Recent progress on synthesis,multi-scale structure,and properties of Y–Si

Abstract

Yttrium silicates (Y–Si–O oxides), including Y₂Si₂O₇, Y₂SiO₅, and Y_4·67(SiO₄)₃O apatite, have attracted wide attentions from material scientists and engineers, because of their extensive polymorphisms and important roles as grain boundary phases in improving the high-temperature mechanical/thermal properties of Si₃N₄ and SiC ceramics. Recent interest in these materials has been renewed by their potential applications as high-temperature structural ceramics, oxidation protective coatings, and environmental barrier coatings (EBCs). The salient properties of Y–Si–O oxides are strongly related to their unique chemical bonds and microstructure features. An in-depth understanding on the synthesis – multi-scale structure-property relationships of the Y–Si–O oxides will shine a light on their performance and potential applications. In this review, recent progress of the synthesis, multi-scale structures, and properties of the Y–Si–O oxides are summarised. First, various methods for the synthesis of Y–Si–O ceramics in the forms of powders, bulks, and thin films/coatings are reviewed. Then, the crystal structures, chemical bonds, and atomic microstructures of the polymorphs in the Y–Si–O system are summarised. The third section focuses on the properties of Y–Si–O oxides, involving the mechanical, thermal, dielectric, and tribological properties, their environmental stability, and their structure–property relationships. The outlook for potential applications of Y–Si–O oxides is also highlighted.

Introduction

Yttrium silicates, or the Y₂O₃–SiO₂ binary system (Fig. 1a ), including Y₂Si₂O₇, Y₂SiO₅, and 2Y₂O₃.3SiO₂, were first discovered from an yttrialite ore yard located in Texas, USA in 1889. Thereafter, similar yttrialite mines were discovered in different places around the world, such as Japan and Russia.^1–3 Although the chemical compositions for these yttrialites were the same for the ores taken from different mines, the reported crystal structures showed a wide diversity. Moreover, the crystal structure also varied depending on the heating history.^1–4 The crystal structure diversity of yttrialites was found to be the result of polymorphisms. The extensive polymorphisms of yttrialite attracted the strong interest of material scientists, such that the polymorphisms and the crystal chemistry of Y–Si–O oxides were widely studied.^4–8 Bondar and Toropov⁹ pointed out that there were three ternary crystalline phases, namely, Y₂SiO₅, Y₂Si₂O₇, and 2Y₂O₃.3SiO₂, with melting points of 1980, 1950, and 1750°C, respectively, in the binary Y₂O₃–SiO₂ system. However, according to the later investigations, 2Y₂O₃.3SiO₂ was stable only in the presence of other divalent or monovalent cations.¹⁰ Later, an apatite structure, Y_4·67(SiO₄)₃O (2·34Y₂O₃.3SiO₂), was determined as the dashed line shown in Fig. 1a . In the solid-state synthesis of Y₂SiO₅, Y_4·67(SiO₄)₃O apatite always appears as second phase.

Phase diagrams of a Y₂O₃–SiO₂ binary system,^23,24 and b Y₂O₃–SiO₂–LiO₂ ternary system⁶⁶

In 1968, Ito and Johnson¹¹ reviewed the crystal structures, X-ray diffraction (XRD) data, polymorphism transformations, and the differences between natural and synthesised yttrium disilicates (Y₂Si₂O₇), wherein six polymorphs were reported, namely, α-, β-, γ-, δ-, y-, and z-Y₂Si₂O₇. These phases could be transformed from one to another at varying temperatures: . After that, a number of reviews on the crystal structure of Y–Si–O oxides were published. Felsche¹² summarised and corrected the phase relationships and the crystal chemistry of rare-earth silicates in 1973; Liddell and Thompson¹⁰ collected and revised the XRD data on Y₂Si₂O₇ in 1986; Becerro and Escudero¹³ reviewed the crystallographic data, including the XRD data, and the ⁸⁹Y and ²⁹Si magic angle spinning – nuclear magnetic resonance (MAS-NMR) spectra, for Y₂SiO₅ and Y₂Si₂O₇ in 2004.

The high melting points of Y–Si–O oxides enable their potential applications as high-temperature structural materials or multifunctional coatings. Recent interest in yttrium silicates was renewed by the roles they play at the grain boundaries of Si₃N₄ and SiC. It was found that the high-temperature mechanical properties and the thermal conductivities of Si₃N₄ and SiC were dramatically improved when yttrium silicates were formed as grain boundary phases.^14–19

In this review, recent progress on novel preparation methods, and the multi-scale structures and properties of Y–Si–O oxides (Y₂Si₂O₇, Y₂SiO₅, and Y_4·67(SiO₄)₃O apatite) are summarised. The first part focuses on synthesis of powders, sintering of bulk ceramics and deposition of coatings/films. Owing to the complex polymorphism of yttrium silicates^{11–13,20–22} and the strict stoichiometric ratio required for the starting materials,^23,24 the synthesis of single-phase yttrium silicates is a major obstacle to the research and application of these oxides. At present, the main preparation methods employed for powders are sol-gel (including the precipitation technique),^25–34 hydrothermal synthesis,^35–38 solid–solid reaction,^39–42 solid–liquid reaction,^43–47 high-P/high-T synthesis,²¹ and molten salt flux growth (or the Czochralski technique);^{20,39,48–51} and for coatings and thin films are chemical vapour deposition (CVD), atomic layer deposition (ALD), radio frequency (rf)-sputtering, magnetron sputtering, pulsed laser deposition (PLD), plasma spraying, electrophoretic deposition (EPD), spin coating, and dip coating,^40,52–57 etc. The features and merits of these methods are summarised in this section.

The second section introduces the crystal structure, chemical bonding, and microstructural characteristics of Y–Si–O oxides. Recently reported crystal structure data of ζ-Y₂Si₂O₇, η-Y₂Si₂O₇, and Y_4·67(SiO₄)₃O apatite are collected. Confusion about the structures of yttrium silicates is also clarified.

In the third section, properties of Y–Si–O ceramics are reviewed. Some yttrium silicates, such as γ-Y₂Si₂O₇ and Y₂SiO₅, have low shear deformation resistance and are tolerant to mechanical damage.^58,59 γ-Y₂Si₂O₇ exhibits low thermal expansion coefficient (TEC, 3·90×10⁻⁶ K⁻¹) and low thermal conductivity (<3·0 W m⁻¹ K⁻¹ above 600 K), while Y₂SiO₅ presents a slightly higher TEC (8·36×10⁻⁶ K⁻¹), but extremely low thermal conductivity (1·34 W m⁻¹ K⁻¹).^60,61 Moreover, Y–Si–O ceramics have been proven to have excellent resistance to hot corrosion, even in strongly basic Na₂CO₃ molten salt,^62,63 which indicates that Y–Si–O ceramics are promising candidates for oxidation-resistance/thermal barrier/environmental barrier coatings (EBCs). Defect formation in Y–Si–O oxides^64,65 will also be presented. Finally, the outlook for potential applications of the Y–Si–O oxides is highlighted in the last section.

