Ultrasonic in-situ reduction preparation of SBA-15 loaded ultrafine RuCo alloy catalysts for efficient hydrogen storage of various LOHCs

Graphical abstract

Highlights

• •RuCo catalyst for efficient hydrogen storage of various LOHCs was prepared by SEA-UR.
• •Higher dispersion and antioxidant capacity of RuCo alloy increase the active sites.
• •Ultrafine particles and electron transfer between Ru and Co improve catalytic activity.
• •Strong coordination effect between RuCo alloy and Si-OH improves catalytic stability.
• •Ru_xCo_y/S15-SU can efficiently catalyze hydrogen storage in NEC, DBT, MBT and acenaphthene.

Abstract

SBA-15-loaded RuCo alloy nanoparticle catalysts (Ru_xCo_y/S15-SU) for the efficient catalysis of hydrogen storage by various liquid organic hydrogen carriers (LOHCs) were prepared via strong electrostatic adsorption (SEA)-ultrasonic in-situ reduction (UR) technology. The above prepared catalysts were subjected to a series of characterization, such as XPS, H₂-TPD/TPR, N₂ adsorption–desorption, ICP, CO-chemisorption, FT-IR, XRD and TEM. Ru³⁺ and Co²⁺ were evenly anchored on the surface of SBA-15 by SEA, and ultrafine RuCo alloy nanoparticles were formed by UR without any chemical reducing or stabilizing agents. The addition of Co enhanced the dispersion and antioxidant capacity of the RuCo alloy NPs with an average particle size of 2.07 nm and increased the number of catalytically active sites. The synergistic effect of ultrafine particle size and electron transfer between Co and Ru improved the catalytic performance of monobenzyltoluene (MBT) for hydrogen storage. SEA-UR technology strengthened the coordination effect between RuCo alloy NPs and Si-OH, which enhanced the catalytic stability. H₂-TPD and H₂-TPR indicated that the addition of Co led to more activated H₂ to produce hydrogen overflow. For the hydrogenation of MBT, the produced Ru₂Co₁/S15-SU showed excellent catalytic performance. The hydrogen storage efficiency of MBT was 99.98 % under 110 °C and 6 MPa H₂ for 26 min, and the TOF was 145 min⁻¹, which is significantly superior to that of Ru/S15-SU catalyst and that reported in the literature. The hydrogen storage efficiency was still as high as 99.7 % after ten cycles, which was much better than that of Ru/S15-SU and commercial 5 wt% Ru/Al₂O₃. Ru₂Co₁/S15-SU is also suitable for efficiently catalyzing hydrogen storage of N-ethylcarbazole, dibenzyltoluene and acenaphthene.

1. Introduction

At present, worldwide energy still relies mainly on fossil fuels. Extensive use of fossil fuels will exacerbate environmental pollution and the greenhouse effect. All the countries in the world are moving toward a low-carbon and clean energy transition, so the search for efficient and renewable clean energy tasks cannot be delayed. Hydrogen energy stands out among various alternative energy sources and is regarded as a clean energy with the greatest development potential in the twenty-first century due to its advantages of abundant reserves, high energy density, cleanliness and nonpollution, diverse storage and utilization methods [1], [2], [3]. Hydrogen has the disadvantages of light mass, flammable and explosive, difficult to store and transport, etc. The most urgent link in the current industrial chain is safely and efficiently storing and transporting hydrogen. In order to carry out the full utilization of hydrogen energy, this restriction factor must be solved [4], [5], [6].

Hydrogen has extremely strict storage and transportation requirements during use. Hydrogen storage technology should consider the cost, hydrogen storage density, safety and other factors, and the development of high volumetric and quality density hydrogen storage technology has been a current research hotspot. The main hydrogen storage technologies that have been applied at this stage include high-pressure gaseous, cryogenic liquids, low-temperature physical adsorption, metal hydrides and LOHC hydrogen storage [7], [8].

High-pressure gaseous hydrogen storage involves compressing hydrogen and storing it in hydrogen storage tanks, which has a simple technology and fast filling and discharging speeds. The bulk density can only reach 40 g/L when the hydrogen pressure reaches 70 MPa, the high pressure container material and process requirements are very high, and there is a risk of leakage [9]. Cryogenic liquid hydrogen storage is achieved by lowering the temperature to −253.15 °C, liquefied the hydrogen, and then stored in an insulated tank. The hydrogen storage density is about as high as 70.8 g/L, but the liquefaction process is difficult, the energy loss is large, and the requirement for container insulation is high, so the application range is small [10]. Physical adsorption hydrogen storage involves the establishment of weak bonds and interactions between H₂ and adsorbing material through intermolecular forces. This technique requires a high surface area or microporous volume of absorbing material, and the dehydrogenation rate is slow under room temperature [11]. The reaction of a metal alloy with hydrogen to form the metal hydride MH_y or the metal solid solution MH_x realized metal hydride hydrogen storages. The easy decay and low recoverability of hydrogen storage materials and poor dehydrogenation performance at low temperatures are difficulties to be solved immediately [12], [13]. LOHC hydrogen storage through reversible hydrogenation and dehydrogenation reactions of unsaturated aromatic compounds, and these hydrogen carriers can be regenerated and recycled after the release of hydrogen [14], [15]. Hydrogen storage in LOHCs is a cycle-closed reaction consisting of three steps: hydrogenation, storage and transportation, and dehydrogenation. LOHCs have a higher flash point and are therefore more stable, safe, higher hydrogen storage density. They can be suitable for direct long-distance transportation using existing oil transportation facilities, which are among the most promising technologies.

LOHCs are the core of organic liquid hydrogen storage technology. Theoretically, all organic materials containing unsaturated bonds can be used as hydrogen storage media but are constrained by the density of hydrogen storage, catalyst, dehydrogenation reaction conditions, reusability, energy savings and environmental protection. Only a few organic liquid materials [16], which are mainly classified as heterocyclic aromatic compounds, cyclic hydrocarbons and other materials, have been applied. N-ethylcarbazole (NEC) is a widely studied heterocyclic aromatic compound due to its higher mass hydrogen storage density and lower dehydrogenation enthalpy, as well as its high safety, cleanliness, nonpollution, and recyclability [17], [18]. Wan et al. [19] studied the reaction kinetics of NEC hydrogenation using Ru/γ-Al₂O₃ catalyst. However, the higher price, high melting point and low thermal stability of NEC limit its large-scale application in industry. Monobenzyltoluene (MBT) and dibenzyltoluene (DBT) are toluene addition compounds that are both used as heat transfer oils in industrial applications. They have many advantages, such as the lower vapor pressure during the hydrogenation reaction, higher mass hydrogen storage densities (6.18 wt% and 6.20 wt%), nontoxic, nonflammable, low cost, as well as being able to be compatible with existing infrastructures to a great extent. They are ideal liquid organic hydrogen carriers for cyclic hydrocarbons and have great potential for commercialization [20]. Kim et al. [21] studied the MBT hydrogenation reaction over ZrO₂-loaded Ru nanocatalysts. It was shown from results that MBT conversion and 12H-MBT selectivity could reach 100 % under 150 °C and 5 MPa H₂. Leinweber et al. [22] studied MBT hydrogenation reaction pathway over Ru/Al₂O₃ catalysts and showed that one side of the benzene ring in MBT attached to the methyl group was first hydrogenated to 6H-MBT (0H-MB was first hydrogenated to 4H-MBT, and then to 6H-MBT), and the other side of the benzene ring was then hydrogenated step by step to 12H-MBT (6H-MBT was hydrogenated directly to 10H-MBT, and finally to 12H-MBT). Other materials, such as acenaphthene and naphthalene, are derived from coal tar. Acenaphthene is hydrogenated and saturated to prepare tetrahydroacenaphthene and perhydroacenaphthene, yielding low-toxicity products that have been widely applied in polymers, pharmaceuticals, and fuels [23]. Naphthalene can be catalytically hydrogenated to produce tetrahydronaphthalene and decahydronaphthalene to enhance its application value, and the hydrogenation of naphthalene to produce decahydronaphthalene is an important reaction in diesel hydrotreating and reforming [24]. Acenaphthene and naphthalene have high mass hydrogen storage densities of 6.10 wt% and 7.24 wt%, respectively, and can be used as LOHCs.

