The development of efficient, selective, and robust heterogeneous photocatalysts for use in organic synthesis is one of the most attractive areas in photocatalysis [1, 2]. Among organic synthetic reactions, reactions, for the selective oxidation of an alcohol to produce the corresponding aldehyde, especially using O₂ as the oxidant, are crucial from the perspective of science and engineering[3, 4]. Various types of semiconductors, such as Bi₂WO₆, Fe₂O₃, and MoS₂, among others, have been reported to be effective for selective alcohol oxidation. The semiconductor TiO₂ is still one of the most active and feasible photocatalyst candidates [5-7]. However, when TiO₂ is used as the photocatalyst for the selective oxidation of alcohols, additives such as photosensitizers and acid co-catalysts are always necessary to further enhance its activity, resulting in rather complex reaction systems [8]. Moreover, the activity and selectivity towards the aldehyde are still limited in such reaction systems. In recent studies, gold nanoparticles with surface plasmon resonance effects have been loaded on TiO₂, resulting in greatly enhanced photoactivity and selectivity towards alcohol oxidation with O₂ as the oxidant. Nevertheless, compared with transition metals, the use of noble metals such as gold greatly increase the cost of the process [9, 10]. Therefore, the development of photocatalysts that can outperform TiO₂ for efficient and selective alcohol oxidation using O₂ as the oxidant is highly desired.

Titanosilicate molecular sieve (TS-1) is a siliceous zeolite containing titanium oxide Ti–O_x (x indicates the number of oxygen atoms coordinated to the Ti atom) species covalently linked to a porous SiO₂ framework [11-13]. It has previously been identified as an oxidation catalyst for organic substrates in traditional catalysis [14]. The application of TS-1 as a photocatalyst has recently received a great deal of attention [15, 16]. The substrate reactivity and product selectivity depend strongly on the nature of the TiO_x species. The tetrahedral Ti–O₄ structure in TS-1, which is distinct from the octahedral Ti–O₆ of TiO₂, could contribute to its unique selectivity towards the oxidation reaction [17, 18]. Additionally, TS-1 has a richer pore system compared to bulk TiO₂, which affords more catalytic sites. The pore structure also allows it to incorporate other components easily to enrich the functionality of the photocatalysts. These advantages make TS-1 a promising photocatalyst candidate for selective oxidation.

Nevertheless, the foremost shortcoming of TS-1 is its limited photoactivity due to its poor charge separation. Furthermore, the surface area of microporous TS-1 is limited. Therefore, the development of high-surface-area TS-1 nanomaterials with improved charge separation for selective organic synthesis, including alcohol oxidation, is highly desirable. In recent research, 2D nanomaterials have attracted tremendous attention, as their 2D morphology endows them with unique and advantageous properties, such as high surface areas, large numbers of active sites on their surfaces, and shorter charge transfer distances as semiconductors, among others [19, 20]. Hence, the fabrication of 2D TS-1 nanosheets is a feasible way to effectively improve the charge separation of TS-1 molecular sieve materials. Another vital issue when TS-1 is applied as a catalyst is mass transfer during the reaction process. Due to the micropore size of approximately 0.55 nm in TS-1, the diffusion of both the reactants and products can be inhibited. If the product molecules created on the catalytic sites cannot gradually diffuse out of the micropore system within the desired time, unwanted products may result, leading to micropore blockage and coke formation [21-23]. Systematic meso-porous nanocatalysts such as MCM-41 [24], SBA-15 [25], and MCM-48 [26] have uniform, lamellar, and unique pore structures (2–50 nm) that can remarkably alleviate diffusion issues. Therefore, the synthesis of lamellar TS-1 molecular sieves with a micro-mesoporous hybrid morphology is becoming a useful technique to overcome poor mass transfer [27, 28]. Based on the considerations above, the fabrication of micro-mesoporous lamellar TS-1 nanosheets for effective photocatalytic selective alcohol oxidation would be meaningful.

To further improve charge separation, electron capturers or shuttles are often introduced to modify semiconductors [29]. Metals, especially noble metals, are most often applied for aerobic alcohol oxidation, and can also function as cocatalysts [30]. However, taking into account the high cost of noble metals and industrial feasibility considerations, transition metals represent promising alternatives. More importantly, the introduced transition metals could function as electron capturers to further facilitate charge transfer. The isomorphic substitution of heteroatoms such as transition metals into the framework can be accomplished successfully by the replacement of the Ti and Si atoms in the skeleton of TS-1 molecular sieves. For example, when Fe [31], Ti [32], V [33], or Ni [34] are introduced into the silica framework isomorphically, the resulting zeolites demonstrate unique catalytic behavior originating from the catalytic activity of the transition elements.

