1. Introduction
Recently, tremendous efforts have been devoted to developing innovative strategies to synthesize nanomaterials with the desired morphologies and properties. Particularly the one-dimensional (1D) structure of TiO
2 nanomaterials exhibits interesting properties compared to other TiO
2 nanoparticles: it has lower carrier recombination rate and higher charge carrier mobility, thanks to the grain boundaries and junctions absence. In fact, the electron diffusion takes place through the junctions between nanoparticles, inducing slower charge transfer by several orders of magnitude [
1]. In addition, it favors light scattering in the photoanode, which increases the light harvesting [
2]. Among the studied morphologies and materials, semi conducting nanostructured materials such as nanowires, nanobelts, and nanotube have received particular attention, due to their use as photoanaode for potential applications in different areas such as photovoltaic [
2], photo catalysis [
3], gas sensing [
4], and water photo-splitting [
5].
Tuning the size and the morphology of materials is becoming a challenging goal in materials science. Over the past few years, various synthesis methods and protocols have been developed to control the semi-conducting nanomaterials morphology, including vapor–liquid–solid (VLS) [
6], solution–liquid–solid (SLS) [
7], template-based synthetic approaches [
8,
9], arc discharge [
10], laser ablation [
11], chemical vapor deposition [
12], microwave [
13,
14], and sol–gel [
15]. Among these synthesis methods, which mostly brought contamination to the synthesis products, the hydrothermal technique has been proven to be a simple and straightforward method using noncomplex apparatus, scalable for large production, with high chemical purity, allowing a large rang of nanomaterial sizes and morphologies [
16,
17]. Furthermore, the morphology of prepared TiO
2 nanomaterials by using hydrothermal method was demonstrated to depend on the concentration of alkaline solution, the synthesis temperature and time, the material precursor used [
17,
18], additives, Pressure, pH, and the reaction medium [
19,
20,
21,
22,
23,
24,
25].
Additionally, the hydrothermal method allows the control of the nanoparticles aggregation [
26]. The most reported strategy to control the morphology of oxide nanomaterials is based on using organic surfactant, which adsorbs on a selected crystallographic plan of growing nucleus, leading to a change of its orientation and growth rate. This results in controlling the morphology of the obtained nanomaterial at the final growth stage [
27,
28]. Additionally, strategies based on aggregation/coalescence of nanomaterials were reported and demonstrated to be efficient in controlling the morphology of the final synthesized powder [
29]. The exfoliation step was also reported to be a crucial step in the formation mechanism of prepared morphologies [
29]. Most of the studies are based on nanomaterials aggregation/coalescence processes, and to the best of our knowledge, very few are based on exfoliation/aggregation/coalescence processes to explain synthesized morphologies. In the case of TiO
2 nanomaterials, there is still a misunderstanding of the mechanism behind the formation of reported morphologies and particularly nanotube, nanobelt, and nanourchin. Some authors claimed that the Na
2Ti
3O
7 nanosheets exfoliation step is the crucial step in the mechanism formation of different morphologies, whereas other authors stated that it is the dissolution of TiO
2 nanoparticles into TiO
6 octahedra, followed by Na
2Ti
3O
7 nucleation and growth, forming a nanosheet in a later stage [
29]. Furthermore, it is well accepted that different polymorphs of TiO
2 nanomaterials are formed by different arrangements of TiO
6 octahedra. In fact, the growth of anatase tetragonal polymorph proceeds through face sharing arrangements of TiO
6 octahedras, whereas the rutile tetragonal phase growth takes place through edge-sharing arrangements. Furthermore, the Brookite phase is obtained by TiO
6 octahedra assembly, sharing their edge and corner; whereas in Ti
2O (B) (bronze) phase, Ti
4+ ion form two distinct geometries with oxygen: octahedron in one case and a square pyramidal in the other. In addition, to homogeneous size and morphology, prepared TiO
2 nanomaterials using hydrothermal method exhibit several characteristics such as high crystallinity, an accurate control of different crystallinity phases from anatase to rutile depending on the synthesis and annealing temperatures, and high specific surface [
30]. It is well accepted that the anatase polymorph possesses a higher band gap energy (3.3 eV) than that of the rutile polymorph (3 eV).
In the present work, different morphologies of TiO2 have been successfully prepared hydrothermally using TiO2 aggregates made of TiO2 nanoparticles as a precursor. The mechanism behind the morphology control of prepared nanomaterials was discussed. It was found that the prepared TiO2 nanomaterials properties were governed by those of TiO2 nanoparticles aggregate. By controlling TiO2 nanoparticles and aggregate sizes, it has been demonstrated that it is possible to control the TiO2 nanobelt sizes.
