1. Introduction
Air pollution and global warming are plaguing the Earth. Developing clean energy and striving to achieve zero carbon emissions have become one of the solutions adopted by countries worldwide [1]. Hydrogen is regarded as a clean and sustainable renewable energy source [2]. However, for the large-scale market application of hydrogen energy, the safety and efficiency of hydrogen storage remain critical issues that need attention [3]. Magnesium hydride (MgH2) boasts a high hydrogen storage capacity (~7.6 wt.% and 109 kg/m3), and magnesium resources are abundantly available [4]-[6]. Nevertheless, in practical applications, MgH2 suffers from sluggish reactions due to the inertness of the magnesium surface, requiring high temperatures of 300˚C - 400˚C for its hydrogen absorption and desorption processes.
Currently, researchers worldwide have found that introducing transition metals (such as titanium, niobium, vanadium, cobalt, nickel, manganese, etc.) and their compounds into MgH2 can improve the reaction kinetics and thermodynamics of MgH2, thereby enhancing its hydrogen storage performance [7]-[9]. For example, the introduction of Nb2O5 into MgH2 significantly enhances the desorption kinetics, exhibiting high catalytic activity [10]-[14]. Studies have shown that when the high-valent Nb (+5) in Nb2O5 is reduced to low-valent Nb (such as NbO2, NbO, etc.), the hydrogen absorption and desorption properties of MgH2 are significantly improved [15] [16]. This is because the multivalent Nb species uniformly dispersed in MgH2 act as hydrogen channels and promote hydrogen diffusion [17]. In 2020, the team led by Professor Yongfeng Liu from the Laboratory of Silicon Materials at Zhejiang University used Nb2O5 hollow spheres (o-Nb2O5) to reduce the initial dehydrogenation temperature of MgH2 to 195˚C, and more than 5.5 wt.% of H2 could be desorbed within 5 minutes at 300˚C [14].
To further improve the hydrogen storage performance of MgH2, various dual transition metal ions have been developed to leverage their role in weakening the Mg-H bond binding energy. For instance, under the synergistic effect of dual transition metals in TiVO3.5, CuFe2O4, and NiFe2O4 [16] [19] [20], reactions between metal oxides and MgH2 generate low-valent oxides or zero-valent metallic elements, increasing catalytic active sites and enhancing the hydrogen storage performance of MgH2. In addition to dual-transition-metal catalysts, researchers have also studied the addition of various carbon nanomaterials. For example, Zou from Shanghai Jiao Tong University et al. used porous carbon nanofibers (pCNF) to self-assemble a scaffold of MgH2/Ni nanoparticles (NPs), enabling MgH2 to absorb hydrogen at 100˚C with a capacity of 2.2 wt.% within 120 minutes [20].
Liu et al. [21] synthesized bamboo-like carbon nanotubes for the self-assembly of MgH2 nanoparticles, which desorbed hydrogen at approximately 220˚C with a hydrogen capacity of 5.79 wt.%. Nobuko Hanada et al. from Waseda University synthesized a MgH2-Nb2O5-CNT composite, which desorbed 5.7 wt.% of hydrogen at 350˚C for 60 minutes [22]. Lan et al. [23] from Guangxi University synthesized an N-Nb2O5@Nb2C-doped MgH2 composite that started desorbing hydrogen at 178˚C and released 6.3 wt.% of hydrogen within 1.95 minutes at 350˚C. These studies have all found that CNT samples show considerable improvements in hydrogen absorption and desorption properties. Thus, it is concluded that CNT samples exhibit higher hydrogen capacity and faster kinetics, facilitating the entry of hydrogen atoms into the interior of Mg particles and their escape.
