Regular Article
Effect of Shape and Composition of Steel Particles on Simultaneous H2 Production and CO2 Fixation
2024 Volume 64 Issue 10 Pages 1607-1614
Details
2024 Volume 64 Issue 10 Pages 1607-1614
To develop a H2 production and CO2 fixation process using scrap iron, the characteristics of iron and steel particles that react efficiently were investigated. The reaction of commercial pure iron and alloyed steel powders were compared, and their reactivity was evaluated based on the specific surface area, apparent density, and crystal lattice strain. The efficient reactivity in porous iron powders was attributed to crevice corrosion. To investigate the effect of alloy composition, we added Ni to pure iron powder by pretreatment, which resulted in enhanced H2 production and CO2 fixation. The results indicated that galvanic corrosion contributes to Fe oxidation, because Fe is less noble than Ni based on their electrode potentials. This study provides guidelines for improving the efficiency of reactions that produce H2 while fixing CO2 using steel scrap.
The problems associated with fossil fuel depletion, global warming, and environmental pollution are intensifying as the global population and energy consumption increase. Thus, many countries are aiming to achieve carbon neutrality by 2050. Therefore, in addition to the reduction of carbon dioxide (CO2) emissions, hydrogen (H2) is attracting attention as a clean and sustainable non-carbon energy source.1) The demand for H2 as a feedstock for fuel-cell vehicles and stationary fuel-cell systems is expected to increase.2) In addition, H2 has industrial applications in manufacturing. Hydrogen ironmaking, where iron ore is reduced directly using H23,4) or the hydrogen carrier NH3,5) is being developed. Therefore, the industrial and commercial applications of H2 are expected to expand significantly in the future.
Currently, steam reforming is the main method for producing H2 commercially.6) This reaction also generates carbon monoxide (CO), which is converted to CO2 by the water–gas shift reaction, while producing more H2. However, this reaction is a strongly endothermic process that consumes large amounts of energy (usually from fossil fuels) and steam to maintain the reaction temperature. Moreover, this process emits large amounts of CO2 and nitrogen oxides (NOx), which means that it cannot be considered a clean energy-production method. The H2 production technologies developed to date tend to rely on the use of fossil fuels. Thus, developing an alternate H2 production method that does not require fossil fuels and can reduce/omit CO2 emissions will significantly contribute to achieving carbon neutrality.
Hydrogen obtained by reforming fossil fuels is called gray hydrogen,7) while that produced from renewable raw materials and energy sources is called green hydrogen.8,9) H2 produced by water electrolysis is carbon-free and truly “green.” In particular, hydrogen production by solar water splitting is considered a promising method because it can directly utilize inexhaustible solar energy.10) Water splitting based on photocatalytic, photoelectrochemical, and photovoltaic–photoelectrochemical hybrid systems has been demonstrated. In particular, hydrogen production by photocatalytic water splitting11) has been extensively researched and has made remarkable progress. Moreover, various types of photocatalysts and system designs have been studied. However, the industrial production of hydrogen by solar water splitting is still limited by low hydrogen-production and solar-conversion efficiencies and the insufficient stability of photocatalysts. Therefore, no existing technologies are immediately available for practical application at the industrial scale.
Water splitting methods based on electrolysis are considered promising for hydrogen production,12) but require the development of a suitable electrode catalyst to ensure efficient hydrogen production. Noble-metal-based electrode catalysts are expensive because of the scarcity of such metals. Therefore, the development of low-cost high-efficiency catalysts that do not contain precious metals is a research focus. However, because few conventional materials show activity and stability comparable to those of platinum-group metals, significant modification and functionalization of the catalyst materials is required before they can be applied to industrial hydrogen production.