Based on all the considerations above, in this work, we first synthesized 2D MFI-framework TS-1 molecular sieves with extra mesopores by a low-temperature hydrothermal method using a tailored diquaternary ammonium surfactant as the structure-directing agent. Subsequently, Ni²⁺ was introduced into the framework of the TS-1 nanosheets by a facile ion-exchange method. The optimized Ni/TS-1 NS provides impressive photoactivity, with a benzyl alcohol (BA) conversion of 84% and benzyl aldehyde (BAD) selectivity of 98% with O₂ as the oxidant under full light irradiation. The BAD yield is approximately six times that of bulk TS-1. The optimized nanosheets also show favorable stability. The mechanism of the photocatalytic aerobic oxidation of BA over Ni/TS-1 NS was carefully investigated using scavenger experiments. This work opens the way to a class of highly selective and robust TS-1-based photocatalysts for the aerobic oxidation and degradation of organics with industrial potential

All reagents were of analytical grade and used as-received without further purification. Deionized (DI) water was used throughout all reactions.

The modified procedure proposed by Choi et al.[28] was used to synthesize the special bifunctional organic surfactant C₂₂H₄₅-N⁺Br-(CH₃)₂-C₆H₁₂-N⁺Br-(CH₃)₂-C₆H₁₃ as follows: 34.5 g (0.200 mol) 1-bromodocosane and 286.3 g (2.000 mol) N, N, N9, N9-tetramethyl-1, 6-diaminohexane were liquefied in 1200 mL of an acetonitrile/toluene mixture (1:1 vol/vol) and aged at 70 ℃ for 10 h. The product was cooled to room temperature, filtered, washed with diethyl ether, and dried in a vacuum oven at 50 ℃ for 10 h. 93.66 g (0.200 mol) of the as-manufactured material and 41 g (0.300 mol) 1-bromohexane were dissolved in 500 mL acetonitrile under moderate agitation and refluxed for 10 h. After cooling at room temperature, the solid was filtered, washed with diethyl ether, and subsequently aged in a vacuum oven at 50 ℃. The as-manufactured quaternary ammonium salt C_22–6–6Br₂ was converted into its hydroxide form by reaction with Ag₂O in DI water. Specifically, a mixture of C_22–6–6Br₂, Ag₂O, and H₂O with an Ag₂O/C_22–6–6Br₂ molar ratio of 1:2:1 was vigorously stirred for 20 h at room temperature. The precipitates of the remaining Ag₂O powder and the formed AgBr were removed by filtration.

Bulk TS-1 was prepared by the hydrothermal method using tetrabutyl orthotitanate (TBOT) with tetraethylorthosilicate (TEOS) and tetrapropylammonium hydroxide (TPAOH, 20% aqua solution) as per reported procedures [27]. Typically, the mixture of TEOS and TBOT with desirable amount was placed in a polyethylene flask and vigorously stirred at ambient temperature for 25 min. The obtained clear solution was added dropwise to an alkali-free aqueous solution of TPAOH under vigorous stirring to avoid the formation of a TiO₂ phase. After TEOS and TBOT had been completely hydrolyzed under continuous stirring for 4 h at room temperature, the mixture was further aged at 360 K to evaporate the alcohols under stirring. The final resulting gels had molar ratios of 2SiO₂:(0.006–0.034)TiO₂:0.16TPAOH:16H₂O. The entire gel mixture was transferred to Teflon-lined stainless-steel autoclaves and then calcined at 445 K for 25 h under static conditions and autogenous pressure. To remove the residual structure-directing reagent or occluded organic species, the as-obtained white product was filtered, washed thoroughly with DI water through centrifugation, and then dried at 383 K for one night. Calcination was performed in air to burn off the occluded organic species at 823 K for 10 h.

The synthesized surfactants were used as zeolite SDAs in their bromide forms or after exchanging Br^– for OH^–. The ion exchange to OH^– was performed by reaction with Ag₂O in DI water. Specifically, C_22–6–6Br₂, Ag₂O, and H₂O were vigorously stirred for 20 h with an Ag₂O/C_22–6–6Br₂ molar ratio of 1.2 at room temperature. The precipitates of the remaining Ag₂O powder and the AgBr formed were removed by filtration. The typically distinctive procedures for TS-1 NS are summarized as follows [27]. The silica source, tetraethylorthosilicate (TEOS, 0.02 mol), was dissolved in 20.16 mL of a DI water aqueous mixture of the diquaternary organic SDA containing 0.0014 mol of C_22–6–6(OH)₂. The synthetic gel was stirred magnetically at room temperature until it became a homogeneous solution (for ca. 20 min). This mixture was cooled in an ice bath, and a freshly prepared solution of 0.0002 mol of titanium(IV) butoxide (TBOT, Aldrich) in 0.003 mol ofn-butyl alcohol was added dropwise. The resultant gel was further homogenized at 60 ℃ for 3 h under magnetic stirring. A solution of 0.0004 mol of titanium(IV) butoxide (TBOT, Aldrich) with 0.006 mol of n-butanol was then introduced slowly. The final molar ratio of the gel mixture was 4.0Si/14:0.03-0.17C_22–6–6(OH)₂: 0.0177TiO₂:101H₂O n-butyl alcohol. The final gel mixture was transferred to a stainless-steel Teflon-coated autoclave and hydrothermally crystallized at 160 ℃ for 15 days under tumbling conditions with the autoclave set at 60 rpm. After the hydrothermal treatment, the TS-1 nanosheet product was filtered, washed with DI water, and dried at 130 ℃. The as-synthesized materials were further calcined to burn off the diquaternary ammonium SDA species at 550 ℃ for 600 min under flowing air.