LOHC hydrogen storage is significantly affected by the choice of active metal. Currently, the active metals commonly used to catalyze the hydrogen storage in LOHCs are Ru, Rh, Pd and Ni. Eblagon et al. [24], [25], [26], [27] carried out the catalytic hydrogen storage of LOHCs over the metals Ru, Rh and Pd to compare their activity. It was found that the metal Ru had the highest catalytic hydrogenation activity. High temperature and H₂ pressure are necessary for LOHCs to have high hydrogen storage capacity. The exploitation of more efficient, low price and green hydrogenation catalysts is imperative. Although noble metal catalysts, such as Ru, exhibit excellent hydrogen storage performance, the higher price of noble metals limits their further development in industry. Although non-precious metals have a low price, the lack of hydrogenation activity or low hydrogenation activity has prevented them from being further investigated. The introduction of non-precious metals into loaded noble metal catalysts is currently among the most widely used routes, which reduces the production cost of the catalysts while improving the hydrogen storage performance of the catalysts due to the metal interaction [28], [29], [30]. Tolek et al. [31] prepared TiO₂ based RuCo bimetallic catalysts using a co-impregnation method. It was found from the results of selective furfural hydrogenation showed that the charge transfer direction from Co to Ru occurs in the RuCo alloy, which promotes the reduction of Ru oxides, and that synergistic effects of the RuCo alloying system as well as the strong interactions between Ru and Co enhance the furfural hydrogenation activity. However, the introduction of non-precious metals into Ru-based catalysts and their use in the hydrogen storage reaction of LOHCs have rarely been reported. Zhu et al. [32] prepared graphene loaded Ru/Co/Co(OH)₂ nanocatalysts and investigated their ability to catalyze naphthalene hydrogen storage reaction. It was shown from the results that the above prepared catalysts exhibited outstanding catalytic activity and decahydronaphthalene selectivity because of the nanosynergistic effect of ruthenium, cobalt and cobalt hydroxides. Qin et al. [33] prepared nitrogen-doped partially graphitized carbon (NGC)-loaded Co@Ru core–shell nanocatalysts and catalyzed NEC hydrogenation. The synergy of the core–shell structure and the formation of NGC resulted in outstanding catalytic performance in NEC hydrogenation. At 130 °C and 6 MPa, the conversion of NEC and selectivity of 12H-NEC reached 100 % and 99.10 %, respectively. After 10 consecutive cycles, the above two terms were maintained at 100 % and 92.22 %, respectively. Pei et al. [34] prepared metastable Ru-B alloy catalyst and catalyzed NEC hydrogenation. The 12H-NEC yield was 99 % under room temperature, which was due to the sub-stabilized Ru-B alloy, distinctive electronic structure, and the high coordination active sites.

Most hydrogen storage catalysts are loaded catalysts that mainly include active components, co-catalysts and carriers. The carrier mainly plays the role of support and dispersion in the catalyst. Commonly used carriers include Al₂O₃, TiO₂, SiO₂, zeolite and activated carbon. Among these, Al₂O₃ and TiO₂ have larger specific surface area, high mechanical strength and other characteristics, which provide better support for the active components. SiO₂, zeolite and activated carbon with porous structures can better disperse the active components. The mesoporous molecular sieve SBA-15 has a regular ordered mesoporous hexagonal pore channel. When SBA-15 is used as a carrier to prepare a loaded catalyst to catalyze the hydrogenation of hydrogen storage materials, the unique ordered mesoporous pore channels of SBA-15 itself will play a domain-limiting role in the reaction, which will lead to the enhancement of the selectivity of the final product.

The preparation process (loading and reduction of active components) of loaded catalysts also has a great effect on their performance in catalyzing hydrogenation reactions. The electrostatic adsorption (SEA) is a method that deprotonates the hydroxyl groups on the carrier surface at a specific pH, generating a negative charge, which causes the metal ions to be adsorbed on the carrier surface by strong electrostatic forces [35], [36]. The advantage of this method is that it expands the range of applicability of many carriers other than molecular sieves and has greater universality. Moreover, the catalysts synthesized by the electrostatic adsorption method present stronger metal-carrier interactions due to the strong binding of opposite charges, resulting in the active species being uniformly dispersed on the carrier surface, resulting in high atom utilization and high catalytic stability [35], [36]. Wang et al. [37] synthesized highly active Pd/SBA-15 catalysts via SEA. The characterization results showed that SEA made the distribution of Pd NPs more uniform and also improved the stability of the catalyst.

Currently, the main reduction technology catalysts are chemical reduction [38], carbothermal reduction [39], hydrogen reduction [40], plasma reduction [41] and ultrasonic assisted reduction [42]. Ultrasound can not only effectively promote mass and heat transfer and shorten the reaction time, but also excite highly reactive hydrogen radicals (•H) from hydroxyl groups (HO–) via its unique acoustic cavitation effect [43], [44]. Moreover, the shear oscillation produced by ultrasound can effectively prevent metal nanoparticles from agglomerating [45], [46], [47]. Based on these advantages, the ultrasonic in-situ reduction technology has been widely used in the synthesis of nanomaterials [48]. Liu et al. [49] prepared NiFe-LDH-loaded Ru nanocatalysts via the ultrasonic in-situ reduction technology. The characterization findings revealed that the abundant HO– on the surface of NiFe-LDH was excited to •H under ultrasonic waves, which in-situ reduced Ru³⁺ to Ru⁰. The Ru NPs were evenly dispersed on the mesoporous surface of NiFe-LDH, which improved its catalytic performance for MBT hydrogenation.

In this paper, SBA-15 supported RuCo alloy nanocatalysts with excellent MBT hydrogenation activity were prepared using SEA-UR technology. The SEA-UR method is simple, efficient, environmentally friendly, low-cost, and has high potential for industrial applications. The optimal conditions for the catalyst preparation process and catalytic MBT hydrogenation reaction were studied. Compared with Ru/S15-SU and commercial 5 wt% Ru/Al₂O₃ catalysts, the Ru₂Co₁/S15-SU catalyst prepared by the SEA-UR method exhibited excellent MBT hydrogenation activity and stability and could effectively reduce the MBT hydrogenation temperature by 30 °C. It showed better hydrogenation performance in catalyzsing the NEC, DBT and acenaphthene hydrogenation reactions, and is highly applicable.

