Study on hydrogen storage properties of Mg–X (X = Fe, Co, V) nano-composites co-precipitated from solution

Abstract

A systematic investigation has been performed on the hydrogen sorption properties of the Mg–X (X = Fe, Co, V) nano-composites co-precipitated from solution through an adapted Rieke method. It is found that the co-precipitated Fe, V or Co has high catalytic efficiency in enhancing the hydrogen sorption kinetics of nano-sized Mg. The Mg–V nano-composite shows faster hydrogen absorption kinetics than the Mg–Fe and Mg–Co nano-composites at lower temperatures. For instance, the hydrogen capacity within 2 h at 50 °C is 4.4 wt% for the Mg–V nano-composite, while for the Mg–Fe nano-composite it is 2.6 wt% and for the Mg–Co nano-composite it is 3.9 wt%. However, the hydrogenated Mg–Fe and Mg–Co nano-composites display significantly lower hydrogen desorption temperatures compared with the hydrogenated Mg–V nano-composite. The hydrogen desorption activation energies of the hydrogenated Mg–Fe and Mg–Co nano-composites are 118.1 and 110.1 kJ mol⁻¹ H₂, much lower than that of the Mg–V nano-composite (147.7 kJ mol⁻¹ H₂). High catalytic effectiveness of the co-precipitated Fe, Co or V depends not only on its intrinsic activity, but also on its distribution state, which may be entirely different from previous composites prepared through physical routes.

---

1. Introduction

In recent years, Mg has been considered to be an attractive hydrogen storage material because of its high hydrogen storage capacity (∼7.6 wt%), high abundance, non-toxicity, high safety and low cost. Nevertheless, the sluggish hydrogen sorption kinetics restricts the practical on-board application of Mg.^1–3 In the past decades, several approaches have been adopted to improve the hydrogen storage properties of Mg, such as (1) alloying Mg with other elements,^4–7 (2) doping with various catalysts^8–12 and (3) reducing Mg particles to nano scale.^13–16

Both experimental and theoretical investigations have shown that the hydrogen sorption kinetics and thermodynamic properties of Mg can be improved when the particle size is reduced to several nm.^17–20 High energy ball milling is the most common used synthesis technique in Mg-based nano-composites. However, ball milled products are often contaminated with traces of metals from vial and balls. Furthermore, the particle size and shape cannot be well controlled. Nanoconfinement receives increasing interests, which may lead to significantly enhanced kinetics and more favorable thermodynamic properties.¹⁷ The confined Mg/MgH₂ in carbon aerogel has shown more than 5 times faster dehydrogenation kinetics than ball-milled MgH₂.²¹ The infiltrated MgH₂/ACF has a significantly reduced reaction enthalpy of 63.8 ± 0.5 kJ mol⁻¹ H₂ compared to the bulk material.²² However, only a portion of pores in nanoporous scaffolds can be filled with MgH₂/Mg, resulting in very low hydrogen capacity for the confined composite. Recently, an adapted Rieke method has been introduced to synthesize Mg nano particles with a desired size.^14,23 For instance, Mg nanocrystals embedded in PMMA with an average diameter of 4.9 ± 2.1 nm enable both the storage of a high density of hydrogen (up to 6 wt% of Mg) and rapid kinetics (loading in <30 min at 200 °C).²³ Mg nanocrystals prepared by Norberg et al. with particle size of 25 nm could absorb 95% of the maximum capacity within 60 s at 300 °C.¹⁴ The adapted Rieke method is a promising method in preparing Mg nanoparticles with desired particle size to dramatically enhance hydrogen sorption kinetics of Mg.

Doping trace catalysts into Mg/MgH₂ is also proved to be an efficient approach to improve the hydrogen sorption properties of Mg. Almost all Mg/MgH₂ doped with catalysts are prepared through physical routes that cannot control the distribution state of catalysts. The catalytic effectiveness depends not only on the intrinsic activity of the catalyst, but also on its distribution state.²⁴ Recently, we have reported that the Mg–Ni and Mg–Ti nano-composites can be prepared though an adapted Rieke method, which is a chemical route to add catalysts into Mg/MgH₂ system.^25,26 The co-precipitated Mg–Ni nano-composite can absorb 85% of its maximum hydrogen capacity within 45 s at 125 °C, and a hydrogen capacity of 5.6 wt% can be obtained within 10 h at room temperature.²⁵ The co-precipitated Mg–Ti nano-composite can absorb 6.2 wt% of hydrogen within 2 h at room temperature.²⁶ In addition to Ni and Ti, transition metals such as Fe, Co and V are proposed as efficient catalysts towards hydrogen sorption of Mg.^27–32 Mg_1−xFe_x films with x = 0.05–0.15 prepared by electron beam co-deposition can absorb more than 3.5 wt% of hydrogen in less than 2 min, and desorb above 3.0 wt% of hydrogen in 15 min at 140 °C.²⁹ Simple mechanical milling under argon of a MgH₂–5 wt% Co composite plus 5 wt% of MWCNTs can absorb 6.5 wt% of hydrogen after 100 s and desorb in 85 min at 250 °C.³¹ MgH₂–5 at% V composite can absorb 5.6 wt% of hydrogen within 250 s at 200 °C, and desorb hydrogen completely within 200 s at 300 °C.²⁸ In this paper, we prepared Mg–X (X = Fe, Co, V) nano-composites through the adapted Rieke method. The distribution states of the transition metals (Fe, Co, V) in Mg/MgH₂ system may be entirely different from previous composites prepared through physical routes. Therefore, the hydrogen sorption properties of the co-precipitated Mg–X (X = Fe, Co, V) nano-composites are investigated and the catalytic mechanisms of Fe, Co and V are also proposed.

