Green Synthesis of Ultrathin MWW Zeolites With Tunable Al‐T Sites Using Nitrogen‐Free Secondary OSDAs for Catalytic Cracking

ABSTRACT

The direct synthesis of ultrathin MWW zeolites with precisely controllable aluminum location represents a significant challenge. Herein, we report a sustainable approach using cycloketones as secondary nitrogen‐free organic structure‐directing agents (NF‐OSDAs) to fabricate two‐dimensional (2D) MWW zeolites. Cyclobutanone (Cy4) directs the formation of ultrathin nanosheets featuring expanded interlayer spacing and tunable aluminum distribution. We demonstrate that Na⁺‐cycloketone complexes selectively occupy specific pore environments, effectively reducing the population of T₆–T₇ framework aluminum (Al_F) sites and migrating to T₂ Al_F sites, promoting external Brønsted acidity. The resulting Cy4‐MWW catalyst exhibits outstanding performance in bulky‐molecule transformations. Mechanistic studies reveal that cycloketones undergo in situ self‐condensation to generate dimeric species that function as pivotal structural directors. Their steric presence affects layer thickness and nanosheet packing, whereas their Na⁺‐coordinated forms strongly interact with T₂ Al_F sites, thereby driving aluminum migration from the confined T₆–T₇ Al_F positions to accessible T₂ Al_F sites. This strategy exhibits broad versatility across a range of cycloketones (Cy4–Cy10), enabling the concurrent tuning of both the MWW architecture and its aluminum distribution.

Nitrogen‐free cycloketones direct the sustainable synthesis of ultrathin MWW zeolites, enabling dual control over nanosheet morphology and precise aluminum location. Through steric hindrance and competitive stabilization, framework aluminum migrates from confined T₆–T₇ to accessible T₂ sites, creating an optimized acidic architecture that achieves exceptional performance in bulky molecule transformations.

1. Introduction

Zeolites with the MWW topology represent a unique class of two‐dimensional (2D) layered materials. They are distinguished by their complementary pore systems, which comprise intralayer 10‐membered ring (10‐MR) sinusoidal channels and interlayer 12‐MR supercages or pockets [1, 2]. This distinctive architecture not only facilitates efficient mass transport but also creates confined “microreactors,” endowing MWW zeolites with exceptional catalytic prowess. Crafting ultrathin MWW nanosheets with enriched surface framework aluminum (Al_F) sites is highly desirable to maximize the exposure of these active sites. This approach simultaneously optimizes intracrystalline diffusion kinetics and active site accessibility [3, 4, 5], which is particularly beneficial for reactions involving bulky molecules.

Substantial efforts have thus been dedicated to the direct synthesis of ultrathin MWW zeolites. While top‐down delamination of layered MWW zeolites has been achieved, it often inflicts irreversible damage to the crystalline framework [6, 7]. Consequently, direct synthesis using sophisticated organic structure‐directing agents (OSDAs) has emerged as a more precise alternative [8], as illustrated in Scheme 1a. Recent advances, exemplified by single‐layer MIT‐1 (Ada‐i‐16) [9], disordered DS‐ITQ‐2 (C₁₆DC₁) [10], delaminated SCM‐1/6 (DCHA) [11], and ultrathin SL‐1/ML‐1 (TPOAC) [12], predominantly rely on complex, multi‐step synthesized, nitrogen‐containing surfactants or ammonium‐based molecules. However, this strategy faces two critical limitations. First, they always rely heavily on complex, multi‐step synthesized, nitrogen‐containing surfactants or ammonium‐based molecules. Their removal necessitates energy‐intensive calcination with the emission of hazardous nitrogen oxides (NO_x), which contradicts the principles of green chemistry. Second, and more fundamentally, a pivotal scientific challenge remains largely unresolved: the precise control of framework aluminum distribution among the eight distinct T‐sites (T₁–T₈) within the unique pore architecture of MWW zeolite (Scheme 1b,c). This site‐specific Al_F placement governs the strength and spatial arrangement of acid sites [13, 14, 15, 16, 17, 18], presenting a key opportunity for rationally tailoring catalytic properties, yet it remains a formidable target via direct hydrothermal synthesis.

SCHEME 1.

(a) Summary of traditional direct synthesis OSDA types, their limitations, and improvement directions. (b) Top‐down [001] view of 10‐MR sinusoidal channels and surface pocket network. (c) The visualization of the MWW framework structure includes T₁–T₈ non‐equivalent T atoms and two independent 12‐MR (red) and 10‐MR (blue) pore structures.

While the epitaxial growth of MWW zeolites on hexagonal‐boron nitride (h‐BN) substrates with the addition of cyclohexanone and HMI as dual OSDAs has been reported to obtain bilayer structures [19, 20], the direct synthesis of MWW zeolites with rationally regulated morphology and Al siting remains unexplored in normal systems. Herein, we report a novel and sustainable synthetic strategy that simultaneously addresses the challenges of morphology control, aluminum location, and framework types without epitaxial substrates. Our approach employs various cycloketones as eco‐friendly secondary nitrogen‐free OSDAs (NF‐OSDAs) in combination with hexamethylenimine (HMI). These cycloketones play a dual structure‐directing role. They not only promote the formation of utrathin nanosheets through in situ self‐condensation into dimeric species that limit the growth of nanosheets along the c‐axis direction and enable unprecedented control over Al_F distribution. The Na⁺‐cycloketone complexes competitively displace H⁺‐HMI, reducing Al occupancy at the T₆–T₇ Al_F sites and generate more accessible T₂ Al_F sites, thereby tailoring the acidic properties. The optimized Cy4‐MWW catalyst exhibits exceptional performance in the cracking of bulky molecules, directly linking the tailored active‐site geometry to enhanced catalytic efficiency. The remarkable versatility of this strategy is also demonstrated across a range of cycloketones, enabling simultaneous control over both the MWW architecture and its aluminum distribution.

