In this chapter, the preparation and the properties of synthetic phases of apatite are given; the geological role is described in Section 7.3. The ideal general formula of an apatite-type oxide may be written as M₁₀(XO₄)₆O₂ (M = alkaline-earth and/or rare-earth element, X = Si, Ge, P, V, …). The structure (Fig. 1) can be described in terms of a “microporous” ¹

Microporous material is defined as containing pores with the diameters >2 nm. The materials with the pore diameter in the range from 2 to 50 nm and higher than 50 nm are termed as mesoporous and macroporous, respectively. In combination with nanotechnology, the term nanoporous material is often used. Despite the fact that there is not clear definition, usually the pores with the size from 0.1 to 100 nm are considered. In other words, nanoporous covers the range from microporous to macroporous [1].

[1] framework (A(1)₄(XO₄)₆) composed of face sharing M(1)O₆ trigonal meta-prismatic columns, which are corner connected to MO₄ tetrahedra. This framework allows some flexibility to accommodate remaining M(2)₆O₂ units [2].

Fig. 1.
Illustration of “microporous” M(1)(XO₄)₆ framework of the apatite (M₁₀(XO₄)₆O₂) structure (a): tetrahedra MO₄, M(1) cation at the center of trigonal meta-prism. Remaining M(2)₆O₂ units occupy the cavities within this framework (b): large spheres are M(2) cations and small spheres are O anions [2].

5.1 Apatite-type lanthanium silicates

During recent decades, oxyapatite ²

Since the prefix “oxy-“ can be explained as containing oxygen or additional oxygen, and the prefix “oxo-“ is used for the functional group or substituent oxygen atom connected to another atom by a double bond (=O), the names oxyapatite, oxy-apatite, oxoapatite and oxo-apatite can be considered as synonyms. In the published literature, the name oxyapatite is the most frequently used (~90%), and the term oxy-apatite is the second (~8%). The names oxoapatite and oxo-apatite are used much rarely (only about ~2%).

Apatite-type silicates described in this chapter can be also named as oxy-britholites (oxybritholites) [84].

-type structure with the general formula: REE_9.33+xSi₆O_26+3x/2 (where REE is rare-earth element) [3],[4], REE_9.33□_0.67(SiO₄)₆O₂ [5] or REE_10−x(SiO₄)₆O_2+y [6] has attracted considerable attention as oxide ion conductors. Apatite-type oxides have attracted much attention as the material for electrolytes in solid oxide fuel cell and sensors (Chapter 10).

In low atomic number rare-earth silicate systems, an apatite phase occurs with a range of stability extending from Ln_4.67(SiO₄)₃O to Ln₄(SiO₄)₃. The stability decreases as the rare-earth atomic number increases, with a mixture of Ln₂SiO₅ and Ln₂Si₂O₇ replacing apatite as the preferred phase assemblage [7],[8],[9],[10],[11].

Apatite-type rare-earth element (REE) silicates of the composition of REE_10−x(SiO₄)₆O_2+y, where REE = La, Nd, Gd and Dy, were prepared by Martínez-González et al [6] via the mechanochemical synthesis (stabilized zirconia planetary ball mill: ball-to-powder ratio ~10:1, 350 rpm for maximum time of 9 h) starting from the stoichiometric mixtures of constituent oxides, REE₂O₃ and SiO₂ (molar ratio = 4:5), followed by post-milling thermal treatment (1500°C for 3 h). The ionic conductivity increases with the increasing size of REE cations. ³

This conclusion is in discrepancy with the findings of Higuchi et al [12] described below.

The mechanochemical synthesis of apatite-type lanthanum silicates from the mixture of La₂O₃ and amorphous silica without post-milling thermal treatment was described by Fuentes et al [4].

Rare-earth element-doped apatite-type lanthanum silicates of the composition of La₉MSi₆O₂₇, where M = Nd, Sm, Gd and Yb, were synthesized by the high-temperature solid-state reaction process by Xiang et al [3]. All rare-earth oxide powders (La₂O₃, Nd₂O₃, Sm₂O₃, Gd₂O₃ and Yb₂O₃) were firstly pre-calcined at 900°C for 2 h in order to achieve complete decarbonation and dehydroxylation before weighing. The stoichiometric mixtures were mechanically mixed in absolute ethanol for 24 h using zirconia milling media at the speed of 400 rpm and dried at 100°C in air. The powder mixture was calcined at 1350°C for 10 h and then ground by hand with an agate mortar and pestle to reduce the particle size. After that, the powders were uniaxially pressed at 20 MPa and then statically cold pressed at 200 MPa for 5 min. The compacts were pressureless sintered at 1650 K for 10 h in air.

Apatite-type phase	Lattice parameters [Å]		M	V	Density	E	σ₀
Apatite-type phase	a	c	[g·mol⁻¹]	[Å³]	[g·cm⁻³]	[eV]	[S·K·cm⁻¹]
La₄SiO₃O₁₂	9.376	6.761	831.87	514.73	—	—	—
La₁₀Si₆O₂₇	9.709	7.176	1989.552	585.82	5.168	0.76	2.51·10⁴
La₉NdSi₆O₂₇	9.729	7.191	1994.886	589.46	5.344	0.79	4.78·10⁴
La₉SmSi₆O₂₇	9.692	7.161	2001.046	582.55	5.405	0.82	1.85·10⁴
La₉GdSi₆O₂₇	9.723	7.187	2007.896	588.41	5.485	0.87	3.24·10⁴
La₉YbSi₆O₂₇	9.701	7.143	2023.686	582.16	5.459	0.98	4.34·10⁵

Table 1.

The properties of apatite-type lanthanium silicates [3],[8].

The lattice parameters and the properties of prepared apatite-type lanthanium silicates are listed in Table 1. All prepared compounds possess hexagonal apatite structure with the space group P6₃/m. The temperature dependence of total electrical conductivity for different compositions is determined using the Arrhenius equation ⁴

The equation of Svante August Arrhenius [13],[14], which predicts that the rate constant k depends on the temperature: k = A exp (-E_a/RT), where A is the frequency (pre-exponential factor), E_a is the activation energy, R is universal gas constant (8.314 J·K⁻¹·mol⁻¹) and T is the thermodynamic temperature [13].

[13],[14],[15] in the following form [3],[5],[16],[17]:

σ T = σ 0 exp − E k B T = σ 0 exp − Δ H m + Δ H a k B T E1

where σ is the total electrical conductivity, σ₀ is the pre-exponential factor related to the effective number of mobile oxide ions, E is the activation energy for the electrical conduction process, k_B is the Boltzmann constant and T is absolute temperature. ΔH_m and ΔH_a denote the migration enthalpy of oxygen ion and the association enthalpy of defects, respectively. The determined activation energy and pre-exponential factor are listed in Table 1. It can be seen that the activation energy gradually increases from La₁₀Si₆O₂₇ to La₉GdSi₆O₂₇. Total electrical conductivity can be calculated from the following equation:

σ = h R S E2

where h is the thickness of the specimen, S is the electrode area of the specimen surface and R is the total resistance including grain and grain boundary resistance. Lanthanum silicates doped with Nd or Yb cations exhibit higher total electrical conductivity than undoped lanthanum silicates. The highest total conductivity value obtained at 500°C is 4.31·10⁻⁴ S·cm⁻¹ for La₉NdSi₆O₂₇. The total electrical conductivity is also a function of partial pressure of oxygen [3].

Fig. 2.
Crystal structure of apatite-type rare-earth element silicate viewed along the c-axis [12].

The measurements using single crystals revealed definite anisotropy of the electrical conductivity of Ln_9.33(SiO₄)₆O₂, that is, the conductivity parallel to the c-axis is larger by one order of magnitude than that perpendicular to the c-axis. This fact clearly indicates that the channel oxide ions not bonded to silicon are the principal charge carriers in apatite-type lanthanum silicates. The structure of apatite-type rare-earth silicate is shown in Fig. 2. SiO₄ tetrahedra are isolated mutually, and Ln ions (REE ions, in general) at 6h sites (sevenfold coordinated site (x,y,¼)) [18] ⁵

According to the Wyckoff notation: the specification of actual coordinates of atoms within the unit or primitive cell, which can be generated by the point-group operations or may be found by reference in the International Tables for Crystallography [18].

form channels, in which oxide ions at 2a sites are located (possess the threefold coordination with rare-earth ions at the 6h sites in the same plane), along the c-axis.

These mobile ions at these sites have much larger anisotropic displacement parameters in the direction of the c-axis than those in the direction of the a-axis, even at room temperature, which reflects high oxide ion conduction along the c-axis. The ninefold coordinate position (4f site (1/3,2/3,z)) is the second site for the accommodation of REE cations in the structure of apatite-type REE silicate [6],[12],[19],[20].

Since the interstitial space provided by these rare-earth ions is the smallest throughout the channel along the c-axis of the apatite structure, the migration of oxide ions through the channel will not be affected significantly even if the sizes of rare-earth ions are varied. It is therefore reasonable that the electrical conductivities of apatite-type rare-earth silicates are independent on the kind of rare-earth elements [12].

Conventional oxide ion conductors are designed on the basis of the oxygen vacancy model by the introduction of aliovalent ⁶

Cation with different valence. Apatite structure shows large flexibility upon the substitution of other aliovalent cations at the ‘Ca’ sites, pentavalent and tetravalent ions such as V⁵⁺, As⁵⁺ and Si⁴⁺ at the ‘P’ site and halide, oxide ions at the ‘OH’ site [21], as was described.

[21] cations. In Ln_9.33(SiO₄)₆O₂, however, cation vacancies are present rather than oxygen vacancies. Therefore, the introduction of cation vacancies into the structure of an oxide material may induce high oxide ion conductivity if the structure has a channel or a plane that can be a path for the migration of oxide ions [12].

Apparent exchange of O(1), O(2) and O(3) oxide ions bonded to Si was observed by ¹⁷O NMR measurement on La_9.33Si₆O₂₆ by Kiyino et al [22], while it was not observed for oxide ion on the isolated site O(4). The results indicate that oxide ions bonded to Si at the position O(1), O(2) and O(3) are the main diffusion species in the oxide ion conductivity.

Trivalent and divalent dopants ⁷

Dopants are also termed as doping agents. It can be defined as an impurity element added to the material structure in low concentration (usually <1 wt.% [23]) in order to alter its properties.