2. Experimental details

2.1 Sample preparation

Previous studies have reported that MgCl₂, FeCl₂, CoCl₂ and VCl₃ can be reduced in ethereal or hydrocarbon solvents using alkali metals.^33–36 In this work, the Mg–X (X = Fe, Co, V) nano-composite was co-precipitated from a homogeneous tetrahydrofuran (THF) solution including anhydrous MgCl₂, transition metal chloride (FeCl₂, CoCl₂ or VCl₃) and lithium naphthalide (LiNp) as the reducing agent. Naphthalene is used as electron carrier to speed up the reaction. The co-precipitation process may follow reactions as:

MgCl₂ + 2Li → Mg(0) + 2LiCl, (1)

FeCl₂ + 2Li → Fe(0) + 2LiCl, (2)

CoCl₂ + 2Li → Co(0) + 2LiCl, (3)

VCl₃ + 3Li → V(0) + 3LiCl. (4)

All reactions were performed in an argon glove-box to avoid the influence of oxygen and moisture. Both the oxygen and water vapour levels inside the glove-box were kept below 1 ppm. In raw materials, X (X = Fe, Co or V)–Mg was kept in weight ratio of 1 : 9. LiNp/THF solution (blackish green) was prepared by stirring vigorously the mixture of naphthalene (12.000 g) and lithium (0.650 g) in the freshly distilled THF (100 ml) at room temperature. A little excess of LiNp was necessary to achieve complete reaction. Anhydrous MgCl₂ (3.808 g) and transition metal chloride (0.245 g FeCl₂, 0.235 g CoCl₂ or 0.334 g VCl₃) were dissolved in the freshly distilled THF (400 ml), stirring vigorously at 60–70 °C to form a transparent solution, and then which was cooled down to room temperature. The mixed solution was dropped into LiNp/THF solution, stirring vigorously for 24 h at room temperature. At last, the resultant product was isolated by centrifugation, washed with freshly distilled THF three times, and dried by vacuum pumping to remove the residual THF.

2.2 Characterization

Structure characterizations of as-prepared samples were performed by X-ray diffraction (XRD) using an apparatus (D/max 2550VL/PCX) equipped with a Cu K_α radiation source. In order to isolate the sample from air, the sample powder was put on a designed XRD holder and sealed with scotch tape in the glove-box. The morphology and microstructure of as-prepared samples were observed by using a JEM-2100F transmission electron microscopy. The weight content of X (X = Fe, Co or V) in the nano-composite was analyzed by energy dispersive spectrometer (EDS). As the co-precipitated Mg–X (X = Fe, Co, V) nano-composites are extremely active, activation process is not necessary for all samples before measurements of hydrogen storage properties. The hydrogen sorption behaviors of as-prepared samples were examined using a Sievert apparatus (Type PCT-2, Shanghai Institute of Microsystem and Information Technology). The auto pressure–composition–temperature (P–C–T) measurements were examined at various temperatures of 250, 275, 300, 325, 350 and 375 °C, respectively. Maximum hydrogen pressure, minimum hydrogen pressure, and time interval for pressure variation (<0.0002 MPa) are set as 4.0 MPa, 0.019 MPa and 20 s, respectively. Hydrogen absorption measurements were examined under a hydrogen pressure of 3 MPa at 50, 75, 100, 125 and 150 °C, respectively. Desorption performance of the hydrogenated nano-composite at 150 °C under a hydrogen pressure of 3 MPa for 2 h was examined by using diffraction scanning calorimetry (DSC, Netzsch STA 449F3) at heating rates of 3, 5 and 10 °C min⁻¹ under Ar gas flow, respectively.

3. Results and discussions

3.1 Microstructure characterization of the Mg–X (X = Fe, Co, V) nano-composites

Fig. 1–3 display the XRD patterns of the Mg–X (X = Fe, Co, V) nano-composites in varying states, respectively. From Fig. 1a, 2a and 3a, almost all broad diffraction peaks of the Mg–X (X = Fe, Co, V) nano-composites match for hexagonal Mg phase, while the characteristic peaks of Fe, Co or V are invisible. A small amount of MgO can be detected, possibly due to oxidation occurred during sample preservation and preparation for analysis. Considering the low solubility of Fe, Co or V in Mg and immiscibility of the component, the characteristic peaks absence of Fe, Co or V is likely due to the fact that Fe, Co or V synthetized from solution is in either amorphous or nano-crystalline. The similar phenomenon is found in Mg–Fe and Mg–V films prepared by electron beam co-deposition.^29,30

Fig. 1 XRD patterns for the Mg–Fe nano-composite in varying states. (a) As prepared, (b and c) hydrogenated at 150 °C and 350 °C for 2 h, and (d) dehydrogenated at 350 °C for 2 h.

Fig. 2 XRD patterns for the Mg–Co nano-composite in varying states. (a) As prepared, (b and c) hydrogenated at 150 °C and 350 °C for 2 h, and (d) dehydrogenated at 350 °C for 2 h.

Fig. 3 XRD patterns for the Mg–V nano-composite in varying states. (a) As prepared, (b) hydrogenated at 150 °C 2 h, and (c) dehydrogenated at 350 °C for 2 h.

