Effect of hydrogen ion dose and sample temperature on hydrogenation of Mg oxides using microwave excited hydrogen plasma

To characterize the treatment condition of the hydrogen plasma, the Γ_i which is obtained from the ion saturation current density was measured with a Langmuir probe. The sample temperature T during the plasma treatment was measured with a thermocouple. Figure 3 shows the spatial distribution of the hydrogen ion saturation current density J_is compared between the direct treatment and the mesh treatment, without and with a mesh electrode, respectively. The gas pressure of hydrogen was 15 Pa. The measured J_is, which is proportional to the plasma density when the electron temperature is constant, exponentially decreased along with the z-axis perpendicular from the microwave launcher. By using mesh electrode, J_is decreased because some charged particles were trapped. In this experiment, the sample was put at z = 60, 80, 115 mm from the microwave launcher. The plasma density at z = 80 and 115 mm was a three-quarters and a half of the density at z = 60 mm, respectively.

Zoom In Zoom Out Reset image size

Figure 4 shows the T of Mg oxides at z = 60, 80 and 115 mm during the plasma treatment compared between the direct treatment and the mesh treatment. The T increased with t and Γ_i . After t = 30 min, T was almost in equilibrium. Focusing on the condition in the same t and z of the direct treatment and the mesh treatment, the T using mesh electrode was reduced by about 20 °C compared to the direct treatment. The hydrogen ion flux to the sample surface is suppressed by the mesh electrode, but neutral particles pass through the mesh electrode. Therefore, the hydrogen ion contributes to increasing T on the sample surface by the effect such as ion bombardment. Considering the results of J_is and T measurement shown in Figs. 3 and 4, the samples were treated by the plasma with the condition of hydrogen ion dose listed in Table I.

Zoom In Zoom Out Reset image size

Figure 5 shows the dependence of T at t = 30 min on the hydrogen ion flux compared between the direct treatment and the mesh treatment. The T increased with the hydrogen ion flux. In the mesh treatment, the relative amount of radical incident on the sample increases when compared with the same ion flux, and the H^*/H⁺ ratio irradiated to the sample becomes high. Therefore, the T with the mesh treatment was higher than the direct treatment. The hydrogen radical also contributes to increasing T around 26 °C by chemical reaction such as hydrogenation and reduction. Theoretically, hydrogen radical generated by the dissociation due to inelastic collision with electrons has and carries an energy of 2.2 eV/atom. ²⁵⁾ The mass of plasma treated samples decreased by 0.46% on average. It is speculated that oxygen atoms in the sample surface were volatilized by reduction or replaced with hydrogen atoms.

Zoom In Zoom Out Reset image size

Figure 6 shows XPS spectra in Mg 2p region of hydrogen and Ar plasma treated Mg oxides. In the XPS analysis, the peaks of MgO, Mg(OH)₂ and metal Mg, which was assigned at the binding energy of 50.25, 49.5 and 48 eV, ^26–29) were considered. The surface oxidized Mg before plasma treatment was composed of amorphous Mg(OH)₂ and MgO. ^30–32) In the plasma treatment, the samples were placed at z = 60 mm where the Γ_i was 4.2 and 1.7 × 10¹⁶ cm⁻² s⁻¹ without and with mesh electrode, respectively. Hydrogen radical flux is considered to be same at the same z. As the results of hydrogen plasma treated sample compared to the untreated sample, MgO peak decreased and Mg(OH)₂ or MgH_x peak appeared. However, it is difficult to distinguish between Mg(OH)₂ and MgH_x from XPS measurements. The intensity of metal Mg peak increased with t in the hydrogen plasma treatment. Moreover, the samples were treated by not only hydrogen plasma but also Ar plasma in which chemically reactive species do not exist. In the Ar plasma treatment, the sample surface is only affected by Ar ion irradiation and bombardment. The metal Mg peak did not appear in Ar plasma treatment. This means that MgO was reduced into metal Mg by the chemically reactive species in hydrogen plasma. Furthermore, the metal Mg peak shifted to the high energy side with increasing t of the direct treatment which is higher Γ_i . Therefore, the peak shift of metal Mg peak was caused by hydrogen ion dose. It is speculated that this peak shift means that the electron density around the Mg atom decreases by the defective binding such as Mg–(OH)_x and Mg–H_x . The Mg peak intensity of the direct treatment was higher than the mesh treatment. This means that the reduction of Mg oxides into metal Mg will progress with Γ_i , or the metal Mg will appear from the dehydrogenation of the MgH₂ as H₂ molecules by hydrogen radical or increasing T by the hydrogen ion dose. The dehydrogenated MgH₂ due to increasing T is converted into Mg which is reported by Nogita et al. ³³⁾ The bulk MgH₂ was dehydrogenated at about 370 °C and 0.01 Pa, while synthesized MgH₂ by hydrogen plasma was dehydrogenated at about 256 °C and 0.2–0.5 Pa. ^34,35) It is speculated that the reduction and oxidation will be simultaneously caused by increasing the hydrogen ion dose.

