Synthesis of Magnetite Particles by Oxidation of Hydroxyl-chloride Green Rust Suspension under Controlled Conditions

Abstract

Magnetite (Fe₃O₄) particles were synthesized by oxidation of a hydroxyl chloride green rust (GR(Cl^–)) suspension at room temperature. The formation process of Fe₃O₄ particles was characterized by X-ray diffraction, magnetization and electrochemical measurements. The results showed that a small amount of fine Fe₃O₄ particles were nucleated when the supernatant solution of the as-synthesized GR(Cl^–) suspension was replaced by deaerated water. By controlling the injection of oxygen gas at room temperature, Fe₃O₄ particles of about 70 nm in diameter formed from such GR(Cl^–) suspension, while goethite (α-FeOOH) particles were mainly obtained from the as-synthesized GR(Cl^–) suspension under the same oxidation conditions. Hence, the saturation magnetization of final oxidation products obtained from the GR(Cl^–) suspension in which the supernatant solution was replaced was about 60 emu/g, which was six times larger than that obtained from the as-synthesized GR(Cl^–) suspension. In the early stage of the oxidation process, the oxidation-reduction potential (ORP) in the GR(Cl^–) suspension in which supernatant solution was replaced was lower than that in the as-synthesized GR(Cl^–) suspension. In addition, the value of pH of the former suspension was higher than that of the latter suspension. It is concluded that the formation of Fe₃O₄ particles is enhanced in solution with relatively low ORP and high pH.

1. Introduction

Magnetite (Fe₃O₄) particles are useful in various applications such as magnetic and biomedical materials because of their favorable magnetic properties and high biocompatibility.^1,2,3,4) The synthesis of Fe₃O₄ particles has been extensively investigated. For example, it has been reported that Fe₃O₄ particles can be obtained by the coprecipitation of ferrous (Fe(II)) and ferric (Fe(III)) ions in an aqueous solution and also by the annealing of hematite (α-Fe₂O₃) in a reduction atmosphere.^{5,6,7,8,9,10)} These synthesis methods usually involve high processing temperature. However, synthesis of monodispersed Fe₃O₄ particles with excellent magnetic properties at low processing temperatures is technologically desirable. Though spinel ferrite particles with a size of about 15 nm in which Fe₃O₄ is the dominant phase have already been examined at 278 K by using a coprecipitation method,¹¹⁾ their saturation magnetization is very low at room temperature.

It is well known that the corrosion products of metallic Fe in aqueous solution are a mixture of Fe₃O₄ and iron hydroxide particles such as goethite (α-FeOOH), akaganeite (β-FeOOH) and lepidocrocite (γ-FeOOH). Interestingly, such corrosion products are formed at room temperature. Given this background, we focused on green rust (GR) as a precursor of Fe₃O₄. GR is known as one of intermediate products in the corrosion process and often observed in the layer between the corrosion products layer and metallic iron substrate.^12,13) GR is an iron hydroxide consisting of Fe(II) and Fe(III).¹⁴⁾ The structure of GR belongs to the class of layered double hydroxides, which is characterized by the stacking of positive iron hydroxide sheets, such as brucite-like [Fe(II)_1–x-Fe(III)_x(OH)₂]^x+, and a negative interlayer compound of anion, such as CO₃^2–, SO₄^2– and Cl^–, and water molecules.^15,16,17,18) A number of studies were dedicated to elucidate the influence of oxidation condition on the transformation of GR to oxidation products.^19,20,21) It has been demonstrated that GR is transformed to the oxidation products, which are mixture of Fe₃O₄, α-FeOOH, β-FeOOH and γ-FeOOH, at room temperature. Hence, it is expected that Fe₃O₄ particles can be synthesized at low processing temperatures by controlling the oxidation of GR. The transformation of GR sensitively depends on oxidation conditions. For example, it has already been pointed out that GR is transformed partially to Fe₃O₄ under slow aerial oxidation condition at room temperature, the rapid aerial oxidation of GR results in the formation of γ-FeOOH.^20,21) According to the potential-pH diagram including Fe₃O₄ and GR,²²⁾ aqueous solution conditions such as potential and pH are considered to be important in the transformation of GR to Fe₃O₄ particles. Although these results are useful in the research field of corrosion science, an effective process to synthesize Fe₃O₄ particles from GR using the potential-pH diagram has not established yet.

