Abstract

BaTiO3 nanoparticles were successfully synthesized by a new coconut water-based sol–gel method using Ba(CH3COO)2 and TiCl3 as the starting salts. The influence of the amount of coconut water and calcination conditions on the barium titanate crystallization was investigated. The resulting nanoparticles were characterized by thermogravimetric and differential thermal analysis, X-ray diffraction, scanning electron microscopy, and Raman and Fourier transform infrared (FTIR) spectroscopies. The ferroelectric tetragonal single phase of BaTiO3 was obtained in samples prepared with a ratio of coconut water volume (mL)/BaTiO3 mass (g) of 25 : 2 and 30 : 2, calcined at 1100°C, which was confirmed by XRD measurements. Crystallites with an average size of about 31 nm for both samples were obtained, and microscopy images revealed the presence of particles in the range of 40 to 60 nm. Raman and FTIR spectra confirmed the dominant tetragonal phase of BaTiO3, meanwhile traces of BaCO3 were identified in FTIR spectra.

1. Introduction

Barium titanate (BaTiO3—BT) is a well-known ferroelectric material widely used in the manufacture of multilayer ceramic capacitors (MLCCs) [1], thermistors [2], and electrooptic devices [3]. BaTiO3 also exhibits piezoelectric, pyroelectric, and catalytic properties, suitable for sonar and infrared detectors [4], oxidation catalysts [5], and photocatalysts [6].

The recent demand for miniaturization of electronic devices requires the dielectric layer thickness to be reduced to a few hundred nanometers. As a consequence, the particle size of the BaTiO3 raw material needs to be less than a few hundred nanometers. To achieve this, an adequate synthesis procedure needs to be employed.

Barium titanate is traditionally prepared by the solid-state mixing of BaCO3 and TiO2. High calcination temperatures and intense ball-milling are required in this procedure to promote slow solid-state diffusion [7]. As an alternative, different wet chemical methods have been employed to produce BaTiO3 nanometer-scale powders including sol–gel, Pechini, sol–emulsion–gel, and solvothermal and coprecipitation methods. These methods allow precursor homogeneity at an atomic level and employ lower calcination temperatures compared with a conventional solid-state method [8]. The main problems in using these wet chemical procedures are the highly toxic and hazardous precursors and solvents, the long duration of the reaction process, and the possibility of contamination by modifying agents (mineralizing and precipitation agents) [8, 9].

At the same time, other innovative routes proposed in the literature for BaTiO3 synthesis frequently use toxic precursors (organometallics) and have multistep reactions. For example, Han and coworkers obtained BaTiO3 in the cubic phase at low temperature using a new LSS (liquid–solid–solution) approach, which involved the preparation of a heterogeneous mixture with three immiscible layers followed by refluxing at 80°C for 2 h. The BaTiO3 nanoparticles were produced in the presence of sodium hydroxide as a catalyst and oleic acid as a capping agent, using an organometallic compound as a Ti source and n-butanol as a solvent [10].

In this way, we have developed a green methodology to synthesize barium titanate nanoparticles via a coconut water-based sol–gel route. This modified sol–gel route is straightforward and has proven to be effective in the synthesis of nanoparticles and thin films [1114]. Due to the high complexity of the coconut water composition, the exact mechanism of sol–gel stabilization and particle formation is still unknown. Some reports state that the large protein chains present in the coconut water can easily bind the metal salts of the solution to form stable micelles [15, 16]. Successful synthesis of gold nanoparticles assisted by a protein mixture [17] and metal oxide nanocrystals using collagen have been recently reported [18], supporting the efficiency of proteins in the production of crystalline nanoparticles. Indeed, metal ions can bind to specific amino acid (aa) ligands of the protein chain depending on the characteristics of the metal ion (e.g., valence state and ionic radius) and the protein chain (e.g., net charge, dipole moment, and polarizability, among others) [19]. On the other hand, sugars may also play a role in this process, since they are the main solute constituents of coconut water [20]. Sucrose and glucose—the more abundant sugars in coconut water—have been used to produce different metal and metal oxide nanoparticles by acting as both chelating agents and fuel during calcination [2123].

