The Catalytic Consequence of Isolated Ni Single-Atoms in BEA Zeolite for Hydrogen Production and Olefin Conversion

Abstract

In our previous work, we fabricated Ni single-atoms within Beta zeolite (Ni₁@Beta-NO₃⁻) using NiNO₃·6H₂O as a metal precursor without any chelating agents, which exhibited exceptional performance in the selective hydrogenation of furfural. Owing to the confinement effect, the as-encapsulated nickel species appears in the form of Ni⁰ and Ni^δ+, which implies its feasibility in metal catalysis and coordination catalysis. In the study reported herein, we further explored the hydrogen production and olefin oligomerization performance of Ni₁@Beta-NO₃⁻. It was found that Ni₁@Beta-NO₃⁻ demonstrated a high H₂ generation turnover frequency (TOF) and low activation energy (E_a) in a sodium borohydride (NaBH₄) hydrolysis reaction, with values of 331 min⁻¹ and 30.1 kJ/mol, respectively. In ethylene dimerization, it exhibited a high butylene selectivity of 99.4% and a TOF as high as 5804 h⁻¹. In propylene oligomerization, Ni₁@Beta-NO₃⁻ demonstrated high selectivity (75.21%) of long-chain olefins (≥C₆⁺), overcoming the problem of cracking reactions that occur during oligomerization using H-Beta. Additionally, as a comparison, the influence of the metal precursor (NiCl₂) on the performance of the encapsulated Ni catalyst was also examined. This research expands the application scenarios of non-noble metal single-atom catalysts and provides significant assistance and potential for the production of H₂ from hydrogen storage materials and the production of valuable chemicals.

1. Introduction

Metal nanoparticles and their compounds represent a significant category of industrial catalysts. However, their application has been severely restricted because of the issue of sintering and deactivation during thermal treatment or catalytic processes [1,2]. In recent years, encapsulating metal species within the micropores of zeolites has been employed to suppress the movement and aggregation of metal species through the confinement effect of the rigid zeolite framework. This strategy can significantly enhance their resistance to sintering [3,4]. Moreover, these catalysts inherit the unique shape-selective catalytic capabilities of zeolites. They have been extensively applied in reactions such as hydrogenation, oxidation, isomerization, cracking, and reforming, demonstrating superior activity, stability, and selectivity [5,6,7]. At present, to address the issue of low metal loading (<2 wt%) within zeolites, zeolite-encapsulated single-atom metal catalysts have been proposed and found to be highly efficient catalysts.

Hydrogen, as a promising alternative energy source, offers a viable solution to the increasingly pressing energy challenges. In recent years, liquid-phase hydrogen storage materials (e.g., borohydride and ammonia borane solution) have attracted extensive research interest because they can release hydrogen rapidly, conveniently, and controllably in aqueous solutions under the action of suitable catalysts [8,9]. Yu et al. [10] employed a direct hydrogen reduction strategy under in situ hydrothermal conditions to successfully encapsulate Rh single-atoms within Silicate-1 zeolite (Rh@S-1) for the hydrolysis of ammonia borane. The as-prepared Rh@S-1-H exhibited a TOF as high as 432 min⁻¹, which is twice that of Rh nanoparticles on S-1 (195 min⁻¹). The aforementioned reports highlight that noble metal-based catalysts encapsulated in zeolites exhibit superior hydrogen production performance. However, there are currently few reports on the use of non-noble metal single-atoms encapsulated in zeolites for hydrogen production reactions.

Light olefin (ethylene or propylene) oligomerization is a series–parallel reaction for the production of a wide range of linear and branched alkylene products (C₄–C₁₂) suitable as transportation fuels, which involves β-scission and co-oligomerization reaction pathways [11]. It has also been frequently used as a probe reaction to identify the confinement effects in zeolite catalysis [12]. Nozik et al. [13] found that the higher ethylene oligomerization activity and selectivity to even-carbon-numbered alkenes of Ga/H-MFI stems from cooperative effects between Ga³⁺ sites and Brønsted (B) acid sites (H⁺). Kitamura et al. [14] found that the Ni²⁺ cations grafted onto the silanols in dealuminated zeolites containing almost no zeolitic acid show high activity for ethylene oligomerization. Therefore, we infer that the combination of Ni single-atoms and acid sites in Beta zeolite should be a good choice for olefin oligomerization. In this study, we focused on the catalytic consequence of Ni₁@Beta-NO₃⁻ in the hydrolysis of NaBH₄, ethylene dimerization, and propylene oligomerization.

For the synthesis of Ni single-atom catalysts within zeolite, we previously fabricated Ni single-atoms within Beta zeolite (Ni₁@Beta-NO₃⁻) using NiNO₃·6H₂O as a metal precursor without any chelating agents [15]. The initial separate nucleation step allows for a strong interaction between Ni²⁺ cations and the alumino-silicate gel, resulting in the formation of nickel silicate and achieving high dispersion. This process is not only related to the pH but also dependent on the anions in the precursors. Escola et al. [16] prepared Ni/Beta using chloride, nitrate, acetylacetonate, and tris(ethylenediamine) nickel(II) chloride (TEDAN) as precursors. It was found that Ni nitrate and acetylacetonate led to smaller and more dispersed particles, thus closer to the acidic sites, driving higher hydrogenation and hydrocracking activity in the hydroreforming of low-density polyethylene thermal cracking oil. In contrast, TEDAN and nickel chloride resulted in larger nickel particles, which hindered the rapid hydrogenation of intermediate species, thus favoring the production of aromatics and branched hydrocarbons. In this study, we synthesized Ni@Beta-Cl⁻ with NiCl₂ as the metal precursor, following the same methods used for the Ni₁@Beta-NO₃⁻ sample. Their catalytic consequence in propylene oligomerization was compared to point out the effect of anions for metal precursors in the encapsulation of Ni single-atoms within zeolite.

