Investigation of Hydrogen Storage Characteristics of MgH 2 Based Materials with Addition of Ni and Activated Carbon

: Magnesium-based materials are promising as hydrogen storage media due to their high theoretical hydrogen absorption capacity, abundance and low price. The subject of this study are the hydrogen sorption characteristics of the composites 80 wt % MgH 2 -15 wt % Ni-5 wt % activated carbon (synthesized from polyolefin wax, a waste product of polyethylene production at low pressure which will be denoted further in the text as POW) and 90 wt % MgH 2 -5 wt % Ni-5 wt % POW, prepared by ball milling under argon atmosphere. Structure, phase and surface composition of the samples before and after hydrogenation are determined by XRD and TEM. The maximum absorption capacity value of the composites at a temperature 573 K and after 60 min. of hydrogenation are 5.3 wt % H 2 for the material with higher Ni content and 5.5 wt % H 2 for the other sample. The presence of both additives—nickel and activated carbon derived from POW—has a positive impact on hydrogenation kinetics and the capacity achieved. The results from TEM characterization, e.g., the polycrystalline SAED (selected area electron diffraction) show the presence of graphite, Mg and monoclinic Mg 2 NiH 4 . hydrogenation kinetics and leads to relatively high absorption capacity. The results from XRD and TEM characterizations after hydrogenation, e.g., the polycrystalline SAED (selected area electron diffraction) show the formation of monoclinic and orthorhombic Mg 2 NiH 4 in the both samples after hydrogenation. Regarding, the possible practical application of these materials, the hydrogen absorption and desorption temperatures should be lower than those, used in this study.


Introduction
Hydrogen is a renewable and clean energy carrier that can store or deliver a tremendous amount of energy. One of the key problems related with its widespread use is its storage. Finding a solution for the hydrogen storage problem is considered to be the foremost challenge for the hydrogen economy. Some metals and intermetallics form hydrides which can reversibly release H2 upon heating. Hydrogen storage in metal hydrides is a safe and cost efficient method. One of the promising candidates for hydrogen storage applications is magnesium. Materials based on this lightweight metal have potential as hydrogen storage media due to their high theoretical hydrogen absorption capacity, abundance and low price. To overcome some hindrances of these materials, like slow hydrogen sorption kinetics and high absorption-desorption temperatures, various additives combined with high energy ball milling are used [1][2][3]. The recently published review by Yartys et al. [3] is focused on the research field of magnesium-based materials for hydrogen energy storage with topics of fundamental and applied studies and a review of the frontiers of both experimental and theoretical research, including nanostructuring, kinetics, thermodynamics and catalysis of magnesium and the hydrides. This review clearly states that Mg-based materials are very attractive for variety of energy applications.
It is established that different carbon compounds, ball milled with Mg or MgH2, reduced hydrogen sorption temperature, improved kinetics and also prevented MgH2 particles growth upon milling and hydrogen sorption cycling [4][5][6][7][8][9][10][11]. A considerable improvement of the hydrogen storage properties of Mg, using ball milling of MgH2 with carbon additives have been reported by Wu et al. [4]. The addition of carbon containing compounds to magnesium leads to milder hydrogen storage temperature and improved hydrogen sorption kinetics of magnesium based materials. Fuster et al. have observed that the catalytic effect of graphite is not related with its morphology, and it is more pronounced when the carbon additive is in higher content and introduced at the beginning of ball milling [4]. The desorption reaction of MgH2, catalyzed by different carbon materials subjected to microwave irradiation have been investigated by Awad et al. [7]. An enhancement of dehydrogenation rate occurs when the amount of carbon additive, microwave power and milling time are increased. Improvement of hydrogen sorption kinetics can be achieved by adding Ni to Mg and the combination between the positive effect of carbon containing additive and Ni, could result in further benefits of hydrogen sorption properties. The Mg2Ni forms Mg2NiH4 with higher reaction rate, absorbs and desorbs hydrogen at lower temperatures than magnesium, but has theoretical hydrogen storage capacity of 3.6 wt % H2. Some studies on preparation, crystal structure of Mg2NiH4 and the catalytic effect of Ni and some other additives to hydrogen storage characteristics of MgH2 are published in a series of articles [12][13][14][15][16][17][18][19]. There are two allotropic forms of Mg2NiH4-a monoclinic low temperature one and other high temperature with a cubic structure. Synthesis of Mg2NiH4 by MgH2 and Ni is reported by Polanski et al. [14] and Martínez-Coronado et al. [15]. Cermak et al. [12] have shown that catalytic efficiency of Mg2NiH4 is considerably higher than that of pure Ni and intermetallic Mg2Ni. These authors also suggested that Mg2NiH4 has un-twinned phase LT1 and micro-twinned phase LT2, two low-temperature modifications, existing below a temperature of 508 K. The phase LT2 is thermodynamically more stable than LT1, and LT1 shows higher catalytic efficiency than LT2 [12]. During the study of the synergistic catalytic effect of SrTiO3 and Ni, on the hydrogen storage properties of MgH2 at a temperature 593 K and pressure of 2.7 MPa, Yahya et al. [16] have observed formation of Mg2Ni and Mg2NiH4, and these active species contribute to improved hydrogen storage properties of the MgH2 system. Hydrogen storage materials based on the Mg(Ni1−xMnx)2 are synthesized by ball milling and studied as materials for solid state hydrogen storage by Gkanas et al. [18]. The duration of ball milling and the amount of Mn have influence on the hydrogen sorption properties. Prolonged ball milling leads to higher content of MgNi2 and MgO and these species reflect negatively on the hydrogen absorption capacity values and kinetics.
In some of our previous investigations, the effect of graphite and activated carbons derived from low-cost agricultural by-products on the hydrogen sorption properties of magnesium is studied [9,11]. The obtained results drove us to the conclusion that the addition of carbon containing compounds in the materials based on magnesium, has positive impact on their hydrogen sorption properties, leading to an enhanced rate of hydrogenation and higher absorption capacity, compared to pure magnesium. Preparation and characterization of MgH2-nickel-carbon composites by ball milling under argon atmosphere, and investigation of their hydrogen sorption properties at different temperatures are performed in this study. Two types of additives to MgH2-carbon containing one synthesized from a waste product, and Ni, with different content, that could lead to formation of Mg2NiH4 are used. The aims of these experiments on the one hand are to synthesize at lower temperature, hydrogen pressure and shorter time of ball milling Mg2NiH4 and on the other hand to explore the effect of activated carbon prepared from waste product as additive to magnesium based hydrogen storage materials. This work contributes to environmental protection via hydrogen storage and also to resource recovery by using waste product for activated carbon synthesis.