Inspired by this background, we are aiming to fabricate a water-splitting system for hydrogen production that can be immediately applied on an industrial scale. In this process, the required energy is supplied by a chemical reaction between a metal and water, and we have proposed using iron in particular.13,14) Iron and steel are basic commodities in society for the construction of buildings and other structures, vehicles, and machines. Because iron resources on Earth are abundant, iron-based products are inexpensively mass-produced and are widely available. The total amount of iron and steel used in various products or stored in a specific form is the steel accumulation, some of which is recovered as scrap iron at the end of its useful life. Because iron is infinitely recyclable, most scrap iron is used as a raw material for new steel products. There is a certain correlation between economic growth and the amount of steel accumulated per capita.15) Large amounts of steel are accumulated in developed countries, such as Japan, the United States of America, Canada, and the United Kingdom, which makes it an essential resource.16) Steel accumulation is expected to continue increasing worldwide, especially in China, India, and other countries that are experiencing rapid development. Moreover, the amount of old scrap steel containing impurities such as copper and tin, which act as tramp elements and cause steel embrittlement, will likely increase.17,18)
Reversible hydrogen-storage methods based on the oxidation of metals by water and subsequent reduction of the formed metal oxides by hydrogen have been studied for some time, with iron as the most promising candidate metal.19,20) This study focuses on H2 gas production using the reaction between Fe obtained from scrap resources, waste CO2, and water, as described by Eq. (1). CO2 accelerates the reaction and is simultaneously immobilized as a carbonate. The production of carbonates is a promising method of carbon dioxide reduction as it is an immobilisation reaction that does not consume energy.21) The reaction in this study has the advantage of proceeding at room temperature and pressure, and hydrogen can be produced simply by adding iron to carbonated water. Because H2 gas is difficult to transport from a hydrogen-production plant, it needs to be converted into a carrier material, such as methylcyclohexane. In contrast, iron is easy to transport and has a high energy density.
| (1) |
This reaction is known as the CO2 corrosion of steel in the oil and gas industry, where the corrosion rate of pipeline steel used to transport natural gas increases in the presence of CO2 and absence of oxygen (O2) because the dissolved CO2 promotes cathodic H2 formation. It has been reported that water containing CO2 significantly increases the corrosion rate compared with strongly acidic solutions.22) CO2 dissolved in water initially forms carbonic acid (H2CO3), which was shown to be directly reduced in a cathodic process, while the H atoms produced during reduction combine to form H2.23) In contrast, it has been reported that H2 production from bicarbonate (HCO3−) discharge exceeds that of other cathodic reactions and occurs at potentials that are more negative than the corrosion potential of pipeline steel.24) The formation of ferrous carbonate (FeCO3) and carbonate complexes ([Fe(CO3)2]2−) has been confirmed in aqueous solutions containing dissolved Fe and CO2.25)
Recently, it has been proposed to react waste iron powder with a CO2 solution to produce FeCO3 for use as a concrete additive, and studies on the reaction conditions and concrete strength have been conducted.26,27) Similar to our concept, Constantinou et al. are researching simultaneous CO2 sequestration and H2 generation using scrap iron and are evaluating the conditions required to accelerate the reaction.28) Because steel is accumulated in large quantities, it is considered a highly available raw material instead of pure iron for large-scale H2 production and CO2 fixation. In our previous study,13) the reaction of grinding sludge generated by grinding high-carbon chromium-bearing steel took twice as long as that of fine pure iron powder (3–5 μm particle size) but exhibited comparable H2 production and CO2 consumption. The lower reaction rate compared to pure iron was attributed to the large particle size (i.e., small specific surface area) of the swarf particles and the introduction of impurities during the grinding process. Despite these limitations, the reaction exhibited high activity, as the decrease in reaction rate was relatively minor. This may be due to the material composition because carbon steel is highly susceptible to oxidation.
We previously conducted experiments using stainless steel as the raw material.13) Stainless steel contains several elements, including Cr, Ni, and C, as a solid solution with Fe. Moreover, stainless steel is defined as an alloy steel with a Cr content of ≥10.5% and C content of ≤1.2%. The corrosion and oxidation resistance of stainless steel can be improved by dissolving various additive elements into the material. Austenitic stainless steels (18% Cr, 8% Ni) tends to have better corrosion resistance compared with ferritic stainless steels (17% Cr).29) However, it is noteworthy that H2 production and CO2 absorption were observed when using austenitic stainless steel, even though it has a significantly lower reaction rate than carbon steel. This indicates that the reactivity significantly depends on the elemental composition of the steel. Scrap iron contains several impurity elements, and thus their effects on the H2 production reaction should be investigated. The present study compares the H2 production reaction using pure iron and alloy steel powders, which are manufactured for industrial use, and evaluates the effects of the particle composition and shape on the reaction.