2. Results and Discussion

2.1. Hydrolysis of NaBH₄

We previously found that, for the Ni₁@Beta catalyst, the hydrogenation of furfural involves not only the activation of hydrogen but also the selective activation of furfural. It should be noted that single-atoms confined within micropores may be subject to diffusion effects, which may not fully reflect the characteristics of isolated single-atoms. To further ascertain the hydrogen activation capability of single-atoms, the NaBH₄ hydrolysis reaction for hydrogen production is used as a probe to evaluate the performance of the single-atom Ni catalyst (Figure 1 and

Table 1. Comparison of the catalytic hydrogen production rate of different samples.

Samples	293 K		298 K		303 K		308 K		313 K
	HGR (mL H2 min−1)	Time (min)	HGR (mL H2 min−1)	Time (min)	HGR (mL H2 min−1)	Time (min)	HGR (mL H2 min−1)	Time (min)	HGR (mL H2 min−1)	Time (min)
Ni/Beta-NO3−	35.4	4.7	50.1	3.4	65.6	2.6	87.8	2.0	109.1	1.6
Ni1@Beta-NO3−	68.5	2.4	81.0	2.1	103.4	1.7	124.8	1.4	145.7	1.2

). As shown in Table 1, the hydrogen generation rates (HGRs) can be improved with the increase in reaction temperature. What is more, Ni₁@Beta-NO₃⁻ exhibits higher HGRs than Ni/Beta-NO₃⁻ at the tested temperatures. In addition to HGR, E_a is a key parameter to evaluate the performance of catalysts for H₂ generation. Kinetic studies revealed a much lower E_a on Ni₁@Beta-NO₃⁻ of 30.1 kJ/mol than Ni/Beta-NO₃⁻ of 42.7 kJ/mol (Figure 1b). Additionally, compared with the other catalysts, even noble metal catalysts, Ni₁@Beta-NO₃⁻ also shows a comparable or even lower E_a and a higher TOF value (331 min⁻¹ at 298 K) than most of the reported catalysts (Figure 1a and

Table 2. A comparison of catalytic activities over various metal catalysts in the hydrolysis of NaBH₄ measured in this work and the ones reported in the literature.

Catalyst	Reaction Condition			Specific Rate (mL H2 min−1 gcat.−1)	TOF (min−1)	Ea		Ref.
	Temperature (K)	NaBH4 (mol L−1)	NaOH (wt%)			Value (kJ mol−1)	Temperature Range (K)
Non-noble metal catalyst
Ni1@Beta-NO3−	298	0.075	No	1620	331	30.1	293–313	This work
Ni@NaY	298	0.1	0.16		27	60.4 ± 3.1	293–313	[17]
Co@C	303	0.125	4		12.9	25.8	303–323	[18]
Co@NaY	298	0.15	10		14.7	34 ± 2	298–318	[19]
Co@ZIF-8	303	0.75	1.6		33.7	62.9	298–313	[20]
Noble metal catalyst
Ru@NaY	298	0.15	No		550	49 ± 2	293–318	[21]
Ru@ZIF-67	303	0.125	7.5		16	36.2	303–333	[22]
Pt@MSN	298	3.17	2		7.6	40.1	298–338	[23]

) [17,18,19,20,21,22,23]. The high TOF and low E_a fully demonstrate that Ni single-atoms in Ni₁@Beta-NO₃⁻ have superior catalytic activity for H₂ generation from the hydrolysis of NaBH₄.

2.2. Ethylene Dimerization over Ni Single-Atom Catalysts

1-Butene is the most in-demand linear α-olefin. Ethylene dimerization to 1-Butene is regarded as a cost-effective route to produce 1-Butene compared to other techniques based on refinery fractionation and steam cracking [24]. The highly active homogeneous ethylene dimerization catalysts are characteristic of a common structural moiety, that is, Ni²⁺ coordinated with two imine groups and two other ligands (halogens) [25]. The catalytic activity of homogeneous Ni-based catalysts is up to 10⁷ g/(mol_Ni h), denoted as 10⁷ h⁻¹ [26], but the catalysts are difficult to separate from the reaction system. In contrast, heterogeneous Ni-based catalysts are easy to separate, but their activity is up to 10⁵–10⁶ h⁻¹. Recently, Ni-containing zeolites with coordinately unsaturated Ni²⁺ centers were applied as effective catalysts for ethylene dimerization [27,28]. In our previous work, we found that Ni single-atoms (Ni₁@Beta) have a superior ability to activate hydrogen, but due to the strong interaction between the framework and Ni, Ni single-atoms exhibit ionic characteristics. Many studies have reported that the B acid sites (BASs) in zeolites play an important role in the dimerization reaction [29,30,31]. In view of this, three types of protonated zeolites were employed for ethylene dimerization. Figure 2 shows the results of the ethylene dimerization of zeolite catalysts with AlEt₂Cl as a co-catalyst. The H-Beta zeolite presents low catalytic activity with a TOF of 317 h⁻¹. After incorporating Ni species into Beta zeolite, the catalytic activity is obviously enhanced. Specifically, Ni₁@H-Beta-NO₃⁻ shows a TOF of 5804 h⁻¹, which is about twice as much as that of Ni/H-Beta-NO₃⁻. Additionally, compared with other heterogeneous Ni-based catalysts [32,33,34], Ni₁@Beta-NO₃⁻ still exhibits the highest TOF under similar reaction conditions (

