Surface Structures and Hydrogenation Properties of Ti–Pd Alloys Immersed in Hydrogen Peroxide

Abstract

This study achieves an increase in Pd concentration on the surface of Ti–Pd alloys using hydrogen peroxide as a green dealloying method, as well as determining the effect of Pd on the hydrogenation of the Ti–Pd alloys. Spontaneous oxidation of Ti in Ti–Pd alloys has been reported to precipitate low-valent Pd on the surface, which can be used for fast δ-TiH₂ formation and as a catalyst for various organic reactions. On the other hand, spontaneous oxidation has limited potential to increase the surface Pd concentration. As most of the Pd remains in the metallic phase, there is a need to increase the utilization of the remaining Pd. H₂O₂ is known to be a green oxidant and also forms complexes with Ti, therefore surface Pd enrichment by dealloying is expected. This study was carried out to investigate the relationship between hydrogenation and surface properties of Ti–Pd alloys by H₂O₂ immersion. The increase in the thickness of Ti oxide layers and the increase in Pd concentration on Ti–Pd alloys were found by H₂O₂ immersion. Model experiments on chip-like specimens showed that the Ti oxide layer retards hydrogen diffusion, while the Pd on the surface has the effect of increasing the hydrogen supply to the metallic phase. Pd on the surface was also found to have an effect on the fast decomposition of H₂O₂. These results indicate that H₂O₂ immersion is effective as a surface treatment to increase the Pd concentration on the surface with reduced Ti dissolution.

1. Introduction

As the world moves towards a hydrogen society, the potential of hydrogen storage materials to improve energy efficiency is being investigated. Among the variety of hydrogen storage materials, some Ti-based hydrogen storage materials are inexpensive and can absorb and desorb hydrogen at low pressure and low ambient temperature.^1–3) The advantage of Ti-based hydrogen storage materials is that, while not reaching the 5.7 mass% achieved in Type IV 70 MPa pressurized H₂ storage tanks with a weight hydrogen storage density of around 1–4 mass%, their volume storage density is 40–70 g H₂/L and the advantages of storing hydrogen in solid state (40 g H₂/L when stored in Type IV 70 MPa pressurized H₂ storage tanks).⁴⁾ In addition, since the hydrogen is stored in a thermodynamic system, the hydrogen pressure in the gas phase can be kept at a low pressure. This character allows long-term storage. Therefore, Ti-based hydrogen storage materials are expected to be used in stationary applications due to their practicality and low cost.^5,6) Despite these advantages, due to the formation of a stable passive oxide film upon exposure to air, Ti-based hydrogen storage materials generally require stringent conditions for initial activation at high temperatures in a vacuum.^7–9) Some researches also carried out surface modification with respect to pure Ti.^10,11) For the Ti–Pd alloys we have reported, as well, the precipitation of low-valent Pd on the surface results in the fast hydrogenation.¹²⁾ The Pd is also reported to act as a stable catalyst on the surface and to proceed various organic chemical reactions.^13–15) However, the amount of Pd precipitated on the surface of Ti–Pd alloys with spontaneous oxide film formation is determined by the inward diffusion of oxygen and outward diffusion of Ti,^16,17) which means that most of the Pd in the alloy would remain in the Ti–Pd alloy phase present in the inner region. Indeed, referring to XPS results, The Pd concentration is reported to decrease slightly from the outermost surface towards the inner regions, and then increase further towards the inner regions.¹²⁾ Therefore, the dealloying method, i.e. actively dissolving Ti on the surface by using hydrogen peroxide, would be effective. Since Ti forms a stable oxide film, a strong acidic solution such as hydrogen fluoride is required to dissolve Ti. In contrast, H₂O₂ has been studied as a green oxidant and has also been used for organic oxidation reactions.^18,19) When a Ti–Pd alloy is subjected to oxidation with H₂O₂ solution, active dissolution of Ti by formation of complexes with H₂O₂ and precipitation of Pd on the surface can be expected. The precipitated Pd on the surface would also act as a catalyst for the decomposition reaction of H₂O₂,²⁰⁾ which may suppress the amount of Ti dissolution. At the same time, when immersed in H₂O₂, enriched surface Pd is expected to act as a catalyst for the dissociation reaction of H₂ molecules at the surface,^21,22) leading to an acceleration of the formation of TiH₂. Therefore, this study reports on the hydrogenation properties and surface conditions of Ti–Pd alloys immersed in H₂O₂ with the purpose of enriching the Pd concentration at the surface of the alloys. The decomposition reactions of H₂O₂ and the effect of Pd on Ti oxides during this immersion treatment are also studied.

