Highly Efficient and Stable Ni-Cs/TS-1 Catalyst for Gas-Phase Propylene Epoxidation with H₂ and O₂

Abstract

The development of non-noble metal catalysts for gas-phase propylene epoxidation with H₂/O₂ remains challenging due to their inadequate activity and stability. Herein, we report a Cs⁺-modified Ni/TS-1 catalyst (9%Ni-Cs/TS-1), which exhibits unprecedented catalytic performance, giving a state-of-the-art PO formation rate of 382.9 g_PO·kg_cat⁻¹·h⁻¹ with 87.8% selectivity at 200 °C. The catalyst stability was sustainable for 150 h, far surpassing reported Ni-based catalysts. Ni/TS-1 exhibited low catalytic activity. However, the Cs modification significantly enhanced the performance of Ni/TS-1. Furthermore, the intrinsic reason for the enhanced performance was elucidated by multiple techniques such as XPS, N₂ physisorption, TEM, ²⁹Si NMR, NH₃-TPD-MS, UV–vis, and so on. The findings indicated that the incorporation of Cs⁺ markedly boosted the reduction of Ni, enhanced Ni⁰ formation, strengthened Ni-Ti interactions, reduced acid sites to inhibit PO isomerization, improved the dispersion of Ni nanoparticles, reduced particle size, and improved the hydrophobicity of Ni/TS-1 to facilitate propylene adsorption/PO desorption. The 9%Ni-Cs/TS-1 catalyst demonstrated exceptional performance characterized by a low cost, high activity, and long-term stability, offering a viable alternative to Au-based systems.

1. Introduction

Propylene oxide is a significant chemical intermediary utilized in the synthesis of organic compounds, including polyether polyols, propylene glycol, and propylene glycol ethers [1,2]. It can be further synthesized into rigid and flexible polyurethane foam for the production of various home items [3]. The gas-phase epoxidation of propylene using hydrogen and oxygen presents considerable economic and environmental benefits owing to its eco-friendly and efficient characteristics [4,5,6]. The creation of a highly efficient catalyst for the gas-phase propylene epoxidation system has posed a significant research challenge in this domain.

In recent decades, extensive studies have been conducted on the gas-phase propylene epoxidation process. Reports indicate that Au nanoparticles exhibit strong coordination with tetrahedrally structured Ti species in catalysts. Researchers have altered the morphological structure of titanium–silicon zeolites [7,8], and various Au-supported systems, including Au/Ti-SiO₂ [9,10,11], Au/Ti-TUD [12], Au/Ti-MCM [13,14], Au/Ti-MWW [15,16,17,18], Au/Ti-SBA-15 [19], Au/TS-2 [20], and Au/TS-1 [21,22,23,24,25,26,27], have been documented. Propylene epoxidation experiments demonstrate that employing catalysts with higher specific surface areas can enhance propylene conversion to varying degrees. Among these, TS-1, characterized by its distinct hydrophobicity and high proportion of tetrahedrally coordinated Ti species, exhibits optimal catalytic performance. Mechanistically, Au nanoparticles activate H₂ and O₂ to generate H₂O₂, while the tetrahedral Ti sites in TS-1 facilitate propylene epoxidation.

Despite the success of Au-based catalysts in research [28,29,30,31,32], challenges such as high costs, low propylene conversion rates, and inadequate stability hinder their commercial implementation. Consequently, developing highly efficient non-noble metal catalysts for this reaction has become imperative. Copper-based [33,34,35] and iron-based [36,37] catalysts, utilizing oxygen as the oxidant, have attracted significant attention. Beyond these, alternative catalyst systems have also been explored. For instance, Li Landong’s group [38] developed a Co@Y molecular sieve catalyst that achieved 25% propylene conversion and 57% PO selectivity at 500 °C. However, PO selectivity remained insufficient for industrial applications compared to Au/TS-1.

To address these limitations, researchers have focused on modifying non-noble metal catalysts with alkali metals. Cu/SiO₂ catalyst synthesized via conventional impregnation initially exhibited negligible epoxidation activity. After NaCl treatment, PO selectivity increased to 44% [39]. Further improvements were achieved by modifying CuO_x/SiO₂ with alkali metals such as Cs⁺ and K⁺ [40]. Ananieva et al. [41] demonstrated that Cs⁺-modified Fe-based catalysts achieved 75% PO selectivity with 10% propylene conversion. Na⁺ modification of Fe catalysts [42] also yielded favorable results, underscoring the universal role of alkali metals in enhancing epoxidation performance.

