E ﬀ ect of Ultrasound Irradiation on the Synthesis of Hydroxyapatite / Titanium Oxide Nanocomposites

: Bioceramic materials, such as hydroxyapatite, Ca 10 (PO 4 ) 6 (OH) 2 , (HAp), can be chemically bound to bone tissue; since they are bioactive and biocompatible. HAp, titanium oxide (TiO 2 ), and hydroxyapatite / titanium oxide (HAp / TiO 2 ) nanocomposite nanoparticles were obtained by ultrasound irradiation assisted by sol-gel and co-precipitation methods at di ﬀ erent time intervals, using Ca(NO 3 ) 2 • 4H 2 O, (NH 4 ) 2 HPO 4 , and TiOSO 4 • xH 2 O as calcium, phosphorus, and titanium sources, respectively. HAp, TiO 2 , and HAp / TiO 2 nanocomposite powders were characterized by X-ray Di ﬀ raction (XRD) and Raman Spectroscopy. The percentages of anatase phase for TiO 2 and of monoclinic and hexagonal phases for HAp were quantiﬁed by Rietveld reﬁnement. Furthermore, sample crystallinity in each material was enhanced by increasing the ultrasound irradiation time. The nanoparticle shape was semi-spherical, agglomerated, and between 17 and 20 nm in size. The agglomeration of particles in the samples was corroborated with a Field Emission Scanning Electron Microscope (FESEM).


Introduction
An example of a natural composite material is bone tissue in vertebrates and humans, whose main components are inorganic (e.g., CaPO 4 , HPO 4 2− , CO 3 2− , Cl − , and metallic ions) and bioinorganic crystal structure, morphology and particle size [26]. On the other hand, ultrasound irradiation is also used to decrease particle size and increase the surface area in TiO 2 nanoparticles. Poiner et al. have indicated that the use of ultrasound irradiation during the synthesis process has an impact on chemical interactions among the reacting species [27,31]. Therefore, ultrasound irradiation was used in this work to obtain HAp/TiO 2 nanocomposites in contrast to other researchers, who have obtained HAp/TiO 2 nanocomposites using other methods [22,36,37].

Hydroxyapatite Preparation
Each chemical reagent used in the experiments was obtained from commercial sources as guaranteed-grade reagents and used without further purification. The synthesis of HAp and TiO 2 was carried out using soft chemistry techniques and co-precipitation and sol-gel methods assisted by ultrasound irradiation to obtain the nanocrystalline materials at time intervals of 15,30,45, and 60 min.
Ca(NO 3 ) 2 •4H 2 O at 99.9% purity (brand Fermont), (NH 4 ) 2 HPO 4 at 99.995% purity (brand Fermont), (NH 4 OH) at 28% purity (brand Aldrich), and Trizma base (NH 2 C(CH 2 OH) 3 (brand Aldrich) were used as reactants for the synthesis of HAp. The solution of Ca(NO 3 ) 2 •4H 2 O 0.04 M was prepared with a Trizma base solution 0.1 M to maintain the pH stable during the co-precipitation method. The hydroxyapatite precursor was obtained by adding 25 mL of Ca(NO 3 ) 2 •4H 2 O solution dropwise into 25 mL of an aqueous solution of (NH 4 ) 2 HPO 4 0.04 M under constant stirring. Immediately after that, small nucleation centers appeared, showing a milky/white solution. The pH value was maintained in a range from 9 to 11 by adding NH 4 OH. Next, the solution obtained was assisted by ultrasound irradiation using the Ultrasonic Processor equipment (Sonics and Materials Inc., Newton, CT, USA) and applying a 20-kHz frequency and a 750 W power with 1 pulsation per 2 s. Time variations of ultrasound irradiation were made at time intervals of 15,30,45, and 60 min. The energy generated after ultrasound irradiation was 10,887; 11,344; 12,667; and 13,835 Joules for each time interval in the reaction system. The resulting homogeneous solutions were filtered, and the products were washed several times with deionized water and ethanol to eliminate salts obtained from the co-precipitation processes. The gel obtained was maintained at room temperature to allow gelification over 24 h. After that, the samples were dried at 150 • C for 6 h to remove water until a xerogel was obtained. Then, the samples were ground in an agate mortar and heated in air in alumina crucibles in a muffle furnace (Prendo model M1) at 550 • C for 6 h.  4 OH dropwise. Subsequently, the CTAB solution 0.05 M was added and stirred vigorously and constantly. The process to synthesize TiO 2 was assisted by ultrasound irradiation at the same time intervals described above. The precipitate was separated from the mother liquor by microfiltration. Then, it was washed with deionized water and ethanol to eliminate any soluble by-products and excessive precursor materials. This gel was maintained at room temperature for 24 h and dried at 150 • C until a xerogel was obtained. The dried gel was ground into fine powder in an agate mortar and heated in air in alumina crucibles in a muffle furnace (Prendo model M1) at 550 • C for 6 h.