This paper reports the synthesis of Heterojunctions Nb2O5-NbC uniformly distributed in a carbon nanofiber matrix. Polystyrene (PS) is introduced as a growth regulator for the Heterojunctions to control the Nb2O5-NbC nanocrystals and prevent agglomeration of Nb2O5-NbC nanocrystal particles. The in-situ formed Nb2O5-NbC Heterojunctions provides abundant active surfaces and interfacial active sites, weakening the interaction between Mg-H bonds and extending the length of Mg-H bonds. The results show that the initial desorption temperature of MgH2 catalyzed by the Nb2O5-NbC Heterojunctions is reduced to 293.6˚C, 52.6˚C lower than that of ball-milled MgH2.
The presence of NbC and Nb2O5 inhibits the grain size of MgH2, providing more active centers and hydrogen diffusion paths. On the other hand, increasing the spinning speed leads to a larger diameter of carbon nanotubes, resulting in a reduction of Nb2O5-NbC Heterojunctionsper unit area of the carbon nanofiber matrix. This increases the kinetic barrier, leading to an increase in the initial desorption temperature of MgH2. However, the above two inferences are speculative interpretations derived from experimental results and literature data, and do not necessarily represent the actual reaction mechanism or situation. The findings of this study provide insights for further improving the hydrogen storage performance of Mg-based hydrogen storage systems.
2. Experimental
2.1. Sample Preparation
Polyacrylonitrile, niobium oxalate and polystyrene are analytical grade, which are purchased from Sigma-Aldrich, Macklin and Alfa Aesar respectively. Dimethylformamide was purchased from Sigma. All chemical reagents are used directly after purchase. The water involved in the experiment is distilled water.
Nb2O5-NbC/CNF-PS [24] was synthesized according to literature reports. The overall preparation process is as follows: firstly, 1.0 g of polyacrylonitrile, 0.5 g of niobium oxalate, 0.5 g of polystyrene are dissolved in 10ml of dimethylformamide, and then stirred at 60˚C for 24 hours. Then, the electrospinning solution was put into a 10 ml plastic syringe equipped with No.21 stainless steel needle, and the feed rate was 0.035 mm/min−1. In the process of electrospinning, the electrospinning solution was placed at a high voltage of 22 kV, and the obtained nanofibers were collected on a copper collector 20 cm away from the needle tip. The collected electrospun films were dried at 280˚C for 2 hours, then annealed at 900˚C for 2 hours at a heating rate of 5˚C/min in a nitrogen atmosphere.Following the feeding method of Sample ①, experiments with feed rates of 0.070 mm/min and 1.05 mm/min were designed separately, yielding Sample ② and Sample ③, respectively. The subsequent preparation procedures for Samples ②-③ were consistent with those of Sample ①: first drying at 280˚C for 2 hours, followed by annealing at 900˚C for 2 hours in a nitrogen atmosphere with a heating rate of 5˚C/min.
Preparation of MgH2-Nb2O5-NbC/CNF-PS: Nb2O5-NbC/CNF-PS Heterojunctions was doped into MgH2 at 10.00 wt.% of the total mass, with MgH2 accounting for 90 wt.% of the total mass. Samples were sealed in a grinding jar (150 mL) made of SUS304 steel inside a glove box. Stainless steel balls with diameters of 10, 5, 2 mm were used in the grinding jar. Mechanical grinding was conducted for 12 hours using a QM-3SP2 planetary ball mill (manufactured in Nanjing). The mass ratio of powder to grinding balls was approximately 120:1, and the grinding speed was 500 rpm.
To prevent sample temperature, rise due to excessive grinding speed, the ball mill rotated forward for 30 minutes, paused for 5 minutes, and then rotated backward for 30 minutes. The obtained sample was denoted as MgH2-10 wt.% (Nb2O5-NbC/CNF-PS). All sample operations were performed in a glove box with a circulating purification system filled with high-purity argon (99.999%), where the water and oxygen contents were both below 0.01 ppm to prevent sample oxidation and contamination. For comparison, pure ball-milled MgH2 was also sealed and ball-milled under the same conditions.