Zoom In Zoom Out Reset image size

To help understanding the hydrogenated Mg such as MgH₂ and Mg(OH)₂ by hydrogen plasma, the samples were analyzed by Raman spectroscopy. Figure 7 shows the Raman spot characterized spectra of hydrogen plasma treated and untreated Mg oxides with each Γ_i . The position of spot analysis was set to where seems to be MgH₂ nucleus by optical micrographs (OM) installed in the NRS-7100. The size of Raman laser spot was 1 μmf. Figure 8 shows the OM images of Raman spot position in hydrogen plasma treated samples under the treatment conditions selected from in Fig. 7(a). Kondo et al. reported the MgH₂ nucleus at initial state of hydrogenation was formed as granular and scattered. ³⁶⁾ As shown in Fig. 7, the Mg oxides before the plasma treatment consisted of Mg(OH)_x at 275 cm⁻¹ and MgO at 595 and 1096 cm⁻¹. ^37–39) The Raman shift of Mg(OH)₂ was assigned at 257 cm⁻¹ corresponding to Raman active phonons E_g. ⁴⁰⁾ The Raman shifts of MgH₂ have three peaks at 311, 953 and 1281 cm⁻¹ corresponding to Raman active phonons B_1g, E_g and A_1g, respectively. ^40,41) As shown in Fig. 7(a) obtained from the direct treatment for 10 min, the Mg oxides was converted into Mg(OH)_x , MgH_x , MgH₂ and finally Mg(OH)₂ with increasing Γ_i . At the lowest Γ_i treatment, which is z = 115 mm, Γ_i = 1.1 × 10¹⁶ cm⁻² s⁻¹ and T = 131 °C shown in Fig. 7(a), the Raman shifts of 365 and 375 cm⁻¹, which are corresponding to MgH_x and Mg(OH)_x , respectively, were observed. As an increase in Γ_i , MgH₂ peaks appeared at Γ_i = 2.7 × 10¹⁶ cm⁻² s⁻¹, and then MgH₂ disappeared at Γ_i = 4.2 × 10¹⁶ cm⁻² s⁻¹. The Raman shifts of 257 and 344 cm⁻¹, which are corresponding to Mg(OH)₂ and MgH_x , respectively, were observed at the highest Γ_i = 4.2 × 10¹⁶ cm⁻² s⁻¹ and T = 219 °C. Interestingly, the peak at 365 or 344 cm⁻¹ which are correspond to MgH_x appeared with increasing Γ_i . It is speculated that MgH_x at 365 cm⁻¹ observed at Γ_i = 1.1 × 10¹⁶ cm⁻² s⁻¹ is formed from an intermediate such as Mg(OH)_x by hydrogen plasma, and MgH_x at 344 cm⁻¹ observed at Γ_i = 4.2 × 10¹⁶ cm⁻² s⁻¹ is formed as a defected MgH₂. Matović et al. reported the Raman shift of 344 cm⁻¹ is related to interaction with independent point defects of MgH₂ by Ar ion dose which causes the ion bombardment to MgH₂. ³⁷⁾ It is speculated that MgH₂ is dehydrogenated into Mg(OH)₂ and MgH_x at 344 cm⁻¹ with increasing T due to Γ_i . Mg(OH)₂ at 257 cm⁻¹ will be generated by the reaction of reactive surface such as defect created by the ion bombardment with moisture in the air or by-products such as H₂O produced from the reduction of Mg oxides. In addition, the structure of plasma treated sample did not change in any sample after about 48 h of air exposure.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