HAp/TiO 2 Nanocomposite Preparation
Grigorieva et al. reported that the mechanical activation of mixtures of two or more samples improved the preparation of nanocomposites [38]. The mixture of HAp and TiO 2 was ground in an agate mortar with an agate pestle at room temperature with no solvent in a 1:1 ratio for 30 min.
The nanocomposites obtained were thermally treated at 550 • C for 6 h. A schematic diagram of the synthesis is illustrated in Figure 1. The HAp and TiO 2 nanoparticles and the HAp/TiO 2 nanocomposite nanoparticles prepared were labelled as HAp, TiO 2 , and HT, respectively. A numeral was placed on the label corresponding to the ultrasound irradiation time used for that sample; for example, HAp-15 corresponds to hydroxyapatite with an ultrasound irradiation time of 15 min.
Crystals 2020, 10, x FOR PEER REVIEW 4 of 13 was placed on the label corresponding to the ultrasound irradiation time used for that sample; for example, HAp-15 corresponds to hydroxyapatite with an ultrasound irradiation time of 15 min.

Characterization
To carry out the X-ray diffraction technique, the equipment used was a Panalytical diffractometer with CuKα1 radiation of 1.5406 Å. Data were collected over the 2θ range from 10 to 80° with a step size of 0.00417. The identification phase of the samples was determined by using the Powder Diffraction File PDF-4+2018 from the International Centre of Diffraction Data (ICDD) [39]. The input data for the Rietveld refinement (space group, cell parameters, and atomic positions) were taken from the database format PDF-4+2018. A background was modeled by a polynomial approach, and a pseudo-Voigt function was selected for the profile form. Then, a zero shift; a scale factor; the unit cell parameters; the U, W, V profile coefficients; and the shape parameters were refined. The Micro-Raman Spectroscopy was measured with a LabRAM HR-Olympus Micro Raman System (HORIBA Jobin Yvon Inc., Edison, NJ, USA) using the line 633 nm of He-Ne laser as excitation source. The compositional powders were determined by wavelength-dispersive X-ray fluorescence (WDXRF), using a Bruker S8 Tiger spectrometer. The compositional analysis was carried out using the Quant-Express method and a previous calibration of the equipment. This measurement was performed after preparing pellets of 5 mm in diameter with an applied pressure of 4 ton/cm 2 . A morphological evaluation of the samples was carried out with a Field Emission Scanning Electron Microscopy (FESEM) measurement (JEOL JSM-7800 F).

X-ray Diffraction
The XRD patterns of HAp, TiO2, and HAp/TiO2 nanocomposite nanoparticles were obtained, and the samples were synthesized at different time intervals of ultrasound irradiation at 550 °C. Other works have reported temperatures between 600 and 1500 °C, high pressures and long reaction times [5,40,41]. Two crystalline phases were detected in the HAp samples, shown in Figure 2a: a hexagonal phase (JCPDS 00-009-0432) and a monoclinic phase (JCPDS 01-089-4405). The TiO2 samples were mainly anatase phase (JCPDS 00-021-1272), shown in Figure 2b. The diffraction patterns were analyzed for the brookite phase (JCPDS 00-029-1360), and it was found not to have presence of this phase. On the other hand, the diffraction patterns of nanocomposites showed a mixture of these phases, and a small peak shift was observed towards values smaller than 2θ. Both the hexagonal and monoclinic phases were found for HAp, whereas the anatase phase was found for TiO2, shown in Figure 2c. The results indicated that ultrasound irradiation time induces more interaction among the chemical species in the solution to obtain the expected precursors in a shorter reaction time.