Preparation of Y–Si–O oxides

Synthesis of powders and sintering of bulk ceramics

In the binary Y₂O₃–SiO₂ phase diagram, Y₂SiO₅, Y₂Si₂O₇, and Y_4·67(SiO₄)₃O, are presented as straight lines (Fig. 1a ),^10,23,24 thus the synthesis of single-phase Y–Si–O compounds needs very strict control of the stoichiometric Y/Si ratio. The extensive polymorphisms also hinder the preparation of single-phase Y–Si–O oxides. Low-temperature polymorphs may appear during cooling. For example, after heating sol-gel prepared Y₂Si₂O₇ powders to 1200°C, a mixture containing α-Y₂Si₂O₇ and β-Y₂Si₂O₇ was obtained.²⁶ The sintering of nearly theoretically dense Y–Si–O bulk is also a challenge because of their poor sinterability and the temperature-dependent phase transformations.

Up to now, the common synthesis methods for Y–Si–O powders are: (i) sol-gel,^25–34 (ii) hydrothermal,^35–38 (iii) solid-state reaction,^28,39–42 (iv) solid–liquid reaction,^43–46 (v) molten salt fluxing growth,^{20,39,48–51} (vi) high-P/high-T synthesis,²¹ etc. For bulk Y–Si–O oxides, the Czochralski technique is used for single-crystals, while pressureless sintering and hot pressing are employed for polycrystalline samples.^45–51 Table 1 summarises the typical methods, processing conditions, and predominant products for the fabrication of Y–Si–O compounds. The overall procedures for these preparation methods are generally described as follows.

Table 1.

Summary of typical preparation methods for Y–Si–O oxides

Method	Processing condition	Contribution	References
Sol-gel	Spherical silica cores from hydrolysis of TEOS coated by yttrium basic carbonate, or vice versa. After ageing, drying and compaction, samples were sintered at temperatures from 1000–1600°C	Y₂Si₂O₇+silica or Y₂Si₂O₇+Y₂O₃ core-shell structure for Y₂Si₂O₇, 1000–1300°C: α-Y₂Si₂O₇; 1400–1500°C: β-Y₂Si₂O₇; 1600°C: γ-Y₂Si₂O₇	Giesche et al.²⁵
	Y(NO₃)₃+TEOS precursors, gelation and heating at temperatures higher than 1050°C	Single-phase α-Y₂Si₂O₇ powders	Diáz et al.²⁶
		Single-phase α-Y₂Si₂O₇ powders	Kahlenberg et al.²⁷
	Y(NO₃)₃+TEOS precursors, gelation for 3 weeks	Mainly α-Y₂Si₂O₇ phase at 1200°C, transformed to β- and then γ-Y₂Si₂O₇ between 1400–1600°C	Parmentier et al.²⁸
	(Y₂O₃+HNO₃) +TEOS precursors, gelation at 70°C, then heated at 1050°C for 3 h	Phase-pure X1-Y₂SiO₅ powders	Wang et al.²⁹
	(YCl+PrOH)+(K-PrOH) mixed solution+TEOS, gelation, then heated at 1500°C for 4h	High-purity Y₂SiO₅ powders	Boyer et al.³⁰
	Y(NO₃)₃+TEOS+Eu(NO₃)₃ precursors, gelation at 50°C for 3 days	Amorphous gel to crystal transformation occurred at 900–1100°C, X1-Y₂SiO₅ transformed to X2-Y₂SiO₅ over 1150–1200°C	Cannas et al.³¹
	YCl₃+EuCl₃+aerogel silica precursors	60 nm Y₂Si₂O₇:Eu at 300–600°C, but decomposed into Y₂O₃:Eu and silica at 700–900°C	Zhou et al.³²
	Spherical silica cores from hydrolysis of TEOS coated by Y₂SiO₅ precursors, then heated at 1000°C for 2 h	SiO₂/Y₂SiO₅:Eu, SiO₂/Y₂SiO₅:Ce/Tb core-shell structures	Lin et al.³³
	Y(NO₃)₃+TEOS+Fe(NO₃)₃ precursors, then heated to 1650°C or 1700°C	Fe-doped Y_4·67(SiO₄)₃O powders	Parmentier et al.³⁴
Sol-gel+hydrothermal	Y(NO₃)₃+TEOS solution, refluxed at 80°C for 6 h, then heat treated at 900–1600°C	Y-, α-, β-, and γ-Y₂Si₂O₇ formed from the precursors by heating at different temperatures	Diáz et al.³⁶
Hydrothermal synthesis	Y(CH₃CO₂)₃+TEOS precursors, pH set at 1, 6·1 and 10·2, respectively, then hydrothermally treated at 170°C for 2 h under autogenous pressure	β-Y₂Si₂O₇ powders at 1300°C	Trusty et al.³⁷
		γ-Y₂Si₂O₇ powders at 1500°C	Trusty et al.³⁷
	Y(NO₃)₃+saponite powder solution heated at 365°C for 6 days in a hydrothermal bomb	y-Y₂Si₂O₇+trace starting powders	Becerro et al.³⁵
	Y(NO₃)₃+smectite powder solution heated at 365°C for 6 days	y-Y₂Si₂O₇ powders, transformed to β-phase at 1050°C	Becerro et al.³⁸
Solid-state reaction	Y₂O₃+SiO₂ starting materials heated to 1100–2200°C for 2 h–68 days	X2-Y₂SiO₅, δ-Y₂Si₂O₇, γ-Y₂Si₂O₇ powders with apatite, unreacted raw materials or other polymorphous impurities	Christensen et al.³⁹
			Ogura et al.⁴¹
			Parmentier et al.²⁸
			Seifert et al.⁴⁰
			Kumar and Drummond⁴²
	Y₂O₃+SiO₂+Fe(NO₃)₃ mixture heated to 1650°C or 1700°C	Fe-doped Y_4·67(SiO₄)₃O with minor Y₂Si₂O₇ impurity powders	Parmentier et al.³⁴
Solid-liquid state reaction	Y₂O₃+SiO₂ starting materials with additives heated to 1100–1600°C	Y₂SiO₅ and γ-Y₂Si₂O₇ powders	Leskelä and Jyrkäs⁴³
			Aparicio et al.⁴⁴
			Sun et al.^46,47
Molten salt flux growth (Czochralski growth technique)	Y₂O₃+SiO₂ starting materials+flux oxides in Ir or Pt crucible heated to form melt, then single-crystal Y₂SiO₅ or Y₂Si₂O₇ was grown from the seed	Y₂SiO₅ and β-, γ- and ζ-Y₂Si₂O₇ single-crystals	Christensen et al.³⁹
			Leonyuk et al.⁴⁸
			Zhang et al.⁴⁹
			Günther and Georg⁵⁰
			Pang et al.⁵¹
			Hartenbach et al.²⁰
Solution combustion	Y(NO₃)₃+SiO₂+CH₆N₄O starting materials were dissolved in solution and boiled at 500°C until they spontaneously burst into flames; then the ashes were annealed at 1350°C	High surface area fine Y₂SiO₅ powders	Bosze et al.⁶⁹
High-energy ball milling	Y₂O₃ or hydrous yttria+SiO₂, ball milled in a high-energy planetary ball mill, then heated at 1100°C	β-Y₂Si₂O₇ powders were obtained after ball milling and heat treatment at 1100°C for 2 h	Tzvetkov and Minkova⁷⁰
Phonochemical method	Y(NO₃)₃+Na₂SiO₃+NaOH aqueous solution put in an ultrasonic generator at 200 W for 60 min, then heated at 900°C for 2 h	Nano-size single-phase Y₂SiO₅, Y_4·67(SiO₄)₃O, and Y₂SiO₅ powders	Huang et al.⁷¹
Pechini method	Aqueous solution of ethyleneglycol, citric acid, Y(NO₃)₃, and TEOS or SiO₂, heated to evaporate nitrogen oxide, then heated to 1100°C	Y₂SiO₅ and Y₂Si₂O₇ powders	Chr. Argirusis et al.⁷²
Hot pressing	Hydrothermal synthesised Y₂Si₂O₇ precursor mixed with LiF fluxing agent, then hot pressed at 10 MPa and 1600°C	86% of theoretical density γ-Y₂Si₂O₇ bulk materials	MacLaren et al.⁴⁵
Hot pressing	Y₂SiO₅ powders synthesised from mixture of Si- and Y-alkoxides, then hot pressed at 1400–1425°C under 27 MPa for 1–4 h in Ar, and then annealed at 1400–1800°C for 1–24 h in air	93·9–99·7 of theoretical density X2-Y₂SiO₅+Y₂O₃	Ogura et al.⁷⁴
High P/High T	Synthesis was performed at 6·0 GPa and 1350°C after decarbonation of Y₂O₃+SiO₂+Na₂CO₃ mixture	Single-crystal η-Y₂Si₂O₇ particles	Kahlenberg et al.²¹
Solid-liquid pressureless sintering	Solid–liquid reaction synthesised γ-Y₂Si₂O₇ and Y₂SiO₅ powders, dry pressed, followed by pressureless sintering at 1400–1500°C.	96–99% theoretical density γ-Y₂Si₂O₇ and Y₂SiO₅ bulk materials	Sun et al.^46,47
Colloidal processing	Y₂Si₂O₇ or Y₂SiO₅ dispersed in distilled water or ethanol to form stable slurry, and then slip-casted, with pressureless sintering at 1400°C	Highly textured and microstructure controlled dense γ-Y₂Si₂O₇ bulk materials	Sun et al.^78,79
Colloidal processing	Y₂O₃+SiO₂ binary suspension was prepared, then slip-casted and sintered at 1600°C for 3–10 h	65% of theoretical density of Y₂SiO₅ from Y₂O₃–SiO₂ suspension, and 90% Y₂SiO₅ bulk obtained with Al₂O₃ sintering aid	Aparicio et al.⁴⁴