[23] have been introduced into the La_9.33(SiO₄)₆O₂ structure according to the following nominal mechanisms [24]:

S i 4 + → M 3 + + 1 3 L a 3 + E3

2 3 L a 3 + → A E E 2 + E4

1 3 L a 3 + → A E E 2 + + 1 2 O 2 − E5

where M = Al, Ga, B, Co, Fe, Mn, … and AEE denotes the alkaline-earth elements (Ca, Sr and Ba). Doping with Al, Ga and B according to the formula: La_9.33+x/3(SiO₄)_6−x(MO₄)_xO₂, via the mechanism in Eq. 3, causes that bulk conductivity increases in up to two orders of magnitude in the case of Al for x = 1 – 1.5. If, however, the sample is stoichiometric on both cation and anion sites, as for La₈Sr₂(SiO₄)₆O₂, the AEE doping reduces the conductivity and increases the activation energy for the conduction compared to La_9.33(SiO₄)₆O₂.

The effect of Fe doping on the electrical properties of lanthanum silicates of the composition of La₁₀Si_6−xFe_xO_27−x/2 (where x = 0.2, 0.4, 0.6, 0.8 and 1.0) was performed by Shi and Zhang [16] via the sol-gel process. Tetraethyl orthosilicate (TEOS), La(NO₃)₃·6H₂O and Fe(NO₃)₃·9H₂O were used as starting materials. Stoichiometric amounts of Fe(NO₃)₃·9H₂O and La(NO₃)₃·6H₂O were dissolved in the mixture of ethanol, acetic acid and distilled water. The appropriate amount of TEOS was added to the solution while continuous stirring. The solution became gradually a purple clear sol. After refluxing at 80°C for 1 – 2 h, the sol transferred to a clear gel. Then, the wet gel was dried at 100°C for 20 h. The gel was heated at 600°C for 4 h to remove water and organic components and to decompose nitrates. In order to get the desired phase, obtained precursor was then calcined at 1000°C for 4 h.

Apatite-type lanthanium silicate	Lattice parameters [Å]		V	E (600 – 800°C)	E (400 – 550°C)
Apatite-type lanthanium silicate	a	c	[Å³]	[eV]
La₁₀Si_5.8Fe_0.2O_26.9	9.725	7.192	589.1	0.78	0.96
La₁₀Si_5.6Fe_0.4O_26.8	9.729	7.208	590.8	0.74	0.95
La₁₀Si_5.4Fe_0.6O_26.7	9.732	7.220	592.2	0.72	0.89
La₁₀Si_5.2Fe_0.8O_26.6	9.735	2.217	592.3	0.74	1.01
La₁₀Si₅FeO_26.5	9.743	7.229	593.5	0.75	1.02

Table 2.

Lattice parameters of Fe-doped apatite-type lanthanium silicates [16].

All synthesized samples have hexagonal lattice structure with the space group of P6₃/m. The lattice parameters of prepared Fe-doped apatite-type lanthanium silicates and the activation energy of conductivities (Eq. 1) for different Fe contents are listed in Table 2. When x = 0.6, La₁₀Si_5.4Fe_0.6O_26.7 exhibits the lowest activation energy. The lattice parameters of La₁₀Si₆O₂₇ (Table 1) and doped specimen (Table 2) show that the values of a, c and V increase with the content of iron. The conductivity of La₁₀Si_6−xFe_xO_27−x/2 is independent of oxygen partial pressure in the range from 0 to 100 kPa, which indicates that the conductivity of all samples is mainly ionic [16].

The oxygen ionic and electronic transport in apatite ceramics with the composition of La₁₀Si_6−xFe_xO_27−x/2 (x = 1 – 2) [25] and La_10−xSi_6−yAl_yO_{27−3x/2−y/2} (x = 0 – 33; y = 0.5 – 1.5) [26],[27] was investigated by Shaula et al In both cases, the essential role of oxygen content on the ionic conductivity of apatite phase was recognized. The ion transference number ⁸

The fraction of total current that is transferred by a given ion is affected by its mobility. The sum of transport numbers for all ions in electrolyte is equal to one [28].

[28] increases with decreasing partial pressure of oxygen. Such behavior indicates that the conduction under oxidizing condition is predominantly of p-type ⁹

The p-type carriers possess typically higher mobility [25].

(with respect to n-type of conductivity). Similar to the foundation of Shi and Zhang [16], the conductivity of these phases is predominantly ionic and almost independent on partial pressure of oxygen. The ion transference numbers are higher than 0.99, while the p-type electronic contribution to total conductivity is about 3% (700 – 950°C, La₁₀Si₄Fe₂O₂₆). The oxygen ionic conductivity should increase with decreasing iron content due to higher concentration of oxygen interstitials.

Another important factor influencing the oxygen diffusion is M-site deficiency, which affects the unit cell volume and may cause the O(5) ion displacement into interstitial sites, thus creating the vacancies in the O(5) sites at fixed total oxygen content. In particular, an enhanced ionic conduction was found in the system La_9.33+x/3Si_6−xAl_xO₂₆, where Al doping is compensated by the A-site vacancy concentration without oxygen content variations [29],[30].

The incorporation of praseodymium in the apatite-type lattice of La_9.83−xPr_xSi_4.5Fe_1.5O_26+δ (x = 0 – 6) decreases the unit cell volume, suppresses the Fe⁴⁺ formation according to Mössbauer

spectroscopy ¹⁰

The technique is based on the Mössbauer effect of recoil-free nuclear resonance fluorescence [31], i.e. the phenomenon of emission or absorption of X-ray photon without the loss of energy. The Mössbauer effect has been detected in a total of 88 X-ray transitions in 72 isotopes of 42 different elements [32]. The ⁵⁷Fe Mössbauer isotope is the most frequently used [33]. The Mössbauer spectroscopy can be used to determine the oxidation states of iron in minerals and to identify the presence of some mineral species in samples of unknown composition [31].

[31],[32],[33],[34] and increases p- and n-type electronic contributions to total conductivity under oxidizing conditions, while the level of oxygen ionic transport at temperatures above 1000 K remains unaffected [35].

Since the size of the conduction channel increases with the Mg doping, the enhancement of the ionic conductivity of lanthanum silicate-based apatites can be reached by optimizing the La content and the Mg doping level at the same time. The ionic conductivities of La₁₀Si_5.8Mg_0.2O_26.8 and La_9.8Si_5.7Mg_0.3O_26.4 at 800°C are 88 and 74 mS·cm⁻¹ with the activation energy of 0.43 and 0.42 eV, respectively [36].

The ionic conduction in cation-deficient apatite La_9.33−2x/3M_xSi₆O₂₆, where M = Mg, Ca and Sr was investigated by Yuan et al [37]. The nature of dopant and the extent of substitution have a significant effect on the conductivity. The greatest decrease in conductivity is observed for Mg doping followed by Ca- and Sr-doped apatites. The effect is ultimately attributed to the amount of oxygen interstitials, which is affected by the crystal lattice distortion arising from the cation vacancies.

The incorporation of additional La₂O₃ into La_9.33(SiO₄)₆O₂ to form La₁₀(SiO₄)₆O₃ or intermediate compositions can most obviously be achieved by filling empty interstitial sites with oxygen. The only alternative scenario would involve the creation of cation vacancies on the Si sublattice, which is unlikely as Si is present as a complex anion. The incorporation of excess of La₂O₃ into La_9.33(SiO₄)₆O₂ can therefore be expressed as [10]:

L a 9.33 0.67 S i O 4 6 O 2 + 1 / 3 L a 2 O 3 → L a 10 S i O 4 6 O 2 O i L a 10 S i O 4 O 3 E6

Thus, in the ideal pure La₁₀(SiO₄)₆O₃, the 4f and 6h sites are fully occupied by La³⁺ ions, while an extra oxygen interstitial is introduced into the lattice to maintain the electroneutrality. The oxygen interstitial may benefit the oxide ion transportation if it is located nearby the [001] direction c-axis of the conventional unit cell. From the space-filling consideration, the most appropriate sites for the oxygen occupation are in this position; however, some distortion of the O 2a sites would be required to accommodate extra oxygen atoms. This could be achieved by decreasing the symmetry from P6₃/m to P6₃ allowing oxygen to move from 0,0,1/4 to 0,0,x. Recent studies suggest that a range of partially occupied (0,0,x) sites may accommodate this extra interstitial oxygen. From this point of view, La₁₀(SiO₄)₆O₃ should exhibit higher conductivity than La_9.33(SiO₄)₆O₂ [10],[38].

Introducing Sr²⁺ cations to the La³⁺ atomic positions, as in the La₁₀(SiO₄)₆O₃ phase, leads to complete elimination of vacancies according to the substitution [39]:

L a 10 S i O 4 6 O 3 → + S r 2 + L a 9 S r 1 S i O 4 6 O 2.5 → + S r 2 + L a 8 S r 2 S i O 4 6 O 2 E7

The substitution of La₂O₃ by SrO, taking into account the charge balance and the oxygen content, can be represented as follows (Kröger-Vink notation ¹¹

The Kröger-Vink notation indicates the lattice position for the point defect species in the crystal and its effective electric charge relative to the perfect lattice: M_Y ^Z is the atomic species M (or vacancy V) that occupies the lattice site Y and possesses the effective charge Z, where the symbols ●, ʹ and × are used for the effective charge +1, -1 and neutral particle, respectively) [40]. For example, Al_i ^●●● is Al³⁺ ion at interstitial site (i), V_Alʹʹʹ is Al³⁺ vacancy, V_O ^●● is O²⁻ vacancy, Sr_Laʹ (Eq. 8) means Sr²⁺ ion replacing La³⁺ at lattice site, Ti_Al ^● means Ti⁴⁺ replacing Al³⁺ at lattice site, eʹ is electron and h^● is the hole. The equation must fulfill the following three rules: mass balance (1), electroneutrality or charge balance (2) and site ratio conservation balance (3) [41].

[40],[41]):

S r O + 1 / 2 O o x + L a L a x → S r L a / + 1 / 2 V o • • + 1 / 2 L a 2 O 3 E8

Lanthanum oxyapatite phases are substantially stable with respect to their binary oxides. The general trend in the formation enthalpies as a function of (La + Sr)/(La + Sr + Si) shows that the apatite phase becomes more energetically stable as the cation vacancy and oxygen excess concentrations decrease. The stoichiometric sample achieved by Sr²⁺ doping, with no cation vacancies or interstitial oxygen atoms, is the most stable composition. The energetics of lanthanum silicate apatite materials (La_9.33+x(SiO₄)₆O_2+3x/2 and La_10−xSr_x(SiO₄)₆O_3−0.5x) depends on lanthanum deficiency and oxygen interstitial¹²

Interstitial sites are sites between normal (equilibrium) atomic positions of ideal lattice atoms [42]. Interstitial atoms and vacancies (lattice site where atom is absent) are the simplest types of point defects in a crystal. A vacancy and interstitial atoms positioned close together are referred to as the Frenkel pair. Apart from the point defects, the line crystal defects (dislocation and disclination) are recognized [43].