As can be seen in Fig. 1b, when the Mg–Fe nano-composite is hydrogenated at 150 °C for 2 h, the hydrogenated Mg–Fe nano-composite is comprised of tetragonal β-MgH₂ as the majority phase, orthogonal γ-MgH₂ and MgO. The phase transformation from β-MgH₂ to γ-MgH₂ is also found in the Mg–Ni and Mg–Ti nano-composites.^25,26 Due to the addition of Fe/FeCl₂, high density defects are likely to form in the nano-composite, and still remain after hydrogenation at 150 °C, leading to the formation of γ-MgH₂ phase. The absence of Fe phase at 150 °C suggests that Fe is still in either amorphous or nano-crystalline states without detectable diffraction intensity. When the hydrogenation is performed at 350 °C, Mg₂FeH₆ and Fe can be found in addition to β-MgH₂ and MgO in Fig. 1c. The appearance of Fe phase at 350 °C suggests that Fe has grown into larger crystalline which can generate detectable diffraction intensity. Polanski et al. have found that MgH₂ forms prior to Mg₂FeH₆ which forms directly by a reaction between the solid MgH₂ and Fe in a temperature range of 350–500 °C under high hydrogen pressure.³⁷ In this work, the formation of Mg₂FeH₆ can be described by the following formula:

2MgH₂ + Fe + H₂ → Mg₂FeH₆. (5)

When the hydrogenated Mg–Fe nano-composite desorbs hydrogen at 350 °C for 2 h, the XRD pattern is shown in Fig. 1d. The dehydrogenated composite is comprised of Mg, Fe and MgO. This is consistent with the result reported by Polanski et al.³⁷ Mg₂FeH₆ decomposes directly at about 340 °C according to the following formula:

Mg₂FeH₆ → 2MgH₂ + Fe + H₂. (6)

As can be seen in Fig. 2b, the hydrogenated Mg–Co nano-composite at 150 °C is comprised β-MgH₂ as the majority phase, along with a small amount of γ-MgH₂ and MgO. This result suggests that the addition of Co/CoCl₂ is also likely to lead the formation of γ-MgH₂ and the co-precipitated Co is still in either amorphous or nano-crystalline states. In addition to β-MgH₂ and MgO, Mg₂CoH₅ LT (LT-low temperature) can be formed in the hydrogenated Mg–Co nano-composite at 350 °C as shown in Fig. 2c. Shao et al. have reported that nanostructured Mg₂CoH₅ can form at 350° in 4 MPa hydrogen pressure from Mg and Co nanoparticles.³⁸ Due to the formation of MgH₂ prior to Mg₂CoH₅, the formation of Mg₂CoH₅ can be described by the following formula:

4MgH₂ + 2Co + H₂ → 2Mg₂CoH₅. (7)

When the hydrogenated Mg–Co nano-composite desorbs hydrogen at 350 °C for 2 h, Mg, Mg₂Co and MgO can be observed in Fig. 2d. Norek et al. have found that Mg₂CoH₅ HT decomposes directly into elemental Mg and Co at a temperature of approximately 350 °C, while the Mg–Co intermetallic compound is formed at temperature of just above 400 °C.³⁹ In this work, the decomposition of Mg₂CoH₅ at 350 °C can be described by the following formulas:

2Mg₂CoH₅ → 4Mg + 2Co + 5H₂, (8)

2Mg + Co → Mg₂Co (9)

The formation of Mg₂Co at 350 °C suggests that the atomic motilities of Mg and Co increase with the particle size reduction of Mg and Co, resulting in lower reaction temperature.

The XRD pattern of the hydrogenated Mg–V nano-composite at 150 °C for 2 h is shown in Fig. 3b. The hydrogenated Mg–V nano-composite is comprised β-MgH₂ as the majority phase, along with a small amount of V₂H and MgO. Compared with the hydrogenated Mg–Fe and Mg–Co nano-composites, γ-MgH₂ is absent. This is possible due to lower density defects formed in the Mg–V nano-composite, which cannot generate the phase transformation from β-MgH₂ to γ-MgH₂. The formation of V₂H shows that V can be partly hydrogenated at 150 °C. The formation enthalpy of V₂H is reported to be −83 kJ mol⁻¹ H₂.⁴⁰ Zahiri et al. have pointed out that the formation of V₂H is expected under temperature/pressure conditions utilized for Mg. And it is reasonable for the formation of V₂H phase at low temperatures.⁴¹ When the hydrogenated Mg–V nano-composite desorbs hydrogen at 350 °C for 2 h, Mg, V and MgO can be detected in the XRD pattern of Fig. 3c. This result suggests that V₂H decomposes during dehydrogenation process.

