Synthesis of organized mesoporous metal oxide films templated by amphiphilic PVA–PMMA comb copolymer

Abstract

The morphology and structure of metal oxide films play pivotal roles in determining their properties in various applications, e.g., catalysis and energy conversion. We present a general synthesis of organized mesoporous metal oxide films with high porosity and good interconnectivity via a sol–gel process using an amphiphilic poly(vinyl alcohol)–poly(methyl methacrylate) (PVA–PMMA) comb copolymer as a structure-directing agent. PVA–PMMA synthesis proceeded via a one-step free radical polymerization, a scalable and economical method for polymer synthesis. Despite a very low combing degree (3 wt%), a large amount of PMMA (25 wt%) was combed to the PVA backbones, resulting in a unique microphase-separated structure. Using a mixed solvent (i.e., dimethyl sulfoxide/tetrahydrofuran) was critical in obtaining the micellization of PVA–PMMA and homogeneous metal oxide films. During the sol–gel process, the hydrolyzed metal precursor selectively interacted with the hydrophilic PVA backbones through coordination, leading to an organized mesoporous structure. Homogeneously well-organized mesoporous structures were obtained for TiO₂, SiO₂, Al₂O₃, and ZrO₂, but particulate structures were formed for Fe₂O₃ and ZnO owing to high reactivity and rapid hydrolysis. Detailed discussion of parameters determining morphology and structure is provided, which is essential for the rational design and reproducible construction of mesoporous metal oxides.

---

Introduction

Mesoporous materials with a large surface area and high porosity have attracted significant attention for various applications including adsorption, separation, catalysis, and electrochemical energy conversion and storage.^1,2 Many methods of generating mesoporous structures have been developed since mesoporous silica was synthesized by Yanagisawa et al.³ One of the common methods of preparing mesoporous metal oxides involves the use of hard templates. In the hard-templating strategy, a porous matrix (e.g., silica or carbon) is impregnated with a precursor solution, and the precursor in the porous channels of the template is transformed into the target material via thermal treatment followed by removal of the template.⁴ Wang et al. used mesoporous silica (SBA-15) as a hard template to prepare hexagonal mesoporous Co₃O₄ for electrochemical energy devices.⁵ Roggenbuck et al. also employed a hard-templating method (CMK-3, mesoporous carbon) to prepare mesoporous MgO.⁶ This hard-templating procedure makes it possible to control the size and morphology of mesostructures by choosing an appropriate hard template. Moreover, mesoporous metal oxides prepared by the hard-templating method possess highly crystalline walls because the template maintains the frame until the structure of the metal oxide is completely formed, after which the template is removed.⁷ The hard-templating method, however, has some drawbacks such as the need for multistep procedures⁸ and the damage to mesostructures that occurs during the elimination of the hard template at high temperature.¹

Another method for synthesizing mesoporous metal oxides is the soft-templating strategy. Based on the sol–gel method, a soft template such as a surfactant or a block copolymer acts as a structure-directing agent to generate a variety of mesoporous metal oxides.⁹ Mesoporous metal oxide films are prepared with the aid of soft templates via the sol–gel method and the evaporation-induced self-assembly (EISA) process.¹⁰ Such a soft-templating approach has the benefits of facile synthetic procedures and applicability to a diverse range of mesostructures, which is contingent on the template and composition of the sol–gel solution. The disadvantages of soft templates, on the other hand, include the high cost of block copolymer templates, and often the formation of amorphous or semi-crystalline walls caused by the poor thermal stability of the template.¹ This is because block copolymers are typically synthesized by anion radical polymerization, which is very sensitive to impurities, and thus, successful synthesis of such polymers requires special expertise. Furthermore, many block copolymer templates completely burn out before reaching the temperature required to form highly crystalline metal oxide walls. For example, the crystal formation temperature is 450 °C for anatase TiO₂, but the degradation temperature of the template is 200–250 °C for poly(ethylene oxide)-b-poly(propylene oxide)-b-poly(ethylene oxide) (Pluronic series),¹¹ 400 °C for poly(ethylene oxide)-b-poly(hydroxybutyrate) (PEO-b-PHB),¹² 350 °C for polystyrene-b-poly(ethylene oxide) (PS-b-PEO),¹¹ and 400 °C for polystyrene-b-poly(4-vinylpyridine) (PS-b-P4VP).¹³ Thus, the development of low-cost polymer templates with higher thermal stability is needed.

Graft or comb copolymers are considered promising alternative soft templates to conventional block copolymers owing to their ease of synthesis and low cost.^14–21 Recently, we suggested a graft copolymer-templating strategy to prepare organized mesoporous TiO₂ films with large surface area, high porosity, and good interconnectivity for photovoltaic applications, using an amphiphilic poly(vinyl chloride)-g-poly(oxyethylene methacrylate) (PVC-g-POEM) comb copolymer.^14,15 This approach has been extended to other mesoporous metal oxides such as SnO₂,^16,17 Al₂O₃,¹⁸ and MgTiO₃.¹⁹ The PVC-g-POEM comb copolymers were synthesized via atom transfer radical polymerization (ATRP) using secondary chlorine atoms as the initiating site, but this process required the use of a catalyst/ligand complex. Because the complex was rather strongly bound to the polymer product, it took time to completely remove it from the product. Therefore, cheaper and more facile methods such as free radical polymerization are required to yield economic benefits and facilitate commercialization.