2. Experimental

2.1. Materials

The reagents and experimental materials used are described in detail in the supplementary material.

2.2. Catalyst preparation

Ru_xCo_y/S15-SU catalysts were prepared from SEA-UR, and the preparation process is described in Fig. 1.

• Step 1: Surface treatment of the carrier SBA-15 molecular sieve.

Fig. 1.

Schematic diagram showing the preparation of the Ru_xCo_y/S15-SU catalysts.

0.5 g of SBA-15 was dispersed in 30 mL of 0.01 M NaOH solution for 30 min, and Si-OH on the surface of SBA-15 reacted with NaOH to form Si-O^--Na⁺ (referred to as S15-ONa), which was separated by centrifugation to obtain the S15-ONa precipitate, which was washed with deionized water and the pH of the supernatant was maintained between 8 and 9.

• Step 2: Preparation of SBA-15 loaded RuCo alloy nanoparticles.

The S15-ONa obtained in step I was dispersed in deionized water, and 0.01 mol/L aqueous solutions of RuCl₃ and CoCl₂ were put simultaneously to the mixture according to the Ru/Co molar ratios(x/y) of 0.5–3. In this process, Ru³⁺ and Co²⁺ are electrostatically adsorbed with Si-O^- in S15-ONa to form (Si-O^-)₃-Ru³⁺ and (Si-O^-)₂-Co²⁺, such that Ru³⁺ and Co²⁺ are adsorbed on the surface of SBA-15. The ultrasonic in-situ reduction of Ru³⁺ and Co²⁺ was carried out under 100–300 W ultrasonication for 0.5–3 h in a N₂ atmosphere. SBA-15-loaded RuCo alloy nanoparticles (Ru_xCo_y/S15-SU) were obtained by centrifugation, washing, and vacuum-drying at 60 °C, where the total loading of Ru metal with Co was 1 wt%.

The Ru/S15-SU and Co/S15-SU catalysts were produced by the above technology for comparison. The loading of both metal Ru and Co was 1 wt%.

2.3. Catalytic hydrogenation experiment

The catalytic hydrogenation performances of MBT, NEC, DBT and acenaphthene were evaluated using a 50 mL stainless steel autoclave. The catalyst and reactants were added to an autoclave reactor according to a mass ratio of (Ru + Co)/LOHCs of 0.3 wt% and 20 mL of cyclohexane was added as solvent. After sealing the reactor, the reactor was replaced three times with H₂, and then heated up to the specified temperature, and then filled with H₂ at the specified temperature reactor. The H₂ pressure changes over time were recorded. When the H₂ pressure doesn't change, the reaction was terminated. The products of LOHC hydrogenation were analyzed via GC gas chromatography. Analysis method and result calculation method were shown in supplementary material.

2.4. Catalyst characterization

The detailed characterization methods are described in the Supporting Materials.

3. Results and discussion

3.1. Optimization of the catalyst preparation process

In order to optimize the optimal molar ratio of Ru to Co and the preparation conditions for the SEA-UR method, the catalytic hydrogenation performance of Ru_xCo_y/S15-SU catalysts prepared with different ultrasonic time and power was evaluated using the MBT hydrogenation reaction as a probe reaction at 110 °C and 6 MPa H₂, and the experimental results are seen in Fig. 2. The actual metal content and dispersion of the metal NPs in Ru_xCo_y/S15-SU were determined via ICP analysis and CO-chemisorption, respectively. The TOF of each catalyst is shown in Table 1.

Fig. 2.

(a) Catalytic MBT hydrogenation profile over the Ru_xCo_y/S15-SU catalysts, and performance of Ru₂Co₁/S15-SU catalyzed MBT hydrogenation prepared with different (b) ultrasonic time and (c) ultrasonic power.

Table 1.

Actual metal content, dispersion and TOF values for the Ru_xCo_y/S15-SU catalysts.

Sample	Actual content (wt%)a		Actual Ru/Co (mole ratio)	D(%)b	TOF(min−1)c
	Ru	Co
Ru3Co1/S15-SU	0.83	0.16	3.02:1	71	105
Ru2Co1/S15-SU	0.76	0.22	2.01:1	76	145
Ru1Co1/S15-SU	0.62	0.36	1.00:1	44	49
Ru1Co2/S15-SU	0.45	0.53	0.495:1	35	26
Ru/S15-SU	0.98			41	33
Co/S15-SU	0.99			34	0

^aActual metal content analyzed by ICP;

^bMetal dispersion measured by CO-chemisorption;

^cThe reaction was carried out under 110 °C, 6 MPa H₂;

From Fig. 2(a) and Table 1, the hydrogenation catalytic performance of MBT over Ru_xCo_y/S15-SU exhibited the tendency to rise and then fall with increasing Co addition. The prepared Ru₂Co₁/S15-SU showed the optimum hydrogenation performance. The conversion of MBT and selectivity of 12H-MBT reached 100 % and 99.90 %, respectively, and hydrogen storage efficiency was 99.98 % after 26 min, which was much higher than that of the monometallic Ru/S15-SU. The TOF value reached 145 min⁻¹. Because Co NPs are very easily oxidized to CoO in air, Co/S15-SU catalyst, hydrogenation reaction of MBT does not basically occur at 110 °C. The catalytic activity of Ru₁Co₂/S15-SU was inferior to that of Ru/S15-SU probably due to the excessive Co content. Some of the Co NPs were oxidized to CoO without interacting with the Ru NPs, which resulted in the agglomeration of the Co NPs, leading to a lower metal dispersion as well as a lower number of active components than that of Ru/S15-SU. Some of the Co in the Ru₁Co₁/S15-SU catalyst interacted with Ru, resulting in a slightly higher metal dispersion than that of Ru/S15-SU, which may also explain why its catalytic performance was slightly better than Ru/S15-SU. The significant improvement in the catalytic performance of Ru₂Co₁/S15-SU compared to that of Ru₁Co₁/S15-SU could be attributed to the introduction of a moderate amount of Co, which lowered the particle size of the metal particles, greatly increased the metal dispersion, and enhanced the antioxidant capacity of the RuCo alloy NPs. The electron transfer between Ru and Co and the greater number of catalytically active sites greatly improved the catalytic performance.

As can be seen from Fig. 2(b), the catalytic MBT hydrogenation performance first increased and then decreased with increasing ultrasonic time, and the optimal ultrasonic time was 1 h. When the ultrasonic time is too short (0.5 h), the number of hydrogen radicals (•H) generated by –OH on the surface of the ultrasonically excited carrier is not enough to reduce the loaded Ru³⁺ and Co²⁺ completely; when ultrasonication is performed for 1 h, the RuCo alloy NPs have formed completely and can be ligated with hydroxyl groups on the surface of SBA-15, which stabilizes the RuCo alloy NPs, and then continued ultrasonication will cause the RuCo alloy NPs to fall off and produce agglomerations, resulting in a decrease in the number of exposed active sites and a decrease in the hydrogenation capacity. As can be seen from Fig. 2(c), with increasing of ultrasonic power, the hydrogenation performance of catalytic MBT also increases first and then decreases, and the optimal ultrasonic power is 300 W. This is due to the fact that too little ultrasonic power produces too little •H and slows the reduction, while too much causes the RuCo alloy NPs to fall off and agglomerate, and also causes the pore structure of SBA-15 to become damaged, which reduces the catalytic MBT hydrogenation reaction performance.