X-ray Powder Diffraction of Ball Milled 80 wt % MgH2-15 wt % Ni-5 wt % POW and 90 wt % MgH2-5 wt % Ni-5 wt % POW
The purchased MgH2 contains some traces of Mg, but the main phase is tetragonal MgH2. X-ray diffraction patterns of ball milled 80 wt % MgH2-15 wt % Ni-5 wt % POW and 90 wt % MgH2-5 wt % Ni-5 wt % POW do not indicate a presence of MgO ( Figure 1a). The phases detected in ball milled composites are MgH2, Ni and some unhydrided Mg. The appearance of new phases is not observed in the samples after ball milling (Figure 1a), which is expected regarding the relatively short milling time.

Hydrogen Absorption and Desorption Characteristics of 80 wt % MgH2-15 wt % Ni-5 wt % POW and 90 wt % MgH2-5 wt % Ni-5 wt % POW
In Figure 2 the hydrogen absorption curves are shown for the 80 wt % MgH2-15 wt % Ni-5 wt % POW, 90 wt % MgH2-5 wt % Ni-5 wt % POW and pure MgH2 and as inset for the first 10 min. of the process of hydrogenation at a temperature of 573 K are also presented. The absorption capacity values obtained at a temperature of 573 K and after 60 min. of hydrogenation are 5.3 wt % H2 for the material with higher Ni content and 5.5 wt % H2 for the other sample. For only 1 min. of hydrogenation the 80 wt % MgH2-15 wt % Ni-5 wt % POW absorbed 1 wt % H2 at a temperature of 573 K versus 1.1 wt % H2 for the 90 wt % MgH2-5 wt % Ni-5 wt % POW. The kinetics hydrogenation curve at a temperature of 573 K for pure magnesium hydride is also presented in this figure. Both samples absorb hydrogen with increased rate and achieved higher absorption capacity than pure MgH2 without additives. The sample with higher Ni content demonstrated increased rate of hydrogenation in the initial stage of absorption, but slightly lower absorption capacity than the other one with 5 wt % Ni at temperatures of 573 K and 473 K. With a decrease of the absorption temperature at 473 K, the hydrogen absorption rate and absorption capacity value are significantly lower for both samples and the absorption capasity is 1.3 wt % H2 for the material with 15 wt % Ni and 1 wt % H2 for the other mixture with 5 wt % Ni. The hydrogen desorption curves for both samples are presented in Figure 3 and as inset for the first 10 min. of the process the curves at temperatures of 623 and 573 K are also presented. The 80 wt % MgH2 -15 wt % Ni-5 wt % POW desorbed faster the absorbed H2 than 90 wt % MgH2-5 wt % Ni-5 wt % POW up to around 15 min. at temperature of 623 K and a pressure 0.15 MPa H2 and only for this sample some desorption with low rate is observed at a temperature of 573 K. The hydrogen desorption curve at 573 K for pure magnesium hydride is also presented in Figure 3. The hydrogen desorption process for the both samples with additives of Ni and activated carbon is with much faster rate up to 15 min. than the rate of desorption for the pure MgH2. After the initial 15 min. of hydrogen desorption the kinetics curves of the 90 wt % MgH2-5 wt % Ni-5 wt % POW and MgH2 are identical. The presence of Ni leads to formation of Mg2NiH4, which has benefits to the kinetics of absorption and desorption, but decreases the absorption capacity.

