Abstract
Photocatalytic materials based on silica-titania (SiO2-TiO2) were synthesized by sol-gel and dip-coating method. TEOS and titanium butoxide were used as precursors of the silica-titania, respectively. A thin film with anatase phase was obtained on the surface of the support. The effect of variables as dispersion mechanism, immersion time, and number of treatment cycles were studied. The materials were characterized using X-ray diffraction, scanning electron microscopy, energy dispersion scanning, and N2 adsorption-desorption. The highest crystallinity of TiO2 on silica, high specific surface area in TiO2-SiO2 materials, and thin film formation were obtained by using a stirring plate and minimum immersion time. The so synthesized catalyst allowed the production of formaldehyde from the photocatalyzed methanol oxidation in a packed-bed reactor.
1. Introduction
Among the advanced oxidation processes (AOPs), heterogeneous photocatalysis has been widely applied in the degradation of organic compounds, hydrogen production from water, reduction of heavy metals, and selective oxidation reactions [1–4]. Selective oxidation of alcohols to aldehydes by photocatalysis is an attractive route because it can be carried out under mild conditions (room temperature, atmospheric pressure, and neutral pH) [5, 6]; particularly, formaldehyde is a highly important intermediate compound in the chemical industry because of its use in the synthesis of adhesives, fertilizers, pyridines, drugs, polyols, and dyes among others [7, 8]. The industrial production of formaldehyde is performed from the oxidation and dehydrogenation of methanol with iron molybdate and silver catalysts, respectively [9, 10].
The most studied semiconductor in the field of heterogeneous photocatalysis is titanium dioxide (TiO2) due to its high oxidative capacity, nontoxicity, low cost, high chemical and physical stability, corrosion resistance, and chemical inertness [11, 12]. TiO2 has been widely used in both powders and thin films [13, 14]. When it is used as suspended particles, a separation step such as filtration or centrifugation is required at the end of the process. For this reason, the immobilization of the particles in an appropriate substrate has attracted great interest [15].
There are several methods for the preparation of TiO2 thin films, and the physicochemical properties strongly depend on the selected method [16–19]. Among all the techniques, the dip-coating sol-gel method is the most widely used because of its good homogeneity, simplicity, low cost, and low temperature during the process [20]. On the other hand, the nature and type of substrate should not be left aside. Amorphous SiO2 has shown to be an excellent support due to its mechanical properties, good adsorption capacity, and high surface area [21, 22].
This work aims to elucidate the effect on morphological, textural, and structural properties of three important variables involved in the synthesis of TiO2 coated SiO2 monoliths. These variables are the mechanism to disperse Ti alkoxide species in an appropriate medium allowing the hydrolysis-condensation processes, dipping time, and number of coating cycles (dip + drying + calcination). In order to evaluate the photocatalytic activity of the synthesized catalysts, the photooxidation of methanol was carried out in a bench-scale continuous-flow packed-bed reactor. This reaction was elected because of its industrial importance and because it is a consecutive reaction, whose selectivity towards intermediate compounds may help to prove the application of TiO2 in selective oxidation processes instead of the organic compounds mineralization. This process could offer the advantage being performed at mild temperature and pressure conditions unlike other existing processes that occur at relatively high temperatures.
2. Experimental
2.1. Preparation of SiO2 Monoliths
Silica dioxide monoliths were synthesized by sol-gel method. The synthesis of the silica support was performed using tetraethyl orthosilicate (TEOS) [Si(OC2H5)4] as alkoxide precursor of the Si sol.
First, the ethanol was added into a beaker and it was maintained under continuous stirring until the temperature reached 60°C. At this point, alkoxide was added and mixed into the beaker for 15 min. After this time, a water and nitric acid solution (1 : 0.0012 molar ratio) was added and the mixture was kept under stirring and keeping the temperature constant for one hour. The molar ratio of water : ethanol : TEOS was of 16 : 4 : 1, respectively.
After that, 2.5 ml of the resulting sol were poured into one container to begin the aging process. This was repeated several times to obtain various monoliths. The container lids were previously drilled to allow solvent diffusion. During this step, alkoxide groups are removed by acid- or base-catalyzed hydrolysis reactions, and link networks O-Si-O are formed in subsequent condensation reactions involving hydroxyl groups [23, 24]. Depending upon the water : alkoxide molar ratio R, pH, temperature, and solvent, condensation leads to different polymeric structures such as linear, entangled chains, clusters, and colloidal particles [25].