Characterization
To carry out the X-ray diffraction technique, the equipment used was a Panalytical diffractometer with CuKα1 radiation of 1.5406 Å. Data were collected over the 2θ range from 10 to 80 • with a step size of 0.00417. The identification phase of the samples was determined by using the Powder Diffraction File PDF-4+2018 from the International Centre of Diffraction Data (ICDD) [39]. The input data for the Rietveld refinement (space group, cell parameters, and atomic positions) were taken from the database format PDF-4+2018. A background was modeled by a polynomial approach, and a pseudo-Voigt function was selected for the profile form. Then, a zero shift; a scale factor; the unit cell parameters; the U, W, V profile coefficients; and the shape parameters were refined. The Micro-Raman Spectroscopy was measured with a LabRAM HR-Olympus Micro Raman System (HORIBA Jobin Yvon Inc., Edison, NJ, USA) using the line 633 nm of He-Ne laser as excitation source. The compositional powders were determined by wavelength-dispersive X-ray fluorescence (WDXRF), using a Bruker S8 Tiger spectrometer. The compositional analysis was carried out using the Quant-Express method and a previous calibration of the equipment. This measurement was performed after preparing pellets of 5 mm in diameter with an applied pressure of 4 ton/cm 2 . A morphological evaluation of the samples was carried out with a Field Emission Scanning Electron Microscopy (FESEM) measurement (JEOL JSM-7800 F).