Recently, we successfully synthesized a single phase of a hydroxyl chloride green rust (GR(Cl^–)) in aqueous solution. The oxidation of such GR(Cl^–) suspension containing several kinds of additional ions has been extensively examined.^{23,24,25,26,27)} It has been reported that a small quantity of Fe₃O₄ particles is nucleated in the GR(Cl^–) suspension by the addition of Cu ions.²³⁾ When oxygen gas was injected in such suspension, the formation of Fe₃O₄ particles was enhanced. It has also been reported that Fe₃O₄ particles are formed in a GR(Cl^–) suspension when the liquid phase of the suspension is replaced by deaerated water before the injection of oxygen gas.²⁸⁾ In the present study, the pretreatment using deaerated water was applied to the synthesis of Fe₃O₄ particles from a GR(Cl^–) suspension at room temperature. The influence of pretreatment using deaerated water on the formation of Fe₃O₄ particles was investigated by X-ray diffraction (XRD), magnetization and electrochemical measurements.

2. Experimental

A GR(Cl^–) suspension was synthesized by adding an aqueous sodium hydroxide (NaOH) solution to a solution containing ferric chloride (FeCl₃) and ferrous chloride (FeCl₂) under continuous Ar bubbling. First, an iron chloride solution was prepared with a [Fe(II)]/[Fe(III)] ratio of 4.5. Subsequently, an aqueous NaOH solution was added to the iron chloride solution in the reaction vessel under magnetic agitation at 278 K. The addition of aqueous NaOH was continued until the [OH^–]/{[Fe(II)]+[Fe(III)]} ratio equaled 1.5.

The as-synthesized GR(Cl^–) suspension was separated into solid particles and supernatant solution by centrifugation, and then the solution was removed. Finally, the remaining solid particles were mixed with deaerated water which made by Ar bubbling; this GR(Cl^–) suspension is hereinafter referred to as the pretreated GR(Cl^–) suspension in distinction from the as-synthesized GR(Cl^–) suspension.

For oxidation of the GR (Cl^–) suspension, a reaction vessel consisting of a 500 mL glass beaker and an airtight acrylic lid with openings for inserting pH, oxidation-reduction potential (ORP) and dissolved oxygen (DO) electrodes, gas inlet/outlet ports, a stirrer port and a sampling port were used. Pretreated GR(Cl^–) and as-synthesized GR(Cl^–) suspensions were oxidized by passing nitrogen gas containing 5% oxygen at a flow rate of 200 mL/min in a water bath maintained at 298 K. The suspensions were extracted at regular time intervals under Ar bubbling to prevent air oxidation. Fully oxidized particles obtained from the pretreated GR(Cl^–) and as-synthesized GR(Cl^–) suspensions were prepared by freeze-drying for more than 24 h.

XRD measurements were performed on a Rigaku RINT-2200 diffractometer using Mo Kα radiation (0.071073 nm) to identify the oxide compounds. Specimens of the oxidation process for XRD measurements were carefully prepared as follows. Suspensions extracted during the oxidation process were divided into liquid and solid phases by centrifugation. In a glove box, solid particles were set in a sample holder for XRD measurements and their surfaces were covered with a glycerol to avoid air oxidation. The morphologies of the solid particles were observed using transmission electron microscope (TEM, HITACHI H-7650). The magnetic properties were measured at room temperature using a vibrating sample magnetometer (VSM, TOEI VSM-C7-10) in a maximum applied field of 10 kOe.

In order to monitor the aqueous solution conditions during the formation of iron oxides, the pH and ORP of the aqueous solution were measured. A TOA DKK IM-55G ion meter was used for automatically measuring the pH and ORP of the GR suspensions as a function of the oxidation time. The value of DO in the aqueous solution was also measured using an oxygen meter (TOA DKK DO METER DO-24P).