The obtained monoliths were then dried from room temperature to 100°C during 14 hours with a slow heating profile to eliminate the solvent. The drying was performed in an Isotemp Vacuum Oven programmable stove model 282 A. The drying treatment was slow to lead the formation of open pores. The drying profile was as follows: 1 h at 40°C, 2.5 h at 50°C, 13 h at 60°C, 2.5 h at 70°C, 3.5 h at 80°C, 2.5 h at 90°C, and 27 h at 100°C. This procedure was performed in order to keep the structure, since a fast drying profile could cause a structure collapse causing cracking of the monolith.
Finally, to provide the monoliths with the appropriate structural and mechanical properties, they were calcined from room temperature (25°C) to 550°C for 6 h at a heating rate of 2.5°C/min using a Jelrus muffle with 2 steps. Amorphous silica compounds without a defined crystalline phase are found at this temperature.
2.2. TiO2 Synthesis
Ethanol, water, titanium butoxide, and diethanolamine (basic catalyst) were used to obtain Ti sols via sol-gel method. Titanium butoxide (Ti[O(CH2)3CH3]4) was dispersed in the ethanol. Immediately, a diethanolamine and water solution was dropped into the volume. It is necessary to maintain a 1 : 1 alkoxide : water molar ratio. Once the solution addition was completed, the agitation was maintained for two hours. After this time, the solution was aged for further two hours without any stirring. Diethanolamine was elected because of its low reactivity during sol-gel process, this makes the hydrolysis reactions slow by favoring the thin film formation [26].
Finally, SiO2 monoliths were immersed into the Ti sol obtained. In this process, the studied variables were the dispersion mechanism (mechanic or ultrasound) and the residence time into the Ti sol. For the former, a stirring plate and an ultrasonic cleaner were utilized as agitation media. Regarding residence time in the Ti sol, this variable was studied at three levels (half, one, and three hours) for each stirring medium. The number of cycles (immersion-drying-calcination) was also studied in order to establish its relationship with the amount of titania on SiO2.
The monoliths coated with titanium species were dried at room temperature for 24 h and calcined under an air flow at 550°C for 5 hours.
2.3. Catalysts Characterization
A Bruker Advance 8 diffractometer was employed to carry out the X-ray diffraction analysis and determine the presence of anatase in the synthesized catalysts. The patterns were obtained using CuKα radiation at 20 kV and 20 mA. Data were collected over 2θ range of 5–50° with a step of 0.5°/min.
A JEOL JSM-6510LV electron microscope coupled with an energy dispersive X-ray spectrometer was employed to observe the surface morphology of the prepared catalysts and to perform elemental analysis of the catalysts.
Autosorb-1 Quantachrome sorption equipment was employed to determine the specific surface area and average pore diameter of the synthesized samples by using liquid nitrogen (77 K). The pore size distributions and the specific surface areas of the materials were estimated by Dubinin-Astakhov (DA) and Brunauer Emmett–Teller (BET) methodologies, respectively.
2.4. Bench-Scale Photocatalytic Reactor
The photooxidation of methanol was performed using a bench-scale continuous packed-bed reactor. An eight-watt UV lamp emitting 254 nm waves was placed right in the center of the reactor. 30 monoliths constituted the catalytic bed. Compressed air was used as carrier gas to help methanol to flow through the packed-bed reactor. The air flow was constant at 50 ml/min. The methanol liquid was heated at 65°C in order to vaporize it. The reactor set-up is depicted in Figure 1.
Identification of formaldehyde by photooxidation of methanol was carried out according to the following methodology: a proper container with a 2,4-dinitrophenylhydrazine (2,4-dnph) solution was placed instead of the condenser (see Figure 1), in such a way that the exit stream was directly bubbled into the solution. This was conducted with all synthesized materials. If there was formaldehyde in such a stream, then a precipitate was observed. This precipitate was the 2,4-dinitrophenylhydrazone, which is the product of the reaction between the 2,4-dnph and the aldehyde as depicted in Figure 2 [27].
The 2,4-dnph solution was prepared as follows: 2 ml of concentrated sulfuric acid was mixed under stirring with 0.4 gr of 2,4-dnph and 3 ml of water until total dissolution appears. At this point, 10 ml of ethanol at 95% are added to the solution.
The quantitative analysis of formaldehyde after photooxidation of methanol was verified by collecting the condensed reaction product in a container at 4°C and analyzed in a Varian GC 3800 using a 52 CP WAX column (30 m × 0.320 mm).