TEOS: tetraethyl orthosilicate.

Sol-gel synthesis of Y₂Si₂O₇ and Y₂SiO₅ ^25–34

Usually, a solution of tetraethyl orthosilicate (TEOS) is added to an aqueous solution of Y(NO₃)₃ to form a transparent sol. Then, a homogeneous Y₂O₃/SiO₂ gel is obtained by drying the sol at temperatures below 600°C. By calcining the gel at 600–1050°C, phase-pure α-Y₂Si₂O₇ or X1-Y₂SiO₅ can be obtained. X2-Y₂SiO₅ can be obtained by further heating the sol-gel prepared X1-Y₂SiO₅ powders to higher temperatures. Via the sol-gel method, Fe-doped Y_4·67(SiO₄)₃O apatite (or Fe_0·2Y₄(SiO₄)₃O_0·2) has also been successfully synthesised from yttrium nitrate, iron nitrate, and TEOS solution.³⁴ The apatite is very stable at ∼1700°C in nitrogen atmosphere, but it becomes metastable in air even at lower temperatures (950–1150°C), and decomposes into Y₂Si₂O₇ and Y₂SiO₅.³⁴

The advantages of the sol-gel method include easy composition control, low synthesis temperature, and suitability for coating on different substrates. For Y₂Si₂O₇, however, preparing single-phase β-, γ-, or δ-Y₂Si₂O₇ is still a challenge. The expensive precursors are also a challenge for volume production.

Hydrothermal method

Single-phase y-Y₂Si₂O₇ can only be synthesised in the presence of a small amount of foreign metal cations, such as Na⁺, K⁺, Mg²⁺, Ca²⁺, Al³⁺, etc., via the hydrothermal method.^35–38 In this process, layered clay, saponite or smectits, is suspended in Y(NO₃)₃ solution, and then heated to 300°C or above in a hydrothermal tank for several days. The solid product collected by filtering was y-Y₂Si₂O₇.^35,38

This route requires extended processing time, on the order of 100 days at 300°C. Autoclaving time can be reduced to approximately 11 days at the high processing temperatures. The long reaction time and the difficulty in product separation from the unreacted solid raw components are the main drawbacks of this method.

Solid-state reaction method

For solid-state reaction synthesis of Y₂Si₂O₇ and Y₂SiO₅, Y₂O₃ and SiO₂ powders are mixed by ball milling and then heated at high temperatures.^28,39–42 High reaction temperatures (usually in the range of 1300–2200°C) and long reaction times (several to hundreds of hours)⁴² are needed. Even so, the synthesis of single-phase γ- or δ-Y₂Si₂O₇ is not easy. One reason is the slow inter-diffusion between the Y₂O₃ and SiO₂ particles, which leads to Y_4·67(SiO₄)₃O, Y₂O₃, or SiO₂ being left as impurities.^34,42 The other reason is that the low-temperature polymorphs appear during the cooling process, unless a very fast quenching rate is employed.⁴² Fe-doped Y_4·67(SiO₄)₃O apatite (or Fe_0·2Y₄(SiO₄)₃O_0·2) was also synthesised by the solid-state reaction of Y₂O₃ and SiO₂ powders.³⁴

Solid-liquid reaction method

Solid–liquid reaction is characterised by lower synthesis temperature and shorter reaction time, since the formation of a low-temperature liquid phase accelerates the mass transportation and improves the chemical reaction between the starting particles.^43–46 Solid–liquid reaction inherits the merits of solid-state reaction, such as low-cost raw materials, simple equipment, and feasibility for volume production.

Leskelä and Jyrkäs⁴³ studied the solid–liquid reaction of Y₂O₃–SiO₂ using different fluxing reagents such as LiF, NaF, PbF₂, NH₄F, NaOH, NaNO₃, MgO, CaO, MgSO₄, and KNO₃, in which lithium compounds, especially lithium fluoride, gave the highest conversion efficiency. High-purity Y₂SiO₅ was obtained at 1100°C with 3 wt.-% LiF. Later, Aparicio et al. ⁴⁴ prepared bulk Y₂SiO₅ with 90% of theoretical density by colloidal processing followed by sintering at 1600°C for 3 h using Al₂O₃ as a sintering aid. MacLaren et al. ⁴⁵ synthesised a γ-Y₂Si₂O₇ with 86% of theoretical density by employing a 3 wt.-% LiF fluxing additive with hot pressing at 1600°C under 10 MPa.