3. Results and Discussion

3.1. Synthesis of Fe₃O₄ Particles at Room Temperature by Controlling Oxidation of Pretreated GR(Cl^–) Suspension

Figure 1 shows XRD patterns of solid particles extracted from the pretreated GR(Cl^–) and as-synthesized GR(Cl^–) suspensions and their final oxidation products. The reference diffraction patterns for α-FeOOH and Fe₃O₄ provided by the International Center for Diffraction Data (ICDD) database are also shown in Fig. 1. The diffraction pattern of the pretreated GR(Cl^–) is very similar to that of as-synthesized GR(Cl^–) and these are identified as a single phase, at least in the resolution of the equipment. On the other hand, diffraction patterns of their final oxidation products are identified as a mixture of α-FeOOH and Fe₃O₄. Note that the intensity of the 311 peak for Fe₃O₄ was much greater than that of the 110 peak for α-FeOOH in the final oxidation products of the pretreated GR(Cl^–) suspension, though the former was smaller than the latter in the final oxidation products of the as-synthesized GR(Cl^–) suspension. Therefore, the amounts of Fe₃O₄ in the final oxidation products of the pretreated GR(Cl^–) suspension are larger than that in the final oxidation products of the as-synthesized GR(Cl^–) suspension.

Fig. 1.

XRD patterns of (a) the solid particles extracted from the pretreated GR(Cl^–) suspension, (b) its final oxidation products and (c) the solid particles extracted from the as-synthesized GR(Cl^–) suspension, (d) its final oxidation products. Reference diffraction patterns of α-FeOOH and Fe₃O₄ provided by the ICDD database are also given. The arrows indicate the 110 peak for α-FeOOH and the 311 peak for Fe₃O₄.

TEM images of the final oxidation products of the pretreated GR(Cl^–) and the as-synthesized GR(Cl^–) suspensions are shown in Fig. 2. Particles of spherical shape and spindle-like shape are observed in the images. The final oxidation products of the pretreated GR(Cl^–) suspension are mainly composed of particles of spherical shape, while particles of spindle-like shape are mainly observed in that of the as-synthesized GR(Cl^–) suspension. In addition, the mean diameter of particles of spherical shape in the final oxidation products of the pretreated GR(Cl^–) suspension is about 70 nm, which is obviously larger than that in the final oxidation products of the as-synthesized GR(Cl^–) suspension. It has been pointed out that Fe₃O₄ and α-FeOOH are of spherical shape and spindle-like shape, respectively, in many cases.¹⁾ Accordingly, Fe₃O₄ particles of about 70 nm in diameter are successfully synthesized at room temperature by controlling the oxidation of the pretreated GR(Cl^–) suspension.

Fig. 2.

Transmission electron micrographs of final products of (a) the pretreated GR(Cl^–) and (b) as-synthesized GR(Cl^–) suspensions.

The magnetization curves at room temperature of the final oxidation products of the pretreated GR(Cl^–) and the as-synthesized GR(Cl^–) suspensions are presented in Fig. 3. The magnetization is increased and saturated by applying magnetic fields. The value of saturation magnetization for the final oxidation products of the pretreated GR (Cl^–) suspension was about 60 emu/g, while that of the as-synthesized GR(Cl^–) suspension was about 10 emu/g. In the crystal structure of Fe₃O₄, there are two kinds of Fe sites, that is, octahedral and tetrahedral sites. Assuming an inverse spinel structure, eight tetrahedral sites are occupied by four Fe(III) and four Fe(II) ions, and four octahedral sites are occupied by Fe(III) ions. Since the magnetic moment of an Fe ion at a octahedral site is coupled antiferromagnetically with that at a tetrahedral site, the saturation magnetization of Fe₃O₄ is 4 μ_B/Fe-atom in a ferrimagnetic state, which corresponds to 96 emu/g. The influence of α-FeOOH on saturation magnetization is negligible because it is in an antiferromagnetic state at room temperature.⁵⁾ Therefore, it is considered reasonable that the fraction of Fe₃O₄ particles accounts for about 60% of the final oxidation products of the pretreated GR(Cl^–) suspension, which is six times larger than those of the as-synthesized GR(Cl^–) suspension.