The above non-precious metal catalysts facilitate the catalytic epoxidation of propylene utilizing molecular oxygen as the oxidizing agent. Nonetheless, molecular oxygen as an oxidizing agent has low selectivity in the epoxidation cycle and is susceptible to side reactions. Furthermore, the utilization of molecular oxygen necessitates elevated reaction temperatures and pressures, imposing specific demands on equipment and energy usage. Consequently, researchers synthesized Ni-based catalysts for the epoxidation of propylene with H₂ and O₂ [43,44,45]. However, these catalysts exhibited suboptimal catalytic performance, falling significantly short of industrial viability requirements.

To improve catalytic performance, in this paper, we used nickel supported on the TS-1 zeolite catalyst modified with the alkali metal Cs⁺, in the gas-phase epoxidation of propylene with H₂ and O₂. A series of Cs-modified catalysts (designated Ni-Cs/TS-1) were synthesized by the deposition precipitation method and the impregnation method. The catalysts were studied by various characterizations, including XRD, TEM, N₂ physisorption, ²⁹SiNMR, H₂-TPR, NH₃-TPD-MS, UV–vis, and so on. In addition, different reaction conditions, such as temperature, space velocity, as well as the pretreatment condition, were also investigated, and the effect of Ni and Cs⁺ loading on the reaction was further systematically studied. Finally, the catalyst was subjected to a long-running test. The 9%Ni-Cs/TS-1 catalyst exhibited unprecedented performance in the epoxidation of propylene, representing a breakthrough for non-noble metal catalysts in epoxidation. Under optimized conditions at 200 °C, the catalyst achieved a 14% propylene conversion with 87.8% PO selectivity and a 382.9 g_PO·kg_cat⁻¹·h⁻¹ PO formation rate, representing state-of-the-art performance among reported Ni/TS-1 catalysts. It can also serve as a good example for other oxidation reactions in H₂ and O₂ atmospheres.

2. Results and Discussion

2.1. Characterization of the Catalysts

Figure 1 and Figure S1 present the X-ray diffraction (XRD) patterns of TS-1, Ni/TS-1, and Ni-Cs/TS-1 catalysts. All materials maintain characteristic MFI topology reflections at 7.9°, 8.8°, 23.1°, 23.9°, and 24.3° after Ni and Cs⁺ deposition, confirming structural integrity. The absence of diffraction peaks attributable to nickel nanoparticles indicates the high dispersion of Ni species.

Figure 2 displays the Ni 2p XPS spectra for reduced Ni/TS-1 alongside fresh, reduced, and spent Ni-Cs/TS-1 catalysts. The Ni⁰ peak intensity at 852.8 eV in reduced 5% Ni/TS-1 without Cs⁺ was significantly lower than in reduced 5% Ni-Cs/TS-1, indicating that Cs⁺ promoted Ni nanoparticle reduction. The Ni 2p_3/2 spectrum of fresh 5% Ni-Cs/TS-1 exhibited two fitted peaks at 856.8 eV and 862.1 eV (shake-up satellite). The peak at 856.8 eV corresponded to the binding energy of Ni(OH)₂, signifying that the surface nickel of the unreduced catalyst existed as Ni²⁺ [46]. Following reduction, four distinct peaks emerged in the Ni 2p_3/2 region. The spent catalyst retained these four spectral features, but the Ni⁰ peak (852.8 eV) showed marked attenuation, indicating that Ni⁰ was consumed during the reaction, confirming its essential role in the catalytic cycle.

Figure S2a displays the Ni 2p XPS spectra of reduced Ni-Cs/TS-1 catalysts with varying Ni loadings. The figure illustrates that, after reduction, four peaks emerged in the Ni 2p_3/2 spectrum of the catalyst surface, exhibiting binding energies of 852.8 eV and 856.0 eV, along with two satellite peaks. As the Ni loading increased, the characteristic peak intensity of Ni⁰ progressively strengthened while that of Ni²⁺ correspondingly diminished. At a 9% loading, the Ni⁰ peak intensity reached its maximum, whereas at an 11% loading, it decreased. The catalytic test results indicated that the optimal catalytic effect occurred at a 9% Ni loading, suggesting that the active site of the catalytic reaction was likely Ni⁰. Figure S2b shows the Ni 2p XPS spectra of reduced Ni/TS-1 catalysts without Cs⁺ modification at equivalent loadings. The characteristic peak associated with Ni⁰ in the unmodified reduced catalyst was less pronounced than in the modified catalyst, demonstrating that Cs⁺ modification promoted the reduction of Ni nanoparticles.