2. Results and Discussion

2.1. Structure and Catalytic Performance of Ultrathin Cy4‐MWW Zeolite

The morphology and textural properties of the Cy4‐MWW zeolite, synthesized using cyclobutanone (Cy4) as a secondary organic structure‐directing agent (OSDA), were systematically investigated. In contrast to the densely packed and rigid “house‐of‐cards” architecture of conventional MCM‐22(P) (Figure S1), Cy4‐MWW(P) exhibits a more open structure consisting of loosely packed nanosheets with a highly pliable, smaller, and thinner morphology (Figure 1a,b). The ultrathin nature of Cy4‐MWW(P) was unequivocally confirmed by high‐resolution TEM (HR‐TEM) and atomic force microscopy (AFM). HR‐TEM analysis revealed a significant reduction in crystallite thickness along the c‐axis to merely 6–9.5 nm, corresponding to 2–3 MWW layers (Figure 1c,d and Figure S2), which is substantially thinner than the 21 nm (∼8 layers) observed for MCM‐22(P) (Figure S3). AFM measurements corroborated the ultrathin nature of Cy4‐MWW(P), revealing a layer thickness of 6.3–9.5 nm (2–3 layers) (Figure S4). Statistical analysis with TEM and AFM further indicated that the Cy4‐MWW nanosheets are predominantly 2–3 layers thick (Figures S5 and S6). This ultrathin and open structure, as opposed to the morphology of MCM‐22(P) with thicker, densely packed nanosheets, favors reduced mass‐transfer resistance between the nanosheets. HR‐TEM also showed that the MWW nanosheets have an expanded layer thickness reaching 2.9 nm (Figure S7), with the surface regions being thicker (∼3.3 nm) than the central zones, indicating possible surface lattice relaxation or species adsorption. Consistent with this altered surface chemistry, Cy4‐MWW(P) exhibited a higher water contact angle than MCM‐22(P) (Figure S8), confirming its more hydrophobic character. The selected‐area electron diffraction (SAED) and FFT pattern (Figure 1c and Figure S9) confirmed the high crystallinity of Cy4‐MWW(P).

FIGURE 1.

(a, b) FE‐SEM images and magnified regions of Cy4‐MWW(P); (c, d) HR‐TEM images and magnified regions of Cy4‐MWW(P); (e) small‐ and wide‐angle XRD spectra of Cy4‐MWW(P) and MCM‐22(P); (f) ²⁹Si MAS NMR spectrum of Cy4‐MWW(P); (g) ²⁷Al MAS NMR spectrum of Cy4‐MWW(P); (h, i) conversion of TiPB and selectivities to Bz and iPB over H‐type MCM‐22 and Cy4‐MWW at different reaction temperatures; (j) catalytic cracking of TiPB and iPB via continuous pulse reaction over H‐type Cy4‐MWW and MCM‐22 zeolites; (k) normalized thermogravimetric curves of LDPE cracking over two H‐type MWW zeolite catalysts and that without catalyst under temperature‐programmed conditions; (l) normalized thermogravimetric curves of LDPE cracking over two H‐type MWW zeolite catalysts during isothermal thermogravimetry at T₁₀ = 330°C (inset: optical photographs of the catalyst and residual substrate); (m) product distribution of catalytic flash pyrolysis of LDPE over two H‐type MWW zeolite catalysts and that without a catalyst at 550°C.

X‐ray diffraction (XRD) patterns confirmed the MWW topology for both Cy4‐MWW(P) and MCM‐22(P) (Figure 1e). Cy4‐MWW(P) shows a weakened (002) diffraction peak, indicating that most of the sample exists as nanosheets with fewer stacked layers along the c‐axis direction, along with the broadening of the (101) and (102) peaks, all consistent with the TEM observations. A larger layer thickness of 2.9 nm, calculated from the (001) peak for Cy4‐MWW(P), versus 2.6 nm for MCM‐22(P), signaling the intercalation of bulky molecules. Upon calcination, the (001) peak shifted to a position matching that of MCM‐22, corresponding to a layer thickness of ∼2.5 nm along with the HR‐TEM results (Figure S10), due to template removal and interlayer condensation. Broader (101) and (102) peaks persisted in the calcined sample (Figure S11) suggest the formation of ultrathin nanosheets. Ar physisorption analyses provided further insight into the textural properties. Cy4‐MWW exhibits a broadened hysteresis loop compared to the type I/IV hybrid isotherm of MCM‐22, indicating enhanced mesoporosity (Figure S12). The sustained uptake at high relative pressure (P/P ₀ > 0.9) suggests capillary condensation in meso/macropores formed by disordered packing of nanosheets, confirming a hierarchical pore structure. While the micropore surface area of Cy4‐MWW is slightly lower, its external surface area increases substantially to 200 m²/g versus the 130 m²/g for MCM‐22, and its disorder index (DI) rises to 0.42 (Table S1), consistent with highly disordered aggregation of the MWW nanosheets. BJH pore size distribution analysis confirms a broader meso/macropore range of 40–120 nm (Figure S12). ICP‐AES analysis (Table S1) shows comparable Si/Al ratios but a significantly higher Na/Al ratio for Cy4‐MWW(P) (0.95) versus MCM‐22(P) (0.41), indicating that Na⁺‐related species may be involved in the electrostatic balance with framework aluminum sites (as discussed later), even though HMI is an indispensable template for directing the crystallization of the MWW framework.