X-ray Diffraction
The XRD patterns of HAp, TiO 2 , and HAp/TiO 2 nanocomposite nanoparticles were obtained, and the samples were synthesized at different time intervals of ultrasound irradiation at 550 • C. Other works have reported temperatures between 600 and 1500 • C, high pressures and long reaction times [5,40,41]. Two crystalline phases were detected in the HAp samples, shown in Figure 2a: a hexagonal phase (JCPDS 00-009-0432) and a monoclinic phase (JCPDS 01-089-4405). The TiO 2 samples were mainly anatase phase (JCPDS 00-021-1272), shown in Figure 2b. The diffraction patterns were analyzed for the brookite phase (JCPDS 00-029-1360), and it was found not to have presence of this phase. On the other hand, the diffraction patterns of nanocomposites showed a mixture of these phases, and a small peak shift was observed towards values smaller than 2θ. Both the hexagonal and monoclinic phases were found for HAp, whereas the anatase phase was found for TiO 2 , shown in Figure 2c. The results indicated that ultrasound irradiation time induces more interaction among the chemical species in the solution to obtain the expected precursors in a shorter reaction time. Moreover, sample crystallinity in each sample was enhanced by increasing the ultrasound irradiation time [31,34,42,43].
Crystals 2020, 10, x FOR PEER REVIEW 5 of 13 Moreover, sample crystallinity in each sample was enhanced by increasing the ultrasound irradiation time [31,34,42,43]. Significant changes in the characteristic peaks of HAp, TiO2, and HAp/TiO2 as a function of ultrasound irradiation were observed in the main results from the XRD analysis. Figure 3 shows a short range of diffraction patterns of the samples. The XRD patterns show slight shifts in the peaks, a wider peak size and different relative intensities. The main (hkl) indices are (002), (102), (210), (211), (112), (300), (202), and (301) for HAp, and (101) and (004) for TiO2, but it was difficult to evaluate the main reflections for HAp/TiO2 due to a peak overlapping. A Rietveld refinement was carried out to evaluate and quantify the phases in each sample. For HAp/TiO2, the diffraction peak width may be attributed to the substitution of Ti 4+ by calcium sites in hydroxyapatite as it has been proposed by several authors [12,16,31,44]. Furthermore, Figure 3c displays the presence of reflections related to  (101) and (004) for TiO 2 , but it was difficult to evaluate the main reflections for HAp/TiO 2 due to a peak overlapping. A Rietveld refinement was carried out to evaluate and quantify the phases in each sample. For HAp/TiO 2 , the diffraction peak width may be attributed to the substitution of Ti 4+ by calcium sites in hydroxyapatite as it has been proposed by several authors [12,16,31,44]. Furthermore, Figure 3c displays the presence of reflections related to TiO 2 (001) and (004)   In addition, a Rietveld refinement of cell parameters for HAp, TiO2, and HAp/TiO2 nanocomposites was carried out, considerer the hexagonal phase to HAp and anatase phase to TiO2, which was consistent with the results reported by Zhou [44]. The quantitative analysis resulted in different percentages and phases present in each sample as a function of the ultrasound irradiation time, shown in Table 1. These results agree with Giardina et al. [45], who have reported that the increased crystallinity is related to the time of ultrasonic processing, generating chemical species that favor the reaction speed in the solution [34]. Table 1 shows only the results of the main phases: monoclinic and anatase. From the Rietveld refinement results, small differences can be observed in the cell parameters for the monoclinic and hexagonal phases for HAp, anatase phase for TiO2, However, in the case of nanocomposites, the interaction between HAp and TiO2 was found to have no effect on the crystalline structure of the hydroxyapatite.
The crystallite sizes (coherent scattering domain (CSD) sizes) were estimated with Scherrer's equation considering the full width at half maximum (FWHM) of the prominent reflection (101) for TiO2, while the reflection selected for HAp and HAp/TiO2 nanocomposite was (002) to avoid a peak overlap between 25.3° and 25.8° (2θ). The average sizes were found to change due to an increased nucleation rate that improved the sample crystallinity obtained at the maximum time interval of ultrasound irradiation [46,47], shown in Figure 4. As a result, the diffraction patterns showed a In addition, a Rietveld refinement of cell parameters for HAp, TiO 2 , and HAp/TiO 2 nanocomposites was carried out, considerer the hexagonal phase to HAp and anatase phase to TiO 2 , which was consistent with the results reported by Zhou [44]. The quantitative analysis resulted in different percentages and phases present in each sample as a function of the ultrasound irradiation time, shown in Table 1. These results agree with Giardina et al. [45], who have reported that the increased crystallinity is related to the time of ultrasonic processing, generating chemical species that favor the reaction speed in the solution [34]. Table 1 shows only the results of the main phases: monoclinic and anatase. From the Rietveld refinement results, small differences can be observed in the cell parameters for the monoclinic and hexagonal phases for HAp, anatase phase for TiO 2 , However, in the case of nanocomposites, the interaction between HAp and TiO 2 was found to have no effect on the crystalline structure of the hydroxyapatite.
The crystallite sizes (coherent scattering domain (CSD) sizes) were estimated with Scherrer's equation considering the full width at half maximum (FWHM) of the prominent reflection (101) for TiO 2 , while the reflection selected for HAp and HAp/TiO 2 nanocomposite was (002) to avoid a peak overlap between 25.3 • and 25.8 • (2θ). The average sizes were found to change due to an increased nucleation rate that improved the sample crystallinity obtained at the maximum time interval of ultrasound irradiation [46,47], shown in Figure 4. As a result, the diffraction patterns showed a decrease in the full width at half maximum, where the crystallite size was expected to increase by the ultrasound irradiation time.

