structure of the oxide powder catalysts were determined by x-ray diffraction (XRD). The surface morphology and particle size of the catalysts materials was examined by scanning electron microscopy (SEM) and transmission electron microscopy (TEM). Energy-dispersive X-ray spectroscopy (EDX) is used to determine the composition of the catalyst mixture. 2.4.2. Electrochemical cheracterization Catalytic layers preparations: The catalyst layer deposited on the carbon paper (AvCarrb 1071-USA) and the anode electrocatalyst loading was 4 mg.cm -2 ; the cathode electrocatalyst loading was 1 mg.cm -2 . To prepare the thin-film electrocatalyst layers, a homogeneous ink composed of powder catalyst, Nafion solution, isopropanol and deionised water were homogenized by stirring and ultrasonicing. The mixture was cast on a cacbon paper surface by sweep method and dried in air, this process was repeated until enough the loading. CV measurements are measured with three purposes: determination of catalytic activation and electrochemical processes on catalytic surfaces: examine the reversibility of the redox process; determination of catalystic degradation. The anodic polarization curve is carried out by scanning the linear potential over time with a constant speed of 0.5 mV.s -1 from - 250 VAg/AgCl to 1.3 VAg/AgCl (compared to the corrosion potential) in 0.5 6 M H2SO4. This method allows the determination of the steady-state current density, the starting potential of OER. Galvanostatic: the catalytic samples were polarized at high current density (200 mA.cm -2 ) in 0.5 M H2SO4 medium to accelerate the dissolution or inactivation process of anode, thereby quickly assessing elctrode’s lifetime. The voltage is recorded over time, the lifetime of the electrode is the measured time until the electrode is destroyed. 2.5. Fabrication and cheracterization of MEA Membrane electrode assembly is fabricated by heat-press method, pressing Nafion film between two diffusion gas layers covered by catalytic ink. The catalytic layer is fabricated by brushing on the cacbon paper surface and dried in air. This process was repeated until reaching enough the electrocatalyst loading of 4mg.cm -2 . In this thesis, a simple PEMWE electrolyte is designed, manufactured and installed. Materials and specifications of PEMWE are given in Table 2.5. Table 2.5. Materials and specifications of PEMWE Component Material Size (mm) MEA Made in section 2.5.1 23 × 23 Gasket Silicon 50 × 50 × 1 Separator plate Graphite AXF- 5Q (Poco) 50 × 50 × 3,2 Shell Acrylic 50 × 50 × 8 Current collector Gold plated copper 50 × 50 × 1 Bolt Stainless steel plastic wrap ⏀5 7 CHAPTER 3. RESULTS AND DISCUSSUIONS 3.1. Fabrication of IrO2 electrocatalyst Observed on both TGA graphs of the two synthetic methods are divided into two stages, the first stage occurs at low temperature, salt mixtures of both methods have a rapid weight reduction. Currently, it is mainly due to the evaporation of water molecules which adsorb physics and in the form of hydrate in Ir's salt. The second stage on the DTA and TGA is the complete thermal decomposition of the salts forming IrO2 powder. For hydrolysis method, this process occurs in the temperature range of 300-394 o C. For Adams method, the decomposition temperature to create IrO2 powder in the range of 350-605.6 o C. The result obtained by spectra of X-ray diffraction spectra have proved that the final product is IrO2 (Figure 3.3 and 3.4). From here, IrO2 catalysts will be furnaced at 300 o C, 400 o C, 500 o C and 600 o C follow hydrolysis method and 400 o C, 500 o C and 600 o C follow Adams method. Fig. 3.1. TGA and DTA diagram of (H2IrCl6.xH2O + NaOH) precursor mixture follows hydrolysis method Fig. 3.2. TGA and DTA diagrams of (H2IrCl6.nH2O + NaNO3) precursor mixture follows Adams method 8 Fig. 3.3. XRD patterns of IrO2 fabricated hydrolysis method Fig. 3.4. XRD patterns of IrO2 fabricated Adams method Figure 3.3 and 3.4 are X-ray diffraction pattern of IrO2 powder samples synthesized by hydrolysis and Adams methods at different furnacing temperatures. At the furnace temperature values lower than 500 o C, the peaks on the X-ray diffraction pattern of both synthesized methods have an unclear peak signal and narrower. This is because the IrO2 formed at these furnacing temperatures has a very