3. Results and Discussion
Various powders were prepared using the alkali hydrothermal synthesis method and varying synthesis temperatures and reaction times. To prepare these powders, TiO
2 aggregates of spherical shape and different sizes were prepared and used as precursors. The FEG-SEM characterization of precursor powders are shown in
Figure 1, and it can be observed that the sizes of spherical aggregates are ranging from 50 to 200 nm.
The XRD method was used to characterize the crystalline phase of TiO
2 aggregate precursors, and the obtained results are depicted in
Figure 2. Several well-resolved peaks were observed and are all assigned to TiO
2 anatase phase (JCPDS No. 21-1272), which is proof of the high purity of the prepared precursor powders. Additionally, Scherer analysis was used to calculate the average crystallite sizes at the half-maximum width of the intense peak corresponding to (101) crystallographic plane, and were found to be 9.8, 24.7, and 30.4 nm, for the synthesis temperatures of 100, 200, and 220 °C, respectively.
White powders were obtained using TiO
2 aggregate precursors whatever the preparation conditions, and their corresponding morphologies are depicted in
Figure 3. As it can be observed, at the synthesis temperature of 100 °C, the morphology of the prepared powder is nanourchin-like with a stretched sheet-like network (
Figure 3a), whereas at a temperature of 150 °C, the morphology is still nanourchin-like but with a more rolled nanosheet-like network (
Figure 3b). From these experiments, it is clear that the temperature increase favors the nanosheet scrolling. This could be explained by the fact that the crystallization enhanced by the temperature increase tends to induce the microstructure to change into rolled nanosheet structure. In fact, to reduce the surface energy of rolled structure, nanosheets reduce the defects and the distortion energy [
31]. At a higher temperature of about 200 °C, the FEGSEM characterization of prepared white powder is depicted in
Figure 3c,d. It can be observed that TiO
2 powder is of nanobelt-like and nanotube morphologies, with monodisperse size. The insert of
Figure 3d shows a sticking of several distinguishable nanobelts along their axis direction, forming bundles of nanobelts as a building unit. It can also be observed that their thickness is homogeneous and it is of about 10 nm, their diameter is ranging from 50 to 100 nm with length of around 10 mm. In addition, the nanobelt surface is smooth at the magnification scale, and no contamination was observed. As indicated in
Figure 3d, some curved nanobelts were observed, which gives an indication about their high elasticity. From the described experiments, it is clear that the synthesis temperature is an important parameter in the morphology control of TiO
2 nanomaterials.
The crystalline structure and phase of prepared TiO
2 nanobelt, nanotube, and nanourchin-like powders were studied by the X-ray diffraction method. The obtained XRD patterns are presented in
Figure 4, and they show well-resolved peaks in the case of nanourchin and nanotube mophologies attributed to (-511) and (020) crystallographic planes of pure TiO
2(B) phase (JCPDS No. 35-0088) (
Figure 4a–c). In the case of TiO
2 with nanobelt morphology, the observed XRD peaks indicates that the prepared powder is a mixture of anatase (JCPDS 21-1272) and brookite (JCPDS 29-1360) phases (
Figure 4d).
Additionally, among all the peaks, the most intense is the one corresponding to (121) crystallographic plane of brookite. Further details of crystallinity are provided by HRTEM depicted in
Figure 5, clearly well resolved lattice planes are shown, and the insert electron diffraction shows well resolved spots (
Figure 5b). These spots are the signature that the individual nanobelt is a single crystal. The interplanar distance of about 0.88 nm measured from HRTEM image is assigned to (100) crystallographic plane of brookite, indicating that the growth takes place along the (100) crystallographic plane, which is in good agreement with the result from XRD experiments in terms of brookite formation.
Furthermore, the chemical composition of the powder was provided by XPS analysis, and the obtained spectra are depicted in
Figure 6. The XPS survey spectrum in
Figure 6a of TiO
2 aggregates precursor shows intense peaks corresponding to O1s and Ti2p core levels, and the very weak intensity of the peak corresponding to Na1s. However, the XPS survey spectrum corresponding to TiO
2 nanobelt-like and nanourchin-like powders (
Figure 6b) shows intense and well resolved peak, corresponding to the core level of Na1s, which is a signature of the formation of sodium titanate (Na
2Ti
3O
7), in addition to those of O1s and Ti2p. It was reported that Na
2Ti
3O
7 is constituted by corrugated strips of edge-sharing TiO
6 octahedra [
29]. The width of each strips is about three-octahedra, and they are connected through their corner to form stepped layers. Within the sticking layers, sodium cations are located at the positions between the layers.