3.2. Catalyst characterization

3.2.1. XPS

XPS characterization of the above prepared catalysts was completed to study chemical states of Ru, Co and O elements. The results are seen in Fig. 3.

Fig. 3.

(a) Full spectrum, (b) Ru 3p, (c) Co 2p, and (d) O 1s XPS spectra of Ru/S15-SU and Ru₂Co₁/S15-SU.

It was known from the Ru 3p spectrum of the catalyst in Fig. 3(b) that for the Ru₂Co₁/S15-SU sample, the binding energies (BE) of 486.9 and 464.8 eV are attributed to Ru⁴⁺, and those at 485.2 and 463.1 eV are attributed to Ru⁰. The BE of Ru⁰ was shifted by 0.7 eV toward higher BE than that of the monometallic Ru/S15-SU catalyst. It was seen from the Co 2p spectrum of the catalyst in Fig. 3(c) that for the Ru₂Co₁/S15-SU sample, BE corresponds to Co²⁺ at 801.2 and 785.8 eV, Co³⁺ at 798.4 and 782.9 eV, and Co⁰ at 796.4 and 780.6 eV. The BE of Co⁰ is shifted toward the lower BE by approximately 1.2 eV compared to that of monometallic Co catalysts. The elevated BE of Ru 3p and decreased BE of Co 2p indicate electron transfer from Ru to Co atoms, showing the existence of strong electronic interactions between Ru and Co in Ru₂Co₁/S15-SU, and the possible formation of RuCo alloys [50], [51]. Among them, the BE of Ru⁴⁺ in Ru₂Co₁/S15-SU is shifted 0.7 eV higher, probably because part of Ru is coordinated with O to form Ru^δ+. It can be seen from the BE data of all the samples in Table 2 that the ratios of Ru⁰/Ru⁴⁺ and Co⁰/(Co²⁺ + Co³⁺) in the Ru₂Co₁/S15-SU catalysts are significantly increased compared with those of the monometallic Ru/S15-SU and Co/S15-SU catalysts, which shows that the formation of RuCo alloy structure makes Ru and Co less susceptible to oxidation.

Table 2.

XPS binding energy data for Co/S15-SU, Ru/S15-SU and Ru₂Co₁/S15-SU.

Catalysts	Ru 3p1/2 (eV)		Ru 3p3/2 (eV)		Co 2p1/2 (eV)			Co 2p3/2 (eV)			Ru0/Ru4+	Co0/ (Co2+ + Co3+)
	Ru0	Ru4+	Ru0	Ru4+	Co2+	Co3+	Co0	Co2+	Co3+	Co0
Ru/S15-SU	484.5	486.2	462.4	464.1	–	–	–	–	–	–	1.67	–
Co/S15-SU	–	–	–	–	802.7	799.9	797.3	787.7	784.2	781.4	–	0.51
Ru2Co1/S15-SU	485.20	486.9	463.10	464.80	801.2	798.41	796.4	786.8	782.89	780.5	2.26	0.83

Fig. 3(d) shows the O 1s spectrum of the catalysts. For the Ru₂Co₁/S15-SU sample, the BE at 533.5 eV corresponds to the complexation of the RuCo alloy with O in HO groups on the surface of the carriers, the BE at 532.7 eV corresponds to the Si-O-Si on SBA-15, and the BE at 532.1 eV corresponds to the O-H in the adsorbed H₂O [52], [53], [54]. The individual BE of O 1s in Ru₂Co₁/S15-SU sample is shifted by about 0.2 eV toward a higher BE than that in the monometallic Ru/S15-SU, suggesting that the coordination between the RuCo alloy and the O in HO groups on the surface of the carrier SBA-15 is stronger in the bimetallic RuCo catalyst. The data in Table S1 show that the peak area ratio of the RuCo-O diffraction peaks in the Ru₂Co₁/S15-SU sample is 0.71, while the peak area ratio of the Ru-O diffraction peaks in the Ru/S15-SU sample is only 0.39. The complexation of the RuCo alloy with O in the carrier Si-O^- shows that the UR process not only induces the reduction of Ru and Co to form an alloy, but also causes the RuCo alloy to complex with O on the carrier to produce a ligand effect, which improves the catalytic stability.

3.2.2. TEM

To study the particle size and metal dispersion of RuCo bimetallic nanoparticles on Ru₂Co₁/S15-SU, Ru₂Co₁/S15-SU was characterized via TEM. The results are seen in Fig. 4.

Fig. 4.

(a) TEM image and (b) particle size distribution of Ru/S15-SU, (c) TEM image and (d) particle size distribution and (e) HRTEM image and (f-h) TEM mapping and (i) line-scan spectrum and (j) EDS spectrum and (k) elemental content of Ru₂Co₁/S15-SU.

It was known from Fig. 4(a-d) that the metal nanoparticle particle size in the Ru₂Co₁/S15-SU catalyst was 2.07 nm, which is smaller than the 3.12 nm in the monometallic Ru/S15-SU. The introduction of Co significantly improved the metal dispersion (Table 1), which obviously improved the hydrogenation performance of the Ru₂Co₁/S15-SU catalyst. It can be seen from HRTEM image in Fig. 4(e), the lattice spacing(d = 0.226 nm) of the RuCo bimetallic nanoparticles is between 0.204 nm for the monometallic Ru (1 0 1) crystallites and 0.244 nm for the monometallic Co (1 1 1) crystallites. It belongs to that for RuCo alloy (1 0 0) crystallites and conforms to Vegard's law, which indicates that the RuCo bimetallic nanoparticles are alloy structures. The line-sweep spectra in Fig. 4(i) further prove that the bimetallic nanoparticles formed on Ru₂Co₁/S15-SU are RuCo alloys, which corresponds to the XPS results. The RuCo alloy NPs are also much uniformly dispersed on the surface of SBA-15 according to the TEM elemental mapping images in Fig. 4(f-h). A higher metal dispersion indicates that there are more active catalytic sites and a higher reactivity of the catalyst [55], [56]. It was known from data in Table 1 that the dispersion of metal alloy NPs in Ru₂Co₁/S15-SU was 76 %, which was much higher than the 41 % of Ru NPs in Ru/S15-SU, suggesting that the addition of a suitable amount of Co can dramatically increase the dispersion of alloy metal NPs and thus improve the catalytic hydrogenation performance. Fig. 4(j) shows the EDS spectrum corresponding to Fig. 4(f). The result of EDS elemental analysis was Ru(mol):Co(mol) = 2.3:1, which was slightly higher than the actual addition amount. The Ru and Co contents in the catalyst were tested by using ICP, as seen in Table 1, which shows that the actual measurement results are in line with the actual addition amount.

3.2.3. XRD

To investigate the structural and morphological changes of the carrier SBA-15 during the preparation process, all the samples were characterized by low and wide angle XRD, as seen in Fig. S1.

As shown in Fig. S1(a), the diffraction peak of SBA-15 appearing at 2θ of 0.93° corresponds to the characteristic diffraction peak of the 2D hexagonal P6mm (1 0 0) mesoporous symmetric structure [57], [58], [59]. Compared with that of SBA-15, the (1 0 0) diffraction peak of the Ru₂Co₁/S15-SU catalyst shifted to a lower angle and the relative intensity of the diffraction peak decreased. This result suggests that microjets and shock waves generated by alkali treatment during electrostatic adsorption and ultrasonication during sonication may lead to shrinkage or structural disorganization of the mesopores of SBA-15. Nevertheless, the (1 0 0) diffraction peak of the Ru₂Co₁/S15-SU catalyst is still clearly observed, indicating that SBA-15 can still maintain an ordered mesoporous structure after electrostatic adsorption and sonication treatment [58].