fine structure or anatas structure. When the furnacing temperature is increased to 500°C and 600°C, the size of the peaks is smaller and the more peak signals that represent the crystal structure. The peaks on the X-ray diffraction pattern have signal peaks at 2θ angle values: 28(110); 35.1(101); 54.3(211) are similar and all peaks match a rutile-structure as indexed. Thus, at the furnacing temperature of 500 o C or more, the IrO2 catalytic material changes from amorphous structure to rutil crystal structure. Fig. 3.5. SEM pictures of IrO2 electrocatalytic powers by hydrolysis method, degree of magnification 80.000 times 9 In order to determine the morphology of the powders, FE-SEM analysis was carried out on the catalyst oxide powders that synthesized by hydrolysis (Fig. 3.5). All powder samples are small and uniform in size and shows the presence of agglomerates composed of fine particles. While increasing the furnace temperature, particle size of IrO2 catalyst was increased. This evidence was in agreement with the X-ray diffraction pattern. Hình 3.6. TEM pictures of IrO2 electrocatalytic powers by Adams method, degree of magnification 80.000 times TEM images of IrO2 samples synthesized by Adams method at different furnace temperatures in Figure 3.6 show that, compared to the hydrolysis method, Adams method gives more uniform distribution of catalyst powders and much smaller size (only a few nanometers). At 400 o C, the catalytic particles were very small and there was no clear grain boundary. This may be because the catalytic powder obtained at this temperature has an amorphous structure. When the temperature increased to 500 o C, the particles were formed clearly and relatively evenly. At 600 o C, catalyst flour appears with more rod-shaped particles, which may be due to the faster growth rate of germ growth, making the crystal size increase. 10 (a) Hydrolysis (b) Adams Fig. 3.7. Cyclic voltammograms of IrO2 powders in solution of 0.5 M H2SO4; scanning potential velocity 50 mV.s -1 Fig. 3.7 is the CV graph of the IrO2 catalytic samples synthesized at different temperatures by two methods of hydrolysis and Adams. In general, CV curves have a similar standard shape of IrO2. However, the oxidation-reduction peaks are not clear except the CV curve of IrO2 furnaced 400 o C according to Adams method. In both methods, the CV of the furnaced at 400 o C has the largest area compared to the furnaced samples at higher temperatures in the same synthesis method, this is due to irO2 heating at 400 o C with an anatas structure, particle size is the smallest and smoothest so their electrochemical activity surface is the largest so the catalytic activity is the largest compared to other samples of the same method. Because the particle size is much smaller, the IrO2 sample furnaced at 400 o C by Adam method has a larger catalytic activity than the synthetic sample by hydrolysis method. The higher the temperature, the lower the CV area means, the lower the activation rate, however the samples synthesized by the Adams method still have greater activation than the sample of hydrolysis method at the same temperature. From the anode polarization curves of the catalytic powders, the electrochemical parameters were calculated and summarized in Table 11 3.1. Based on the data in Table 3.1, it is found that IrO2 oxide is furnased at 500 o C by the Adams method has been catalyzed to OER in the highest due to the lowest Eoer. Table 3.1. Electrochemical parameters for IrO2 powders furnaced different temperatures Mẫu io (µA/cm2) Eoer (mV) Charge total (C.cm -2 .mg -1 ) 400 o C-Hydrolysis 51 1232 26.25 500 o C-Hydrolysis 50 1253 22.25 600 o C-Hydrolysis 33 1262 12.00 400 o C- Adams 16 1190 30.50 500 o C- Adams 12 1220 23.00 600 o C- Adams 22 1222 13.75 Thus, the catalyst samples synthesized by Adams method have smaller particle size so they have better catalytic activity, are more stable and catalyze the oxygen evolution reaction better than the synthetic catalysts by hydrolysis method. Furnacing temperature also greatly affects the structure, activity and durability of the catalyst, at a furnace temperature of 400 o C, the catalysts have a smaller size, better catalytic activity but less stable more than c furnaced samples at temperatures of 500 o C and 600 o C. Optimal