3. Results and Discussion
SiO2 monoliths with a diameter of 15 mm and thickness of 1 mm approximately were obtained following the methodology described in the previous section. Figure 3 shows such monoliths. It can be observed that they are totally transparent. This is expected to allow an excellent light transmittance and an appropriated use in photocatalytic reactions using UV light.
In total, 8 samples were characterized in order to decide at what conditions the monoliths for the methanol photooxidation should be synthesized. From the 8 monoliths, 7 were coated with TiO2 while the left one was only SiO2. The nomenclature used to name the samples is explained in Table 1.
Figure 4 shows the diffraction patterns with respect to the agitation mechanism and number of immersion-drying-calcination cycles. The obtained crystalline phase is mainly anatase, and this is expected because of the calcination temperature (550°C) [28]. Usually, when the calcination temperature is increased to more than 550°C, the anatase phase is observed to gradually change into a rutile phase with a larger particle size that results unfavorable for photocatalytic degradation reactions [29].
The average crystallite size of samples was estimated using the Scherrer’s equation through the full width at half maximum of the anatase (101) peak (see Table 2). Based on these results, it can be observed that the agitation mechanism has a significant effect on crystallinity since the monoliths coated in the ultrasound bath exhibit the smallest crystalline size (Figure 4(d)) than that of prepared with mechanical stirring (Figure 4(c)) to the same number of cycles (3 cycles) and residence time in the sol (3 hours). The diffraction pattern for commercial Degussa P-25 is included on the top right corner of Figure 4 for reference purposes. It can also be observed that the obtained SiO2 is amorphous (Figure 4(e)). As can be seen in Table 2, the number of treatment cycles decreases crystal growth. In addition, the increase in immersion time for samples with 3 treatment cycles decreased the crystallinity. The sample with the largest crystalline size corresponds to 3 treatment cycles and half an hour of immersion (sample 3ST12P). It is worth mentioning that more than 3 cycles were unsuccessfully attempted since monoliths got broken, with the exception of those with 1 hour immersion time. In this case, monoliths did not stand the fifth cycle.
Regarding samples 3ST1B and 3ST12B, these were discarded because the ultrasound influenced the structure to an extent that the SiO2 monolith was broken. This may be ascribed to the vibrational movements caused by ultrasonic, causing the structure to become weaker. This is a consequence of the immersion under ultrasound presence during TiO2/SiO2 monoliths preparation as well as the decrease in crystallinity.
Figure 5 (3ST12P sample) shows the surface morphology of TiO2 film obtained after 3 cycles of immersion-drying-calcination treatments. The residence time was half an hour for each of the immersion processes. The film exhibits a homogeneous morphology with a lineal growing up with almost totally flat surface. The EDS analysis shows the Ti presence. Figure 6 (3ST1P sample) presents the surface of the TiO2 film obtained after 3 cycles of immersion-heating-calcination treatments and one hour of residence time for each immersion process.
Figure 7 (3ST3P sample) illustrates the characteristic morphology of a TiO2 film after 3 treatment cycles and 3 hours of immersion under mechanical stirring. The comparison of Figures 7 and 5 make it clear that the amount of material deposited on the surface increases with immersion time. This increase is not related to the crystalline growth of anatase on the surface as has been evidenced by XRD analysis. The clusters presented in Figure 7 can be ascribed to the time given to the monolith in the sol where hydrolysis and condensation reactions are occurring, so the longer the immersion time, the larger the agglomerate.
By comparing Figures 5–7, it can be observed that the residence time has a significant effect on the morphology of the coating. A time higher than half an hour (Figures 6 and 7) promotes the appearance of microcracks, and therefore the film loses homogeneity. These microcracks may be ascribed to a stress effect during the drying-calcination treatment due to thermal shrinkage and expansion phenomena. The stress is caused by chemical reactions during the drying and thermal expansion coefficients difference between the support (5 x 10−7°C) and TiO2 film (2.1-2.8 × 10−6°C) [30]. It seems that the effect of this phenomenon becomes stronger when the amount of TiO2 increases due to the drying and calcination process. The crystallinity of the sample shown in Figure 5 may be related to the absence of cracking since with the growth of the crystal and densification of the film, the compressive stresses are reduced.