Various mass percentage ratios (1%, 2%, 5%, and 10%) (xNi/TS-1, x represents the mass percent ratio of Ni to TS-1) of Ni species were deposited on TS-1 NS by an ion-exchange method [35]. Two grams of TS-1 NS was dispersed in 50 mL DI water at a pH range of up to pH = 11, after which the desired quantity of nickel nitrate was added. The synthetic gel was vigorously stirred under moderate agitation for one and a half nights, then subjected to centrifuge filtration and aged at 355 K. Transition-metal (V, Fe, Co, and Cr) modified TS-1 NS were also prepared using the same method for comparison.

Powder X-ray diffraction (XRD) was carried out with a D8 Advance diffractometer (Bruker, Germany) using Cu K_α radiation (α = 0.15418 nm) at an accelerating voltage of 30 kV and emission current of 20 mA. The composition and elemental states of the samples were confirmed through X-ray photoelectron spectroscopy (XPS) using a Kratos-Axis Ultra DLD apparatus with an Al (mono) X-ray source. The binding energies of the samples were calibrated with respect to the signal of adventitious carbon with a binding energy equal to 284.55 eV. The morphology and structure of the samples were examined using electron micrographs taken on a JEOL JEM-2100 TEM operated at 200 kV, while scanning electron microscopy (SEM) was performed on a Hitachi S-4800 microscope. The UV-vis diffuse reflectance spectra (UV-Vis DRS) were measured with a Shimadzu UV-2550 spectrometer using BaSO₄ as a standard. O₂-temperature-programmed desorption (TPD) was performed using a chemisorption analyzer apparatus, the TP 5080 Chemisorb, which was equipped with a thermal conductivity detector (TCD). In a typical experiment, about 50 mg of the sample was preheated at 300 ℃ for 60 min to remove adsorbed gases, including H₂O molecules, and then cooled to 30 ℃ under a flow of ultra-pure He at a flow rate of 30 mL min^-1. Subsequently, highly pure O₂ gas at 30 ℃ was introduced at a flow rate of 40 mL min^–1 for 6 min. The excess weakly physically adsorbed O₂ was then removed by exposure to an ultra-pure He flow at a flow rate of 30 mL min^-1 at 30 ℃ for 60 min. The temperature was increased to 700 ℃ at a heating rate of 10 ℃ min^-1 under a He flow of 30 mL min^-1. Steady-state surface photovoltage spectroscopy (SS-SPS) measurements of the samples were carried out under a controlled atmosphere using a custom-built apparatus equipped with a lock-in amplifier (SR830) synchronized with a light chopper (SR540, USA). The powder sample was sandwiched between two indium-tin-oxide (ITO) glass electrodes, which were kept in a sealed container with a controlled atmosphere. Monochromatic radiation was produced by passing the light from 500 W xenon lamps (CHF XQ500W, Global Xenon lamp power) through a double prism monochromator (SBP300) to obtain monochromatic radiation. The Brunauer-Emmett-Teller (BET) surface areas and average pore sizes of the samples were obtained using a Micromeritics Tristar II 3020 system (Atlanta, GA, USA). The samples were degassed under a flow of N₂ at 180 ℃ for 3 h prior to measurements. For the evaluation of the BET surface area, a ST-2000 constant volume adsorption apparatus was utilized. The analysis of hydroxyl radicals was performed as follows. Each sample (0.05 g) was dispersed in 40 mL of a 0.001 mol L^–1 coumarin solution in a beaker. To achieve adsorption-desorption equilibrium, the mixture was magnetically stirred for 10 min before irradiation. Each sample was illuminated for one hour with UV-visible-light irradiation. After centrifugation, a certain amount of each sample was poured into a Pyrex glass cell for measurement of the fluorescence of 7-hydroxycoumarin at 390 nm excitation and 470 nm emission wavelengths using a spectrofluorometer (Perkin-Elmer LS55). UV Raman spectra were recorded on a custom-built UV Raman spectrometer using a Jobin-Yvon T64000 triple-stage spectrograph with a spectral resolution of 2 cm^-1. The 290 nm line from a Coherent Innova 300 Fred laser was used as an excitation source. The power of the 290 nm line at the samples was less than 1.0 mW. The Fourier transform infrared (FT-IR) spectra of the M/TS–1 samples were recorded using an FT-IR spectrometer (Nicolet Nexus 870) with a resolution of 4 cm^-1 and 64 scans in the region 4000–400 cm^-1.