catalytic selection is based on the best combination of specific characteristics of catalysts such as activity, durability, good catalytic ability for OER as well as production costs, and furnace temperature of 500°C is the right temperature to furnace catalyst powder by Adams method. In order to further optimize the activity surface of the catalytic powder, melting temperature of NaNO3 at 308 o C, so it should be maintained at this melting temperature for a while so that the melting NaNO3 will completely react with H2IrCl6 to form IrO2 with high performance. Applying this improvement, the IrO2 catalyst was 12 synthesized by the Adams method of improvement according to the two-step heating process, specifically as follows: furnace at 325 o C for 30 minutes with a heating ramp of 5 °C min-1 then raising the furnace level up to 500 o C with a heating ramp of 5°C.min-1 and keep at this temperature for 1 hour. X-ray diffraction pattern of two IrO2 samples furnaced in one step and two step modes have the same intensity and peak peaks are similar and all peaks match a IrO2 rutile-structure as indexed. This proves that the samples that burn both of these thermal modes form IrO2 rutile structure. However, the furnaced sample in two thermal steps has a wider peak footprint meaning that the catalytic particles size obtained are smaller. TEM images of these samples also show that the two stage samples has smaller (only a few nanometers in size), more uniform particles, and large particles almost do not appear. Table 3.2. Electrochemical parameters for IrO2 powders furnaced two stage Parameter io (µA/cm2) Eoer (mV) Charge total (C.cm -2 .mg -1 ) IrO2 500 o C 16 1220 23.0 IrO2 325 oC‒500oC 22 1220 29.5 CV line shapes of two furnace modes are similar. The two-step furnaced catalytic powder sample has a more charge total than the one- step furnaced catalyst implies that the sample has a better activation capacity and the reactions take place on the electrode surface easily and more advantageous. This proves that the two stage furnacing regime improves the activation of IrO2 catalytic powder. From here, the catalysts will be fabricated according to the two step calcination process, the manufacturing process is shown in Figure 3.14. 13 Fig. 3.14. Process of fabrication of IrO2 powder catalytic material 3.2. Fabrication of IrxRu(1-x)O2 electrocatalyst From the TGA and DTG diagrams when furnaced salt mixtures (H2IrCl6.nH2O + NaNO3) and (RuCl3.mH2O + NaNO3) in the air, it is possible to determine the temperature range of 400-600 o C is the appropriate temperature range for furnaced of a mixture of salt into IrxRu(1-x)O2, combined with experimental part 3.1 (proved that the temperature of 500 o C is the optimal temperature to synthesize IrO2) we give the appropriate temperature to synthesize IrO2, Ir0.9Ru0.1O2, 14 Ir0.8Ru0.2O2, Ir0.7Ru0.3O2, Ir0.6Ru0.4O2, Ir0.5Ru0.5O2 and RuO2 are 500°C following to the improved Adams method of using are two salts H2IrCl6.nH2O and RuCl3.mH2O precursors. Fig. 3.17. XRD patterns of IrxRu(1-x)O2 Fig. 3.17 shows XRD patterns of IrxRu1−xO2 powders. As can be observed that the diffraction patterns of all samples have well-defined peaks, narrow pics’ width and rutile structures. Diffusion spectra of pure ruthenium oxide and pure iridium oxide are distributed in nearly identical 2θ angles due to their similar structures. The diffraction peaks of the oxide mixture have both all the diffraction angles of the two pure oxides, however these peaks slowly toward the peaks of the RuO2 oxide as the gradual increase of the ruthenium concentration. Which may indicate that when rutheni was added, a lattice was modificated, the solid phase was formatted and crystal size increases. The average crystallite sizes for IrxRu1−xO2 powders were estimated using Scherrer equation, the results are 2.4 nm and 3.1 nm for IrO2 and RuO2, respectively. As the crystallization temperature of RuO2 is lower than that of IrO2, the crystallite size increases when increasing the RuO2 amount in the mixed materials. The Ir0.7Ru0.3O2 catalyst was the highest 15 crystallinity showed that the solid phase was best formed at this molar ratio. The result of the EDS analysis of IrxRu1−xO2 powder samples synthesized at x values different. The results show that for all samples have presence