2. Experimental Procedure

2.1 Sample preparation

Ti–Pd alloys were prepared by the Ar arc melting method. Pure Ti buttons were prepared by arc melting pure sponge Ti (99.7 mass%) (Nilaco, Japan) before forming the Ti–Pd alloy. Subsequently, Ti–xPd alloys (x = 0–1 mol%) were alloyed by arc melting varying amounts of pure Ti button and pure Pd sheet (99.98 mass%) (Tanaka kikinzoku kogyo K.K., Japan) under Ar using a water-cooled copper furnace. Each alloy was re-melted at least 10 times to increase compositional homogeneity. These ingots were milled in an axial depth of cut of 0.1 mm on a milling machine to obtain turnings.

Each of the turnings (0.5 g) was immersed in 10 mL of H₂O₂ (30 mass%) (Fujifilm wako pure chemical corporation, Japan) for 64.8 ks at room temperature (specimen:H₂O₂ = 1:20) under atmospheric conditions. The Ti–Pd alloy of the turnings after immersion in H₂O₂ is abbreviated by the subsequent addition of H₂O₂ (e.g. Ti_H₂O₂). These samples were first cleaned with methanol and then rinsed with acetone. Nitrogen gas was used for final drying.

A chip-like specimens (3 mm × 3 mm × 1 mm) was used for the observation and analysis of the surface properties and for the modelling experiments. Field emission scanning electron microscopy (FE-SEM) (JSM-6700F, JEOL Ltd., Japan) and X-ray photoelectron spectroscopy (XPS) (PHI5000 Versaprobe, ULVAC-PHI, Inc., Japan) were used to observe and analyses the surface properties. Chip-like specimens were cut using a micro cutter with a SiC grinding wheel. The final surface was prepared dry using SiC waterproof abrasive paper (of #320, #800, #1000 and up to #2000).

2.2 Evaluation of initial activation properties, constituent phases, surface properties and dissolution

Initial activation properties were evaluated using a Sievert’s apparatus: 500 mg samples were heated in a stainless-steel reactor from ambient temperature to 673 K at 5 K min⁻¹ and an initial hydrogen pressure of 3.0 MPa.

X-ray diffraction (XRD) (MiniFlex II, Rigaku Corporation, Japan) was used for phase identification. After hydrogenation, the turnings were ground with a pestle and mortar to obtain powdered samples. The measurement conditions were as follows: scanning speed of 2.0° min⁻¹, sampling width of 0.005°, scanning range of 10° to 90°, divergence slit of 0.625°, and no scattering slit and no receiving slit. A Ni filter was used as a filter. A semiconductor radiation detector was used as the detector.

XPS analysis was performed using AlKα radiation with a 1486.6 eV excitation source. The take-off angle for photoelectron detection was 45° from the sample surface. The bonding energies were calibrated by the C 1s peak at 284.5 eV and Ag 3d_5/2 at 368.3 eV. The intensity of the photoelectron peak was estimated by subtracting the background from the spectrum using the Shirley method. Peak fitting and spectral integration were performed using a commercial software package (CasaXPS; Casa Software, Teignmouth, UK).

The dissolution rate was evaluated by inductively coupled plasma atomic emission spectrometry (ICP-AES) (ICPS-7510, Shimadzu Corporation, Japan) using a radiofrequency power of 1.2 kW, pure argon as the measuring gas and a carrier gas flow rate of 0.70 L min⁻¹.

The H₂O₂ decomposition reaction was evaluated using a gas burette. 0.05 g of turnings were weighed and immersed in 1 mL of H₂O₂ (30 mass%) at room temperature for 64.8 ks in the ratio specimen:H₂O₂ = 1:20. Since 1 mL of H₂O₂ produces 109 mL of oxygen gas, 0.5 g of turnings requires 10 mL of H₂O₂, which would exceed the volume of the gas burette. Therefore, 0.05 g of turnings was used. Readings were taken at 30 second intervals. Saturated NaCl solution was used as the solution.

