中国矿业大学2012届本科毕业设计（论文） 第 31 页
The exponential statistical scattering becomes more intense with increasing temperature and follows the sequence oxidizing > inert > reducing atmosphere. Accordingly, the radius of gyration and n, describing the shape of the statistically scattering pores and their number respectively, are significantly smaller for the samples treated in the reducing and
inertatmospheres. The most important structural parameter of the porous system is its lattice parameter calculated on the basis of the scattering angle of the main Bragg-reflection. The initial hexagonal lattice parameter a is 21.5 nm for the as-made sample. The development of the lattice parameter in dependence of the temperature of further heat treatment is shown in Figure 4. In this graph measurements of samples are considered which have been treated at 150 and 200 °C. These temperatures are sufficient to remove the ethanol from the samples but not high enough for complete template removal. Additionally, only the samples without subsequent calcination in air are considered. The mesoporous lattice is shrinking independently from the gas atmosphere exponentially with temperature to avalue of 12 nm. Thus, if the atmosphere is oxidizing, inert or reducing has no influence on the lattice shrinkage. Nitrogen adsorption isotherms and the determined pore sizes are shown in Figure 5. All samples show typical reversible type IV
adsorption isotherms. The samples calcined in N2 and H2/N2 atmospheres show similar behavior regarding the total pore volume which is slightly higher than the one of the samples calcined in air. The pore diameters show narrow distribution of the pore sizes around 6 nm. One peculiarity can be observed in the measurements of the samples heat treated at 350 °C, which show a measurement artifact at approximately 3.9 nm (cavitation). The sample treated in the reducing atmosphere shows an additional maximum of the pore diameters at 4.5 nm. The SAXS
measurements (Figure 3) of this sample show an increase in intensity, indicating an increase in the electron density difference from 350 to 450 °C calcination temperature. It has to be noted that these samples underwent the treatment under air to remove the template. That means that in the sample treated in the reducing atmosphere at 350 °C some reaction between the template and the TiO2 has taken place which modifies the pores. The true nature of this effect remains unclear, but it is conceivable that the template induces chemical reactions in the outer surface areas of the particles under the influence of the reducing atmosphere and that these reactions cannot proceed fully into the grains, thus leading to two different pore sizes, an inner and an outer one, after the final calcination step.
Figure 2. Mean crystallite sizes determined as the Lorentzian scattering volume by the integral breadth of the broadening functions in the Rietveld refinements.
中国矿业大学2012届本科毕业设计（论文） 第 32 页
Figure 3. SAXS pattern of the samples treated at different temperatures (black: 350 °C; red: 450 °C; green: 500 °C) with different gas atmospheres (circles:air; triangles up: nitrogen; triangles down: N2/H2) after subsequent calcination at 350 °C in air (graphs shifted vertically for the sake of clarity).
Figure 4. Development of the lattice parameter a of the mesopore’s hexagonal lattice. Here additional samples are considered which have been treated at temperatures 150 and 200 °C which are well below the temperature at which the template is removed but high enough to remove the ethanol of the samples.
中国矿业大学2012届本科毕业设计（论文） 第 33 页
Figure 5. Nitrogen adsorption/desorption isotherms measured at ?196 °C and the pore size distributions calculated from desorption isotherms by using the BJH method.
Figure 6. TEM micrographs (left: TEM bright-field (except g: STEM dark-field), middle: HRTEM, right: selected area electron diffraction (SAED)) of the samples treated at 450 °C (upper row: air, middle row: N2, lower row: N2/H2).
3.3. TEM.
中国矿业大学2012届本科毕业设计（论文） 第 34 页
The samples treated at 450 °C under the different gas atmospheres have been further
examined by transmission electron microscopy. Representative micrographs are shown in Figure 6. All three samples show porosities with an opening diameter of approximately 5 nm, which complies very well with the results obtained by N2-gas adsorption measurements. The porosities show a channel-like parallel ordering over small areas and the pore walls consist out of TiO2 particles with a size of 5 to 10 nm. The pore walls in the sample treated in N2/H2 atmosphere consist out of slightly smaller particles. The HRTEM micrographs and the SAED measurements show only crystalline material; the samples do not contain any amorphous TiO2. So it can be concluded that the crystallization processes, as observed by XRD measurements, are occurring uniformly inthe OMM. The initial points for the crystallization of anatase or rutile are interfaces between crystallites as shown in Figure 7 (red arrow), showing a HRTEM picture from the sample calcined under N2/H2 atmosphere. Such interfaces have been considered by Penn and Banfield33 as energetically favorable sites for the crystallization of new phases.
