Structural and photophysical properties of hydroxyapatite doped with lanthanide ions Propriedades e fotofísicas da hidroxiapatita dopada com íons lantanídeos

Hydroxyapatite (HAp) [Ca 10-x Ln x (PO 4 ) 6 (OH) 2 ] were obtained through the precipitation method so their photophysical properties could be studied. From the Scanning Electron Microscopy analyzes it was observed that samples doped with lanthanide ions (Ln 3+ ) showed similarity with the morphology of pure HAp, whereas Rietveld-refined X-ray diffractograms showed that the matrix appears as the major phase (≈70%) while (≈30%) of a second phase of tricalcium phosphate (TCP) was due to the different doping levels. Adsorption isotherms obtained by the BET technique indicated that all the synthesized materials are classified as mesoporous. The emission spectra of the materials exhibit bands typical of the Ln 3+ ions incorporated into the matrix structure at 573 nm ( 5 D 0 →F 0 ), 600 nm ( 5 D 0 →F 1 ), 628 nm ( 5 D 0 →F 2 ), and 698 nm ( 5 D 0 → F 4 ) for HAp/Eu, 337-360 nm (5d → 4f) for HAp/Ce and at 425 nm for HAp/Gd.


INTRODUCTION
The mineral apatite, Ca 5 (PO 4 ) 3 X 2 (X=F, OH, Cl), is the most abundant natural phosphate on Earth. Its chemical structure contains many replacements between cations and anions -practically half of the periodic table can be incorporated to its atomic arrangement, which is one of the most interesting properties of apatites [1]. In biological systems, the apatite variety called hydroxyapatite, Ca 10 (PO 4 ) 6 (OH) 2 , consists of the major mineral component of bones and teeth [2]. Because of its high chemical and crystallographic similarity to the bone structure, synthetic HAp is considered a bioactive substance, since it binds strongly to the host bone tissue. In addition, it is a biocompatible, osteoconductive, non-immunogenic and non-toxic material [3]. These characteristics make HAp a suitable biomaterial for medical applications, with it being widely used as bone grafts and highly promising for clinical use [3].
The replaced apatites are generally produced by synthesis; in the case of HAp, different routes are used in their production such as the reaction sol-gel [4,5], microwave irradiation [6,7] and precipitation reaction [8,9], the latter being the most used. The ease of performing ionic exchanges in the HAp structure can be attributed to its broad surface area, besides the existence of a metastable hydrate layer on the surface of the nanocrystals whose ionic bonds are weak [10,11].
Considering the characteristics presented by HAp as a good host material, the introduction of trivalent lanthanide ions (Ln 3+ ), due to its optical properties and great applicability, has aroused the interest of several research groups. ZENG et al. [12] developed a polymeric biosensor doped with Ce 3+ and Tb 3+ ions capable of detecting H 2 O 2 , indicating potential for use in the determination of glucose. CANTARELLI et al. [13] proposed the incorporation of Gd 3+ , Yb 3+ , Er 3+ and Gd 3+ , Yb 3+ , Tm 3+ ions in CaF 2 nanoparticles producing biocompatible nanoprobes for biomedical images, which were efficient for surgical applications. In our research group, various studies involving the spectroscopic study of materials containing lanthanide ions have already been developed, such as SOUZA et al. [14]. The authors carried out the doping of a polymer with Eu (III) and Tb (III) complexes and obtained red and green solid state emitters. LIMA et al. [15] synthesized complexes of lanthanide ions (Eu 3+ , Tb 3+ and Gd 3+ ) with ligands derived from dicarboxylic acids to study the effect of the substituents on the luminescence of lanthanide ions in the Eu 3+ and Tb 3+ complexes.
Thus, the objective of the present work was to evaluate the incorporation of different Ln 3+ ions into an inorganic matrix (HAp) and to study its photophysical properties for the development of a luminescent biomaterial with potential for application in the medical field.