Fig. 3.

Magnetization curves at room temperature of the final oxidation products of the pretreated GR(Cl^–) and as-synthesized GR(Cl^–) suspensions.

To enhance the magnetic property, the purification of Fe₃O₄ particles in the final oxidation products of the pretreated GR(Cl^–) suspension has been examined. In this study, the difference of the particle size between Fe₃O₄ and α-FeOOH was focused on. Since the size of Fe₃O₄ particles is seen to be larger than that of α-FeOOH particles in Fig. 2(a), it is expected that only Fe₃O₄ particles remain in the final oxidation products by washing with HCl solution. Figure 4 shows XRD patterns of the final oxidation products of the pretreated GR(Cl^–) suspension after washing with 0.0, 2.5 and 5.0 mol/L HCl solutions. The diffraction peaks of α-FeOOH become smaller with increasing HCl concentration, indicating the purification of Fe₃O₄ particles in the final oxidation products. Figure 5 shows the influence of the HCl solution concentration on the saturation magnetization at room temperature of the final oxidation products after washing. The inset shows the TEM image of the final oxidation products washed with 5.0 mol/L HCl solution. The fraction of α-FeOOH particles is seen to decrease by washing. The average size of Fe₃O₄ particles after washing is about 60 nm, which is slightly smaller than that before washing. Thus, as shown in Figs. 4 and 5, in comparison with Fe₃O₄ particles, it is confirmed that α-FeOOH particles are more likely to be dissolved by the HCl solution. As it turned out, the value of saturation magnetization for the final oxidation products washed with 5.0 mol/L HCl solution reached about 70 emu/g, which is about 20% greater than that before washing.

Fig. 4.

XRD patterns of the final oxidation products of the pretreated GR(Cl^–) suspension after washing with (a) 0.0, (b) 2.5 and (c) 5.0 mol/L HCl solutions. The arrows indicates the 110 peak for α-FeOOH.

Fig. 5.

Influence of HCl solution concentration on the saturation magnetization at room temperature of the final oxidation products after washing.

3.2. Influence of Pretreatment using Deaerated Water on GR(Cl^–) Suspension

Pretreatment using deaerated water is useful for the synthesis of Fe₃O₄ particles from the GR(Cl^–) suspension as discussed in the previous section. It is expected that a small amount of fine Fe₃O₄ particles is nucleated in the GR(Cl^–) suspension by such pretreatment, although the XRD pattern of the pretreated GR(Cl^–) suspension is very similar to that of the as-synthesized GR(Cl^–) suspension, as seen in Fig. 1. In this section, characterization of magnetic properties is examined to clarify the effect of the pretreatment using deaerated water, because GR(Cl^–) and Fe₃O₄ are in a paramagnetic and a ferrimagnetic state, respectively, at room temperature.⁵⁾

Figure 6 indicates the magnetization curves at room temperature of solid particles obtained by centrifugation from the pretreated GR(Cl^–) and the as-synthesized GR(Cl^–) suspensions. In a glove box, the pretreated GR(Cl^–) suspension was kept at room temperature, and solid particles extracted from it at different times were set into a sample holder. The solid particles of as-synthesized GR(Cl^–) suspension show a linear magnetization curve due to the paramagnetism. A non-linear magnetization curve is observed for the solid particles of the pretreated GR(Cl^–) suspension extracted at 1 h. Such tendency becomes clear in the magnetization curve of the solid particles extracted at 2 h. Moreover, the magnetization curve of the solid particles extracted at 68 h shows a hysteresis loop.

Fig. 6.

Magnetization curves at room temperature of the solid particles extracted from the as-synthesized GR(Cl^–) suspension, together with those extracted from the pretreated GR(Cl^–) suspensions at the different times of 1, 2, and 68 h.