Here, we demonstrate a general synthesis of a variety of mesoporous metal oxide films in the size range 30–50 nm, including TiO₂, SiO₂, Al₂O₃, ZrO₂, and SnO₂, via a sol–gel method using an amphiphilic poly(vinyl alcohol)–poly(methyl methacrylate) (PVA–PMMA) comb copolymer. The PVA–PMMA comb copolymer was synthesized via free radical polymerization, which is cheaper and easier than the conventional living radical polymerization method. The synthesis of PVA–PMMA was confirmed by Fourier transform infrared spectroscopy (FT-IR), proton nuclear magnetic resonance (¹H-NMR), differential scanning calorimetry (DSC), X-ray diffraction (XRD), and thermogravimetric analysis (TGA). The general mechanism of mesoporous structure formation was proposed on the basis of FT-IR, XRD, scanning electron microscopy (SEM), and transmission electron microscopy (TEM).

Experimental

Materials

PVA (M_w = 85 000–124 000 g mol⁻¹, 99%), methyl methacrylate (MMA, 99%), PMMA (M_w ∼ 120 000 g mol⁻¹), ceric(IV) ammonium nitrate (CAN, ≥98.5%), titanium(IV) isopropoxide (TTIP, 97%), hydrogen chloride solution (HCl, 37 wt%), aluminum tert-butoxide (ATB), tetraethyl orthosilicate (TEOS, 98%), zirconium(IV) tert-butoxide (ZTB, 80 wt% in 1-butanol), tin(II) chloride (SnCl₂, 98%), iron(III) nitrate nonahydrate (INN, ≥98%), zinc acetate dihydrate (ZAD, ≥98%), 1-propanol (anhydrous, 99.7%), triethanolamine (TEA, ≥99%), glacial acetic acid (CH₃COOH, 99.9%), and nitric acid (HNO₃, 70%) were purchased from Aldrich. Tetrahydrofuran (THF, ≥99.5%), methanol (MeOH, 99.9%), and ethanol (EtOH, 99.9%) were obtained from J. T. Baker. Dimethyl sulfoxide (DMSO, 99.9%) was purchased from Duksan. All solvents and chemicals were used as received without any purification or treatment. Fluorine-doped tin oxide (FTO) coated glass was purchased from Pilkington, and Si wafer was purchased from France and Entegris, USA.

Synthesis of PVA–PMMA comb copolymer

The PVA–PMMA was synthesized by Ce(IV)-initiated free radical polymerization. A typical synthetic process is as follows. First, 3 g of PVA was dissolved in 95 ml of DMSO in a round bottomed flask with stirring at 60 °C for 2 h. After cooling the solution to room temperature, 7 g of MMA was added and the solution was mixed until homogeneous. A 0.1 M CAN solution (0.25 g of CAN in 5 ml of DMSO) was then added to the above solution under vigorous stirring, and the flask was sealed with a rubber septum and purged with nitrogen for 1 h. The polymerization reaction was allowed to proceed at 60 °C for 20 h. The synthesized polymer was precipitated by methanol three times and was dried at 50 °C for one day in an oven.

Preparation of mesoporous metal oxide films

The PVA–PMMA was dissolved in a 6 : 1 (v/v) DMSO : THF mixed solvent. Two types of metal precursors were used to prepare sol solutions: metal alkoxides (TTIP, TEOS, ATB, and ZTB as the Ti, Si, Al, and Zr precursors, respectively) and metal salts (SnCl₂, INN, and ZAD as the Sn, Fe, and Zn precursors, respectively). The metal precursor solutions were prepared as follows:

(i) Ti precursor solution. 0.75 ml of HCl was slowly added to 1.5 ml of TTIP under vigorous stirring. Then, 0.75 ml of MeOH was added to the TTIP solution.¹⁵ The TTIP solution was then aged for 30 min (the volume ratio of [TTIP] : [HCl] : [MeOH] was 1 : 0.5 : 0.5).

(ii) Si precursor solution. 1 ml of TEOS and 1 ml of DI water were dissolved in 2 ml of absolute EtOH under vigorous stirring followed by slowly adding 0.4 ml of HCl.²² The TEOS solution was aged for 20 min (the volume ratio of [TEOS] : [EtOH] : [H₂O] : [HCl] was 1 : 2 : 1 : 0.4).

(iii) Al precursor solution. 0.25 g of ATB was added to 3 ml of absolute EtOH. 0.6 ml of HCl and 0.23 ml of CH₃COOH were then added to the ATB solution²³ and it was aged for 30 min (the molar ratio of [ATB] : [EtOH] : [HCl] : [CH₃COOH] was 1 : 50 : 20 : 4).

(iv) Zr precursor solution. 0.64 g of ZTB and 5 ml of absolute EtOH were mixed. The ZTB solution was stirred vigorously, and 0.174 ml of HCl and 0.38 ml of CH₃COOH were added dropwise.²² The solution was aged for 30 min (the molar ratio of [ZTB] : [EtOH] : [HCl] : [CH₃COOH] was 1 : 46 : 3.3 : 4).