Sun et al. ^46,47 prepared single-phase γ-Y₂Si₂O₇ and X2-Y₂SiO₅ powders and bulk materials via the solid–liquid reaction method using LiYO₂ as additive at 1400–1500°C. From the Y₂O₃–SiO₂–LiO₂ phase diagram (Fig. 1b ), the reaction temperature between Y₂O₃ and SiO₂ can be lowered by eutectic reactions induced by LiO₂.⁶⁶ Sun et al. observed the formation of Li-containing liquid phase at ∼1000°C.^46,47 Figure 2 presents the significance of the solid–liquid reaction in the synthesis of Y–Si–O oxides by comparing the differential scanning calorimetry (DSC) of the Y₂O₃/SiO₂ and Y₂O₃/SiO₂/LiYO₂ mixtures (Fig. 2a ), and the corresponding XRD patterns after DSC test (Fig. 2b ).^46,47 Without LiYO₂, exothermic peaks only resulted from crystallisation of SiO₂. The addition of LiYO₂ changed the reactions between the Y₂O₃ and SiO₂. γ-Y₂Si₂O₇ and X2-Y₂SiO₅ were respectively detected in the XRD patterns in the samples with different Y₂O₃/SiO₂ ratios. Figure 2c–d shows the XRD patterns of single-phase γ-Y₂Si₂O₇ (Fig. 2c ) and Y₂SiO₅ (Fig. 2d ) synthesised via the solid–liquid reactions at 1400 and 1500°C, respectively. Sun et al. pointed out the direct use of LiO₂ might not be as effective as LiYO₂ because LiO₂ might be evaporated at high temperatures.⁶⁷ LiYO₂ presents a lower vapour pressure, only partially evaporates at 1600°C⁶⁸ and had been used in the liquid phase sintering of Si₃N₄.

a Differential scanning calorimetry (DSC) thermographs of Y₂O₃/SiO₂ and Y₂O₃/SiO₂/LiYO₂ powder mixtures heated to 1450°C at the rate of 2°C min⁻¹; b the corresponding X-ray diffraction (XRD) patterns after the DSC test; XRD patterns of c γ-Y₂Si₂O₇ and d Y₂SiO₅ with the addition of 3 mol.-% LiYO₂ after holding at 1400°C for 4 h^46,47

Molten salt flux growth method (Czochralski technique)

Y₂O₃ and SiO₂ powders are mixed with low melting point flux oxides such as Bi₂O₃, V₂O₅, Li₂O, Na₂O, Li₂Mo₂O₇, etc., and then melted in Ir or Pt crucibles. By pulling the seed crystal from the melt slowly, single-crystals of Y₂Si₂O₇ and Y₂SiO₅ 40–140 mm in diameter and 50–150 mm in length can be obtained.^{20,39,48–51} A similar method, i.e. the glass crystallisation method was also reported for the synthesis of yttrium silicates from the melt. In this method, Y₂O₃ and SiO₂ powders were melted in a crucible at ∼2100°C, and amorphous glass was obtained after cooling down. The glass was then annealed at different temperatures in air to obtain the final crystals such as β-, γ-, and δ-Y₂Si₂O₇.^14,42

High-pressure/high-temperature (high-P/high-T) synthesis

η-Y₂Si₂O₇ was discovered as a by-product in high-P/high-T synthesis of pyroxene-type NaYSi₂O₆.²¹ The starting materials were SiO₂, Y₂O₃, and Na₂CO₃, and the synthesis was performed at 6·0 GPa and 1350°C in a Walker-type multi-anvil device for ∼216 h. This new type of polymorphous shows a different crystal structure from those reported by Ito and Johnson¹¹ and Feslche.¹²

Some other methods for the synthesis of Y–Si–O powders can also be found in the literature, such as the solution combustion method,⁶⁹ high-energy ball milling,⁷⁰ the phonochemical method,⁷¹ the Pechini method,⁷² etc. (Table 1).

Pressureless sintering of bulk Y–Si–O ceramics

Sun et al. ^46,47 successfully obtained dense, single-phase polycrystalline γ-Y₂Si₂O₇ and X2-Y₂SiO₅ bulk materials by pressureless sintering of powders synthesised by the solid–liquid reaction method. Sintering atmosphere is very important for the pressureless sintering of Y–Si–O oxides.⁴⁶ Figure 3a–b presents the strong influence of atmosphere on the sintering behaviour of γ-Y₂Si₂O₇. When constant rate sintering was conducted in oxygen at 5 K min⁻¹ to 1500°C, densification was obviously enhanced, and the final density reached 97·5% of the theoretical. The densification was dramatically reduced in argon. This phenomenon was explained based on the solubility of the gases, as shown in the schematic (Fig. 3a ), which affects pore elimination in the final stage of sintering. Compared to oxygen, the solution and outward diffusion of nitrogen (the major component of air) and argon atoms in oxide ceramics are negligible.⁷³ Therefore, the elimination of the pores was much more difficult during sintering in air or argon, resulting in relatively lower densities. Figure 3c–d shows the γ-Y₂Si₂O₇ and Y₂SiO₅ bulk materials obtained via pressureless sintering.

Sintering behaviour of γ-Y₂Si₂O₇ in different atmospheres: a scheme of sintering at different stages; b shrinkage curves of γ-Y₂Si₂O₇ in air, oxygen, and argon; c optical image of sintered γ-Y₂Si₂O₇ bulk material; d optical image of sintered Y₂SiO₅ bulk material

Hot pressing of bulk Y–Si–O ceramics

MacLaren et al. ⁴⁵ prepared bulk γ-Y₂Si₂O₇ by hot pressing hydrothermally synthesised amorphous Y₂Si₂O₇ powders with 3 wt.-% LiF additive. A dark coloured 2 mm thick plate with a density of 3·48 g cm⁻³ (86% of the theoretical) was obtained at 1600°C under 10 MPa. The dark colour indicates the formation of a high concentration of oxygen vacancies during the sintering in Ar.

Ogura et al. ⁷⁴ obtained bulk Y₂SiO₅ by hot pressing Y₂SiO₅ powders that were synthesised from a mixture of Si- and Y-alkoxides. Hot pressing was carried out at 1400–1425°C under 27 MPa for 1–4 h in Ar. After annealing the hot-pressed samples in air at 1400–1800°C for 1–24 h, samples with 93·9–99·7% of theoretical density were obtained. The major phase was X2-Y₂SiO₅, accompanied by a small amount of Y₂O₃.

Colloidal processing of bulk Y–Si–O ceramics

Colloidal processing, including slip casting, pressure casting, and tape casting, has been widely used in shape-forming of ceramics.^75–77

Colloidal processing of Y₂SiO₅

Aparicio et al. ⁴⁴ studied the colloidal processing of Y₂SiO₅. Concentrated aqueous suspensions were prepared by milling Y₂O₃, SiO₂, or Al₂O₃ powder mixtures. An alkali-free strong base, c hydroxide (TMAH), was used as dispersant to achieve slip stabilisation. A carbonic acid based polyelectrolyte (Dolapix CE-64) was used to provide electrosteric stabilisation. Green bodies of the binary and ternary composites with 47–50% relative densities were obtained by slip casting. The final density reached 73% of the theoretical density after sintering at 1670°C, with the composition of Y₂Si₂O₇ together with Y_4·67(SiO₄)₃O as a second phase. The solid-liquid sintering was achieved with Al₂O₃ additive, e.g., 48Y₂O₃–49SiO₂–3Al₂O₃, and the final density reached 90% of theoretical density after sintering at 1600°C for 3 h. Y₂SiO₅ was the major phase with small amounts of 3Y₂O₃·5Al₂O₃ and Y₂Si₂O₇ as impurities.