Table 3. A comparison of catalytic activities over various Ni-based catalysts in ethylene dimerization measured in this work and those reported in the literature.

Samples	Temperature (K)	Pressure (MPa)	Selectivity(%)			TOF (h−1)	Ref.
			C4	C6	C8
Ni1@Beta-NO3−	298	1.0	99.4	0.6	0	5804	This work
1D-S-Ni	298	1.0	95.3	3.6	0.3	4464	[32]
Ni-UMOFNs-190	298	1.0	74.5	2.9	12.9	3571	[33]
Ni-BDC	298	1.0	78.1	1.5	15.0	2429	[33]
Ni/Y	308	4.0	92.0	5.0	2.0	~4200	[34]

). This is due to the fact that Ni₁@H-Beta-NO₃⁻ possesses atomically dispersed Ni species, which provide extremely high percentages of exposed catalytically active surfaces to ensure high catalytic activity. Moreover, under the investigated conditions, Ni₁@H-Beta-NO₃⁻ shows high selectivity for butene, affording 99.44% butene and 0.56% hexenes, without higher oligomers, which is also higher than the catalysts listed in Table 3. The high selectivity of butene can be ascribed to the steric limitation effect of the zeolite structure, which plays an important role in shape selectivity because shape selectivity is apparent, resulting in the formation of a broader range of low-carbon hydrocarbons (C₄, C₆, C₈, etc.) when the active sites of the external surface are compromised.

2.3. Propylene Oligomerization

The ethylene dimerization above shows that the conversion of olefins depends on Ni as the active center; the BASs showed a bit of activity (Figure 2). We found that Ni single-atoms possess superior olefin conversion activity. It should also be noted that the BASs, as active sites for the isomerization, cracking, and aromatization of olefins, has been realized. Here, propylene oligomerization to larger α-olefins as a probe reaction was carried out, and the Ni₁@Beta-NO₃⁻, without ion-exchanging treatment with NH₄Cl to generate BASs, was used as an oligomerization catalyst to explore the performance of single-atoms. Concurrently, as a comparison, the sample prepared by using the NiCl₂ precursor was also included.

2.3.1. Catalytic Performance

Figure 3a,b show the relationship between gas hourly space velocity (GHSV) and propylene conversion, as well as the selectivity of various products over H-Beta and Ni₁@Beta-NO₃⁻ at 453 K, respectively. When the GHSV decreases from 600 h⁻¹ to 200 h⁻¹, propylene conversion for both H-Beta and Ni₁@Beta-NO₃⁻ increases by approximately 79%. This indicates that reducing the GHSV is beneficial for enhancing propylene conversion. The high conversion of H-Beta over Ni₁@Beta-NO₃⁻ is attributed to the abundance of protonic acid sites in H-Beta. They enable the combination of propylene with protons, leading to dimerization and trimerization, yielding products including C₆, C₉, and so on. Concurrently, isomerization, β-scission, cyclization, and other reactions also take place. In contrast, the Ni₁@Beta-NO₃⁻ sample possesses some Lewis acid sites (LASs) generated by the Ni species, but with few BASs. On the other hand, a low GHSV is detrimental to the formation of long-chain olefins because of the cracking reactions. For H-Beta, at higher GHSV, the short contact time between reactants and the catalyst leads to a decreased probability of cracking reactions of oligomerization products. Consequently, with the increasing GHSV, the selectivity of <C₃, C₄, and C₅ shows a declining trend, while that of C₆, C₇, C₈, and C₉⁺ shows an upward trend. The pattern of selectivity variation for each product in Ni₁@Beta-NO₃⁻ is largely consistent with that of H-Beta. It is noteworthy that in H-Beta, the selectivity for <C₃ significantly increases at GHSV = 200 h⁻¹, reaching 42.7%, higher than other products. For Ni₁@Beta-NO₃⁻, the selectivity for <C₃ remains at a very low level, suggesting that the active sites for oligomerization and cracking reactions in H-Beta are primarily the BASs. The high acidity is prone to cause C₆ cracking into C₂ and C₄. In contrast, in Ni₁@Beta-NO₃⁻, the reaction active sites are LASs, and C₆ is less likely to undergo cracking to produce C₄ and C₂. Overall, H-Beta is more prone to cracking reactions compared to Ni₁@Beta-NO₃⁻.