(v) Sn precursor solution. Unlike the aforementioned methods, the SnCl₂ (0.21 g) was directly added to the comb copolymer solution. After 15 min, 0.17 ml of DI water was added to the polymer solution under rapid stirring. The mixture was then aged under mild stirring for 12 h ¹⁶ (the molar ratio of [SnCl₂] : [H₂O] was 1 : 8.3).

(vi) Fe precursor solution. 0.8 g of INN and 7.1 ml of absolute EtOH were mixed. After mixing, 0.4 ml of DI water and 0.33 ml of HNO₃ were added to the INN solution,²⁴ which was then aged for 30 min (the molar ratio of [INN] : [EtOH] : [H₂O] : [HNO₃] was 1 : 60 : 10 : 4).

(vii) Zn precursor solution. 0.423 g of ZAD was added to 5 ml of 1-propanol. 0.25 ml of TEA was then added dropwise to the ZAD solution,²⁵ and the solution was aged for 30 min (the molar ratio of [ZAD] : [1-propanol] : [TEA] was 1 : 33.3 : 1).

To determine the proper ratio of metal precursor to comb copolymer for organized mesostructures, two experiments were performed in sequence: (1) different amounts of precursor solution (0.2, 0.6, and 1.0 ml) with a constant concentration (1 wt%) of comb copolymer and (2) different concentrations of comb copolymer (1, 3, and 5 wt%) with a constant amount of precursor solution. The PVA–PMMA/metal precursor hybrids were aged for different times depending on the type of precursor. The aging time and specific conditions for the sol–gel process are shown in Table 1. After aging, each solution was spin-coated onto Si wafer and FTO glass. The samples were then dried in an oven at 50 °C for 1 h before calcining at 500 °C for 30 min to remove the comb copolymer template.

Table 1 Constituents of metal precursor solutions used in the sol–gel method and characteristics of various mesoporous metal oxide films

	Metal precursor solution		Aging time	Transparency	Pore diameter
	Precursor	Solvent
TiO2	TTIP	MeOH, HCl	3 h	O	20–50 nm
SiO2	TEOS	EtOH, HCl, H2O	3 h	O	30–50 nm
Al2O3	ATB	EtOH, HCl, CH3COOH	5 h	O	30–50 nm
ZrO2	ZTB	EtOH, HCl, CH3COOH	3 h	O	30–50 nm
SnO2	SnCl2	H2O	12 h	Δ	20–50 nm
Fe2O3	INN	EtOH, H2O, HNO3	24 h	X	—
ZnO	ZAD	1-Propanol, TEA	3 h	O	—

---

Characterization

FT-IR spectra were collected using an FT-IR spectrometer (Spectrum 100, PerkinElmer, USA) in the frequency range 4000–500 cm⁻¹. The ¹H-NMR spectrum was obtained using a 600 MHz high-resolution NMR spectrometer (ADVANCE 600, Bruker, Germany). DSC curves were measured using a differential scanning calorimeter (DSC-Q1000, TA Instrument, U.K.) at a scan rate of 10 °C min⁻¹ in the temperature range 50–250 °C. The thermal stability of the comb copolymer was studied by TGA, using a TGA/DSC 1 (Mettler Toledo, Korea) with a ramp rate of 10 °C min⁻¹ from room temperature to 600 °C in an air atmosphere. XRD patterns of the samples were obtained by powder X-ray diffractometry (D8 ADVANCE, BRUKER, Germany) with a Cu cathode operated at 40 kV and 40 mA, in the 2θ range 5–80°. The structure and morphology of the metal oxide films were characterized by field emission (FE)-SEM (JSM-7001F, JEOL Ltd.) and TEM (JEM1010, JEOL, Japan).

Results and discussion

Synthesis and characterization of PVA–PMMA comb copolymer

Scheme 1 shows the reaction mechanism of PVA–PMMA comb copolymer formation by free radical polymerization initiated by ceric ammonium nitrate (CAN), which costs one tenth of the price of the conventional ATRP process. The chemical structure of PVA–PMMA was confirmed by FT-IR spectroscopy, as shown in Fig. 1 in which the absorption spectra of neat PVA, MMA monomer, and PVA–PMMA are plotted as a function of wavenumber. The sharp absorption bands at 1721 and 1638 cm⁻¹ for the MMM monomer are assigned to the stretching vibrations of the carbonyl bond (C=O) and the carbon–carbon double bond (C=C), respectively. A broad absorption band at around 3272 cm⁻¹ was observed for neat PVA due to the hydroxyl (OH) stretching vibration. The PVA–PMMA comb copolymer exhibited characteristic absorption bands of PVA and MMA, including the hydroxyl group (OH) and carbonyl group (C=O) stretching. The stretching vibration of the C=C bond at 1638 cm⁻¹ completely disappeared in PVA–PMMA, indicating that there was no remaining MMA monomer in the final product. Upon copolymerization, there were slight band shifts for the OH stretching band from 3272 to 3327 cm⁻¹ and for the C=O stretching band from 1721 to 1725 cm⁻¹. These are attributed to weakened hydrogen bonding or dipole–dipole interactions in the PVA–PMMA compared with the neat PVA and MMA monomers, due to intermingling of the long hydrophobic PMMA chains with the long hydrophilic PVA chains.

