The influence of the amount of hydrogen fluoride (HF) on product formation from the hydrothermal reaction of titanium butoxide and concentrated HF is investigated. Low HF contents lead to a preference for the formation of small TiO2 nano-particles, medium HF contents lead to a preference for TiO2 nanosheets, and high HF contents lead to a preference for large TiOF2 particles. Meanwhile, TiO2 nanosheets display higher activity in photocatalytic hydrogen generation than that of smaller TiO2 nanoparticles ; this demonstrates the higher photocatalytic activity of (001) facets over others. The synergistic effect between TiO2 nanosheets and TiOF2 particles could improve the performance of TiO2 nanosheets owing to possible charge separation over their interface, although TiOF2 particles themselves barely show any activity.
Keywords : hydrogen fluoride · hydrothermal synthesis · nanostructures · photochemistry · titanates
Rapidly depleting fossil-fuel reserves and skyrocketing oil prices have urged the development of sustainable clean- energy solutions to maintain our highly civilized and energy- demanding society. Owing to its activity in splitting water into hydrogen and oxygen under UV-light irradiation, TiO2 has at- tracted substantial interest for renewable energy applications. However, because UV light accounts for less than 5 % of total solar irradiation, TiO2's overall efficiency for solar-driven photo- catalysis is very limited owing to its large band gap (3.0- 3.2 eV) in the UV region. Our recent discovery of black TiO2 nanocrystals by hydrogenation[3a] provides a new approach for improving its optical and photocatalytic activity, apart from tra- ditional metal/nonmetal doping and dye/semiconductor cou- pling approaches.[3b-d] TiO2 nanocrystals treated by hydrogen have displayed superior performance as materials for lithium- ion batteries, supercapacitors, fuel cells, field emission, and microwave absorption, in addition to their dramatic structural, optical, and photocatalytic property changes.
On the other hand, it is known from theoretical calculations and experimental studies that various facets of TiO2 crystals have different thermodynamic and photocatalytic activi- ties.[10-13] For example, (001) facets of anatase TiO2 have been reported to possess higher photocatalytic activities over other facets. The theoretical surface formation energy is 0.44, 0.53, 0.90, and 1.09 J m^2 for relaxed, unreconstructed TiO2 (101), (100), (001), and (110) facets, respectively. The thermodynam- ic stability follows the order (101) >(100)>(001) > (110). Most available anatase TiO2 crystals are dominated by the ther- modynamically stable (101) facets (more than 94 %, according to the Wulff construction), rather than the much more reactive (001) facets. In reality, the (101) and the (100)/(010) planes together with (001) are found on the surface of powdered ana- tase crystals. However, this thermodynamic order breaks down under various conditions. For example, the (100) facet is the most stable for oxygenated anatase TiO2 surfaces and (001) is most energetically preferred for fluorine-terminated surfaces. Fluorine-terminated, (001)-faceted anatase TiO2 single crystals or nanosheets have been obtained by reacting HF with titanium tetrafluoride,[12c] and with the addition of 2- propanol, respectively.[12d] In this reaction, TiOF2 was found to be a precursor formed after 6 h of reaction at 200 8C[13a] for anatase TiO2 nanosheets, which normally need 24 h of reac- tion.[13b] TiOF2 was the only product from the first reaction stage and then it dissolved gradually into the solvent. Subse- quently, anatase TiO2 nanosheets began to appear in the solid precipitates.[13b] Alternatively, fluorine-terminated, (001)-faceted, small TiO2 nanosheets have been made from Ti(OBu)4 and a concentrated solution of HF.[13c] We recently studied the influ- ence of reaction time on the formation of TiO2 nanosheets from this reaction and found that small TiO2 nanoparticles were formed at the very beginning of the reaction, then were transformed into large TiOF2 crystals, and finally turned into (001)-faceted TiO2 nanosheets.[13h] This method seems to be very efficient in generating large quantities of TiO2 nanosheets, and it is thus of great fundamental and practical interest to reveal more information to understand the mechanism of this reaction and the product formed.