Table 4. Product yields of propylene polymerization.

Samples	GHSV (h−1)	Conversion (wt%)	Yield (%)
			<C3	C4	C5	C6	C7	C8	C9+
H-Beta	600	13.9	0.03	3.54	2.88	3.24	3.04	0.15	1.03
Ni1@Beta-NO3−	450	19.3	0.05	2.21	2.79	2.31	8.88	0.54	2.56
Ni@Beta-Cl−	400	21.1	0.03	3.88	3.60	2.92	6.59	0.81	3.26
Ni1@Beta-NO3−	300	79.8	0.17	58.31	18.24	1.09	0.16	0.50	0.33
Ni/Beta-NO3−	300	79.6	0.19	40.36	34.33	3.67	0.66	0.17	0.21

presents the yield of various products for different catalysts at similar conversion rates. Figure 3c presents the selectivity of various products for H-Beta, Ni₁@Beta-NO₃⁻, and Ni@Beta-Cl⁻ at conversion of 13.9%, 19.3%, and 21.1%, respectively. It can be observed that for H-Beta, the selectivity for C₄, C₅, C₆, and C₇ is relatively similar, indicating that these are the primary products of the reaction. For Ni₁@Beta-NO₃⁻ and Ni@Beta-Cl⁻, the selectivity for C₇ is the highest, followed by C₄, C₅, C₆, and C₉⁺, with their selectivity being relatively close. This suggests that in H-Beta, C₇ may crack to C₃ and C₄, or it may first polymerize with other products and then undergo cracking. In contrast, Ni₁@Beta-NO₃⁻ and Ni@Beta-Cl⁻ exhibit weaker acidity, which favors the formation of long-chain olefins, leading to the low selectivity for C₄ and C₅. Figure 3d illustrates the selectivity of the products for Ni₁@Beta-NO₃⁻ and Ni/Beta-NO₃⁻ at conversion of 79.8% and 79.5%, respectively. Both of them primarily produce C₄ and C₅. The difference is that the selectivity for C₄ is much greater than C₅ in Ni₁@Beta-NO₃⁻, while they are similar in Ni/Beta-NO₃⁻. This may be due to the fact that C₄ and C₅ primarily originate from the cracking of long-chain olefins (C₉) at the metal active sites on Ni/Beta-NO₃⁻. Ni₁@Beta-NO₃⁻ has higher cracking activity because of Ni species, which possess both metallic and acidic properties. Consequently, C₄ may not only originate from C₉ but also potentially from C₆, C₇, and C₈. In summary, introducing metal species into the zeolite channels can generate new LASs. It is helpful to overcome the cracking issues occurring during oligomerization using H-type zeolite, thereby enhancing the selectivity of the long-chain olefins.

From the aforementioned results, it can be observed that Ni₁@Beta-NO₃⁻ and Ni@Beta-Cl⁻ exhibit similar product selectivity. However, when GHVS = 400 h⁻¹, the conversion of Ni@Beta-Cl⁻ is only 21.1%, significantly lower than that of Ni₁@Beta-NO₃⁻ (~40.6%). This discrepancy arises because the nickel species in Ni₁@Beta-NO₃⁻ interact strongly with the framework, making them less reducible, whereas in Ni@Beta-Cl⁻, the majority of the nickel species can be reduced, reducing the active centers for the oligomerization of olefins to Ni⁺ or Ni²⁺ [35].

2.3.2. Discussion on the Effect of Metal Precursors for Ni@Beta Zeolite Synthesis

Figure 4a presents the X-ray diffraction (XRD) patterns of Na-Beta and Ni@Beta-Cl⁻. All samples exhibit characteristic peaks of Beta zeolite, indicating that the introduction of Ni species does not disrupt the structure of the Beta zeolite. No diffraction peaks related to Ni species are observed, suggesting that the lattice distortion of Beta zeolite after encapsulation with metal can be neglected. There are no significant Ni particles in the samples, indicating strong dispersion of Ni. According to the X-ray fluorescence (XRF) results in

Table 5. Physicochemical properties of Na-Beta, Ni₁@Beta-NO₃⁻, and Ni@Beta-Cl⁻.

Samples	Relative Crystallinity (%) a	SiO2/Al2O3 b	Ni Content (wt%) b	Surface Area (m2/g) c			Volume (cm3/g) c
				SBET	Smeso	Smicro	Vtotal	Vmeso	Vmicro
Na-Beta	100	56.5		654	154	500	0.29	0.04	0.25
Ni1@Beta-NO3−	85	60.9	1.28	697	193	504	0.32	0.27	0.05
Ni@Beta-Cl−	79	63.5	1.36	672	174	498	0.30	0.06	0.24

^a Calculated based on XRD results, with the relative crystallinity of Na-Beta specified as 100%. ^b Measured based on XRF analysis. ^c The specific surface area was obtained using the Brunauer–Emmett–Teller (BET) method; the micropore surface area and micropore volume were calculated using the t-plot method; and the mesopore volume was calculated by subtracting the micropore volume from the total volume.