Scheme 1 PVA–PMMA comb copolymer synthesis via free radical polymerization.

Fig. 1 FT-IR spectra of neat PVA, MMA monomer, and PVA–PMMA comb copolymer.

The actual mass ratio, combing ratio, and regularity of PVA–PMMA were verified by NMR spectroscopy. The solution was prepared by dissolving PVA–PMMA comb copolymer in deuterated DMSO (DMSO-d₆) at 1 wt%, and the ¹H NMR spectrum was obtained at room temperature (Fig. 2a). The PVA homopolymer has triad tacticity,²⁶ which arises from the relative stereochemistry of three adjacent chiral centers, depending on the configuration of the hydroxyl group (OH) of PVA, as shown in Fig. 2b. The ratio of triad tacticity played a key role in determining the regularity of the macromolecular structure, which could be crystalline or amorphous in nature. The protons of OH (b + b′) in PVA were observed at 4.20, 4.45, and 4.65 ppm, and each resonance indicates an isotactic (mm), heterotactic (mr or rm), or syndiotactic (rr) triad, respectively.²⁷ The ratio of triad tacticity was 21/50/29 (mm/mr/rr), demonstrating that all of the possible triads are randomly distributed,²⁸ as the probability of an ideally random chain is approximately 0.25.²⁹ These results suggest that the PVA–PMMA comb copolymer has an atactic structure of amorphous nature. The resonance of the methylene proton (a + a′) in PVA appeared at 1.29–1.50 ppm, and the resonance 3.80–3.89 ppm could be assigned to a methine proton (c) in PVA. In addition, the signals at 0.75 and 0.94 ppm indicate the methyl proton (f) in PMMA. The methylene (a + a′) and methine groups (c) in PVA, and the methyl group (f) in PMMA, play significant roles in determining the actual mass ratio and the combing ratio. The actual mass ratio of PVA to PMMA was determined to be approximately 75 : 25, whereas the PMMA side chains were only combed from 3.4 wt% of the PVA backbones. Such a low combing degree (3.4 wt%) together with the relatively large amount of PMMA (25 wt%) in the PVA–PMMA comb copolymer might be responsible for the formation of the microphase-separated structure.

Fig. 2 (a) ¹H NMR spectrum of PVA–PMMA comb copolymer. Chemical structure of PVA–PMMA (inset). (b) Triad tacticity of PVA–PMMA.

The DSC curves of neat PVA, PMMA homopolymer, and PVA–PMMA copolymer are shown in Fig. S1.† The PMMA homopolymer showed a glass transition temperature (T_g) of 118 °C without melting, indicating an amorphous structure with a lack of regularity. On the other hand, the PVA homopolymer exhibited a T_g of 76 °C as well as a strong endothermic T_m at 220 °C, revealing its semi-crystalline nature. For PVA–PMMA comb copolymer, two distinct T_g values were observed at 86 and 107 °C, which are due to the motions of the PVA and PMMA chains, respectively. The slight shift in T_g for PVA–PMMA compared with those of the homopolymers (PVA and PMMA) is attributable to the partial miscibility/intermingling of PVA backbones with PMMA side chains due to their secondary bonding interactions. Nonetheless, the presence of two obvious T_g values is a clear indication of the microphase-separated structure of the PVA–PMMA comb copolymer.^14,19 The disappearance of T_m in PVA–PMMA demonstrates that the combing of PMMA from PVA results in a change from a semi-crystalline to an amorphous state, which arises from the intervention of amorphous PMMA long chains in the regularly ordered PVA crystalline chains.

XRD is a very useful tool for identifying the atomic and molecular structure of polymers, especially their crystalline nature. Fig. 3 presents the XRD patterns for neat PVA, PMMA homopolymer, and PVA–PMMA comb copolymer. The sharp peak at 19.7° for neat PVA indicates a semi-crystalline nature. Meanwhile, there were only broad peaks for the PMMA homopolymer, implying an amorphous phase of the polymer. The main peaks for the PVA and PMMA homopolymers were all observed in the PVA–PMMA comb copolymer; the main peak at 19.6° and the shoulder peak at 13.9° are attributed to the PVA backbones and PMMA side chains, respectively. The sharp peak at 19.6° became rather broad and its intensity was significantly reduced, indicating a large decrease in crystallinity, which is consistent with the DSC analysis above.

Fig. 3 XRD patterns of PVA, PMMA, and PVA–PMMA comb copolymer.

TGA is a method to determine the thermal characteristics of materials related to the decomposition mechanism and organic content of samples. The TGA curves of neat PVA, PMMA homopolymer, and PVA–PMMA comb copolymer are presented in Fig. S2.† For the thermogram of PVA–PMMA comb copolymer, there were four major decomposition steps. The first degradation step appeared in the range 100–204 °C, which is attributed to the evaporation of strongly adsorbed water. The second transition was in the range 204–272 °C and is ascribed to chain stripping of the PVA in the comb copolymer. The third step occurred the temperature range 272–374 °C, which is due to the characteristic decomposition of PMMA in PVA–PMMA. Finally, the polyene intermediate formed during the second transition was decomposed above 388 °C, and residual PVA–PMMA was completely degraded above 480 °C. Each step from the first to the final accounted for 8.9, 31.6, 24.8, and 34.7% of the weight loss in PVA–PMMA, respectively. The ratio of weight loss in PVA (31.6 and 34.7 for the 2nd and 4th step, respectively) and PMMA (24.8 for the 3rd step) was approximately 3 to 1, which is consistent with the actual mass ratio in the NMR spectrum above. Although PMMA was completely pyrolyzed at around 400 °C, the PVA–PMMA was maintained up to 480 °C, which could be of great help in forming highly crystalline mesostructures via the sol–gel process.