It is well known that ferromagnetic and ferrimagnetic materials exhibit a magnetization curve with hysteresis. However, their small particles exhibit the superparamagnetism, where the magnetic moment of each particle is free to fluctuate in response to thermal energy, while the individual atomic moments maintain their ordered state.²⁾ Such behavior leads to a non-linear magnetization curve with no hysteresis^2,5) because thermal fluctuation of the magnetic moment is too fast. Thus, the non-linear magnetization curve of the pretreated GR(Cl^–) suspension extracted at 1 and 2 h highly suggests the presence of superparamagnetic fine Fe₃O₄ particles in the suspension. Consequently, it is probable that there is a small quantity of fine Fe₃O₄ particles nucleated by pretreatment, which grow in the pretreated suspension as time passes without any injection of oxygen gas.

3.3. Change in Oxidation Process of GR(Cl^–) Suspension by Control of Oxidation

To investigate the formation process of Fe₃O₄ particles, solid particles extracted from the pretreated GR(Cl^–) suspension for different oxidation times were analyzed by XRD measurements. In addition, the values of ORP, pH and DO in the pretreated GR(Cl^–) suspension were measured during the injection of oxygen gas because the precipitation of iron oxides are affected by conditions of aqueous solution. Figures 7 and 8 show XRD patterns of solid particles extracted from the pretreated GR(Cl^–) and as-synthesized GR(Cl^–) suspensions, respectively, at different oxidation times. Since the specimens for XRD measurements contain not only solid particles but also aqueous solution and glycerol, the XRD patterns reveal a broad background. In the case of the pretreated GR(Cl^–) suspension, the diffraction peak assigned to 311 of Fe₃O₄ appears at 30 min of oxidation. Its intensity increases with increasing oxidation time. Diffraction peaks assigned to α-FeOOH are observed at 120 min of oxidation. Thus, Fe₃O₄ is formed preferentially in the pretreated GR(Cl^–) suspension at a relatively early stage. In the case of the as-synthesized GR(Cl^–) suspension, diffraction peaks corresponding to α-FeOOH are observed at 30 min of oxidation. Diffraction peaks corresponding to Fe₃O₄ appear at 60 min of oxidation. Therefore, pretreatment using deaerated water enhances the formation of Fe₃O₄ at a relatively early stage, suggesting that fine Fe₃O₄ particles nucleated by pretreatment act as seeds.

Fig. 7.

XRD patterns of solid particles extracted from the pretreated GR(Cl^–) suspension at the different oxidation times of 0, 30, 60, 120 and 240 min at 298 K. The arrows indicate the 110 peak for α-FeOOH and the 311 peak for Fe₃O₄.

Fig. 8.

XRD patterns of solid particles extracted from the as-synthesized GR(Cl^–) suspension at the different oxidation times of 0, 30, 60, 120 and 240 min at 298 K. The arrows indicate the 110 peak for α-FeOOH and the 311 peak for Fe₃O₄.

The oxidation time dependence of (a) DO, (b) ORP and (c) pH in the pretreated GR(Cl^–) and as-synthesized GR(Cl^–) suspensions are shown in Fig. 9. The points marked by closed circles and squares in the figure indicate oxidation times at which specimens for XRD measurements in Figs. 7 and 8 were extracted. The value of DO of the as-synthesized GR(Cl^–) suspension is very small up to about 80 min of oxidation because the dissolved oxygen is consumed to oxidize GR(Cl^–). When GR(Cl^–) is oxidized completely, the value of DO is sharply increased and saturated. Thus, the oxidation of the pretreated GR(Cl^–) suspension is completely achieved at about 180 min of oxidation, which is longer than that of the as-synthesized GR(Cl^–) suspension.

Fig. 9.

Oxidation time dependence of the (a) DO, (b) ORP and (c) pH in the pretreated and as-synthesized GR(Cl^–) suspensions. The points marked by closed circles and squares indicate oxidation times at which specimens for XRD measurements in Figs. 7 and 8 were extracted.

When the oxidation of the pretreated GR(Cl^–) suspension proceeds, the values of ORP and pH gradually increase and decrease, respectively, with increasing oxidation time. A significant change of ORP and pH curves at about 180 min of oxidation results from the increase of DO. It should be noted that a slight increase of the ORP curve and a slight decrease of the pH curve are observed at 120 min of oxidation. Since the precipitation of α-FeOOH particles mainly begins at oxidation times above 120 min, as shown in Fig. 7, a slight increase of ORP and decrease of pH at 120 min of oxidation is related to the formation of α-FeOOH. On the other hand, when the oxidation of the as-synthesized GR(Cl^–) suspension proceeds, the value of ORP increases markedly up to –100 mV at a relatively early stage and then gradually decreases with increasing oxidation time up to about 80 min. At this time, the value of pH drastically decreases up to about 6.5 and then gradually increases. These behaviors of ORP and pH of the as-synthesized GR(Cl^–) suspension are similar to those of previous studies.²³⁾ Before the injection of oxygen gas, the pretreated GR(Cl^–) suspension exhibits lower ORP and higher pH in comparison with the as-synthesized GR(Cl^–) suspension. These tendencies are maintained during oxidation. Consequently, it is highly suggested that fine Fe₃O₄ particles nucleated by pretreatment grew preferentially in the solution with relatively low ORP and high pH.

4. Conclusion

Fe₃O₄ particles were synthesized by oxidation of GR(Cl^–) suspension at room temperature. In this process, the pretreated GR(Cl^–) suspension was made in such a way that the solution of as-synthesized GR(Cl^–) suspension was replaced by deaerated water, and then oxygen gas was injected at room temperature. After pretreatment, a small quantity of fine Fe₃O₄ particles was nucleated in the suspension. By controlling the injection of oxygen gas into the suspension, Fe₃O₄ particles were formed preferentially because of the growth of fine Fe₃O₄ particles nucleated by pretreatment. As it turned out, Fe₃O₄ particles of about 70 nm in diameter were obtained from the pretreated GR(Cl^–) suspension although the as-synthesized GR(Cl^–) suspension mainly transformed to α-FeOOH particles under the same oxidation conditions. Note that the saturation magnetization of the final oxidation products obtained from the pretreated GR(Cl^–) suspension was about 60 emu/g, six times larger than that obtained from the as-synthesized GR(Cl^–) suspension. In addition, it increased up to about 70 emu/g by purification using HCl solution, suggesting that the fraction of Fe₃O₄ particles in the final oxidation products is about 70%.

When the nucleated fine Fe₃O₄ particles grew in response to the injection of oxygen gas into the pretreated GR(Cl^–) suspension, the value of ORP was lower than that in the as-synthesized GR(Cl^–) suspension. In addition, the value of pH of the former was higher than that of the latter. Consequently, the formation of Fe₃O₄ particles are enhanced in solution with relatively low ORP and high pH.

The Fe₃O₄ particles have attached much attention, because these are promising candidates for magnetic fluid, magnetic hyperthermia and contrast agent for magnetic resonance Imaging (MRI) based on nuclear magnetic resonance (NMR). Since various synthesis techniques of Fe₃O₄ particles are developed for these applications, the present results on oxidation process of GR(Cl^–) are considered to be useful to synthesize Fe₃O₄ particles at low processing temperature.

Acknowledgements

This research was partially supported by a Grant-in-Aid for Scientific Research Funds from the Japan Society for Promotion of Science (No. 23360276 and No. 23760620).