As shown in Fig. S1(b), diffuse peaks with 2θ angles in the range of 15-37° are clearly observed for all the samples, which are attributed to the amorphous skeleton of silica in SBA-15 and are characteristic diffraction peaks of SBA-15 [57], [60], [61]. For all, no characteristic diffraction peaks belonging to Ru (JCPDS Card No. 06–0663) or Co (JCPDS Card No. 05–0727) were observed in the XRD spectra, which is attributed to the lower Ru and Co loadings below the XRD detection limit [62], [63] or to the highly homogeneous dispersion of metal Ru or Co NPs within the backbone of SBA-15 [64].

3.2.4. N₂ adsorption–desorption

As an ordered mesoporous silica material, the pore properties of SBA-15 have an important influence on the diffusion of reactants and the dispersion of metal nanoparticles. The pore structures of all the samples were investigated using N₂ adsorption–desorption, as shown in Fig. 5 & Table 3.

Fig. 5.

(a) N₂ adsorption–desorption isotherms and (b) pore size distributions of the samples.

Table 3.

Structural parameters of each sample.

Samples	BET surface area (m2/g)	Pore volume (cm3/g)	Average pore diameter (nm)
SBA-15	671.25	1.082	6.448
Co/S15-SU	606.51	1.037	7.155
Ru/S15-SU	605.54	1.041	7.106
Ru2Co1/S15-SU	634.41	1.046	7.079

It was known from Fig. 5(a) that all the samples are type IV isotherms, and all of them have H3 hysteresis return lines, which indicates that all the samples still maintain typical mesoporous structures. Moreover, the SEA-UR method did not change the mesoporous structure of the carrier, which was also corroborated by the XRD results. The data in Fig. 5(b) and Table 3 show that the specific surface area of both the monometallic and bimetallic catalysts reduced and the average pore size enhanced compared to SBA-15, which was due to the reason that a small amount of NaOH reacted with free and marginal SiO₂ in the pores to form Na₂SiO₃ during the SEA process. According to our previous studies, when metal catalysts are prepared by SEA using SBA-15 as a carrier, metal nanoparticles are anchored to Si-OH on the inner surface [65], [66]. The pore volume of both the monometallic and bimetallic catalysts decreased, which indicated that the Ru or Co NPs and RuCo alloy NPs successfully entered the pores of the SBA-15, which was confirmed by TEM characterization.

Compared to the monometallic Ru and Co catalysts, the increased specific surface area of Ru₂Co₁/S15-SU is attributed to the introduction of Co which improves the metal dispersion and the accumulation of highly dispersed RuCo alloy NPs produces more intergranular mesopores [67]. The slightly higher pore volume of the Ru₂Co₁/S15-SU catalyst was because the RuCo alloy NPs had a much smaller particle size (2.07 nm) and were highly dispersed on the inner surface of the pores, as confirmed by the particle size distribution, mapping, and CO-chemisorption results obtained via TEM characterization. The average pore size of Ru₂Co₁/S15-SU is slightly lower than that of the monometallic Ru/S15-SU and Co /S15-SU because there are more RuCo alloy NPs in the bimetallic catalyst that are anchored to the inner surface of the pores, and the RuCo alloy NPs will cover more smaller pore diameters of the holes in the carrier SBA-15.

3.2.5. FT-IR

To study the variation in the functional groups on the surface of SBA-15 during catalyst preparation, the above samples were analyzed via FT-IR, as shown in Fig. 6.

Fig. 6.

(a) FT-IR spectra of all the samples and (b) partial graph of the region 700–1400 cm⁻¹.

As shown in Fig. 6(a), bending, symmetric and antisymmetric stretching of the Si-O-Si vibrations present in the SBA-15 backbone are observed at 462, 806 and 1082 cm⁻¹, respectively, in all the samples [68]. The absorption peak at 967 cm⁻¹ belongs to the symmetric telescopic vibration of the silanol Si-O-H group [69]. The peaks at 3430 cm⁻¹ and 1630 cm⁻¹ are due to the telescopic and bending vibrations of O-H in H₂O physically adsorbed by the sample.

Compared with SBA-15, the peak intensities at 3430 cm⁻¹,1630 cm⁻¹ and 967 cm⁻¹ in all samples were weakened to varying degrees, which was due to the disruption of Si-OH on the SBA-15 surface after alkali treatment and the disruption of O-H groups in the physically adsorbed H₂O by the acoustic cavitation effect or micro-jet from ultrasonication treatment.

The magnified FT-IR spectrum in Fig. 6(b) shows that the vibrational peak intensity of S15-ONa at 967 cm⁻¹ is the weakest, which was attributed to the formation of Si-O^--Na⁺ from the reaction of Si-OH at the SBA-15 surface with NaOH. The flame atomic absorption spectroscopy results of S15-ONa in Table S2 show that the sodium content was 0.95 wt%, which also indicates Si-O^--Na⁺ formation. Upon the addition of RuCl₃ and CoCl₂ solutions, Ru³⁺ and Co²⁺ ions electrostatically adsorbed with Si-O^--Na⁺ to form (Si-O^-)₃-Ru³⁺ and Co²⁺-(Si-O^-)₂. The slight peak intensity increase at 967 cm⁻¹ for monometallic Ru and Co catalysts, which was due to the generation of hydrogen radicals (•H) by H₂O under the transient high temperature and pressure from ultrasonication to reduce a portion of (Si-O^-)₃-Ru³⁺ or Co²⁺-(Si-O^-)₂, while the generated free H⁺ was attracted by a portion of the not electrostatically adsorbed Si-O^- to regenerate Si-OH, which led to a slight increase in the amount of Si-OH.

The peak intensity of the Si-OH group in Ru₂Co₁/S15-SU further raised, which was attributed to the fact that during UR process, •H generated by H₂O first reduced Ru³⁺ to Ru, generating more free H⁺ to regenerate Si-OH with Si-O^- and part of the Si-OH under ultrasonic conditions to generate •H and Si-O^-. Afterwards, •H reduced Co²⁺-(Si-O^-)₂ to Co–(Si-O)₂ and simultaneously generated more free H⁺ to regenerate Si-OH with Si-O^-, which further increased the peak intensity of Si-OH. The changes in the Si-OH absorption peaks in the FT-IR spectra indicated that Ru³⁺ and Co²⁺ were successfully electrostatically adsorbed, and the (Si-O)₃-Ru and Co–(Si-O)₂ generated by ultrasonic in-situ reduction interacted to produce the (Si-O)₃-RuCo-(Si-O)₂ alloy. The RuCo alloy undergoes a coordination effect with Si-OH on SBA-15, which is firmly anchored to the carrier surface and greatly improves the catalytic hydrogenation stability of the Ru₂Co₁/S15-SU catalyst, as evidenced by the O 1s spectra in the XPS characterization.