Herein, we further investigated the influence of the amount of HF on the final products in a typical hydrothermal reaction with titanium butoxide and concentrated HF towards TiO2 nanosheets. We found that, at very low HF contents, small ana- tase TiO2 nanoparticles were formed, at medium HF contents, TiO2 nanosheets were formed, and at high HF contents, large TiOF2 particles were formed. We also investigated their photo- catalytic activity on photocatalytic hydrogen generation from a solution in water/methanol.
Results and Discussion
Anatase TiO2 nanosheets were prepared from the hydrother- mal reaction of titanium butoxide and concentrated HF (48 wt %) at 200 8C for 24 h. The amount of HF was changed from 0.2 to 2.0 mL, while the amount of titanium butoxide was kept at 0.6 mL. Figure 1 A provides the XRD patterns of the TiO2 samples synthesized with various amounts of HF. First, all samples contained highly crystalline phases based on their strong diffraction peaks. Second, the structures of the final products were clearly influenced by the amount of HF added to the reaction. If the amount of HF added was less than 0.8 mL, only TiO2 was formed (Figure 1 A, curves a-c). If the amount of HF added was between 0.8 and 1.2 mL, a mixture of both TiO2 and TiOF2 was obtained (Figure 1 A, curves d-f). If the amount of HF added was equal to or more than 1.5 mL, pure TiOF2 formed (Figure 1 A, curves g-h). Thus, the amount of HF added to the reaction played an important role in the structure of the final products from the reaction of titanium butoxide and concentrated HF under hydrothermal conditions in this study.
Changes to the average crystalline lengths estimated from the XRD patterns along the ,  , and  directions of TiO2 nanomaterials and along the  and  directions of synthesized TiOF2 are shown in Figure 1 B as a function of the amount of HF. The average size of the primary crystalline do- mains or crystals can be calculated by using the Scherrer for- mula: t = (kl)/(bcos q), in which t is the mean size of the or- dered (crystalline) domains, k is the shape factor with a typical value of 0.9, l is the X-ray wavelength, b is the line broadening full width at half maximum (FWHM) peak height in radians, and q is the Bragg angle. With 0.2 mL of HF, the average length of TiO2 was around 6.6, 5.4, and 7.1 nm in the  ,  and  directions, respectively. Thus, these TiO2 nano- particles were nearly spherical, as determined from the similar sizes in these directions. With 0.4 mL of HF, the average length of TiO2 was around 11.3, 4.5, and 28.1 nm in the ,  , and  directions, respectively. Anatase TiO2 has a tetrahedral structure with the same size in the  and  directions. Thus, these TiO2 nanoparticles were elongated in the  and  directions (28.1 nm) and shortened in the  directions (4.6 nm). TiO2 nanosheets therefore formed with a length of 28.1 nm and a height of 4.6 nm. With 0.6 mL of HF, the average length of TiO2 was around 11.6, 4.5, and 26.4 nm in the  ,  , and  directions, respectively. Similarly, TiO2 nano- sheets were obtained, with a length of 26.4 nm and a height of 4.5 nm. With 0.8 mL of HF, the average length of TiO2 was around 12.4, 4.2, and 26.4 nm in the  ,  , and  di- rections, respectively, whereas the average length of TiOF2 was 27.1 nm along the  direction. Further increasing the amount of HF to 1.2 mL gave increased average lengths of TiO2 along the  and  directions and of TiOF2 along the  direction. If the amount of HF increased to 1.5 mL or above, only TiOF2 was obtained and the size of the average length of TiOF2 along the  direction continued to increase. Based on the above XRD analysis, the influence of the amount of HF studied on the final products of the hydrothermal reac- tion of titanium butoxide and concentrated HF at 200 8C for 24 h can be divided into four regions, as shown in Figure 1 B: region I : 0.2 mL of HF or less for near-spherical TiO2 nanoparti- cles; region II : 0.4-0.6 mL of HF for pure TiO2 nanosheets ; re- gion III : 0.8-1.2 mL of HF for a mixture of TiO2 nanosheets and TiOF2 particles ; and region IV: 1.5-2.0 mL of HF for pure TiOF2 particles.