As shown in Fig. 7(b) obtained from the direct treatment for 30 min, the Mg oxides was converted into MgH₂ and finally MgH_x and Mg(OH)₂ with increasing Γ_i . As an increase in t from 10 min to 30 min shown in Figs. 7(a) to 7(b), Mg(OH)_x and MgH_x at 365 cm⁻¹ were converted into MgH₂ at the lowest Γ_i = 1.1 × 10¹⁶ cm⁻² s⁻¹, and MgH₂ was converted into Mg(OH)₂ and MgH_x at Γ_i = 2.7 × 10¹⁶ cm⁻² s⁻¹. Therefore, the hydrogen ion dose affects the hydrogenation into MgH₂ from Mg oxides and also the dehydrogenation into Mg(OH)₂ and MgH_x from MgH₂. Figures 7(c) and 7(d) show the Raman spectra of the mesh treatment for 10 and 30 min, respectively. As shown in Fig. 7(c) obtained from the mesh treatment for 10 min, the Mg oxides was converted into Mg(OH)_x , MgH_x and MgH₂ with increasing Γ_i . At the highest hydrogen ion flux treatment with Γ_i = 1.7 × 10¹⁶ cm⁻² s⁻¹, MgH₂ peaks was still observed and more clearly than the treatment with Γ_i = 0.8 × 10¹⁶ cm⁻² s⁻¹. As discussed in Figs. 7(a) and 7(b), it is speculated that the dehydrogenation of MgH₂ is not only reduction by hydrogen ions but also an increase in T. However, T also increases with the hydrogen radical in the discussion of Fig. 5. In the plasma treatment with same Γ_i between the direct treatment and the mesh treatment in Figs. 7(a) and 7(c), the hydrogen radical flux of the mesh treatment is relatively higher than the direct treatment. The Mg(OH)_x and MgH_x at Γ_i = 1.1 × 10¹⁶ cm⁻² s⁻¹ in Fig. 7(a) were converted into MgH₂ at Γ_i = 0.8 × 10¹⁶ cm⁻² s⁻¹ in Fig. 7(c). Therefore, it is speculated that hydrogen radical also affects the hydrogenation of Mg oxides. As shown in Fig. 7(d) obtained from the mesh treatment for 30 min, the Mg oxides was converted into MgH₂ and Mg(OH)₂ with increasing Γ_i . This tendency was similar to the result for 10 min shown in Fig. 7(a). Compared to the treatment in the same hydrogen radical flux at z = 115 mm in Figs. 7(b) and 7(d), the MgH₂ peaks of the direct treatment in Fig. 7(b) was more clearly. It means that MgH₂ nucleus becomes high density by the higher ion flux treatment.

Figure 9 shows the condition map of plasma treated samples at z and t axis with isolines of T and the hydrogen ion dose Γ_i t by the direct and the mesh treatment. The experimental conditions in which Raman peaks (311, 953 and 1281 cm⁻¹) corresponding to MgH₂ are observed by the same measurement method as Fig. 7 are plotted as blue squares. The gray squares indicate the experimental conditions in which Mg(OH)_x was observed at 257 cm⁻¹ or 275 cm⁻¹ in Raman spectroscopy. MgH₂ nucleus at initial state of hydrogenation appeared under the experimental conditions with Γ_i t range of 1.3–3.2 × 10¹⁹ cm⁻² and 0.2–2.1 × 10¹⁹ cm⁻² for the direct and the mesh treatment, respectively. Excessive ion and radical irradiation cause the increase in sample temperature and accelerates the dehydrogenation of MgH₂. MgH₂ was not observed when the sample temperature was higher than about 200 °C. The minimum temperature for producing MgH₂ is also important conditions. In this experiment, the minimum temperature required to produce MgH₂ was about 150 °C. Therefore, the hydrogenation was caused by the synergistic effect of T and ion does, and which are expected to be optimized.