Solid‐state NMR spectroscopy provided direct evidence for the local structural modifications induced by Cy4. Deconvolution of the ²⁹Si MAS NMR spectrum for Cy4‐MWW(P) resolved six resonances at δ = −99, −104, −110, −113, −116, and −119 ppm (Figure 1f), corresponding to Q³[Si(1Al) and (SiO)₃SiOH] and Q⁴[T₂, T_1,3,4,5, T₈, T₇, T₆] silicon species in different environments [14]. The intensified Q³ signal results in a Q³/Q⁴ ratio of 0.12, higher than that of MCM‐22(P) (Figure S13a), reflecting an increased density of surface silanols resulting from an enlarged external surface. Further analysis reveals a decrease in silicon population at Q⁴ T₂ and T_1,3,4,5 sites, accompanied by an increase at T₆–T₇ sites (Table S2). More critically, two‐dimensional ²⁷Al multiple‐quantum MAS (MQ MAS) spectroscopy (Figure S14) provided high‐resolution separation of the overlapping quadrupolar‐broadened signals, which, together with high‐field (600 MHz, 14.1 T) ²⁷Al MAS NMR, revealed a pronounced redistribution of framework aluminum (Al_F). Deconvolution of the spectral asymmetry, guided by the ²⁷Al MQ MAS data, identified four Al_F species at δ _iso = 61, 57, 54, and 49 ppm (Figure S15 and Table S3), which are assigned to the T₂, T₃, T_1,4,5,8, and T₆‐T₇ sites, respectively [14, 21, 22]. Quantitatively, Cy4‐MWW(P) exhibits a significant increase in Al occupancy at the accessible T₂ Al_F sites, from 12.4% to 22.2% and a concurrent decrease at the confined T₆–T₇ Al_F sites, from 14.6% to 10.3% (Figure 1g and Figure S13b), yielding a markedly lower (T_{1,3,4,5,6,7,8})/T₂ ratio of 3.5 versus 7.1 for MCM‐22(P) (Table S4). This suppression of Al incorporation at T₆–T₇ Al_F sites and aggregation at T₂ Al_F sites suggests the probable displacement of structure‐stabilizing H⁺‐HMI by Na⁺‐cycloketone complexes during crystallization, as inferred from elemental analysis and prior studies [21]. To directly verify this hypothesis, we employed two complementary solid‐state NMR methods. ¹H{²³Na} transfer of populations in double‐resonance (TRAPDOR) MAS NMR (Figure S16) showed that the α‐H signal of cyclobutanone (∼3.1 ppm) is significantly attenuated under ²³Na irradiation, proving that Na⁺ is in close spatial proximity to cyclobutanone—that is, a Na⁺‐cycloketone complex is formed. Additionally, ¹H‐²⁷Al double‐heteronuclear multiple‐quantum coherence (D‐HMQC) NMR (Figure S17) revealed a clear cross‐peak between the T₂ Al site (δ _iso ≈ 61 ppm) and α‐H of Na⁺‐cycloketone complex (∼3.3 ppm), whereas no such correlation was observed for the T₆–T₇ Al site. These results provide direct experimental evidence that the Na⁺‐cycloketone complex selectively interacts with the T₂ site, driving the preferential incorporation of aluminum at this accessible position. The net effect is a strategic relocation of Brønsted acid sites from confined pore junctions to more accessible supercage environments, thereby creating an acidic site distribution highly beneficial for reactions involving bulky molecules.

To probe the consequent alterations in acidic properties, the as‐synthesized zeolites were first converted to their H‐type via NH₄ ⁺ exchange and subsequent calcination, followed by systematic characterization and catalytic tests. In situ FTIR spectroscopy revealed a higher concentration of external and interlayer silanols (bands at ∼3740 and 3726 cm⁻¹ [23, 24]) for Cy4‐MWW (Figure S18), consistent with its thinner nanosheets, enhanced structural openness, and incomplete interlayer condensation. This was further supported by a lower water contact angle of 24.8° versus 32.1° for MCM‐22 (Figure S19), while the opposite trend was observed for the as‐synthesized precursors (Figure S8). NH₃‐TPD profiles showed comparable total acid amounts and acid strength distributions for both zeolites (Figure S20 and Table S5). However, probe molecule FTIR spectroscopy revealed critical differences. Pyridine (Py) adsorption (Figure S21) shows bands at 1545 cm⁻¹ (Brønsted acid sites, BAS) and 1454 cm⁻¹ (Lewis acid sites, LAS) [25]. The BAS/LAS ratio for Cy4‐MWW is 3.47 versus 4.23 for MCM‐22 (Table S6), indicating a relatively stronger intrinsic Lewis acidity and slightly lower framework integrity. Moreover, adsorption of the bulky 2,6‐di‐tert‐butylpyridine (2,6‐DTBPy, ∼1.1 nm), which probes exclusively surface acid sites [26, 27], showed a significantly more intense band at 1617 cm⁻¹ for Cy4‐MWW (Figure S22). This superior Brønsted acidity on the external surface may result from its open framework structure coupled with the redistribution of Al_F to more accessible T‐sites.