PEC measurements were performed using a CHI660D electrochemical workstation (Chenhua Instrument, Shang–hai, China) in a conventional three-electrode configuration. The working electrode was prepared from the sample and had an active area of ca. 1 cm². A Pt foil and an Ag/AgCl electrode were used as the counter and reference electrode, respectively. An aqueous Na₂SO₄ solution with a concentration of 0.5 M was used as the lectrolyte. PEC I-V curves were measured at set potentials of –0.2 V to 1.8 V. The photocurrent densities at different excitation wavelengths were measured at an applied potential of 0.4 V. Monochromatic light was obtained by passing light from a 500 W xenon lamp through a monochromator (CM110, Spectral Products).

Standard photocatalytic reaction: The photocatalyst (100 mg) and the reactant benzyl alcohol (BA, 200 μmol) were added to toluene (5 mL) in a three-neck flask (25 mL), and the flask was sealed with rubber septum caps. The photocatalyst was dispersed well by ultrasonication for 5 min, and O₂ was bubbled through the solution for 20 min. The flask was equipped with a reflux condenser cooled with water and immersed in a temperature-controlled water bath (30 ± 2 ℃), and photoirradiated at λ> 400 nm using a 300 W Xe lamp; magnetic stirring was applied during the reaction. After the reaction, the same number of moles of dodecane with BA were added to the mixture as the external standard. The photocatalyst powder was separated from the liquid by filtration. The liquid mixture was analyzed using an Agilent gas chromatograph 6890 equipped with an HP-5 capillary column (30 m long and 0.32 mm in diameter, packed with silica-based supelcosil) and a flame ionization detector (FID). The injector temperature was 250 ℃, and the spilt was 0.1 μL. The column head pressure of the carrier gas (helium) during the analysis was maintained at 22.57 psi. The temperature program was as follows: 50 ℃ to 180 ℃; 20 ℃ min^–1; hold for 4 min. By-products were identified using gas chromatography-mass spectrometry (GC-MS, Agilent, GC 6890N, MS 5973 inert).

For the recyclability test, the photocatalyst was recovered by centrifugation and washed with 50 mL acetone and dried at 60 ℃ overnight. After drying, the same amount of the recovered photocatalyst was reused in a standard photocatalytic reaction as described above.

For the scavenger experiments with the photocatalysts, very small amounts of silver nitrate, benzoquinone, and triethanolamine were utilized in the standard reaction system for the photocatalytic aerobic oxidation of BA to investigate the active species during the BA transformation process.

The photocatalytic activities of as-prepared samples were measured in the degradation of 2, 4-DCP under UV-visible-light irradiation. In each experiment, 0.2 g of the photocatalyst was added to 100 mL of a solution of 2, 4-DCP (100 mg L^–1) in a beaker and then stirred in the dark for 30 min in order to attain adsorption-desorption equilibrium. Thereafter, the suspension was irradiated with a 150 W Xe lamp. During this irradiation period, a specific amount of the sample was collected at regular intervals and centrifuged to remove the photocatalyst; subsequently, the amount of pollutant was measured with a Shimadzu Model UV-2550 Spectrophotometer (Kyoto, Japan).

The as-designed diquaternary ammonium-type surfactant C_22–6–6 is composed of a long-chain alkyl group (C₂₂) and two quaternary ammonium groups spaced by a C₆ alkyl linker. The diammonium head group acted as an effective structure-directing agent for the MFI zeolite, while the hydrophobic interactions between the long-chain tails induced the formation of mesoscale micellar structure. With this surfactant, ultrathin TS-1 nanosheets were formed at the hydrophilic part of the micelles while the hydrophobic tails restricted the excessive growth of zeolites [28]. The crystallinity of the as-synthesized TS-1 NS was characterized by XRD and compared with that of bulk TS-1 (Fig. 1(a)). In the XRD patterns of the samples, the characteristic MFI diffraction peaks at 7.9°, 8.9°, 23.2°, and 24.1°, which are allocated to the [101], [202], [501], and [303] planes, respectively, were observed [27]. The locations of the peaks for TS-1 NS were basically consistent with those of the bulk TS-1. However, the TS-1 NS diffraction peaks were less intense and exhibited oriented growth of the crystals along the a-axis. This occurred because crystal growth along the b-axis was prevented by the hydrophobic tails of the designed surfactant, as reported by Choi et al.[28]. The broadened XRD peaks indicated that TS-1 NS was composed of nano-sized crystals. The UV-vis DRS were utilized to observe the optical absorption of the as-produced samples. Bulk TS-1 and TS-1 NS had similar absorption edges, as shown in (Fig. 1(b)).

Compared with that of bulk TS-1, the light absorption range of TS-1 NS was somewhat expanded, which was strongly associated with its ultrathin two-dimensional morphology. The band gap energy of TS-1 NS was calculated using the equation hν = A(hν – E_g)^n/2 where ν, E_g, and A are the light frequency, band gap energy, and a constant, respectively, and n is 4 for the indirect transition. As shown in the SEM image (Fig. 1(c)), TS-1 NS was observed to be a disordered assembly of thin platelets, whereas bulk TS-1 was obtained with a single-crystalline morphology (Fig. S1). The morphology of the TS-1 NS was similar to that previously reported for unilamellar MFI [27, 28]. The TEM image in (Fig. 1(d)) shows that the thickness of TS-1 NS was ca. 2 nm, which corresponds to the lattice parameter of the b-axis. Intersheet mesopores were observable between neighboring nanosheets. In contrast, no mesopores existed in the bulk TS-1; only small micropores with long diffusion pathways were present in the single crystals. Moreover, the BET area of TS-1 NS far surpassed that of bulk TS-1 (485.0 m² g^–1), reaching 1340.6 m² g^–1, which was attributed to its two-dimension morphology and the produced mesopores, which had an average pore size of ca. 3.2 nm.