Zoom In Zoom Out Reset image size

Figure 10 shows the 2D Raman mapping characterization in growth of the hydrogenated Mg nucleus by hydrogen plasma treatment at z = 80 mm. The spatial resolution of the Raman mapping analysis was 1 μm. The scanned area in Fig. 10(b) was 100 μm², and others were 49 μm² which were optimized in the range of 36–100 μm² for the size of synthesized MgH₂ nuclei which position were searched by OM images. In the mapping measurements, the peak intensity of MgH₂ at a Raman shift of 953 cm⁻¹ was divided into mapping colors with a maximum peak intensity of red and a minimum intensity of black. The maximum peak intensity depended on the plasma treatment condition and the relative value of the maximum peak intensity is indicated above the scale bar. Considering the S/N ratio in the measurement, the MgH₂ nucleus area was identified as yellow and higher color part in the mapping measurement image. As shown in Fig. 10, MgH₂ nucleus area was expanded with Γ_i t. The largest MgH₂ nucleus area in all treatment conditions was obtained at z = 80, t = 20 min and Γ_i t = 3.2 × 10¹⁹ cm⁻² of the direct treatment shown in Fig. 10(b). The largest MgH₂ nucleus area on the oxidized Mg surface was 14.9 μm² estimated from the image analysis by using ImageJ. ²⁴⁾

Zoom In Zoom Out Reset image size

Figure 11 shows the relationship between the MgH₂ nucleus area on the oxidized Mg surface and the hydrogen ion dose. The MgH₂ nuclei were observed around 150 °C–200 °C. The MgH₂ nucleus area increased with increase in hydrogen ion dose in the direct and the mesh treatment. The straight line in Fig. 11 is obtained by the least squares method of the direct treatment plot, which means that the area of the nucleus is proportional to the hydrogen ion dose. The growth rate of MgH₂ nucleus will correlate with hydrogen ion dose. It is speculated that the MgH₂ nucleus area on the oxidized Mg surface was expanded by hydrogen ion dose to the Mg which reduced by the hydrogen plasma. Pranevicius et al. reported that the MgH₂ conversion under low-temperature condition from Mg requires the hydrogen diffusion into the bulk with not only high hydrogen ion flux irradiation but also thermal diffusion process. ⁴²⁾ The problem of the MgH₂ nucleus growth from Mg is a low hydrogen diffusivity in MgH₂. ^36,43) The hydrogen transport processes at grain boundaries, such as sample surface defects, are activated under high hydrogen ion flux irradiation. ^42,44) As a result of XPS analysis shown in Fig. 6, Mg oxides were reduced by hydrogen plasma with increasing the hydrogen ion dose. Hydrogen is transported by the defects which is formed by the high hydrogen ion dose to the reduced Mg oxides. Interestingly, the MgH₂ nucleus area of the mesh treatment which is higher H^*/H⁺ ratio was larger than that of the direct treatment in the lower hydrogen ion dose range under ∼2 × 10¹⁹ cm⁻². MgH₂ nuclei were synthesized in the range of Γ_i t and T shown in Fig. 9. As shown in Figs. 4 and 5, both of hydrogen ion and radical contribute to increasing T. Therefore, in this low hydrogen ion dose range, thermal dehydrogenation will be avoided, and growth of MgH₂ nuclei by hydrogen ion or radical will be observed. It is speculated that the hydrogen radical is also contributed to the nucleation and growth of MgH₂ by diffusing at grain boundaries along with hydrogen ion. MgH₂ nuclei were synthesized in the range of Γ_i t and T shown in Fig. 9.

Zoom In Zoom Out Reset image size