Atomized and reduced iron powders, employed in this study, are used as raw materials in powder metallurgy.30,31) Atomized iron powder is produced by reducing iron ore in a blast furnace to produce pig iron, blowing high-pressure oxygen in a converter furnace to produce molten steel slurry, and atomizing the molten steel slurry using high-pressure water or gas. Reduced iron powder is produced by reducing iron ore or mill scale with coke, followed by heat treatment in an H2 atmosphere. Mill scale is an oxide film formed by the reaction with the O2 in air at high temperature during ironmaking. The shape, apparent density, and porosity of water-atomized and reduced iron powders are different, which is expected to affect their reactivity.
In steel utilization, crevice corrosion32) and galvanic corrosion33) are some of the causes of steel degradation and require countermeasures. The present study focused on crevice corrosion and galvanic corrosion, and their effects on the redox reaction between Fe and carbonated water. Crevice corrosion occurs when a micron-sized gap exists between metals or between a metal and another material, forming a concentration cell in water due to the dissolved oxygen gradient across the gap.34) In environments with CO2 instead of air, the concentration gradient of CO2 in the water is believed to promote crevice corrosion more than the oxygen concentration gradient. Furthermore, galvanic corrosion occurs when dissimilar metals are in direct contact and form a galvanic cell between them.35) In this study, the effect of coexisting metals, including Cr, Ni, and Mo, was investigated. In particular, Ni and Mo have more noble potentials than Fe, and are hypothesized to promote Fe corrosion and increase its reactivity. Conversely, Cr has a lower potential than Fe and is hypothesized to inhibit Fe corrosion and decrease its reactivity. In this study, Ni was intentionally added as a dopant and the effects of galvanic corrosion on Ni-doped iron powder and alloy steel powder samples were evaluated. As a result, we were able to increase the reaction rate compared to our previous report,13) thereby obtaining insights for increasing the reaction efficiency to promote the practical use of this system.
The reaction procedure used in this study was similar to that employed in our previously reported experiments.13,14) Pure iron powders (KB-90, K-100T, and 303A-60) and alloy steel powders (SIGMALOY2010, 4100V, 30CRV, and 20CRV) manufactured by JFE Steel Corporation (Japan)36) were used as raw materials. KB-90 and K-100T powders are produced by reducing iron ore and mill scale, respectively, while 303A-60 is a water-atomized powder. SIGMALOY 2010 is a partially alloyed steel powder produced by adding an alloy powder to the surface of the water-atomized iron powder. 4100V, 30CRV, and 20CRV are alloyed steel powders produced by adding alloying elements, such as Cr and Mo, to molten steel. Table 1 lists the compositions of the seven JFE steel samples. These powders were sieved to classify the particles into sizes of <45, 45–106, and 106–180 μm.
| Type | Sample name | Chemical composition |
|---|---|---|
| Pure iron powder | KB-90 | 87.0% ≤ M.Fe |
| K-100T | 90.0% ≤ M.Fe | |
| 303A-60 | 99.0% ≤ T.Fe | |
| Alloyed steel powder | SIGMALOY 2010 | Fe– 2%Ni–1%Mo |
| 4100V | Fe–1%Cr–0.8%Mn–0.3%Mo | |
| 30CRV | Fe–3%Cr–0.3%Mo–0.3%V | |
| 20CRV | Fe–2%Cr–0.8%Mn–0.2%Mo–0.2%S |
M.Fe: metallic Fe, T.Fe: total Fe
The powders were characterized by field-emission scanning electron microscopy (FE-SEM; Hitachi SU-8230), powder X-ray diffraction (XRD; Rigaku, MiniFlex 600, Cu Kα), Kr gas adsorption, and mercury intrusion porosimetry (Micromeritics, Autopore IV9520) to quantify the pore volume before the reaction experiments. To quantitatively analyze the crystal phases and evaluate the crystal lattice strain, the Whole Powder Pattern Fitting (WPPF) method was used with integrated powder XRD software (Rigaku, PDXL Ver. 2.8.4), which performs profile fitting of the entire pattern based on crystal structure information.