The catalytic superiority of Cy4‐MWW, stemming from its tailored acidity and porosity, was demonstrated in the cracking of bulky molecules. For 1,3,5‐triisopropylbenzene (TiPB, ∼0.95 nm), which cannot access the MWW micropores [28], Cy4‐MWW exhibited excellent cracking activity and high selectivity toward light aromatics across all evaluated temperatures (280–350°C), highlighting superior external site accessibility (Figure 1h,i). In continuous pulse tests, Cy4‐MWW maintained 54% TiPB conversion after 30 reaction cycles (Figure 1j), significantly outperforming MCM‐22 and highlighting its anti‐coking capability and relatively stable framework (Figures S23–S26). For the smaller isopropylbenzene (iPB, ∼0.67 nm), Cy4‐MWW also showed enhanced stability, retaining > 99% conversion for over 11 pulses (Figure 1j). In the catalytic cracking of low‐density polyethylene (LDPE), Cy4‐MWW exhibited remarkable efficiency. Under programmed heating, its T₅₀ was 370°C, which is 13°C and 95°C lower than MCM‐22 and non‐catalytic pyrolysis, respectively (Figure 1k, Table S7). Isothermal cracking further showed Cy4‐MWW achieved 90% conversion in half the time required for MCM‐22 (52 vs. 120 min, Figure 1l). Flash pyrolysis‐GC analysis at 550°C (Figure 1m) revealed that Cy4‐MWW produced high selectivity to C_2–C₄ hydrocarbons (72%) and low yield of C₁₆₊ residues (11%), vastly outperforming MCM‐22 (40% and 25%). This performance is attributed to the abundant and accessible acid sites in Cy4‐MWW with unique framework aluminum distribution and the highly open external surface structure, facilitating efficient cleavage of bulky molecules [29, 30].

2.2. Mechanism of Morphology Control and Aluminum Location Control With Cycloketone

To elucidate how the cycloketone simultaneously directs both the morphology and aluminum redistribution observed in Cy4‐MWW, we first extended our investigation to cyclopentanone (Cy5) and cyclohexanone (Cy6) for a comparative study. Thermogravimetric analysis (TGA) of Cyx‐MWW(P) (x = 4, 5, and 6) showed a higher total mass loss compared to MCM‐22(P) (Figure S27 and Table S8), consistent with the presence of additional organic composition [17, 31, 32]. This was corroborated with solid‐state ¹³C MAS NMR spectra, which displayed distinct signals assignable to the cycloketones [33, 34, 35], alongside those of HMI signals (Figure S28) [17, 18]. Further evidence came from ATR‐FTIR spectroscopy, revealing the distinctive C═O stretching vibrations of the cycloketones (Figure S29) [36, 37, 38, 39]. Elemental analyses provided further evidence for the incorporation of cycloketones, demonstrating that cycloketones competitively displace HMI within the zeolites, a conclusion directly supported by the correspondingly lower HMI contents and higher C/N ratios (Table S9) [40].

The incorporation of different cycloketones led to distinct and correlated changes in morphology and composition. Contrary to expectation based on molecular size, FE‐SEM imaging revealed that Cy5/Cy6‐MWW(P) consists of more rigid, densely packed nanosheets (Figure S30) compared with Cy4‐MWW(P) (Figure 1a,b). AFM measurements revealed that Cy5/Cy6‐MWW(P) consist of thicker nanosheets of 14 and 19.2 nm (∼5 and ∼7 layers, respectively) and have a lower average layer thickness of approximately 2.7–2.8 nm (Figure S31). This is consistent with the TEM layer thickness statistics shown in Figure S32, and this counterintuitive result immediately suggested that a simple size‐filling model is inadequate.

Concurrently, a pronounced redistribution of framework aluminum was observed. The deconvolution of high‐field ²⁷Al MAS NMR spectra revealed a decrease in T₂ Al_F site occupancy from 22.2% in Cy4‐MWW(P) to 14.0% in Cy6‐MWW(P) (Figure S33 and Table S10). In parallel, the combined population of the T₆–T₇ Al_F sites increased from 10.3% to 13.2%, while the (T_{1,3,4,5,6,7,8})/T₂ ratio increased from 3.5 to 6.2 (Figure 2a). Complementary ²⁹Si MAS NMR data confirmed the corresponding framework modification, revealing a decrease in the population of Si‐Q⁴ sites at T₆–T₇ positions from 24.6% in Cy4‐MWW(P) to 22.9% in Cy6‐MWW(P), alongside an increase at T₂ sites from 22.4% to 24.1% (Figure S34 and Table S2). Notably, Cy4‐MWW(P) possesses a higher Na/Al ratio (0.95) than Cy6‐MWW(P) (0.51) (Table S11), which may be linked to its greater enrichment of aluminum at the more accessible T₂ sites.

FIGURE 2.

(a) The Al_F T site proportion of different MWW zeolite samples and the ratio of different Al_F sites; (b) DFT Gibbs free energy analysis of Cy4–Cy6 and HMI combined with different ions on T₆, T₇, and T₂ Al sites in the MWW zeolite framework; (c) DFT simulation results of the changes in system Gibbs free energy caused by OSDAs, including Cy4 and HMI species occupying the 12‐MR and 10‐MR pores near the T₆ and T₇ Al site.