3.2.6. H₂-TPR

To study the metal reducibility, the role of Co introduction, and the interaction between Ru and Co in the above catalysts, the precursors of the monometallic and bimetallic catalysts were analyzed using hydrogen programmed temperature reduction (H₂-TPR), as seen in Fig. 7.

Fig. 7.

H₂-TPR spectra of Ru_xCo_y/S15-SU, Co/S15-SU & Ru/S15-SU catalyst precursors.

From Fig. 7, the Co/S15-SU catalyst showed a broad reduction peak between 200–500 °C, which was fitted by three peaks. The peak at 222 °C is attributed the reduction of free CoO, that at 306 °C is attributed the reduction of Co³⁺ to Co²⁺, and that at 408 °C is attributed the reduction of the Co species that interact with Si-OH on the SBA-15 surface. Ru/S15-SU showed reduction peaks at 125 °C and 489 °C, respectively, representing the reduction of free Ru (RuO₂) and the reduction of Ru which strongly interacts with Si-OH on the SBA-15 surface [66].

Compared to the monometallic Ru and Co catalyst precursors, the low-temperature reduction peak of the bimetallic RuCo catalyst precursor is between 125 °C and 222 °C because free Ru species are first reduced to Ru monomers at lower temperatures. The dissociated hydrogen adsorbed on the Ru monomers spills over into the free Co oxides, resulting in the Co oxides being reduced to Co monomers at lower temperatures [70]. The peak area of the low-temperature reduction peak was slightly higher than that of the monometallic Ru catalyst precursor, suggesting that the introduction of Co allowed more free Ru species to produce hydrogen spillover, which facilitated the reduction of free Co oxides. The high-temperature reduction peak temperatures of the bimetallic RuCo catalyst precursor were lower than those of the monometallic Ru (489 °C) and Co (408 °C) catalyst precursors due to the generation of RuCo alloys in the bimetallic catalysts, where mutual electron transfer between Ru and Co occurs and interactions take place, which makes the reduction of RuCo alloys easier. The peak area of the high-temperature reduction peak was significantly higher than that of the monometallic Ru and Co catalyst precursors, which was attributed to the smaller particle size and more homogeneous dispersion of the alloy nanoparticles in the bimetallic RuCo catalysts, which had more RuCo alloy active metal sites. And the RuCo alloy produces a greater hydrogen overflow effect, which causes the alloy to undergo a stronger coordination effect with the hydroxyl group on the carrier SBA-15.

When the Ru/Co(mol) ratio is 3, the number of RuCo alloys formed is also smaller due to the lower actual content of Co, and the remaining Ru remains in its monometallic form. The reduction peak at 138 °C is attributed to the reduction of free Ru species (RuO₂), the reduction peak at 186 °C is attributed to the reduction of a small amount of free RuCoO_x species, and the reduction peak at 380 °C is attributed to the reduction of a small amount of RuCo alloy that strongly interacts with Si-OH on the surface of the carrier SBA-15. The reduction peak at 422 °C is attributed to the reduction of Ru species that interact strongly with Si-OH on the surface of the carrier SBA-15, and this reduction temperature is lower than that of monometallic Ru is due to the introduction of Co, which allows for increased hydrogen spillover to the Ru species, thus facilitating the reduction of the Ru species.

When the Ru/Co (mol) ratio is 2, the reduction peak at 130 °C is attributed to the reduction of the free Ru species (RuO₂). The reduction peak at 163 °C is attributed to the reduction of free RuCoO_x species, the reduction peak at 210 °C is attributed to the reduction of a very small amount of free CoO, and the reduction peak at 321 °C is attributed to the reduction of Ru-Co alloys that interact with the hydroxyl groups on the carriers (coordination effect). The percentage of the peak area of the reduced peak at 210 °C is only 5.15 %, and all the remaining Co species form RuCo alloys with Ru.

With a further increase in the Co content, when the Ru/Co (mol) ratio is 0.5, the actual content of Co is too much and the actual content of Ru is less, resulting in the formation of only a small amount of RuCo alloys, with the remaining Co species remaining in the monometallic form. The reduction peak at 152 °C is attributed to the reduction of free RuCoO_x species and the reduction peak at 201 °C is attributed to the reduction of free CoO species. The reduction peak at 285 °C is attributed to the reduction of RuCo alloys interacting with the carriers. This decrease in reduction temperature compared to that of Ru₂Co₁/S15-SU is partly due to the lower amount of RuCo alloys formed and partly due to the weaker interaction between Ru and Co in the RuCo alloys formed. The reduction peak at 360 °C is attributed to the reduction of Co species interacting with the carrier.

H₂-TPR characterization findings demonstrated that the introduction of Co into the bimetallic catalyst reduced the difficulty of metal reduction, and the generated RuCo alloy interacted more strongly with SBA-15, which improved the hydrogenation stability of the bimetallic catalyst.

3.2.7. H₂-TPD

To study the metal dispersion and interactions between the metals and the carriers, the above catalysts were characterized by CO-chemisorption and hydrogen programmed temperature desorption (H₂-TPD), as shown in Table 1 and Fig. 8.

Fig. 8.

H₂-TPD spectra of Ru₂Co₁/S15-SU, Co/S15-SU & Ru/S15-SU.

In general, the temperature and peak area of the hydrogen desorption peak directly represent the intensity and amount of hydrogen adsorbed by the metal loaded on the catalyst. The hydrogen desorption temperature indirectly reflects the interaction strength of hydrogen species with active metal sites, while the number of catalytic active sites and the metal dispersion correspond to the hydrogen desorption peak area [71]. Weak adsorption of hydrogen on the surface of the metal does not sufficiently activate the reactants, while stronger adsorption of hydrogen results in the generation of a stable structure with the reactants. Compared with the monometallic Ru and Co catalysts, the main peak temperature of the Ru₂Co₁/S15-SU catalyst (471 °C) is between those of the monometallic Ru catalyst (445 °C) and the Co catalyst (508 °C), which shows that the RuCo alloys were generated in Ru₂Co₁/S15-SU. The electron transfer between Ru and Co leads to the enhancement of the hydrogen adsorption by the metal Ru and the decrease in hydrogen adsorption by the metal Co, which results in an equilibrium of hydrogen adsorption at the active sites of RuCo alloys and is favorable for the catalytic hydrogenation reaction. The hydrogen desorption peak areas of the Ru₂Co₁/S15-SU catalyst were all more than those of the monometallic Ru/S15-SU and Co/S15-SU catalysts, which indicated that more active sites existed on Ru₂Co₁/S15-SU and the metal dispersion was higher. And the CO-chemisorption data also indicated that the metal alloy dispersion in the bimetallic catalysts was 76 %, which was much higher than the 41 % of Ru/S15-SU and 34 % of Co/S15-SU. TEM characterization similarly showed that the RuCo alloys were highly dispersed in the bimetallic catalysts.