Scanning transmission electron microscope (STEM) micrographs of the Mg–X (X = Fe, Co, V) nano-composites are shown in Fig. 4a, 5a and 6a, respectively. It is observed that the Mg–Fe or Mg–Co nano-composite is composed of irregular shaped particles aggregating together with their particle sizes ranging from 10 to 20 nm. In contrast, plate-shaped Mg–V particles with the thickness ranging from 10 to 20 nm are stacked together, forming a shape like “chicken claws” structure. The morphology of the Mg–V nano-composite is similar with the reduced pure Mg.²⁵ Above results indicate that the co-precipitated Fe or V can dramatically change the morphology of Mg particles and reduce the particle size of Mg. The corresponding selected area electron diffraction (SAED) patterns of the Mg–X (X = Fe, Co, V) nano-composites have been shown in Fig. 4b, 5b and 6b, respectively. The diffraction rings or points can be indexed with Mg and MgO phase. The formation of MgO phase is likely due to oxidation occurred during TEM sample preparation. This result is consistent with XRD analysis. EDS elemental mapping is performed to qualitatively evaluate the distribution of Fe, Co or V in Mg particles. For the Mg–Fe nano-composite, EDS elemental maps of Mg and Fe are showed in Fig. 4c and d, respectively. Fig. 5c and d displays EDS elemental maps of Mg and Co for the Mg–Co nano-composite. EDS elemental maps of Mg and V for the Mg–V nano-composite are showed in Fig. 6c and d. It is observed that the co-precipitated Fe, V or Co is homogeneously distributed on the surface or inside Mg particles. The special distribution state for the co-precipitated Fe, V or Co may be entirely different from previous composites prepared through physical routes. The actual Fe, Co or V weight content in the corresponding nano-composite is determined to be around 2.9, 5.8 and 2.2 wt% by EDS, respectively. The TEM micrographs of the Mg–X (X = Fe, Co, V) nano-composites hydrogenated at 150 °C for 2 h are shown in Fig. 7a–c, respectively. In the corresponding SAED patterns showed in the insets of Fig. 7a–c, the diffraction rings or points can be indexed with Mg and MgO phase. However, XRD analysis has shown that hydrogenated Mg–X (X = Fe, Co, V) nano-composites consist of β-MgH₂ phase as the majority phase. This is attributed to the hydrogen release from nano-sized MgH₂ during the TEM measurement under a high vacuum condition and the exposure to the electron beam.⁴² As can be seen, the morphology of hydrogenated nano-composite is similar to the as-prepared nano-composite. The hydrogenated Mg–Fe or Mg–Co nano-composites are mainly composed of particles with irregular shape. The particle size of the hydrogenated Mg–Fe nano-composite is slightly smaller than that of the hydrogenated Mg–Co nano-composite, ranging from 10 to 20 nm. The hydrogenated Mg–V nano-composite is composed of fine plates with their thickness ranging from 10 to 30 nm.

Fig. 4 (a) STEM micrograph and (b) SAED pattern of the Mg–Fe nano-composite, and along with (c and d) EDS elemental maps of Mg and Fe.

Fig. 5 (a) STEM micrograph and (b) SAED pattern of the Mg–Co nano-composite, and along with (c and d) EDS elemental maps of Mg and Co.

Fig. 6 (a) STEM micrograph and (b) SAED pattern of the Mg–V nano-composite, and along with (c and d) EDS elemental maps of Mg and V.

Fig. 7 TEM micrographs of (a) Mg–Fe nano-composite, (b) Mg–Co nano-composite and (c) Mg–V nano-composite hydrogenated at 50 °C, and insets in (a–c) show the corresponding SAED patterns.

3.2 Hydrogen sorption properties of the Mg–X (X = Fe, Co, V) nano-composites

The hydrogen storage thermodynamic and kinetic properties of the Mg–X (X = Fe, Co and V) nano-composites are evaluated by PCT technique. The PCT curves of the Mg–X (X = Fe, Co, V) nano-composites measured at 250, 275, 300, 325, 350, and 375 °C in a hydrogen pressure range from 0.02 to 4 MPa are showed in Fig. 8a–c, respectively. As can be seen in Fig. 8a and c, single flat plateaus owing to the formation and decomposition of MgH₂ phase are present across the whole range of the hydrogen content at each temperature for both of the Mg–Fe and Mg–V nano-composites. In contrast, PCT desorption curves of the Mg–Co nano-composite measured at 350 °C and 375 °C clearly consist of two plateaux: the lower one is owing to the decomposition of Mg₂CoH₅ and the higher associates with the decomposition of MgH₂ (ref. 7) as shown in Fig. 8b. For the Mg–Fe and Mg–V nano-composites, the maximum hydrogen absorption capacity at the temperatures ranging from 250 to 375 °C is 6.2–6.4 wt%, while the maximum hydrogen absorption capacity of the Mg–Co nano-composite at the same temperature range is 5.8–6.0 wt%. This result is attributed to the high weight content of Co in the nano-composite. At lower temperatures, it is obvious that the PCT desorption profiles of the Mg–X (X = Fe, Co and V) nano-composites have more data points than those measured at high temperatures, which is due to the slow hydrogen desorption kinetics of the nano-composite at low temperatures. Such kinetic artifacts, together with the thermodynamic effects, contribute to the relatively large hysteresis observed in the PCT curves measured at low temperatures. Based on the absorption/desorption plateaus obtained from PCT curves, the corresponding van't Hoff plots (ln P versus T⁻¹) for the Mg–X (X = Fe, Co, V) nano-composites are used to estimate hydrogenation and dehydrogenation enthalpies and entropies, as shown in the insets of Fig. 8a–c. For the Mg–Fe nano-composite, the hydrogenation enthalpy (ΔH_ab) and entropy (ΔS_ab) are determined to be −74.4 kJ mol⁻¹ H₂ and 135.3 J mol⁻¹ K⁻¹ H₂ while the dehydrogenation enthalpy (ΔH_de) and entropy (ΔS_de) are 77.3 kJ mol⁻¹ H₂ and 139.7 J mol⁻¹ K⁻¹ H₂, respectively. For the Mg–Co nano-composite, ΔH_ab and ΔS_ab are −74.3 kJ mol⁻¹ H₂ and 136.8 J mol⁻¹ K⁻¹ H₂ for hydrogen absorption, while ΔH_de and ΔS_de are 77.3 kJ mol⁻¹ H₂ and 140.3 J mol⁻¹ K⁻¹ H₂ for hydrogen desorption. For the Mg–V nano-composite, ΔH_ab and ΔS_ab are −77.5 kJ mol⁻¹ H₂ and 139.6 J mol⁻¹ K⁻¹ H₂ for hydrogen absorption, while ΔH_de and ΔS_de are 77.4 kJ mol⁻¹ H₂ and 138.5 J mol⁻¹ K⁻¹ H₂ for hydrogen desorption. The values of enthalpies for the Mg–X (Fe, Co, V) nano-composites are quite close to those of the standard values for Mg (±74.7 kJ mol⁻¹ H₂)⁴³ and the reduced pure Mg (ΔH_ab = 74.8 kJ mol⁻¹ H₂, ΔH_de = 76.3 kJ mol⁻¹ H₂).²⁵ In addition, the values of entropies are also close to the classical value of 135 J mol⁻¹ K⁻¹ H₂ given in Sandia National Lab database. The change in hydrogen sorption thermodynamics of Mg with regard to its particle size, especially in nm scale, has been reported in some previous works.^44–48 However, a widely acceptable conclusion on the threshold of particle size for such a “nano size effect” is still not drawn yet. Above results indicate that the hydrogenation and dehydrogenation thermodynamics of the Mg–X (X = Fe, Co, V) nano-composites are not changed by the addition of Fe, Co or V.

