Novel and facile synthesis of transparent-monolithic TiO₂ gels by sol-gel method based on an esterification reaction

The esterification reaction leads to progressive formation of water and isopropyl acetate in situ. The infrared spectrometer was used to get a visual evidence that a chemical change has occurred in an isopropyl acetate condensation reaction between isopropanol and acetic acid: C3H7OH+CH3COOH→CH3COOC3H7+H2O$${{\rm{C}}_3}{{\rm{H}}_7}{\rm{OH}} + {\rm{C}}{{\rm{H}}_3}{\rm{COOH}} \to {\rm{C}}{{\rm{H}}_3}{\rm{COO}}{{\rm{C}}_3}{{\rm{H}}_7} + {{\rm{H}}_2}{\rm{O}}$$(1)

The water produced in situ by the esterification reaction is difficult to identify because its vibration bands are masked by the presence of hydroxyl groups (–OH) of isopropanol and acetic acid. Thus, to highlight the esterification reaction, we followed the formation of isopropyl acetate, which is produced in the same proportions as the water. FT-IR spectra of acetic acid, isopropanol, isopropyl acetate and water are shown in Fig. 2.

Fig. 2

The main bands of the isopropyl acetate (spectrum c) obtained after subtraction of the bands of the two reactants are given in Table 1 and correspond to those indicated in the literature [18]. It is well known that esterification is a reversible reaction limited by the low equilibrium conversion and slow reaction rate between reagents. Therefore, it was not surprising to find some unreacted acetic acid and isopropanol in the esterification mixture. The unreacted acetic acid was determined by titration of the solution with NaOH (1 M) (Table 2).

The esterification mixture obtained at 75 °C during 60 min (described above) was mixed with a desired amount of Ti(OC₃H₇)₄ precursors under stirring for 30 min. The experiments were performed with Ti = 0.034 mol, molar ratio AcOH/Ti = 5.5 and H₂O/Ti = 4.5 (Table 1). After intimate mixing, a drop of the solution was placed on KBr pellet for FT-IR measurement (Fig. 3). As shown in the spectrum (Fig. 3c), the bands at 1006 cm^-1, 1127 cm^-1 and 1168 cm^-1, 2965 cm^-1 and 2870 cm^-1, characteristic of Ti(O–C) stretching vibration of isopropoxy groups OC₃H₇ bonded to titanium ions have disappeared [19].

Fig. 3

This excludes partially the presence of acetate ligands or hydrolyzed species, such as Ti(OC₃H₇)_x(OOCCH₃)_yOH and Ti(OC₃H₇)_xOH_y in the mixture [20]. The formation of –OH groups (bands at 3400 cm^-1 and 1633 cm^-1) in Ti[(OH)_y(OOCCH₃)_x] species is then favored with high hydrolysis rates for a medium amount of water. The band at 1633 cm^-1 is strongly screened by the dominant bands at 1553 cm^-1 and 1453 cm^-1. This result is similar to that published by Ivanda et al. [21] for the modification of Ti(IV)-isopropoxide through an esterification reaction between acetic acid and alcohol. The authors proposed that the hydroxylation of titanium alkoxide with water molecules generated in situ by esterification provided excellent control of the hydrolysis and condensation rate. The bridging acetates in Ti[(OH)_y(OOCCH₃)_x]

species have appeared in FT-IR spectrum. The peaks at 1029 cm^-1, 1049 cm^-1, 1351 cm^-1, 1453 cm^-1, 1553 cm^-1, 1607 cm^-1 (inset) and 1614 cm^-1 are assigned to adsorbed acetic acid on Ti sites [22–25]. In particular, the bands at 1614 cm^-1, 1553 cm^-1 and 1453 cm^-1 reveal that the acetate species (asymmetric v_asym(COO) and symmetric v_sym(COO) stretching vibrations) resulting from the acetic acid dissociation may bind to the Ti as a bidentate acetate ligands. The frequency separation (Δν = 100 cm^-1) between the doublets at 1553 cm^-1 and 1453 cm^-1 indicates a coordination in bidentate geometry and suggests a bridging configuration. The peaks at 1795 cm^-1, 1296 cm^-1 and 1410 cm^-1 are due to the free AcOH molecules with monomeric and dimeric (1720 cm^-1) forms [26]. Based on the above results, we suggest that we now have a polymeric titanuim acetate, such as Ti[(OH)_y(OOCCH₃)_x] much less reactive species, with x and y between 0 and 2, from which all OC₃H₇ groups have completely disappeared (942 cm^-1) [27]. After gelification of the transparent solution under controlled conditions, the gels obtained are transparent.

Fig. 3d gives the FT-IR spectrum recorded for the same solution (described above) after reaction time 120 min. Due to complexity of the experiment, the titration of unreacted AcOH was very difficult at this stage. However, we can observe that the acetic acid interacted more strongly with the Ti atoms. This strong interaction has been further confirmed by the increase of the bands at 1029 cm^-1, 1049 cm^-1, 1553 cm^-1 and 1453 cm^-1, assigned to acetate groups bonded to Ti atoms. The weak band at 1345 cm^-1 is attributed to (CH₃) bending of acetate adsorbed onto the surface of titania.

The frequency shift of the band centered in the range of 1725 cm^-1 to 1714 cm^-1 together with the increase of the band at 1290 cm^-1 suggest that some monodentate acetate ligands that exhibit these bands in the same region should be linked to Ti atoms. They can also be attributed to free AcOH but their intensity seems to be too high to be due to AcOH only. Furthermore, the band at 1248 cm^-1, characterstic of isopropyl acetate molecule rapidly increases with increasing reaction time with respect to isopropyl acetate before modification (spectrum b). This result reveals that the isopropanol C₃H₇OH (945 cm^-1) originated from Ti(OC₃H₇)₄ can participate in the esterification reaction 1. Based on the law of chemical equilibrium (law of mass action), the equilibrium of Fischer esterification can be driven to the left (formation of isopropyl acetate and water) by using an excess of isopropanol. Therefore, it is not surprising to detect a large amount of isopropyl acetate in the medium.