The DFT calculations were performed in an attempt to elucidate the aluminum redistribution introduced by cycloketones. From a thermodynamic perspective, the Gibbs free energy (Δ_r G) of interaction for Na⁺‐cycloketone complexes with Al_F sites is significantly less favorable than that for H⁺‐HMI (Figure 2b), although Na⁺ enhances the interaction of cycloketones with Al_F sites (Table S12). This implies that Na⁺‐cycloketone complexes are unlikely to displace the more strongly bound H⁺‐HMI and would not be expected to alter the aluminum distribution. This theoretical prediction stands in contrast to the experimental observation of decreased Al occupancy at T₆–T₇ sites alongside enrichment at the T₂ site. Even if we assume that the smaller Cy4 could, due to its size, access the 10‐MR channel and partially displace H⁺‐HMI near the T₆–T₇ Al_F sites (Figure S35), the calculations based on four representative ion‐OSDA pair configurations in the adjacent 12‐MR and 10‐MR pores indicate that introducing Na⁺‐Cy4 consistently lowers the total interaction energy relative to systems containing H⁺‐HMI (Figure 2c and Table S13), leading to inferior stabilization for T₆‐T₇ Al_F sites. Notably, the interaction of Cy4 with the T₂ Al_F is also lower than that of HMI, which cannot explain the experimental observation of T₂ enrichment. Moreover, the larger Cy6, which is hindered from entering the 10‐MR channel, interacts primarily with the T₂ site, as its binding free energy is similarly less favorable than that of H⁺‐HMI, thus failing to explain the experimentally observed Al redistribution. In addition, the current model cannot rationalize the observed morphological differences. Although Cy6 has a larger molecular size than Cy4 and would be expected to impose greater spatial constraints, it unexpectedly leads to thicker nanosheets and smaller interlayer spacing.

A critical clue to this possibility emerged from the synthesis process itself. The distinct color yellowing observed in the synthesis solutions, which also occurred in control experiments under identical pH conditions without silicon and aluminum sources (Figure 3a), indicated chemical transformations of the cycloketones. As pure Cy4 and Cy6 reagents are colorless or pale yellow (Figure 3b), this pronounced yellowing suggested the in situ formation of new organic species via plausible reactions involving the cycloketones. To elucidate the mechanistic foundation of the observed structural evolution, we first investigated the potential reaction pathways involved in the cycloketone‐induced MWW zeolite synthesis. Two reactions were considered: Stork enamine formation between the cycloketones and HMI [41, 42] or aldol self‐condensation of the cycloketones themselves (Figure 3c) [43, 44, 45, 46]. Analysis of the synthesis liquor by GC‐MS definitively excluded the formation of Stork enamine products. Instead, it identified aldol self‐condensation as the primary route, generating dimeric species (Figure S36). The dimeric products varied depending on the distinct reactivity of each cycloketone. It is therefore proposed that these in situ formed dimers function as space‐filling structural motifs within the interlayer region of the zeolite precursor. Their steric presence disrupts the conventional face‐to‐face packing of MWW nanosheets by inducing disorder and ultimately promoting the formation of more meso/macro pores in the MWW zeolite.

FIGURE 3.

(a) Optical images of Cyx and HMI in silicon‐ and aluminum‐free solution system at different times; (b) optical images of different pure OSDAs reagents and zeolite precursor powders obtained using different cycloketones as the secondary OSDA; (c) Stork(1) and Aldol(2) reactions of HMI with different cycleketones; (d1–d4) FE‐SEM images of MWW zeolite precursors with different ratios of HMI and Cy6 or using DiCy6 as the secondary OSDA; (e1–e4) HR‐TEM images of MWW zeolite precursors with different ratios of HMI and Cy6 or using DiCy6 as the secondary OSDA with [001] direction; (f1–f4) HR‐TEM images of MWW zeolite precursors with different ratios of HMI and Cy6 or using DiCy6 as the secondary OSDA along the tilted [001] direction; (g) small‐ and wide‐angle x‐ray diffraction spectra of different MWW zeolite samples; (h) Ar sorption isotherms of different MWW zeolite samples, including the enlargement of the low‐relative‐pressure region.

Building upon the identification of in situ formed dimers, we further refined the synthesis strategy by modulating the HMI/cycloketone ratio. HR‐TEM and FE‐SEM analysis revealed distinct morphological changes when HMI/Cy6 ratio changed from 1:1 (1 eq) to 1:3 (3 eq). The nanosheets adopted a disordered packing mode (Figure 3d) from SEM results, markedly distinct from the house‐of‐cards morphology of Cy6‐MWW(P) and tended to become progressively thinner. The nanosheet thickness obtained by TEM analysts decreased from Cy6‐MWW(P) about 6–7 layers to Cy6‐MWW(P)‐3eq as few as 2–3 layers (Figure 3e,f and Figure S37), with a thinner average layer thickness of ∼2.5 nm. Besides, XRD peak evolution in (002) and (310) peaks showed a clear topological transition from MCM‐22(P) to the MCM‐49 type (Figure 3g and Figure S38). Consistently, in situ FTIR spectroscopy revealed lower intensity of interlayer silanol bands at 3726 cm⁻¹, indicating that interlayer condensation already occurred during synthesis (Figure S18). Correspondingly, with the Cy6 concentration increased, the calcinated samples external surface area rose progressively from 143 to 324 m²/g, due to the formation of thinner nanosheets, and the DI increased markedly from 0.30 to 0.71 (Figure 3h, Figure S39, Table S14). Concurrently, the total pore volume also grew from 0.58 to 0.93 cm³/g. The relationship between the measured external surface areas and the statistical layer thickness distributions is in reasonable agreement with the theoretical predictions (Figure S40 and Table S15). These changes were consistently supported by the fact that a higher concentration of Cy6 promotes the in situ formation of dimeric species (Figure S41), which were related to the crystallization toward thinner MWW nanosheets.

However, the exogenous addition of a high concentration of dimer, such as pure commercial DiCy6 reagent, impaired crystallinity (Figure 3e,g) and resulted in a reduced BET surface area and total pore volume (Figure 3h, Figure S39, Table S14). Analysis of the synthesis mixture revealed that the DiCy6‐MWW system was dominated by the dimer, alongside traces of trimeric and monomeric Cy6, confirming the dimer (DiCy6) as the primary active species (Figure S42). Besides, the poor crystallinity of DiCy6‐MWW directly led to weak interaction with the exogenous dimers. This resulted in their facile removal upon washing, manifesting in the characteristic light color of the precursor powder compared to Cy6‐MWW(P)‐3eq (Figure 3b). TGA further supported the structural evolution of those samples, showing a reduction in the high‐temperature mass loss associated with hydroxyl condensation (Figure S43 and Table S8), which is consistent with the possible formation of an MCM‐49‐like topology.