To study the photocatalytic activities of the TS-1 based nanomaterials, the selective aerobic oxidation of BA under UV-vis light irradiation was selected as a benchmark reaction. As shown in (Fig. 2), bulk TS-1 showed some photoactivity in the target reaction, with its BA conversion and BAD selectivity reaching 19% and 72%, respectively. In contrast, the TS-1 NS exhibited much more favorable photoactivity. Its BA conversion increased to 44% and its BAD selectivity reached up to 90%. Ultrathin TS-1 NS with mesopores has a greatly enlarged surface area and showed a great advantage compared to microporous bulk TS-1 in terms of photoactivity. This was apparently because the mesopores and enlarged surface area of ultrathin TS-1 NS could expose more catalytic sites while shortening the mass transfer distance.

In addition to the factors above, the charge separation situation of a photocatalyst is a key factor affecting its photoactivity. The charge separations of TS-1-based materials were studied by measuring their SS-SPS responses. It is widely accepted that the signal intensity of SPS has a positive relationship with the charge separation [36, 37]. As depicted in (Fig. 3(a)), TS-1 NS exhibited a stronger SPS response than bulk TS-1, demonstrating its significantly improved charge separation; this is reasonable since ultrathin TS-1 NS had a greatly shortened charge transfer distance.

Moreover, the amount of the hydroxyl radicals formed was also used to determine the photogenerated charge separation in the photocatalytic system through the coumarin method [38]. Specifically, luminescent 7-hydroxy-coumarin is produced by the reaction of coumarin with hydroxyl radicals, and can be detected by fluorescence spectroscopy (FS). The intensity of the fluorescence spectra is directly proportional to the charge separation[39].

As shown in (Fig. 3(b)), the signal of TS-1 NS was much higher than that of the bulk TS, which was in good agreement with the SS-SPS results. Based on the above, the greatly improved charge separation of TS-1 NS contributes to the enhanced photoactivity for the aerobic oxidation of BA. Synthetically, the enlarged surface area, improved mass transfer, and improved charge transfer originating from the shortened charge transfer distance benefit the photocatalytic aerobic selective oxidation of BA.

Although TS-1 NS showed enhanced photoactivity compared to conventional bulk TS-1, there was some room to improve its charge separation. Moreover, effective catalytic sites are necessary for activating reactants to facilitate their conversion. For aerobic alcohol oxidation, the activation of O₂ is indispensable, as it is the speed-limiting step[29]. To address the complications above, the introduction of supplementary components is required. Noble metals such as Au and Pd have been widely reported to function as effective catalytic centers for photocatalytic alcohol oxidation in both traditional catalysis and photocatalysis [40]. However, due to the high cost and requirements for potential industrial applications, transition metals have become an alternative choice for this application. Therefore, in this work, we utilized Ni salts to introduce Ni species into the framework of the TS-1 NS molecular sieves by a facile ion exchange method [29].

NiNO₃ was used as the Ni source to introduce different amounts of Ni species to modify TS-1 NS (labeled as xNi/TS-1 NS, where x = 1, 2, 5, or 10). As shown in (Fig. 4(a)), for the samples 1Ni/TS-1 NS and 2Ni/TS-1 NS, only the characteristic peaks for TS-1 NS were observed in the XRD patterns, indicating that nickel oxide was not formed when limited amounts of Ni were introduced. It was inferred that the Ni was embedded into TS-1 NS to form amorphous Ni species by replacing some of the Si and Ti atoms. However, when the amount of Ni added was further increased, i.e., in 5Ni/TS-1 NS and 10Ni/TS-1 NS, diffraction peaks appeared at 2θ values of 43.2° and 43.2°; these were attributed to the formation of NiO (JCPDS Card No. 87–0712) [34]. In (Fig. 4(b)), it is obvious that the introduction of Ni species did not change the optical adsorption of TS-1 NS. A SEM image of 2Ni/TS-1 NS was acquired to observe the effect of introducing Ni species on its morphology (Fig. 4(c)). Noteworthily, 2Ni/TS-1 NS exhibited a very distinct morphology compared with that of TS-1 NS. After the ion exchange process, the intact nanosheet assemblies of TS-1 NS separated into thinner ones that loosely aggregated together.