Materials
The following analytical grade chemical reagents were used in the synthesis of

Synthesis of the Hydroxyapatite Matrix
Hydroxyapatite was obtained by the precipitation method [16], with a molar ratio of 1.66 Ca/P, according to the reaction represented by equation (1). Initially, 250 mL of a 0.1 mol.L -1 CaSO 4 .½H 2 O solution was placed under stirring and then a 3.0 mol.L -1 NH 4 OH solution was added until pH = 10. Subsequently, 250 mL of a 0.06 mol.L -1 (NH 4 ) 2 HPO 4 solution was added at a flow rate of 20 mL.min -1 . The pH was maintained around 10 through the addition of 3.0 mol.L -1 NH 4 OH until the end of the synthesis. After the addition of all reagents, the system remained under magnetic stirring for one hour for complete homogenization of the mixture. Then, the system stood still for the sedimentation of the denser component. The material produced was vacuum-filtered and washed with deionized water until neutral pH and then oven-dried at 100 °C for 24 hours. The dried material was triturated and then calcined at 900 °C for 2 hours with a heating rate of 10 °C.min -1 .

Synthesis of the Doped Hydroxyapatite Matrix with Lanthanide ions
The process of synthesis of the HAp doped with lanthanide ions followed the same methodology used in the preparation of pure HAp. For each of the doped materials, HAp/Eu, HAp/Ce and HAp/Gd, 2.0 % of the molar amount of Ca 2+ ions was replaced by the respective Ln 3+ ion in the form of chlorides. The addition was made during the stirring step of calcium sulphate, following the same procedures described in item 2.1.2.

Characterizations
The synthesized materials were characterized by several techniques, such as Scanning Electron Microscopy with X-ray Dispersive Energy Analysis (Hitachi TM1000 and Vega 3 Tescan) used to observe the sample morphology. The structure of the materials was characterized by infrared spectra obtained by FTIR equipment, Model: Spectrum Two, in the range of 4000-400 cm -1 wave numbers. Samples were prepared as KBr pellets using mechanical pressing.
The Brunauer-Emmett-Teller (BET) surface area measurements were performed by the Micrometrics ASAP 2420 surface area analyzer. The photoluminescent properties of the solid-state materials were investigated in a Jobin-Yvon Ramanor U-1000 double-monochromator spectrofluorometer. For excitation, a Jobin-Yvon monochromator model H-10 was used, using a Xe-Hg lamp (150W).