The presence of Ni on TS-1 shifted the Ti2p_3/2 binding energy from 460.6 eV without Ni to 459.6 eV with Ni, as shown in Figure 3. This shift indicates strong Ni-Ti interactions [47,48], suggesting Ti-O-Ni bond formation. Furthermore, in Cs⁺-modified samples, the Ti2p_3/2 signal shifted further from 459.6 eV to 458.7 eV. This progression reflects electron withdrawal from Ti by adjacent Ni and Cs atoms, which reduces electron density around Ti and consequently lowers its binding energy.

TEM imaging revealed that TS-1 exhibited a well-defined crystal structure (Figure 4a). Figure 4b confirms metallic Ni nanoparticles displaying lattice fringes of 2.08 Å, corresponding to the Ni(111) plane. Figure 4c shows unmodified 5%Ni/TS-1, where Ni nanoparticles appear agglomerated with large particle sizes. In contrast, Figure 4d–g demonstrate Cs⁺-modified Ni-Cs/TS-1 catalysts at 5%, 7.5%, 9%, and 11% loadings, revealing strong Ni-Cs⁺ interactions. Cs⁺ modification yielded uniformly distributed Ni nanoparticles with a narrow size distribution, indicating enhanced Ni dispersibility. The particle size analysis (>200 nanoparticles) confirmed uniform Ni nanoparticles averaging 3–4 nm. The 9%Ni-Cs/TS-1 catalyst exhibited the smallest average size (3.04 nm), attributed to the optimal equilibrium between Ni, Cs⁺, and TS-1. This minimal particle size correlated with superior catalytic performance, suggesting the Ni nanoparticle size critically influences efficiency. Figure 4h indicates that the Ni nanoparticle size on the spent 5%Ni-Cs/TS-1 catalyst is 3.48 nm, while Figure 4i shows the same catalyst after 150 h of operation, compared to 3.33 nm for the fresh catalyst, suggesting no significant sintering occurred.

The N₂ adsorption–desorption isotherm and pore size distribution are shown in Figure 5. Four lag rings are visible in TS-1, which suggests the presence of a multi-level pore structure (microporous + mesoporous + macroporous), primarily due to macroporosity and crystal accumulation mesopores, suggesting that the isotherm of the catalyst corresponds to type I(b). After loading the metal, due to the partial blockage of the macroporous structure by the metal, the Ni/TS-1 and Ni-Cs/TS-1 catalysts exhibited small hysteresis loops in the high-pressure region (p/p₀ = 0.6–0.95), suggesting that the isotherm of the catalysts corresponds to type IV. All samples were composed of micropores and some mesopores within the particles (Figure S3). The pore size distribution of the samples indicated that the majority of pores in the TS-1 and Ni/TS-1 samples were less than 1 nm in diameter, characteristic of micropores (Figure S4).

Table 1. Summary of structural characterization and physicochemical properties of TS-1, Ni/TS-1, and Ni-Cs/TS-1 samples with different Ni loadings.

Samples	SBET a (m2/g)	VTotal b (cm3/g)	Vmicro b (cm3/g)	Pore Size c (nm)	Ni Loading d (wt.%)	Ti Loading d (wt.%)
TS-1	463	0.48	0.07	4.0	-	1.08
1%Ni/TS-1	424	0.64	0.12	3.0	0.56	1.09
1%Ni-Cs/TS-1	364	0.60	0.11	3.0	0.66	1.05
3%Ni/TS-1	419	0.64	0.13	3.3	2.04	1.20
3%Ni-Cs/TS-1	352	0.58	0.11	3.1	1.70	1.14
5%Ni/TS-1	386	0.50	0.13	3.1	3.12	1.17
5%Ni-Cs/TS-1	367	0.47	0.08	4.2	3.43	1.27
7.5%Ni/TS-1	435	0.51	0.05	4.0	5.50	1.14
7.5%Ni-Cs/TS-1	311	0.40	0.04	4.2	4.16	1.09
9%Ni/TS-1	418	0.63	0.10	3.1	5.91	1.23
9%Ni-Cs/TS-1	305	0.51	0.08	3.9	6.68	1.26
11%Ni/TS-1	408	0.42	0.10	4.2	7.46	1.15
11%Ni-Cs/TS-1	363	0.51	0.03	3.0	8.81	1.12

^a S_BET (total surface area) was calculated by the BET method. ^b V_Total (total volume of porous structure) and V_micro (volume of microporous structure) were calculated by the t-plot method. ^c Pore size was calculated by the BJH method. ^d Determined by ICP-OES.

demonstrates that metal loading reduced the catalyst’s specific surface area, total pore volume, and micropore volume. This phenomenon may be associated with the partial obstruction of the catalyst’s pores by Ni and Cs⁺ species [49].