, after the introduction of Ni using the two-step crystallization method, the SiO₂/Al₂O₃ of Ni@Beta-Cl⁻ is higher than that of Na-Beta, indicating that the addition of Ni affects the crystallization process of the zeolite. It is also higher than that of Ni₁@Beta-NO₃⁻, suggesting that anions also play a non-negligible role in the crystallization process. Furthermore, the actual Ni content in Ni@Beta-Cl⁻ is higher than the theoretical value, which is due to the fact that the introduction of metal precursors increases the difficulty of crystallization of the zeolite, and not all silicon species are converted into zeolite products during the conversion of alumino-silicate gel. Figure 4b presents the N₂ adsorption–desorption isotherms for various samples. The isotherms for all samples are of the I type, characteristic of typical microporous materials, indicating a rich microporous structure. Compared to Na-Beta, the total surface area and total pore volume of Ni@Beta-Cl⁻ and Ni₁@Beta-NO₃⁻ both increased. This may be due to the fact that metal cations can participate in the crystallization process of zeolites, playing a role in the structural guidance and charge balance of the framework. Figure 4c presents the transmission electron microscopy (TEM) image and EDS mapping of Ni@Beta-Cl⁻. It can be observed that the two-step crystallization method did not result in the formation of large agglomerated particles after the introduction of Ni, indicating their uniform dispersion and encapsulation within the Beta zeolite. Additionally, the regular crystal structure of the zeolite suggests that the addition of Ni via the two-step crystallization method does not affect the crystal structure of the zeolite, consistent with the XRD results. The EDS results further demonstrate that the two-step crystallization method enables a uniform dispersion of Ni particles throughout the entire Beta zeolite crystal. The above results indicate that during the hydrothermal synthesis of zeolite, the in situ introduction of a metal precursor can effectively encapsulate the metal and maintain its small particle size and high dispersion.

Pyridine adsorption (FT-IR-Py) and ammonia temperature-programmed desorption (NH₃-TPD) were employed to analyze the acidity of the zeolite samples. As shown in Figure 5a, two absorption bands at 1450 cm⁻¹ for LASs and 1545 cm⁻¹ for BASs are observed for Na-Beta, Ni₁@Beta-NO₃⁻, and Ni@Beta-Cl⁻. Upon the introduction of Ni, the peak corresponding to the LASs significantly increased in intensity, while the peak associated with the BASs decreased, leading to a reduction in the ratio of B/L (

Table 6. The FT-IR-Py and NH₃-TPD results for Na-Beta, Ni₁@Beta-NO₃⁻, and Ni@Beta-Cl⁻.

Samples	Total Acid Amount (mmol/g) a	Total L Acid Amount (mmol/g) b	Total B Acid Amount (mmol/g) b	B acid Amount/ L Acid Amount b	Strong L Acid Amount (mmol/g) c	Strong B Acid Amount (mmol/g) c
Na-Beta	0.18	0.08	0.10	1.25	0.05	0.10
Ni1@Beta-NO3−	0.20	0.14	0.06	0.43	0.11	0.06
Ni@Beta-Cl−	0.17	0.10	0.07	0.70	0.10	0.09

^a Determined based on NH₃-TPD. ^b Determined based on the adsorbed pyridine amount at 473 K in the FT-IR-Py spectra. ^c Determined based on the adsorbed pyridine amount at 623 K in the FT-IR-Py spectra.

). Further NH₃-TPD analysis was conducted to investigate the quantity and strength of the acidic centers of the catalysts. As depicted in Figure 5b, all samples exhibit two desorption peaks, with the peak in the range of 453–483 K attributed to weak acid sites and the peak between 513 and 593 K corresponding to medium–strong acid sites. The total acidity of the Beta zeolite increased after the incorporation of Ni, which is attributed to the generation of new LASs associated with Ni²⁺ and NiO [36], consistent with the findings from the FT-IR-Py analysis. The differences in acid amount and strength between Ni@Beta-NO₃⁻ and Ni@Beta-Cl⁻ arise from the fact that different anions alter the dissolution rate of the alumino-silicate gel and the size of the template agent.