Colloidal processing of γ-Y₂Si₂O₇

Sun et al. ^78,79 systematically studied the colloidal processing of γ-Y₂Si₂O₇, which involved surface chemistry, the stabilisation mechanism of γ-Y₂Si₂O₇ in aqueous and ethanol-based suspensions, slip casting behaviour, and microstructure control of the bulk materials via the combination of strong magnetic field alignment during slip casting and one- or two-step sintering techniques in pressureless sintering.

The surface state of oxide particles is very important in colloidal processing. Figure 4a demonstrates that the surface groups are ≡Si–OH and = Y–OH hydroxyl groups on the γ-Y₂Si₂O₇ particles identified by Fourier transform infrared (FTIR) spectroscopy. In aqueous solution, the surface groups interact with solvent molecules and reach an ionic equilibrium (Fig. 4b ).⁷⁹ Figure 4c presents the ζ-potential curves of the aqueous γ-Y₂Si₂O₇ suspensions as a function of pH, with or without dispersants. The aqueous γ-Y₂Si₂O₇ suspensions without dispersant remain stable for a short term at pH 9–11·5 by electrostatic repulsion, but stability can be destroyed by the release and re-adsorption of Y³⁺, Y(OH⁾²⁺, and Y₂(OH)₂ ⁴⁺ ions. The suspensions with 1 dwb% polyethylenimine (PEI) have excellent long-term stability in the pH range of 4–11·5. As shown in Fig. 4c , in the high pH range near the isoelectric point (IEP) of γ-Y₂Si₂O₇ (pH = 10·8), the stability of the suspensions is dominated by the steric effect; at pH≈7, the stability is contributed by the electrosteric effect because of the ionisation of PEI; in the strongly acidic range, the stability mechanism is dominated by electrostatic and/or the depletion (or structural) potential.

a Fourier transform infrared (FTIR) spectra of hydrated groups on γ-Y₂Si₂O₇ particles, b ζ-potential of the γ-Y₂Si₂O₇ suspensions, c aqueous yttrium and silicon ion equilibrium diagram as a function of pH, d microstructure of the γ-Y₂Si₂O₇ surface perpendicular to the magnetic field direction after one-step pressureless sintering of the green bodies that were slip-casted in 12 T magnetic field; e microstructure of the γ-Y₂Si₂O₇ surface perpendicular to the magnetic field after two-step pressureless sintering of the green bodies slip-casted in 12 T magnetic field; and f the corresponding X-ray diffraction (XRD) patterns collected from the surface perpendicular to the magnetic field direction of γ-Y₂Si₂O₇ after two-step sintering^78,79

Sun et al. ⁷⁸ also prepared well-dispersed ethanol-based γ-Y₂Si₂O₇ suspensions using PEI as dispersant. Similar to the PEI stabilised aqueous Y₂Si₂O₇ suspensions, a high-affinity adsorption between cationic PEI and γ-Y₂Si₂O₇ in ethanol was observed. Slip casting of γ-Y₂Si₂O₇ ethanol suspensions was carried out with or without 12 T strong magnetic field.⁷⁸ After slip casting, the magnetic field led to a strong preferred orientation in the γ-Y₂Si₂O₇ green bodies. After a conventional one-step sintering in air at 1500°C for 90 min, 97·5% theoretical density was achieved, and the texture index was 0·74 for the 12 T sample, with an average grain size of more than 100 μm (Fig. 4d ). The two-step sintering scheme involves heating to 1300°C, then immediately cooling down to 1000°C and holding for 20 h, which allows the production of 12 T samples with a fine grain size of 4·5 μm and a perfect orientation, with a Lotgering orientation factor of 0·90 (Fig. 4e–f ).

Preparation of Y–Si–O coatings/films

Oxidation resistant coatings on SiC–C/C composites

Carbon/carbon (C/C) composites are attractive for high-temperature applications such as re-entry vehicles.⁸⁰ For most of these applications, oxidation resistant coatings are needed.⁸¹ SiC is a reliable coating on C/C composites because of the formation of a stable SiO₂ film on SiC below 1773°C, however, the vaporisation of SiO₂ via SiO formation, limits its high-temperature applications.⁸¹ Therefore, developing an outer erosion resistant layer on SiC coating is necessary. Y–Si–O oxides are favourable outer erosion resistant coatings on SiC because of their high melting points, close TECs to SiC, and low vaporisation rate. Up to now, a number of versatile Y–Si–O coating preparation methods have been reported, which are summarised in Table 2. The following is a summary of the details for the preparation of Y–Si–O coatings on C/C composites.

Table 2.

Summary of preparation methods for Y–Si–O coatings/thin films

Method	Processing condition	Contribution	Possible application	References
Slurry dip coating	Aqueous Y₂O₃+SiO₂ mixed slurry spread on SiC substrate, then heated at 1500–1600°C in air or argon	100–200 μm Y₂SiO₅ or Y₂Si₂O₇ or Y₂SiO₅+Y₂Si₂O₇ composite layer formed on the SiC substrate	Oxidation resistant coating for SiC–C/C composites	Kondo et al.⁵⁶
				Webster et al.⁸²
				Aparicio and Duran^83,84
	Aqueous Si+Y₂O₃+PVA binder slurry prepared and coated on substrate, sintered at 1500°C; then a borosilicate glass was applied on yttrium silicate layer	glass/Y₂SiO₅/SiC multilayer coating formed on C/C composite substrate	Oxidation resistant coating for SiC–C/C composites	Huang et al.^85,86
			Oxidation resistant coating for SiC–C/C composites	Smeacetto et al.⁸⁷
PLD	Y₂SiO₅:Ce was cold pressed and sintered as a target, PLD performed on silicon or silica substrate, followed by annealing at 400–1200°C	A 300–400 nm Y₂SiO₅:Ce thin film obtained after annealing	Blue phosphor thin film	Sun and Kwok⁵³
PLD		A 300–400 nm Y₂SiO₅:Ce thin film obtained after annealing	Blue phosphor thin film	Coetsee et al.⁸⁹
Rf-deposition	Y₂SiO₅:Ce was cold pressed and sintered as a target, then thin film was deposited by using a modified rf-magnetron, then post heat treated at 900–1100°C	Y₂SiO₅:Ce thin film obtained after annealing	Blue phosphor thin film	Bondar¹⁰⁵
	Y₂O₃+SiO₂ mixed and pressed into copper plate as target; films were rf-deposited on stainless steel and glass substrate	Transparent amorphous Y₂O₃–SiO₂ thin film was obtained in oxygen atmosphere with 0–60% Y₂O₃ content		Hanada et al.⁵⁴
	200–1000 Å yttrium film were sputtered on Si (100), then oxidised between 500–900°C	40–4500 Å Y–O–Si thin and thick films were obtained after post annealing	High-κ film for semiconductor devices	Chambers et al.^57,100,102
Thermal plasma spraying	50:50 Y₂O₃+SiO₂ mixed powders were used for coating preparation	200 nm Y₂SiO₅ coating was prepared on SiC–C/C composite	Oxidation resistant coating for SiC–C/C composites	Ogura et al.⁹⁰
	Y₂Si₂O₇+SiO₂ and Y₂Si₂O₇+Y₂SiO₅ powders were used for plasma spraying	100–200 μm yttrium silicates coatings were deposited on SiC–C/C composites	Oxidation resistant coating for SiC–C/C composites	Seifert et al.⁴⁰
	SiO₂–Y₂O₃ powders for plasma spray in different mol.-% compositions were synthesised at 1600°C for 3 h	90 μm yttrium silicate coatings were deposited on SiC–C/C composites	Oxidation resistant coating for SiC–C/C composites	Huang et al.⁹¹
EPD	Hydrothermally synthesised ultrafine α-Y₂Si₂O₇ powders were used for EPD, then sintered at 1050°C for 3 h	90 μm α-Y₂Si₂O₇ thick coating was obtained on SiC–C/C composites	Oxidation resistant coating for SiC–C/C composites	Kaya et al.⁵⁵
EPD	Pechini method synthesised Y₂SiO₅ and Y₂Si₂O₇; the mixture was dispersed in isopropanol for EPD, and then annealed at 1530°C	15–30 μm Y₂Si₂O₇/Y₂SiO₅ thick coating was obtained on SiC–C/C composites	Oxidation resistant coating for SiC–C/C composites	Argirusis et al.^72,92
Microwave sintering	SiO₂+Y₂O₃ outer layer with BAS glass+Y₂O₃ inner layer was painted on SiC–C/C substrate, then microwave sintered at 1500°C for 2 h	Y₂Si₂O₇ outer layer with Celsian+Y₂SiO₅ inner layer multilayer coating on SiC–C/C composites	Oxidation resistant coating for SiC–C/C composites	Zheng et al.⁸⁸
CVD or ALD	solutions of tris(tert-butoxy)silanol in mesitylene and tris(tert-butyl(trimethylsilyl)amido)yttrium in mesitylene were separately pumped into tee joint to flow into the same heated tube, with the substrates of silicon and glassy carbon placed inside	Films with approx. composition of Y₂Si₂O₇ were formed. For ALD, the thickness was about 0·3 nm per cycle	High-κ film for semiconductor devices	Gordon et al.
CVD or ALD			High-κ film for semiconductor devices	US 6969539 B2¹⁰³