Transmission Electron Microscopy Characterization
Transmission electron microscopy characterization is made for the 80 wt % MgH2-15 wt % Ni-5 wt % POW after hydrogenation. In Figure 4, a bright field micrograph of the composite 80 wt % MgH2-15 wt % Ni-5 wt % POW after hydrogenation is presented, along with polycrystalline SAED (selected area electron diffraction) patterns. The presence of monoclinic Cc Mg2NiH4 phase, graphite, MgH2 and magnesium are detected. A particle diameter size distribution of 80 wt % MgH2-15 wt % Ni-5 wt % POW sample, prepared by using Image J program is made (not presented). The average particle diameter derived from this TEM characterization is 8 nm.  The X-ray diffraction patterns for both samples after hydrogenation at a temperature of 573 K and pressure of 1 MPa H2 are shown in Figure 1b and confirmed the presence of MgH2 as main phase, Mg2NiH4 in monoclinic and orthorhombic forms and some Mg, Ni and traces of MgO.

Discussion
House et al. [17] proved the formation of Mg2NiH4 by selected enlarged peaks from the X-ray diffraction patterns of the MgH2 catalyzed with Ni in the sample ball milled for 10 h. Ball milling for 1h with the parameters used by us for the both prepared samples does not lead to the appearance of Mg2NiH4 or other new phases in the diffraction patterns (Figure 1a). The MgO is not detected by Xray diffraction analysis after ball milling and probably it is presented but below the detection limit.
The composite containing 15 wt % Ni at 573 K after 60 min of hydrogenation reaches absorption capacity of 5.3 wt % H2 and for the other sample with only 5 wt % Ni the capacity is 5.5 wt % H2. Practically the obtained absorption capacity values are very close and the difference between them is insignificant. The rate of hydrogenation process at a temperature of 573 K is increased for the 80 wt % MgH2-15 wt % Ni-5 wt % POW. At the initial stage of hydrogenation up to 2 min. the absorption capacity values are: for the sample with 15 wt % Ni-1.8 wt % H2 and 1.5 wt % H2 for the sample with lower amount of Ni. Decreasing the hydrogenation temperature to 473 K reflects negatively on the hydrogen absorption capacity and kinetics for the both samples. However, the mixture 80 wt % MgH2-15 wt % Ni-5 wt % POW has demonstrated higher absorption capacity and increased absorption rate at a temperature of 473 K than 90 wt % MgH2-5 wt % Ni-5 wt % POW. Both samples have demonstrated improved kinetics of hydrogen absorption than MgH2 without any additives. The final absorption capacity values at 573 K do not differ significantly, e.g., for the 90 wt % MgH2-5 wt % Ni-5 wt % POW is 5.5 wt % H2, for the 80 wt % MgH2-15 wt % Ni-5 wt % POW is 5.3 wt % H2 and for MgH2 is 5.1 wt % H2 after 60 min of hydrogenation. Regarding hydrogen absorption curves, the ternary hydride Mg2NiH4 is formed after only few cycles of hydrogenation/dehydrogenation at T = 573 K and P = 1 MPa for absorption and T = 623 K and P = 0.15 MPa for desorption. Comparing these results with some obtained before lead to the conclusions that adding nickel has great effect on the kinetics. For the first 2 min of hydrogenation process at same experimental conditions 95 wt % MgH2-5 wt % POW absorbed 1.1 wt % H2 [11] which is around only 60% of the capacity obtained by the 80 wt % MgH2-15 wt % Ni-5 wt % POW for the same time and experimental conditions. About half of the absorbed hydrogen is released after 28 min of dehydrogenation by the 80 wt % MgH2-15 wt % Ni-5 wt % POW at temperature 623 K and a pressure 0.15 MPa H2. Most probably due to the higher content of Mg2NiH4 in the mixture of MgH2 and 15 wt % Ni shows faster desorption, but only up to 15 min from the process. Some very slow desorption at 573 K is detected only for the 80 wt % MgH2-15 wt % Ni-5 wt % POW and this should be associated also with higher Ni content and as a result increased Mg2NiH4 presence. The MgH2 shows very slow hydrogen desorption kinetics at 623 K up to first 15 min of the process, but after that its desorption curve overlap this of the 90 wt % MgH2-5 wt % Ni-5 wt % POW. At same experimental conditions 95 wt % MgH2-5 wt % POW desorbed hydrogen with slower kinetics for the first 5 min of the process than these composites with addition of nickel [11]. These positive effects on the kinetics can be attributed to the Mg2NiH4 formation. Definitely, when this ternary hydride is in higher amount in the samples based on MgH2 thus reflects positively on the sorption kinetics and lead to lower hydrogenation and dehydrogenation temperatures.