Formation of mesoporous metal oxide films using a PVA–PMMA template

The surface images of mesoporous TiO₂, SiO₂, Al₂O₃, and ZrO₂ films templated by the PVA–PMMA comb copolymer were characterized using FE-SEM, as shown in Fig. 4 and S3–S5.† These four metal oxides showed high porosity, good interconnectivity, and homogeneously well-organized mesopores with a size of 30–50 nm. As the concentration of PVA–PMMA increased, the pore size gradually increased, and sometimes large macropores were even formed (Fig. S3e and f and S4c–f†). This phenomenon can be explained by the ratio of PVA–PMMA to metal precursor; in the case of the appropriate ratio, the hydrophilic metal precursors located in the vicinity of hydrophilic PVA chains were of great help in developing the regular mesopores by increasing the gap between hydrophilicity and hydrophobicity, and thus well-organized mesoporous structures were obtained. However, an imbalance in the ratio between PVA–PMMA and metal precursors led to denser surface (comb copolymer < precursor) or irregular macropores (comb copolymer > precursor). As shown in Fig. 4d and S6,† the mesoporous TiO₂, SiO₂, Al₂O₃, and ZrO₂ films were all transparent.

Fig. 4 FE-SEM images of mesoporous TiO₂ prepared using different PVA–PMMA concentrations at low and high magnification: (a) 1%, (b) 3%, and (c) 5%, and (d) photographs of the TiO₂ films.

The surface of SnO₂, Fe₂O₃, and ZnO films were also examined by FE-SEM analysis, as shown in Fig. S7–S9.† A mesoporous structure was obtained for SnO₂ films under some conditions, despite the overall morphologies being heterogeneous with cracks. The number of mesopores and their size decreased with the concentration of PVA–PMMA due to the lack of metal precursor (SnCl₂). In the case of Fe₂O₃, many cracks were observed at low magnification, and Fe₂O₃ nanoparticles were interconnected along the template, because the hydrolyzed Fe precursors interacted with the PVA–PMMA template. The formation of nanoparticular Fe₂O₃ instead of a mesoporous film might result from rapid nucleation due to the high reactivity toward hydrolysis of the Fe precursor.^30,31 This rapid hydrolysis may induce the formation of clusters or nanoparticles. This phenomenon also occurred in the ZnO system. The size and number of nanoparticles of ZnO increased with increasing Zn precursor due to the high hydrolysis rate of the precursor and the low crystallization temperature. The crystallization of ZnO at low temperature disrupted the interaction energies at the inorganic–organic interface because the interaction energies are generally dominated by crystallization energies.²⁵ The ZnO films were transparent, whereas SnO₂ and Fe₂O₃ were translucent, as shown in Fig. S10.† The pore diameter range and transparency of each metal oxide are presented in Table 1.

The crystallinity of the metal oxides TiO₂, SiO₂, Al₂O₃, ZrO₂, SnO₂, Fe₂O₃, and ZnO was characterized by XRD analysis, as shown in Fig. 5. The peak locations and relative intensities for the metal oxides are cited from the Joint Committee on Powder Diffraction Standards (JCPDS) database. The diffraction peaks located at 25.24°, 37.92°, 48.08°, 53.96°, and 54.88° in the TiO₂ XRD pattern correspond to the (101), (004), (200), (105), and (211) planes of anatase TiO₂, respectively (JCPDS no. 21-1272). The mesoporous ZrO₂ film exhibited diffraction peaks centered at 30.24° and 50.48°, which are assigned to the (101) and (200) planes of tetragonal ZrO₂ (JCPDS no. 80-0965). The XRD pattern of the SnO₂ sample showed peaks for the tetragonal phase of SnO₂ (JCPDS 41-1445) at 2θ = 26.72°, 34.08°, 38.2°, 51.72°, and 54.68°, corresponding to the (110), (101), (200), (211), and (220) planes of the tetragonal phase of SnO₂, respectively (JCPDS no. 41-1445). The diffraction peaks centered at 24.16°, 33.08°, 35.56°, 40.92°, 49.44°, 54.12°, and 57.40° in the Fe₂O₃ XRD pattern correspond to the (012), (104), (110), (113), (024), (116), and (018) planes of the hematite phase (α-Fe₂O₃, JCPDS no. 33-0664). The ZnO nanoparticles contained the diffraction peaks of hexagonal ZnO located at 31.84°, 34.44°, 36.36°, 47.76°, and 56.60°, which correspond to the (100), (002), (101), (102), and (110) planes, respectively (JCPDS no. 36-1451). On the contrary, no strong peaks were detected for the SiO₂ and Al₂O₃ patterns, indicating a lack of crystallinity; this is because both SiO₂ and Al₂O₃ often exhibit amorphous behavior under 900 °C.^32,33 Moreover, the main broad peak of PVA–PMMA comb copolymer at 19.6° completely disappeared in all XRD patterns, indicating that the polymer template was completely removed by calcination at 500 °C.