The main desorption peaks for the monometallic and bimetallic catalysts were fitted to three peaks using a Gaussian function. The peak I is attributed to the desorption of hydrogen adsorbed on the surface of the Ru, Co or RuCo alloys. The area of peak I for the Ru₂Co₁/S15-SU catalyst was much larger than that for the monometallic Ru and Co catalysts, which also indicates more RuCo active sites. The peak III is thought to be a hydrogen overflow phenomenon resulting from hydrogen desorption from strong chemisorption with the catalyst surface. The higher peak intensity in Ru₂Co₁/S15-SU could be due to the RuCo alloy on the surface of the catalysts promoted hydrogen spillover, suggesting that more hydrogen was absorbed on the surface and that there was better metal dispersion. The desorption peak II, corresponds to the desorption of hydrogen chemisorbed at the Si-OH junction between the metal Ru, Co and RuCo alloys and the carrier SBA-15 surface. The intensity of the desorption peak for the Ru₂Co₁/S15-SU catalyst was higher than that for the monometallic catalysts, which suggested that there were greater metal-carrier interactions and stability for the bimetallic catalyst. H₂-TPD characterization data demonstrated that the addition of Co improved the dispersion of the metal particles, further strengthened the interaction between the RuCo alloy and SBA-15, enhanced the catalytic stability. The adsorption of the RuCo alloy with hydrogen occurs at equilibrium, which is favorable for catalytic hydrogenation.

3.3. SEA-UR process mechanism

Without any chemical reducing or stabilizing agents, SBA-15 supported RuCo alloy catalysts were prepared using the SEA-SU method. In conjunction with our previous report, there were not enough hydrogen radicals (•H) released from water to reduce metal cations under ultrasound [49]. It is known that a portion of •H originated from the surface –OH of SBA-15, as evidenced by the weakened intensity of the Si-OH peak at 967 cm⁻¹ in the FT-IR spectrum. The fabrication of Ru₂Co₁/S15-SU by the SEA-UR method involved the following steps. After SBA-15 surface was treated with NaOH, Si-O^--Na⁺ formed on its surface. The electrostatic adsorption of Si-O^- ions with Ru³⁺ and Co²⁺ generates (Si-O^-)₃-Ru³⁺ and Co²⁺-(Si-O^-)₂, respectively.

According to the literature, the reduction electrode potentials of Ru³⁺ and Co²⁺ are 0.40 eV and −0.28 eV, respectively, and the reduction potential difference of Ru³⁺ is larger than that of Co²⁺, and the reduction rate of Ru³⁺ is faster than that of Co²⁺ [72]. Under the action of ultrasound, –OH in water is first dissociated into strongly reducing •H, which first reduces (Si-O^-)₃-Ru³⁺ to generate (Si-O)₃-Ru and H⁺, which recombines with Si-O^- to generate Si-OH, which is then dissociated to produce •H. •H reduces the Co²⁺-(Si-O^-)₂ around the nucleated Ru NPs to Co–(Si-O)₂, and Co–(Si-O)₂ is fused with the first reduced (Si-O)₃-Ru phase during nucleation to form the alloy structure of (Si-O)₃-Ru-Co–(Si-O)₂. The RuCo alloy had a coordination effect with HO groups on the SBA-15 surface, which led to uniformly dispersed anchoring on the SBA-15 carrier, thus greatly enhancing the catalytic performance of the bimetallic catalyst.

Si-OH + NaOH → Si-O^--Na⁺ + H₂O

3Si-O^--Na⁺ + Ru³⁺ (Si-O^-)₃-Ru³⁺

2Si-O^--Na⁺ + Co²⁺ → Co²⁺-(Si-O^-)₂

H₂O → •H + •OH.

(Si-O^-)₃-Ru³⁺ + 3•H → (Si-O)₃-Ru + 3H⁺

H⁺ + Si-O^- → Si-OH

Si-OH → Si-O• + •H

Co²⁺-(Si-O^-)₂ + •H → Co–(Si-O)₂ + 2H⁺

(Si-O)₃-Ru + Co–(Si-O)₂ → (Si-O)₃-RuCo-(Si-O)₂

3.4. Catalytic hydrogenation performance of LOHCs

We compared the catalytic MBT hydrogenation performances of the optimal Ru₂Co₁/S15-SU catalyst and the commercial 5 wt% Ru/Al₂O₃ catalyst. Ru₂Co₁/S15-SU was much better than the commercial 5 wt% Ru/Al₂O₃ and the specific analyses are detailed in the Supplementary Material.

3.4.1. Optimization of the MBT hydrogenation reaction conditions

In order to optimize the MBT hydrogenation reaction conditions over Ru₂Co₁/S15-SU, experiments were completed at different temperatures and pressures. The TOF values of the catalysts at different temperatures were calculated and compared with Ru/S15-SU and commercial 5 wt% Ru/Al₂O₃ catalysts, as shown in Fig. 9 and Table S3.

Fig. 9.

(a) Influence of reaction temperature on hydrogen storage of MBT over Ru₂Co₁/S15-SU, comparison of MBT (b) conversion and hydrogen storage efficiency and (c) selectivity at different temperatures for commercial 5 wt% Ru/Al₂O₃, Ru/S15-SU & Ru₂Co₁/S15-SU, (d)effect of reaction pressure on hydrogen storage of MBT over Ru₂Co₁/S15-SU, comparison of MBT (e) conversion and hydrogen storage efficiency and (f) selectivity at different pressures for commercial 5 wt% Ru/Al₂O₃, Ru/S15-SU & Ru₂Co₁/S15-SU.

From Fig. 9(a-c) and Table S3, the MBT conversion, hydrogen storage efficiency and TOF tended to increase with the rise of reaction temperature. Only at 80 °C for 80 min, the MBT conversion can reached 100 % and the hydrogen storage efficiency has reached 99.49 %. The full hydrogen storage time is 26 min at 110 °C, and the reaction rate does not increase much when the temperature is increased, possibly because an increase in temperature will reduce the amount of dissolved hydrogen in the solvent. For commercial 5 wt% Ru/Al₂O₃ and Ru/S15-SU catalysts, complete hydrogen storage can be achieved at 110 °C for 80 min, which indicates that the Ru₂Co₁/S15-SU catalyst prepared by the SEA-UR method has excellent catalytic performance for MBT hydrogenation at lower temperatures. The reaction temperature for MBT hydrogenation can be effectively reduced by 20 °C.

The effect of H₂ pressure on the performance of MBT hydrogenation was researched at 110 °C and compared with Ru/S15-SU and commercial 5 wt% Ru/Al₂O₃ catalysts. From Fig. 9(d-f), the MBT conversion and hydrogen storage efficiency continue to increase with increasing H₂ pressure, thus greatly improving the catalytic performance of the bimetallic catalyst.

From Fig. 9(e, f), compared with Ru/S15-SU and commercial 5 wt% Ru/Al₂O₃ catalysts, the hydrogen storage efficiency of MBT over Ru₂Co₁/S15-SU catalysts can reach 99.57 % at a lower reaction pressure (4 MPa), which demonstrated that the excellent reactivity of the Ru₂Co₁/S15-SU catalyst was excellent reactivity. The hydrogen storage efficiency of MBT over Ru₂Co₁/S15-SU catalyst could reach 98.85 % under milder reaction conditions (70 °C and 6 MPa), because of the synergistic effect generated by the electron inter-transfer between Ru and Co in the bimetallic catalysts, the better dispersion and the greater number of highly active sites.