[42],[43] concentrations, and the cation vacancy concentrations appear to be the dominant factor in energetics [39].

Fig. 3.
Schematic diagram of preparation of lanthanum silicate by a sol-gel process [44].

The schematic diagram of the sol-gel process used by Tao and Irvine [44] for the preparation of apatite-type lanthanium silicates is shown in Fig. 3. The room-temperature structure is hexagonal, the space group is P6₃ or P6₃/m, with a = 9.722 and c = 7.182 Å for La₁₀(SiO₄)₆O₃ and a = 9.717 and c = 7.177 Å for La_9.33(SiO₄)₆O₂, i.e. the cell volume of La₁₀(SiO₄)₆O₃ is a little greater than that of La_9.33(SiO₄)₆O₂. Both compositions exhibit high ionic conductivity, although the grain boundary resistance is the dominant feature in the impedance spectrum of both. In general, the conductivity of La₁₀(SiO₄)₆O₃ is higher than that of La_9.33(SiO₄)₆O₂ and this indicates that oxygen interstitials may be introduced into the apatite lattice of La₁₀(SiO₄)₆O₃, which may benefit the oxygen ion transportation [44].

The La₁₀Si₆O₂₇ nanopowders with apatite structure were synthesized by the Li et al [45] co-precipitation method. After the calcination at 900°C and then removing of La₂O₃ by acid washing, the pure stoichiometric La₁₀Si₆O₂₇ nanopowders are obtained. The oxyapatite ceramics with the density higher than 95% can be obtained at rather low sintering temperature of 1300°C, and it has comparable total conductivity with the samples sintered at 1650°C from the powders prepared by solid-state reaction.

La₂O₃ and TEOS in stoichiometric amount were used by Masubuchi et al [46] as the starting materials for the preparation of both powder and film of apatite-type La_9.33(SiO₄)₆O₂ via the alkoxide hydrolysis. Lanthanum oxide was dissolved in HNO₃ (6 mol·dm⁻³) and mixed with ethanol. Then, stoichiometric amount of TEOS in ethanol was added (La:Si = 9.33:6) to this solution. The precursor solution was obtained by refluxing for one night. This solution was heated to gelating on the hot plate followed by calcination and annealing in powder preparation. Either quartz glass or Pt foil substrate was dipped to the gelatinous solution and dried for the film preparation. It was calcined at 500°C for 1 h to remove the organic contents and then fired at 1000°C for 10 h. This preparation steps were repeated to increase the film thickness. The film showed preferred orientation of the apatite crystal in thinner film. The conductivity of sintered body was lower in about one order of magnitude than the value of single crystal perpendicular to c-axis [46].

The synthesis and the conductivities of Ti-doped apatite-type phases of the composition of (La/Ba)_10−x(Si/Ge)₆O_26+z, where Ti substituted at the Si/Ge site, were reported by Sansom et al [47]. The conductivities were shown to be the highest for the samples containing either cation vacancies or oxygen excess, which is consistent with previous studies of apatite-type oxide ion conductors. However, the Ti doping was shown to generally decrease the conductivity in comparison with equivalent samples containing only Si/Ge at the tetrahedral sites, with the greatest decrease for Si-containing samples.

Vanadium-doped oxyapatite phases of the composition of La_10−xV_x(SiO₄)₆O_3+x were prepared by Yuan et al [48] via the sol-gel method. The apatite phase begins to form at 800°C, which is much lower than in the case of conventional solid-state synthesis method. The best conductivity of La₉V(SiO₄)₆O₄ is 1.67·10⁻² S·cm⁻¹, which is significantly higher than that for lanthanum silicate oxides (1.19·10⁻² S·cm⁻¹). The valence ion V⁵⁺ doped for La³⁺ does lead to the formation of hexagonal apatites even with high oxygen contents.

The phase La₅Si₂BO₁₃ [49],[50],[51] crystallizes with apatite-related structure (Fig. 4) with the space group P6₃/m and the cell parameters a = 9.5587 Å, c =7.2173 Å and Z = 2. The composition of these apatite-like compounds can be also expressed via the general formula: La_9.33+xSi_6−2yB_3yO₆, where 0 ≤ x ≤ 0.67. At limiting compositions x = 0.67, La(2) site is fully occupied, and the formula referred to the unit cell is La₁₀Si₄B₂O₂₆ or, more simply, La₅Si₂BO₁₃.

Fig. 4.
The crystal structure of La₅Si₂BO₁₃ (a) [51] and the upper section of the ternary phase diagram La₂O₃-SiO₂-B₂O₃ at 1100°C (b) [49].

The comparison with other apatite-like structures shows lower distortion in the M(1) polyhedron and unusually short bond length from La in the M(2) site and O(4) oxygen in the column site (2.303 Å). These results can be explained in view of the presence of trivalent La and divalent O, respectively, in the M(1) and M(2) sites and in the column anion site, whereas, in apatites, these sites are occupied by divalent and monovalent ions, respectively [49].

The preparation of La-Si-O apatite-type thin films was described by Vieira et al [52] with Si/(La + Si) atomic ratios ranging from 0.36 to 0.43 being produced via the magnetron sputtering in reactive Ar/O discharge gas. The apatite-type lanthanum silicate phase was formed in all as-deposited films upon the annealing at 900°C for 1 h. The lanthanum silicate films obtained by annealing the as-deposited films with lower Si/(La + Si) atomic ratios have a preferential orientation with the c-axis perpendicular to the substrate, while low-intensity diffraction peaks ascribed to La₂Si₂O₇ phase were detected in the films deposited with higher Si content. Preferentially oriented films have higher activation energy and lower ionic conductivity, as the ionic conductivity measurements were performed in the direction perpendicular to the c-axis. The highest ionic conductivity was obtained for the film deposited with a Si/(La + Si) atomic ratio of 0.42, with a value of 1.2 × 10⁻² S·cm⁻¹ at 750°C. By the incorporation of oxygen in the as-deposited films, the silicon segregation upon annealing was avoided.

Fig. 5.
Primary phase diagram of ternary system La₂O₃-Ga₂O₃-SiO₂ around LGS [53].

The formation of ternary compound with apatite structure in the system La₂O₃-Ga₂O₃-SiO₂ (Fig. 5) was first reported by Wang and Uda [53]. The apatite phase, which precipitates from the melt of the composition around that of stoichiometric La₃Ga₅SiO₁₄ (LGS), can be described by the formula: La₁₄Ga_xSi_9−xO_39−x/2, where 0 ≤ x ≤ 3.5. Since there is a large field for the formation of solid solutions with the range extending from La₁₄Si₉O₃₉ to Ga₂O₃, some Si⁴⁺ sites are probably substituted by Ga³⁺.

The liquidus surface of LS(G) was determined to be the field on the Ga₂O₃-poor side of boundary curve ABCD. The liquidus surface of LS(G) covers the stoichiometric composition of LGS. In this field, the crystallization of LS(G) aciculae was observed in all samples that were heated to temperatures above 1500°C. The liquidus volume of LGS is denoted by the field BCEF. It seems to be a narrow field in the composition between the liquidus surfaces of LS(G) and Ga₂O₃. E and F are eutectic points, where LGS + LaGaO₃ + Ga₂O₃ + liquid and LGS + Ga₂O₃ + La₂Si₂O₇ + liquid were found, respectively [53].

The CaO-La₂O₃-SiO₂-P₂O₅ phase diagram was investigated by El Ouenzerfi et al [54] in order to determine a domain inside which all points correspond to pure apatitic oxyphosphosilicates with the general formula: Ca_xLa_y(SiO₄)_6−u(PO₄)_uO_t. The defined domain (Fig. 6) is only a part of the whole existence domain of the apatitic structure, but it allows to prepare pure apatitic samples with well-controlled composition.

Fig. 6.
Britholite stability domain in the CaO-SiO₂-La₂O₃-P₂O₅ quaternary system [54].

For these samples, the continuous change of the stoichiometry of each element proves that it exists as a solid solution including the oxygen content. This observation completes the literature data where britholites are presented as limited to three series corresponding to the stoichiometry [54],[55]:

Sr_2+xLa_8−x(SiO₄)_6−x(PO₄)_xO₂ with 0 ≤ x ≤ 6 ¹³
Wanmaker et al [55] reported the synthesis of apatite-type compounds of the composition:

(a) M²⁺(II)M(III)_8−x(SiO₄)_6−x(PO₄)_xO₂, where 0 ≤ x ≤ 6;

(b) M(II)_3+xM(III)_6−x(SiO₄)_6−x(PO₄)_x, where 0 ≤ x ≤ 1.4;

(c) M(II)_4+xM(III)_6−x(SiO₄)_6−x(PO₄)_xO, where 0 ≤ x ≤ 6.

where M(II) = Ca, Sr, Ba, Mg, Zn or Cd and M(III) Y or La. The paper also contains structural data for several other newly prepared oxy-britholites, including Zn₂La₈(SiO₄)₆O₂, BaMgY₈(SiO₄)₆O₂, Zn₂Y₈(SiO₄)₆O₂, Cd₂Y₈(SiO₄)₆O₂, Ca₄La₅(SiO₄)₅(PO₄) and Ba₄La₅(SiO₄)₅(PO₄).
;
Sr_3+xLa_6−x(SiO₄)_6−x(PO₄)_x with 0 ≤ x ≤ 1.5;
Sr_4+xLa_6−x(SiO₄)_6−x(PO₄)_xO with 0 ≤ x ≤ 6.

Inside this domain, the solid solution continuously varies between pure phosphate apatites Ca_xLa_y(PO₄)₆O_t and pure silicate apatites Ca_xLa_y(SiO₄)₆O_t and also between oxyapatites Ca_xLa_y(SiO₄)_6−u(PO₄)_uO_t and nonoxyapatites Ca_xLa_y(SiO₄)_6−u(PO₄)_u.

During the investigation of the kinetics of solid-state sintering ¹⁴

The densification rate is considered as the function of temperature (T) and mean grain size (D_m). Constant A depends on the surface energy (γ_sg) of grains, on the apatite molar volume (Ω) and on average coefficient of diffusion of limiting species D. This relationship can be written as follows [56]:

(a) d ρ d t = AD T D m n

The coefficient of diffusion D is thermally activated:

(b) D = D 0 exp − E a R T

where D₀ is the pre-exponential coefficient of diffusion, R is the universal gas constant and E_a is the apparent activation energy of diffusion of the rate limiting process. Exponent n in Eq. (a) depends on the mechanism of transport of the limiting species governing the kinetics of densification.

of strontium-doped apatite-type lanthanum silicates (Sr_xLa_10−xSi₆O_27−x/2) under isothermal conditions (1250 – 1550°C), Bonhomme et al [56] recognized that the densification mechanism of the apatite ceramics was controlled by the diffusion of rare-earth element (La) at the grain boundaries. This process showed the activation energy of 470 kJ·mol⁻¹.