The 80 wt % MgH2-15 wt % Ni-5 wt % POW consists of fine particles and they enhance the hydrogen diffusion and hydride phase formation, thus could be beneficial of the hydrogen sorption properties. The TEM characterizations of 80 wt % MgH2-15 wt % Ni-5 wt % POW are confirmed the formation of ternary hydride Mg2NiH4 with low temperature monoclinic structure.
Some published results by other authors also confirmed the formation of Mg2NiH4 starting by MgH2 and Ni and in some studies other additives were used. The Mg-Ni-Fe alloy obtained by reactive ball milling showed fast hydrogenation rate at the beginning of the process and absorbed 3.3 wt % H2 at temperature 593 K and a pressure of 1.2 MPa H2 [19]. The X-ray diffraction patterns of Mg-Ni-Fe after hydrogenation/dehydrogenation cycling confirmed a presence of Mg2Ni. A study of the catalytic effect of SrTiO3 and Ni on the hydrogen storage properties of the MgH2 have been carried out by Yahya et al. [16]. At temperature of 593 K and a pressure 2.7 MPa H2 the Mg2NiH4 were formed and help to improve the hydrogen storage properties of MgH2 system. House et al. [17] prepared MgH2 materials catalyzed with Ni by ball milling with different duration. As milling time increased up to 10 h, more Ni was transformed into Mg2NiH4 and this cause degradation of hydrogen absorption capacity value. The theoretical absorption capacity value of Mg2Ni is around twofold lower than that of magnesium and this negative effect is expected. Recently, Gkanas et al. [18] reported their investigation on a new hydrogen storage materials of the type Mg2(Ni1−xMnx)2. The stored hydrogen by these intermetallic phases ball milled for 10 or 20 h were around 1 wt % H2 at a temperature of 573 K and a twofold higher hydrogen pressure than this used in our study, e.g., 2 MPa. The hydrogen desorption was observed at temperatures 573 K and 523 K and at a pressure 0.05 MPa H2. These materials demonstrated fast absorption and desorption kinetics, and also desorption at lower temperatures, but with very low sorption capacity. The parameters of ball milling and the conditions of hydrogenation/dehydrogenation used by us are different, but by an easier and faster way we obtained materials that demonstrate relatively fast sorption rate and higher capacity. As is illustrated by the obtained results and characterizations both monoclinic and orthorhombic Mg2NiH4 type phases are formed in the samples and have positive effect on the absorption and desorption kinetics.
The combination of these two types of additives-nickel and activated carbon POW results to improved hydrogen sorption properties for the 80 wt % MgH2-15 wt % Ni-5 wt % POW and 90 wt % MgH2-5 wt % Ni-5 wt % POW. These investigations confirmed the appearance of Mg2NiH4 at these shorter ball milling treatment, lower temperatures and pressure reported in the literature. Moreover, in the both samples MgNi2 is not detected after 10th cycle of hydrogenation at a temperature of 573 K and pressure 1 MPa H2. The MgNi2 phase does not form hydride and consume a part of magnesium, e.g., reflecting negatively on the capacity and the kinetics. Starting from MgH2 and Ni mixtures eliminated the formation of the MgNi2 phase. The complete reaction of nickel and MgH2 to Mg2NiH4 is not observed and some Ni is detected in the both samples after hydrogenation. It seems that this method of Mg2NiH4 synthesis is easier, faster and gives the opportunity to avoid the MgNi2 appearance.
In these materials based on MgH2 further improvement of hydrogen sorption kinetics probably could be observed at different ball milling parameters or/and for the samples with other composition, e.g., with increased amount of nickel. The last option should be carefully précised because of diminution of the absorption capacity when the nickel content is increased.