Figure 6 presents the ²⁹Si solid-state NMR spectra of TS-1, Ni/TS-1, and Ni-Cs/TS-1 catalysts. All samples showed a peak at −116 ppm, primarily attributed to [Si-O-Ti] bonds from titanium incorporation into the framework [50,51,52]. Peaks at −113 ppm and −104 ppm corresponded to Q⁴ [Si(OSi)₄] and Q³ [Si(OSi)₃OH] species, respectively [53]. The Q³/Q⁴ intensity ratios were 0.087 for TS-1 and 0.07 for Ni/TS-1. For Cs⁺-modified Ni-Cs/TS-1, this ratio decreased markedly to 0.058. This reduction indicates fewer surface [Si-OH] groups from the alkali metal incorporation, leading to an increase in lattice defects, demonstrating improved catalyst hydrophobicity [54]. Enhanced hydrophobicity in Ni-Cs/TS-1 facilitates propylene adsorption and PO desorption while suppressing side reactions, thereby increasing the PO formation rate [55].

The surface acidity of the catalyst significantly influences propylene gas-phase epoxidation. NH₃-TPD-MS was employed to characterize the surface acidity of TS-1, Ni/TS-1, and Ni-Cs/TS-1 catalysts, as shown in Figure 7. All samples exhibited a weak NH₃ desorption peak at approximately 170 °C, corresponding to weakly acidic sites, indicating that all the catalysts primarily comprised weakly acidic sites. TS-1 displayed only weak acidity with minimal ammonia desorption (0.35 μmol/g), suggesting limited surface acid sites. In contrast, Ni/TS-1 showed significantly increased acidity: ammonia desorption from weak acid sites rose to 5.49 μmol/g, while an additional desorption peak (1.00 μmol/g) emerged at ~345 °C. This medium–strong acidity indicates substantial acidification after Ni incorporation. However, the Cs modification’s acid site quantities decreased substantially: weak sites diminished to 0.32 μmol/g, moderate sites to 0.06 μmol/g, and strong sites to 0.23 μmol/g, confirming the overall acidity reduction [56]. These results demonstrate that alkali metals modulate acid site density and strength on Ni/TS-1, optimizing the acidic profile and enhancing catalytic performance [57].

Ultraviolet–visible absorption techniques were employed to investigate the binding interactions of titanium with the silica framework [58]. The UV–vis analysis (Figure 8) indicated that the primary absorption peak of TS-1 was situated at 210 nm, signifying the presence of tetrahedrally coordinated Ti species. Compared to TS-1, the Ni/TS-1 sample exhibited a new characteristic absorption peak at 310 nm. This signal corresponds to TiO₂, likely resulting from the dissolution and recrystallization of some framework titanium under alkaline synthesis conditions. The absorption peak observed at 274 nm is associated with octahedrally coordinated titanium species. Its formation mechanism can be attributed to the etching effect of OH⁻ on the TS-1 framework within the alkaline preparation environment. After loading Ni and Cs⁺, the intensities of the absorption peaks near 210 nm and 274 nm decreased. This indicates that the alkali metals etched the TS-1 framework, leading to an increase in lattice defects. Consequently, the hydrophobicity improved, which is consistent with the results from the XRD patterns and ²⁹Si NMR spectra (Figure 1 and Figure 6). After loading the Ni metal, the Ni/TS-1 sample exhibited an expanded absorption band near 500 nm, attributed to the surface plasmon resonance properties of Ni nanoparticles.