A powder sample (2.0 g) was sealed in a 295 mL polypropylene bottle with carbonated water (5 mL), which was prepared using purified water and equilibrated with 101.3 kPa CO2 (99.5%) from a high-pressure gas cylinder. Subsequently, the bottle was filled with 101.3 kPa CO2 and sealed under anaerobic conditions. The carbonated water and iron powder were mixed by shaking the bottle at 323 K in a constant-temperature shaker to maintain the solid material in suspension and allow the mass-transport rate to plateau. The plateau in the mass-transport rate was confirmed by measuring the relationship between shaking rate and reaction rate beforehand. According to the reaction shown in Eq. (1), the pressure inside the reaction vessel should remain constant because the amount of consumed gaseous CO2 is equivalent to that of generated gaseous H2. During the reaction, gas was sampled (0.1 mL) at regular intervals, and the concentrations of CO2 and H2 were analyzed by gas chromatography (GC; Shimadzu GC-8A, Porapak N 2 m and Porapak Q 1 m columns) using Ar as the gas carrier and a thermal conductivity detector.
2.2. Effect of Ni Addition to Pure Iron Powder on the ReactionUsing chemical-grade pure iron powder (>99%, particle size: 3–5 μm, High Purity Chemicals, Japan) as the raw material, the effect of Ni addition was evaluated. First, 0.9 g of Ni(NO3)2·6H2O (>99.9%, Fujifilm Wako Chemicals, Japan) was dissolved in 15 mL of acetone (99%, Fujifilm Wako Chemicals, Japan) in a polypropylene bottle, to which 3.0 g of the pure iron powder was added. The mixture was shaken at 2000 rpm for 4 d at a temperature of 298 K using a constant-temperature shaker. Subsequently, the mixture was placed in a vacuum dryer and dried at 353 K for 5 h. Excess Ni(NO3)2 on the iron powder was removed by washing with 15 mL of acetone, and the resultant solid–liquid mixture was separated by centrifugation at 5000 rpm for 15 min. The recovered solid phase was named “Ni-treated sample”. X-ray photoelectron spectroscopy (XPS; Surface Science Instruments, SSX-100, Al Kα, 6.5 kV) was performed to confirm that Ni doping was successful.
In addition, iron powder without Ni(NO3)2·6H2O was shaken in acetone under the same conditions, collected, and named “Acetone-treated sample”. “Ni-mixed sample” was prepared by adding 0.1 g of Ni(NO3)2·6H2O to 2.0 g of pure iron powder, followed by dry mixing with a mortar pestle for 15 min. These samples were characterized by XRD and then used as reactants for the reaction, which was performed using the same procedure as mentioned above to enable comparison with the pure iron powder.
For all iron powder samples, H2 formation was observed along with a decrease in the CO2 concentration as the reaction progressed over time. As a typical example, the changes in the gas concentration over time in the experiment using pure iron powder (KB-90, 45–106 μm) is shown in Fig. 1(a). Additionally, the formation of FeCO3 in the solid phase was observed by XRD, confirming that the reaction proceeded according to Eq. (1). Since the rate of CO2 decrease over time and the rate of H2 production were approximately equal to that predicted by Eq. (1), hereafter we discuss the progress of the reaction based on the amount of produced H2. Figure 1(b) shows the relationship between the particle size and H2 production rate of the commercial iron and steel powders. The reported particle size is the median diameter (D50) of the particle size distribution we measured using a laser diffraction and scattering system. For each sample, the amount of H2 produced was calculated from the H2 concentration measured by GC. During the early stage of the reaction, the linear portion of the time-dependent H2 concentration curve was fit using the least-squares method to determine the reaction rate (pseudo-zero order reaction). When the reaction rates were compared with those previously reported by us for pure iron powder,13) the reaction rates were more than one order of magnitude slower in all samples. The reason is obvious: the iron powders used in the previous experiments were smaller in particle size (3–5 μm) and had a much larger specific surface area than the present iron powders.