Typical TEM and SEM images of each of these regions are shown in Figure 2. These observations apparently supported the conclusions drawn from the XRD results. Figure 2 A and B displays the TEM and SEM images, respectively, of the TiO2 nanoparticles synthesized with 0.2 mL of HF. These TiO2 nano- particles were nearly spherical in shape and their size was around 5-10 nm, as seen from the TEM image. The SEM image in Figure 2 B confirmed this observation. Because the particle size of these TiO2 nanocrystals matched well with the crystal- line size deduced from the XRD pattern, these TiO2 nanocrys- tals were thus most likely single crystals with a high crystallini- ty. The TEM and SEM images of TiO2 synthesized with 0.6 mL of HF, as shown in Figure 2 C and D, respectively, clearly dis- played that TiO2 nanosheets formed under these conditions. The thickness of individual TiO2 nanosheets was around 5- 6 nm, which was consistent with the crystalline length along the  direction from XRD analysis. The length of individual TiO2 nanosheets was between 50 and 100 nm, which was 2- 3 times the crystalline length in the  direction from XRD analysis. This suggested that larger TiO2 nanosheets were likely to be formed with a few smaller crystalline grained TiO2 nano- sheets with the orientational attachment of the (100) facets in between. One example of such attachment along the (100) facets is shown in Figure 2 C by the red circle. Meanwhile, larger bundles with larger thicknesses were formed from the stacking of TiO2 nanosheets on the (001) facet.
Figure 2 E and F displays typical TEM and SEM images of the samples obtained with 1.2 mL of HF. The samples were com- monly seen as mixtures of separated TiO2 nanosheets (marked as 1), stacked TiO2 nanosheet bundles (marked as 2), and large TiOF2 particles (marked as 3). The size of the TiO2 nanosheets ranged from 100 to 300 nm in length, probably from the merg- ing of several TiO2 single-crystalline nanosheets because the average crystalline length of these TiO2 nanosheets was around 31.6 nm from XRD analysis. The size of the large TiOF2 particles was around 300-400 nm ; this apparently resulted from the aggregation of smaller nanosheets or TiOF2 crystalline grains (about 40 nm in length from XRD analysis). From the TEM image shown in Figure 2 E, it seemed that stacking of TiO2 nanosheets led to the formation of large TiOF2 particles. Typical TEM and SEM images of samples obtained with 2.0 mL of HF are shown in Figure 2 G and H, respectively. Only large, nearly cuboid TiOF2 particles were formed sizes in the range of 250- 600 nm. Energy-dispersive X-ray spectroscopy (EDX) analysis of the samples made with 0.6, 1.2, and 2.0 mL of HF confirmed the chemical nature of the TiO2 nanosheets and the large TiOF2 particles (see the Supporting Information). In a brief summary, at low HF contents, the reaction of titanium butoxide and con- centrated HF preferred the formation of small TiO2 nanoparti- cles ; at medium HF contents, TiO2 nanosheets were formed by the orientational attachment of small TiO2 nanoparticles along their (100) facets ; and at very high HF contents, large TiOF2 particles were formed by the aggregation of TiO2 nanosheets.
All samples were highly crystallized, based on the strong dif- fraction patterns in the XRD results. This conclusion is also con- firmed from the high-resolution TEM (HRTEM) images shown in Figure 3. Figure 3 A shows the HRTEM image of TiO2 nanoparti- cles synthesized with 0.2 mL of HF. All of the TiO2 nanoparticles were highly crystalline based on well-resolved lattice fringes within the nanoparticles. Each nanoparticle was of one crystal- line domain because the arrangement of lattice fringes was consistent throughout the whole nanoparticle. The size of the nanoparticles was around 5 nm. This observation matched well with the results from XRD analysis. Figure 3 B shows the HRTEM image of the TiO2 nanosheets synthesized with 0.6 mL of HF. The lattice fringes were clearly resolved, which suggest- ed that these TiO2 nanosheets were highly crystalline. Fig- ure 3 C shows the HRTEM image of the edge of one large TiOF2 particle synthesized with 1.2 mL of HF. The polycrystalline nature of the particle was ob- served from several different crystalline domains within the particle. A closer look at the edge of the particle suggested that the particle could be formed from the merging of TiO2 nanosheets because two crystal- line TiO2 nanosheets were seen on the edge of this particle and the thickness was around 5 nm, which matched well with the XRD results and the HRTEM image in Figure 3 B. Meanwhile, some disordered phases were seen on the edges of the particle and between the crystalline do- mains within the particle. Fig- ure 3 D shows the HRTEM image of the corner of one large TiOF2 particle synthesized with 2.0 mL of HF. A well-crystallized lattice and single domain was observed throughout the particle ; howev- er, some disordered phases also existed on the edges of the par- ticle.