Raman Spectroscopy
A Raman spectroscopy was performed on all the samples to confirm the mixture of phases in the HAp/TiO2 nanocomposites. The Raman spectra showed an increase in the intensity of the phononic modes characteristic of each material in relation to the increase in the ultrasound irradiation time, confirming the results obtained by X-ray diffraction. Figure 5a shows an intense band at 142.7 cm −1 (E1g), as well as the characteristic bands at 196 cm −1 (Eg), 396 cm −1 (B1g), 517 cm −1 (A1g) and 639 cm −1 (Eg), which were attributed to the anatase phase of TiO2 [14]. The Raman analysis also confirmed a band at 964.6 cm −1 attributed to the symmetrical stretching mode of phosphate (ν1). On the other hand, the bands situated at 430.2 and 451 cm −1 correspond to the bending mode of the O-P-O bond in phosphate (ν2), while the bands at 582.6 and 595.1 cm −1 were assigned to the asymmetrical bending mode (ν4) [48][49][50]. The bands at 493 and 709.8 cm −1 were not assigned in these spectra because they were attributed to another phosphate group, shown in Figure 5b

Raman Spectroscopy
A Raman spectroscopy was performed on all the samples to confirm the mixture of phases in the HAp/TiO 2 nanocomposites. The Raman spectra showed an increase in the intensity of the phononic modes characteristic of each material in relation to the increase in the ultrasound irradiation time, confirming the results obtained by X-ray diffraction. Figure 5a shows an intense band at 142.7 cm −1 (E 1g ), as well as the characteristic bands at 196 cm −1 (E g ), 396 cm −1 (B 1g ), 517 cm −1 (A 1g ) and 639 cm −1 (E g ), which were attributed to the anatase phase of TiO 2 [14] . The Raman analysis also confirmed a band at 964.6 cm −1 attributed to the symmetrical stretching mode of phosphate (ν 1 ). On the other hand, the bands situated at 430.2 and 451 cm −1 correspond to the bending mode of the O-P-O bond in phosphate (ν 2 ), while the bands at 582.6 and 595.1 cm −1 were assigned to the asymmetrical bending mode (ν 4 ) [48][49][50]. The bands at 493 and 709.8 cm −1 were not assigned in these spectra because they were attributed to another phosphate group, shown in Figure 5b. Raman spectra show the same phononic modes of HAp and TiO 2 for the nanocomposite only with small position shifts (2-4 cm −1 ). Figure

Morphological Analysis
Other researchers have reported that the size and morphology of the hydroxyapatite particles are strongly related to synthetic methods and parameters such as pH, temperature and use of chelating agents and surfactants [52][53][54]. A FESEM micrograph of HAp, shown in Figure 6a, obtained at 60 min of ultrasound irradiation shows a semi-spherical shape with agglomerates and a size between 17 and 20 nm. These results agree with those of Nathanael et al., who state that the synthesis of hydroxyapatite crystals at nanometric scale shows particle agglomeration [55]. Nevertheless, TiO2 also exhibits a spherical shape with an average particle size of 18 nm in diameter and a nanoparticle agglomeration due to aggregation of primary particles, shown in Figure 6b. Lin et al. have reported a study where they indicate that the agglomeration and size distribution are related to fast mixing and synthesis methods, such as the mechanochemical, microwave, chemical precipitation, hydrothermal and sol-gel methods, due to fast nucleation and particle aggregation, which avoid Ostwald ripening [56]. Figure 6c shows the HAp/TiO2 nanocomposite nanoparticles, where the formation of agglomerates with a spherical shape and an average particle size of 20 nm can be observed. The interaction between HAp and TiO2 can favor the growth of spherical particles. An alternative to avoid particle agglomeration is to increase the concentration of surfactants that inhibit it.