The developed method has important advantages over conventional methods, including low toxicity, simplicity of processing, and exemption of the need for organic solvents and sophisticated equipment. Its environmentally friendly character is derived from the fact that it uses nitrates, carbonates, and chlorides instead of organometallic compounds as metal sources and employs coconut water as a natural polymerization agent. Additionally, the coconut itself can be completely consumed along its supply chain, reducing the production of waste. For example, coconut oil (extracted from the dried flesh) and coconut husk and shell have shown great potential in clean energy production, such as biodiesel [24] and biomass fuel [25].

In this study, the influence of the calcination temperature and the amount of coconut water on the crystallization of the BaTiO3 perovskite phase were evaluated. The produced nanoparticles were characterized by means of simultaneous differential thermal analysis (DTA) and thermogravimetry (TG), X-ray diffraction (XRD), field emission gun scanning electron microscopy (FEG-SEM), Fourier transform infrared spectroscopy (FTIR), and Raman spectroscopy.

2. Experimental

For BaTiO3 powder synthesis via the alternative sol–gel route proposed, Ba(CH3COO)2 and TiCl3 (15% solution in HCl) were used as starting materials. Initially, stoichiometric amounts of Ba(CH3COO)2 were dissolved in distilled water. The volume of water was fixed at 5 mL/g of BaTiO3 produced, which was enough to completely dissolve the salt. Subsequently, 3.34 mL of TiCl3 acid solution was added to the Ba solution under magnetic stirring. Filtered coconut water was then added to the mixture, followed by stirring for 30 min. Different samples were produced according to the volume/mass ratio (mL/g) of coconut water (CW) and BaTiO3 (BT) powder produced as follows: , 10 : 2, 15 : 2, 20 : 2, 25 : 2, and 30 : 2. The final solutions were dried at 200°C overnight with continuous magnetic stirring and then preheated at 400°C for 5 h. Finally, the obtained xerogels were calcined at 1000 or 1100°C for 4 h.

DTA/TG measurements of the preheated samples were performed in a simultaneous DTA/TG system (NETZSCH STA 449 F1 Jupiter) at a heating rate of 10°C/min in N2 flow. The powders’ crystalline structure was investigated by XRD using Co Kα radiation (Rigaku RINT 2000/PC), in a 2θ range from 20° to 80°, with a scan speed of 1°/min in continuous mode. The size of the crystallites was estimated from the full width at half maximum (FWHM) of diffraction peaks using the Scherrer equation. Particle size and morphology were determined from scanning electron microscopy (JEOL JSM-6510LV) images. Raman spectra were recorded in a WITec alpha300 equipped with a Nd:YAG laser (2.33 eV, 532 nm) and a Nikon lens (20x; ). The scattered light was detected with a spectrometer equipped with a diffraction grating at 1800 I/mm and a charge-coupled device (CCD). For each spectrum, three accumulations of 30 s were performed, with an incident power on the sample surface of 630 μW. The FTIR spectra were acquired using Bruker VERTEX 70v equipment with a Platinum ATR accessory and evacuated chamber. A total of 128 scans with a resolution of 2 cm−1 were acquired.

3. Results and Discussion

Figure 1 presents the DTA/TG data for a sample preheated at 400°C. The curves can be divided into three stages. The first stage is characterized by endothermic peaks and 15% of mass loss, which are associated with desorption of gas and water molecules from particle surfaces [26, 27]. In the second stage, a large exothermic peak between 400 and 900°C with a very low loss of mass is observed, which can be related to the formation of intermediate phases and solid-phase transition in the sample [28]. The temperature range of 200 to 600°C is a typical region of organic degradation, and the low loss of mass observed in this region indicates a low content of remainder organic matter in the preheated sample. The exothermic peak in the third stage at around 1100°C corresponds to BaTiO3 crystallization, followed by a mass loss of about 7% due to release of HCl, O2, and H2 species. At the end of the reaction, the total weight loss was about 23%.

Figure 2 presents the diffraction patterns for samples prepared with a CW : BT ratio of 25 : 2 thermally treated at (a) 400°C, (b) 1000°C, and (c) 1100°C. For the sample annealed at 400°C (pattern (a)), a mixture of intermediate Ba and Ti phases, such as titanium oxide (PDF number 75-0315), titanium chloride (PDF number 74-0444), and barium chloride (PDF number 74-1971), could be detected. Nonidentified peaks are also present. This mixture of intermediate phases is maintained in the sample treated at 1000°C, with small differences in the peak intensity ratios as can be observed in pattern (b), which corroborates the thermal events that occurred in the second stage of the DTA/TG curves (Figure 1). In pattern (c), the perovskite phase of BaTiO3 (PDF number 79-2265) was obtained after calcination at 1100°C for 4 h in accordance with the DTA/TG results (third stage of Figure 1).