PVA: polyvinyl alcohol; PLD: pulsed laser deposition; rf: radio frequency; EPD: electrophoretic deposition; BAS: barium aluminosilicate; CVD: chemical vapor deposition; ALD: atomic layer deposition.

Slurry dip coating

Preparation of Y–Si–O coatings on SiC–C/C comfposites from aqueous or organic slurry is the simplest yet efficient method, which is especially suitable for coatings on large-size components and for volume production.^56,82–88 The slurries for slip coating or dip coating usually consist of a Y₂O₃+SiO₂ starting powder mixture, synthesised Y₂SiO₅, Y₂Si₂O₇, and their composites, or sol-gel precursors. For dip coating, excellent slurry fluidity is necessary to ensure homogeneous coating. The coating can be obtained by dipping the SiC–C/C substrates into the slurry, followed by drying and sintering. Owing to the presence of porosity in the as-prepared coatings, and the cracking and spallation of the coating during service, Webster et al. ⁸² found that the SiC/Y₂SiO₅ coatings could not provide effective protection from oxidation. It has been found that an Y₂SiO₅+Y₂Si₂O₇ mixture outer layer has a better performance because of its lower porosity, less cracking and spallation in service.⁸³ An intermediate functional glass coating was also tried for sealing the cracks during service. Aparicio and Duran⁸⁴ prepared crack free 50Y₂O₃:SiO₂ sol-gel coatings on soda-lime glass substrates with a critical thickness of 80 nm. Homogeneous coatings with a thickness of 2 μm were also prepared by multilayer processes on SiC-coated C/C composite substrates. Unfortunately, the performance of this sol-gel coating was not examined.

Huang et al. and Smeacetto et al. ^85–87 developed multilayer coatings consisting of a Y–Si–O intermediate layer and a borosilicate glass outer layer on SiC–C/C composites. The slurry for dip coating was a mixture of Si, Y₂O₃, distilled water, and polyvinyl alcohol (PVA) binder. The coating was obtained after drying at 80°C and sintering at 1500–1600°C for 1 h in argon. Then, the coated samples were pre-oxidised in air at 1500–1600°C for 2–10 h. Finally, a layer of borosilicate glass was applied directly on their surfaces by heating at 1500°C for 2 h in Ar. Owing to the sealing effect provided by the borosilicate glass, the SiC/yttrium silicate/glass multilayer coating exhibited good oxidation protection for C/C composites in air at 1600°C for 202 h, with a weight loss of 2·87×10⁻³ g cm⁻². Zheng et al. ⁸⁸ reported another type of multilayer protective coating obtained by separately dipping the SiC–C/C substrates into barium aluminosilicate (BAS) glass+Y₂O₃ and Y₂O₃+SiO₂ suspensions. After drying and sintering at 1500°C for 2 h, a multilayer film was obtained, consisted of a Y₂Si₂O₇ outer layer and a Celsian+Y₂SiO₅ inner layer. The weight loss of the double layer coated SiC–C/C composite was only 1·22% after 150 min oxidation at 1400°C.

Plasma spray coating

Plasma spraying is a very important technique that features the easy deposition of dense, homogeneous, large-size, complex shapes, and thick protective coatings.^40,89–91 Seifert et al. ⁴⁰ tested plasma-sprayed yttrium silicate coated SiC–C/C in a plasma wind tunnel at 1350°C for 23 min. The results showed that the weight change of the SiC–C/C composites with the yttrium silicate protective coating was 10% less than for the samples with only the CVD-SiC coating, and was approximately 6 times (MAN Technology) and 10 times (Astrium) lower than for the C/SiC composites from industrial suppliers. The yttrium silicate coating lost its ability to protect, however, when the temperature was higher than 1700°C, where a sudden coating failure was observed by the formation of a blister at the coating/SiC interface. Huang et al. ⁹¹ reported that plasma spraying of single or multilayer coatings with the compositions of 1·5SiO₂·Y₂O₃, 1·5SiO₂·Y₂O₃/Y₂O₃, and 2SiO₂·Y₂O₃/1·5SiO₂·Y₂O₃/SiO₂·Y₂O₃ gave better oxidation-resistance than the SiO₂·Y₂O₃ coatings, since the former coatings could provide protection for SiC–C/C composites at 1500°C for 73 h.

Electrophoretic deposition

Kaya et al. ⁵⁵ prepared thick Y₂Si₂O₇ coatings on SiC/SiC composites from an α-Y₂Si₂O₇ suspension via EPD. The solid loading was 20 vol.-% with 5 wt-% LiF sintering aid. The pH of the suspension was 9·7, where the Y₂Si₂O₇ particle surfaces were negatively charged and deposited on the positive electrode. At constant voltage of 8 V, continuous and homogeneous α-Y₂Si₂O₇ protective coatings with a thickness of 90 μm on SiC/SiC composites were obtained after deposition for 1·5 min followed by sintering at 1050°C for 3 h under flowing argon.