Materials and Methods
The activated carbon is prepared from a polyolefin wax, a waste product of polyethylene production at low pressure, submitted to pyrolysis and steam activation, and used as additive in the preparation of MgH2 based materials. More detailed procedure for activated carbon synthesis can be found elsewhere [9]. The MgH2 with purity 98% and Ni powder (99.7%) are purchased from Sigma Aldrich (St. Louis, MO, United States).

Ball Milling in a Planetary Mill
Mixtures of MgH2, Ni and carbon additive, in the corresponding ratio of 80 wt % MgH2-15 wt % Ni-5 wt % POW and 90 wt % MgH2-5 wt % Ni-5 wt % POW, are ball milled under Ar atmosphere in a planetary mono mill Pulverisette 6 Fritsch (Fritsch, Idar-Oberstein, Germany). The parameters of ball milling are: the rotation speed of 200 rpm for duration of 1 h, ball to sample weight ratio 10:1 in a stainless steel vial (volume ca. 80 cm 3 ) and balls with diameter 10 mm.

Hydrogen Sorption Measurements
Absorption and desorption kinetics of the composites, denoted, are studied with volumetric Sievert's type apparatus at different temperatures. The absorption measurements are carried out at temperature 473 and 573 K, and a pressure of 1 MPa H2, while desorption is investigated at T = 623 and 573 K, and P = 0.15 MPa. The both samples are subjected to at least 10 cycles of hydrogenation/dehydrogenation and the best cycle at each temperature is presented in Figures 2 and  3. The hydrogen sorption kinetics curves for the 80 wt % MgH2-15 wt % Ni-5 wt % POW are published before in [20].

X-ray Diffraction and TEM Analysis
The composition of the starting compounds and of the samples, after ball milling and hydrogenation, are determined by XRD with Cu Kα radiation (Powder X-ray Diffractometer Bruker D8 Advance with a LynxEye detector, Bruker, Karlsruhe, Germany). The sample with higher amount of Ni after hydriding is examined also by TEM (TEM HR STEM JEOL JEM 2100 with GATAN Orius 832 SC1000 CCD Camera, JEOL, Tokyo, Japan) at accelerating voltage of 200 kV. The preparation procedure of the specimen is consisted of dispersing them in ethanol by ultrasonic treatment for 6 min. The suspensions are dripped on standard holey carbon/Cu grids. The sample is exposed briefly to air during transfer to the TEM holder. Some part of TEM characterizations were published before in [20].

Conclusions
The maximum achieved absorption capacity values at a temperature of 573 K and after 60 min. of hydrogenation are 5.3 wt % H2 for the material with higher Ni content and 5.5 wt % H2 for the other sample. The presence of both additives-Ni and activated carbon derived from a polyolefin wax, a waste product of polyethylene production at low pressure-has a positive impact on hydrogenation kinetics and leads to relatively high absorption capacity. The results from XRD and TEM characterizations after hydrogenation, e.g., the polycrystalline SAED (selected area electron diffraction) show the formation of monoclinic and orthorhombic Mg2NiH4 in the both samples after hydrogenation. Regarding, the possible practical application of these materials, the hydrogen absorption and desorption temperatures should be lower than those, used in this study. Funding: This research was funded by Bulgarian Ministry of Education and Science-National science fund (Project DFNI-KП-06-H27/9, 08/12/2018) and part of the experiments were performed with equipment included in the National Infrastructure SI ESHE (NI SEVE) supported by the Bulgarian Ministry of Education and Science under grant agreement № DO1-160/28.08.18.

Conflicts of Interest:
The authors declare no conflict of interest.