of the elements Ir, Ru and O2 and the composition measured is very close to the composition target proving that the Adams’ method is suitable for synthesis of catalyst mixture with the desired composition. TEM measurements were performed to analyze the morphology and the particles size of the synthesized materials. The morphology for all the samples contain aggregated particles. The particles present a cuboid shape and have sizes ranging from 2 to 50 nm. IrO2 had the smallest particle size (about 2-5 nm) but it was inserted lager cuboid shape (about 10-40 nm). While the RuO2 particles were more uniform and larger (10-15 nm). As RuO2 was gradually added, the surface morphology of the particles gradually changed, the size of the particle increased slightly (but still smaller than the pure RuO2 particles) and the quanlity of the lager cuboid shape particles was decreased, in Ir0.5Ru0.5O2 had a few large cuboid particles. Proving that when added, Ru interacts with Ir to form a solid solution that flattens lager cuboid shape particles, making the morphology of mixtrure oxides more uniform. The paramaters from the CV plot of IrxRu1−xO2 catalyst mixtures were listed in Table 3.4. RuO2 oxide has the most activity, IrO2 oxide has the lowest activity, the oxide mixtures have similar shape and medium activity and the activation area increases as the ruthenium is added to the mixture. 16 The activity degradation of the catalyst mixtures were shown by the reduction in the area of CV curves after 1000 th cycles in 0.5 M H2SO4 at a scanning speed of 50 mV s -1, the results also shown in Figure 6 and Table 2. Results showed that although IrO2 had lower reactivity, it was more resistant to RuO2, which is consistent with previous studies. IrO2 was reduced to 4.6 % activity after 1000 th cycles while RuO2 was reduced to 19.8 %. As RuO2 was gradually added, activity degradations increased but these increases was insignificant compared to the large reduction of pure RuO2. EOER ,io parameters from polarisation (table 3.4). Similar to previous study, RuO2 is the most active, its corresponding starting potential of oxygen evolution is the lowest (1100 mV) and the catalytic properties are optimized and IrO2 is the lowest active. For mixture oxides, the starting potential of oxygen evolution decreases with the increase of RuO2. There result are similar to the result of the CV measurements. This imply that the presence of the ruthenium in the electrocatalyst and it can be insert the latice to the formation of a common valence band with iridium oxide, thus promotes the oxygen evolution reaction. However, RuO2 is also known to be less stable than IrO2, in the acid media RuO2 can corrode to RuO4 so the corrosion of the mixture oxides increase with increase of ruthenium content. Thus, a solid solution with a moderate Ru content on the outercatalyst surface should represent a good compromise between activity and stability. In this study, the optimun Ru content is 30 % mol, because at this mol ratio Ir0.7Ru0.3O2 has the best combination of activity and stability. 17 Table 3.4. Electrochemical parameters of IrxRu1−xO2 Catalyst io (µA/cm2) EOER (mV) Charge total (C.cm -2 .mg -1 ) Activity degradatio(%) IrO2 22 1220 29.5 4.6 Ir0.9Ru0.1O2 11 1200 32.8 5.1 Ir0.8Ru0.2O2 14 1180 33.7 6.1 Ir0.7Ru0.3O2 16 1140 38.2 7.8 Ir0.6Ru0.4O2 20 1160 34.6 9.5 Ir0.5Ru0.5O2 24 1150 36.9 10.2 RuO2 42 1100 38.9 19.8 Fig. 3.27. Polarisatiom curves of IrxRu1−xO2 electrocatalytic power in 0.5 molL -1 H2SO4 electrolyte, 1 mV.s -1 Thus, it can be seen that the modified Adams method has produced IrxRu1−xO2 catalyst mixture with relatively uniform size, small size of nanometer size and rutile structure, with good activity and durability for oxygen evolution reaction at anode. The addition of ruthenium formed a solid solution between Ir and Ru which significantly improved the surface morphology as well as the size of the catalytic particles thereby improving the catalytic activity without significantly reducing the durability for OER. Ir0.7Ru0.3O2 catalyst mixture has the best crystallization because at this molar ratio the solid solution formation between Ru and Ir is the highest. Ir0.7Ru0.3O2 particles have fairly uniform morphology for highest activity and average 18 durability among catalytic mixtures. It can be said that the catalyst mixture with the molar ratio (Ir: Ru) = (7: 3) shows the best combination of activation and stability, so we chose the mixture Ir0.7Ru0.3O2 for further research. 