Fig. 8 PCT curves of (a) Mg–Fe nano-composite, (b) Mg–Co nano-composite and (c) Mg–V nano-composite measured at different temperatures, and insets in (a–c) show the corresponding van't Hoff plots. Ab-absorption, De-desorption.

Fig. 9a, c and e presents the hydrogen absorption curves of the Mg–X (X = Fe, Co, V) nano-composites measured at the temperatures ranging from 50 to 150 °C under 3 MPa hydrogen pressure, respectively. It has been reported that pure Mg prepared through the adapted Rieke method absorbs not more than 2.6 wt% within 2 h at 125 °C.²⁵ Compared to the reduced pure Mg, a significant improvement on hydrogen absorption kinetics is observed at relatively low temperatures for the Mg–X (X = Fe, Co, V) nano-composites. For instance, the Mg–Fe nano-composite can absorb hydrogen of 2.6 wt% within 2 h at 50 °C, and 85% of its maximum hydrogen capacity can be obtained within 1008 s at 150 °C as shown in Fig. 9a. For the Mg–Co nano-composite, 3.9 wt% of hydrogen can be absorbed at 50 °C within 2 h, while more than 2285 s is needed to obtain 85% of its maximum hydrogen capacity at 150 °C, as shown in Fig. 9c. In contrast to the Mg–Fe and Mg–Co nano-composites, the Mg–V nano-composite shows much better hydrogen absorption kinetics at low temperatures. As can be seen in Fig. 9e, the Mg–V nano-composite can absorb 4.4 wt% of hydrogen after 2 h at 50 °C, and 85% of its maximum hydrogen capacity can be obtained within 196 s at 150 °C. Such excellent hydrogen absorption kinetics at low temperature for the Mg–X (X = Fe, Co, V) nano-composites indicate that the addition of Fe, Co or V has drastic effects on accelerating the hydrogenation rate of Mg. The improved hydrogen absorption kinetics at low temperatures can be further investigated by calculating the activation energy (E_a) of the hydrogenation reaction. It has been established that a number of steps are involved for hydrogen absorption in metals: H₂ adsorption onto the surface, H₂ dissociation, H chemisorption, H migration from surface to bulk, H diffusion and nucleation of hydride, and growth of hydride phase.⁴⁹ The E_a summarizes overall energy barrier of all steps during the hydrogen absorption process. The hydrogen absorption data can be analyzed by applying the Johnson–Mehl–Avrami–Kolmogorov (JMAK) model, which is one of the most effective ways to describe the nucleation and growth model, and the linear equation is described as follows:

ln[−ln(1 − α)] = η ln k + η ln t, (10)

where α is the fraction of Mg transformed into MgH₂ at time t, k an effective kinetics parameter and η is the Avrami exponent or reaction order. Based on the experimental data, the linear fitting can be done by plotting ln[−ln(1 − α)] vs. ln t at different temperatures. Then the values of k and η can be obtained by calculating the values of η (the slope) and the η ln k (intercept) of that straight line at each temperature. E_a for the absorption is usually evaluated according to Arrhenius equation:

k = Aexp(−E_a/RT), (11)

where A is a temperature independent coefficient, R the universal gas constant, and T is the absolute temperature. The Arrhenius type linear fittings of ln k vs. 1000/T⁻¹ for the Mg–X (X = Fe, Co, V) nano-composites are drawn in Fig. 9b, d and f, respectively. The E_a values are therefore estimated to be 38.9 kJ mol⁻¹ H₂ for the Mg–Fe nano-composite, 45.4 kJ mol⁻¹ H₂ for the Mg–V nano-composite and 30.3 kJ mol⁻¹ H₂ for Mg–Co nano-composite, which are much lower than those of Mg Nanocrystals with particle size of 25 nm (122 kJ mol⁻¹ H₂),¹⁴ Mg–4Fe at% thin film deposited by electron beam evaporation (56 ± 3 kJ mol⁻¹ H₂)⁵⁰ and Mg–V nanoparticles prepared by hydrogen plasma–metal reaction (71.2 kJ mol⁻¹ H₂).⁵¹ It is worth noting that, though the Mg–V nano-composite shows the fastest hydrogenation rate, the E_a of the Mg–V nano-composite is slightly larger than that of the Mg–Fe or Mg–Co nano-composite. The similar phenomenon was also observed in the Mg–TM–La (TM = Fe, Ti, Ni) composite powders,⁵² which is a result of different values of A in different hydrogen storage systems. The improved hydrogen absorption kinetics of the Mg–X (X = Fe, Co, V) nano-composites are mainly attributed to two factors: nano size effect and catalytic effects from the co-precipitated Fe, Co or V. Nanosize particles can shorten the diffusion length of hydrogen and increase active surface area to provide more nucleation sites for the formation of MgH₂.⁵³ It has been suggested that transition metals are good catalysts to enhance the dissociation of hydrogen molecule, due to their electron configurations.⁵⁴ Zheng et al. have reported that the Fe and V layers percolating the Mg act as “highway” to deliver hydrogen throughout the bulk, due to the bcc transition metals with faster hydrogen diffusion rate at low temperatures.^29,55 For the Mg–V nano-composite, V₂H can act as a hydrogen pump.⁵⁶ The interfaces between Mg and X (X = Fe, Co or V) can act as nucleation sites for MgH₂. For the co-precipitated transition metal, the high catalytic effectiveness is attributed to not only its intrinsic activity but also the special distribution state. All of these positive factors contribute to the excellent hydrogen absorption kinetics of the Mg–X (X = Fe, Co, V) nano-composites.