A previous study has demonstrated that the excellent catalytic performance of Ni₁@Beta-NO₃⁻ is closely related to the electronic properties of the isolated Ni sites [15]. The temperature-programmed reduction of hydrogen (H₂-TPR) was conducted on Ni@Beta-Cl⁻ and Ni₁@Beta-NO₃⁻ (Figure 6a). Ni@Beta-Cl⁻ exhibits two H₂ consumption peaks at 753 K and 876 K, which are attributed to the reduction of NiO to metallic Ni [37,38] and the reduction of isolated Ni cations within the zeolite pores, respectively. However, the reduction temperature is significantly lower than that of Ni₁@Beta-NO₃⁻ (923 K), indicating weaker metal–support interactions in Ni@Beta-Cl⁻ compared to Ni₁@Beta-NO₃⁻ [39,40]. Additionally, the calculated nickel reduction degrees for Ni@Beta-Cl⁻ and Ni₁@Beta-NO₃⁻ are determined to be 62.9% and 23.4%, respectively, also indicating a stronger metal–support interaction in Ni₁@Beta-NO₃⁻ than that in Ni@Beta-Cl⁻. X-ray photoelectron spectroscopy (XPS) analysis was further utilized to characterize the electronic states of Ni in Ni₁@Beta-NO₃⁻ and Ni@Beta-Cl⁻. The XPS spectrum of the Ni 2p_3/2 for Ni₁@Beta-NO₃⁻ (Figure 6b) only shows one characteristic peak at 856.1 eV, which is attributed to the Ni²⁺ species. Unlike Ni₁@Beta-NO₃⁻, Ni@Beta-Cl⁻ exhibits two main peaks centered at 851.5 eV and 855.4 eV, corresponding to Ni⁰ and Ni²⁺ species, respectively [41,42]. This indicates that only a portion of the Ni species in Ni@Beta-Cl⁻ carry a positive charge, suggesting a weaker interaction between Ni and O atoms. Notably, the peak of Ni²⁺ in Ni₁@Beta-NO₃⁻ is shifted to a slightly higher binding energy compared to Ni@Beta-Cl⁻. This positive shift indicates a buildup of positive charges on the Ni species, which also points to enhanced electron interaction between Ni and O atoms. To further characterize the electronic configurations of Ni species in Ni₁@Beta-NO₃⁻ and Ni@Beta-Cl⁻, Fourier transform infrared spectroscopy of CO adsorbed (FT-IR-CO) was conducted at low temperatures of −100 K, with CO being gradually introduced (Figure 6c,d). As for Ni@Beta-Cl⁻, the adsorption peaks at positions of 2020 and ~1900 cm⁻¹ are attributed to linear and bridge adsorption of CO on metal Ni nanoparticles, respectively [43,44]. Notably, such two signals were not apparent for Ni₁@Beta-NO₃⁻, further confirming that the Ni species in Ni₁@Beta-NO₃⁻ are atomically dispersed. The vibrational characteristics of CO adsorbed near 2168 cm⁻¹ on Ni₁@Beta-NO₃⁻ and Ni@Beta-Cl⁻ are associated with the formation of carbonyl involving Na⁺ [45,46]. In addition, for Ni₁@Beta-NO₃⁻, a new peak appears at 2202 cm⁻¹, which can be assigned to the adsorption of CO on isolated Ni²⁺ sites [28,47]. However, this was not observed in Ni@Beta-Cl⁻, indicating that the degree of dispersion of Ni species and the content of Ni²⁺ in Ni@Beta-Cl⁻ are both lower than those in Ni₁@Beta-NO₃⁻.

Based on the above results, the preparation process of Ni@Beta is influenced by the anionic species in the precursor solution. The anions present in the nickel precursor directly influence the SiO₂/Al₂O₃ and pore structure, subsequently leading to variations in the content and charge environment of Ni. It was shown that the dispersion of Ni species and the content of Ni²⁺ in Ni@Beta-Cl⁻ were both lower than those in Ni₁@Beta-NO₃⁻. Compared to Cl⁻ anions, NO₃⁻ anions have a plane triangle structure, high electronegativity of oxygen atoms, and a strong ability to form hydrogen bonds with hydroxyl or water molecules. This means that they can easily interact with alumino-silicate gel. Accordingly, it is beneficial for the dispersion of their counter ions (Ni²⁺) because the movement of nitrate ions is accompanied by Ni²⁺ cations.

3. Materials and Methods

3.1. Materials for Catalyst Synthesis

Silica gel (99 wt% SiO₂), tetraethylammonium hydroxide (TEAOH, 25 wt% in H₂O), and diethylaluminum chloride (AlEt₂Cl, A.R.) were purchased from Shanghai Aladdin Biochemical Technology Co., Ltd., Shanghai, China. Sodium hydroxide (NaOH, 99%) was obtained from Sinopharm Chemical Reagent Co., Ltd., Shanghai, China. Sodium metaaluminate (NaAlO₂, 45 wt% Al₂O₃) was obtained from the Tianjin Jinke Fine Chemical Research Institute, Tianjin, China. Nickel (II) nitrate hexahydrate (Ni(NO₃)₂·6H₂O, 98%) was obtained from Tianjin Fuchen Chemical Reagent Factory, Tianjin, China. Anhydrous Ni chloride (NiCl₂, 98%) was obtained from Thermo Fisher Scientific Co., Ltd., Shanghai, China. Ammonium chloride (NH₄Cl, 99.5%) was purchased from Shanghai Macklin Biochemical Co., Ltd., Shanghai, China. Sodium borohydride (NaBH₄, 98%) was obtained from J&K scientific Co., Ltd., Beijing, China. Toluene was purchased from Tianjin Fuyu Chemical Co., Ltd., Tianjin, China. Ethylene was purchased from Zhangjiakou Chuangqi Gas Co., Ltd., Zhangjiakou, China. Propylene was obtained from Beijing Qianxi Jingcheng Gases Co., Ltd., Beijing, China. Ethanol (C₂H₅OH, 99%) was obtained from Beijing Tongguang Fine Chemical Co., Ltd., Beijing, China. The above materials were used without further purification.