3.4.2. Kinetic study of the MBT dehydrogenation reaction

The kinetics of the MBT hydrogenation reaction catalyzed by Ru₂Co₁/S15-SU were studied. According to previous reports, the effects of external and internal diffusion can be neglected due to the high rotational speed (900 rpm) and ultrafine particle size (2.12 nm) of RuCo alloy nanoparticles during the reaction when calculating MBT hydrogenation kinetics [73], [74]. The Ea of Ru₂Co₁/S15-SU catalyzed MBT hydrogenation reaction was 41.51 kJ/mol based on the first-order kinetic model (r = -dc/dt = k P_H2) (Fig. 10), at 6 MPa H₂ and 90–120 °C. A comparison of the Ea of Ru₂Co₁/S15-SU and the reported catalysts for catalyzing MBT hydrogenation is shown in Table S4. As shown in Table S4, compared with the reported catalysts, the Ru₂Co₁/S15-SU catalysts have relatively mild reaction conditions, higher hydrogen storage and lower apparent activation energies.

Fig. 10.

(a) Reaction rate constants of MBT hydrogenation for Ru₂Co₁/S15-SU catalysts at 90-120 °C & (b) lnk-1/T curves.

3.4.3. Catalytic stability of the MBT dehydrogenation reaction

To verify that Ru₂Co₁/S15-SU is a heterogeneous catalyst for the catalytic hydrogenation of MBT, a thermal filtration experiment was carried out, as seen in Fig. 11(a).

Fig. 11.

(a) Thermal filtration experiments of Ru₂Co₁/S15-SU; cycle test of (b) Ru₂Co₁/S15-SU & (c) Ru/S15-SU.

After MBT hydrogenation started for 10 min, the autoclave was removed from the heating jacket and cooled quickly to terminate the reaction. The catalyst was separated by centrifugation, and the supernatant was heated to the above reaction temperature and continued to react for 70 min. From Fig.11(a), the hydrogen storage efficiency remained unchanged throughout the thermal filtration experiments, which indicated that the leaching of metal nanoparticles did not occur during the hydrogenation reaction.

We examined the stability of Ru₂Co₁/S15-SU in catalyzing MBT hydrogenation under 110 °C and 6 MPa H₂, and compared it with the monometallic Ru/S15-SU, and the results are shown in Fig. 11(b-c). From Fig. 11(b) that Ru₂Co₁/S15-SU catalyst maintained its high hydrogenation performance after ten cycles, with a 12H-MBT selectivity of 98.11 % and a hydrogen storage efficiency of 99.70 %. In contrast, after ten cycles, the monometallic Ru/S15-SU catalyst showed only 12H-MBT selectivity of 69.54 % and hydrogen storage efficiency of 95.17 %.

Ru₂Co₁/S15-SU catalyst has excellent stability, which is due to the fact that the introduction of Co promotes the stronger anchoring effect of the RuCo alloy with HO groups on the SBA-15 surface. TEM images of all the catalysts after ten cycles are shown in Fig. S3. The average particle size of the RuCo alloy NPs in Ru₂Co₁/S15-SU increased from 2.12 nm to 3.44 nm after ten cycles of use, whereas the Ru NPs in the monometallic Ru/S15-SU underwent significant agglomeration and the particle size increased to 4.90 nm, which result in obviously decrease in catalytic performance of Ru/S15-SU. The agglomeration phenomenon occurs because the metal nanoparticles prepared by the electrostatic adsorption ultrasonic in-situ reduction method are attached to the inner surface of the pores of the SBA-15 carrier, and as the number of recycling cycles increases, the interaction of some metal nanoparticles with the surface of the carrier becomes weaker, and detachment occurs. Due to the narrow pores of SBA-15, the shed metal nanoparticles cannot be discharged from the pores in time, and they will gather around the un-shed metal nanoparticles and form large agglomerates, which makes the particle size of the metal nanoparticles become larger.

3.4.4. Investigation of the adaptability of Ru₂Co₁/S15-SU catalyst

To further study the adaptability of Ru₂Co₁/S15-SU for other LOHCs, the catalytic hydrogenation performances of NEC, DBT and acenaphthene were investigated with the same amount of catalyst as MBT.

The performance of Ru₂Co₁/S15-SU catalyzed NEC hydrogenation was investigated between 80 °C and 110 °C, as shown in Fig. 12(a, b). As shown in Fig. 12(a, b), the catalytic hydrogenation performance of NEC improved with the rise of reaction temperature. The hydrogen storage efficiency was 99.90 %, and the NEC could be completely hydrogenated at a reaction temperature of 100 °C for 26 min.

Fig. 12.

(a) Hydrogen storage curves and (b) product selectivity and hydrogen storage efficiency for Ru₂Co₁/S15-SU catalyzed NEC hydrogenation, (c) hydrogen storage curves and (d) product selectivity and hydrogen storage efficiency for Ru₂Co₁/S15-SU catalyzed DBT hydrogenation, (e) hydrogen storage curves and (f) product selectivity and hydrogen storage efficiency for Ru₂Co₁/S15-SU catalyzed acenaphthene hydrogenation at different temperatures.

The Ru₂Co₁/S15-SU catalyzed DBT hydrogenation performance was investigated between 100 °C and 160 °C, as shown in Fig. 12(c, d). As shown in Fig. 12(c, d), the catalytic performance of DBT hydrogenation enhanced with the rise of reaction temperature. The hydrogen storage efficiency was 98.83 % at 160 °C for 80 min.

The catalytic acenaphthene hydrogenation performance of Ru₂Co₁/S15-SU was investigated between 110 °C and 140 °C, as shown in Fig. 12(e, f). It can be shown from Fig. 12(e, f), the catalytic performance of acenaphthene hydrogenation enhanced with the rise of reaction temperature. The hydrogen storage efficiency was 93.13 % at 140 °C for 80 min.

In summary, the Ru₂Co₁/S15-SU catalyst prepared by the SEA-UR method is suitable for the hydrogen storage of various LOHCs and shows good adaptability.

4. Conclusion

In this study, RuCo alloy catalysts with high activity, high stability and strong applicability were prepared by an environmentally friendly method. The hydrogen storage efficiency of MBT over the Ru₂Co₁/S15-SU catalyst was 99.98 % at 110 °C and 6 MPa H₂, and 98.85 % at 70 °C. Compared with Ru/S15-SU and Ru/Al₂O₃, the Ru₂Co₁/S15-SU catalyst effectively reduced the MBT hydrogenation temperature by 30 °C. The Ru₂Co₁/S15-SU catalyst also showed strong applicability, achieving hydrogen storage efficiencies of 99.90 %, 98.83 % and 93.13 % for NEC, DBT and acenaphthene, respectively. The excellent catalytic hydrogenation performance of the Ru₂Co₁/S15-SU catalysts is because introduction of Co improves the dispersion and antioxidant capacity of the RuCo alloy NPs. The formation of the RuCo alloy structure improved the strong interactions of Ru and Co from electron transfer, and the stronger coordination effect of the well-dispersed RuCo alloy and OH on the surface of SBA-15, which resulted in the catalysts having a more excellent stability. The SEA-UR technology provides a promising preparation pathway for loaded alloy catalysts with low-cost, high-activity, high-stability, and high-applicability in the field of LOHCs hydrogen storage.

CRediT authorship contribution statement

Taiyi Liu: Writing – original draft, Visualization, Software, Investigation, Formal analysis, Data curation. Wei Wu: Funding acquisition, Formal analysis, Conceptualization, Methodology, Project administration, Resources, Supervision, Writing – review & editing. Xuefeng Bai: Writing – review & editing, Supervision, Resources, Conceptualization, Formal analysis, Funding acquisition, Methodology, Project administration.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article: