Synthesis and characterization of new

Solid State Ionics 160 (2003) 183 – 195 www.elsevier.com/locate/ssi

Synthesis and characterization of new oxyboroapatite. Investigation of the ternary system CaO–P2O5 –B2O3 R. Ternane a,*, M.Th. Cohen-Adad b, G. Boulon b, P. Florian c, D. Massiot c, M. Trabelsi-Ayedi a, N. Kbir-Ariguib d a

Laboratoire de Physico-Chimie Mine´rale (LPCM), De´partement de Chimie, Faculte´ des Sciences de Bizerte, 7021 Zarzouna, Bizerte, Tunisia b Laboratoire de Physico-Chimie des Mate´riaux Luminescents, UMR-CNRS 5620, Universite´ Claude Bernard Lyon I, 43 Bd du 11 Novembre 1918, 69622 Villeurbanne cedex, France c C.R.P.H.T-C.N.R.S., 1D, Av. Recherche Scientifique, 45071 Orle´ans cedex 2, France d Laboratoire des Proce´de´s Chimiques, I.N.R.S.T., BP 95, Hammam-Lif, 2050, Tunisia Received 25 September 2002; received in revised form 25 February 2003; accepted 22 March 2003

Abstract The CaO – P2O5 – B2O3 phase diagram was investigated at 1200 jC in order to determine the existence domain of pure apatitic oxyborophosphates with the general formula Cax(PO4)yBzOt. The prepared compounds were chemically analyzed and then characterized by X-ray diffraction, infrared absorption and Raman scattering spectroscopies. 11B and 31P magic angle spinning (MAS) NMR experiments suggest that boron is introduced as 2-fold coordinated boron BO2 in the channels of the apatitic structure and as triangular BO33 groups in the channels and substituting for PO4 groups. The present data show that a continuous solid solution is obtained inside the apatitic defined domain. D 2003 Elsevier Science B.V. All rights reserved. Keywords: Boron; Oxyapatite; Infrared spectroscopy; Raman scattering spectroscopy; MAS nuclear magnetic resonance

1. Introduction The apatite structural family, represented by M10 (AO4)6X2, includes a large class of minerals and synthetic compounds crystallizing in a few different, but related, hexagonal and pseudohexagonal structures. M is most often an alkaline earth ion, the tetrahedral group generally involves the (PO4), (AsO4),

* Corresponding author. Tel.: +216-72-59-19-06; fax: +216-7259-05-66. E-mail address: [email protected] (R. Ternane).

(VO4), (SiO4) or (GeO4) species and X is usually a halide, hydroxide, oxide or sulfide ion. The extensive and diverse chemistry of these materials attests to the remarkable stability of this lattice structure [1,2]. It is known that single crystals with the apatite structure are effective host materials for luminescent [3,4] and laser [5,6] applications. The structural complexity of borophosphate compounds is influenced by localized bonding arrangements of the principal components, B2O3 and P2O5. In the structure of the borate, the boron atoms can form planar BO3 groups with 3-oxygen atoms by trigonal sp2 bonds, but also tetrahedral BO4 groups with 4-

0167-2738/03/$ - see front matter D 2003 Elsevier Science B.V. All rights reserved. doi:10.1016/S0167-2738(03)00161-9

184

R. Ternane et al. / Solid State Ionics 160 (2003) 183–195

Fig. 1. Synthesis of oxyboroapatite showing the phase evolution with temperature (sample a). (a) X-ray diffraction patterns (major lines are labeled as follows: Ca3(BO3)2, n h-Ca3(PO4)2, apatite), (b) Raman spectra.

y


185

Table 1 Stoichiometry of polyphasic compounds in the limit oxyboroapatite – CaO Compound stoichiometry determined by chemical analysis* or by weight losses calculation**

(b) 10.5CaO (63.6%)5.9PO2.5 (35.8%)0.1BO** 1.5 (0.6%) (c) 10.2CaO (63%)5.7PO2.5 (35.2%)0.3BO*1.5 (1.8%) (d) 10.5CaO (63.8%)5.3PO2.5 (32.2%)0.65BO*1.5 (4%) (e) 11.5CaO (65.7%)4.6PO2.5 (26.3%)1.4BO** 1.5 (8%)

Existent phases in the compound Phase

Calculated amounts (%)

Apatite CaO Apatite CaO Apatite CaO Apatite CaO

94.88 5.12 88.88 11.12 86.2 13.8 85.8 14.2

Stoichiometry of pure apatitic phase

10.4CaO (63.4%)5.9PO2.5 (36%)0.1BO1.5 (0.6%) 10.1CaO (62.8%)5.7PO2.5 (35.4%)0.3BO1.5 (1.8%) 10.4CaO (63.6%)5.3PO2.5 (32.4%)0.65BO1.5 (4%) 11.3CaO (65.3%)4.6PO2.5 (26.6%)1.4BO1.5 (8.1%)

Table 2 Stoichiometry of polyphasic compounds in the limit oxyboroapatite – Ca3(PO4)2 Compound stoichiometry determined by chemical analysis


(f) 10CaO (62.5%)6PO2.5 (37.5%)

Apatite h-Ca3(PO4)2 (g) 9.5CaO (61.1%)5.9PO2.5 (38%)0.15BO1.5 (0.9%) Apatite h-Ca3(PO4)2 (h) 9.0CaO (59.8%)5.7PO2.5 (37.9%)0.35BO1.5 (2.3%) Apatite a-Ca3(PO4)2 (i) 8.8CaO (59.5%)5.6PO2.5(37.8%)0.4BO1.5 (2.7%) Apatite a-Ca3(PO4)2


Calculated amounts (%) 86.95 13.05 96.45 3.55 79.94 20.06 48.3 51.7

10.1CaO (62.7%)6PO2.5 (37.3%) 9.5CaO (61.1%)5.9PO2.5 (38%)0.15BO1.5 (0.9%) 9.0CaO (59.8%)5.7PO2.5 (37.9%)0.35BO1.5 (2.3%) 8.76CaO (59.8%)5.5PO2.5 (37.5%)0.4BO1.5 (2.7%)

Table 3 Stoichiometry of polyphasic compounds in the limit oxyboroapatite – Ca2B2O5 Compound stoichiometry determined by weight losses calculation

(j) 8.5CaO (58.2%)3.3PO2.5 (22.6%)2.8BO1.5 (19.2%)



Apatite Ca2B2O5

87.18 12.82


8.6CaO (58.9%)3.4PO2.5 (23.3%)2.6BO1.5 (17.8%)

Table 4 Stoichiometry of polyphasic compounds in the limit oxyboroapatite – Ca3(BO3)2 Compound stoichiometry determined by chemical analysis* or by weight losses calculation**

(k) 8.6CaO (59.3%)4.0PO2.5 (27.6%)1.9BO*1.5 (13.1%) ** (14.4%) 8.6CaO (58.9%)3.9PO2.5 (26.7%)2.1BO1.5 (l) 8.4CaO (58.3%)4.0PO2.5 (27.8%)2.0BO*1.5 (13.9%) ** (12.8%) 9.0CaO (60.4%)4.0PO2.5 (26.8%)1.9BO1.5 (m) 9.1CaO (60.3%)3.8PO2.5 (25.1%)2.2BO*1.5 (14.6%)



Apatite Ca3(BO3)2 Apatite Ca3(BO3)2 Apatite Ca3(BO3)2

89.3 10.7 97.37 2.63 84.46 15.54


8.6CaO 8.5CaO 8.5CaO 9.0CaO 9.1CaO

(58.9%)4.2PO2.5 (58.6%)4.0PO2.5 (58.6%)4.0PO2.5 (60.4%)4.0PO2.5 (60.3%)4.0PO2.5

(28.8%)1.8BO*1.5 (12.3%) ** (13.8%) (27.6%)2.0BO1.5 (27.6%)2.0BO*1.5 (13.8%) ** (12.8%) (26.8%)1.9BO1.5 (26.5%)2.0BO*1.5 (13.2%)

186


Table 5 Stoichiometry of polyphasic compounds in the limit: oxyboroapatite – Ca3(BO3)2 – Ca2B2O5, oxyboroapatite – CaO – Ca3(BO3)2 Compound stoichiometry determined by chemical analysis* or by weight losses calculation**



(n) 7.9CaO (56.9%)2.0PO2.5 (14.4%)4.0BO*1.5 (28.7%) 8.5CaO (58.6%)2.0PO2.5 (13.8%)4.0BO** 1.5 (27.6%) (o) 13.5CaO (69.3%)3.3PO2.5 (16.9%)2.7BO*1.5 (13.8%) 14.4CaO (70.3%)3.4PO2.5 (16.6%)2.7BO** 1.5 (13.1%)

Apatite

57.14

Ca3(BO3)2 Ca2B2O5 Apatite

32.57 10.29 61.09

CaO Ca3(BO3)2

30.36 8.55

oxygen atoms by tetrahedral sp3 bonds [7]. Besides, the borates can contain not only the above two simple groups, but also many complex groups, such as the symmetrical B3O6 boroxol ring, the infinite chain (BO2)n containing only 3-fold coordinated borons [8] and the groups containing interconnected BO3 and BO4 [9]. Among borophosphate compounds, glasses were investigated intensively, and particularly, sodium borophosphate glass-ceramic containing crystalline apatites (calcium fluorapatite [10] or hydroxyapatite [11,12]) in order to enhance the physical and chemical properties such as the microhardness and water resistance for use as biomaterials or ultraviolet edge absorption. This improvement was attributed to the ‘‘cross linkage’’ of phosphate and/or borophosphate groups by the Ca2 + cations. Such ‘‘anomalous’’ behaviour associated to an improvement of the ultraviolet absorption was also observed when sodium tetraborate is added to sodium metaphosphate glasses [13]. Very little is known concerning crystallized boroapatites. A monophase field has been suggested in the ternary system CaO –P2O5 –B2O3 at 1200 jC [14] and at 900 jC [15]. Compounds with stoichiometry Ca9 + yNax(PO4)6Bx + 2yO2 (9 + y + x < 10, x + 2y < 1) have been mentioned [16] and the structure of fluxgrown crystals with composition Sr9 + yNax(PO4)6 Bx + 2yO2 has been reported showing a space group P3¯ [17]. Single crystals with stoichiometry Ca9.5 + 0.5x {(PO4)6 x(BO3)x}{(BO2)1 xOx} have been grown and used for a structure determination. It was proposed an apatitic structure with the space group P3¯ [18,19].


7.9CaO (56.8%)2.3PO2.5 (16.6%)3.7BO*1.5 (26.6%) 8.5CaO (58.6%)2.4PO2.5 (16.6%)3.6BO** 1.5 (24.8%) 13.4CaO (69.1%)3.4PO2.5 (17.5%)2.6 BO*1.5 (13.4%) 14.3CaO (70.4%)3.5PO2.5 (17.3%)2.5BO** 1.5 (12.3%)

In a previous paper [20], we studied the incorporation of boron in hydroxyapatite Ca10(PO4)6(OH)2 at 1000 jC. It was shown that a monophasic solid solution can be obtained up to P/B ratio = 7.22. For higher content, Ca3(BO3)2 and CaO were observed as secondary phases. The infrared and Raman measurements and 11B magic angle spinning (MAS) NMR experiments suggest that boron is introduced as 2-fold coordinated boron BO2 in the channels of the apatitic structure and as triangular BO33 groups substituting PO4 and OH groups leading to a AB-type borohydroxyapatite. The present paper deals with the reinvestigation of the phase diagram CaO – P2O5 –B2O3 at 1200 jC and the structural characterization of oxyboroapatites.

Table 6 Stoichiometry of polyphasic compounds in the limit: Ca3(PO4)2 – liquid, Ca3(PO4)2 – Ca2P2O7 – liquid, Ca3(PO4)2 – Ca2P2O7 Compound stoichiometry determined by chemical analysis



(p) 8.9CaO (59.7%) 6PO2.5 (40.3%)BO1.5

a-Ca3(PO4)2 h-Ca3(PO4)2 liquid a-Ca3(PO4)2 a-Ca2P2O7 liquid h-Ca3(PO4)2 a-Ca3(PO4)2 a-Ca2P2O7

82.65 17.35 – 75.47 24.53 – 51.55 38.65 9.80

(q) 8.1CaO (57.5%) 6PO2.5 (42.5%)BO1.5 (r) 8.3 CaO (58%) 6PO2.5 (42%)BO1.5


Fig. 2. CaO – PO2.5 – BO1.5 phase diagram at 1200 jC. (a) Present work, (b) L: Liquid.

187

.: Present work, n: data of Bauer [14], E: data of Ito et al. [18].

188


2. Experimental procedure 2.1. Powder samples synthesis Powder samples were prepared by the standard solid-state reaction using starting materials CaCO3 (Prolabo, z 99.5%), Ca2P2O7 (Aldrich, 99.9%) and H3BO3 (Prolabo, z 99.5%) with defined proportions. The mixtures were pressed at about 3 T cm 2 into pellets of 2.5 cm in diameter and fired under air first at 400 jC during 12 h, then at 800 jC during 24 h. The pellets were calcined at 1000 jC for 24 h, at 1100 jC for 16 h and finally at 1200 jC for 16 h. The heat treatment was performed in a platinum crucible with intermediate grindings in order to eliminate volatile compounds (H2O and CO2) and to obtain complete reaction and well-crystallized specimens. At the end of the heat treatment, the samples were quenched, the crucible being removed from the furnace and put on a ceramic plate in dry atmosphere. It was verified that the same state was obtained when the sample was cooled in the furnace, excluding the formation of new phase at lower temperature. With not very strong quenching, the cooling rate of samples was higher than the diffusion kinetics in the solid state and the analyses carried out at room temperature approximately reflect the sample state at 1200 jC. The nominal composition was calculated after each step of the heat treatment through the weight losses (especially the B2O3 losses). Some samples were chemically analyzed by plasma emission spectroscopy (ICP-AES) in order to confirm the calculated value.

frequencies of 11B and 31P of 96.3 and 121.5 MHz, respectively. The spinning speed was 15 kHz for all nuclei. The 11B spectra were recorded using a single pulse acquisition scheme with pulse flip angle of 10j, a radio frequency field mRF of 40 kHz and a recycle delay between 20 and 40 s depending on the sample spin lattice relaxation time. Between 2000 and 5000 scans were acquired, due to the rather small amount of sample available (less than 50 mg). Chemical shifts have been reported relative to BF3OEt2 (Et = C2H5) for 11B and to a 0.85 M H3PO4 solution for 31P. Isotropic chemical shift diso, quadrupolar coupling constant CQ and quadrupolar asymmetry parameter gQ have been extracted by simulation using a modified version of the Bruker WINFIT program taking into account spinning side bands of the central transition whenever needed [21]. The Multiple Quantum MQ-MAS experiment has been carried out using a shifted echo pulse sequence [22]. The radio frequency field for excitation and mixing pulses was 100 kHz giving pulse lengths of 3.2 and 1.0 As, respectively. Selective refocusing pulse was 20 As (with mRF of 12.5 kHz) and refocusing delay of one rotor period (i.e. 72 As). The recycle delay was adjusted to 5 s and 480 scans were acquired for each t1 increments, the first dimension evolution being synchronized with the spinning speed (i.e. 14 kHz).

3. Results and discussion The formation of apatite phase was followed by Xray diffraction as a function of the synthesis condi-

2.2. Characterization Phases were identified from their X-ray diffraction patterns using a Philips PW3710 diffractometer and ˚ ). Raman spectra the Cu Ka1 radiation (k = 1.5406 A were recorded at room temperature using a DILOR XY microspectrometer equipped with a CCD detector and a Spectra Physics Argon Laser (excitation at 514.5 nm). Infrared spectra were recorded between 4000 and 350 cm 1 using a Perkin Elmer FTIR Spectrum 1000 device and pellets prepared by mixing together 1 mg of sample and 300 mg of dry KBr. Magic angle spinning (MAS) NMR experiments have been carried out on a Bruker DSX 300 spectrometer operating at principal field of 9.4 T with Larmor

Table 7 Stoichiometry of pure oxyboroapatite compounds Stoichiometry of pure apatitic phase determined by chemical analysis* or by weight losses calculation** (a) 10.3CaO (63.1%)5.6PO2.5 (34.3%)0.42BO*1.5 (2.6%) 9.7CaO (61.8%)5.5PO2.5 (35%)0.5BO1.5 ** (3.2%) (s) 10.0CaO (62.4%)5.6PO2.5 (34.9%)0.43BO1.5 * (2.7%) 10.1CaO (62.8%)5.5PO2.5 (34.1%)0.5BO1.5 ** (3.1%) (t) 9.5CaO (61.3%)5.0PO2.5 (32.25%)1.0BO1.5 * (6.45%) 9.4CaO (61%)5.0PO2.5 (32.5%)1.0BO1.5 ** (6.5%) (u) 9.0CaO (60.4%)4.6PO2.5 (30.9%)1.3BO1.5 * (8.7%) 9.1CaO (60.3%)4.7PO2.5 (31.1%)1.3BO1.5 ** (8.6%)


189

Fig. 3. (a) X-ray diffraction pattern of the monocrystalline fiber. (b) Calculated X-ray diffraction pattern of oxyboroapatite based on the results of Ito et al. [18]. (c) X-ray diffraction patterns of oxyboroapatites for different boron content.

190


tions. Equilibrium was considered to have been reached when the X-ray patterns showed no change with successive heat treatments. Fig. 1a describes the evolution of X-ray patterns for a mixture with the ratio P/B = 13 (sample a). At 1000 jC, characteristic lines of apatitic structure, Ca3(BO3)2 and h-Ca3(PO4)2, were observed. After 16 h at 1200 jC, the reaction was complete and all lines could be indexed in the apatite structure with the P63/m space group. At 1100 jC and then at 1200 jC (Fig. 1b), the disappearance of the Raman band at 3570 cm 1 attributed to the stretching mode vibration (mS) of OH groups confirms the formation of oxyapatite and not hydroxyapatite. Several compounds were prepared with various compositions chosen on the basis of previous studies concerning boroapatites [14,18]. The X-ray diffraction lines were indexed by comparison with literature data. With polyphasic samples, it is not easy to obtain the chemical composition of the phases even for a ternary system. Then we delimited the tentative monophasic field using an iterative method. The phases were identified using X-ray diffraction, and in a first step, the stoichiometry of the apatitic phase was calculated from the relative intensity of the major lines of each phase and from the mass balance of each element. In a second step, a sample having the previous composition was prepared to control its monophasic state with the accuracy given by the Xray measurements. If the sample was not exactly monophasic, the composition of the apatitic phase was calculated again and the stoichiometry controlled as previously. Furthermore, pure samples, chosen in different directions, have been prepared in order to confirm the limits of the domain determined through the previous calculation. Our results were also compared with literature data. They show the existence of four limits where the oxyboroapatite is in equilibrium with CaO, Ca3(PO4)2, Ca2B2O5 and Ca3(BO3)2, respectively. There is a domain where a liquid phase is observed. The global stoichiometry of the prepared samples and the composition of pure apatitic phases are reported in Tables 1 –6. The oxyboroapatite domain in the CaO –P2O5 – B2O3 ternary system at 1200 jC is presented in Fig. 2a. It is based on the b, c, d, e, o, k, l, m, n, j, i, h, g and f compounds as defined in Tables 1 – 6. Inside this

domain, the X-ray diffraction patterns shows that all samples (denoted a, s, t and u in Fig. 2a) are single apatitic phases and their compositions are given in Table 7. As observed in Fig. 2b, this defined domain includes the results of Ito et al. [18] but is more extended than the one proposed by Bauer [14]. A monocrystalline fiber was also elaborated by the LHPG (Laser Heated Pedestal Growth) technique. This method was described in detail elsewhere [23].The fiber composition agrees with the limit proposed for the monophase domain (Fig. 2b). 3.1. Structural considerations The apatite structure mainly belongs to the P63/m space group but the space group P3¯ was proposed for some oxyboroapatites such as Sr9.402Na0.209(PO4)6 B0.996O2 [17] and Ca9.64(P5.73B0.27O24)(BO2)0.73 [18]. The X-ray diffraction pattern of the monocrystalline fiber elaborated in this work and the calculated pattern from the results of Ito et al. [18] are given in Fig. 3a and b, respectively. In our work, the lack of the (00S ) reflexions with S odd (expected in the P3¯ space ˚ group), especially the (001) reflexion at 6.905 A (2h = 12.81j) (Fig. 3), allows to consider that the monocrystalline fiber grew in the hexagonal symmetry with the P63/m as space group.

Table 8 Infrared and Raman band wave numbers (cm for oxyboroapatite (sample a)

1

) and assignments


Fig. 4. Vibrational spectra of oxyboroapatite (sample a). (a) Infrared spectrum, (b) Raman spectrum.

191

192


In the whole monophase domain, the powder X-ray patterns can be indexed in the same P63/m space group. When the boron content increases, the splitting between the two major lines (211) and (112) decreases. At high boron content, these two lines overlap as it is seen for the monocrystalline fiber (Fig. 3c). 3.2. Vibrational infrared and Raman spectra The assignments for infrared (IR) and Raman spectra of a pure oxyboroapatite (sample a) (Table 8) have been done using literature data concerning apatites [2], oxyboroapatites [16,19], hydroxyboroapatites [20], some borates such as Ca3(BO3)2 [24] formed in samples when the boron content increases and rare-earth borophosphates [25]. The IR spectrum (Fig. 4a) is dominated by four strong lines situated at 1040 and 1090 cm 1 and at 571 and 603 cm 1 and assigned to the antisymmetric stretching m3 and the antisymmetric bending m4 modes of the PO43 ions, respectively. One can note the presence of the bands assigned to both BO33 groups and linear BO2

Fig. 5. line).

11

groups. The bands at 1304, 1253, 1208 cm 1 and 784, 771, 755 cm 1 are attributed to the antisymmetric stretching m3 and the symmetric bending m2 modes of the BO33 groups, respectively. The pair of weak peaks at 2002 and 1932 cm 1 can be attributed to the 10 B – O and 11B – O antisymmetric stretching m3 mode of the linear BO2 groups, respectively [16,19,26]. No any characteristic band of HPO 42 groups was observed in the range of 865 cm 1. In the Raman spectrum (Fig. 4b), the strongest line situated at 962 cm 1 is attributed to the symmetric stretching m1 mode of PO43 ions, while the weak line at 912 cm 1 is assigned to the symmetric stretching m1 mode of the BO33 groups. The absence of band at 850 cm 1 corresponding to the O –O vibration evidences that the synthesized samples are oxyapatites with O2 ions and not peroxyapatites [27]. One can note the lack of the characteristic bands of OH groups located at 630 cm 1 (mL: librational mode) in the IR spectrum and at 3570 cm 1 (mS: stretching mode) in both Raman and IR spectra. These results confirm the purity of the investigated samples in the precision limits of the measurements.

B MQ-MAS NMR spectrum of oxyboroapatite (sample s): experimental spectrum (continuous line) along with simulations (dashed


193

3.3. NMR measurements Two samples (a, s) belonging to the oxyboroapatite monophase domain and the monocrystalline fiber were investigated. Their respective compositions are: (a) 10.3CaO5.6PO2.50.42BO1.5 (s) 10.0CaO5.6 PO2.50.43BO1.5 (fiber) 9.0CaO4.0 PO2.51.9BO1.5 3.3.1. 11B NMR The line broadening generated by second-order quadrupolar effects in the NMR spectra of quadrupolar nuclei introduces serious limitations in the resolution of chemically inequivalent sites. These secondorder anisotropies can be removed from the powder spectra by combining magic angle spinning (MAS) and bidimensional Multiple Quantum (MQ) spectroscopy as shown in Fig. 5 for the sample (s). This experiment shows clearly two distinct boron environments, despite the low signal-to-noise due to the small amount of sample available and the long relaxation time. The following parameters can be extracted from the one-dimensional MAS slices taken at the peak maxima: diso(1) f23.1 ppm, CQ(1) f2.69 MHz, gQ(1) f0.0 and diso(2) f22.8 ppm, CQ(2) f2.59 MHz, gQ(2) f0.0. These parameters have been used as starting point for the simulation of the one-pulse MAS NMR spectra of this sample obtained at two different principal magnetic fields. It was found to be necessary to add a third component to obtain the simulations shown in Fig. 6. The results of the simulation of the three experimental spectra corresponding to the samples (a, s and fiber) (Fig. 6) are presented in Table 9. The three sites evident for sample (s) exist for the two other samples. It should be noted that the site labelled 1 presents a quadrupolar coupling constant significantly more important which could explain why it has not been Table 9 Interaction parameters of the distinct

Fig. 6. 11B MAS NMR spectra of oxyboroapatites: experimental spectra (continuous line) along with simulations (dashed line). (a) Sample a, (b) sample s, (c) fiber.

11

B sites in oxyboroapatites

Site diso no. (ppm) F 0.2

CQ (MHz) F 0.02

gq F 0.05

(fiber) F3 (%)

1 2 3

3.53 2.75 2.61

0.00 0.05 0.00

3 (3) 18 66 (68) 53 31 (29) 29

15.0 21.5 23.7

(Sample s) F3 (%)

(Sample a) F3 (%) 33 39 28

194


observed in the MQ-MAS experience. From the shape of the peaks and from the values of diso and CQ, the two signals (nos. 2 and 3) can be attributed to triangular and regular symmetry BO3 units. The site no. 1 exhibits an isotropic chemical shift lower than the one of BO3 groups and higher than the one of BO4 groups, a more important quadrupolar coupling constant than the one typically observed for these two units and an asymmetry parameter (gQ = 0.0) suggesting a cylindrical symmetry environment. Using ab initio calculations, we assumed [20] that the site no. 1 is attributed to a 2-fold coordinated boron. It results that borate groups substitution would occur as linear BO2 groups in the channels (site no. 1) and as BO33 groups on PO4 and channels sites leading to a AB-type apatite (sites nos. 2 and 3). 3.3.2. 31P MAS NMR In free boron apatite, only one 31P site at diso = 2.8 ppm is observed, as expected from literature data [28]. In oxyboroapatites, the spectra of the samples are reported in Fig. 7 and the results of simulations in Table 10. All the spectra are dominated by a major signal (noted no. 2) centered around diso = 2.8 ppm for the samples a and s and around diso = 2.4 ppm for the fiber, which can be identified with the P environment in the free boron apatite. The influence of the boron introduction on the phosphorus seems to be evidenced by the existence of a new site localized at around diso = 4.2– 5.4 ppm (noted no. 1 in Table 10). According to Ducel et al. [12], this chemical shift range (3.5 – 5.0 ppm) could agree with an assignment to BOPO32 groups. The broadening of the peaks and the increase of the full width at half maximum Dm1/2 with boron content

Table 10 Results of simulation of 31P MAS NMR spectra in oxyboroapatites

Fiber Sample s

Fig. 7. 31P MAS NMR spectra of oxyboroapatites: experimental spectra (continuous line) along with simulations (dashed line). (a) Sample a, (b) sample s, (c) fiber.

Sample a

Site

diso (ppm) F 0.2

Dm1/2 (Hz) F 20

%F5

1 2 1 2 3 1 2 3

5.0 2.4 5.4 2.8 0.7 4.2 2.8 0.9

420 360 271 262 188 391 234 174

55 45 13 84 3 17 79 4


increasing traduce an increase of the disorder induced by the cell deformation. 4. Conclusion The existence domain of single-phase oxyboroapatites Cax(PO4)yBzOt was defined in the CaO – B2O3 – P2O5 ternary system. Vibrational infrared and Raman spectra are given and all lines have been assigned. 11B and 31P MAS NMR investigations have been used for better understanding of the structure. The shape of peaks and the values of diso and CQ suggest that borate groups are introduced as linear BO2 and as triangular BO33 units with regular symmetry. As two kinds of orthoborate sites are observed, it can be assumed that these groups substitute for both PO4 and channels sites while metaborate groups would be only located in channels along the c-axis.

Acknowledgements The authors are grateful to Professor A. Durif (L.E.D.S.S., University Joseph Fourier, Grenoble, France) for fruitful discussions.