3.2. Catalyst Preparation

3.2.1. Synthesis of Beta Zeolite

The Beta zeolite was prepared using a conventional hydrothermal synthesis method. In a typical process, 27.9 g TEAOH and 0.45 g NaOH were added to deionized water and stirred until a clear solution was formed. Subsequently, 0.8 g NaAlO₂ was added and stirring was continued for 30 min. Then, 15 g silica gel was slowly added to the above solution and stirred for 3 h to achieve a homogeneous alumino-silicate gel with a molar composition of SiO₂/Al₂O₃/TEAOH/H₂O/Na₂O = 1/0.014/0.19/7.21/0.042. The as-prepared gel was transferred into a 100 mL Teflon-lined stainless-steel autoclave, placed in an oven at 413 K, and crystallized for 48 h. After crystallization, the solid products were separated via centrifugation, washed with deionized water until the pH of the filtrate was ~7, and dried at 373 K for 12 h. Finally, the Beta zeolite was obtained through calcination in the air at 823 K for 6 h with a ramping rate of 2 K min⁻¹).

3.2.2. Synthesis of Ni₁@Beta-NO₃⁻

For the synthesis of Ni₁@Beta-NO₃⁻, an alumino-silicate gel with a molar composition of SiO₂/Al₂O₃/TEAOH/H₂O/Na₂O = 1/0.014/0.19/7.21/0.042 was produced. It was placed in an autoclave at 273 K for 24 h and then the autoclave was cooled to ambient temperature. Next, the 0.75 g Ni(NO₃)₂ precursor was dissolved into 1 g deionized water and added to the autoclave. The autoclave was sealed again, heated to 413 K, and kept at this temperature for 36 h. The as-obtained solid product was filtered, washed with water several times, and then dried at 373 K overnight, followed by calcination at 823 K for 6 h. Finally, the sample was reduced at 30 mL min⁻¹ 10% H₂/N₂ flow at 773 K (ramping rate: 5 K min⁻¹) for 2 h. Before exposure to air, all of the samples were passivated at 303 K for 1 h under 30 mL min^–1 1% O₂/N₂ flow.

3.2.3. Synthesis of Ni@Beta-Cl⁻

Ni@Beta-Cl⁻ was prepared with the same procedure as Ni₁@Beta-NO₃⁻ using 0.61 g NiCl₂ as a precursor.

3.2.4. Synthesis of Ni/Beta-NO₃⁻

Ni/Beta was prepared using an impregnation method. In a typical procedure, 5 g Beta zeolite was slowly added to a beaker containing 40 mL of a Ni(NO₃)₂ solution (0.021 mol L⁻¹). The mixture was stirred uniformly at 333 K for approximately 6 h until the solvent evaporated. Subsequently, the obtained sample was dried overnight at 373 K. The calcination, reduction, and passivation steps were identical to those used for Ni@Beta-Cl⁻.

3.2.5. Synthesis of H-Beta, Ni₁@H-Beta-NO₃⁻, and Ni@H-Beta-Cl⁻

The Beta, Ni₁@Beta-NO₃⁻, and Ni@Beta-Cl⁻ zeolites were treated with ammonium ion-exchange in 1 mol/L NH₄Cl solution at 363 K for 2 h. The ion-exchanged zeolites were filtered and washed with deionized water and then calcined at 823 K for 6 h. The ion-exchanging process was repeated twice to obtain H-Beta, Ni₁@H-Beta-NO₃⁻, and Ni@H-Beta-Cl⁻.

3.3. Catalyst Characterization

XRD analysis was conducted using a Bruker D8 Advance diffractometer with a Cu-Kα radiation source (40 kV, 40 mA) (Malvern Panalytical, Malvern, UK). The elemental composition of the zeolite samples was determined via XRF analysis, which was conducted on a PANalytical AxiosMAX analyzer (Malvern Panalytical, Malvern, UK). TEM images of the zeolite samples were obtained using a FEI Tecnai G² F20 microscope equipped with an EDS spectrometer (FEI Company, Hillsboro, OR, USA). The textural properties of zeolite samples were recorded via N₂-sorption using a Kubo X1000 setup (Bjbuilder, Beijing, China) operating at 77 K. Before sorption measurements, the samples were treated at 623 K for 4 h under vacuum to eliminate adsorbed substances. The BET equation was used to calculate the specific surface area of the zeolite sample. The micropore surface area and the micropore volume were calculated using the t-plot method. XPS analysis was performed using a Fluorolog-Tau3 spectrometer (HORIBA, Kyoto, Japan), with the C 1s peak at 284.6 eV used as a reference for calibration. The H₂-TPR and NH₃-TPD were conducted on a chemisorption instrument (Huasi, Yueyang, China, DAS-7000), which was equipped with a TCD detector. FT-IR-Py and FT-IR-CO were performed using a Bruker Tensor II equipped with an MCT detector, produced by Bruker (Billerica, MA, USA).

3.4. Catalyst Evaluation

3.4.1. Hydrolysis of NaBH₄

Hydrolysis of NaBH₄ was conducted in a 150 mL small reactor with a quartz glass inner sleeve at 298 K under atmospheric pressure. The outlet of the reactor is connected to a CS200-A flow meter, which is connected to a computer to monitor the HGR (mL/min) during the hydrolysis process. In a typical manner, 0.05 g catalyst and 25 mL deionized water were introduced into the glass container and thoroughly mixed by using a magnetic stirrer. The reactor was purged with N₂ for 30 min before the reaction started, then heated to the reaction temperature, and 25 mL 75 mM NaBH₄ solution (corresponding to a maximum amount of H₂ gas of 7.5 mmol = 168 mL at 298 K) was introduced into the glass container via a syringe.