References

• 1)   K.  Raj and  R.  Moskowitz: J. Magn. Magn. Mater., 85 (1990), 233.
• 2)   Q. A.  Pankhurst,  J.  Connolly,  S. K.  Jones and  J.  Dobson: J. Phys. D, Appl. Phys., 36 (2003), R167.
• 3)   P.  Tartaj,  M.  Del Puerto Morales,  S.  Veintemillas-Verdaguer,  T.  González-Carreño and  C. J.  Serna: J. Phys. D, Appl. Phys., 36 (2003), R182.
• 4)   B.  Jeyadevan: J. Ceram. Soc. Jpn., 118 (2010), 391.
• 5)   R. M.  Coronell and  U.  Schwertmann: The Iron Oxides, 2nd ed., Wiley-VCH, Weinheim, (2003).
• 6)   Y. S.  Kang,  S.  Risbud,  J. F.  Rabolt and  P.  Stroeve: Chem. Mater., 8 (1996), 2209.
• 7)   C. -Y.  Hong,  I. J.  Jang,  H. E.  Horng,  C. J.  Hsu,  Y. D.  Yao and  H. C.  Yang: J. Appl. Phys., 81 (1997), 4275.
• 8)   G. F.  Goya,  T. S.  Berquó,  F. C.  Fonseca and  M. P.  Morales: J. Appl. Phys., 94 (2003), 3520.
• 9)   A.  Pineau,  N.  Kanari and  I.  Gaballah: Thermochim. Acta, 447 (2006), 89.
• 10)   S.  Sun and  H.  Zeng: J. Am. Chem. Soc., 124 (2002), 8204.
• 11)   S.  Bhagwat,  H.  Singh,  A.  Athawale,  B.  Hannoyer,  S.  Jouen,  B.  Lefez,  D.  Kundaliya,  R.  Pasricha,  S.  Kulkarni and  S.  Ogale: J. Nanosci. Nanotechnol., 7 (2007), 4294.
• 12)   S.  Suzuki,  E.  Matsubara,  T.  Komatsu,  Y.  Okamoto,  K.  Kanie,  A.  Muramatsu,  H.  Konishi,  J.  Mizuki and  Y.  Waseda: Corros. Sci., 49 (2007), 1081.
• 13)   S.  Suzuki,  S.  Suzuki,  K.  Inoue,  E.  Matsubara and  Y.  Waseda: Surf. Interface Anal., 40 (2008), 307.
• 14)   J. -M. R.  Génin,  G.  Bourrié,  F.  Trolard,  M.  Abdelmoula,  A.  Jaffrezic,  P.  Refait,  V.  Maitre,  B.  Humbert and  A.  Herbillon: Environ. Sci. Technol., 32 (1998), 1058.
• 15)   L.  Simon,  M.  François,  P.  Refait,  G.  Renaudon,  M.  Lelaurain,  J.-M. R.  Génin: Solid State Sci., 5 (2003), 327.
• 16)   J. -M. R.  Génin,  C.  Ruby and  C.  Upadhyay: Solid State Sci., 8 (2006), 1330.
• 17)   P.  Refait,  M.  Abdelmoula,  J. -M. R.  Génin: Corros. Sci., 40 (1998), 1547.
• 18)   J. -M. R.  Génin and  C.  Ruby: Solid State Sci., 6 (2004), 705.
• 19)   T.  Misawa,  W.  Suëtaka and  S.  Shimodaira: Bull. Chem. Soc. Jpn., 42 (1969), 3339.
• 20)   T.  Misawa,  K.  Hashimoto and  S.  Shimodaira: Corros. Sci., 14 (1974), 131.
• 21)   T.  Misawa: Zairyo-to-Kankyo, 50 (2001), 538.
• 22)   P.  Refait,  A.  Géhin,  M.  Abdelmoula,  J. -M. R.  Génin: Corros. Sci., 45 (2003), 659.
• 23)   F.  Nagata,  K.  Inoue,  K.  Shinoda,  I.  Bergmann,  V.  Šepelák and  S.  Suzuki: Mater. Trans., 50 (2009), 2557.
• 24)   K.  Shinoda,  S.  Fujieda,  S.  Tsuri and  S.  Suzuki: Mater. Corros., 63 (2012), 343.
• 25)   F.  Nagata,  K.  Inoue,  S.  Fujieda,  K.  Shinoda,  S.  Suzuki and  S.  Tsuri: Mater. Corros.,62 (2011), 1100.
• 26)   G.  Sahoo,  S.  Fujieda,  K.  Shinoda and  S.  Suzuki: Corros. Sci., 53 (2011), 2446.
• 27)   G.  Sahoo,  S.  Fujieda,  K.  Shinoda,  S.  Yamaguchi,  M.  Korosaki and  S.  Suzuki: Corros. Sci., 53 (2011), 4001.
• 28)   F.  Nagata,  K.  Inoue,  K.  Shinoda and  S.  Suzuki: ISIJ Int., 49 (2009), 1730.