The ternary phase diagram Al₂O₃-SiO₂-La₂O₃ at 1300°C (Fig. 7) was investigated by Mazza and Ronchetti [57]. La₁₄Si₉O₃₉ was described by Kuz’min and Belov [58] as an apatite-like structure of hexagonal symmetry (space group P6₃/m). Isomorphous compounds were also reported for Nd [59], Ce [60] and Sm [58]. The La₁₄Si₉O₃₉ compound extends its stability range in the interior of the phase diagram, forming the solid solution of the type La_14+1x/3Si_9−xAl_xO₃₉, which is stable from x = 0 to x = 1.5. This substitution stoichiometry (Al + 1/3La ↔ Si) can be described as a tetrahedral Al for Si substitution on the 6h position and contemporary occupation of vacant La sites [57].

Fig. 7.
Ternary phase diagram at 1300°C in air [57].

5.2 Apatite-type lanthanium germanates

Lanthanum germanate and silicate apatite-based materials, both undoped and with partial substitution of, for example, Al, B instead of Si and Sr in place of La, are promising oxide ion conductors with potential applications as high-temperature solid electrolytes. Considerable uncertainties remain over the stoichiometry, the defect structure and the conductivity variations within various apatite systems, partly caused by the fact that the La:(Ge, Si) ratio is variable, giving rise to the solid solutions in the undoped systems as well as to the solid solutions formed by partial replacement of La and/or (Si, Ge) together with, depending on the solid solution mechanism, variations in oxygen content [24].

The preparation of single crystal of apatite-type lanthanium germanate of the composition of La_9.33Ge₆O₂₆ was reported by Nakayama and Sakamoto [61]. The mixtures of La₂O₃ and GeO₂ were well mixed in ethanol under an atomic ratio of La:Ge = 9.33:6 using a ball mill. The mixture was dried and then calcined in air at 1000°C for 2 h. The resulting La_9.33Ge₆O₂₆ powders were further ball milled into finer powders. After the pre-sintering, the closely packed La_9.33Ge₆O₂₆ powders were heated at 1300°C for 2 h in air, and the surface of the specimen was mirror polished. The polished surface of polycrystalline La_9.33Ge₆O₂₆ ceramics was then bonded to a <001> face of La_9.33Ge₆O₂₆ seed crystal prepared by the Czochralsky (Section 4.2) method. On heating of the bonded sample at 1525 – 1550°C, continuous grain growth of polycrystalline La_9.33Ge₆O₂₆ occurred and the single crystal was gradually grown from the seed crystal into the polycrystalline region [61].

Apatite-type lanthanium germanate possesses hexagonal structure with the space group P6₃/m and the lattice parameters: a = 9.9256 and c = 7.2900 Å, V = 621.97 Å³ and Z = 2. The calculated density of the phase is 2.148 g·cm⁻³. Similar to apatite-type lanthanium silicate (La_9.33Si₆O₂₆) described above, the structure of La_9.33Ge₆O₂₆ (Fig. 8) contains two different sites for atoms of La. The La(1) and La(2) sites are located at 4f and 6h, respectively. While the La_9.33Ge₆O₂₆ single crystal showed little anisotropy in conductivity, the conductivity of La_9.33Si₆O₂₆ single crystal gave 100 times higher value parallel to the c-axis than that perpendicular to the c-axis at each temperature (Section 5.1) [61].

Fig. 8.
Hexagonal structure proposed for the apatite-type phase of La_9.33X₆O₂₆, where X = Si and Ge [61].

The selective doping of La_9.33+x(GeO₄)₆O_2+3x/2 with Y leads to the stabilization of hexagonal lattice, even at high oxygen contents. Furthermore, this has the effect of enhancing the low-temperature conductivities [62]. Depending on the composition, the cell can be either hexagonal or triclinic, with the evidence of reduced low-temperature conductivities for the latter, attributed to increased defect trapping in this lower symmetry cell. In summary, it was shown that the series La₈Y₂(GeO₄)_6−x(GaO₄)_xO_3−x/2 can be prepared for 0 ≤ x ≤ 2 with all samples showing the hexagonal symmetry, compared to the series without Y co-doping, La₁₀(GeO₄)_6−x(GaO₄)_xO_3−x/2, for which all compositions display the triclinic symmetry [24],[62],[63].

The effect of Ga doping of the oxygen stoichiometric series containing the cation vacancies, La_7.33+y/3Y₂(GeO₄)_6−y(GaO₄)_yO₂ (0 ≤ y ≤ 2), single-phase samples was obtained for y ≥ 1.0, with small impurities observed at lower Ga contents. The conductivities were shown to increase with increasing cation vacancy content, reaching the values of ≈0.02 S·cm⁻¹ at 800°C, which are similar to the oxygen excess series. These results are in agreement with previous reports on the apatite systems, which showed that the oxide ion conductivity was maximized in samples containing the oxygen excess and/or the cation vacancies [24], [62],[63].

The series of apatite-type silicates/germanates of the composition of La_8+xSr_2−xSi₆O_26+x/2 (0 ≤ x ≤ 1) and La_8+xSr_2−xGe₆O_26+x/2 (0 ≤ x ≤ 2) were prepared from high-purity La₂O₃, SrCO₃, SiO₂ and GeO₂ by Orera et al [64] via the thermal treatment of these components mixed in the stoichiometric ratio.

The extent of, and the structural changes within, the apatite domain in the LaO_1.5-GeO₂-SrO ternary system at 1100°C was studied and the single-phase samples were obtained for La_{9.33+x−2y/3}Sr_y(GeO₄)₆O_2+1.5x with x = 0.17 and 0.34. The hexagonal to triclinic transition is clearly associated with increasing oxygen content rather than with filling the La sites by the addition/substitution of Sr into the structure. The limits of undoped solid solution are ~0.17 ≤ x ≤0.5 at 1100°C [24].

The hydrothermal synthesis of apatite-type compound NaRE₉(GeO₄)₆O₂ (RE = Nd, Pr) with the hexagonal structure of the space group of P6₃/m was described by Emirdag-Eanes et al [65]. The structure is composed of REO₇ and REO₉ polyhedra as well as GeO₄ tetrahedra (Fig. 10). The unit cell dimensions are: a = 9.782(1) Å, c = 7.083(1)Å (T = 293 K) and V = 587.0(2)Å3 for REE = Nd and a = 9.802(1) Å, c = 7.116(1)Å (T = 293 K) and V = 592.1(2)Å3 for REE = Pr.

Fig. 9.
Apatite solid solutions in the LaO_1.5-GeO₂-SrO ternary system [24]: pure phases are indicated by filled circles and the presence of secondary phases is shown by half-filled circles.

Fig. 10.
The unit cell view with GeO₄ and NdO₇ (a) and GeO₄ and NdO₉ down the c-axis. NdO₇ and NdO₉ are dotted polyhedra, and GeO₄ are lined polyhedra [65].

The high-temperature flux method for the preparation of single crystal of hexagonal NaLa₉Ge₆O₂₆ apatite-type germanate (space group P6₃/m, a = 9.883, c = 7.267 Å and Z = 1) was used by Takahashi et al [66]. The crystal structure (Fig. 11) was found to be similar to that of silicate oxyapatite NaY₉Si₆O₂₆. The 4f cation sites are occupied disorderedly by La and Na. On the other hand, the 6h cation sites are occupied by La only. This compound constitutes a new member of the oxyapatite-type structure family with the composition given by general formula: A_x Ln_10−xB₆O₂₄O_3−x.

Fig. 11.
The structure of sodium lanthanum germanate NaLa₉Ge₆O₂₆ [66].

Fig. 12.
Coordination environment of La atoms [66].

The coordination environments of La atoms by O atoms are shown in Fig. 12. La(1) atom on the 4f site is coordinated by nine O atoms. It is linked to three O(1) atoms in a distance of about 0.2479 nm, to three O(2) atoms in a distance of about 0.2561 nm and to three O(3) atoms in a distance of about 0.2924 nm. Because the distance between La(1) and O(3) is relatively large, La(1) atom can be also regarded to be in sixfold coordination, the environment of which is fairly distorted from an ideal octahedron. On the other hand, La(2) atom, which occupies the 6h position, is coordinated by seven O atoms, that is, O(1), O(2), O(4) and O(3). The distances of those two types of bonds between La(2) and O(3) atoms are 0.2618 nm×2 and 0.2431 nm×2, respectively. Those between La(2) and O(1), O(2) and O(4) are 0.2732, 0.2521 and 0.23281 nm, respectively [66].

5.3 Apatite-type borates

Two high terbium content apatites Tb₅Si₂BO₁₃ (a = 9.2569 Å, c = 6.8297 Å, V = 506.83 Å³ and Z = 2) and Tb_4.66Si₃O₁₃ (a = 9.493 Å, c = 6.852 Å, V = 534.70 Å³ and Z = 2) were prepared by Chen and Li [67] via spontaneous crystallization and synthesized with high purities and excellent crystallinities by the sol-gel process. Both compounds are isostructural with P6₃/m space group and exhibit paramagnetic behavior down to 2 K. Owing to high Tb³⁺ ion concentrations and good transmittance in the range from 500 to 1500 nm, Tb₅Si₂BO₁₃ and Tb_4.66Si₃O₁₃ may be promising magneto-optical materials in the visible-near-IR range.

Both Tb₅Si₂BO₁₃ and Tb_4.66Si₃O₁₃ contain two distinct sites for Tb³⁺ cations, which are depicted in Fig. 13. Tb(2) is at the 4f site, which is on a threefold axis and coordinated by nine oxygen ions. However, for Tb_4.66Si₃O₁₃, the Tb(2) site is not fully occupied but leaves one of six Tb(2) positions randomly vacant while fully occupied in Tb₅Si₂BO₁₃. In contrast, Tb(1) at the 6h site is fully occupied and sevenfold coordinated in both Tb₅Si₂BO₁₃ and Tb_4.66Si₃O₁₃. Moreover, in Tb₅Si₂BO₁₃, one third of Si is disorderly occupied by B, which gives rise to extra 1/3 Tb³⁺ ion for the charge balance. The Tb(2)O₉ polyhedron consists of Tb(2) and nine oxygens along the c-axis. Six Tb(1) comprise a sixfold channel parallel to the c-axis. It is worth to note that the channel is considered to play an extremely important role in oxide ion conductivity [67].