Figure 4 A shows the global XPS results for samples obtained in this study. The amount of flu- orine increased in the order of 0.2 (TiO2 nanoparticles) < 0.6 (TiO2 nanosheets) in the F 1 s binding energy indicated shift- ing of the surface Ti^F species on the TiO2 nanoparticles to the Ti^F bonds in the bulk TiOF2 particles. Figure 4 C shows the core-level O 1 s XPS spectra. TiO2 nanoparticles (0.2 mL) showed a typical major O 1 s peak at 530.0 eV from O2^ in the lattice of TiO2 and a small peak at 531.9 eV from the OH groups on the surface.[3a,9b] For TiO2 nanosheets and TiOF2 par- ticles, there was another small peak near 533.3 eV, which matched well with oxygen in the adsorbed water on the surfa- ce.[15e] This might suggest that the high concentration of fluo- rine on the surface could increase water adsorption on the sur- face. Figure 4 D shows the core-level Ti 2p XPS spectra, which suggested Ti4 + ions for all samples.[3a,9b] The binding energies of both Ti 2 p3/2 (from 458.8 to 459.4 eV) and Ti 2 p1/2 (from 464.5 to 465.2 eV) increased slightly from TiO2 nanoparticles to TiOF2 particles. This could be attributed to the increased bond strength of the Ti^F bond in TiOF2 particles compared with the Ti^O bond in TiO2 nanoparticles.
Figure 5 A shows the reflectance spectra of the samples ob- tained in this study. Clearly, all samples displayed a main ab- sorption in the UV region and a long-tail absorption in the visi- ble-light region. The absorption in the UV region was attribut- ed to the electronic transition over the bulk band gap of TiO2 nanoparticles/nanosheets and/or TiOF2 particles, and the ab- sorption in the visible-light region was associated with the transition involved with defects or possible fluorine dopant in the bulk and surface of the nanoparticles. EDX and XPS analy- ses suggested that the TiO2 nanoparticles and nanosheets were likely to be doped with some amounts of fluorine. As de- termined from the midpoint rapid drop in the reflection curves near the UV edge, the band gaps of all samples were similar, with values of around 3.3 eV. This is slightly larger than the bulk value of anatase TiO2 (3.2 eV), possibly owing to size con- finement for the TiO2 nanoparticles and nanosheets, but is likely to be the bulk value for TiOF2 particles. Careful examina- tion revealed that the TiOF2 particles had a slightly larger band gap than TiO2 nanoparticles or nanosheets. The amount of light absorbed in the visible-light region follows the order of 0.2 (TiO2 nanoparticles; Figure 5A a) >0.8 (a mixture of TiO2 nanosheets/TiOF2 particles; Figure 5A d) >1.0 (a mixture of TiO2 nanosheets/TiOF2 particles; Figure 5A e) > 0.4 (TiO2 nanosheets; Figure 5A b), 0.6 (TiO2 nanosheets; Figure 5A c), 1.2 (a mixture of TiO2 nanosheets/TiOF2 particles; Figure 5A f) >2.0 (TiOF2 par- ticles; Figure 5Ah) >1.5 mL (TiOF2 particles; Figure 5Ag). This could suggest that the amount of defects within these samples also follows the same order.
The FTIR spectra of the samples are shown in Figure 5 B. The IR absorption can be divided into three regions : the bands be- tween 2800 to 3700 cm^1, the band around 1645 cm^1, and the bands below 1450 cm^1. The broad bands centered at 3250 cm^1 were due to O^H stretching vibrations of strongly adsorbed H2O on the various facets of the TiO2 nanoparti- cles. The peak at 1645 cm^1 was from the O^H bending of H2O. The band near 1428 cm^1 was likely to be due to Ti^O vibrations on the (001) facets because only samples containing TiO2 nanosheets had this feature. The band at 986 cm^1 was due to Ti^F vibrations in the TiOF2 nanoparticles because this band was only seen in the TiOF2-containing samples. The band centered at 903 cm^1 was attributed to Ti^O vibrations not specifically on the (001) facets in the TiO2 or TiOF2 nano- particles because this band was seen with similar intensities for all samples.