Argirusis et al. ^72,92 studied the impedance spectra of Y–Si–O suspensions in EPD coatings. The impedance spectrum for the pure isopropanol was a single semicircle, which corresponded to the specific conductivity of (1·00±0·02)×10⁻⁷ S cm⁻¹ and a dielectric constant of ϵ_r = 25·6±3·7. The resistances of the Y–Si–O suspensions obtained from the impedance spectra were much lower, resulting in a specific conductivity of (1·02±0·01)×10⁻⁴ S cm⁻¹ and a dielectric constant of ϵ_r = 31·7±2·9. During deposition, the observed total resistance R _suspension+R _deposition continuously increased with time and coating thickness. For the dispersion of Y₂SiO₅, Y₂Si₂O₇, and 70 wt-% Y₂Si₂O₇+30 wt-% Y₂SiO₅ composite in isopropanol, a coating with an average thickness of 5 μm per deposition step at 60 V/1 min could be obtained on SiC–C/C composites. The impedance spectrum study demonstrated that the kinetics of EPD was mainly controlled by mass transport through the growing deposit. Further oxidation tests were carried out on the SiC–C/C composites with 15–30 μm Y–Si–O protective coatings, and they showed that the EPD Y–Si–O coatings provided good protection at 1500°C for more than 100 h.

Environmental barrier coatings on SiC–C/C substrate

Si-based ceramics, such as SiC_f/SiC composites and monolithic Si₃N₄, exhibit superior high-temperature strength and durability, demonstrating their potential to revolutionise gas turbine engine technology.^93,94 A key stumbling block to realising Si-based ceramic as turbine hot section components is their lack of environmental durability in high velocity combustion environments, resulting in unacceptably high recession rates.^94,95 This can be prevented by EBCs. The candidate EBCs are mullite, yttrium stabilized c (YSZ), and BSAS (1–xBaO–xSrO–Al₂O₃–2SiO₂, 0≤x≤1).^96–98 Some rare-earth silicates with the formulae Re₂Si₂O₇ and Re₂SiO₅, where Re is a rare-earth element, have potential as promising candidates for EBCs, due to their low TEC and chemical stability.^93,99

Multilayer EBCs consisting of mullite with an Y₂SiO₅ top layer were prepared by Lee et al. ⁹³ using the atmospheric pressure plasma spraying on SiC and Si₃N₄ and then annealed at 1300°C for 20 h. The volatilisation of the EBC top coating was tested at 1300°C and 1400°C in 90% H₂O-balanced O₂. A severe reaction between mullite and Y₂SiO₅ was observed after testing at 1400°C for 46 h, turning the entire EBC into a layer of Y₂O₃–Al₂O₃–SiO₂ bubbles. This implies that Y₂SiO₅ may not be so competitive with Yb₂SiO₅, Sc₂SiO₅, and Lu₂SiO₅ when mullite is used as the base layer. A very recent report shows that γ-Y₂Si₂O₇ has a very close TEC to those of SiC and Si₃N₄ and is promising as a single-layer environmental barrier coating.⁶⁰

High-κ gate dielectric thin films

Sputtering of amorphous yttrium silicate thin film

Chambers and Parsons⁵⁷ suggested that yttrium silicate possesses excellent thermodynamic, dielectric, and structural properties that make it attractive as a high-κ candidate. Moreover, Y₂Si₂O₇ and Y₂SiO₅ exhibit low lattice mismatch with silicon substrate. Chambers et al. ^57,100–102 sputtered yttrium films in a two-chamber vacuum system from a Y-metal target with an rf-power of 420 W on silicon substrate. Amorphous Y–Si–O film was formed by annealing the yttrium film in vacuum for 20 min, followed by annealing in air for 20 min at 900°C. The thickness of the Y–Si–O film varied between 40 and 1000 Å, depending on the thickness of the sputtered yttrium film. The electric characterisation in Al/Y–Si–O/Si capacitors suggested that there was a large charge density accumulation in the capacitors, which might have resulted from the competition between the yttrium/silicon reaction and the oxidation, leaving unsatisfied bonds. The leakage current for the n-type capacitors was less than that of silicon dioxide with equivalent capacitance, but was greater than that expected for a SiO₂ film of similar physical thickness, which was consistent with the smaller barrier expected for yttrium silicate compared to SiO₂.

CVD or ALD of yttrium silicates

Metal silicates can be deposited on a heated substrate by the reaction of vapours of alkoxysilanols or alkylphosphates with reactive metal amides, alkyls, or alkoxides. Such deposition techniques are referred as CVD or ALD.¹⁰³ For the CVD of yttrium silicate, a solution of tris(tert-butoxy)silanol in mesitylene and a solution of tris(tert-butyl(trimethylsilyl)amido)yttrium in mesitylene were separately pumped into a tee joint and flowed into the same heated tube. Substrates of silicon and glassy carbon placed inside the tube were coated with yttrium silicate. Films with the approximate composition of Y₂Si₂O₇ were formed. The refractive indexes of films deposited on silicon were found to be about 1·6 by ellipsometry.

For the ALD of yttrium silicate, almost the same procedure was employed as that for CVD, except that the precursors were injected in alternate pulses spaced 5 seconds apart, instead of continuously. A film of composition similar to Y₂Si₂O₇ was deposited with uniform thickness along the whole length of the heated zone. The thickness was about 0·3 nm per cycle.

Deposition of Y₂SiO₅:Ce phosphoric thin film

Phosphors are widely used in fluorescent lamps, cathode-ray tubes (CRT), and X-ray intensifier screens.¹⁰⁴ Y₂Si₂O₇ and Y₂SiO₅ are important luminescent and laser host materials for various rare-earth activators owning to their excellent thermal and chemical stability. The PLD Y₂SiO₅:Ce phosphoric thin films prepared by Sun and Kwok⁵³ were in the range of 300–400 nm on Si(100) substrate or SiO₂ coated Si substrate. Besides PLD, multilayer thin film oxide phosphors were also prepared via ion-plasma deposition by Bondar¹⁰⁵