Figure 9 compares the H₂-TPR profiles of unmodified and Cs⁺-modified Ni/TS-1 catalysts. The 5%Ni/TS-1 catalyst exhibited a primary reduction peak at 370 °C, corresponding to the reduction of Ni²⁺ to Ni⁰. The introduction of Cs⁺ shifted this peak to higher temperatures (ΔT = 35 °C), indicating strengthened metal–support interactions. This shift arose from electron donation from Cs⁺ to adjacent Ni species, which stabilized Ni²⁺ through increased electron density at the Ni-O bonds. Consequently, greater thermal energy was required for the Ni²⁺ reduction, consistent with the observed temperature elevation. Furthermore, the strengthened Ni-O-Ti bonds and altered H₂ dissociation pathway kinetically hindered the reduction process, requiring higher activation energy, as observed in H₂-TPR. When the Ni loading increased from 5% to 9%, the Ni reduction temperature decreased. After Cs⁺ modification, that temperature increase became less pronounced. This correlation likely arose from Ni particle size differences. The average Ni particle sizes of 5%Ni-Cs/TS-1 and 9%Ni-Cs/TS-1 were 3.33 nm and 3.04 nm (Figure 4). Based on TEM observations indicating an increase in easily reducible small Ni particles with a higher Ni loading, the area of the low-temperature peak (0–200 °C) increased accordingly. Concurrently, a portion of nickel may become incorporated into the support lattice, requiring reduction at significantly higher temperatures (500–800 °C).

Figure 10 presents the C₃H₆-TPD-MS results for the TS-1, Ni/TS-1, and Ni-Cs/TS-1 catalysts. The absence of the C₃H₆ desorption peak on TS-1 indicated weak propylene adsorption. Ni/TS-1 showed two desorption peaks at ~210 °C and 351 °C. A comparison of peak areas revealed that Ni-Cs/TS-1 exhibited the largest propylene adsorption peak area. This demonstrates comparatively superior propylene adsorption capacity, which facilitates PO desorption [59] by promoting surface reaction kinetics, thereby enhancing PO production, mitigating catalyst deactivation, and improving stability.

2.2. Catalytic Performance

The catalytic efficacy of Ni/TS-1 and 5%Ni-Cs/TS-1 for gas-phase propylene epoxidation with H₂ and O₂ was evaluated in a continuous-flow quartz reactor. Before testing, the catalysts underwent reduction treatment. The optimal Ni loading was determined through a systematic performance evaluation.

The experimental conditions were systematically evaluated. Through a comparison of propylene oxide (PO) selectivity, PO formation rate, hydrogen utilization efficiency, and propylene conversion, the optimal parameters were established as follows: reduction temperature of 500 °C (Figure S6), reducing gas composition of N₂:H₂ = 18:1 (25 mL/min) (Figure S7), and reduction duration of 6 h (Figure S8). Space velocity effects were additionally investigated (Figure S9). When space velocity increased from 10,000 to 14,000 mL·g⁻¹·h⁻¹ using the 5%Ni-Cs/TS-1 catalyst, the PO formation rate initially increased then subsequently declined. At high space velocities (>12,000 mL·g⁻¹·h⁻¹), shortened reactant–catalyst contact time decreased propylene conversion. Consequently, 12,000 mL·g⁻¹·h⁻¹ optimized both economic and catalytic efficiency for propylene epoxidation. Based on the PO formation rate optimization, the reaction gas ratio was maintained at N₂:H₂:C₃H₆:O₂ = 7:1:1:1 (Figure S10).

Figure 11 demonstrates that the unmodified Ni/TS-1 catalyst, with various Ni loadings, exhibited significantly low propylene conversion, PO selectivity, hydrogen utilization efficiency, and PO formation rate. This poor performance was attributed to nickel particle agglomeration and inadequate reduction. To address this, we modified the Ni/TS-1 catalyst with CsOH·H₂O and optimized the Ni:Cs molar ratio (Figures S11–S13). Catalysts with fixed Ni concentrations (3%, 5%, 7.5%) and varying Cs additions (corresponding to Ni:Cs ratios of 1:1, 1:0.5, and 1:0.25) were evaluated. The catalyst with a Ni:Cs ratio of 1:0.5 demonstrated optimal catalytic performance.

Figure 12 presents the performance of Cs⁺-modified catalysts with varying Ni loadings. Compared to the unmodified catalysts, the Cs⁺-modified catalysts exhibited significantly enhanced propylene conversion, PO selectivity, hydrogen utilization efficiency, and PO formation rate. This enhancement was attributed to Cs⁺ promoting the transformation of the Ni coordination structure to a tetrahedral configuration and facilitating the reduction of Ni²⁺ to Ni⁰, as evidenced by XPS data. Ni⁰ served as the active site for propylene epoxidation and oxygen activation. After Cs⁺ modification, the adsorption capacity for propylene could be further improved, enhancing the hydrophobicity of the catalyst. Cs⁺ modification has regulatory mechanisms on multiple aspects, including Ni particle size, surface properties, and propylene adsorption capacity, etc. Catalytic performance improved when increasing the Ni loading up to 9 wt%, after which it declined with further increases in Ni content. Consequently, the 9%Ni-Cs(1:0.5)/TS-1 catalyst demonstrated optimal performance.