RESULTS AND DISCUSSION
The FTIR spectra for both pure HAp and HAp doped with the different lanthanide ions are shown in Figure  1. As observed, HAp/Eu, HAp/Ce and HAp/Gd have the same spectral profiles. Vibrations of 569 to 609 cm -1 are attributed to O-P-O (٧ 4 ) deformation in PO 4 3or angular deformation O-P-O (٧ 4 ) in HPO 4 2 ; 954 to 956 cm -1 for symmetrical stretching P-O (٧ 1 ) of the PO 4 3group; 1032 to 1100 cm -1 for asymmetric stretching of PO 4 3-(٧ 3 ) or stretching PO 3 (٧ 6 ) in HPO 4 2-; low intensity bands were verified at 1630 to 1633 cm -1 corresponding to the deformation H-O-H indicating the presence of H 2 O; 636 to 643 cm -1 are OHhydroxyl vibrations and 3575 to 3579 cm -1 for hydroxyl OH -(٧ 5 ) stretching, in accordance with the standards verified by YANG et al. [17].
The presence of OHgroups found in the samples indicates that HAp has binding sites suitable for incorporation of drug molecules; therefore, the material presents an important property for its use as a support for drugs [18].  Figure 2 shows the powder XRD pattern obtained for the samples: HAp/Eu, HAp/Gd and HAp/Ce, where it can be seen that HAp is the major phase; however, the incorporation of lanthanide ions into the matrix structure caused the appearance of a peak at 2θ = 31.0º attributed to phosphate tricalcium phase (TCP). It was verified that the appearance of this phase can be explained by the introduction of the Ln 3+ ions in the HAp lattice, which occurs by replacing the Ca 2+ ions in two different crystallographic sites of the unit cell, turning it into Ca 10-x Ln x (PO 4 ) 6 (OH) 2 , [19] as shown in Figure 3. The unit cell of HAp in the hexagonal phase has 10 calcium ions at two different non-equivalent crystallographic sites. Of these, four ions are at site I (Ca1) and another six located at site II (Ca2). The incorporation of the Ln 3+ ions: Eu 3+ , Ce 3+ and Gd 3+ into the HAp structure of can occur through the occupation of both sites [21], Figure 3  The mechanism of charge compensation for the substitution of Ca 2+ by Ln 3+ in the HAp molecule in addition to the heat treatment can result in destabilization of the structure and formation of the TCP phase [22]. The presence of a second phase in the HAp causes an increase in its rate of degradation, and may be an important characteristic for its application in the medical field [23].
In the HAp/Ce diffractogram, a peak observed in 2θ = 28.5º was not compatible with HAp and also did not appear for HAp/Eu and HAp/Gd. From the Ritveld refinement data, it was possible to identify the peak attributed to cerium oxide (CeO 2 ), indicating that Ce 3+ ions were not thoroughly incorporated into the matrix structure. Comparing the ionic rays of Ca 2+ to those of the dopants, Ca 2+ = 0.106 nm, Eu 3+ = 0.098 nm, Ce 3+ = 0.107 nm and Gd 3+ = 0.097 nm [24], it is observed that Eu 3+ and Gd 3+ occupy the vacancies generated by the Ca 2+ ions more easily because their rays are smaller. On the other hand, the ionic ray of Ce 3+ is very close, but higher than that of Ca 2+ , and may limit its incorporation into the HAp lattice. MORAIS et al. [25] identified characteristic peaks for crystalline cerium oxide in the same region for the sample of HAp doped with the ion, confirming the results obtained in the diffractogram of Figure 2-b (iii). The influence of the dopants on the appearance of TCP in the material was proven by the quantification of the phases present for each of the samples. According to Table 01, that the values of HAp/Eu and HAp/Gd are very close to the TCP phase, whereas HAp/Ce presents a considerably lower value, demonstrating the complete nonincorporation of the Ce 3+ ions into the HAp lattice. Micrographs of pure HAp are shown in Figure 4, where particles with round shapes and uniform distribution are observed for the non-calcined material, Figure 4-a. For the calcined HAp, the morphology shows changes as a function of the heat treatment; in this case, the particles start a mass diffusion process, causing their size to increase and a change in their shape due to grain growth [9]. GYORGY et al. [26] and KAMALATHAN et al. [27] observed the densification of HAp powders due to the growth of grains attributed to the diffusion of the particles as a function of the sintering temperature. In Figure 5, micrographs of the samples of HAp doped with Ln 3+ ions are observed. The images of the samples doping with Eu 3+ and Ce 3+ , (Figures 5-a and 5-c) show similarity between their morphologies and that observed for the matrix. However, spherical particle clusters were observed for the HAp/Gd sample, Figure 5-e. Also shown in Figure 5 are the EDS spectra for the various samples. In addition to the identification of the main elements that compose HAp as calcium (Ca) and phosphorus (P), doping is confirmed from the low intensity peaks attributed to Ln 3+ ions: Eu 3+ , Ce 3+ e Gd 3+ incorporated into the HAp structure. The carbon-associated peak is due to the carbon tape used to fix the sample to the sample holder.  