With the exception of KB-90, the reaction rate increases with decreasing particle size. This is especially evident for SIGMALOY2010. Thus, it is concluded that a smaller particle size (larger specific surface area) enhances the reaction. However, for the other samples, the particle size dependence of the reaction rate was small, probably because the irregular particles exhibit a poor correlation between the measured particle size and specific surface area. Furthermore, parameters other than specific surface area (such as pore volume) have a significant effect on reactivity, and these parameters may be almost independent of the particle size.
The H2 production rates of the KB-90 and K-100T reduced iron powders were higher than those of the 303A-60 atomized iron powder. This is because the reduced powders are more porous, as shown in the SEM images in Fig. 2, resulting in larger specific surface areas (Fig. 3(a)) on which the reaction can occur. The high porosity was confirmed by the apparent densities (Fig. 3(b)), which were smaller for KB-90 and K-100T than for 303A-60. The apparent density of KB-90 was particularly small, which is mainly due to its low purity. This is attributed to KB-90 containing a low metallic iron content (Table 1) and a large amount of iron oxide (as discussed later) and impurities containing light elements. A high porosity indicates that numerous small voids are present on the surface (Fig. 2). Figure 3(c) shows that K-100T, and especially KB-90, had larger pore volumes than 303A-60, and consequently exhibited a significantly higher H2 production rate. Typically, the oxidation of metals in the presence of water in air is accelerated by crevices (fractures).3) This is because it is difficult for the oxygen in air to penetrate the crevice, resulting in an oxygen concentration gradient between the surface and bottom of the crevice that forms a local oxygen concentration cell. In this study, based on the CO2 gradient, it is proposed that the reaction is promoted by a mechanism similar to that of crevice corrosion.
SIGMALOY2010 is a partially alloyed steel powder, which is prepared by metallurgically coating pure iron powder particles with the alloying elements (Ni and Mo). SIGMALOY2010 exhibited the strongest dependence of the H2 production rate on particle size. Assuming that the shape of the iron powder is spherical with a smooth surface and the H2 production rate is proportional to the surface area, then the rate will be inversely proportional to the particle size. This relationship was calculated for SIGMALOY2010 based on the reaction rate of 150-μm particles and is plotted in Fig. 1 (orange dashed line). The measured reaction rates are lower than the predicted values at smaller particle sizes.
Figure 4 shows the XRD patterns of SIGMALOY2010, and Table 2 lists the analysis results obtained using the WPPF method. In the table, Rwp and S are confidence factors representing the degree of agreement between the experimental and calculated data determined by the Rietveld method. By least-squares fitting, the parameters of the crystal structure are optimized to minimize the residual sum of squares, and the smaller Rwp and the closer S is to 1, the better the fitting can be judged to have been performed. The diffraction angles and intensities of each crystalline phase were refined, yielding a scaling factor that was used to determine the weight fraction of each phase in the mixture. Additionally, based on the refinement results, the lattice strains were calculated using a Halder–Wagner plot. The XRD results show that the diffraction line widths of α-Fe were narrower for smaller particle sizes (Fig. 4). This indicates that the lattice strain is smaller for smaller particles (Table 2), which demonstrates that the crystal quality is high and physically stable. Thus, the SIGMALOY2010 samples with a small particle size have lower reactivity than would be expected from the particle size.