Figure 8 (4ST1P sample) shows an image of a TiO2 film after 4 treatment cycles and 1 hour of immersion under mechanical stirring. The comparison of the EDS analysis of this with that in Figure 6 confirms that the amount of TiO2 is a direct function of the number of dip coating/heat treatment cycles and that the extent and frequency of the microcracks increase with the number of cycles [31, 32].
The final percentage in weight gained of TiO2 by the SiO2 monoliths is shown in Table 3. It can be observed that the increase in immersion time (samples 3ST12P and 3ST1P) favors the amount of TiO2 deposited on the monoliths. On the other hand, the increase in the number of treatment cycles (samples 3ST1P and 4ST1P) decreases the final percentage of weight gained of TiO2.
Table 4 shows the specific surface area and average pore diameter of synthesized materials. All samples presented type I isotherms and average pore sizes of 18 Å, which according to the IUPAC classification corresponds to materials with microporous texture. It can be seen that the pure SiO2 monolith presented the highest surface area (339 m2/g). As the immersion time of the monoliths increases, the surface area decreases by about 50% (3ST3P sample). This decrease can be attributed to the amount of TiO2 on the SiO2 surface. Although the surface area decreases, the pore size distribution and the mean diameter are maintained.
It can be said that the synthesis conditions in which better crystallinity of the anatase phase is obtained; the highest specific surface area as well as a better uniformity of the formed film are with half an hour immersion, 3 cycles and stirring plate as dispersion mechanism. Therefore, these conditions were used to synthesize 30 monoliths to pack the bed reactor in order to perform the photocatalytic oxidation of methanol.
The minimum air flow rate to carry the methanol gas through the reactor was established as 50 ml/min. Two sets of experiments by triplicate were performed; one set without catalyst and the other one with the catalyst. In the former case, no formaldehyde was detected by the employed analysis method (2,4-dnph), and therefore the production of formaldehyde by photolysis was discarded. In the latter set of experiments, formaldehyde was identified. The analytical technique for the qualitative analysis of formaldehyde with a 2,4-dnph solution was carried out. The formation of micelles as precipitates is due to the formation of 2,4-dinitrophenylhydrazone indicating the presence of formaldehyde during the photocatalytic reaction. A precipitate indicating the presence of the aldehyde was only observed with the material 3ST12P. This does not mean that the other materials did not have photoactivity but that this could be so high that total methanol oxidation rather than selective oxidation was attained. To determine the amount of formaldehyde formed during photooxidation of methanol, the first condensed reaction product was evaluated by gas chromatography (GC). The final concentration of produced formaldehyde corresponds to a value of 457 micromol/L (13.7 mg/L). This result is superior to those obtained with other titania-silica systems [33].
4. Conclusions
SiO2 monoliths coated with thin films of TiO2 anatase phase were successfully prepared by using a dip-coating sol-gel method. The dispersion mechanism, the immersion time, and the number of dip-coating cycles of the SiO2 monoliths into the Ti sol were found to affect both the morphology and crystallinity of the TiO2 deposit. SiO2 monoliths coated with crackle-free TiO2 films were obtained after three dip-coating cycles, with a dip time of 30 minutes. It can also be concluded that mechanical stirring should be preferred over ultrasound dispersion since the former favors the structural stability of the monolith and increases the film crystallinity, while the ultrasound dispersion method leads to monolithic structure breakage and also increases film crackles. Immersion time diminished both TiO2 film and homogeneity. Immersion time and number of cycles also affect the surface area and deposit crystallinity. The surface area of the SiO2-TiO2 materials was decreased when the immersion time increased, which is related to the amount of TiO2 on the SiO2 surface. The highest anatase phase crystallinity and specific surface area were obtained after 3 dip-coating cycles and half an hour of immersion under mechanical stirring. Under these preparation conditions, the attained surface area was 241 m2/g and the crystallite size was 19.4 nm. The weight percentage gained by SiO2 monoliths was 14.3% (TiO2 film).
A formaldehyde concentration of 13.7 mg/L was attained at mild conditions of pressure and temperature in a continuous flow reactor packed with SiO2 monoliths coated with TiO2 anatase films prepared with 3 dip-coating cycles and 0.5 hours of immersion time.
Conflicts of Interest
The authors declare that there is no conflict of interests.
Acknowledgments
The authors are grateful to UAEMex for the financial support through project 4373/2017/CI and to CONACYT (project 269093). R. Regalado would like to thank CONACYT for the financial support and to CCIQS from UAEM for the granted support.