2.2. Structural Characterization
The phase composition of all materials was characterized by X-ray diffraction (XRD, D8 Advance, Bruker AXS) using filtered Cu Kα radiation. The detection angle 2θ ranged from 20˚ to 80˚ with a step size of 0.02˚. A field-emission scanning electron microscope (FE-SEM; JEOL7500FA, Tokyo, Japan) was used for observation.
2.3. Property Evaluation
The temperature-dependent hydrogen desorption behavior was measured using a self-made temperature-programmed desorption (TPD) system. Approximately 20 - 30 mg of the sample was loaded into a stainless-steel tubular reactor, which was then connected to the TPD system. The sample was gradually heated from room temperature to 450˚C at a rate of 2˚C/min. Typically, 20 - 30 mg of the sample was placed in the stainless-steel tubular reactor, with a thermocouple inserted into the reactor to monitor the real-time sample temperature. Hydrogen desorption was conducted in a non-isothermal mode, where the sample was heated at a heating rate of 2˚C/min under primary vacuum until complete hydrogen release. The temperature-programmed desorption (TPD) system was calibrated using standard pure hydrogen gas. The hydrogen storage capacity was calculated based on the ideal gas state equation and the sample mass. The initial desorption temperature was defined as the temperature at which the hydrogen desorption amount began to increase significantly, while the peak desorption temperature was the temperature corresponding to the maximum desorption rate. All temperature-programmed desorption experiments were repeated at least twice to ensure the reproducibility and reliability of the results.
3. Results
3.1. Preparation of Nb2O5-NbC/CNF-PS Heterojunctions
Nb2O5-NbC/CNF-PS heterojunctions were fabricated via electrospinning combined with high-temperature calcination under nitrogen atmosphere, using polystyrene as a nanocrystal growth regulator (Figure 1). The preparation procedure was as follows: niobium oxalate, polyacrylonitrile, and polystyrene were stirred at 80˚C for 24 h to achieve uniform dissolution in dimethylformamide. The precursor solution was electrospun at three feeding rates: 0.035 mm/min (Sample ①), 0.070 mm/min (Sample ②), and 1.05 mm/min (Sample ③). The as-spun fibrous film was pre-oxidized at 280˚C (heating rate: 5˚C/min) in air, then calcined at 900˚C at the same heating rate under nitrogen protection, followed by natural cooling to obtain the final composite sample.
Figure 1. Schematic illustration of the synthesis strategy for ultrafine Nb2O5-NbC/CNF-PS.
For temperature-programmed desorption (TPD) tests, 20 - 30 mg of each sample was loaded into a stainless-steel tubular reactor in a glove box, with a thermocouple inserted to monitor sample and furnace temperatures. TPD tests were performed under vacuum (initial pressure < 0.0001 bar) with a heating rate of 2˚C/min to the preset target temperature.
3.2. XRD Analysis of Nb2O5-NbC/CNF-PS Heterojunctions
Figure 2. (a), (b) MgH2-10 wt.% (Nb2O5-NbC/CNF-PS) after ball milling and after hydrogen desorption.
XRD characterization was conducted on the as-prepared Nb2O5-NbC/CNF-PS catalyst (Figure 2). Characteristic diffraction peaks located at 34.74˚, 40.41˚, 58.36˚, and 69.7˚ were detected, matching the standard card of NbC (PDF#00-10-181). Diffraction peaks at 28.31˚, 29.06˚, 36.49˚, 37˚, 54.97˚, and 55.33˚ were also observed, corresponding to the standard pattern of Nb2O5 (PDF#00-27-1313). After ball milling: the composite consists of MgH2, Nb2O5, NbC. After dehydrogenation: the main phases are Mg, Nb2O5, NbC. The results confirm that the Nb2O5/NbC heterojunction is well maintained during the dehydrogenation process, which supports our conclusion.