In
Figure 7a, it is important to note that nanourchin-like nanoparticles show more enrolled nanosheet with more dense structure, as a consequence of the annealing process. The EDS analyses have been performed to determine the chemical composition of TiO
2 nanoparticles, after just synthesis, or after washing and annealing processes. In
Figure 7 the obtained EDS spectra are depicted; it should be noted that, on the EDS spectrum of TiO
2 nanoparticles, after synthesis shows the presence of Na peak
Figure 7b, whereas it is absent in the spectrum after the washing and annealing processes in
Figure 7c. In fact, during the washing processes of Na
2Ti
3O
7 by HCl, Na
+ ions were exchanged by H
+ ions. These results are a clear evidence of the important role played by Na
+ ions in the formation of TiO
2 nanobelts, nanotube, and nanourchin morphologies.
The details of TiO
2 nanobelt and nanotube formation mechanisms are further investigated by using high resolution transmission electron microscopy (HRTEM). The influence of hydrothermal reaction time on the morphology of prepared TiO
2 nanomaterials is studied at 15, 180, and 360 min. At short reaction time of about 15 min, the morphology of prepared powder is mainly stretched nanosheet-like, with some minor rolled sheet. Closer analysis of prepared powder (
Figure 8) shows different stages of the same formation mechanism.
In fact, the observed nucleation stage can be considered as an integrated growth process of nanobelt structure, from aggregates made of nanoparticles of about 20 nm diameter to nanobelt of several micrometers in length. Similar evolution was observed by other authors [
4,
29]. Thus, we may assume that the morphologies shown in
Figure 8 and
Figure 9 represent different stages of the nanobelt growth process.
It can be observed that at the earlier stage (reaction time of 15 min) of the nanobelt growth process, coalesced nanoparticles coexist with nanosheet like particles, indicated by zones in
Figure 8a,c. Nanoparticles were located at the nanotube edges (region 1 and 2 in
Figure 8c), and beside this simple attachment, an alignment of coalesced nanoparticles takes place (region 3 in
Figure 8c). In addition, the nanosheet shows both stretched and rolled structures. The indicated region 4 in
Figure 8c shows the starting process of nanosheet rolling. However, all these steps are a consequence of different nanobelt growth stages, which will evolve in a later stage to a nanotube structure observed in
Figure 9 and nanobelt structure shown in
Figure 10. However, at closer inspection of the nanosheet structure at an earlier stage, with a synthesis time of 15 min, we find that it presents an assembly of nanoparticles, whose sizes range from 5 to 20 nm, as indicated in selected region of
Figure 8a,c. This proves that these nanoparticles and aggregates are the primary building units for the nanosheet formation process. Furthermore, it is well accepted in the literature that the key point for the formation of nanobelt-like structure is the formation of sodium titanate nanobelt intermediate, in which the sodium ion (Na
+) is inserted into space between TiO
6 octahedra layers, balancing their negative charges [
4,
29]. From the present experiments, it can be inferred that the aggregate of TiO
2 nanoparticles split up into nanosheets as a consequence of Na
+ insertion and their rolling in a second stage to form nanotube in an intermediate stage. Typical TEM and HRTEM patterns of TiO
2 nanotube are depicted in
Figure 9, with similar structure of nanotube obtained using TiO
2 nanoparticles in terms of asymmetrical walls. It can be seen that the nanotube exhibits four layers on one side and two layers on the other (
Figure 9e), which indicates that the nanotubes are formed by the scrolling of several layers of nanosheet, as previously observed by other authors. The interplane on both sides is of 0.36 nm, which corresponds to the (010) crystallographic plane, and is the characteristic of monoclinic H
2Ti
3O7. It was reported for the same materials that the nanotube growth takes place along the (010) direction. Additionally, the interlayer distance between rolled nanosheets is about 0.76 nm closer to different reported values [
29].
From XPS and EDX analysis in
Figure 6 and
Figure 7, it is clearly demonstrated that the sodium ions are incorporated in the TiO
2 nanobelt, nanotube, and nanourchin, which suggest that it plays a role in their formation mechanisms. These observations indicate that nanobelts are formed by an orderly sticking of nanosheet and their coalescence in later stage; whereas nanourchins are formed by random assembly of the nanosheets.