The BET method was employed for the specific surface area by using pure Kr gas (ASAP2010, Shimadzu Corporation, Japan).

2.3 Model experiment

In order to investigate the effect of the oxide film on the Ti surface, specimens in the form of chips have been used. To investigate the effect of natural oxide film, pure Ti was immersed in pure water for 1.8 ks and 3.6 ks (subscript +PW(time)).²³⁾ The surfaces were polished with #150 SiC abrasive paper before immersion in pure water (subscript MP). Pd was also deposited on the surface using a magnetron sputtering apparatus with a thickness of 5 nm (subscript wPd). Four samples were prepared according to the above-mentioned processes. Ti_MP, Ti_MPwPd, Ti_MP+PW(1.8)wPd and Ti_MP+PW(3.6)wPd. To investigate the effect of H₂O₂ immersion treatment, Ti–Pd samples were immersed for 64.8 ks under atmospheric conditions with a fixed weight/volume ratio of H₂O₂ solution of 1:20. These samples are referred to as H₂O₂ in parentheses. Experiments without Pd deposition have also been carried out, but are not shown in the results, because the hydrogenation could not be confirmed.

Hydrogenation of the chip-like specimen was evaluated by high-pressure differential scanning calorimetry (HP-DSC) (HP-DSC8230, Rigaku Corporation, Japan). The conditions were as follows: hydrogen inlet pressure 3.0 MPa, temperature range from 373 K to 723 K and a temperature increase/decrease rate of 10 K min⁻¹.

3. Results and Discussions

A representative relationship between the hydrogen/metal ratio and the reaction time profiles of Ti–Pd alloys immersed in H₂O₂ is shown in Fig. 1. For reference, data for untreated Ti–Pd alloys are also shown in the same figure.¹²⁾ Hydrogenation of untreated Ti tended to be faster with increasing Pd addition, and a similar trend was observed for Ti–Pd alloys immersed in H₂O₂, with Ti–1.0Pd and Ti–0.5Pd tending to form hydrides faster. On the other hand, even for pure Ti, immersing in H₂O₂ resulted in faster hydrogenation compared to untreated Ti. α-Ti can be hydrogenated by gradual hydrogenation around H/M = 0.1–0.4, followed by drastic hydrogen absorption. Significant heat generation was observed during the drastic hydrogen absorption. The reason for the faster hydrogenation achieved by immersion in H₂O₂ is predicted to depend on the oxide/metal phase ratio contained in the sample. As discussed later, results supporting the effect of Ti oxide in delaying the delivery of hydrogen to the internal residual metallic phase were obtained from hydrogenation experiments on chip-like specimens. Referring to the phase diagram, α-Ti to cubic δ-TiH₂ (with peak attributed to TiH_1.924) forms a miscibility gap hydride.²⁴⁾ Kværndrup et al. also report that the hydrogenation of Ti in a flow of hydrogen gas leads to the formation of δ-TiH₂ by an abrupt transition through the β-Ti with the appearance of an exothermic peak.²⁵⁾ The hydrogenation of Ti–Pd alloys shown in Fig. 1, under conditions maintained at a constant temperature of 673 K, leads to the gradual solid solution of hydrogen occurring in Ti and finally to the formation of δ-TiH₂. The diffusion of hydrogen in α-Ti was reported to be fast and diffusive into the metallic phase.^26–28) From these facts, the limit of solid solution region to induce an abrupt phase transformation from metallic phase to δ-TiH₂ is considered to be a decrease in the ratio of metallic phase within the oxide film. The following results also support the above. On a chip-like specimens, the hydrogenation reaction could not be confirmed within the operating time when hydrogenation was carried out using a Sievert’s apparatus at 673 K and 3 MPa H₂.

Fig. 1

Representative relationship between the hydrogen/metal ratio and the reaction time profiles of Ti–Pd alloys immersed in H₂O₂.

The XRD profiles measured on the samples after hydrogenation are shown in Fig. 2. All samples were found to form δ-TiH₂, with no obvious change in lattice parameters with Pd concentration as in Ti–Pd alloys, but with compositional variations in data. This is due to the fact that for δ-TiH₂, TiH_1.971 and TiH_1.924 have been reported. Both have a $Fm\bar{3}m$ crystal structure, but the position of the diffraction peaks varies with hydrogen content. Hadjixenophontos et al. also reported that when δ-TiH₂ is formed, the position of the diffraction peak shifts to the high angle side at low hydrogen concentration and to the low angle side at high hydrogen concentration.²⁹⁾

Fig. 2

XRD profiles of Ti–(0–1)Pd alloys which immersed in H₂O₂ and hydrogenated.