Morphological Analysis
Other researchers have reported that the size and morphology of the hydroxyapatite particles are strongly related to synthetic methods and parameters such as pH, temperature and use of chelating agents and surfactants [52][53][54]. A FESEM micrograph of HAp, shown in Figure 6a, obtained at 60 min of ultrasound irradiation shows a semi-spherical shape with agglomerates and a size between 17 and 20 nm. These results agree with those of Nathanael et al., who state that the synthesis of hydroxyapatite crystals at nanometric scale shows particle agglomeration [55]. Nevertheless, TiO 2 also exhibits a spherical shape with an average particle size of 18 nm in diameter and a nanoparticle agglomeration due to aggregation of primary particles, shown in Figure 6b. Lin et al. have reported a study where they indicate that the agglomeration and size distribution are related to fast mixing and synthesis methods, such as the mechanochemical, microwave, chemical precipitation, hydrothermal and sol-gel methods, due to fast nucleation and particle aggregation, which avoid Ostwald ripening [56]. Figure 6c shows the HAp/TiO 2 nanocomposite nanoparticles, where the formation of agglomerates with a spherical shape and an average particle size of 20 nm can be observed. The interaction between HAp and TiO 2 can favor the growth of spherical particles. An alternative to avoid particle agglomeration is to increase the concentration of surfactants that inhibit it.

Compositional Analysis
Quantitative WDXRF data of the samples show the nominal values of the sample composition confirmed by the presence of Ca 2+ , P 5+ , Ti 4+ and other elements occurring at trace levels. Table 2 shows the values for each element and the presence of trace elements that occur due to possible contamination. The Ca/P ratio of hydroxyapatite varied from 1.80, 1.64, 1.88, and 1.76 for each sample. A semi-quantitative analysis presents values higher than the stoichiometric Ca/P ratio of 1.67 in the hydroxyapatite. McConnell et al. reported that this phenomenon may be attributed to volatile compounds formed during the thermal treatment or to the presence of other compounds containing calcium (e.g., Ca(PO4)2) or carbonates (e.g., CaCO3 and carbonate hydroxyapatite). Another reason for the Ca/P ratio to exceed this value may be due to the substitution of tetrahedral hydroxyls or phosphate groups in the hydroxyapatite [57]. Accordingly, the results described above show that a non-stoichiometric hydroxyapatite was obtained.

Compositional Analysis
Quantitative WDXRF data of the samples show the nominal values of the sample composition confirmed by the presence of Ca 2+ , P 5+ , Ti 4+ and other elements occurring at trace levels. Table 2 shows the values for each element and the presence of trace elements that occur due to possible contamination. The Ca/P ratio of hydroxyapatite varied from 1.80, 1.64, 1.88, and 1.76 for each sample. A semi-quantitative analysis presents values higher than the stoichiometric Ca/P ratio of 1.67 in the hydroxyapatite. McConnell et al. reported that this phenomenon may be attributed to volatile compounds formed during the thermal treatment or to the presence of other compounds containing calcium (e.g., Ca(PO 4 ) 2 ) or carbonates (e.g., CaCO 3 and carbonate hydroxyapatite). Another reason for the Ca/P ratio to exceed this value may be due to the substitution of tetrahedral hydroxyls or phosphate groups in the hydroxyapatite [57]. Accordingly, the results described above show that a non-stoichiometric hydroxyapatite was obtained.
The results described above show that the TiO 2 was incorporated into the HAp network as a nanocomposite. It should be emphasized that the time intervals of ultrasound irradiation are directly associated with the crystalline phase percentages obtained. Furthermore, the methodology promotes the formation of nanomaterials in shorter reaction time intervals than other methods, where high temperatures, high pressures and long reaction times are used. However, it is important to adjust the thermal treatment to reach a suitable stoichiometry in future research experiments.

Conclusions
In this work, HAp, TiO 2 , and HAp/TiO 2 nanocomposites were successfully obtained by the sol-gel and co-precipitation methods assisted by ultrasound irradiation at different time intervals of 15, 30, 45, and 60 min at 550 • C, revealing a mixture of phases and an increased crystallinity as a result of the ultrasound irradiation time. The percentages of the crystalline phases were quantified and the effect on the network parameters were observed by Rietveld refinement. All of the materials showed a morphological shape ranging from semi-spherical to spherical particles and an average size between 17 and 20 nm. The agglomeration of particles was caused by fast nucleation in the co-precipitation process. Consequently, it is convenient to encourage the nanoparticle dispersion by increasing the surfactant concentration in order to reduce the aggregation of nanoparticles. The results obtained indicate that sample crystallinity can be controlled and modified by managing the ultrasound irradiation time.