The size dependence of the TiO
2 nanobelt on the size of TiO
2 aggregate precursor was demonstrated. Different sizes of TiO
2 aggregate precursors were used to prepare TiO
2 nanobelt, and the obtained results are depicted on
Figure 10. It can be observed that the nanobelt length tends to increase with the increasing of the TiO
2 aggregate precursor size. Additionally, the TiO
2 nanobelt width increases from 50 to 200 nm (
Figure 10), when the TiO
2 aggregate precursor size increases from 50 to 200 nm (
Figure 1). This confirms that TiO
2 nanoparticles play a role in the formation of different observed morphologies. In fact, if we assume that the formation of observed morphologies goes through the TiO
2 dissolution and precipitation, the TiO
2 nanoparticles size will not have any effects on the final nanoparticle size. Additionally, the observation of TiO
2 nanoparticles during the nanotube formation supports the mechanism through which sodium ions (Na
+) induce exfoliation of TiO
2 aggregates by insertion into the space between TiO
6 octahedra layers and their coalescence to form nanosheets at later stage. Furthermore, the present results provide additional arguments to support some reported works in the literature and contradict others [
30,
32], in which it was claimed that during the hydrothermal synthesis process, TiO
2 is dissolved through Ti–O–Ti bonds breaking and formation of sodium titanate nanosheet [
29], which is converted to hydrogen titanate during the washing step and at a later stage to TiO
2 nanobelt after the annealing process.
It can be seen from the XRD results that the nanobelt powder, at different synthesis stages (
Figure 11), shows a changing of crystalline structure. The TiO
2 aggregates precursor is of anatase phase, with tetragonal structure, in which TiO
6 octahedra are sharing their face and get stacked in a one-dimensional zigzag chain. During the synthesis of Na
2Ti
3O
7 nanobelts, a crystalline transition takes place, and TiO
2 anatase phase is transformed into an orthorhombic structure. In fact, the formation of sodium titanate nanobelt intermediate is obtained through the insertion of sodium ion (Na
+) into the space between TiO
6 octahedra layers, inducing the distortion of the initial structure.
From these XRD results obtained at the synthesis temperature of 100 °C, it can be inferred that the anatase TiO
2 aggregate structure changes are a consequence of Na
+ insertion and a strong repulsion between Na
+ ions, which induces a distortion of the anatase crystalline structure. Similar behavior is observed with the insertion of Na+ ion in the case of Na ion batteries charging/discharging cycles [
33]. However, after the washing step with hydrochloric acid solution, the H
2Ti
3O
7 nanobelts are obtained as a consequence of proton exchange processes of sodium trititanate. From
Figure 11, it can be seen that this exchanging of steps and the resulting orthorhombic structure of H
2Ti
3O
7 (JCPDS Card No. 47-0124) are accompanied by some XRD peak modifications, in terms of the intensity enhancement of some peaks, and their decrease for some others [
34,
35]. These modifications indicate the distortion of the initial structure after ion exchanges. Additionally, after the annealing process and the removal of protons, a mixture of anatase (JCPDS 21-1272) and TiO
2-B (JCPDS 35-0088) phases is obtained at the synthesis temperature of 100 °C. The obtained XRD pattern is similar to that obtained for the same mixture by Beuvier et al. [
36]. A phase transition was observed when the morphology changed from nanotube to nanobelt, but with different compositions than those obtained at the synthesis temperature of 200 °C. It was reported by Zhang et al. that the TiO
2 nanoparticle size has a strong impact on the phase transformation during the growth of coalesced nanoparticles [
37]. In addition, the temperature also plays an important role in the phase transformation of TiO
2 nanoparticles [
26]. However, as when the temperature is changed the coalescence and/or growth of TiO
2 nanoparticles take place, both the temperature and the size contribute to the phase transformation and a formation of different phase mixtures depending on the used synthesis temperature 100 and 200 °C. Furthermore, as it can be observed from
Figure 4 and
Figure 10, the peaks corresponding to the anatase phase are of lower intensity, which indicates that both of the latter synthesis temperatures produce a lower proportion of anatase, in agreement with different reported works in the literature [
36]. During the synthesis process at a given temperature, the phase is also changed due to the insertion of different ions, and it is not necessary to dissolve and precipitate TiO
2 octahedra. Furthermore, from these results, it is worth noting that the synthesis temperature plays a crucial role in the phase control of prepared nanobelt powders.