Fig. 9 Hydrogen absorption curves and the corresponding ln k–1000/T⁻¹ plot of (a and b) Mg–Fe nano-composite, (c and d) Mg–Co nano-composite and (e and f) Mg–V nano-composite measured at different temperatures under 3 MPa hydrogen pressure.

In order to investigate the desorption performance of the hydrogenated Mg–X nano-composites, DSC curves of the hydrogenated Mg–X (X = Fe, Co, V) nano-composites at heating rates of 3, 5, 10 °C min⁻¹ are showed in Fig. 10a–c, respectively. As can be seen, there is a very broad endothermic peak at all heating rates for the hydrogenated Mg–Fe or Mg–Co nano-composite, which is attributed to hydrogen desorption from both β-MgH₂ and γ-MgH₂. As shown in Fig. 10a, peak temperatures of DSC curves for the hydrogenated Mg–Fe nano-composite at heating rates of 3, 5, 10 °C min⁻¹ are determined to be 276.1, 287.7, 300.9 °C, respectively. And the onset dehydrogenation temperature at heating rate 3 °C min⁻¹ is 246.5 °C. For the hydrogenated Mg–Co nano-composite, peak temperatures and onset dehydrogenation temperatures of DSC curves at different heating rates are close to those of the hydrogenated Mg–Fe nano-composite, as shown in Fig. 10b. In contrast, DSC curve of the hydrogenated Mg–V nano-composite at all heating rates shows a sharp endothermic peak corresponding to the dehydrogenation of β-MgH₂. Due to the relatively low V content in the Mg–V nano-composite (2.2 wt%), the desorption peak of V₂H phase is hardly revealed in the DSC profile of the hydrogenated Mg–V nano composite. As seen in Fig. 10c, peak temperatures of DSC curves for the hydrogenated Mg–V nano-composite at heating rates of 3, 5, 10 °C min⁻¹ are determined to be 305.4, 314.9, 327.5 °C, respectively. The onset dehydrogenation temperature at a heating rate of 3 °C min⁻¹ is 278.2 °C. For the hydrogenated pure Mg, the onset dehydrogenation temperature at heating rate 3 °C min⁻¹ is about 311.1 °C.²⁵ Compared with the hydrogenated pure Mg, the hydrogenated Mg–X (X = Fe, Co, V) nano-composites show superior hydrogen desorption properties, due to the addition of Fe, Co or V in Mg system. In addition, the hydrogen desorption properties of the hydrogenated Mg–Fe and Mg–Co nano-composites are better than that of the hydrogenated Mg–V nano-composite. The improved desorption property can be further understood by calculating the activation energy of the dehydrogenation reaction of the hydrogenated Mg–X nano-composite. The activation energy, E_d, for the hydrogen desorption mechanism of the Mg–X nano-composites is obtained from DSC measurements at different heating rates by the Kissinger equation:

ln(β/T_p²) = A − E_d/(RT_p), (12)

where β is the heating rate, T_p is the peak temperature, and A is a linear constant. The E_d values for the Mg–X nano-composites can be estimated from the slope of the plot of ln(β/T_p²) vs. 1000/T_p as shown in the insets of Fig. 10a–c. The E_d value for dehydrogenation of the hydrogenated Mg–Fe nano-composite is determined to be 118.1 kJ mol⁻¹ H₂, for the hydrogenated Mg–Co nano-composite that is 110.1 kJ mol⁻¹ H₂, while for the hydrogenated Mg–V nano-composite that is 147.7 kJ mol⁻¹ H₂. The E_d value for the hydrogenated Mg–Fe or Mg–Co nano-composite is much lower than that of the Mg–V nano-composite and the hydrogenated pure Mg (E_d = 147.4 kJ mol⁻¹ H₂).²⁵ In this work, nanosize effect alone cannot explain the improved hydrogen desorption property for the hydrogenated Mg–X nano-composite. The co-precipitated Fe, V or Co is thus responsible for the improvement in the hydrogen desorption kinetics. Giusepponi et al. have pointed out that the higher coordination of Fe compared to Mg in the hydride is able to destabilize the crystalline structure and raise the probability that H atoms start diffusion toward the interface.⁵⁷ The nucleation of Mg phase occurs in the interface of MgH₂ and co-precipitated X (X = Fe, Co or V), thus leading to a significant improvement of the hydrogen desorption kinetics of MgH₂. In addition, γ-MgH₂ phase formed in the hydrogenated Mg–Fe or Mg–Co nano-composites can destabilize β-MgH₂ (ref. 58) and thereby improve the hydrogen desorption kinetics. Therefore, the co-precipitated Fe, Co or V has high catalytic efficiency in enhancing the hydrogen desorption kinetics of nano-sized Mg.