We tested the photocatalytic performance of these nanoma- terials in solar hydrogen generation from water. Figure 6 A shows the time profiles for hydrogen generated over time for these nanomaterials. Figure 6 B shows the averaged hydrogen generation rate within the 6 h tests. Clearly, the observed amount of hydrogen generated increased linearly in the 6 h tests, although the amount of hydrogen generated by the TiOF2 particles was barely seen. The photocatalytic activity of the various TiO2 nanomaterials followed the order of 1.5 (Fig- ure 6A g), 2.0 (TiOF2 particles; Figure 6Ah) < 0.2 (TiO2 nanoparti- cles; Figure 6A a)< 1.2 (TiO2 nanosheets/TiOF2 particles; Fig- ure 6A f) < 0.6 (TiO2 nanosheets; Figure 6Ac), 0.8 (TiO2 nano- sheets/TiOF2 particles; Figure 6Ad) < 1.0 (TiO2 nanosheets/TiOF2 particles; Figure 6A e) < 0.4 mL (TiO2 nanosheets; Figure 6A b).
The surface specific area (SSA) of spherical particles could be estimated according to the formula SSA = 6/1D, in which 1 is the density of the particle and D is the diameter of the parti- cle. Because the size of the TiO2 nanosheets (50-100 nm in length, 5-6 nm in thickness) is much larger than that of the TiO2 nanoparticles (5-7 nm), and the photocatalytic hydrogen generation rate of the TiO2 nanoparticles (Figure 6B a; 0.2 mL) was smaller than that of the TiO2 nanosheets (Figure 6B b, c), this suggested that the (001) facets in TiO2 might have had higher photocatalytic activity than other facets. The higher photocatalytic activity of sample (Figure 6B b; TiO2 nanosheets) over sample (Figure 6B c; TiO2 nanosheets) could be related to the difference in the distribution of the defects within these two samples. Meanwhile, the defects within the samples may also contribute important roles to the photocatalytic per- formance because the TiO2 nanoparticles seemed to have in- volved more defects, although the bulk band gaps of TiO2 nanoparticles and nanosheets were similar. In addition, it seemed that the TiOF2 particles (Figure 6B g and h) had little activity in generating hydrogen, possibly owing to their smaller SSA and mismatched electronic band structures for hydrogen generation from a solution in methanol/water. The higher ac- tivity of mixtures of TiO2 nanosheets and TiOF2 particles (Fig- ure 6B d and e) over that of TiO2 nanosheets could be attribut- ed to synergistic effects between the TiO2 nanosheets and TiOF2 particles ; even mixtures seemed to have more defects than the TiO2 nanosheets. The proposed band alignment is schematically shown in Figure 7. Based on the UV/Vis spectra, the band gap of TiOF2 particles is slightly larger than that of TiO2 nanosheets. The synergistic effect could be attributed to the better charge separation capability in the mixture relative to that of the pure phase, in which excited electrons migrate to one phase and excited holes migrated to the other phase owing to the difference between the conduction and valence band edges of these two phases (TiO2 and TiOF2), which is sim- ilar to those observed in mixtures of anatase/rutile TiO2 com- posites.