3.3. Fabrication of IrRuMO2 ternary catalyst ( M = Ti, Sn, Co) Fig. 3.28. XRD patterns of IrRuTiO2; IrRuSnO2; IrRuCoO2. Purpose of this study insert third component to Ir0.7Ru0.3O2 mixture catalyst to reduce the precious metal to fabricate IrRuMO2 which still keep its activity and durability for OER. Substances insertion with an initial molar ratio of Ir: Ru: M = 1: 1: 1 (M = Ti; Sn; Co) Figure 3.29 is X-ray diffraction of IrRuTiO2; IrRuSnO2; IrRuCoO2 and Ir0.7Ru0.3O2 catalytic powders. It can be seen that the diffraction spectrum of all 3 mixes has the main peaks of Ir0.7Ru0.3O2. The peak signal of IrRuTiO2 is most clearly, then IrRuCoO2 and IrRuSnO2. This proves that the catalyst with M = Ti has the highest crystal degree. The reason may be that Ti has the same crystal size as Ir and Ru, so it is easy to mix with Ir and Ru at crystal level to form a solid solution. The width of the peaks shows that the crystal of IrRuSnO2 is the smallest, then IrRuCoO2 and IrRuTiO2, that are also consistent with the TEM image in Figure 3.29. 19 Fig. 3.29. TEM micrographs of IrRuMO2 (M = Ti, Sn, Co), electrocatalytic, magnification 80.000 times The values of electrochemical determined from the graphs are listed in Table 3.5. The activity of catalyst also changed with increasing activity towards Ti <Sn <Co. This is because the electrical conductivity of these substances also increases (resistivity of ρTiO2 = 108 Ωcm; ρSnO2 = 107 Ωcm; ρCo3O4 = 104 Ωcm). However, The activity degradation after 1000 scanning cycles of IrRuCoO2 is highest because Co3O4 is a strong oxidation, so it is not sustainable in acidic environment which is facilitated by PEMWE. Thus, considering the catalytic durability, IrO2 is still the ideal catalyst for OER. The mixture of Ir0.7Ru0.3O2 is the best combination between the durability of IrO2 and the activation of RuO2. When replacing Ir and Ru expensive than cheaper element, Ti is a good choice. To ensure performance and durability when applying an anode catalyst to a single PEMWE, we decided to select the catalytic mixture of Ir0.7Ru0.3O2 for further tests. Table 3.5. Electrochemical parameters of IrRuMO2 Catalyst io (µA/cm2) Eoer (mV) Charge total (C.cm -2 .mg -1 ) Activity degradation(%) Ir0.7Ru0.3O2 16 1140 38.2 7.8 IrRuSnO2 38 1220 33.7 19.4 IrRuTiO2 28 1220 26.7 16.7 IrRuCoO2 49 1138 42.6 18.84 20 3.4. Fabrication and cheracterization of single PEMWE The components of the single PEMWE are designed to include: separator plate with water channels, current collector, shell and gasket, then they are assembled. Figure 3.39 is the image of a single PEMWE and figure 3.40 is a single PEMWE test system. Fig. 3.39. Single PEMWE Fig. 3.40. Single PEMWE system The energy conversion efficiency of PEMWE is determined by the properties of MEA. The properties of MEA will be affected by parameters such as catalytic density, ionomer content and especially the method of manufacturing MEA. In the hot pressing method used to make MEA, the pressing process at high temperature plays a very important role. There are three important parameters in the pressing process that will affect the properties of MEA: pressure, temperature and pressing time. In this thesis, the pressing time and temperature are fixed at 180 s and 130 o C, the pressure is changed from 18-22 kg.cm -2 . Figure 3.43 is a V-I graph of a single PEMWE with MEA pressed at different pressures. Observing on V-I, it is found that when the pressure decreases, the polarization curves tend to shift to the left, indicating that the pressure is reduced, the voltage of PEMWE increases. During the average current density stage, the slope of line V-I of MEA fabricated at 20 and 22 kg.cm -2 is lower than that of MEA manufactured at other pressure values, meaning that the internal 21 resistance of MEA is made at pressures of 18 and 24 kg.cm -2 will be higher than the internal resistance of MEA fabricated at pressures of 20 and 22 kg.cm -2 . This may be due to the loose cohesion between the diaphragm and the catalyst layer at the pressure of 18 kg.cm -2 , while at the pressure of 24 kg.cm -2 , this cohesion is much more compact