Fig. 10 DSC curves of (a) hydrogenated Mg–Fe nano-composite, (b) hydrogenated Mg–Co nano-composite and (c) hydrogenated Mg–V nano-composite at heating rates of 3, 5, 10 °C min⁻¹, and insets in (a–c) show the corresponding ln(β/T_p²)–1000/T_p plots.

4. Conclusions

In the present work, the Mg–X (X = Fe, Co, V) nano-composites were co-precipitated through an adapted Rieke method. The microstructure features and hydrogen sorption properties were carefully investigated. The main conclusions are as follows:

(1) STEM and TEM observations reveal that the Mg–Fe or Mg–Co nano-composite is composed of irregular shaped particles aggregating together with their size ranging from 10 to 20 nm, while plate-shaped Mg–V particles with the thickness ranging from 10 to 20 nm are stacked together, forming a shape like “chicken claws” structure.

(2) The Mg–V nano-composite show faster hydrogen absorption kinetics than the Mg–Fe and Mg–Co nano-composites at lower temperatures. For instance, a hydrogen capacity within 2 h at 50 °C is 4.4 wt% for the Mg–V nano-composite, while for the Mg–Fe nano-composite that is 2.6 wt% and for the Mg–Co nano-composite that is 3.9 wt%.

(3) The hydrogen desorption temperature of the hydrogenated Mg–Fe or Mg–Co nano-composite is lower than that of the hydrogenated Mg–V nano-composite. The E_d values of Mg–Fe and Mg–Co nano-composites are 118.1 and 110.1 kJ mol⁻¹ H₂, much lower than that of the Mg–V nano-composite (147.7 kJ mol⁻¹ H₂).

(4) Based on the PCT measurements, the hydrogenation and dehydrogenation enthalpies of Mg–X (X = Fe, Co, V) nano-composites were close to the standard values for Mg (±74.7 kJ mol⁻¹ H₂), indicating that thermodynamic properties of nano-composites are not changed by the addition of Fe, Co or V.

(5) High catalytic effectiveness of the co-precipitated Fe, Co or V depends not only on its intrinsic activity, but also on its distribution state.

Acknowledgements

The authors would like to thank the support from the Science and Technology Committee of Shanghai under nos 10JC1407700 and 11ZR1417600, and ‘Pujiang’ project (no. 11PJ1406000). This work is partly supported by Research Fund for the Doctoral Program of Higher Education of China (no. 20100073120007) and from the Shanghai Education Commission (no. 12ZZ017).

