Effect of SiO2 amount on heterogeneous base catalysis of SiO2@Mg–Al layered double hydroxide

The effects of SiO2 amount on the base catalysis of highly active finely crystallized Mg–Al type layered double hydroxides prepared by the co-precipitation method with coexistence of SiO2 spheres, denoted as SiO2@LDHs, were investigated. With the Si/(Mg + Al) atomic ratios of 0–0.50, the highest activity for the Knoevenagel condensation was observed in the case of Si/(Mg + Al) = 0.17, as the reaction rate of 171.1 mmol g(cat)−1 h−1. The base activity increased concomitantly with decreasing LDH crystallite size up to Si/(Mg + Al) atomic ratio of 0.17. However, above the Si/(Mg + Al) atomic ratio of 0.17, the reaction rate and TOFbase were decreased although the total base amount was increased. Results of TEM-EDS and 29Si CP-MAS NMR suggest that the co-existing SiO2 causes advantages for dispersion and reduction of the LDH crystallite to improve the base catalysis of SiO2@Mg–Al LDH, whereas the excess SiO2 species unfortunately poisons the highly active sites on the finely crystallized LDH crystals above a Si/(Mg + Al) atomic ratio of 0.17. According to these results, we inferred that the amount of spherical SiO2 seeds in the co-precipitation method is an important factor to increase the base catalysis of SiO2@LDHs; i.e. the control of Si/(Mg + Al) atomic ratio is necessary to avoid the poisoning of highly active base sites on the LDH crystal.


Introduction
Heterogeneous solid catalysts have been widely used in various chemical reactions in the chemical industry, exhaust gas puri-cation, and environmentally friendly reactions. They are more suitable for reactions at high temperatures than homogeneous catalysts, and are removed easily from the reactor. In particular, solid acid catalysts are used in many important processes related to petroleum rening and petrochemical production. Consequently, numerous studies have specically examined solid acid catalysts. On the other hand, fewer efforts have been devoted to heterogeneous base catalysts. 1 Because general base sites on a solid base catalyst are readily poisoned by moisture and carbon dioxide in the atmosphere, 2 the development of a promotion strategy for solid base catalysts is more difficult than for solid acid catalysts.
Layered double hydroxide, generally called LDH, is a wellknown layered clay mineral that is known to act as a unique solid base catalyst. Actually, LDH comprises brucite-like positively charged two-dimensional sheets denoted as [M 1Àx 2+ M x 3+ (OH) 2 ] x+ and interlayer parts denoted as A x/n nÀ $mH 2 O, where A nÀ corresponds to interlayer anions such as carbonate and hydroxide. The positively charged sheets and the interlayers are alternately laminated to compensate for the charge between sheets. [3][4][5] Base sites on LDH are mainly regarded as identical Brønsted basic OH À and HCO 3 À anions, which are adsorbed onto the LDH surface. These unique basic sites can act even in an air atmosphere and can exhibit basic characteristics without pretreatment. 4 As a solid base catalyst, LDH is well-known to catalyze various organic transformations such as aldol condensation, [6][7][8][9] Knoevenagel condensation, 10-13 epoxidation, [14][15][16] and transesterication. [17][18][19] Recent studies have indicated that LDHs can promote advanced environmentally friendly reactions such as biomass-derived saccharide conversion 13,20-25 and photocatalytic conversion of CO 2 in an aqueous solution. [26][27][28] Therefore, the development of highly active LDH catalysts is eagerly sought. The main base sites on the LDH surface are generally regarded as adsorbed anions located at the corner and edge of a crystal. 29 However, the anions in the interlayer space cannot participate in the chemical reactions because of the high charge density of the LDH layers and the high contents of anionic species and water molecules, resulting in strong interlayer electrostatic interactions between the sheets. 4,30 Therefore, the delamination of LDH nanosheets [30][31][32][33][34][35][36] and the ne crystallization of LDH on appropriate carriers [37][38][39][40][41][42] have been conducted to increase the number of exposed active base sites. As-prepared LDH materials also have been evaluated carefully to assess their characteristics and utility as photocatalysts and electrocatalysts, 43,44 high active base catalysts for Knoevenagel condensation 45 and epoxidation, 46 magnetic separation of proteins, 37 pseudocapacitance, 38 ame retardancy of epoxy resins, 40 and as adjuvants. 41 Nevertheless, no report describes a study of improvement of base catalysis of LDH itself by ne crystallization followed by in situ growth method of SiO 2 @LDH nanoparticles.