Fig. 5 XRD patterns of mesoporous metal oxide films (TiO₂, SiO₂, Al₂O₃, ZrO₂, Fe₂O₃, SnO₂, and ZnO).

The formation of mesoporous metal oxide films was developed by two processes: sol–gel and EISA, as shown in Scheme 2. Microphase-separation of the amphiphilic PVA–PMMA comb copolymer was enhanced by dissolving in appropriate solvents. In developing micellar mesostructures, hydrophilic PVA chains formed in the outside as the surface of micelles while hydrophobic PMMA chains formed the core of micelles. During a sol–gel process, hydrophilic metal precursors were arranged in along PVA chains due to preferential interaction. This PVA–PMMA/precursor hybrid was spin-coated on FTO glass and was self-assembled via evaporation of a volatile solvent. After calcination, mesoporous metal oxide films were obtained.

Scheme 2 Mechanism for the formation of well-organized mesoporous metal oxide film using a PVA–PMMA comb copolymer by sol–gel process.

The selection of solvent was of pivotal importance for the comb copolymer-templating approach. In our study, the mixed solvent (DMSO : THF, 6 : 1 v/v) was chosen to dissolve the structure-directing agent, i.e., PVA–PMMA. The solubility parameter (δ) provides a numerical estimate of the affinity between the different materials; i.e., similar δ values for two materials represents good affinity and miscibility.^34,35 The δ values of DMSO, PVA, PMMA, and THF are 13.1, 12.9, 10.0, and 9.5 cal^1/2 cm^−3/2, respectively, showing the tendency that materials with a high polarity have high δ values. According to solubility tests, DMSO dissolves all three polymers (PVA, PMMA, and PVA–PMMA), whereas THF only dissolves PMMA, which is consistent with the above δ estimates. This indicates that DMSO and THF are good and poor solvents for PVA–PMMA comb copolymer, respectively. Upon casting from DMSO, the PVA–PMMA comb copolymer was uniformly distributed (Fig. 6a), because the chains could obtain a stretched-out conformation based on the strong interaction between the solvent and the polymer. When a small amount of THF was added to DMSO, the PVA–PMMA comb copolymer agglomerated to form micellar aggregates approximately 50 nm in size (Fig. 6b), which are responsible for generating organized mesoporous structures (Fig. 6c and d). Similar images of spatial distribution were observed for the PVA–PMMA micelles (in Fig. 6b) and the TiO₂ film pores (in Fig. 6d), indicating the effectiveness of PVA–PMMA comb copolymer as a structure directing agent. Furthermore, the addition of THF was crucial in obtaining homogeneous films on the substrate (inset of Fig. 6d). The THF helps DMSO to easily vaporize by forming a miscible mixture, due to a high volatility of THF,³⁶ leading to homogeneously organized self-assembly of organic–inorganic hybrids on the substrate.³⁷

Fig. 6 TEM images of PVA–PMMA comb copolymer cast from (a) DMSO only, and (b) DMSO/THF mixed solvent, and SEM images of mesoporous TiO₂ films templated by PVA–PMMA comb copolymer using (c) DMSO only, and (d) DMSO/THF mixed solvent inset: photos of mesoporous TiO₂ films prepared using (left) DMSO only, and (right) DMSO/THF mixed solvent.

A variety of mesoporous metal oxide films were prepared via the sol–gel method, which involves the hydrolysis of a metal precursor followed by polycondensation. Two types of metal precursors, i.e., metal alkoxides and metal salts (chloride, nitrate, and acetate), were used in our experiment. It is expected that the metal cations of the precursors strongly and preferentially interact with the hydroxyl (–OH) groups in the PVA³⁸ through coordination bonding. Metal alkoxides such as TTIP, TEOS, ATB, and ZTB underwent a hydrolysis process upon the addition of water to the precursor solution. The hydrolyzed precursors condensed with each other, which led to the formation of M–O–M bonds.^39,40 The stoichiometric coefficients and valence electrons in the metal ions were not considered for convenience.

Hydrolysis: M–OR + H₂O → M–OH + ROH (1)

Polycondensation: M–OH + M–OH → M–O–M + H₂O, M = metal, R = alkyl group (C_xH_2x+1) (2)

Similarly, metal salts such as stannous chloride, iron nitrate nonahydrate, and zinc acetate dihydrate follow the same process of M–O–M bond formation, but undergo an additional process prior to hydrolysis. When the metal salts were dissolved in alcohols or water, the metal salts underwent dissociation. In the presence of water, metal anions were in equilibrium with acids and OH⁻ anions were generated. These anions combined with metal cations or metal sources through nucleophilic attack, forming metal hydroxides (M–OH). Condensation of M–OH led to the formation of M–O–M bonds as a source of coordination with PVA.^41,42

Dissociation I: M–X ↔ M⁺ + X⁻ (3)

Dissociation II: X⁻ + H₂O ↔ XH + OH⁻ (4)