The results of the XPS analysis of Ti–Pd alloys immersed in H₂O₂ are shown in Fig. 3. As expected, the Pd concentration on the surface increased in Ti–1.0Pd. The binding energy of Pd in the Ti–Pd alloy immersed in H₂O₂ is independent of the Pd concentration and is 335.1 eV. However, in untreated Ti–Pd alloys, the Pd valence tends to shift to higher valence positions with increasing Pd concentration, in agreement with previous studies.¹²⁾ Figure 4 shows the SEM-EDS measurements of Ti–Pd alloys after immersion in H₂O₂. Immersion in H₂O₂ resulted in a rough surface and the oxygen concentration increased from a few % to several tens.

Fig. 3

XPS profiles at Pd of Ti–Pd alloys as polished and immersed in H₂O₂.

Fig. 4

SEM-EDS measurements of Ti–Pd alloys after immersion in H₂O₂.

Figure 5 shows the relationship between the immersion time in H₂O₂ and the amount of gas generated, and as the Pd concentration on the surface of the sample increases with the dissolution of Ti, faster decomposition of H₂O₂ due to the catalytic effect of Pd can also be assumed. The amount of gas generated showed a faster plateau region as the Pd concentration on the Ti–Pd alloys increased. The decomposition reaction of H₂O₂ is assumed to proceed according to the following reaction equation.³⁰⁾

  

\begin{equation} \text{H$_{2}$O$_{2}$($l$)} = \text{H$_{2}$O($l$)} + \frac{1}{2}\text{O$_{2}$($g$)} \end{equation} (1)

  

\begin{equation*} \Delta_{\text{r}}H{{}^{\circ}} = -98.05\,\text{kJ$\,$mol$^{-1}$} \end{equation*}

  

\begin{equation*} \Delta_{\text{r}}S{{}^{\circ}} = 62.90\,\text{J$\,$K$^{-1}{}\,$mol$^{-1}$} \end{equation*}

Fig. 5

Relationship between the immersion time in H₂O₂ and the amount of gas generated.

The amount of oxygen that could be produced from 30 mass% H₂O₂ is 109 mL, and from 35 mass% H₂O₂ 132 mL, however the amount of gas produced was greater than expected. This extra gas production could be hydrogen gas produced by Ti oxidation. Detailed component analysis by mass spectrometer may be required to analyse its composition. Figure 6 shows the amount of Ti and Pd dissolved in H₂O₂, which decreases with increasing Pd concentration in Ti–Pd alloys. This can be understood as a result of the faster termination of gas generation from H₂O₂ as shown in Fig. 5, i.e. the faster H₂O₂ decomposes. The results for H₂O₂ decomposition and Ti dissolution also support an increase in surface Pd concentration as shown by the XPS results in Fig. 3. Figure 7 shows the relationship between Pd concentration and specific surface area upon immersion in H₂O₂. The results also indicate that immersion in H₂O₂ results in a rough surface.

Fig. 6

Amount of Ti and Pd dissolved in H₂O₂.

Fig. 7

Relationship between Pd concentration and specific surface area upon immersion in H₂O₂.

Two effects, an increase in the amount of Ti oxide film and an increase in the amount of Pd precipitation, are shown as a result of immersion in H₂O₂. The increase in the amount of the oxide film and the decrease in the amount of metallic phase remaining in the inner part of the turnings. Increasing the surface Pd concentration leads to increasing the amount of hydrogen supplied to α-Ti. The thickness of the oxide layer content prevents the diffusion of hydrogen into the metallic phase, but the decrease in the proportion of metallic phase in the turnings and the increase in surface Pd concentration cause fast hydrogenation. Therefore, H₂O₂ can have both a retarding and promoting effects on the hydrogenation of α-Ti.

In order to verify the influence of the oxide film and the Pd on the surface, a model experiment was carried out on chip-like specimens. Chip-like specimens were analyzed for Ti_MP, Ti_MPwPd, Ti_MP+PW(1.8)wPd, Ti_MP+PW(3.6)wPd and Ti–(0–1.0)Pd(H₂O₂)wPd.