We studied the influence of the amount of HF on product for- mation for the hydrothermal reaction of titanium butoxide and concentrated hydrogen fluoride through XRD, SEM, and TEM measurements. We found that the products of this hydrother- mal reaction largely depended on the amount of HF added to the reaction. Low HF contents led to a preference for the for- mation of small TiO2 nanoparticles, medium HF contents led to a preference for the formation of TiO2 nanosheets, and high HF contents led to a preference for the formation of large TiOF2 particles. Such structural transformations from small TiO2 nanoparticles to TiO2 nanosheets and further to TiOF2 particles, was likely to be based on an orientational attachment on the (100) facets of the nanoparticles and stacking on the (001) facets of the nanosheets, respectively, possibly caused by HF in the reaction. Detailed XRD, TEM, and SEM analyses displayed experimental evidence for such a structural transformation mechanism. EDX analysis showed that the TiO2 nanosheets contained a certain amount of fluorine dopants, which could facilitate the structural transformation from TiO2 nanosheets to large TiOF2 particles. We also studied the photocatalytic activity of these samples in photocatalytic hydrogen generation from solutions in methanol/water. We found that TiO2 nanosheets displayed higher activity in photocatalytic hydrogen genera- tion than that of smaller TiO2 nanoparticles ; this demonstrated the higher photocatalytic activity of (001) facets over others. The synergistic effect between TiO2 nanosheets and TiOF2 par- ticles could improve the performance of TiO2 nanosheets owing to the possibility for better charge separation over the interface, although TiOF2 particles barely showed any activity. This study thus may provide us with more information on the fundamental structural evolution of TiO2 nanosheets in the hy- drothermal reaction of titanium butoxide and concentrated HF, and may inspire the new development of better photocatalysts for photocatalytic hydrogen generation towards the realization of sustainable clean-energy solutions in relieving or solving our energy crisis.
Anatase TiO2 nanosheets were prepared as follows : In a typical syn- thetic procedure, Ti(OC4H9)4 (5 mL) and HF (0.6 mL, 48 wt %) were mixed in a Teflon-lined 20 mL autoclave at RT and then kept at cer- tain reaction temperatures for 24 h. The reaction temperature was changed from 100 to 240 8C to study the influence of reaction tem- perature on the formation of TiO2 nanosheets. After the hydrother- mal reaction, the precipitates were collected and washed with de- ionized (DI )water and ethanol several times and finally dried in an oven at 100 8C overnight.
XRD was performed by using a Rigaku Miniflex XRD instrument with CuKa as the X-ray source (wavelength= 1.5418 ^). The TEM study was performed on a FEI Tecnai F200 transmission electron microscope. The electron accelerating voltage was 200 kV. A small amount of sample dispersed in water was dropped onto a thin holey carbon film and dried overnight before TEM measurements. The SEM images were obtained by using an FEI Quanta 200 field- emission scanning electron microscope with an EDX unit. The nanoparticles dispersed in ethanol were dropped onto an alumi- num sample stage and naturally dried. The UV/Vis reflectance spec- tra were collected with an Agilent Cary 60 UV/Vis spectrometer with an optical reflectance fiber unit. MgO powder was used as the reference material. The FTIR spectra were collected by using a Thermo-Nicolet iS10 FTIR spectrometer with an attenuated total reflectance (ATR) unit.
Photocatalytic hydrogen generation test
The photocatalytic activities of the samples were determined by measuring photocatalytic hydrogen generation under simulated solar light irradiation. The solar simulator had a 150 W Xe lamp with an AM 1.5 air mass filter to simulate natural sunlight. Photoca- talysts (20.0 mg) were added to a mixture of DI water (50 mL) and methanol (50 mL). A suitable amount of a solution of H2PtCl6 was added to the solution to make 0.6 wt % of Pt on the photocatalysts under solar irradiation. The solution was purged with pure argon for 30 min in a closed glass reactor. An
X.C. acknowledges support from the
 a) A. Fujishima,
 a) J.-
 a) A. Vittadini, A. Selloni,
 a) U. Diebold, Surface Sci. Rep. 2003, 48, 53 ; b)
Published online on
Yuliang Zhang,[a, b] Mingwei Shang,[c, d] Yifan Mi,[e] Ting Xia,[a] Petra Wallenmeyer,[a] James Murowchick,[f] Lifeng Dong,[c, d] Qiao Zhang,[e] and
E-mail : email@example.com
[f] Prof. Dr.
Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/cplu.201402036.
Hispanic #1 Breaking News for Entrepreneurs, Professionals and Small Business Owners - HispanicBusiness.com
SEPTEMBER 22, 2014
|Customer Service||Policy||Advertise||Site Tools|
|HispanicBusiness Services||Editorial Guidelines||Ad Specifications||Magazine Archive|
|List Rental Agreement||About Us||Media Kits||Post Your Event|
|HireDiversity Job Search||Editorial Calendar||Research|
© 2014 HispanicBusiness Inc. All Rights Reserved.