| Particle size (μm) | Chemical composition (%) | Crystal lattice strain (%) | Rwp (%) | S | ||
|---|---|---|---|---|---|---|
| α-Fe | Ni | α-Fe | Ni | |||
| −45 | 88.3(3) | 11.7(3) | 0.091(5) | 0.535(9) | 30.7 | 0.978 |
| 45–106 | 97.2(3) | 2.8(3) | 0.27(2) | 0.44(5) | 32.2 | 0.988 |
| 106–180 | 95.9(3) | 4.1(3) | 0.37(2) | 0.1(2) | 27.9 | 0.944 |
The percentage of α-Fe in SIGMALOY2010 decreases with decreasing particle size (Fig. 4, Table 2), which is another reason for the deviation from the inversely proportional curve (Fig. 1). Prior to the present experiment, we confirmed that insignificant amounts of H2 are produced when Ni is used as the reaction feedstock instead of Fe. Some of the alloying element particles, such as Ni, that were coated on the surface of the iron particles are thought to have detached from the surface. Because the alloying powder particles are small, they were distributed within the powder sample with a smaller particle size, which may have resulted in an reduced α-Fe fraction.
Figure 5 shows the XRD patterns of KB-90 and Table 3 lists the corresponding WPPF analysis results. KB-90 exhibited a decrease in the H2 production rate with decreasing particle size (Fig. 1), which is attributed to the decrease in the α-Fe content and increase in the amount of oxides. Because iron oxidation proceeds from the particle surface, smaller particles with a higher surface-to-volume ratio have a higher oxide fraction.
| Particle size (μm) | Chemical composition (%) | Crystal lattice strain (%) | Rwp (%) | S | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
| α-Fe | FeO | α-Fe2O3 | Fe3O4 | α-Fe | FeO | α-Fe2O3 | Fe3O4 | |||
| −45 | 65(5) | 24(2) | 2(7) | 9(1) | 0.32(1) | 0.5(1) | 1.21(9) | 1.1(2) | 29.6 | 1.06 |
| 45–106 | 90(2) | 3.6(8) | 1(2) | 5(1) | 0.265(4) | 1.44(3) | 0.95(9) | 3.0(2) | 27.7 | 0.932 |
| 106–180 | 91(2) | 3.2(7) | 1(1) | 4.5(8) | 0.266(4) | 1.56(2) | 1.11(9) | 1.9(3) | 27.7 | 0.931 |
Additionally, the effects of alloying elements were analyzed. The improved H2 production rate of SIGMALOY2010 compared to that of 303A-60, which is also a water-atomized iron powder but without alloying treatment, may be due to the chemical effects of the alloying elements (Ni and Mo). This effect is attributed to galvanic corrosion, in which the less-noble metal corrodes more rapidly when dissimilar metals are in electrical contact. Thus, Ni and Mo, which have larger electrode potentials than Fe, may enhance the ionization of Fe and thereby improve its reactivity. This result is consistent with the conclusion from corrosion resistance studies that Zn and Fe corrosion is enhanced when in contact with Ni powder.33)
The H2 production rates of 4100V, 30CRV, and 20CRV are superior to that of 303A-60, although they are all produced using the same water-atomization method. Because the manufacturing method is identical, the particle shapes are similar, and no significant differences were observed when comparing their XRD patterns (Fig. 6) and corresponding crystal lattice strains (Table 4). Therefore, it is proposed that the physical effects are small. The acceleration of the H2 production rate is therefore attributed to a chemical effect; that is, the effects of the alloying elements. These are completely alloyed steel powders with high Cr and Mn contents, which are less-noble metals than Fe. Thus, the electrode potential of the alloy was decreased by alloying, making it more susceptible to corrosion, thereby increasing the reaction rate.