Ti–(0–1.0)Pd(H₂O₂), a sample immersed in H₂O₂ after mechanical polishing with no Pd deposited on the surface, was also analyzed but was not shown because of the absence of the exothermic heating peak expected during hydrogenation. The DSC profiles are shown in Figs. 8 and 9. From Fig. 8, comparing Ti_MP and Ti_MPwPd, no exothermic peak was detected in Ti_MP, while the exothermic peak was detected by Pd deposition in Ti_MPwPd, Ti_MP+PW(1.8)wPd and Ti_MP+PW(3.6)wPd. This exothermic peak is caused by the whole reaction (2) until δ-TiH₂ is formed.³⁰⁾

  

\begin{equation} \text{Ti($s$)} + \text{H$_{2}$($g$)} = \text{TiH$_{2}$($s$)} \end{equation} (2)

  

\begin{equation*} \Delta_{\text{r}}H{{}^{\circ}} = -144.35\,\text{kJ$\,$mol$^{-1}$} \end{equation*}

  

\begin{equation*} \Delta_{\text{r}}S{{}^{\circ}} = -131.728\,\text{J$\,$K$^{-1}{}\,$mol$^{-1}$} \end{equation*}

Fig. 8

DSC profiles of Ti_MP, Ti_MPwPd, Ti_MP+PW(1.8)wPd and Ti_MP+PW(3.6)wPd and OM images after DSC analysis.

Fig. 9

DSC profiles Ti-x mol%Pd(H₂O₂)_MPwPd (x = 0∼1.0) and OM images after DSC analysis.

The appearance of the exothermic peak during Ti hydrogenation is in agreement with previous reports.^12,29) With increasing immersion time in pure water, the exothermic peak shifted to the higher temperature side. In addition, the sample with the exothermic peak had crack formation after the DSC measurement. The α-Ti to δ-TiH₂ phase transformation causes a volume expansion of 24%, calculated from the lattice volumes (#44-1294 and #25-0982). No surface cracks were observed after DSC measurements on samples where no exothermic peaks were detected.

Some reported that the formation of an oxide film on the Ti surface retards the hydrogenation of Ti, and the results support their findings.^31–33) In Ti–(0–1.0)Pd(H₂O₂)wPd, an exothermic peak was detected during the cooling period in Ti–1.0Pd(H₂O₂)wPd, whereas no exothermic peak was detected in Ti(H₂O₂)wPd, Ti–0.2Pd(H₂O₂)wPd, and Ti–0.5Pd(H₂O₂)wPd. The results of SEM-EDS showed that the Ti–Pd alloys immersed in H₂O₂ had more oxygen on the surface than the untreated Ti–Pd alloys (Fig. 4). The amount of hydrogen atoms to be supplied to the metal phase is increased by the Pd present on the surface. This reaction continues during the cooling process, although at a reduced rate from a kinetic point of view. These results support that the surface oxide film retards the hydrogen supply to the metallic phase, while the Pd effectively increases the hydrogen supply to the metallic phase and promotes the dissociation reaction of H₂ molecules. The chip-like specimens were more difficult to hydrogenate and to observe the extent of a clearly exothermic reaction compared to the turnings.

The thickness of the oxide film formed on the Ti–Pd alloy immersed in H₂O₂ and the aspect ratio of the remaining metallic phase inside depend on the thickness of the sample. Hydrogenation of the 3 mm thick chip sample was confirmed by accelerated HP-DSC. The increase in the residual metallic phase, which increases the limit of soluble hydrogen, is responsible for the difference in hydrogen absorption results due to the different thicknesses. Hydrogenation could not be confirmed for the chip-shaped sample because the amount of hydrogen supplied was insufficient to form δ-TiH₂ during the operating time. This result is also in agreement with the results in support of the fact that the presence of Pd on the sample surface increases the appropriate hydrogen supply to the metallic phase.

4. Conclusion

The objective of this study was to increase the utilization of Pd in Ti–Pd alloys and to determine the effect of surface treatment by H₂O₂ immersion on the hydrogenation and surface properties of Ti–Pd alloys due to the selective dissolution of Ti in H₂O₂ and the faster decomposition of H₂O₂ by Pd. Increasing surface Pd concentration and decreasing dissolved Ti in H₂O₂ were observed with increasing Pd concentration in Ti–Pd due to fast decomposition of H₂O₂ by precipitated Pd.