In the TEM image in (Fig. 4(d)), it can be observed that the thickness of the nanosheets did not change compared to those of TS-1 NS, while the interspace between the nanosheets became larger due to the ion exchange process. However, according to the XRD results, the change in the morphology during ion exchange did not affect the crystallinity of TS-1 NS. As shown in Table S1, the BET surface area of the xNi/TS-1 NS samples (where x = 2, 5, or 10) decreased slightly compared with that of TS-1 NS. For 2Ni/TS-1 NS, the specific area decreases from 1340.6 to 1165.9 cm² g^–1. The major suppression of the BET surface area in Ni-modified TS-1 NS could be due to restricted structural collapse of the Si framework [31-34]. Moreover, when the Ni loading amount was increased (i.e., in 5Ni/TS-1 and 10Ni/TS-1), the specific area of TS-1 NS decreased more obviously due to the blocking effect of the NiO particles formed, as evidenced by XRD.

The chemical states and structural features of xNi/TS-1 NS were investigated by XPS, FTIR, and UV-Raman spectra using 2Ni/TS-1 NS as a representative and compared to those of TS-1 NS. To verify the chemical environments of the main elements of 2Ni/TS-1, the XPS spectra of TS-1 NS and 2Ni/TS-1 were collected. As shown in (Fig. 5(a)), the peak for Ni 2p_3/2 was located at 857.2 eV, and was assigned to Ni²⁺ with tetrahedral geometry. This binding energy was larger than the binding energy values of NiO, possibly due to the Ni species accepting electrons from TS-1 NS. Additionally, the Ti 2p_3/2 spectra for TS-1 NS and 2Ni/TS-1 NS were deconvoluted to determine the contribution of Ti⁴⁺ species in the TS-1 NS framework in (Fig. 5(b)). In conventional bulk TS-1, the binding energy for Ti 2p_3/2 has been reported to be approximately 460.1 eV [41, 42]. For TS-1 NS, the binding energy of the main Ti 2p_3/2 peak was 458.8 eV, indicating that the electronic cloud density of the periphery of the Ti 2p core level decreased. This might have been due to the Ti atoms adopting a tetrahedral geometry in the framework of TS-1 NS, which was different than the situation for bulk TS-1. After introducing the Ni species, the binding energy of the main Ti 2p_3/2 peak further decreased to 458.4 eV. This clearly indicated that the introduction of Ni changed the electronic environment of the Ti atoms. More specifically, the electronic cloud of Ti atoms was drawn by Ni species. The FTIR spectra of TS-1 NS and 2Ni/TS-1 NS are depicted in (Fig. 5(c)). The intense band at 554 cm^-1 was assigned to the vibration of the double five-membered ring units and was considered to be a fingerprint of the MFI structure [34]. The sample TS-1 NS exhibited a characteristic band at 980 cm^-1, which was attributed to the collective vibrations of Si–O–Ti bonds or Si–O bonds disturbed by the existence of titanium atoms in the TS-1 molecular sieve framework [36]. For 2Ni/TS-1 NS, this band was shifted to 967 cm^-1, indicating that the introduction of Ni weakened the Si–O–Ti bonds or Si–O bonds. This might have been due to the charge transfer from tetrahedrally coordinated Ti–O. The introduction of Ni species changed the electronic charge of the Ti atoms, which further confirmed that the strength of the Ti–O bond decreased and the electron cloud density surrounding Ti decreased, which was in good agreement with the XPS results. Additionally, UV-Raman spectroscopy is a helpful technique to identify transition metals in the framework of zeolites and mesoporous silica and has been applied specifically for the structural analysis of TS-1 molecular sieves, as reported by Li et al. [18]. The low fluorescence interference and strong resonance Raman enhancement make the recognition of the framework Ti in TS-1 NS it possible at low concentrations under UV excitation [43]. The UV-Raman spectra of TS-1 NS and 2Ni/TS-1 NS obtained using a 290 nm laser line as the excitation source are depicted in (Fig. 5(d)).

For TS-1 NS, the band at 296 cm^-1 was assigned to the vs (Si–O–Si) modes of the five-membered rings of the MFI structure, and is also observed in the UV-Raman spectrum of silicate [18, 43]. The band at 1125 cm^-1 was attributed to tetrahedrally coordinated framework titanium species in a local [Ti(OSi)₄] unit. More specifically, this band was attributed to the asymmetric stretching vibration of Ti–O–Si [44-48]. The Raman band at 681 cm^-1 was attributed to "TiO₆" species, indicating the existence of a small fraction of titanium with a hexahedral structure [49]. The small band at 923 cm^–1 originated from silicate species with no correlation with titanium [50]. After the introduction of Ni, the bands assigned to TS-1 NS were preserved; in particular, the band at 1122 cm^-1 that was identified as a fingerprint for TS-1 NS showed a shift of 3 cm^-1. This indicated that after the introduction of Ni, the tetrahedral coordination of the Ti centers became less rigid compared with that of TS-1 NS. However, a small new band at 901 cm^-1 appeared, which was attributed to NiO, verifying the existence of Ni species in the skeleton of TS-1 NS.