Fig. 13.
Unit cell of Tb₅Si₂BO₁₃ (a) and coordination environment of kinds of Tb cations (b) [67].

The structure and optical properties of noncentrosymmetric borate RbSr₄(BO₃)₃ (RSBO) was described by Xia and Li [68]. RSBO can be viewed as a derivative of the apatite-like structure. Based on the anionic group approximation, the optical properties of the compound are compared to those of the structure-related apatite-like compounds with the formula “A₅(TO_n)₃X”. When the structures of all apatite-like crystals are presented in orthorhombic unit cell, the arrangements of planar anionic BO₃ groups are all similar to one-third of the BO₃ groups aligned perfectly parallel at corner- and face-centered locations, whereas the other two-thirds of BO₃ groups are distributed differently.

Europium borate fluoride, Eu₅(BO₃)₃F with an apatite-type structure, was synthesized by Kazmierczak and Höppe [69] as single-phase crystalline powder starting from europium oxide, europium fluoride and boron oxide at 1370 K. Eu₅(BO₃)₃F crystallizes in the space group Pnma.

Fig. 14.
The crystal structure of Sr₁₀[(PO₄)_5.5(BO₄)_0.5](BO₂) (1): (a) the projection along [001] showing the channels formed by Sr3 (gray triangles) and the positions of the XO₄ tetrahedra (gray; Z = 11/12 P + 1/12B) and (b) the side view with emphasized [BO₂]⁻ groups and the coordinating trigonal antiprism formed by Sr3. The comparison of the arrangement of [BO₂]⁻ and F- ions within the Sr channels of Sr₁₀(PO₄)_5.5(BO₄)_0.5(BO₂) (c(1)) and Sr₁₀(PO₄)₆F₂ (c(2)) [70].

The structure of single crystal of strontium phosphate orthoborate metaborate (Sr₁₀[(PO₄)_5.5(BO₄)_0.5](BO₂)) that was grown from the melt by Chen et al [70] is shown in Fig. 14(a). The phase crystallizes in the space group P3 with the cell parameters a = 9.7973 Å, c = 7.3056 Å, V = 607.19 Å³ and Z = 1. The crystal structure is closely related to apatite and contains linear metaborate groups, [BO₂]⁻ (point group D_∞h, B-O = 1.284 Å), taking positions within the channels running along the threefold inversion axis. Strontium sites are found to be fully occupied, while [PO₄]³⁻ tetrahedra are partially replaced by[BO₄]⁵⁻ groups.

The comparison of the nearest neighbors around [BO₂]⁻ and F⁻ located within the channels is shown in Fig. 14(b,c). F⁻ ions (0,0,1/4) are situated on the mirror plane in the center of Sr triangle. As a result, constant F…F distances of 3.64 Å (a/2) are observed along [001] (b). In Sr₁₀[(PO₄)_5.5(BO₄)_0.5](BO₂), the incorporation of boron atoms between two O atoms draws these atoms closer (d(O-B-O) = 2.57 Å) and at the same time increases the gaps between two neighboring [BO₂]⁻ units (d(O…O) = 4.73 Å), which results in alternating O…O distances along the c-axis (a) [70].

By replacing Mn in YCa₃(MnO)₃(BO₃)₄ with trivalent Al and Ga, two new borates with the compositions of YCa₃(MO)₃(BO₃)₄ (M = Al, Ga) were prepared via the solid-state reaction by Yu et al [71]. The phases are isostructural to gaudefroyite with the hexagonal space group P6₃/m. The cell parameters of a = 10.38775 Å, c = 5.69198Å for the Al-containing compound and a = 10:5167 Å, c = 5:8146 Å for the Ga analogue were obtained from the refinements. The structure is constituted of AlO₆ or GaO₆ octahedral chains interconnected by BO₃ groups in the ab plane to form a Kagomé-type lattice ¹⁵

The Kagomé lattice (d) is one of the most interesting lattices in 2D, especially in materials in which the Kagome lattice is built from magnetic ions. Each of its vertices touches a triangle, hexagon, triangle and hexagon (the planes of corner-sharing equilateral triangles). The vertices correspond to the edges of the hexagonal (honeycomb) lattice (c), which in turn is the dual of triangular lattice (can be derived from triangular lattice by periodical removal of ¼ sites) (b). Since it has the same coordination number (z = 4), the Kagomé lattice is also related to the square lattice (a) [72], [73]. Numerous Kagomé compounds built from stacked Kagomé layers were found in Alunite (Jarosite, KFe³⁺₃(SO₄)₂(OH)₆, (e)) family of minerals [74].

[72],[73],[74], leaving trigonal and apatite-like tunnels. It was found that most rare-earth and Cr, Mn ions can be substituted into the Y³⁺ and M³⁺ sites, respectively, and the preference of rare-earth ions to be located in the trigonal tunnel is correlated to the sizes of the M³⁺ ions.

5.4 Other apatite-type REE silicates

Hexagonal apatite-type phase of the composition of Pr₉K(SiO₄)₆O₂ (space group P6₃/m, a = 9.6466 Å and c = 1136 Å, V = 573.28 Å³, ρ_calc = 5.48 g·cm⁻³ and Z = 1) was synthesized by Werner and Kubel [75] in a potassium fluoride flux. Potassium fills one (4f) of two metal positions present in the structure (Fig. 15) with the occupancy factor of 25%. The remaining positions of this site (Pr2/K2) are occupied by praseodymium.

Fig. 15.
The perspective view of Pr₉K(SiO₄)₆O₂ along the c-axis: regular planar coordination of the channel oxygen O(4) and the SiO₄ tetrahedra are marked in gray, and the unit cell is outlined [75].

Oxygen from the silicate groups forms a coordination polyhedron (ninefold) in the shape of a distorted threefold capped trigonal prism. These face sharing [(Pr2/K2)O₉]-polyhedra build up chains, which are interconnected via the SiO₄ groups. The resulting channel framework accommodates sevenfold oxygen-coordinated praseodymium (Pr1), attached to the inside of the tubes that are aligned parallel to the c-axis. Oxide ions O4, located on the longitudinal axis of the channels, exhibit anomalously high atomic displacement parameters along the c-direction [75].

Single crystals of apatite-type Nd_9.33(SiO₄)₆O_2, Pr_9.33(SiO₄)₆O₂ and Sm_9.33(SiO₄)₆O₂ were described in Section 4.2.2. The structure of samarium orthosilicate oxyapatite (Sm₅(SiO₄)₃O, Fig. 16) was resolved by Morgan et al [76]. The phase crystallizes in hexagonal system with the space group P6₃/m and the cell parameters: a = 9.4959 Å, c = 7.0361 Å, c:a = 0.7410 and V = 549.46 Å³. The preparation and the structure of single crystal of strontium tetrapraseodymium tris(silicate) oxide (SrPr₄(SiO₄)₃O), which was grown by the self-flux method using SrCl₂, was described by Sakakura et al [77].

Fig. 16.
The structure of Sm₅(SiO₄)₃O viewed along the c-axis [76].

The M(2) sites are almost exclusively occupied by praseodymium. The complete series of apatite-like compounds REE_9.33□_0.67[SiO₄]₆O₂, LiREE₉[SiO₄]₆O₂ and NaREE₉[SiO₄]₆O₂ were synthesized by Felsche [78] with REE: La → Lu. Apatite-type neodymium silicates doped with various cations at the Si site, Nd₁₀Si₅BO_27−δ (B = Mg, Al, Fe, Si), were synthesized by Xiang et al [79] via the solid-state reaction.

The crystal growth and the structure of three new neodymium-containing silicates, Na_0.50Nd_4.50(SiO₄)₃O, Na_0.63Nd_4.37(SiO₄)₃O_0.74F_0.26 and Na_4.74Nd_4.26(O_0.52F_0.48)[SiO₄]₄, prepared using the eutectic mixture of KF/NaF were investigated by Latshaw et al [80]. Na_0.50Nd_4.50(SiO₄)₃O and Na_0.63Nd_4.37(SiO₄)₃O_0.74F_0.26 adopt the apatite structure and crystallize in hexagonal space group P6₃/m, while Na_4.74Nd_4.26(O_0.52F_0.48)[SiO₄]₄ crystallizes in tetragonal space group I-4 and exhibits rare-earth mixing on the sodium site. The unit cell parameters of the crystals are:

Na_0.50Nd_4.50(SiO₄)₃O: a = 9.5400 Å and c = 7.033 Å;
Na_0.63Nd_4.37(SiO₄)₃O_0.74F_0.26: a = 9.5533 Å and c = 7.0510 Å;
Na_4.74Nd_4.26(O_0.52F_0.48)[SiO₄]₄: a = 12.1255 Å and c = 5.4656 Å.

Fig. 17.
Bond distances (Å) and anisotropic displacement in Gd_4.33Ho_4.33(SiO₄)₆(OH)₂: note that the exaggerated anisotropic displacement of O(3) is attributable to the high proportion of vacancies in REE(1) and near-equatorial distribution of strong bonds to Si and REE(2) [81].

Double REE silicate Gd_4.33Ho_4.33(SiO₄)₆(OH)₂ with the hydroxylapatite structure was synthesized by Wang et al [81] using the piston-cylinder high-pressure apparatus at the pressure of 2.0 GPa and the temperature of 1450°C. Since they have nearly identical chemical character, two REE cations (Ho and Gd) are distributed randomly among the M(1) and M(2) sites, and the charge balance is maintained by the cation vacancies in M(1). The presence of two different REE cations in the same compound might promote better understanding of the cooperative effects of ions under the solid-state conditions (Fig. 17).

The structure analysis reveals that the hexagonal compound crystallizes in usual apatite space group P6₃/m with lattice parameters a = 9.3142 Å and c = 6.7010 Å. REE(1) atoms are connected to nine oxygen atoms with the REE(1)-O bond distances ranging from 2.317 to 2.751 Å (mean 2.478 Å) and REE(2) atoms are connected to seven oxygen atoms with the RE(2)-O bond distances ranging from 2.223 to 2.690 Å (mean 2.393 Å). Oxygen anion O(4) in the apatite channel is located on the 6₃ axis and coordinated with three REE(2) cations arranged in a tricluster perpendicular to the c-axis. An isotropic displacement parameter was used for O(4), and H atom was assumed to ride on it. OH⁻ anions are stacked in regular column in the apatite channel, and in locally ordered structure, their polar direction is flipped in neighboring channel [81].

The formation of apatite-type phases of the composition of KNd₆(SiO₄)₆O₂ from the glass precursor (4K₂O-Nd₂O₃-17SiO₂) during the hydrothermal experiments (500°C and 825 bar) carried out at KOH molarities of 6 or greater was reported by Haile et al [82]. High temperatures, high pressures and long times tended to favor the synthesis of this apatite-type phase over the K₈Nd₃Si₁₂O₃₂OH phase. In comparison with the potassium system, the concentration of NaOH required for the synthesis of NaNd₆(SiO₄)₆O₂ phase (system 4Na₂O-M₂O₃-17SiO₂, where M = Nd and Y) is very low. The formation of apatite-type phases in the 4Na₂O-Y₂O₃-17SiO₂ system was not observed [83].

Two series of strontium-lanthanum apatites, Sr_10−xLa_x(PO₄)_6−x(SiO₄)_xF₂ and Sr_10−xLa_x(PO₄)_6−x(SiO₄)_xO with 0 ≤ x ≤ 6, were synthesized by Boughzala et al [84] via the solid-state reaction in the temperature range of 1200 – 1400°C:

3 S r C O 3 + 6 − x / 2 S r 2 P 2 O 7 + S r F 2 + x S i O 2 + x L a O H 3 → S r 10 − x L a x P O 4 6 − x S i O 4 x F 2 + 3 C O 2 + 3 x / 2 H 2 O E9

and

4 S r C O 3 + 6 − x / 2 S r 2 P 2 O 7 + x S i O 2 + x L a O H 3 → S r 10 − x L a x P O 4 6 − x S i O 4 x O + 4 C O 2 + 3 x / 2 H 2 O E10

where x = 0, 1, 2, 4 and 6. Sr₂P₂O₇ was synthesized by the following reaction at 900°C:

2 S r C O 3 + 2 N H 4 2 H P O 4 → S r 2 P 2 O 7 + 2 C O 2 + 3 H 2 O + 4 N H 3 E11

The raw meal was prepared via mixing SrCO₃, La₂O₃, SiO₂, SrF₂ and (NH₄)₂HPO₄ in required stoichiometric amounts (0 ≤ x ≤ 6). The mixture was ground in an agate mortar, pressed to pellets and calcined at the temperature of 900°C for 12 h under the flow of argon (Sr_10−xLa_x(PO₄)_6−x(SiO₄)_xF₂) and oxygen (Sr_10−xLa_x(PO₄)_6−x(SiO₄)_xO). The product was ground and pressed again in order to improve its homogeneity. Next, thermal treatment was performed at the temperature of 1200 and 1400°C (depending on the content of SiO₂) for 12 h. The samples were heated and cooled with the rate of 10°C·min⁻¹. The incorporation of La³⁺ and SiO₄ ⁴⁻ ions into the apatite structures, i.e. the substitution of the pair La³⁺ and SiO₄ ⁴⁻ for Sr²⁺ and PO₄ ³⁻, induced an increase of parameter a and decrease of parameter c (Fig. 18) [84].

The formation of nanocrystalline Ce-Yb mixed silicate-type oxyapatite of the composition of Yb_yCe_9.33−y(SiO₄)₆O₂ via the solid-state synthesis was described by Małecka and Kępiński [85]. The phase was identified as an intermediate formed during the synthesis of Ce-Yb silicates.

Fig. 18.
The variation of a and c parameters of Sr_10−xLa_x(PO₄)_6−x(SiO₄)_xF₂ and Sr_10−xLa_x(PO₄)_6−x(SiO₄)_xO phases with the value of x [84].

Different compositions of apatite-type La₁₀Si_6−xW_xO_27+δ ceramics were prepared successfully by Xiang et al [86] via the high-temperature solid-state reaction route. Doping with W⁶⁺ is beneficial to the removal of La₂SiO₅ impurity phase. When the doping content of W⁶⁺ is more than 0.1, the rod-like grains of La₁₀Si_6−xW_xO_27+δ ceramics are replaced gradually by equiaxed apatite-type grains, and randomly shaped convex La₆W₂O₁₅ particles appear at the grain boundaries. While doping with Nb⁵⁺ leads to the hexagonal-phase La_9.5Ge_5.5Nb_0.5O_26.5, the addition of MO⁶⁺ leads to the compound La_9.5Ge_5.5Mo_0.5O_26.75 with triclinic symmetry [87].

5.5 Apatite-type yttrium silicates

5.5.1 Yttrium silicates

The formation of the phase with the composition (Y₄Si₃O₁₂, Y₄(SiO₄)₃ or 2Y₂O₃·3SiO₂ [88]), which is stable between 1650 and 1950°C, was reported by Toropov and Bondar [89] in the binary system of Y₂O₃-SiO₂ and by Toporov and Fedorov [90] in the ternary system of CaO-Y₂O₃-SiO₂ (Fig. 19(a)). This was the first reported occurrence of the phase with the composition between yttrium orthosilicate (Y₂SiO₅, oxyorthodsilicate, YSO, Y₂O₃:SiO₂ = 1:1 [91]) and yttrium disilicate Y₂Si₂O₇ (yttrium pyrosilicate, YPS, 1:3). The structure of this phase was described as the garnet type [88].

Fig. 19.
Phase equilibrium in the system Ca₂SiO₄-Y₄(SiO₄)₃ according to Toporov and Fedorov [90] and Y₂O₃-SiO₂ phase diagram [92].

Since then, authors have disagreed about the existence of such a phase because the attempts to make it starting with yttria and silica powders resulted in the formation of only Y₂SiO₅ and Y₂Si₂O₇ [93]. This phase was not reported either by other studies of Y₂O₃-SiO₂ system [92],[94],[95], which contains two compounds. Y₂SiO₅ andY₂Si₂O₇ were found, with two (A and B) and five (y, α, β, γ and δ, also called y, B, C, D and E [96]) polymorphs, respectively. The first has a congruent melting, whereas the second has an incongruent one (Fig. 19(b)).

Nevertheless, the formation of oxyapatite phase of the composition of Y_4.67□_0.33(SiO₄)₃O (7:9) prepared via the oxidation of nitrogen apatite Y₅(SiO₄)₃N was reported by other authors [7], [97]. The apatite-like phase Y_4.67(SiO₄)₃O possesses hexagonal structure with the space group P6₃/m, a = 9.368 and c = 6.735 Å [78],[98]. The specific gravity of the phase is 4.39 g·cm⁻³ and the hardness on the Mohs scale is 5 – 7 [99].

Since the structure of YSO containing two different types of anions includes the [SiO₄]⁴⁻ complex ion and an additional non-silicon-bonded oxygen ion (NBO), it could be written as Y₂(SiO₄)O. This compound also displays two structure types of monoclinic symmetry with different linking of O-Y₄ tetrahedra. Low-temperature X₁ phase and high-temperature X₂ phase belong to the space groups of P2₁/c (Z = 2) and C2/c (Z = 8), respectively [91].

The samples of the composition of Y₄(SiO₄)₃, and similar ones containing small amount of iron oxide, corresponding to an overall composition of Fe_0.2Y₄(SiO₄)₃O_0.2, were produced by the mixed powder method and by the sol-gel route using yttrium nitrate (Y(NO₃)₃·5H₂O), TEOS (tetraethylorthosilicate) and iron nitrate (Fe(NO₃)₃·9H₂O) by Parmentier et al [7]. Nitrate was dissolved in ethanol/water mixture (volume ratio 7:3), the amount of the latter being controlled to give final Si concentration. Iron nitrate was added at this stage in calculated amounts corresponding to the final iron-doped apatite composition. The solution was stirred for a few hours and TEOS was added to give the appropriate silicon content and then the solution was placed in an oven at 60°C until the gelation occurred. The gel was dried at 80°C and calcined at 600°C for 1 h.

Powders prepared by the two routes were uniaxially pressed into pellets and treated to temperatures up to 1650°C in air in a Pt crucible, or for the heat treatments at 1700°C, carbon element furnace was used, and the samples were heated in a BN-lined crucible in nitrogen atmosphere. Iron appears to have two roles depending on the temperature; it stabilizes the apatite phase at high temperatures when produced by the sol-gel route and catalyzes the decomposition of sol-gel-derived apatite at low temperatures [7].

A new phase of yttrium magnesium silicate having the apatite structure was prepared by Suwa et al [100] at 1500°C in air. Its chemical composition can vary from (Y₄Mg)Si₃O₁₃ to (Y_4.33Mg_1.13)Si_3.34O₁₃. The hexagonal unit cell dimensions a₀ and c₀ of (Y₄Mg)Si₃O₁₃ are 9.298 ± 0.002A and 6.635 ± 0.001A, respectively, and its axial ratio c/a is 0.714. It is optically uniaxial negative with ε = 1.810 ± 0.005, ω = 1.820 ± 0.005 and ω – ε = 0.010. The cleavages parallel and perpendicular to the c-axis were recognized. The formation of the apatite-type phase of the composition of NaY₉Si₆O₂₆ in the ternary system Na₂O-Y₂O₃-SiO₂ was also reported by Lee et al [101].

Fig. 20.
Calculated liquidus surface of the Y₂O₃-Al₂O₃-SiO₂ system: three-phase equilibria with liquid phase (thick lines), liquidus surfaces for various solids (labeled area) and isothermal section (dotted lines, temperature in hundreds °C) [92].

The phase diagram of Al₂O₃-SiO₂-La₂O₃ system (Fig. 20) can be compared with the Y₂O₃-Al₂O₃-SiO₂ ternary diagram examined by Bondar and Galakhov in 1964 [102]. The latter represents the only other example of REE₂O₃-Al₂O₃-SiO₂ phase diagram found so far in the literature. The authors identified the liquidus surface of the whole ternary field, but they failed to elucidate the subsolidus phase relationships among different binary compounds. Due to the much smaller ionic size of Y³⁺ ion with respect to La³⁺ ion (1.18 and 1.015 Å for the eightfold coordination, respectively [103]) and lower bond-valence parameter (2.019 and 2.172 Å [104]), the stability of the binary compositions is substantially altered [57].

The β-alumina-like phase LaAl₁₁O₁₈ is no longer stable, while the garnet-like phase Y₃Al₅O₁₈ and Y₄Al₂O₉ monoclinic compound exist. The lacunar apatite-like phase Y₁₄Si₉O₃₉ reported by Wills et al [105] does not appear in the Y₂O₃-Al₂O₃-SiO₂ ternary diagram; however, a compound with similar Y/Si atomic ratio, namely Y₄Si₃O₁₂, also reported by Wills et al [105], does. Since the formation of Y₄Si₃O₁₂ phase was not confirmed, it may be stabilized by impurities [57],[92].