An earlier study revealed a co-precipitation method for preparation of small-crystallized LDH catalysts on SiO 2 nanospheres and explored the superior base catalyses for the Knoevenagel condensation of benzaldehyde compared with conventional LDHs. 47 This method is applicable for the preparation of SiO 2 @LDH nanoparticles with various compositions and element combinations: i.e. SiO 2 @M 2+ -M 3+ LDH (M 2+ : Mg 2+ or Ni 2+ , M 3+ : Al 3+ or Ga 3+ , and M 2+ /M 3+ : 1 or 3). Various characterizations of SiO 2 @LDH nanoparticles using XRD, TEM-EDS, and 29 Si CP-MAS NMR techniques revealed that the co-existence of small SiO 2 sphere (ca. 40 nm diameter) surface generated the starting points of LDH growth via Si-O-M covalent bond formation, leading to the formation of ne-crystallized LDH and enhancement of base catalysis for the Knoevenagel condensation of benzaldehyde with ethyl cyanoacetate. However, the roles of Si-O-M covalent bonds in the necrystallization of LDH and base catalysis have not been explored well. Therefore, in this paper, we investigated the base properties and structural parameters of as-prepared SiO 2 @Mg-Al LDH materials with Mg 2+ /Al 3+ ¼ 3, and discussed the base catalysis with different Si to (Mg + Al) ratios to reveal the mechanism of our strategy.
Spherical SiO 2 (40 nm) was prepared according to descriptions in earlier reports. 41,46 First, 96 mmol of TEA and 2 mL of TEOS were combined in a 200 mL eggplant ask. The two-phase mixture was heated in an oil bath at 363 K for 20 min without stirring. When the mixture was removed from the oil bath, 26.0 mL of an aqueous solution (2.8 wt%) of CTAB pre-heated at 333 K was added immediately as a structure-directing agent in a condensation process. Then, it was stirred continuously for 24 h at room temperature. Thereaer, the resulting mixture was added to 50 mL of ethanol to obtain colloidal aqueous suspension. The obtained precipitate was centrifuged for 5 min at 4000 rpm. Aer decantation, the sediment was re-dispersed through vigorous stirring in 50 mL of an ethanolic solution of ammonium nitrate (20 g L À1 ), and then reuxed for 1 h. This procedure was repeated three times. The same operation was performed with a solution of concentrated hydrochloric acid in ethanol (5 g L À1 ) to replace the ammonium ions. The nal sediment was washed with ethanol, and then dried in vacuo. The obtained spherical SiO 2 powder was calcined at 823 K under 1 L min À1 of air ow for 6 h.
The SiO 2 @(Z)LDH catalysts (Z: desired Si/(Mg + Al) atomic ratio) were prepared via an in situ co-precipitation method according to a previous report. 42 Spherical SiO 2 (40 nm) was dispersed in 20 mL of water using ultrasound treatment. Aer 30 min, 0.96 mmol of Na 2 CO 3 was added to the solution. Then, aer an additional 5 min of sonication was conducted, 19.2 mL of metal nitrate aqueous solution ([Mg] + [Al] ¼ 0.075 M) was slowly dropped into the spherical SiO 2 dispersed solution, followed by stirring at room temperature. The pH was maintained at 10.0 by an aqueous NaOH solution (1 M) during titration. The obtained suspension was stirred for an additional 1 h. Aer the resulting paste was ltered, it was washed with 1 L of water and ethanol. Then, it was dried at 383 K overnight. The Si/(Mg 2+ + Al 3+ ) atomic ratios were varied from 0 to 0.50 whereas the Mg/Al atomic ratio was adjusted to 3.

Reaction
Knoevenagel condensation of benzaldehyde with ethyl cyanoacetate was proceeded in a 20 mL Schlenk tube under an N 2 ow (30 mL min À1 ). The reaction was performed using 1.0 mmol of benzaldehyde, 1.2 mmol of ethyl cyanoacetate, 10 mg of catalysts and 3 mL of toluene at 313 K. The obtained products were analyzed using a GC-FID (GC-2014, Shimadzu Corp.) equipped with a polar column (DB-FFAP, Agilent Technologies Inc.).