Figure 7. HRTEM of TiO2 powder calcined at 450 °C (N2/H2 atmosphere) showing an interface between two crystallites (see arrow).
4. SUMMARY AND CONCLUSIONS
Here examinations on the influence of different gas atmospheres and temperatures during the calcination of ordered mesoporous TiO2 nanoparticles made by a facile one-pot synthesis were presented. Regarding the pore-wall material it was shown that crystallization of the TiO2 modifications can be influenced by an additional heat treatment step under N2 or N2/H2 gas atmospheres. By comparison of the temperature dependent phase relationships in the respective samples it is clear that the temperature of the rutile formation can be reduced to 450 °C (inert atmosphere). The growing behavior of the pore wall material is homogeneous over the samples; no evidence for a partitioning of the pore wall material in amorphous and crystalline areas could be found. The phase analysis of the powder diffraction data shows clearly that anatase content remains constant with temperature and is fed by APM and consumed by rutile whenit is formed. These results along with the TEM observations makes it clear that the interfaces between crystallites are the energetically favorable place in these materials for the beginning of phase transition from APM to anatase to rutile. If the anatase-rutile transition temperature is reduced by different gas atmospheres, these interfaces and the overall surface energy effects of the nanosized crystallites are more and more important. In the oxidizing atmosphere the APM
中国矿业大学2012届本科毕业设计（论文） 第 35 页
crystallite boundaries are the initial point for the crystallization of anatase. Due to the fact that the oxidizing conditions inhibit the development of rutile, the anatase can grow further until all the APM material is consumed. The inert N2 and the reducing N2/H2 atmospheres reduce the temperature at which rutile can develop. In these cases the here obtained results of the phase compositions of the samples lead to the following picture of the development of the different phases: At first anatase forms from the APM at its crystallite boundaries. When anatase
crystallites are present, boundaries involving anatase are the initial points for the crystallization of rutile. The observations made for the phase relationships are indicating that without anatase a development of rutile would not take place. The use of N2/H2 or N2 atmospheres can effectively reduce the crystallite size by 2 nm compared to the calcination in air. This leads to a better
ordering of the mesoporous system than in the case of the calcination under air, which is clearly shown by the presented SAXS measurements, where even at temperatures of 500 °C small Bragg intensities can be observed. In the case of N2/H2 the temperature at which the template is completely burned off is increased by at least 50 °C. This effect goes along with two kinds of pores of different size of unclear origin. In the case of the smaller ones it may be a possible
explanation that the template reacts with the TiO2 and that in this case it is not possible to burn it off completely in the following calcination. An increase of the temperature in the heat treatment to 450 °C inhibits this effect. The examinations carried out here showed that the phase relationships in the TiO2-system are complex and can be easily influenced by experimental parameters like the gas atmosphere, with regard to photocatalytical applications, which need a high anatase content and a well ordered pore system, the application of inert or reducing gas atmospheres is able to achieve a good compromise between anatase crystallinity and retained order of the pore system. Additionally it is clear from the results obtained here that the
knowledge of bulk material behavior (that is temperature dependent phase relations) cannot be easily transferred on nanostructured materials. This goes along with studies on the influence of the surface energies in TiO2 nanoparticles (as described in the Introduction) and with the work of Kirsch et al.,31 who describe that the limiting factor for the crystal growth of anatase in nanostructured TiO2 is the temperature and not the time.
ASSOCIATED CONTENT Supporting Information
X-ray diffraction, Raman spectroscopy, and TG/DTA. This material is available free of charge via the Internet at http:// pubs.acs.org.
AUTHOR INFORMATION Corresponding Author
*E-mail: lrobben@uni-bremen.de. Present Address
?University of Bremen, Solid State Chemical Crystallography, Leobener Strasse/NW2, 28359 Bremen, Germany.
Notes
The authors declare no competing financial interest ACKNOWLEDGMENTS
The authors wish to thank C. Hubsch, Institut fur Werkstoffkunde, University of Hanover, for the TG/DTA measurements.