The immersion of the Ti–Pd alloy in H₂O₂ resulted in an increase in the thickness of the Ti oxide layer formed on the surface. Model experiments were carried out on chip-like specimens and the results support that the Ti oxide layer has the effect of delaying the diffusion rate of hydrogen supply to the metallic phase. The presence of Pd on the surface also promotes the dissociation reaction of hydrogen gas, resulting in the faster formation of δ-TiH₂ due to the increased supply of sufficient hydrogen.

Acknowledgments

This research was financially supported by the Japan Institute of Metals and Materials, the Frontier Research Grants and the Kansai University Grant-in-Aid for progress of research in graduate course, 2023.

REFERENCES

• 1)   G.  Sandrock: J. Alloy. Compd. 293–295 (1999) 877–888. doi:10.1016/S0925-8388(99)00384-9
• 2)   B.  Sakintuna,  F.  Lamari-Darkrim and  M.  Hirscher: Int. J. Hydrogen Energ. 32 (2007) 1121–1140. doi:10.1016/j.ijhydene.2006.11.022
• 3)   L.  Schlapbach and  A.  Züttel: Nature 414 (2001) 353–358. doi:10.1038/35104634
• 4)   Y.  Kojima: Int. J. Hydrogen Energ. 44 (2019) 18179–18192. doi:10.1016/j.ijhydene.2019.05.119
• 5)   R.C.  Bowman and  B.  Fultz: MRS Bull. 27 (2002) 688–693. doi:10.1557/mrs2002.223
• 6)   C.  Corgnale,  B.  Hardy,  T.  Motyka,  R.  Zidan,  J.  Teprovich and  B.  Peters: Renew. Sustain. Energy Rev. 38 (2014) 821–833. doi:10.1016/j.rser.2014.07.049
• 7)   M.V.  Lototskyy,  V.A.  Yartys,  B.G.  Pollet and  R.C.  Bowman: Int. J. Hydrogen Energ. 39 (2014) 5818–5851. doi:10.1016/j.ijhydene.2014.01.158
• 8)   J.J.  Reilly and  R.H.  Wiswall: Inorg. Chem. 13 (1974) 218–222. doi:10.1021/ic50131a042
• 9)   G.D.  Sandrock and  P.D.  Goodell: J. Less-Common Met. 73 (1980) 161–168. doi:10.1016/0022-5088(80)90355-0
• 10)   K.  Shinzato,  Y.  Nakagawa,  S.  Yang,  S.  Isobe,  T.  Shibayama,  H.  Miyaoka and  T.  Ichikawa: ACS Appl. Energy Mater. 5 (2022) 951–957. doi:10.1021/acsaem.1c03348
• 11)   K.  Shinzato,  S.  Hamamoto,  H.  Miyaoka and  T.  Ichikawa: J. Phys. Chem. C 123 (2019) 19269–19274. doi:10.1021/acs.jpcc.9b02014
• 12)   R.  Kondo,  S.  Nakamichi,  R.  Azuma,  Y.  Takahashi,  Y.  Obora and  H.T.  Takeshita: Mater. Trans. 59 (2018) 1911–1914. doi:10.2320/matertrans.M2018240
• 13)   M.  Utsunomiya,  R.  Kondo,  T.  Oshima,  M.  Safumi,  T.  Suzuki and  Y.  Obora: Chem. Commun. 57 (2021) 5139–5142. doi:10.1039/D1CC01388B
• 14)   R.  Azuma,  S.  Nakamichi,  J.  Kimura,  H.  Yano,  H.  Kawasaki,  T.  Suzuki,  R.  Kondo,  Y.  Kanda,  K.  Shimizu,  K.  Kato and  Y.  Obora: ChemCatChem 10 (2018) 2378–2382. doi:10.1002/cctc.201800161
• 15)   Y.  Takahashi,  R.  Kondo,  M.  Utsunomiya,  T.  Suzuki,  H.T.  Takeshita and  Y.  Obora: ChemCatChem 11 (2019) 2432–2437. doi:10.1002/cctc.201900318