To study the influence of the amount of coconut water (CW : BT ratios) and calcination conditions on the crystallization of the BaTiO3 phase, diffractograms of the powders synthesized with different CW : BT ratios and calcined at 1000°C for 4 h were examined (see Figure 3). Intermediate phases are present in all samples, demonstrating that the thermal treatment employed (1000°C) was insufficient to obtain single-phase BaTiO3 in all samples.

A different behavior was observed for samples calcined at 1100°C. The diffractograms are presented in Figure 4. For the sample prepared with a CW : BT ratio of 20 : 2 (pattern (a)), the BaTiO3 main peaks are of low intensity and appear with various unidentified peaks. The samples that were prepared with a CW : BT ratio of 25 : 2 (pattern (b)) and 30 : 2 (pattern (c)) show that the perovskite BaTiO3 phase was achieved. In both diffraction patterns, an asymmetrical peak at is clearly observed. The splitting or asymmetry of reflection in this region is a result of distortion in the unit cell, characteristic of tetragonal phase BaTiO3.

The crystallite size of these two samples was estimated by a broadening of the diffraction peaks using the Scherrer equation, which is given by

In this equation, is the Scherrer constant, is the wavelength of X-rays ( Å for Co Kα radiation), is the peak position, and is the full width at half maximum (FWHM) of the peak. The crystallite size determined for both samples was averaged at 31 nm.

Some authors had previously produced pure and doped barium titanate using the conventional sol–gel method. Hao et al. [29] used the standard sol–gel procedure to prepare pure and Ag, La-codoped BaTiO3 samples calcined at 1100°C, similar to the temperature used in the present study. The powders were crystallized in the (paraelectric) cubic structure of barium titanate, and considerable amounts of barium carbonate (a common by-product of barium titanate production) were detected by XRD. Cernea et al. [30] produced Ce-doped BaTiO3 also at 1100°C. The sample presented a diffraction pattern consistent with the cubic phase (no peak splitting of the tetragonal phase could be detected) and various nonidentified peaks of low intensity. The cubic BaTiO3 phase has also been obtained at lower calcination temperatures using a conventional sol–gel procedure [31, 32].

Some drawbacks arising from the conventional sol–gel methodology can be seen in all of the above-mentioned studies: they use highly toxic alkoxide precursors and organic solvents as well as multistep procedures involving pH adjustment and reflux and require highly controlled hydrolysis and polycondensation reactions. On the other hand, the modified sol–gel method with coconut water is a one-step procedure that uses metallic salts as the chemical precursor, distilled water as the solvent, and coconut water as the polymerization agent of the departure solution. No toxic organic solvents, surfactants, and chelating or intermediate precursors are necessary, which highlights the low toxicity and eco-friendly character of the proposed alternative sol–gel methodology for large-scale production.

Figure 5 displays FEG-SEM images of a sample prepared with a CW : BT ratio of 25 : 2 annealed at 1100°C for 4 h taken from different regions of the powder dispersion. A predominant rod-like morphology can be observed in both regions with structures that are 200–400 nm in length. The nonequiaxial growth of BaTiO3 particles with a tetragonal crystalline structure has also been recently reported using BaTiO3 powders produced by a hydrothermal method [33], sonochemical synthesis followed by thermal treatment [34], and a solvothermal method using ethylene glycol as a solvent [35]. Rod-shaped particles are particularly advantageous in catalysis applications. The surface-to-volume ratio is higher in nanorods and nanowires compared to nanospheres, which guarantees a higher density of active sites exposed for surface catalytic reactions [36].

The tetragonal polymorph of perovskite BaTiO3 is ferroelectric—it retains spontaneous polarization, which is reversible under an applied electric field [37]. Unlike the cubic structure of BaTiO3, the tetragonal structure is capable of storing a large amount of potential energy due to the distortion of the Ti4+ ions from a centrosymmetric to an asymmetric position within the TiO6 octahedron (as illustrated in Figure 6), resulting in high values of dielectric constant [38]. It is the ferroelectric character of the material that is responsible for the various practical applications of barium titanate ceramics.