Characterizations
X-ray diffraction patterns (XRD) were collected using a Smar-tLab (Rigaku Corp.) with a Cu Ka X-ray source (40 kV, 30 mA). The LDH (003) and (110) crystallite sizes were calculated using the Scherrer equation: D hkl ¼ Kl/(b cos q) (K: Scherrer number (0.9), l: incident ray wavelength (0.1542 nm), b: peak width at half height (rad), q; Bragg angle). 29 Si cross polarization magic angle spinning nuclear magnetic resonance ( 29 Si CP-MAS NMR) measurements were obtained by an Avance III 500 (Bruker Analytik GmbH) in a 4 mm ZrO 2 rotor. The spinning rate was 8 kHz. The 29 Si chemical shis are referenced to hexamethylcyclotrisiloxane (taken to be at d ¼ À9.6875 ppm). Transmission electron microscopeenergy dispersive X-ray spectroscopy (TEM-EDS) elemental mapping analytical techniques were done with a JEM-ARM200F (JEOL) at 200 kV. Inductively coupled plasmaatomic emission spectrometry (ICP-AES) was operated by an iCAP 6300 Duo (Thermo Fisher Scientic Inc.) to estimate the actual amount of precipitated M(OH) x and SiO 2 in asprepared SiO 2 @Mg-Al LDHs with various Si/(Mg + Al) atomic ratio.

Results and discussion
Optimization of Si/(Mg + Al) ratio in SiO 2 @LDH We prepared SiO 2 @LDHs with various loading amounts of SiO 2 to ascertain the optimized SiO 2 @LDH structure for high catalytic reactivity. The sphere morphology and the diameter of SiO 2 were conrmed from SEM and TEM observations, as presented in an earlier report. 47 The correlation between Si/(Mg + Al) atomic ratio, base catalysis, and structural properties of SiO 2 @LDHs were investigated in the range of 0-0.50 on Si/(Mg +   Table 1.  Table S1 (see ESI †). Among various Si/(Mg + Al) atomic ratio from 0 to 0.50, actually, the 0.17 was found to be the best catalyst with a reaction rate of 171.1 mmol g(cat) À1 h À1 . This reaction rate is 2.2 times higher than conventional LDH prepared with the same co-precipitation method without SiO 2 seeds (Si/(Mg + Al) ¼ 0).
The XRD patterns and crystal properties of SiO 2 @(Z)LDHs are portrayed respectively in Fig. 2 and Table 2. All prepared catalysts showed an LDH-originated diffraction pattern. The intensity of LDH originated peaks decreased in accordance with Si loading amount, whereas that of amorphous SiO 2 increased slightly. Lattice parameters a and c, respectively calculated from LDH (003) and (110) diffraction peaks, are almost identical among Si/(Mg + Al) atomic ratios of 0-0.50. This result indicates clearly that these SiO 2 @(Z)LDHs have the same LDH crystal unit. However, the crystallite size of LDH is unquestionably affected by Si/(Mg + Al) atomic ratio. The crystallite size of D(003) is reduced in accordance with the Si loading amount. The crystallite size of D(110) is almost identical in the region among Si/(Mg + Al) ratio of 0-0.13, but it is reduced from ca. 13 nm to 8 nm when a Si/(Mg + Al) ratio increased. In our earlier research, it was revealed that the co-precipitation method with co-existence of spherical SiO 2 caused dispersion of starting points of LDH crystal growth on the SiO 2 surface through the Si-O-Al and Si-O-Mg covalent bonds to lead generation of necrystallized LDH nanocrystal. 47 Fig. 3(A) shows that the spherical SiO 2 (40 nm) showed three peaks at À91, À100 and À109 ppm, which respectively correspond to Q 2 , Q 3 , and Q 4 species 48-50 where Q n designated the Si-centered tetrahedral structural species; Q refers to silicon atom and n denotes the number of bridging oxygens. Furthermore, SiO 2 @(Z)LDHs showed broad resonance between À70 to À115 ppm, which include some peaks attributed to Q 0 and/or Q 1 (À60 to À83 ppm) 49 and a Si-centered tetrahedral structure that possesses Si-O-Al and Si-O-Mg bonds (À73 to À105 ppm). 42 These results suggest that, in the case of lower Si/(Mg + Al) atomic ratio (<0.13), the LDH crystal is immobilized onto the SiO 2 surface through the Si-O-Al and Si-O-Mg bonds to inhibit ab-face stacking without reducing the plane crystallite size. This result seems to be attributable to the lower number of starting points of LDH crystal growth: the amount of metal constituting one crystal did not change compared with conventional LDH prepared without SiO 2 seeds (Si/(Mg + Al) ¼ 0). Actually, the proportion for the 29 Si CP-MAS NMR peaks attributed to Si-O-Mg and Si-O-Al bonds on the SiO 2 @(0.13)LDH is only #40%, whereas that of SiO 2 @(0.17)LDH is #61%, as shown in Fig. 3 and Table 3. Therefore, we infer that the number of Si-O-Mg and Si-O-Al bonds on the SiO 2 surface deeply affected  a The crystallite sizes of LDHs were calculated by Scherrer equation: D hkl ¼ Kl/(b cos q) (K: Scherrer number (0.9), l: incident ray wavelength (0.1542 nm), b: peak width at half height (rad)).