Figure 6. According to the IUPAC classification, these curves can be identified as type IV isotherms, which is characterized as a mesoporous material, that is, it has pores with a diameter between 2 and 50 nm [28]. With regard the surface areas BET (S BET ) obtained for all samples, it was verified that the Eu 3+ and Gd 3+ ions caused a great reduction in S BET (HAp/Eu: S BET = 41.44 m 2 .g -1 and HAp/Gd: S BET = 37.95 m 2 .g -1 ) when compared to pure material (S BET = 79.09 m 2 .g -1 ). On the other hand, the Ce 3+ doped matrix (S BET = 74.79 m 2 .g -1 ) showed a value very close to that of pure HAp, corroborating with the XRD data. Based on these data it is proposed that the Eu 3+ and Gd 3+ ions may also be occupying the interstices of the crystals of HAp or influencing the decrease of the unit cell as a function of their smaller rays. In Figure 7, the excitation and the emission spectra for the HAp are shown. The excitation spectrum (Figure 7-a) shows a broad band with two peaks at 310 nm and 350 nm (λ Em = 410 nm). The emission spectrum (λ Ex = 350 nm) shows a broad band with maximum 428 nm (Figure 7-b). In the work of ZHANG et al. [29], an emission with λ max at 428 nm was also obtained for HAp microspheres, in this case, as Ca 2+ ions or PO 4 3groups are not able to generate photoluminescence, the study of the mechanism of the HAp synthesis indicated that the observed emission was attributed to the presence of impurities or defects in the host lattice of the material. Another possibility is that this emission originates from O-Ca charge transfer band. The HAp/Eu excitation spectra were obtained from emission monitoring at 611 nm ( 5 D 0 → 7 F 2 ) and 573 nm ( 5 D 0 → F 0 ), Figure 8-a. Both spectra present thin bands related to intraconfiguration transitions f-f. However, the excitation spectrum with emission monitoring at 573 nm exhibits a wide excitation band between 250 and 380 nm, attributed to a charge transfer transition between Eu 3+ and O 2- [30]. Different excitation wavelengths were used to obtain the emission spectra of the HAp/Eu sample, Figure 8-b. In these, changes in the spectral profile are verified as a function of the excitation wavelength, indicating that it is possible to selectively excite Eu 3+ ions with different symmetry sites, corroborating with the results of XRD and confirming the presence of more than one phase in the material and, consequently, different chemical environments. The emission spectra, with excitation at 335 nm, 522 nm and 565 nm, exhibit spectral profile with bands related to the transitions of the Eu 3+ ion at: 573 nm ( 5 D 0 → F 0 ), 600 nm ( 5 D 0 → F 1 ), 622 and 628 nm ( 5 D 0 → F 2 ). For the spectra obtained with excitation at 394 nm, 464 nm and 533 nm, the most intense emission bands were observed at: 611 nm ( 5 D 0 → F 2 ) and 698 nm ( 5 D 0 → F 4 ), in addition to less intense emissions at 573 nm , 577 nm, 617 nm, 622 nm and 628 nm [21,31]. The presence of Eu 3+ ions in two dis-tinct environments in the host matrix can be explained by the existence of two different sites for the Ca 2+ ions in the crystalline structure of HAp, called CaI and CaII (Figure 3). The CaI site has symmetry (C 3 ) and the CaII site, symmetry (Cs). According to MARTIM et al. [20] Eu 3+ ions preferentially occupy the Ca1 site of HAp due to the better positioning to compensate for loads between the ions; however, the sample calcination process provides thermal energy for the system, inducing diffusion of the Eu 3+ ions from the CaI site to the CaII site from 400 °C. The results agree with the migration mechanism of the ions through the matrix lattice, since the samples were calcined at 900 ºC and the emissions obtained refer to the presence of Eu 3+ ions simultaneously in the two different sites. The characteristic emission lines observed in the spectra show the potential for use of the material in applications linked to its photoluminescence monitoring, such as probe to investigate the degradation process of the material. For HAp/Ce, the excitation and emission spectra are shown in Figure 9. The excitation spectra (λ Em = 360 nm) exhibit three bands with maxima at 268, 295 and 315 nm. The emission spectrum shows two convoluted bands with maxima at 337 nm and 360 nm, Figure 9-b. Both the emission and excitation bands are attributed to interconfiguration transitions 5d → 4f of the cerium ions. In the same way that ions Eu 3+ , Ce 3+ ions replace the two different Ca 2+ sites in the HAp lattice, in agreement with the two emission peaks observed [32]. The spectra for HAp/Gd are shown in Figure 10. The material was excited using the λ max at 343 nm, Figure 10-a, for the acquisition of its emission spectrum shown in Figure 10-b, where it is observed a wide emission band at 425 nm. Since the energy difference between the ground state 8 S 7/2 and the first excited state 6 P 7/2 of the Gd 3+ ion corresponds to the ultraviolet region [33] and the emission band observed for HAp/Gd appears in blue, this emission can be attributed to the matrix. On the other hand, it is known that gadolinium is a paramagnetic element widely used as a contrast agent in magnetic resonance diagnostics, so the incorporation of Gd 3+ ions in HAp adds the image detection feature to the material [34].

CONCLUSION
In this work, the synthesis of hydroxyapatite doped with different lanthanide ions was presented. The results obtained from the characterizations showed the complete incorporation of Eu 3+ and Gd 3+ ions and partial Ce 3+ ions into the crystalline structure of the matrix. The material doped with Eu 3+ ions indicated a great photoluminescent potential, with intense emission peaks characteristic of the europium transitions. The HAp/Ce sample also demonstrated photoluminescence brought by the Ce 3+ ions inserted into the HAp, while for HAp/Gd it was suggested photoluminescence attributed to the matrix. This research demonstrated that all the lanthanide ions studied had properties that allow the generation of images of the doped HAp, as well as the obtaining of a biphasic bioceramic HAp/TCP luminescent, making these materials good candidates for application as biosensors.

ACKNOWLEDGMENT
We acknowledge to Dr. Leonis L. da Luz for helpful in the photoluminescence data aquisition.