| Designation | Crystal lattice strain (%) | Rwp (%) | S |
|---|---|---|---|
| α-Fe | |||
| 4100V | 0.29(1) | 31.06 | 0.8892 |
| 30CRV | 0.27(1) | 31.12 | 0.8977 |
| 20CRV | 0.246(7) | 30.82 | 0.8787 |
| 303A-60 | 0.237(9) | 31.20 | 0.8980 |
XPS analysis was performed on Ni-treated sample, which was prepared by immersion in an acetone solution containing nickel nitrate, and a Ni 2p signal was observed. Because N derived from nitrate ions was not clearly detected, it is concluded that the Ni ions reacted with the Fe surface, resulting in Ni coating. Figure 7 shows the changes in the H2 concentration over time for Ni-treated, acetone-treated, Ni-mixed samples, and the pure iron powder. Ni-treated and acetone-treated samples exhibited improved reactivity and a higher final H2 production compared to pure iron powder. Moreover, the reactivity of acetone-treated sample was improved despite the lack of Ni doping, which may be due to the increase in the specific surface area as the particles were mechanically broken down and reduced in size by stirring in acetone. Furthermore, the full-width-at-half-maximum of the XRD peaks of Ni-treated and acetone-treated samples were larger than those of pure Fe (Fig. 8), suggesting that the mechanical stimulation caused a decrease in crystallinity, i.e., a disordered atomic arrangement and decrease in the crystallite size. In addition, as listed in Table 5, the crystal lattice strains of Ni-treated and acetone-treated samples were larger than that of pure iron powder, which is thought to have increased their reactivities. The iron powder reagent contains iron nitride, and consequently, ammonia was produced with H2.37)
| Designation | Chemical composition (%) | Crystal lattice strain (%) | Rwp (%) | S | ||||
|---|---|---|---|---|---|---|---|---|
| α-Fe | Fe4N | Fe3N | α-Fe | Fe4N | Fe3N | |||
| Ni-treated | 83.7(7) | 2.9(4) | 13.4(6) | 0.86(5) | 1.09(5) | 1.0(1) | 23.4 | 0.928 |
| Acetone-treated | 84.1(8) | 2.8(5) | 13.1(7) | 0.94(9) | 1.5(2) | 0.3(2) | 18.5 | 0.943 |
| Pure Iron | 87(2) | 3(2) | 10.3(8) | 0.53(2) | 1.1(2) | 0.6(1) | 17.9 | 0.941 |
The reactivity of Ni-treated sample was more improved than that of acetone-treated sample. Therefore, the effect of Ni doping is clear. As mentioned previously, the reactivity of Fe was improved by contact with relatively more noble metals. Ni-mixed sample was less reactive than pure iron powder, indicating that the simple physical mixing of the pure iron powder with Ni(NO3)2 salt did not improve the reactivity. For Ni-mixed sample the low H2 production is mainly due to the slow increase in the H2 production rate during the initial stages of the reaction, that is, the presence of an induction period, which was also observed for Ni-treated sample. This is because the added Ni2+ oxidized Fe on the surface of the iron particles (instead of Ni2+ being reduced to metallic Ni), and also because the
Although there was a delay in the onset of the reaction of Ni-treated sample due to the induction period, the reaction rate was significantly increased, indicating that the galvanic reaction effectively promotes H2 production. These results are consistent with our previous report13) stating that carbon steel is more reactive than pure iron. When carbon steel has a high C content, i.e., when it contains cementite as a separate phase, the corrosion potential of the phase is more positive than that of pure iron, which increases the corrosion rate of pure iron38) via a galvanic corrosion effect.
The optimal conditions for efficient H2 production and CO2 fixation from iron powder were discussed. It was confirmed that the reaction rate is dependent on the morphology of the iron particles, which varies based on the manufacturing process of the iron powder. The reaction rate was enhanced by increasing the porosity, specific surface area, and crystal lattice strain. In particular, crevice corrosion is a major driving force for the reaction of Fe with CO2 due to the formation of a concentration cell derived from the CO2 concentration gradient between the surface and bottom of the crevice. In addition to these physical effects, chemical effects, such as the presence of impurities, change the reactivity of iron. It was confirmed that electron emission from Fe is promoted by galvanic corrosion when the Fe is in contact with a metal that is electrically more noble than itself. Therefore, it is feasible to adjust the electrical conditions of iron powder or to use scrap iron under appropriate conditions to promote the H2 production and CO2 fixation reactions.
There are no conflicts to declare.
This study was supported by a Grant-in-Aid for Scientific Research (Grant no. JP21K05132) from the Japan Society for the Promotion of Science and by JKA through its promotion funds from KEIRIN RACE, Japan.