Importantly, this structural evolution driven by dimer formation was also accompanied by a systematic redistribution of framework aluminum. High‐field ²⁷Al MAS NMR analysis of samples synthesized with Cy6 and DiCy6 revealed a distinct trend: as the molecular size increased from monomeric Cy6 to dimeric DiCy6, the Al_F occupancy at T₂ sites rose substantially from 14.0% to 23.9%, while the population of T₆‐T₇ Al_F sites decreased from 13.2% to 9.3% (Figure 4a,b and Table S16). This aluminum redistribution directly correlated with the progression toward the MCM‐49 like topology. Systematic variation of the HMI/Cy6 ratio further supported this relationship, with increasing Cy6 concentration progressively elevating T₂ Al_F site occupancy and depleting T₆–T₇ Al_F sites (Figure 4a,b and Table S16). Elemental analysis also confirmed that higher cycloketone proportions or direct dimer addition both led to sodium enrichment within the different MWW zeolites (Figure 4c). For catalytic cracking, Cy6‐MWW‐3eq with a significantly larger external surface area (324 m²/g) than Cy4‐MWW and maintaining a similar T₂ Al_F fraction, exhibits even higher catalytic activity in TiPB and LDPE cracking (Figures S44 and S45, Table S7). This enhancement correlates with its higher amount of external Brønsted acid sites, as evidenced by stronger 2,6‐DTBPy adsorption (Figure S46), underscoring that enlarging the external surface area can also boost catalytic performance.

FIGURE 4.

(a) The ²⁷Al MAS NMR spectra of Cy6‐MWW(P) zeolites with different HMI/Cy6 ratios and using DiCy6 as a secondary OSDA, and (b) the proportions of Al_F T sites; (c) Na/Al ratios of different MWW zeolite samples with ICP‐AES analysis; (d) DFT Gibbs free energy analysis of HMI and cycloketone or its dimers combined with different ions on T₂ Al sites in MWW zeolite framework; (e) adsorption configuration of Cy6 and DiCy6 located in the 12 ‐MR pocket of MWW zeolite from the x and z directions. The distance between carbonyl oxygen and Na⁺ is labeled. The dipole moments (μ) of Cy6 and DiCy6 and the corresponding transferred charges (q) upon incorporation into the 12‐MR of MWW zeolite are presented.

Interestingly, DFT calculations provided a molecular‐level rationale for these observations if following the dimer model. The assessment of basicity via proton affinity revealed that the dimeric species exhibit enhanced basicity compared to their monomeric precursors (Figure S47, left). The calculations on the interaction with T₂ Al_F sites showed that Na⁺‐coordinated dimers possess markedly stronger Gibbs free energies for T₂ aluminum than the monomeric Na⁺‐cycloketone complexes (Figure 4d, left; Table S17). Notably, for certain configurations, the stabilization provided by Na⁺‐dimer complexes at T₂ Al_F sites can even surpass that of H⁺‐HMI. This is particularly evident for Cy4, where its hydroxylated dimer (DiCy4(2)), especially in its sodium‐coordinated form, demonstrates superior basicity and a significantly enhanced interaction with T₂ Al_F sites with the assistance of the hydroxyl group compared with that of the dehydrated dimer of DiCy4(1) (Figure S48 and Table S18). Such enhanced interaction facilitates the enrichment of aluminum at T₂ sites in Cy4‐MWW(P) as observed in the experimental results. Upon dimerization for Cy6, DiCy6 exhibits enhanced van der Waals and chemical binding interactions at T₂ Al_F sites, arising from enlarged molecular size and increased dipole moment, respectively (Figure 4e and Table S18). Unlike Cy4, Cy6 has a lower tendency to dimerize due to its lower ring strain, and therefore a higher concentration of Cy6 is required to generate a sufficient amount of the dimeric species for T₂ Al_F enrichment. DFT calculations (Figure S49) also reveal that when T₂ Al is present, the adjacent surface silanols (e.g., T₁‐OH) strengthen their interaction (−1.85 eV) with excess surface Na⁺‐cycloketone dimers due to the more negatively charged via an inductive effect. This leads to an excessive increase in the Na/Al ratio and cycloketone content as well as to thinning of the MWW zeolite nanosheets. This computational insight resolves the earlier puzzle posed by the monomer model, identifying the in situ generated, Na⁺‐associated dimeric species as the key agents responsible for both the preferential enrichment of aluminum at T₂ sites and the steering of crystallization toward the MCM‐49‐like topology and thinner morphology.