Our catalytic studies revealed that the major oxidation products over Ni/TS-1 were propanal, acrolein, and CO₂. The modification with Cs⁺ significantly enhanced PO selectivity on Ni-Cs/TS-1. The kinetic analysis indicated that under Ni/TS-1 catalysis, both acrolein and PO formed as primary products via allylic oxidation and epoxidation pathways, respectively. However, PO underwent isomerization to allyl alcohol, which was further oxidized to acrolein or transformed to propanal; the subsequent deep oxidation of these intermediates yielded CO₂. Critically, Cs⁺ inhibited the isomerization of PO, thereby suppressing its transformation into secondary oxidation products. This blockage constitutes the primary mechanism for the enhanced PO selectivity following Cs⁺ modification.

Figure 13 presents the propylene epoxidation performance of Cs⁺-modified catalysts with varying Ni loadings (5–11 wt%) across 180–230 °C. Catalytic efficiency increased with the Ni loading up to 9 wt%, beyond which it decreased. Titanium (Ti) sites catalyzed propylene activation, while nickel (Ni) sites facilitated peroxide formation from H₂ and O₂. The optimized Ni/Ti molar ratio at 9 wt% Ni generated sufficient peroxides for epoxidation without excessive formation that would promote C₃H₆ over-oxidation to CO₂. At 11 wt% Ni, however, excessive Ni species physically obstructed Ti active sites on the TS-1 support. This site blockage suppressed propylene oxide formation efficiency. As the temperature rises, the selectivity of by-products escalates, whilst the selectivity of PO diminishes. While elevated temperatures enhance propylene conversion, they also increase the likelihood of by-product formation, leading to CO₂ becoming the predominant product. Consequently, we designated 200 °C as the ideal reaction temperature. At 200 °C, the propylene conversion rate of 9%Ni-Cs/TS-1 catalyst was 14%, the PO selectivity was 87.8%, and the formation rate of PO was 382.9 g_PO·kg_cat⁻¹·h⁻¹. Based on a comparative analysis of research on catalysts for propylene epoxidation to propylene oxide, the 9%Ni-Cs/TS-1 catalyst developed in this work exhibited superior performance compared to existing Au-based catalysts reported in recent years and all other Ni-based catalysts, as shown in Table S1.

The stability of the 5%Ni-Cs/TS-1 catalyst was evaluated at 200 °C and a space velocity of 7000 mL·g_cat⁻¹·h⁻¹ (Figure 14). The catalyst maintained high activity for the initial 30 h and demonstrated stable performance over 150 h. After 150 h on stream, the average propylene oxide (PO) formation rate reached ~200 g_PO·kg_cat⁻¹·h⁻¹ with consistent hydrogen efficiency. High-resolution transmission electron microscopy (HRTEM) confirmed the negligible growth of Ni particles during the reaction, with particle sizes comparable to pre-reaction measurements. These results establish the robust stability of the 5%Ni-Cs/TS-1 catalyst under continuous operation.

3. Materials and Methods

3.1. Materials

Tetraethyl orthosilicate (Sinopharm Chemical Reagent Co., Ltd., AR, Shanghai, China), tetrapropylammonium hydroxide (Adamas, 40 wt% in water, AR, Shanghai, China), Tween 20 (Acros, AR, Beijing, China), Ni(NO₃)₂·6H₂O (99.99%, Sigma-Aldrich, St. Louis, MO, USA), tetrabutyl titanate (Alfa, AR, Ward Hill, MA, USA), isopropanol (Acros, 99.5%, AR, Beijing, China), CsOH·H₂O (Aladdin, AR, Beijing, China), NaOH (99.99%, Sigma-Aldrich, St. Louis, MO, USA), and deionized water.

3.2. Synthesis of TS-1

TS-1 (Si/Ti = 60) was synthesized via conventional hydrothermal synthesis [60]. Briefly, 2.0 g of Tween 20 and 18.8 g of tetrapropylammonium hydroxide (TPAOH) solution were dissolved in 32 mL of deionized water under stirring and agitated in a 40 °C water bath. Then, 35.6 g of tetraethyl orthosilicate (TEOS) was added to the mixture and vigorously stirred for 4 h to form silica sol. Separately, 0.48 g of tetrabutyl titanate (TBOT) was homogenized with 4.5 g of isopropanol (IPA), then added dropwise to the silica sol. The mixture was stirred for 2 h and subsequently de-alcoholized at 70 °C. The resulting gel was transferred to a Teflon-lined autoclave and crystallized at 170 °C for 72 h. The solid product was recovered by centrifugation, washed with deionized water, and dried at 100 °C. Finally, the material was calcined at 550 °C for 8 h to yield TS-1 as white powder.