Fig. 4 shows the FT-IR spectra recorded for the xerogels obtained after drying the previous gels at T = 90 °C (where T > T_boiling isopropyl-acetate). Compared with Fig. 3, clear bands around 846 and 640 cm^-1 can be observed in Fig. 4, which can be assigned to the vibration of the Ti–O–Ti moiety in titanium oxide network, suggesting the formation of TiO₂. Simultaneously, the absorption bands of the acetate ligands at 1714 cm^-1 and 1290 cm^-1 disappear indicating that the monodenetate has transformed into the structure of Ti–O–Ti by a ploycondensation reaction [28]. It must be pointed out that the intensities of the peaks at 1553 cm^-1 and 1453 cm^-1 noticeably decreased and have shifted toward smaller wavenumbers, ultimately reaching 1534 cm^-1 and 1435 cm^-1. It can be assumed that the decrease in intensity results from breaking of the acetate bridges.

Fig. 4

The powder X-ray diffraction (XRD) patterns of the samples are shown in Fig. 5. No obvious diffraction peaks are observed in the XRD pattern of the sample synthesized for 120 min, suggesting that the xerogels adopt an amorphous phase. Another possible reason is that the content of crystalline phase in the sample is below the detection limit of this technique. To confirm this supposition, we evaluated the XRD of a new sample prepared with high acetic acid content (AcOH/Ti = 10) and dried at 90 °C. From the spectra in Fig. 5 (b to d) it can be seen that this sample is composed of amorphous phase, unknown Ti-containing complex and a small, but not negligible fraction of anatase (Pearson’s Crystal Data PCD# 1003622). The XRD pattern is similar to that observed by Liao et al. [29]. They prepared anatase TiO₂ catalysts of different morphologies by solvothermal reactions of titanium n-butoxide and acetic acid (AcOH/Ti = 180) at different reaction temperatures of 120 °C to 180 °C. They showed that the sample prepared at 170 °C was composed of anatase TiO₂ and an unknown Ti-complex Ti₆O₆(AcO)₆(OBu)₆. Further study should be carried out to identify this complex. The synthesis route followed in this work confirms the formation of Ti-complex even at AcOH/Ti = 10, which has not been reported elsewhere. However, only anatase TiO₂ can be observed for the sample calcined at 250 °C and 450 °C.

Fig. 5

UV-Vis absorption spectra of transparent TiO₂ xerogel and TiO₂ anatase (commercial) are shown in Fig. 6. As expected, TiO₂ anatase characteristic absorption occurs in UV region in a wavelength of around 300 nm to 400 nm (3.2 eV). For transparent TiO₂ xerogel, the absorption spectrum changes dramatically in relation to neat anatase TiO₂ and spawns over the visible region (400 nm to 700 nm), suggesting transparency of the material in the visible range. The results are similar to those reported in the literature for synthesized polymer TiO₂ [30, 31], ZnO [32] and ZrO₂ [33] films in an organic phase made of polycarbonate or polyesters, methyl methacrylate, such as polycaprolactone, poly(D, L-lactic acid), oleic acid and carboxylic acid. Mono- and bidentate complexes can be formed between titanium oxo groups and organic acid after condensation and oligomerization, providing a good homogeneity in the final material and maintaining transparency in the visible range.

Fig. 6

As reported in the preceding sections, all FT-IR features typical of Ti(OC₃H₇)₄ bonds have completely disappeared. This suggests that we now have a polymeric titanuim acetate Ti[(OH)_y(OOCCH₃)_x]. This species is less hydrolyzed than Ti(OC₃H₇)_xOH_y, increasing therefore the time of gelation. Upon aging for a week in glass tubes (Fig. 7a), the viscosity of the transparent solution gradually increased and stabilized to form transparent and yellowish gels. The gel appearance has not changed with changing of volume. There are two reasons responsible for this result. One possible reason is that the low temperature causes a decrease in reaction rate of esterification and low generation of water in situ. Another reason is that the presence of small water content cannot hydrolyze and condense easily the new Ti[(OH)_y(OOCCH₃)_x] complex. The later species is slightly soluble and can be dissolved in situ in H₂O from esterification for a week without any further hydrolysis reaction. Hence, the formation of precipitate is suppressed. For practical use, the same transparent solution was coated on a glass Petri dish using the same experimental conditions described above (Fig. 7b). After drying at 90 °C over night, the sample was crashed in the form of shiny granules under the effect of heat flow (Fig. 7c).

Fig. 7

Doeuff et al. [34] showed that the transparent monolithic gel can be obtained only when external water is intentionally added to the transparent solution containing the titanium alkoxide and acetic acid. The reported experimental conditions were: AcOH/Ti = 1.5, H₂O/Ti = 4 and Ti = 0.8 mol⋅L^-1. This is not the case of this study. We focused only on the esterification reaction as an elegant resource to generate water homogeneously in situ.

When esterification reaction still advances at room temperature, the viscosity gradually increases up to a certain limit, indicating the presence of higher amount of water produced in situ from the condensation between excess of C₃H₇OH and unreacted acetic acid. In such case, there is an optical gel transformation from transparent to monolithic transluscent gel within three days (Fig. 7d₁). Note that the viscosity of the transluscent gel can be maintained by adding 3 or 6 mL of external water (Fig. 7d, sample 3 and sample 4). Because the esterification reaction is reversible, the equilibrium (reaction 1) is driven to the left by adding water [35]. The gel remains transluscent without any precipitate.