3.3. SEM Morphology of Nb2O5-NbC Heterojunctions
SEM characterization was performed on the precursor samples before calcination. All three samples displayed uniform fibrous structures. For Sample ① (0.035 mm/min), the carbon nanofibers were thin, with a diameter range of 80-150 nm (Figure 3(a), Figure 3(b). Sample ② (0.070 mm/min) exhibited thicker fibers with a diameter range of 150-230 nm (Figure 3(b), Figure 3(d). Sample ③ (1.05 mm/min) possessed the thickest fibers, with a diameter range of 800-1000 nm (Figure 3(c), Figure 3(f)). All samples showed hollow fibrous structures.
Figure 3. (a) (c) (e) and (b) (d) (f) SEM images of as-synthesized Nb2O5-NbC/CNF-PS at different spinning speeds before calcination.
After high-temperature calcination, the morphology of Nb2O5-NbC/CNF-PS composites was further observed by SEM. Nb2O5-NbC particles were embedded in the carbon fiber skeleton. Sample ① showed an ordered carbon fiber skeleton with uniformly dispersed Nb2O5-NbC particles (Figure 4(a), Figure 4(b)). Sample ② presented a relatively disordered fiber skeleton with slight particle agglomeration (Figure 4(c), Figure 4(d)). Sample ③ had a messy fiber skeleton with severe particle agglomeration and large bare fiber areas (Figure 4(e), Figure 4(f)).
Figure 4. (a) (c) (e) and (b) (d) (f) SEM images of as-synthesized Nb2O5-NbC/CNF-PS at different spinning speeds after calcination.
3.4. Hydrogen Desorption Performance of MgH2 Catalyzed by Nb2O5-NbC/CNF-PS Heterojunctions
Figure 5. (a) TPD curves of MgH2 under the catalysis of Nb2O5-NbC/CNF-PS at different spinning speeds, including ball-milled MgH2. (b) The derivative curves of H2 desorption of Nb2O5-NbC/CNF-PS at different spinning speeds, including ball-milled MgH2. (c) TPD curves of MgH2 under the catalysis of Nb2O5-NbC/CNF-PS at 0.035 mm/min spinning speeds, including ball-milled MgH2. (d) The derivative curves of H2 desorption of Nb2O5-NbC/CNF-PS at 0.035 mm/min spinning speeds, including ball-milled MgH2.
TPD tests were carried out to evaluate the hydrogen desorption performance of pure ball-milled MgH2 and MgH2-10 wt.% (Nb2O5-NbC/CNF-PS) composites. All samples were heated to 450˚C at 2˚C/min under vacuum, and the corresponding TPD curves and derivative curves were recorded (Figure 5).
For pure ball-milled MgH2, the initial hydrogen desorption temperature was 346.2˚C, and the hydrogen desorption capacity at 450˚C was 7.48 wt.%. For the catalyzed samples, the initial desorption temperatures were 293.6˚C (Sample ①), 302.9˚C (Sample ②), and 320.2˚C (Sample ③), with hydrogen desorption capacities of 6.48 wt.%, 6.38 wt.%, and 6.35 wt.% at 450˚C, respectively.
The hydrogen desorption peak temperatures derived from TPD curves were 367.4˚C for pure MgH2, 338.3˚C for Sample ①, 341.8˚C for Sample ②, and 354.5˚C for Sample ③. Compared with pure MgH2, Sample ① showed a 52.6˚C reduction in initial desorption temperature and a 29.1˚C reduction in desorption peak temperature. At 346.2˚C, the MgH2-10 wt.% (Nb2O5-NbC/CNF-PS) composite released over 4.89 wt.% of H₂, while pure MgH2 just initiated hydrogen desorption.
4. Discussion
The XRD results confirmed the successful fabrication of Nb2O5-NbC heterojunctions via electrospinning and high-temperature calcination, with the characteristic peaks of both Nb2O5 and NbC clearly detected without impurity phases. This heterostructure is critical for the subsequent catalytic modification of MgH2, as the synergistic effect between Nb2O5 and NbC has been proven to effectively boost the hydrogen storage kinetics of Mg-based materials.