3.3. Catalyst Preparation

A 0.2 g nickel precursor (5 wt% solution) was dissolved in 20 mL of deionized water. Then, 1.0 g of TS-1 was added to the solution. Moreover, a 0.5 M NaOH solution was added dropwise under stirring until pH 10–11 was achieved. The mixture was stirred for 2 h at room temperature. The solid product was collected by filtration, washed thoroughly with deionized water, and dried under vacuum at 100 °C for 12 h. This yielded the 1 wt% Ni/TS-1 catalyst.

Ni/TS-1 catalysts with higher loadings (3–11 wt%) were prepared similarly by adjusting nickel precursor quantities. These catalysts were subsequently modified via incipient wetness impregnation with CsOH·H₂O aqueous solutions, maintaining a fixed Ni:Cs molar ratio of 1:0.5 for all Ni-Cs/TS-1 samples. After 4 h of impregnation, the materials were dried at 80 °C. Before catalytic testing, all catalysts were pretreated in 5 vol% H₂/N₂ at 500 °C for 6 h.

3.4. Catalyst Characterizations

The crystalline structure of the samples was assessed using X-ray powder diffraction (Empyrean, Malvern Panalytical, Almelo, The Netherlands) run at 60 kV and 55 mA. UV–vis spectrum acquisition employed a Lambda 750 spectrometer (PerkinElmer, Waltham, MA, USA). Nickel content quantification was performed using a Shimadzu Corporation ICPS-8100 spectrometer (Shimadzu, Kyoto, Japan). Surface elemental composition and associated chemical states were evaluated through X-ray photoelectron spectroscopy (XPS) on a Thermo Scientific ESCALAB 250Xi instrument (Thermo Fisher Scientific, Waltham, MA, USA), referencing the C 1s peak at 284.8 eV as an internal standard. Physical structure properties were examined via nitrogen physisorption isotherms measured with a Micromeritics ASAP 2020 HD88 analyzer (Micromeritics, Norcross, GA, USA). Before analysis, all specimens underwent activation at 250 °C for 6 h. Ammonia temperature-programmed desorption mass spectrometry (NH₃-TPD-MS) experiments utilized a Micromeritics ASAP 2920 unit (Micromeritics, Norcross, GA, USA). Approximately 0.1 g of material was gradually placed into a reactor tube and conditioned at 200 °C for 2 h under a 30 mL·min⁻¹ helium stream. Following cooling to 50 °C, the sample was exposed to a 30 mL·min⁻¹ NH₃/He mixture (1 vol%) for 60 min. Subsequent heating to 100 °C under a 30 mL·min⁻¹ helium purge for 90 min eliminated physisorbed ammonia. Finally, the temperature was ramped to 600 °C at 10 °C·min⁻¹. Evolved species were detected by the coupled mass spectrometer (MKS Cirrus 2). Ammonia desorption quantities were calculated based on peak areas calibrated against pulses of known NH₃ concentration. C₃H₆-TPD-MS measurements followed an analogous process on the same equipment. High-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM, FEI Tecnai G2 F20, Thermo Fisher Scientific, Hillsboro, OR, USA) facilitated the nickel particle analysis. Size determination involved randomly measuring 300 distinct particles. The ²⁹Si solid-state NMR spectra were recorded on a Bruker AVANCE III HD spectrometer (400 MHz, Bruker, Billerica, MA, USA) applying the following conditions: π/2 pulse length of 1.5 μs, relaxation delay of 5 s, 820 scans accumulated, and a spinning rate of 10 kHz. Hydrogen temperature-programmed reduction (H₂-TPR) was performed using an AutoChem II 2920 chemisorption analyzer (Micromeritics, Norcross, GA, USA) with a 10% H₂/Ar gas mixture.