5.5.2 AM and AEE-yttrium orthosilicate oxyapatites

Alkaline metals (AM) and alkaline-earth element oxyapatites (oxybritholites) are described in this chapter. Phosphate minerals of the apatite supergroup possess strong affinity for strontium [106]. The apatite-type phase of the composition of Sr₂Y₈Si₆O₂₆ (Sr₂Y₈(SiO₄)₆O₂) was prepared by Zuev et al [107] via the two-stage calcination of mixture of SrCO₃, Y₂O₃, Eu₂O₃ and SiO₂ in air in order to investigate the spectral characteristics of Sr₂Y₈(SiO₄)₆O₂:Eu polycrystals. The structure of apatite phase is shown in Fig. 21. The sol-gel synthesis and the characterization of Sr₂Y_8(1−x)Eu_8xSi₆O₂₆ solid solution doped by Eu, where x = 0.01 – 0.4, was described by Karpov and Zuev [108]. The formation of calcium analogue (Ca₂Y₈(SiO₄)₆O₂, calcium-yttrium-silicate oxyapatite) was observed during the crystallization of SiO₂-Al₂O₃-CaO-Na₂O-K₂O-F-Y₂O₃ glass ceramics [109] and during the degradation of advanced environmental barrier coatings [110]. The phase Ca₄Y₆(SiO₄)₆O₂ was often prepared in order to investigate its luminescent properties [111],[112].

Fig. 21.
The structure of Sr₂Y₈(SiO₄)₆O₂ [107].

Fig. 22.
The structure of NaY₉(SiO₄)₆O₂ oxyapatite (the perspective view along the c-axis) [114].

The precipitation of NaY₉(SiO₄)₆O₂ apatite-type compound (sodium nonayttrium hexakis(silicate) dioxide) in the SiO₂-B₂O₃-Al₂O₃-Y₂O₃-CaO-Na₂O-K₂O-F glass-ceramics system (Section 10.3.8) was described by van’t Hoen et al [113]. The hexagonal structure of NaY₉(SiO₄)₆O₂ oxyapatite (Fig. 22) was resolved by Gunawardane et al [114]. The phase crystallizes in P6₃/m space group with the cell parameters a = 9.334, c = 6.759 Å, c:a = 0.7241, V = 509.97 Å and Z = 1).

Lithium yttrium orthosilicate (LiY₉(SiO₄)₆O₂, lithium nonayttrium hexakis(silicate) dioxide) crystallizes in centrosymmetric space group P6₃/m. The structure closely resembles those of fluorine apatite. There are two different crystallographic sites for Y³⁺ ion, which are coordinated by seven and nine O atoms. One-fourth of the nine-coordinated site is occupied by Li atoms, thus maintaining the charge balance. Si atom occupies the tetrahedral site [115].

The preparation, the properties and the effect of sintering additives of hexagonal (P6₃/m) strontium-yttrate-silicate oxyapatite (oxybritholite2) with the composition of SrY₄(SiO₄)₃O as the main product of sinter-crystallization process, in which the non-equilibrium melt was formed in the temperature interval from 1300 to 1550°C in the SrO-Y₂O₃-SiO₂ system, was described by Ptáček et al [116]. The formation of non-equilibrium melt is facilitated by borate fluxes, alkaline fluxes and talc. The apparent activation energy and the frequency factor of the sinter-crystallization process were determined to be 1525 kJ mol⁻¹ and 1.04·10⁴⁵ s^-1, respectively. The material shows low value of linear thermal expansion coefficient of (1.1 ± 0.1)·10⁻⁶°C⁻¹ in the temperature range from 25 to 850°C.

The course of synthesis can be expressed by the following reaction formula [116]:

A E E C O 3 + 2 Y 2 O 3 + 3 S i O 2 → A E E Y 4 S i O 4 3 O + C O 2 g E12

This reaction 12 is too general to describe formed intermediates (SrSiO₃, Sr₂SiO₄, SrY₂O₄, … ¹⁶

Detailed description of formed intermediates can be found in work [116].

) and the process of sinter-crystallization of apatite:

A E E C O 3 + 2 Y 2 O 3 + 3 S i O 2 → int ermediates → n o n − equilibrium melt → A E E Y 4 S i O 4 3 O + C O 2 g E13

Since the formation of SrY₄(SiO₄)₃O proceeds thorough non-equilibrium melt phase, the effect of sintering additives such as borate fluxes, fluorides and carbonates of alkaline metals as well as talc was investigated. Sintering additives facilitate the formation of melt phase and increase the length of sinter-crystallization interval. The expansion after the thermal decomposition of strontium carbonate is reduced as well. Calcinate, treated to the temperature lower than the temperatures of sinter-crystallization interval, has hydraulic activity. Therefore, it can be applied in special composite cements as an activator for latent hydraulic and pozzolanic materials.

Fig. 23.
SEM picture and WDX analysis of hexagonal crystal of CaY₄(SiO₄)₃O apatite phase.

Effect on sinter-crystallization process	Decreasing intensity or temperature of effect →
Expansion before sinter-crystallization	Pure >> NaF >> Talc ≈ Li₂CO₃ ≈ Li₂B₄O₇ >> Na₂CO₃ > LiBO₂ > K₂CO₃
Firing shrinkage (sample treated to 1600°C)	LiBO₂ > Li₂B₄O₇ ≈ Pure > NaF > K₂CO₃ ≈ Na₂CO₃ > Talc > Li₂CO₃
Initial temperature of sinter-crystallization	Pure ≈ Na₂CO₃ ≈ Talc > Li₂B₄O₇ > K₂CO₃ ≈ NaF > LiBO₂ > Li₂CO₃
Maximum rate of sinter-crystallization	Li₂CO₃ > Pure > Li₂B4O₇ ≈ Talc ≈ NaF ≈ K₂CO₃ ≈ Na₂CO₃ > LiBO₂
Length of interval of sinter-crystallization	LiBO₂ >> Li₂B₄O₇ ≈ NaF > Li₂CO₃ > pure > Talc > Na₂CO₃ > K₂CO₃

Table 3.

The influence of sintering additive on the behavior during thermal treatment [116].

After the process of sinter-crystallization, the reactivity of glassy phase with water drops. A significant benefit of talc is the fact that the glassy phase surrounding the crystals of apatite phase becomes resistant against the influence of water with this sintering additive. Furthermore, magnesium is not being incorporated into the structure of apatite phase during the crystallization of SrY₄(SiO₄)₃O from non-equilibrium melt. The influence of sintering additives on the behavior during the thermal treatment is summarized in Table 3 [116].

The important feature of this compound is the formation of colored center after the exposition to X-ray radiation (Fig. 24); hence, the prepared material is an important candidate for optical applications, sensors and dosimeters.

Fig. 24.
Semitransparent disc of SrY₄(SiO₃)₄O sintered specimen (a) and formation of colored center on X-ray irradiated area (Cu(Ka), 40 kV and 30 mA) [116].

On the other hand, this reaction also indicates that the synthesis of individual apatite analogues (AEEY₄(SiO₄)₃O, where AEE = Ca, Sr and Ba) and their solid solutions proceeds via similar ical intermediates formed in the temperature range, which is affected by the thermal stability of AEE carbonates that increases in the order: CaCO₃, SrCO₃ and BaCO₃.

Fig. 25.
SEM picture of hexagonal crystal of CaY₄[SiO₄]₃O apatite phase.

While the synthesis of CaY₄(SiO₄)₃O leads to well-developed hexagonal crystals (Fig. 25), the attempts for the preparation of BaY₄(SiO₄)₃O phase were not successful. This synthesis leads to well-developed crystals of yttrium orthosilicate (Y₂SiO₅) surrounded by BaO-Y₂O₃-SiO₂ glassy phase (Fig. 26).

Fig. 26.
SEM of Y₂SiO₅ crystal surrounded by barium-containing glassy phase.

The investigation of this system leads to the conclusion that non-limited Ca²⁺ ↔ Sr²⁺ substitution can be performed in the binary system of (Ca-Sr)Y₄[SiO₄]₃O. On the contrary, the BaY₄[SiO₄]₃O analogue of CaY₄[SiO₄]₃O and SrY₄[SiO₄]₃O apatite cannot be prepared; therefore, the extent of Ca²⁺ ↔ Ba²⁺ and Sr²⁺ ↔ Ba²⁺ substitutions is limited to 28 ± 4 and 38 ± 4%, respectively. The field of ternary solid solutions in the AEEY₄[SiO₄]₃O system, where AEE = Ca, Sr and Ba, is shown in Fig. 27.

Fig. 27.
The miscibility in the AEE Y₄(SiO₄)₃O system.

5.5.3. N-apatite

The main secondary phases in Jänecke prism¹⁷

Jänecke prism is used to visualize the phase relationships among α-sialon, β-sialon and other phases in the M-Si-Al-O-N system. α- and β-sialons are isostructural with α- and β-Si₃N₄, respectively. The substitution of Al-O for Si-N in β-Si₃N₄ yields β-sialon with general formula: Si_6−xAl_zO_zN_8−z (0 < z < 4.2). The structure is built up by Si and Al tetrahedra coordinated with oxygen and nitrogen. The unit cell contains two Si₃N₄ units. α-sialons are solid solutions based on the α-Si₃N₄ structure, with the general formula: M^p+_xSi_12−(m+n)Al_(m+n)O_nN_16−n, where M is metal ion such as Li, Ca, Ba, Y and RE with a valence of p+ and index m = px [117].

[117] for Si₃N₄-Al₂O₃-SiO₂-Y₂O₃-YN-AlN system ¹⁸

Yttria is often used additive to improve the sintering behavior of Si₃N₄ [124].

are shown in Fig. 28. The formal exchange of oxygen by nitrogen leads to the compounds of N-apatite (Y₁₀(SiO₄)₆N₂, H-phase ¹⁹

In dependence on the system composition, the general composition of H-phase (N-apatite) can be written as (M,REE)₁₀(SiO₄)₆N₂. The specification of cations then leads to the names such as Mg-Nd-N-apatite [123].

), N-melilite ²⁰

The melilite-type structure (tetragonal mineral melilite ((Ca,Na)₂(Al,Mg,Fe²⁺)(Si,Al)₂O₇)) is sorosilicate from the group of melilite, first described (Capo di Bove, near Rome, Italy) in 1976 and named from the Greek words meli “honey” and lithos “stone”). Y₂Si₃O₃N₄ was described by Fang et al [118]. N atoms fully occupy the bridging site (2c) and O atoms fully occupy the terminal site (4e) with 2 O and 6 N atoms at the bridging 8f site. The preferential distribution of O and N atoms at the 8f site results in two different local coordinations of Y and three different types of Si atoms.