References [1] D. McConnell, Apatite. Its Crystal Chemistry, Mineralogy, Utilization and Geologic and Biologic Occurrence, SpringerVerlag, Wien, 1973. [2] J.C. Elliot, Structure and Chemistry of the Apatites and Other Calcium Orthophosphates, Elsevier, Amsterdam, 1994. [3] R.G. Pappalardo, J. Walsh, R.B. Hunt, J. Electrochem. Soc. 130 (1983) 2087. [4] K.H. Butler, Fluorescent Lamp Phosphors, University Park, PA, University Press, Pennsylvania State, 1986. [5] J.P. Budin, J.C. Michel, F. Auzel, J. Appl. Phys. 50 (1979) 641.

195

[6] S.A. Payne, L.D. Deloach, L.K. Smith, W.L. Kway, J.B. Tassano, W.F. Krupke, B.H.T. Chai, G. Loutts, J. Appl. Phys. 76 (1) (1994) 497. [7] A.F. Wells, Structural Inorganic Chemistry, 4th ed., Oxford University Press, Oxford, 1975. [8] A. Rulmont, M. Almon, Spectrochim. Acta, Part A 45 (5) (1989) 603. [9] J.S. Ysker, R. Bohlhoff, H. Bambauer, W. Hoffmann, Z. Kristallogr. 132 (1970) 457. [10] T. Kitsugi, T. Yamamuro, T. Nakamura, S. Yoshii, T. Kokubo, Biomaterials 13 (6) (1992) 393. [11] J.F. Ducel, J.J. Videau, Mater. Lett. 18 (1993) 69. [12] J.F. Ducel, J.J. Videau, K.S. Suh, J. Senegas, Phys. Status Solidi, A 144 (1994) 23. [13] J.F. Ducel, J.J. Videau, Mater. Lett. 13 (1992) 271. [14] H. Bauer, Bull. Soc. Chim. Fr. (1968) 1703. [15] P. Ramamoorthy, T.J. Rockett, J. Am. Ceram. Soc. 57 (1974) 501. [16] C. Calvo, R. Faggiani, J.C.S. Chem. Comm. (1974) 714. [17] C. Calvo, R. Faggiani, N. Krishnamachari, Acta Crystallogr., B 31 (1975) 188. [18] A. Ito, H. Aoki, M. Akao, N. Miura, R. Otsuka, S. Tsutsumi, J. Ceram. Soc. Jpn. 96 (3) (1988) 305. [19] A. Ito, H. Aoki, M. Akao, N. Miura, R. Otsuka, S. Tsutsumi, J. Ceram. Soc. Jpn. 96 (6) (1988) 707. [20] R. Ternane, M.Th. Cohen-Adad, G. Panczer, C. Goutaudier, N.K. Ariguib, M.T. Ayedi, D. Massiot, P. Florian, J. Alloys Compd. 333 (1 – 2) (2002) 62. [21] D. Massiot, F. Fayon, M. Capron, I. King, S. Le Calve´, B. Alonso, J. Durand, B. Bujoli, Z. Gan, G. Hoatson, Magn. Reson. Chem. 40 (2002) 70. [22] D. Massiot, B. Touzo, D. Trumeau, J.P. Coutures, J. Virlet, P. Florian, P.J. Grandinetti, Solid State Nucl. Magn. Reson. 6 (1) (1996) 73. [23] R. Ternane, G. Boulon, Y. Guyot, M.Th. Cohen-Adad, M.T. Ayedi, N.K. Ariguib, Opt. Mater. 22 (2) (2003) 117. [24] S.D. Ross, J. Mol. Spectrosc. 29 (1969) 131. [25] Y. Shi, J. Liang, H. Zhang, J. Yang, W. Zhuang, G. Rao, J. Solid State Chem. 129 (1997) 45. [26] G. Herzberg, Infrared and Raman Spectra of Polyatomic Molecules, Van Nostrand, Princeton, NJ, 1964. [27] J.C. Trombe, Ann. Chim. 8 (1973) 335. [28] W.P. Rothwell, J.S. Waugh, J.P. Yesinowski, J. Am. Chem. Soc. 102 (1980) 2637.