SEM observations revealed that the electrospinning feeding rate exerted a significant influence on the diameter and dispersion of the carbon nanofibers, as well as the distribution of Nb2O5-NbC particles. With the increase of feeding rate, the fiber diameter increased sharply from 80 - 150 nm to 800 - 1000 nm, accompanied by aggravated particle agglomeration and disordered fiber skeleton. This morphological evolution is attributed to the excessive solution supply at higher feeding rates, which hinders the uniform stretching of fibers during electrospinning and leads to uneven nucleation and growth of heterojunction particles during calcination. The reduced specific surface area and fewer active sites caused by particle agglomeration directly weakened the catalytic efficiency of the composite, which was consistent with the deteriorating hydrogen desorption performance of Samples ② and ③.
The hydrogen desorption data demonstrated that doping Nb2O5-NbC/CNF-PS heterojunctions significantly lowered the hydrogen desorption temperature and improved the dehydrogenation kinetics of MgH2, especially for Sample ① prepared at 0.035 mm/min. The optimized catalytic performance can be explained by two key factors. Firstly, polystyrene decomposition during high-temperature calcination generates gas bubbles, which not only promote the partial conversion of Nb2O5 to NbC but also inhibit the aggregation of heterojunction particles, ensuring uniform dispersion of active sites in the carbon fiber framework. Secondly, high-energy ball milling facilitates the uniform mixing of the catalyst with MgH2, breaks the surface passivation layer of MgH2, exposes fresh active surfaces, increases lattice defects and internal stress, and shortens the hydrogen diffusion path, thereby reducing the dehydrogenation activation energy of MgH2.
Compared with pure ball-milled MgH2, the initial desorption temperature of Sample ① decreased by 52.1˚C, and the peak desorption temperature dropped by 29.1˚C, verifying the remarkable synergistic catalytic effect of Nb2O5-NbC heterojunctions. The slight decrease in hydrogen desorption capacity of catalyzed samples is mainly due to the mass fraction of the catalyst in the composite, which does not affect the overall catalytic superiority of the material.
Despite the enhanced dehydrogenation performance of MgH2 catalyzed by Nb2O5-NbC/CNF-PS, there are still limitations in this study. The hydrogen adsorption performance and cycling stability of the composite have not been systematically investigated, and the long-term catalytic durability under repeated hydrogen absorption-desorption cycles remains to be verified. In future work, the catalyst dosage will be further optimized, and the cycling performance and hydrogen absorption kinetics will be tested in depth. Additionally, in-situ characterization techniques will be adopted to reveal the catalytic mechanism of Nb2O5-NbC heterojunctions on MgH2 at the atomic scale, providing theoretical support for the development of high-performance Mg-based hydrogen storage materials.
5. Conclusions
In summary, this work employed electrospinning to in-situ construct Nb2O5-NbC Heterojunctionswithin carbon nanofiber (CNF) matrices via polystyrene (PS) introduction, aiming to enhance the hydrogen storage performance of Mg-based materials. Results indicate that the synthesized Nb2O5-NbC/CNF-PS nanocomposite exhibits a low initial hydrogen desorption temperature of 293.6˚C and a high hydrogen storage capacity of 6.48 wt.%. While ball-milled MgH2 begins desorption at 346.2˚C, the MgH2-10 wt.% (Nb2O5-NbC/CNF-PS) composite releases over 4.89 wt.% H₂ at this temperature, with its initial desorption temperature reduced by 52.6˚C.
Scanning electron microscopy (SEM) revealed that decreasing electrospinning speed reduces carbon nanotube diameter and promotes uniform distribution of Nb2O5-NbC Heterojunctionson CNFs. Based on the above results, it can be demonstrated that the Nb2O5-NbC heterojunctions used as catalysts can effectively improve the hydrogen storage properties of MgH2. This PS-assisted synthesis strategy for Nb2O5-NbC/CNF offers new insights into scalable and cost-effective industrial production of carbon-based materials.
Funding
This work was financially supported by the 13th Graduate Education Innovation Fund of Wuhan Institute of Technology (No. CX2021023).