Hydrolysis: M⁺ + OH⁻ → M–OH (5)

Polycondensation: M–OH + M–OH → M–O–M + H₂O, M = metal, X = Cl₂, (C₂H₃O₂)₂, (NO₃)₃ (6)

These M–OH and M–O–M bonds can interact with the hydrophilic PVA main chains in the PVA–PMMA comb copolymer through coordination interactions. The PVA–PMMA comb copolymer forms micellar mesostructures in DMSO/THF mixed solvent, as shown in the above TEM image (Fig. 6b). When the metal precursor solution was added to the PVA–PMMA solution, PVA–PMMA/precursor hybrids were generated, after which the precursor underwent hydrolysis and polycondensation, mainly in the domains located adjacent to the PVA chains due to the coordination interactions between the OH in PVA and the Ti–OH or Ti–O–Ti bonds. After spin coating onto the substrate and calcining at high temperature (e.g., 450 °C), the metal precursor was transformed into a mesoporous metal oxide film and the comb copolymer template was completely removed.

The coordination interactions between PVA and the metal precursors were confirmed by shifts in the absorption band of the hydroxyl (OH) groups in PVA, as can be seen in Fig. 7. The PVA–PMMA/precursor hybrids were prepared by drying at 70 °C for 3 days to eliminate the solvent effect (without calcination at 450 °C) and used for characterization. The other characteristic bands noted in Fig. 7 are assigned in more detail in Table S1.† The broad –OH stretching band at 3336 cm⁻¹ observed for the PVA–PMMA comb copolymer shifts toward lower wavenumbers for all samples. This is because the O–H bonds of PVA become weak due to the coordination with the M–O–M bonds of the precursor. On the other hand, the sharp C=O stretching band at 1724 cm⁻¹ hardly shifts, suggesting negligible specific interactions between the hydrophobic PMMA side chains and M–OH or M–O–M bonds. These results are consistent with the proposed mechanism for the formation of PVA–PMMA/metal precursor hybrids. Furthermore, the absorption bands of metal hydroxides and metal oxides were observed simultaneously for all of the samples, demonstrating that metal precursors partially underwent the hydrolysis and condensation process.

Fig. 7 FT-IR spectra of PVA–PMMA comb copolymer with different kinds of metal precursors.

Conclusions

In summary, we conducted a systematic study of the formation of various kinds of organized mesoporous metal oxide films through the sol–gel method using an amphiphilic PVA–PMMA comb copolymer template synthesized via free radical polymerization. The PVA–PMMA comb copolymer possessed an amorphous, microphase-separated structure due to a very low combing degree (3 wt%) and the large fraction (25 wt%) of hydrophobic PMMA side chains, as characterized by FT-IR, NMR, DSC, XRD, and TGA analysis. The following mechanism of mesostructure formation was proposed: (1) the micellization of the PVA–PMMA comb copolymer occurred by using a DMSO/THF (6 : 1) mixed solvent due to mismatch of interactions between the polymer and the solvent, as revealed by TEM and SEM analysis. (2) The hydrophilic metal precursors or hydrolyzed precursors preferentially and selectively interact with the hydrophilic PVA backbones via coordination, as confirmed by FT-IR spectroscopy. With the appropriate ratio of comb copolymer to precursor, organized mesoporous structures were formed with high porosity, good interconnectivity, and uniform mesopores in the range 30–50 nm, as observed for TiO₂, SiO₂, Al₂O₃, and ZrO₂. However, slow crystallization at low temperature and rapid hydrolysis of the metal precursor for Fe₂O₃ and ZnO resulted in a particulate structure without mesopores. Nevertheless, the PVA–PMMA comb copolymer can be a promising template for the formation of organized mesoporous metal oxides as an alternative to conventional block copolymer templates due to its low cost, facile synthesis, and ability to form highly crystalline metal oxides.

Acknowledgements

This work was supported by the Center for Advanced Meta-Materials (NRF-2014M3A6B3063716) and the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIP) (NRF-2015R1C1A1A01053807).