3.5. Catalytic Tests

Gas-phase propylene epoxidation with O₂ and H₂ was performed in a vertical fixed-bed quartz reactor (with an inner diameter of 6 mm). First, 0.6 g of catalyst was loaded between quartz wool plugs. Before reaction, the catalyst was pretreated in 5% H₂/Ar at 500 °C for 6 h with a heating rate of 10 °C/min. The reaction gas mixture (N₂:H₂:O₂:C₃H₆ = 7:1:1:1 vol%) was then fed at 10,000–14,000 mL·g⁻¹·h⁻¹. During the experiment, the reaction temperature was elevated from room temperature to 180 °C at 2 °C/min, followed by sampling at intervals of 10 °C from 180 °C to 230 °C. In the stability test, the reaction temperature was elevated from room temperature to 200 °C at a rate of 2 °C/min and sustained until the end of the experiment.

All effluent components from the reactor were monitored in real time using Agilent 7890B gas chromatographs (Agilent Technologies, Wilmington, DE, USA). A flame ionization detector (FID) equipped with an HP-INNOWAX column and high-purity helium carrier gas quantified acrolein, propene, acetone, acetaldehyde, propene oxide, and propanal. Carbon dioxide and propane concentrations were measured by a thermal conductivity detector (TCD) with Hayesep Q and 5A molecular sieve columns, also employing high-purity helium as the carrier gas. Hydrogen content was assessed via a separate TCD (5A molecular sieve column, high-purity argon carrier gas). The calculations for C₃H₆ conversion, H₂ efficiency, and product selectivity were defined as follows:

$${\mathrm{C}}_3{\mathrm{H}}_6 \mathrm{conversion}={\displaystyle \frac{\mathrm{moles} \mathrm{of} ((\mathrm{PO}+\mathrm{Pr}+\mathrm{Ac}+\mathrm{An})\times 3+\mathrm{Ad}\times {2+\mathrm{CO}}_2)}{\mathrm{moles} \mathrm{of} ({\mathrm{C}}_3{\mathrm{H}}_6 \mathrm{in} \mathrm{the} \mathrm{reaction} \mathrm{gases} \times 3)}}\times 100\%$$

$$\mathrm{Product} \mathrm{selectivity}={\displaystyle \frac{\mathrm{mole} \mathrm{of} (\mathrm{Product}\times \mathrm{number} \mathrm{of} \mathrm{carbon} \mathrm{atoms} \mathrm{in} \mathrm{product} \mathrm{molecule})}{\mathrm{moles} \mathrm{of} \left(\right(\mathrm{PO}+\mathrm{Pr}+\mathrm{Ac}+\mathrm{An})\times 3+\mathrm{Ad}\times 2+{\mathrm{CO}}_2)}}\times 100\%$$

$$\begin{array}{l}{\mathrm{H}}_2 \mathrm{efficiency}={\displaystyle \frac{\mathrm{mole} \mathrm{of} \mathrm{PO}}{{\mathrm{mole} \mathrm{of} \mathrm{H}}_2 \mathrm{converted}}}\times 100\%\\ \end{array}$$

4. Conclusions

This work employed the alkali metal Cs⁺ as a modifier to enhance the non-precious metal Ni/TS-1 catalyst via impregnation. The catalytic efficacy of Cs⁺ in gas-phase propylene epoxidation was systematically examined. Mechanistic insights were elucidated through multiple characterization techniques. Cs⁺ significantly enhanced the density of active sites and promoted propylene adsorption and PO desorption by improving the reducibility of Ni species, modulating the acidity/basicity, and optimizing the support–metal interaction. It was elucidated that Cs⁺ enhanced Ni/TS-1 performance through a triple mechanism, “size control–electron modification–hydrophobicity enhancement”, transcending conventional single-modification strategies. Optimal pretreatment conditions and reaction conditions were established: a reduction temperature of 500 °C, reducing gas composition of N₂:H₂ = 18:1, and reduction duration of 6 h. Using a space velocity of 12,000 mL·g⁻¹·h⁻¹, the reaction gas ratio was maintained at N₂:H₂:C₃H₆:O₂ = 7:1:1:1. The 9% Ni-Cs/TS-1 catalyst achieved 14% propylene conversion at 200 °C with 87.8% PO selectivity and a 382.9 g_PO·kg_cat⁻¹·h⁻¹ PO formation rate.

Ni-Cs/TS-1 is a superior non-noble metal catalyst. Based on the comparative analysis of catalysts for propylene epoxidation with H₂ and O₂ to PO, the 9%Ni-Cs/TS-1 catalyst developed in this work demonstrates performance metrics surpassing all other reported Ni-based catalysts and even exceeds those of noble Au-based catalytic systems. Moreover, the employment of the non-noble metal nickel catalyst presents considerable economic benefits. This study paves the way for employing non-noble metal catalysts that exhibit catalytic properties akin to those of gold-based catalysts in epoxidation processes.