The photoactivities of the as-synthesized Ni-modified TS-1 NS samples for the selective oxidation of BA were examined using O₂ as the oxidant under full light irradiation. In (Fig. 6(a)), it can be observed that among the different amounts of Ni tested, the best photocatalytic performance was obtained for 2Ni/TS-1 NS, which exhibited the highest BA conversion of ca. 78.9% and BAD selectivity of 98.8%. Apparently, the introduction of Ni species greatly enhanced the photoactivity relative to that of pristine TS-1 NS. Nevertheless, the photoactivity was suppressed when the Ni loading was further increased. The effect of different Ni sources was also studied by using Ni(CH₃COO)₂·4H₂O and NiCl₂·6H₂O (labelled as S–2 and S–3). The XRD spectra of the samples prepared using the same amount of these Ni precursors as was used in 2Ni/TS-1 NS were similar, as shown in (Fig. S3(a)). Among the three Ni salts, NiNO₃ (labelled as S–1) was found to be the optimum choice, affording the best photocatalytic performance (Fig. S4). Therefore, it was proven that Ni-modified TS–1 NS nanocomposites are effective photocatalysts for the aerobic selective oxidation of alcohols.

Another vital index to evaluate a heterogeneous photocatalyst is its stability. As shown in (Fig. 6(b)), both the BA conversion and the BAD selectivity of 2Ni/TS-1 NS were maintained for five runs with negligible loss of photoactivity.

According to the UV-vis spectra analysis in (Fig. 4(b)) and (Fig. S2(b)), the introduction of Ni species did not extend the light absorption. Next, the charge separation situation of 2Ni/TS-1 NS was examined using SS-SPS and FS measurements, which were compared with those of TS-1 NS. As shown in (Fig. 7(a)), after introducing the Ni species, the intensity of the SPS signal was greatly improved. Among all the Ni-modified samples, 2Ni/TS-1 NS showed the most intense SPS signal, which was consistent with the photoactivity results. This demonstrated that that the improved charge separation achieved by the introduction of Ni contributed to the enhanced photoactivity of 2Ni/TS-1 NS, as the Ni species could function as an electron capturer [37, 38]. Moreover, among the samples prepared using the same loadings of the three distinct Ni sources, the SPS signal intensities of 2Ni (S–2)/TS-1 NS and 2Ni (S–3)/TS-1 NS were weaker than that of 2Ni/TS-1 NS (Fig. S4a). Correspondingly, the photoactivities of the three samples showed a positive relationship with the intensities of the SPS signal. Additionally, FS of the produced hydroxyl radicals were also collected for the different Ni-modified samples (Fig. 7(b)) and (Fig. S4(b)). The intensities of the FS followed the same trend as the SS-SPS responses, with 2Ni/TS-1 NS showing the greatest FS intensity. Therefore, based on the above, the excellent photoactivity of 2Ni/TS-1 NS was strongly associated with the role of Ni species in capturing electrons.

Moreover, the excellent photoactivity of 2Ni/TS-1 could be extended to photocatalytic pollutant degradation. As shown in (Fig. S6(a)), using 2, 4-DCP as a model pollutant, 2Ni/TS-1 showed superior photoactivity compared with TS-1 NS and bulk TS-1, similarly to in selective BA oxidation. After 4 h of photocatalytic reaction, the degradation conversions using bulk TS-1, TS-1 NS, and 2Ni/TS-1 were 30%, 51%, and 75%, respectively (Fig. S6(b)).

Surface catalytic processes are another key factor affecting overall photoactivity. In the initial design of the photocatalyst, the introduction of Ni species was expected to afford catalytic activity benefiting the photocatalytic aerobic oxidation of alcohols. To verify and investigate the catalytic activity of the Ni species, other transition metals were also studied for comparison. For the 2M/TS-1 NS samples (M = Ni, V, Fe, Co), modification with all the transition metals enhanced the photoactivity of TS-1 NS to distinct extents, with 2Ni/TS-1 NS showing the best photocatalytic performance (Fig. 8). The XRD patterns of the 2M/TS-1 NS samples (M = Ni, V, Fe, Co) are depicted in (Fig. S5(a)). Using this same loading amount for all the transition metals, the crystalline structure of TS-1 NS was not interrupted after modification. The light absorption of the 2M/TS-1 NS samples was characterized by UV-visible spectroscopy (Fig. S5(b)). The introduction of Ni and V species increased the light absorption in the UV range, while the introduction of Fe and Co only slightly increased the light absorption in the visible-light range.

Moreover, the charge separation situation was also studied for the 2M/TS-1 NS samples using SS–SPS and FS (Fig. 9(a)) and (Fig. 9(b)). As shown in the SPS spectra in (Fig. 9(a)), among all the samples, 2Ni/TS-1 NS showed the strongest signal intensity, indicating that its charge separation situation was the most favorable.

Noteworthily, the SPS signal intensities of the 2M/TS-1 NS samples were found to exhibit a positive relationship with their photoactivities. Moreover, their FS spectra were measured to further support the enhanced charge separation of the as-prepared samples. Interestingly, 2Ni/TS NS showed the greatest amount of hydroxyl radicals among all the 2M/TS-1 NS samples as determined using FS (Fig. 9(b)). Based on the results above, the introduction of all the transition metals improved the light absorption and charge separation to different extents. On this basis, the catalytic activities of the metals, especially that of Ni, were studied further. In the aerobic oxidation of alcohol, oxygen activation is very important for effective oxidative conversion. The reduction of O₂ utilizes the photogenerated electrons, which improves the charge separation. Furthermore, the oxygen species produced by the O₂ reduction would lead the aerobic BA oxidation process. Therefore, additional catalytic sites that can activate O₂ are necessary.