References

1. Y. Ren, Z. Ma and P. G. Bruce, Chem. Soc. Rev., 2012, 41, 4909–4927.
2. N. Yang, F. Pang and J. Ge, J. Mater. Chem. A, 2015, 3, 1133–1141.
3. T. Yanagisawa, T. Shimizu, K. Kuroda and C. Kato, Bull. Chem. Soc. Jpn., 1990, 63, 988–992.
4. D. Gu and F. Schuth, Chem. Soc. Rev., 2014, 43, 313–344.
5. G. Wang, H. Liu, J. Horvat, B. Wang, S. Qiao, J. Park and H. Ahn, Chem.–Eur. J., 2010, 16, 11020–11027.
6. J. Roggenbuck, G. Koch and M. Tiemann, Chem. Mater., 2006, 18, 4151–4156.
7. W. Yue and W. Zhou, Prog. Nat. Sci., 2008, 18, 1329–1338.
8. S.-W. Cao and Y.-J. Zhu, Nanoscale Res. Lett., 2011, 6, 1.
9. Y. Yue, T. Umeyama, Y. Kohara, H. Kashio, M. Itoh, S. Ito, E. Sivaniah and H. Imahori, J. Phys. Chem. C, 2015, 119, 22847–22854.
10. B. P. Bastakoti, S. Ishihara, S.-Y. Leo, K. Ariga, C.-W. Wu and Y. Yamauchi, Langmuir, 2014, 30, 651–659.
11. X. Jiang, N. Suzuki, B. P. Bastakoti, W.-J. Chen, Y.-T. Huang and Y. Yamauchi, Eur. J. Inorg. Chem., 2013, 2013, 3286–3291.
12. N. M. Nursam, X. Wang and R. A. Caruso, J. Mater. Chem. A, 2015, 3, 24557–24567.
13. A. Guiet, T. Reier, N. Heidary, D. Felkel, B. Johnson, U. Vainio, H. Schlaad, Y. Aksu, M. Driess, P. Strasser, A. Thomas, J. Polte and A. Fischer, Chem. Mater., 2013, 25, 4645–4652.
14. S. H. Ahn, J. H. Koh, J. A. Seo and J. H. Kim, Chem. Commun., 2010, 46, 1935–1937.
15. S. H. Ahn, D. J. Kim, W. S. Chi and J. H. Kim, Adv. Funct. Mater., 2014, 24, 5037–5044.
16. J. T. Park, S. H. Ahn, D. K. Roh, C. S. Lee and J. H. Kim, ChemSusChem, 2014, 7, 2037–2047.
17. J. T. Park, C. S. Lee and J. H. Kim, Nanoscale, 2015, 7, 670–678.
18. H. Jeon, S. Ahn, J. Kim, Y. Min and K. Lee, J. Mater. Sci., 2011, 46, 4020–4025.
19. D. K. Roh, S. J. Kim, H. Jeon and J. H. Kim, ACS Appl. Mater. Interfaces, 2013, 5, 6615–6621.
20. C. Vannucci, I. Taniguchi and A. Asatekin, ACS Macro Lett., 2015, 4, 872–878.
21. P. Bengani, Y. Kou and A. Asatekin, J. Membr. Sci., 2015, 493, 755–765.
22. N. Suzuki, T. Kimura and Y. Yamauchi, J. Mater. Chem., 2010, 20, 5294–5300.
23. K. Niesz, P. Yang and G. A. Somorjai, Chem. Commun., 2005, 1986–1987.
24. W. Hamd, S. Cobo, J. Fize, G. Baldinozzi, W. Schwartz, M. Reymermier, A. Pereira, M. Fontecave, V. Artero, C. Laberty-Robert and C. Sanchez, Phys. Chem. Chem. Phys., 2012, 14, 13224–13232.
25. M.-H. Hong, C.-S. Park, S. Shin, H. H. Cho, W.-S. Seo, Y. S. Lim, J.-K. Lee and H.-H. Park, J. Nanomater., 2013, 2013, 6.
26. T. Ikejima, N. Yoshie and Y. Inoue, Macromol. Chem. Phys., 1996, 197, 869–880.
27. K. Yamada, T. Nakano and Y. Okamoto, Polym. J., 1998, 30, 641–645.
28. A. Narumi, H. Baba, T. Akabane, Y. Saito, S. Ohno, D. Togashi, K. Enomoto, M. Kikuchi, O. Haba and S. Kawaguchi, Macromolecules, 2015, 48, 3395–3405.
29. M. J. Miri, B. P. Pritchard and H. N. Cheng, J. Mol. Model., 2010, 17, 1767–1780.
30. W. Stöber, A. Fink and E. Bohn, J. Colloid Interface Sci., 1968, 26, 62–69.
31. K. M. S. Khalil, H. A. Mahmoud and T. T. Ali, Langmuir, 2008, 24, 1037–1043.
32. L. Zhao, N. Li, A. Langner, M. Steinhart, T. Y. Tan, E. Pippel, H. Hofmeister, K. N. Tu and U. Gösele, Adv. Funct. Mater., 2007, 17, 1952–1957.
33. K. Matori, L. Wah, M. Hashim, I. Ismail and M. Zaid, Int. J. Mol. Sci., 2012, 13, 16812.
34. C. S. Brazel and S. L. Rosen, Fundamental principles of polymeric materials, John Wiley & Sons, 2012.
35. J. Y. Lim, C. S. Lee, J. M. Lee, J. Ahn, H. H. Cho and J. H. Kim, J. Power Sources, 2016, 301, 18–28.
36. J. Li, M. Rodrigues, A. Paiva, H. A. Matos and E. G. de Azevedo, J. Supercrit. Fluids, 2007, 41, 343–351.
37. L. Mahoney and R. Koodali, Materials, 2014, 7, 2697.
38. Y. Lu, G. Deng, F. Miao and Z. Li, Carbohydr. Res., 2004, 339, 1689–1696.
39. M. Simonsen and E. Søgaard, J. Sol-Gel Sci. Technol., 2010, 53, 485–497.
40. I. A. Rahman and V. Padavettan, J. Nanomater., 2012, 2012, 15.
41. M. Habibi and M. Khaledi Sardashti, Int. J. Biomed. Nanosci. Nanotechnol., 2008, 4, 13–16.
42. D. Kumar, H. Singh, S. Jouen, B. Hannoyer and S. Banerjee, RSC Adv., 2015, 5, 7138–7150.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra13844f

---