Based on the comprehensive experimental evidence and DFT calculation results, we propose that governing the selection and application of cycloketones by their molecular size and proportion enables the rational construction of different distinct MWW‐type zeolites. The samples directed by HMI/cycloketone exhibit the characteristics of T₂ Al_F site enrichment and T₆–T₇ Al_F site depletion, accompanied by the formation of nanosheets of varied thickness with changed interlayer spacing. As shown in Scheme 2a,b, the addition of specific cycloketones drives the as‐synthesized zeolite to the T₂ Al_F enriched ultrathin MWW topology. This route is exemplified by small, highly self‐condensing cycloketones such as Cy4 (Route 1), which constitute the major organic component. Their small size grants them access to the 10‐MR sinusoidal channels. Here, Na⁺‐Cy4 complexes are likely to displace H⁺‐HMI, promoting the escape of Al_F from the confined T₆–T₇ sites. Simultaneously, the in situ formed hydroxylated dimer DiCy4(2), combined with Na⁺, occupy the interlayer region or the external surface with 12‐MR pockets and exhibits a stronger interaction affinity for T₂ Al_F sites compared to H⁺‐HMI. The combined effect of monomer‐mediated T₆–T₇ Al_F site depletion and dimer‐driven T₂ Al_F site enrichment promotes the aluminum distribution different from that of MCM‐22(P). However, the non‐hydroxylated dimers of DiCy4(1), with lower interaction affinity for the T₂ Al_F sites, act primarily as steric hindrances, or so‐called “physical spacers,” to induce an expansion of the interlayer spacing to ∼2.9 nm, while their presence on the external surface further contributes to the formation of thinner, flexible nanosheets consisting of only 2–3 MWW layers. Furthermore, the lower self‐condensation propensity of Cy6 or Cy5 at the same concentrations results in a less pronounced effect compared to Cy4, corresponding to less interlayer expansion with an average layer thickness of 2.7–2.8 nm and the formation of thicker nanosheets of approximately 5–7 layers.

SCHEME 2.

Direct synthesis of ultrathin MWW, MCM‐22(P), and MCM‐49‐like zeolites by cycloketone‐mediated control of (a) morphology and (b) aluminum distribution (for clarity, HMI has been omitted from part of the skeleton); (c) dimer or large‐sized cycloketone‐induced T₂ Al site enrichment enhances interlayer Si–OH condensation during the hydrothermal synthesis.

Correspondingly, both increasing the Cy6 proportion and directly adding commercial DiCy6 can shift the synthesis toward Route 2, which tends to yield T₂ Al_F enriched, thin‐layered MWW zeolites with MCM‐49 topology. In this regime, dimers become the predominant organic species. Their accumulation on the external surface imposes steric hindrance that limits c‐axis direction growth and leads to thinner nanosheets (Figure S50). The strong interaction of those Na⁺‐dimers with T₂ Al_F sites, driven by enhanced basicity, plays a decisive role in directing the framework. Critically, the action of Na⁺‐dimers diverges depending on their strength of interaction with T₂ Al_F sites. Different from the dimers DiCy4(1) acts as “physical spacers” with weaker interaction with T₂ Al_F sites in Route 1, the dimers of larger DiCy6 complexes near the 12‐MR supercages within Route 2 primarily operate as “chemical cross‐link promoters” for their stronger interaction for T₂ Al_F sites. This pronounced T₂ Al_F enrichment electronically activates the neighboring silanol groups such as T₁‐OH, by increasing their oxygen electron density compared with that of T₂‐Si sites_. The enhanced nucleophilicity of these activated T₁‐OH groups promotes dehydration condensation between adjacent layers, thereby facilitating the formation of covalent interlayer connections (Scheme 2c). Consequently, Route 2 generates an MCM‐49‐like topology with ultrathin nanosheets and reduced interlayer spacing, directly supported by the few‐layered morphology observed in high‐concentration Cy6 systems.

2.3. Extending Cycloketone OSDAs With Larger Molecules to Synthesize MWW Zeolite

Based on the established synthetic mechanism of MWW zeolite in the system with cycloketone as the secondary OSDAs, we systematically extended this strategy to those including dicarbonyl cyclohexanones (m‐BiCy6 and p‐BiCy6) and larger‐ring ketones (Cy8 and Cy10). A clear correlation was observed between molecular structure and self‐condensation propensity (Figure 5a). The coloration of the precursor powders and synthesis mixtures, ranging from white for Cy10 to dark gray for p‐BiCy6, arising from conjugation in the oligomeric products, served as a visual indicator of the condensation extent (Figure 5b and Figure S51). XRD analysis confirmed that MCM‐49 like topology was successfully yielded in the systems containing large scale cycloketones (Cy8, Cy10) or high steric hindrance (p‐BiCy6), but it failed crystallization in that containing m‐BiCy6 (Figure 5c). TGA analysis further indicated a trend toward the MCM‐49 like topology in the Cy8 and Cy10 systems, as evidenced by a pronounced reduction in the mass loss associated with interlayer condensation (Figure S52 and Table S19).

FIGURE 5.

(a) The product of possible self‐condensation of different cycloketones during the hydrothermal process (structure within the black dashed line was not generated); (b) optical images of different pure cycloketone reagents and zeolite precursor powders obtained using different cycloketones as the secondary OSDA; (c) XRD results of sample precursors synthesized using different cycloketones as the secondary OSDA; (d) Ar sorption isotherms of different MWW zeolite samples, including the enlargement of the low relative pressure region; (e–j) HR‐TEM images of MWW zeolite precursors using Cy8,Cy10 and p‐BiCy6 as the secondary OSDA; (k) the ²⁷Al MAS NMR spectra of MWW zeolites with different cycloketones as secondary OSDA.

Characterization further revealed that the zeolite derived from the bulky, non‐polymerizing adamantanone (Cy10) exhibited superior textural properties, including high BET surface area and external surface area, and pronounced hierarchical architecture with packed nanosheets (Figure 5d, Figure S53, Table S20). TEM imaging provided direct morphological evidence: Cy10‐MWW comprised highly crystalline, hexagonal nanosheets with a bilayer thickness of ∼5 nm (Figure 5g,h, Figures S54 and S55), establishing it as the optimal OSDA for fabricating ultrathin MCM‐49 like crystals with only two connected MWW layers (i.e., a single cell thinness). In contrast, p‐BiCy6‐MWW was surrounded by polymeric aggregates that suppressed crystal growth (Figure 5i,j), and Cy8 with the in situ formed bulky dimer like DiCy8 yielded a lower crystallinity of zeolite product (Figure 5e,f) due to its hindrance for zeolite growth. These findings indicate that excessive self‐condensation can compromise structural integrity.