reduction of the crystallite size not only in the stacking direction but also in the plane direction when below Si/(Mg + Al) < 0.17-0.25. Correlation between the base amount and catalytic activity for Knoevenagel condensation over SiO 2 @(Z)LDHs is presented in Table 4. Although the base amount of SiO 2 @(0.13)LDH was lower than that of conventional LDH, the reaction rate and the apparent TOF per base site (TOF base ) for SiO 2 @(0.13)LDH are higher than those of LDH. The D(003) of SiO 2 @(0.13)LDH was smaller than LDH, whereas D(110) of SiO 2 @(0.13)LDH and LDH are almost identical, as shown in Table 2. Therefore, these indicated that the immobilization of LDH crystal onto SiO 2 with inhibition of the ab-face stacking led to increase in the number of highly active base sites located on the surface LDH layer. Above the Si/(Mg + Al) atomic ratio of 0.13, a base amount increased in accordance with Si/(Mg + Al) ratio from 0.32 to 0.49 mmol g(cat) À1 . Furthermore, the activity was maximized at Si/(Mg + Al) ratio of 0.17 with the reaction rate of 171.1 mmol g(cat) À1 h À1 . It is particularly interesting that the reaction rate per obtained LDH phase and TOF base were also maximized at Si/ (Mg + Al) ratio of 0.17 with the reaction rate of 193.6 mmol g(LDH) À1 h À1 and TOF base of 450 h À1 . Above the Si/(Mg + Al) ratio of 0.17, both the reaction rate per LDH phase and TOF base were decreased respectively to 158.4 mmol g(LDH) À1 h À1 and 238 h À1 at Si/(Mg + Al) ratio of 0.50. These results strongly suggest that the Si/(Mg + Al) atomic ratio affects not only the LDH crystallite size and base amount but also the type of base sites and these fractions. 51 The LDH crystallite size of SiO 2 @(0.50)LDH is at least smaller than that of SiO 2 @(0.17)LDH. Therefore, the base catalysis of SiO 2 @(0.50)LDH is expected to be better than that of SiO 2 @(0.17)LDH if the base catalysis is only inuenced by the crystallite size. 29 Si CP-MAS NMR spectra showed that the proportion of terminal Si-OH species assigned as Q 0 and/or Q 1 decreased in accordance with Si/(Mg + Al) atomic ratio, as shown in Table 3, indicating rst that a surface Si-O-Si bond is cleaved to generate terminal Si-OH species and then that these act as cross-link point with Mg and Al ions. Consequently, when there are the excess free terminal Si-OH species in the solution aer the generation of SiO 2 @LDH, these excess Si-OH species cover the LDH crystal to produce Si-O-Mg and Si-O-Al covalent bonds. Although the base amount is increased even the region from SiO 2 @(0.13)LDH to SiO 2 @(0.50)LDH, this phenomenon might take place only with difficulty on the inferior base sites located at a at plane of LDH. However, the decrease of TOF base strongly suggests that the high active base sites are poisoned by Si species in the case of a higher Si/(Mg + Al) atomic ratio.