[118] (Y₂Si₃O₃N₄, M-phase), N-wollastonite (YSiON₂, K-phase) and N-woehlerite (Y₄Si₂O₇N₂, J-phase). The latter one forms a complete solid solution with Y₄Al₂O₉ (YAM) of the composition of Y₄Si_2−xAl_xO_7+xN_2−x (Jss-phase) [119],[120],[121],[122],[123],[124].

Fig. 28.
Representation of the Si₃N₄-Al₂O₃-SiO₂-Y₂O₃-YN-AlN phase equilibrium.

Y₁₀(SiO₄)₆N₂ (N-apatite, H-phase, (Y,Si,□)₁₀[Si(O,N)₄]₆(O,N,□)₂ [125], silicon-yttrium oxynitride) was first identified by Rae et al [126] as a compound with the compositional mixture of 10Y₂O₃·9SiO₂·Si₃N₄ that was stable up to 1750°C. There were other suggested compositions, such as Y₁₀Si₇O₂₃N₄ [127]. Later work by Gauckler et al [128] established N-apatite as a stoichiometric compound with the formula unit of Y₁₀[SiO₄]₆N₂ and the apatite structure (space group P6₃/m). The lattice constants of the hexagonal cell were reported to be a = 9.638 Å and c = 6.355 Å. The electronic structure and bonding of the complex ceramic crystal Y₁₀(SiO₄)₆N₂ was studied by Ching et al [121]. This crystal is an insulator with direct band gap of 1.3 eV. It has some unique properties related to one-dimensional chain structure in the c-direction and planar N-Y bonding in the x-y plane.

The ternary phase diagrams of the Si₃N₄-Y₂O₃-SiO₂ [123] and Si₃N₄-La₂O₃-SiO₂ systems [129] are shown in Fig. 29(a) and (b). The apatite phase is able to form various solid solutions that may influence the development of strength in silicon nitride densified by yttria [130].

Fig. 29.
Phase relationships in the systems Si₃N₄-SiO₂-Y₂O₃ [123] (a) and Si₃N₄-SiO₂-La₂O₃ at 1700°C and 1550°C (dashed lines) [129] (b).

The hexagonal lanthanum N-apatite phase of the composition of La₅(SiO₄)₃N (isostructural with apatite) can be prepared from the mixture of La₂O₃ and Si powder sintered at temperatures in the range from 900 to 1200°C under the flow of nitrogen. The melting temperature of this phase was determined to be ~1600°C. It was observed that continuous heating and addition of Pd into the reaction mixture favored the formation of La₅(SiO₄)₃N. Prolonged heating of this compound yields La_4.67(SiO₄)₃O [129],[131],[132],[133],[134],[135]. The absorption bands observed in infrared spectrum of lanthanum oxynitrides are introduced in Table 4.

Wavenumber [cm⁻¹]	Mode	Frequency [cm⁻¹]	[cm⁻¹]
225	δ (Si-O) (A₁)	730	ν (Si-N)
270	Si-O₂	800	SiO₂
327	δ (Si-O) (B₂)	840	Si-N
337	Si-O-Si (A₁)	872	SiO₄ (ν₁)
380	SiO₄ (ν₄)	909	ν (Si-N) vs
376 – 385	δ (Si-N) (sh)	930	Si-O-N_x
396	SiO₄ (ν₄)	940	ν Si-N vs
432	δ (Si-N) (sh)	960
448	Si-N s or O-Si-O bend.	905
462	SiO₄ (ν₄)	934	SiO₄ (s) (ν₃)
475	SiO₄ (ν₄)	978	SiO₄ (s) (ν₃)
490	ν Si-O (A₁)	980	ν Si-N (s)
542	ν Si-N (s)	1060	ν Si-O (ν₃)
600	ν Si-O	1090	ν Si-N (sh)
648	ν Si-O (w)	1130	ν Si-O (B₁)
679	ν Si-O m (B₂)	—	—

Table 4.

The absorption bands observed in infrared spectrum of lanthanum oxynitrides [131].

Mitomo et al [129] used the mixture of Si₃N₄ and La₂O₃ powder heated in a 10 mm diameter carbon die using a hot-pressing apparatus. In some compositions and at temperatures above 1600°C, liquid phases were squeezed out of the die by applied pressure resulting in a change in the overall composition. The compacted powder was therefore heated at the pressure of 20 MPa up to 1400°C and the temperature was then raised to 1700°C without the pressure. The specimen was kept at 1700°C for 1 h and then allowed to cool.

The sintering temperature of Si₃N₄ with La₂O₃ additions is 1700 to 1800°C. Heating of powder mixture of various Si₃N₄/L₂O₃ ratios at 1700°C results in the formation of 2Si₃N₄·La₂O₃, La₅(SiO₄)₃N ²¹

The presence of La₅(SiO₄)₃N is inevitable in the production of high-density materials by liquid-phase sintering; therefore, the amount of La₅(SiO₄)₃N and glassy phase must be minimized to obtain materials with good high-temperature strengths [129].

and β-Si₃N₄ or a glass. The reactions occurring during heating were determined as follows [129]:

S i 3 N 4 + L a 2 O 3 → 1200 − 1250 ° C S i 3 N 4 + L a 4 S i 2 O 7 N 2 + LaSi O 2 N → 1400 − 1500 ° C LaSi O 2 N + S i 3 N 4 → 1650 − 1750 ° C 2 S i 3 N 4 ⋅ L a 2 O 3 + liquid → 20 ° C 2 S i 3 N 4 ⋅ L a 2 O 3 + L a 5 S i O 4 3 N + glass E14

The results of Sakai et al [136] indicate that N-apatite and N-diopside containing grain boundary phase may improve the oxidation resistance of silicon nitride. Since the oxidation of Si₃N₄ leads to the formation of protective SiO₂ layer on the surface:

S i 3 N 4 + 3 O 2 → 3 S i O 2 + 2 N 2 E15

the silicon nitride shows excellent oxidation resistance. Formed SiO₂ then reacts with the grain boundary constituents to form silicates:

S i O 2 + Me O x → MeSi O 2 + x E16

but in the case of MgO, the formed layer did not act as protection [136]. Y_4.67(SiO₄)₃O apatite (britholite phase²) is formed as the oxidation product of silicon yttrium oxynitride (H-phase) in the temperature range from 700 to 1400°C [137],[138].

5.6. REE vanadocalcic apatite

The synthesis and physicochemical study of rare-earth-containing vanadocalcic oxyapatites where the pair Ca²⁺ and □ was substituted by Ln³⁺ and 1/2O²⁻ was described by Benmoussa et al [139]. This substitution leads to lanthanum and praseodymium dioxyapatites Ca₈Ln₂(VO₄)₆O₂, where Ln = La and Pr. Regarding rare earths such as neodymium, samarium, europium, gadolinium and terbium, the Ln³⁺ content limit varies from one REE to another. It decreases when the REE ionic radius declines.

5.7. Apatite-type yttrium phosphates

The following compositions having the apatite structure were prepared by Wanmaker et al [140]:

Me(II)_2+xMe(III)_8−x(SiO₄)_6−x(PO₄)_xO₂, where 0 ≤ x ≤ 6;
Me(II)_3+xMeIII_6−x(SiO₄)_6−x(PO₄)_x, where 0 ≤ x ≤ 1.5;
Me(II)_4+xMe(III)_6−x(SiO₄)_6−x(PO₄)_xO, where 0≤ x ≤ 6.

with Me(II) = Ca, Sr, Ba, Mg, Zn or Cd and Me(III) = Y or La. Among these, there are several new compounds, e.g. Zn₂La₈(SiO₄)₆O₂, BaMgY₈(SiO₄)₆O₂, Zn₂Y₈(SiO₄)₆O₂, Cd₂Y₈(SiO₄)₆O₂, Ca₄La₅(SiO₄)₅(PO₄) and Ba₄La₅(SiO₄)₅(PO₄). The crystallographic parameters were determined and their luminescence was studied. The most efficient activator proved to be trivalent antimony, especially in the compositions of type I. At 77°K, an emission band at about 400 nm was observed in many of these apatites.

The humidity-sensitivity of yttrium-substituted calcium oxyhydroxyapatites was studied by Owada et al [141]. The logarithm of the electrical resistance of present sensors decreased linearly with increasing relative humidity (RH) from 30 to 65%. The resistance of [Ca_9.0Y_1.0](PO₄)₆[O_1.5□_0.5] with the largest OH vacancy content was about one order of magnitude lower than that of calcium hydroxyapatite. It was found that the larger the ratio of surface hydroxyl groups per unit surface area in the sample, the lower the resistance and the higher the amount of OH vacancies.

A ceramic proton conductor was obtained in the solid solutions of yttrium-substituted oxyhydroxyapatite (Ca_10−xY_x)(PO₄)₆((OH)_2−x−2yO_x+y□_y) [142]. Using the hydrogen concentration cells, it was confirmed that the specimens with the composition of x ≤ 0.65 have the protonic transference number (t_i) is equal to one, while the values of t_i of specimens with 0.65 < x < 1 were smaller than one. The conduction properties were also dependent on the composition of apatites. At x = 0.65, the conductivity (σ) showed the maximum value (5·10⁻⁴S·cm⁻¹ at 800°C) in the relationship between σ and x, while the activation energy was the lowest (about 1.0 eV) at corresponding x. The applicability of proton conductive apatite for a fuel cell was discussed in Section 10.4.

Fig. 1.

5.1 Apatite-type lanthanium silicates

Table 1.

Fig. 2.

Table 2.

Fig. 3.

Fig. 4.

Fig. 5.

Fig. 6.

Fig. 7.

5.2 Apatite-type lanthanium germanates

Fig. 8.

Fig. 9.

Fig. 10.

Fig. 11.

Fig. 12.

5.3 Apatite-type borates

Fig. 13.

Fig. 14.

5.4 Other apatite-type REE silicates

Fig. 15.

Fig. 16.

Fig. 17.

Fig. 18.

5.5 Apatite-type yttrium silicates

5.5.1 Yttrium silicates

Fig. 19.

Fig. 20.

5.5.2 AM and AEE-yttrium orthosilicate oxyapatites

Fig. 21.

Fig. 22.

Fig. 23.

Table 3.

Fig. 24.

Fig. 25.

Fig. 26.

Fig. 27.

5.5.3. N-apatite

Fig. 28.

Fig. 29.

Table 4.

5.6. REE vanadocalcic apatite

5.7. Apatite-type yttrium phosphates

Erik Vakil

Chapters authored

Related collaborators

Edna Reiche

Luciana Costa

Andrea Name Colado Simão

Thatiane Sampaio

Luiza Mendonça

Nitin Saksena

Dominic Dwyer

Bin Wang

Leonor Huerta

Cesar Noe Cortes-Rubio