The electrochemical reduction curves for the 2M/TS-1 NS samples were measured in an O₂-bubbled system, and are shown in (Fig. 10(a)). It can clearly be observed that the reduction of O₂ was much more favorable on 2Ni/TS-1 NS than on the 2M/TS-1 NS samples. These results indicate that the introduction of Ni species could better facilitate the reduction of O₂ compared to Fe, Co, and V; thus, their catalytic abilities followed the predicted order. It was inferred that Ni species could effectively activate the O₂ molecules. This view was further proved by means of the O₂ temperature-programmed desorption curves (Fig. 10(b)). Normally, desorption peaks at temperatures below 300 ℃ are attributed to physical O₂ adsorption, while those observed above 300 ℃ are attributed to chemical adsorption. The introduction of the transition metals clearly promoted both the physical and chemical adsorption of O₂ molecules significantly. Among all the samples, 2Ni/TS-1 NS adsorbed the most oxygen; this result was in good agreement with the electrochemical reduction measurements. The above information clearly demonstrates that the as-introduced Ni species have superior ability to improve charge separation as an electron capturer and to activate O₂ as catalytic sites compared with the other transition metals, and as a result, it exhibited the best photoactivity towards the photocatalytic aerobic oxidation of BA.

To further investigate the reaction mechanism of the 2Ni/TS-1 NS-catalyzed photocatalytic aerobic oxidation of BA, scavenger experiments were carried out. Specifically, additional scavengers, namely AgNO₃ for the photogenerated electrons, triethanolamine (TEA) for the photogenerated holes, and benzoquinone (BQ) for ∙O^2- radicals, were introduced to the reaction system for the generalized photocatalytic reaction using 2Ni/TS-1 NS as the photocatalyst [38].

As shown in Fig. 11, the addition of AgNO₃, BQ, and TEA all suppressed the reaction to distinct extents, indicating that the 2Ni/TS-1 NS-catalyzed photocatalytic oxidation of BA was a complex process involving several reactive species. The introduction of AgNO₃ decreased the photoactivity compared to that of the blank experiment. This proved that the photogenerated electrons would induce further reactions. Specifically, the addition of BQ significantly retarded the reaction, indicating that ∙O^2- radicals played key roles during the oxidation process.

Moreover, the addition of TEA had a moderate slowing effect on the reaction, indicating that photogenerated holes also participated in the oxidation. ∙O^2- radicals and their derivative ∙OOH radicals have been reported to dominate in the oxidation of alcohols [51]. Therefore, based on the above, it was inferred that under full light irradiation, the photogenerated electrons of 2Ni/TS-1 NS would reduce O₂ to produce ∙O^2- radicals, which would then produce ∙OOH radicals. The ∙OOH radicals would further react with alcohol cations originating from dehydrated BA by the holes to finally produce BAD. The hypothetical mechanism for the photocatalytic oxidation of BA on 2Ni/TS-1 NS is illustrated in Scheme 1.

In summary, we have successfully fabricated ultrathin TS-1 molecular sieve nanosheets (ca. 2 nm) by a low-temperature hydrothermal method using the tailored diquaternary ammonium-type surfactant C_22–6–6 as the structure directing agent. The as-synthesized TS-1 nanosheets had a greatly enlarged surface area with mesopores and improved charge separation compared with traditional TS-1 molecular sieves. For these reasons, they were effective and robust photocatalysts for the aerobic oxidation of alcohols. The functionalization of TS-1 NS with Ni cations by ion-exchange further enhanced the photoactivity of the catalyst in the aerobic oxidation of BA as a typical reaction. A high selectivity towards BAD (98.9%) was achieved, which represented an approximately six-fold increase in the BAD yield compared to that of bulk TS-1; this activity was maintained for five runs. The excellent photoactivity of Ni/TS-1 NS was attributed to the significantly enlarged surface area of its two-dimensional morphology with extra mesopores and greatly improved charge separation. The as-introduced Ni species could capture photoelectrons to further improve the charge separation, and also functioned as catalytic sites for the activation of O₂. An overall photocatalytic reaction mechanism was proposed based on the scavenger experiments. Specifically, the ∙O^2– radicals formed by the reduction of O₂ subsequently form ∙OOH radicals to further oxidize the dehydrated BA by the photogenerated holes and then produce BAD selectively. This work has developed titanosilicate molecular sieve-based nanomaterials as low-cost, highly selective, and robust photocatalysts for aerobic oxidation and the degradation of organic compounds, with good prospects of being extended to the industrial applications.

Acknowledgments

We are grateful to the Key Laboratory of Functional Inorganic Material Chemistry (Heilongjiang University), Ministry of Education.