• 16)   K.  McNamara,  S.  Beloshapkin,  K.M.  Hossain,  M.S.  Dhoubhadel and  S.A.M.  Tofail: Appl. Surf. Sci. 535 (2021) 147621. doi:10.1016/j.apsusc.2020.147621
• 17)   J.  Ševčíková,  D.  Bártková,  M.  Goldbergová,  M.  Kuběnová,  J.  Čermák,  J.  Frenzel,  A.  Weiser and  A.  Dlouhý: Appl. Surf. Sci. 427 (2018) 434–443. doi:10.1016/j.apsusc.2017.08.235
• 18)   R.  Noyori,  M.  Aoki and  K.  Sato: Chem. Commun. (2003) 1977–1986. doi:10.1039/b303160h
• 19)   R.A.  Sheldon: Chem. Commun. (2008) 3352–3365. doi:10.1039/b803584a
• 20)   M.A.  Hasnat,  M.M.  Rahman,  S.M.  Borhanuddin,  A.  Siddiqua,  N.M.  Bahadur and  M.R.  Karim: Catal. Commun. 12 (2010) 286–291. doi:10.1016/j.catcom.2010.10.001
• 21)   W.  Dong,  G.  Kresse and  J.  Hafner: J. Mol. Catal. A Chem. 119 (1997) 69–76. doi:10.1016/S1381-1169(96)00470-0
• 22)   T.  Miyazawa,  M.  Kurihara,  S.  Ohno,  N.  Terashima,  Y.  Natsui,  H.  Kato,  Y.  Kato,  A.  Hashimoto,  T.  Kikuchi and  K.  Mase: J. Vac. Sci. Technol. A 36 (2018) 051601. doi:10.1116/1.5037023
• 23)   M.  Yan,  Z.-Q.  Huang,  Y.  Zhang and  C.-R.  Chang: Phys. Chem. Chem. Phys. 19 (2017) 2364–2371. doi:10.1039/C6CP06974F
• 24)   K.  Wang,  X.  Kong,  J.  Du,  C.  Li,  Z.  Li and  Z.  Wu: Calphad 34 (2010) 317–323. doi:10.1016/j.calphad.2010.07.001
• 25)   F.B.  Kværndrup,  M.A.J.  Somers and  T.L.  Christiansen: Metall. Mater. Trans. A 52 (2021) 4997–5003. doi:10.1007/s11661-021-06444-w
• 26)   A.V.  Bakulin,  T.I.  Spiridonova,  S.E.  Kulkova,  S.  Hocker and  S.  Schmauder: Int. J. Hydrogen Energ. 41 (2016) 9108–9116. doi:10.1016/j.ijhydene.2016.03.192
• 27)   H.  Wipf,  B.  Kappesser and  R.  Werner: J. Alloy. Compd. 310 (2000) 190–195. doi:10.1016/S0925-8388(00)00945-2
• 28)   P.  Millenbach and  M.  Givon: J. Less-Common Met. 92 (1983) 339–342. doi:10.1016/0022-5088(83)90500-3
• 29)   E.  Hadjixenophontos,  L.  Michalek,  M.  Roussel,  M.  Hirscher and  G.  Schmitz: Appl. Surf. Sci. 441 (2018) 324–330. doi:10.1016/j.apsusc.2018.02.044
• 30)  M.W. Chase, Jr., C.A. Davies, J.R. Downey, Jr., D.J. Frurip, R.A. McDonald and A.N. Syverud: NIST JANAF Thermochemical Tables, (American Institute of Physics, USA, 1998) pp. 1–1951.
• 31)   W.  Mao,  M.  Wilde,  S.  Ogura,  J.  Chen,  K.  Fukutani,  H.  Matsuzaki and  T.  Terai: J. Phys. Chem. C 122 (2018) 23026–23033. doi:10.1021/acs.jpcc.8b06893
• 32)   Y.  Zeng,  J.J.  Noël,  P.R.  Norton and  D.W.  Shoesmith: J. Electroanal. Chem. 649 (2010) 277–285. doi:10.1016/j.jelechem.2010.06.022
• 33)   Y.  Shimohata,  Y.  Hamamoto,  K.  Nishi and  K.  Tanabe: Fusion Eng. Des. 173 (2021) 112833. doi:10.1016/j.fusengdes.2021.112833