Dark-eld TEM images and results of EDS elemental mapping of SiO 2 @(Z)LDHs are presented in Fig. 4. In the case of lower Si/(Mg + Al) atomic ratio such as 0.13 and 0.17, the LDH crystal is generated with covering the SiO 2 phase to form a SiO 2 core -LDH shell-like structure, as shown in Fig. 4(A)-(J). Furthermore, results show that the boundary between SiO 2 phase and LDH phase becomes ambiguous in accordance with the increase of Si/(Mg + Al) atomic ratio ( Fig. 4(K)-(T)). These indicate that rst the LDH crystal grows up from the SiO 2 surface to generate the immobilized SiO 2 core -LDH shell structure. If there are excess dissolved SiO 2 species possessing  Accordingly, we infer the correlation between Si/(Mg + Al) atomic ratio and base catalysis of prepared SiO 2 @(Z)LDHs as follows: (i) below Si/(Mg + Al) atomic ratio of 0.13, the LDH crystal is just immobilized onto the SiO 2 surface with inhibition of the ab-face stacking. The exposed corner and edge located at the surface layer act as highly active base sites. (ii) Fine crystallization occurs not only in the stacking direction, but also in the plane direction to increase the amount of base sites, especially base sites with high activity up to Si/(Mg + Al) atomic ratio of 0.17. (iii) Above a Si/(Mg + Al) atomic ratio of 0.17, excess terminal Si-OH species covered the highly active base sites to produce Si-O-Mg and Si-O-Al covalent bonds, thereby lowering the activity.

Conclusions
The effects of SiO 2 loading amounts on the crystallite sizes, basicity, and catalytic activity of SiO 2 @LDH catalysts were investigated. SiO 2 @LDHs were prepared using co-precipitation with the coexistence of various amounts of spherical SiO 2 with particle sizes of ca. 40 nm. The XRD results suggest that the LDH crystallite size of D(003) is simply reduced in accordance with the Si loading amount. Furthermore, the crystallite size of D(110) is almost identical below Si/(Mg + Al) of 0.13, although it is reduced above 0.13. Base catalysis of SiO 2 @LDHs was evaluated using Knoevenagel condensation of benzaldehyde and ethyl cyanoacetate. Both the reaction rate and the apparent TOF per base site were increased in the region between Si/(Mg + Al) atomic ratio of 0-0.17, whereas the base amount is increased linearly from the Si/(Mg + Al) atomic ratio of 0.13 to 0.50 with reduction of the LDH crystallite size. The results of 29 Si CP-MAS NMR and STEM-EDS suggest that a surface Si-O-Si bond is cleaved, generating terminal Si-OH species that act as a crosslink point with Mg and Al ions to form an immobilized SiO 2 core -LDH shell structure. However, when the amount of Si becomes excessive with respect to Mg and Al ions, the excess Si-OH group forms Si-O-Mg and Si-O-Al covalent bonds with LDH crystal to cover the LDH shell. From these results, we inferred the effect of SiO 2 amount on heterogeneous base catalysis of SiO 2 @Mg-Al LDH as follows: (i) below Si/(Mg + Al) atomic ratio of 0.13, the highly active base sites located at the corner and edge of surface layer are exposed by immobilization of LDH crystals onto the SiO 2 surface, (ii) up to Si/(Mg + Al) atomic ratio of 0.17, the number of exposed highly active base sites is increased in accordance with reduction of LDH crystallite, and (iii) above Si/(Mg + Al) atomic ratio of 0.17. Excess terminal Si-OH species covered the highly active base sites through the Si-O-Mg and Si-O-Al covalent bonds to decrease the activity. This study elucidated the correlation between SiO 2 amount, crystal properties, and the basicity of ne-crystallized SiO 2 @LDH catalysts to present a new technique to improve the base catalysis of the widely used LDH material. Aer optimization of the Si/(Mg + Al) atomic ratio in SiO 2 @LDHs, the Si/ (Mg + Al) atomic ratio of 0.17 was found to be the best catalyst, with the reaction rate of 171.1 mmol g(cat) À1 h À1 , which is a 2.2 times higher value than that of the conventional LDH prepared with same protocol in absence of SiO 2 seed agents.

Conflicts of interest
There are no conicts to declare.