The TOF was calculated based on the quantity of Ni metals in the catalysts when the conversion of NaBH₄ reached 100%. The calculation equation used was as below:

$$\mathrm{T}\mathrm{O}\mathrm{F}={\displaystyle \frac{n_{{\mathrm{H}}_2}}{n_{\mathrm{N}\mathrm{i}}·t}}$$

where $n_{{\mathrm{H}}_2}$ is the mole number of released H₂, n_Ni is the mole number of Ni atoms in the catalyst based on the bulk composition, and t is the completion time of the reaction in minutes.

For kinetic studies, hydrolytic dehydrogenation of the NaBH₄ reaction was also carried out at 293, 303, 308, and 313 K in order to obtain the E_a. The E_a value can be calculated via the following Arrhenius equation:

$$lnr=ln{k}_0-{\displaystyle \frac{E_a}{R}}·{\displaystyle \frac{1}{T}}$$

where r is the HGR, T is the hydrolysis temperature in Kelvins, R is the gas constant (8.314 J mol⁻¹ K⁻¹), and k₀ signifies the pre-exponential factor. By fitting the Arrhenius plot of lnr versus 1/T, the slope can be clearly obtained via −E_a/R.

3.4.2. Ethylene Dimerization

Ethylene dimerization was performed in a 100 mL stainless steel autoclave, which was purged three times with N₂ and once with 1 atm ethylene. Then, the as-prepared catalysts and AlEt₂Cl were dispersed into toluene to prepare a suspension and injected into the reactor under N₂. Next, the autoclave was pressurized with ethylene to 10 atm and kept constant during the reaction. After 0.5 h of reaction, the reactor was cooled to 253 K with ice-cold ethanol solution and depressurized. An organic layer was separated from the reaction mixture and analyzed via gas chromatography (GC).

The TOF was calculated according to the following equation:

$$\mathrm{T}\mathrm{O}\mathrm{F}={\displaystyle \frac{n_{{{\mathrm{C}}_2\mathrm{H}}_4}}{n_{\mathrm{N}\mathrm{i}}·t}}={\displaystyle \frac{m_{{{\mathrm{C}}_2\mathrm{H}}_4}}{{M·n}_{\mathrm{N}\mathrm{i}}·t}}$$

where $m_{{{\mathrm{C}}_2\mathrm{H}}_4}$ (g) is the mass of C₂H₄ consumed, which is equal to the yield of oligomer products (g). M (g/mol) is the molar mass of C₂H₄, n_Ni is the mole of Ni atoms in the catalyst based on bulk composition, and t is the reaction time (h).

3.4.3. Propylene Oligomerization

Propylene oligomerization was carried out in a stainless steel fixed-bed reactor (inner diameter: 10 mm, length: 360 mm). In general, the catalyst was pressed into sheets at 20 MPa and then crushed and screened into particles (0.250–0.425 mm). Finally, the catalyst was loaded into the reactor. The catalyst was reduced by hydrogen gas (10 mol% H₂/N₂, 30 mL/min) at 573 K (ramping rate 2 K/min) for 2 h. Mixed gas (25 mol% propylene/N₂) was fed into the reactor, and numerical results were obtained under the reaction condition that the reactive temperature was maintained at 453 K, the operation pressure was 1 atm, and GHSV = 200–600 h⁻¹ was varied. The products were measured using an online GC (HuiFen, Beijing, China, GC-7820) equipped with an HP-PONA capillary column connected to a flame ionization detector (FID). The conversion (X) of propylene and selectivity (S) of the products were calculated as follows:

$$X\left(\%\right)={\displaystyle \frac{\sum n_i{c}_i}{\left(\sum n_i{c}_i+c_{propylene}\right)}}\times 100$$

$$S\left(\%\right)={\displaystyle \frac{n_i{c}_i}{\sum n_i{c}_i}} \times 100$$

where i represents the product (i ≤ C₃, C₄, C₅, C₆, C₇, C₈, and C₉), n_i represents the number of carbon atoms in product i, and c_i represents the molar concentration of product i.

4. Conclusions

In summary, benefiting from a unique coordination environment and maximum atomic utilization of encaged Ni single-atoms within Beta zeolite, compared with other heterogeneous catalysts, the obtained Ni₁@Beta catalyst not only exhibits good catalytic activity and selectivity in hydrogen generation reactions (TOF = 331 min⁻¹ and E_a = 30.1 kJ/mol) and the dimerization of ethylene (butylene selectivity = 99.4% and TOF = 5804 h⁻¹) but also propylene oligomerization (long-chain olefin (≥C₆⁺) selectivity = 75.21%). These findings provide a general and promising configuration of metal atoms. The prepared catalysts will be of great potential for hydrogen production and the synthesis of clean liquid fuels involving metal or metal ions as active sites. Moreover, the effect of the anions of metal precursors for encapsulation has been identified. Compared with Ni(NO₃)₂, the introduction of NiCl₂ leads to a weaker interaction between Ni species and the zeolite framework, and larger nickel clusters are formed.