Notes and references

1. I. Jain, C. Lal and A. Jain, Int. J. Hydrogen Energy, 2010, 35, 5133–5144.
2. B. Sakintuna, F. Lamari-Darkrim and M. Hirscher, Int. J. Hydrogen Energy, 2007, 32, 1121–1140.
3. L. Schlapbach and A. Züttel, Nature, 2001, 414, 353–358.
4. L. Zaluski, A. Zaluska and J. Ström-Olsen, J. Alloys Compd., 1995, 217, 245–249.
5. H. Shao, Y. Wang, H. Xu and X. Li, J. Solid State Chem., 2005, 178, 2211–2217.
6. X. Zhang, R. Yang, J. Qu, W. Zhao, L. Xie, W. Tian and X. Li, Nanotechnology, 2010, 21, 095706.
7. J. Chen, H. Takeshita, D. Chartouni, N. Kuriyama and T. Sakai, J. Mater. Sci., 2001, 36, 5829–5834.
8. W. Oelerich, T. Klassen and R. Bormann, J. Alloys Compd., 2001, 315, 237–242.
9. N. Hanada, T. Ichikawa and H. Fujii, J. Phys. Chem. B, 2005, 109, 7188–7194.
10. T. Sadhasivam, M. Sterlin Leo Hudson, S. K. Pandey, A. Bhatnagar, M. K. Singh, K. Gurunathan and O. Srivastava, Int. J. Hydrogen Energy, 2013, 38, 7353–7362.
11. T. Ma, S. Isobe, Y. Wang, N. Hashimoto and S. Ohnuki, J. Phys. Chem. C, 2013, 117, 10302–10307.
12. M. Lototskyy, J. Sibanyoni, R. Denys, M. Williams, B. Pollet and V. Yartys, Carbon, 2013, 57, 146–160.
13. Y. Liu, J. Zou, X. Zeng, X. Wu, H. Tian, W. Ding, J. Wang and A. Walter, Int. J. Hydrogen Energy, 2013, 38, 5302–5308.
14. N. S. Norberg, T. S. Arthur, S. J. Fredrick and A. L. Prieto, J. Am. Chem. Soc., 2011, 133, 10679–10681.
15. H. Shao, Y. Wang, H. Xu and X. Li, Mater. Sci. Eng., B, 2004, 110, 221–226.
16. J. Huot, G. Liang, S. Boily, A. Van Neste and R. Schulz, J. Alloys Compd., 1999, 293, 495–500.
17. T. K. Nielsen, F. Besenbacher and T. R. Jensen, Nanoscale, 2011, 3, 2086–2098.
18. R. W. Wagemans, J. H. van Lenthe, P. E. de Jongh, A. J. van Dillen and K. P. de Jong, J. Am. Chem. Soc., 2005, 127, 16675–16680.
19. M. Paskevicius, D. A. Sheppard and C. E. Buckley, J. Am. Chem. Soc., 2010, 132, 5077–5083.
20. V. Berube, G. Radtke, M. Dresselhaus and G. Chen, Int. J. Energy Res., 2007, 31, 637–663.
21. S. Zhang, A. F. Gross, S. L. Van Atta, M. Lopez, P. Liu, C. C. Ahn, J. J. Vajo and C. M. Jensen, Nanotechnology, 2009, 20, 204027.
22. Z. Zhao-Karger, J. Hu, A. Roth, D. Wang, C. Kübel, W. Lohstroh and M. Fichtner, Chem. Commun., 2010, 46, 8353–8355.
23. K.-J. Jeon, H. R. Moon, A. M. Ruminski, B. Jiang, C. Kisielowski, R. Bardhan and J. J. Urban, Nat. Mater., 2011, 10, 286–290.
24. P. Wang, X.-D. Kang and H.-M. Cheng, J. Phys. Chem. B, 2005, 109, 20131–20136.
25. Y. Liu, J. Zou, X. Zeng, X. Wu, D. Li and W. Ding, J. Phys. Chem. C, 2014, 118, 18401–18411.
26. Y. Liu, J. Zou, X. Zeng and W. Ding, RSC Adv., 2014, 4, 42764–42771.
27. G. Liang, J. Huot, S. Boily, A. Van Neste and R. Schulz, J. Alloys Compd., 1999, 292, 247–252.
28. G. Liang, J. Huot, S. Boily, A. Van Neste and R. Schulz, J. Alloys Compd., 1999, 291, 295–299.
29. S. Zheng, K. Wang, V. P. Oleshko and L. A. Bendersky, J. Phys. Chem. C, 2012, 116, 21277–21284.
30. S. Zheng, Z.-P. Li and L. A. Bendersky, ACS Appl. Mater. Interfaces, 2013, 5, 6968–6974.
31. M. G. Verón, H. Troiani and F. C. Gennari, Carbon, 2011, 49, 2413–2423.
32. J. Bystrzycki, T. Czujko and R. Varin, J. Alloys Compd., 2005, 404, 507–510.
33. R. D. Rieke and S. E. Bales, J. Am. Chem. Soc., 1974, 96, 1775–1781.
34. A. V. Kavaliunas, A. Taylor and R. D. Rieke, Organometallics, 1983, 2, 377–383.
35. J. L. Dye, A. S. Ellaboudy and K.-L. Tsai, Alloy of AuZn AuCu or ZnCu, Google Patents no. 6350408, 2002.
36. G. L. Rochfort and R. D. Rieke, Inorg. Chem., 1986, 25, 348–355.
37. M. Polanski, T. K. Nielsen, Y. Cerenius, J. Bystrzycki and T. R. Jensen, Int. J. Hydrogen Energy, 2010, 35, 3578–3582.
38. H. Shao, H. Xu, Y. Wang and X. Li, J. Solid State Chem., 2004, 177, 3626–3632.
39. M. Norek, T. K. Nielsen, M. Polanski, I. Kunce, T. Płociński, L. Jaroszewicz, Y. Cerenius, T. R. Jensen and J. Bystrzycki, Int. J. Hydrogen Energy, 2011, 36, 10760–10770.
40. M. Gonzalez-Silveira, R. Gremaud, A. Baldi, H. Schreuders, B. Dam and R. Griessen, Int. J. Hydrogen Energy, 2010, 35, 6959–6970.
41. B. Zahiri, M. Danaie, X. Tan, B. S. Amirkhiz, G. A. Botton and D. Mitlin, J. Phys. Chem. C, 2012, 116, 3188–3199.
42. D. A. Sheppard, M. Paskevicius and C. E. Buckley, J. Alloys Compd., 2010, 492, L72–L74.
43. K.-F. Aguey-Zinsou and J.-R. Ares-Fernández, Energy Environ. Sci., 2010, 3, 526–543.
44. W. Liu and K.-F. Aguey-Zinsou, J. Mater. Chem. A, 2014, 2, 9718–9726.
45. E. J. Setijadi, C. Boyer and K.-F. Aguey-Zinsou, Phys. Chem. Chem. Phys., 2012, 14, 11386–11397.
46. P. Vajeeston, P. Ravindran, M. Fichtner and H. Fjellvag, J. Phys. Chem. C, 2012, 116, 18965–18972.
47. K. C. Kim, B. Dai, J. K. Johnson and D. S. Sholl, Nanotechnology, 2009, 20, 204001.
48. V. A. Vons, A. Anastasopol, W. J. Legerstee, F. M. Mulder, S. W. H. Eijt and A. Schmidt-Ott, Acta Mater., 2011, 50, 3070–3080.
49. J. Fernandez and C. Sanchez, J. Alloys Compd., 2002, 340, 189–198.
50. Z. Tan, C. Chiu, E. J. Heilweil and L. A. Bendersky, Int. J. Hydrogen Energy, 2011, 36, 9702–9713.
51. T. Liu, T. Zhang, C. Qin, M. Zhu and X. Li, J. Power Sources, 2011, 196, 9599–9604.
52. J. Zou, H. Guo, X. Zeng, S. Zhou, X. Chen and W. Ding, Int. J. Hydrogen Energy, 2013, 38, 8852–8862.
53. X. Yao, Z. Zhu, H. Cheng and G. Lu, J. Mater. Res., 2008, 23, 336–340.
54. G. J. Kubas, Science, 2006, 314, 1096–1097.
55. S.-A. Jin, J.-H. Shim, Y. W. Cho and K.-W. Yi, J. Power Sources, 2007, 172, 859–862.
56. W. Oelerich, T. Klassen and R. Bormann, J. Alloys Compd., 2001, 322, L5–L9.
57. S. Giusepponi and M. Celino, Int. J. Hydrogen Energy, 2013, 38, 15254–15263.
58. S. X. Zhou, W. X. Ran, M. J. Yang, D. X. Wang, G. Q. Chen, Y. Zhang, Z. Y. Han and Q. Q. Zhang, Adv. Mater. Res., 2013, 724, 1033–1036.

---