Although the identification of crystalline structures by XRD measurements was carried out (probing the global structure of the sample), Raman spectroscopy was also employed as it is a highly sensitive technique, more suited to probing the local structure of a material on an atomic scale based on vibrational symmetry. Figure 7 presents the Raman spectra for 25 : 2 and 30 : 2 (CW : BT ratio) samples, both calcined at 1100°C. Both samples showed similar spectra consistent with those reported in the literature for BaTiO3 crystals produced by a solid-state method [39]. The dominant tetragonal phase is confirmed by the presence of a sharp and distinct peak at 303 cm−1 assigned to E(TO) mode, which indicates asymmetry within the TiO6 octahedron of BaTiO3 on a local scale [40]. The peaks at 198, 246, 512, and 704 cm−1 also indicate the formation of the tetragonal phase [41, 42]. Among these, the peaks at 303 and 703 cm−1 disappear above the Curie temperature, which is the stable region for the cubic phase [43]. This means that it is possible to discern between the cubic and tetragonal phases by the presence of these two peaks in the Raman spectrum. Finally, the asymmetry in the 515 cm−1 peak suggests the presence of coupling TO modes associated with the tetragonal phase [42]. Thus, the Raman experiments support the XRD measurements regarding the structural phases present in both samples.

The FTIR spectra of samples prepared using CW : BT ratios of 25 : 2 and 30 : 2 shown in Figure 8 provide complementary information to the Raman experiments. Both samples have a broad absorption band at 3000–3500 cm−1 (symmetric and asymmetric stretching vibrations of O–H) and double peaks at 1600 and 1632 cm−1 (symmetric stretching of carboxylate groups and bending vibration of H–O–H, resp.) with a higher intensity in the sample prepared with a 30 : 2 CW : BT ratio [44]. The BaTiO3 phase is confirmed by the peak at 510 cm−1, which is related to Ti–O absorption in the barium titanate lattice [45]. A strong absorption at 1450 cm−1 can be observed for the sample prepared with a CW : BT ratio of 25 : 2, which—as well as the ones at 1750 and 850 cm−1—is due to asymmetric stretching of the carbonate ions (CO32−) of BaCO3 [45]. BaCO3 could not be detected by XRD in the 25 : 2 sample. Its presence, however, was evidenced by FTIR—a sensitive technique that can detect traces of impurities such as barium carbonate in very small quantities (~0.6 wt%) [46]. Theoretically, the origin of barium carbonate in BaTiO3 samples produced by sol–gel methodologies is due to the reaction between hydrolyzed barium species and atmospheric CO2 [47]. According to Arvanitidis and coworkers [48], barium carbonate is a stable compound whose decomposition is completed above 1600 K (~1327°C). Hence, its presence in the samples calcined at 1100°C is reasonable.

4. Conclusions

In this work, a green methodology for the preparation of BaTiO3 nanoparticles is suggested. The proposed modified sol–gel method employs filtered coconut water as the polymerization agent of the precursor solution, and it proves to be highly efficient. The samples produced using CW : BT ratios of 25 : 2 and 30 : 2 and calcined at 1100°C for 4 h presented the single crystalline phase of BaTiO3. In the XRD patterns, the tetragonal form of BaTiO3, evidenced by the splitting of the (200)/(002) peak, was detected in both samples. The Raman spectra confirmed the presence of the dominant tetragonal phases. Traces of barium carbonate were detected in the 25 : 2 (CW : BT ratio) sample. The average crystallite size was estimated at 31 nm by the Scherrer equation, while electron micrographs showed samples composed of round particles with diameters of 40–60 nm and elongated structures of 200–400 nm in length. The results demonstrate a potential application in the manufacture of electrooptic devices using the BaTiO3 nanoparticles produced in this work.

Data Availability

The XRD, FTIR, and Raman data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest

There are no conflicts of interest to declare.

Acknowledgments

This work was supported by Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) (Grant nos. 310282/2013-6 and 304419/2015-0), Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), and Fundação de Apoio à Pesquisa e à Inovação Tecnológica do Estado de Sergipe (Fapitec/SE). The authors thank the CMNano-UFS microscopy center for microscopy facilities.