Crucially, this series of zeolites provided definitive validation of the proposed aluminum redistribution mechanism. ²⁷Al MAS NMR analysis revealed an intensified trend in Al_F siting with larger or dimer‐favoring cycloketones. While the stabilizing effect of monomeric Cy8 on the T₂ Al_F site is a litter weaker than that of H⁺‐HMI, its dimeric form (DiCy8) exhibits the strongest interaction. Furthermore, Cy10 cannot form a dimer; it itself also demonstrates a potent interaction with the T₂ Al_F site (Figure 4d, right; Table S17). Consequently, Cy8‐MWW and Cy10‐MWW exhibited remarkably high T₂ Al_F site occupancies of 26.1% and 27.2%, respectively, accompanied by a drastic decrease in T₆–T₇ Al_F populations to 8.7% and 6.7% (Figure 5k and Table S21). This pronounced T₂ Al_F enrichment was paralleled by a significant increase in sodium content, with Na/Al ratios reaching 1.5–2.0 for these samples, the highest values observed in this study (Table S22). This trend underscores that cycloketones with larger steric structures are highly effective in directing Al_F to more accessible T₂ sites, even at equal ketone/HMI ratios, and that their enhanced basicity strongly promotes sodium incorporation within the framework (Figure S47, right). Owing to its higher T₂ Al_F fraction compared to Cy4‐MWW, Cy10‐MWW exhibits higher catalytic activity in bulky molecule cracking, demonstrated by its high amount of external Brønsted acid sites (Figures S44–S46 and Table S7). The extent of self‐condensation and the size‐compatibility of the products with the 12‐MR supercages were found to critically modulate this effect. For instance, p‐BiCy6 forms large oligomers (e.g., pentamers) that likely reside on the external surface of nanosheets rather than within the 12‐MR supercages (Figure S50), resulting in a more moderate Al_F redistribution (T₂ ∼18.5% and T₆–T₇ ∼12.1%). These results collectively confirm that the enhancement of cycloketone self‐condensation or increase of molecular size/dimensions facilitates T₂ Al_F site redistribution and thereby steers crystallization toward the specific MWW framework configuration. From a techno‐economic perspective of the synthesis progress (Figure S56), the NF‐OSDAs cycloketone‐directed synthesis of ultrathin MWW zeolites shows promising viability. Based on laboratory‐scale reagent prices, the estimated direct cost per kilogram of catalyst is approximately $27–29 (Table S23), with the series of typical cycloketones (e.g., Cy4, Cy6, and Cy10) itself accounting for only 2%–8% of the total cost (Figure S57). The synthesis employs standard equipment and common reagents, and the procedure is robust and reproducible. Moreover, the use of N‐free secondary OSDAs also offers a greener route (reduced NO_x emissions) and economically attractive approach for preparing high‐performance MWW zeolites.

3. Conclusion

This study has established a versatile and sustainable synthesis strategy for MWW zeolites in which various cycloketones are used to simultaneously control framework morphology and aluminum distribution. By employing simple cycloketones as eco‐friendly secondary nitrogen‐free OSDAs (NF‐OSDAs) to exhibit a dual structure‐directing function. They govern morphology through steric modulation, wherein in situ formed dimers induce the thinness of the nanosheets or even direct bilayered MCM‐49 crystallization and simultaneously steer the precise placement of Al_F from confined T_6–T₇ junctions to more accessible supercage T₂ environments. Critically, the enrichment of aluminum at T₂ sites is driven by the strong interaction of in situ generated Na⁺‐coordinated dimers with enhanced basicity. This site‐specific aluminum migration, governed by competitive stabilization between Na⁺‐cycloketone complexes and protonated HMI, creates an optimized acidic architecture rich in more accessible acid sites. The resulting zeolites demonstrate exceptional performance in bulky molecule cracking, directly linking tailored active‐site geometry to enhanced catalytic efficiency. This work provides not only a general and sustainable route to customize zeolite properties but also fundamental insights into aluminum location control, opening new avenues for the rational design of advanced catalytic zeolite materials.

Author Contributions

Haocheng Zhang: conceptualization, methodology, data curation, investigation, validation, formal analysis, visualization, writing – original draft, writing – review and editing. Tianyu He: writing – original draft, writing – review and editing, validation, software, data curation. Chuang Liu: investigation, validation, methodology. Linhai He: formal analysis, investigation, methodology. Zhaoqi Ye: writing – review and editing, formal analysis, writing–original draft. Kexin Yan: writing – review and editing, investigation. Yifan Zhang: writing – review and editing. Zhendong Wang: resources, supervision, writing – review and editing. Shutao Xu: supervision, investigation, methodology, resources. Hongbin Zhang: writing – review and editing, supervision, resources, methodology, investigation, funding acquisition, formal analysis. Xiao‐Ming Cao: writing – review and editing, software, data curation, supervision, resources, project administration, investigation, funding acquisition. Yahong Zhang: project administration, writing – review and editing, resources, supervision, funding acquisition, investigation. Yi Tang: conceptualization, methodology, investigation, writing – review and editing, project administration, resources, supervision, formal analysis, funding acquisition.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File: anie73482‐sup‐0001‐SuppMat.docx.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.