Polymorphs, crystal structure, defects and microstructure of Y–Si–O oxides

Polymorphs and phase transformations

Yttrium disilicate presents up to six polymorphs: y, α, β, γ, δ, and probably z, according to Ito and Johnson.¹¹ The low-temperature phases (y and z) are probably stabilised by impurities.^10–13 Liddell and Thompson¹⁰ suspected that z-Y₂Si₂O₇ may actually be a kind of hydrated yttrium disilicate or Thanlenite with the chemical formula Y₂Si₂O_7.1/3H₂O. Recently, two new polymorphs of Y₂Si₂O₇, which were indexed as η- and ζ-Y₂Si₂O₇, were reported as by-products in the synthesis of other compounds.^20,21 The temperatures reported for polymorph transitions and the phase-stable ranges vary considerably from author to author.^{10–13,28,39,50,106,107} Figure 5 illustrates the widely accepted polymorph transformations of yttrium disilicate by Ito and Johnson.¹¹ The different densities: 4·30, 4·03, 4·04, and 4·11 g cm⁻³ for α, β, γ, and δ, respectively,⁴² suggest that the polymorph transformations are accompanied by volume change. Thereafter, different transition temperatures: , were reported by Felsche.¹² Recent studies seem to be in contradiction to these results, however.¹⁰⁸ For example, Parmentier et al. ²⁸ and Becerro et al. ³⁸ demonstrated that the γ-phase formed below 1350°C and was associated with β-phase; while the β-phase usually appeared at 1300°C and was present in samples quenched from 1600°C. In addition, Parmentier et al. ²⁸ did not succeed in synthesis of the δ-phase at 1600°C, which should be the final product after annealing above 1580°C according to Felsche.¹² Similarly, a yttrium silicate sample sintered by Fukuda and Matsubara¹⁰⁹ at 1600°C and subsequently quenched in air was γ-phase instead of δ-phase (the Y₂Si₂O₇ in Ref. 109 is referred to as ‘δ-phase’, but based on the crystal structure parameters provided by the authors, it should be γ-phase). The phase transition of y-Y₂Si₂O₇ to high-temperature polymorphs also varied from author to author. Dinger et al. ¹¹⁰ observed a direct transition from y- to δ-Y₂Si₂O₇ during the crystallisation of Y₂O₃–SiO₂–AlN glass, and Vomacka and Babushkin¹⁰⁷ reported a transition from y- to β-Y₂Si₂O₇ at 1300°C in the crystallisation of Y₂O₃–Al₂O₃–SiO₂ glass containing ZrO₂. Afterwards, Becerro et al. ³⁸ established the y- to β-Y₂Si₂O₇ transition by annealing hydrothermally synthesised y-Y₂Si₂O₇, following the sequence of . Maclaren et al. ¹¹¹ studied the phase transformation of hydrothermally synthesised y-Y₂Si₂O₇ using selected area electron diffraction (SAED) technique. They claimed that the transformation temperature from y- to α-Y₂Si₂O₇ was between 1100 and 1200°C instead of 1050°C. Diaz et al. ²⁶ noticed that γ-phase transformed to α-phase with C2/m symmetry below 1445°C. The α-phase reported in their paper had a similar structure to that described by Batalieva et al. ¹¹² Actually, the reported ‘α-phase’ in both papers should be ‘β-phase’. Moreover, some works show that there is irreversible phase transformation between γ-Y₂Si₂O₇ and low-temperature phases such as α- or β-Y₂Si₂O₇. Kumar and Drummond⁴² studied the crystallisation behaviour of Y₂O₃–SiO₂ melt, where the transition from β- to γ-phase occurred at 1400°C, and γ- to δ-phase at 1600°C, while no γ-phase to low-temperature phase transformation was observed during cooling. Sun et al. ⁴⁶ prepared single-phase γ-Y₂Si₂O₇ that maintained a good structural stability between room temperature and 1550°C. Parmentier et al.,²⁸ however, observed the transitions of δ→γ→β and δ→β phase transitions. Maier et al. ¹⁰⁸ studied the formation and phase transformation by two different methods. One was the crystallisation of silicate melt, where the crystalline phase appeared to change as δ→γ→β with decreasing temperature. The other was the solid-state synthesis of mixed oxide powders, where the initial crystalline phase was α-Y₂Si₂O₇ and transitions from α to β, and then to γ were observed. However, no any reversible phase transition from γ to β or to α during cooling was observed. It was concluded that such a reversible phase transition was strongly dependent on the synthesis route. Although there are only two polymorphisms for Y₂SiO₅, X1- and X2-Y₂SiO₅, the reported phase transition temperature from X1 to X2 was still found to vary over a wide temperature range from 850–1200°C.^113,114 Therefore, the phase formation and stability, or the phase transition, still need to be studied. The main reason for this ambiguity is the susceptibility of yttrium silicates to the preparation method, impurity, and heating history.

Polymorph transformations of yttrium disilicate reported by Ito and Johnson¹¹

Crystal structure and chemical bonding

Liddell and Thompson¹⁰ summarised the XRD data for all the reported Y–Si–O phases in 1986. Table 3 lists the crystal structure data for the five existing polymorphs of Y₂Si₂O₇ that were refined from the XRD data proposed by Liddell et al., together with the newly reported ζ- and η-Y₂Si₂O₇ and those for Y₂SiO₅ and Y_4·67(SiO₄)₃O.^11,115–119 Controversies on the crystal structures still exist in the literature, however. Although γ-Y₂Si₂O₇ has been well characterised, the space group representation differs in the reports^10,36 because of the inconsistency between the chosen designation and the reported Miller indices. The designations P2₁/a, P2₁/c and P2₁/n are different Hermann–Mauguin representations for the same space group (Schönflies notation), denoting different orientations of the symmetry operation of the group with respect to the chosen labels for the cell axes.^36,120 For X2-Y₂SiO₅, the space group was also reported as both C2/c and B2/b because of the different selection of the basis vectors.

Table 3.

Space group, unit cell parameters and density for the polymorphs of Y–Si–O oxides

Polymorph	Space group	Unit cell parameters						Density (g cm⁻³)	<Si–O_b–Si in Si₂O₇	References
Polymorph	Space group	a (Å)	b (Å)	c (Å)	α (degree)	β (degree)	γ (degree)	Density (g cm⁻³)	<Si–O_b–Si in Si₂O₇	References
y-Y₂Si₂O₇	P2₁/m	7·50	8·06	5·02	90	112·0	90	4·08	134°	Batalieva and Pyatenko¹¹⁵
		7·44	8·06	5·06	90	111·9	90		134°	Becerro et al.³⁸
		5·03	8·07	7·34	90	108·6	90	4·06		Heward et al.¹¹⁶
α-Y₂Si₂O₇	PĪ	6·59	6·64	12·03	94·5	91	91·8	4·38		Dolan et al.¹¹⁷
α-Y₂Si₂O₇	PĪ	6·59	6·63	12·03	94·5	89·1	88·2	4·39	121°, 131·4°	Kahlenberg et al.²⁷
β-Y₂Si₂O₇	C2/m	6·88	8·97	4·72	90	101·7	90	4·03	180°	Günther and Georg⁵⁰
γ-Y₂Si₂O₇	P2₁/n	5·58	10·86	4·70	90	96·0	90	4·04	180°	Christensen et al.³⁹
δ-Y₂Si₂O₇	Pna2₁	13·66	5·02	8·15	90	90	90	4·11	158°	Dias et al.¹¹⁸
ζ-Y₂Si₂O₇	P2₁/m	5·04	8·06	7·33	90	108·6	90	4·08	156°	Hartenbach et al.²⁰
η-Y₂Si₂O₇	P1	6·63	6·58	35·92	91·1	94·5	91·7	4·42		Kahlenberg et al.²¹
X1-Y₂SiO₅	P2₁/c	9·01	6·98	6·63	90	106·7	90	3·60		Wang et al.²⁹
X1-Y₂SiO₅	P2₁/c	8·50	6·82	6·38	90	104	90			Heward et al.¹¹⁶
X2-Y₂SiO₅	C2/c	10·34	6·69	12·38		102·5		4·44		Maksimov et al.¹¹⁹
Y_4·67(SiO₄)₃O	P6₃/m	9·37	/	6·73						Parmentier et al.³⁴

Ching et al. and Wang et al. ^121–123 carried out comprehensive theoretical investigations on the Y–Si–O system via density functional theory calculations. Some experimental studies also were carried out on the crystal structures of yttrium disilicate.^{10,27,28,34,36,38,50,106,124–126} The cell parameters were obtained by refining XRD data collected from Y₂Si₂O₇ powders.^{10,27,28,38,50,125} Becerro, Parmentier, and Díaz et al.^{28,35,36,38,106,124,126} used ⁸⁹Y MAS-NMR and ²⁹Si MAS-NMR spectroscopy to clarify and differentiate chemically distinct Y and Si environments in each polymorph. According to the ⁸⁹Y MAS-NMR data for y- and δ-Y₂Si₂O₇, only one crystallographically distinct Y site exists in the y- and δ-Y₂Si₂O₇ unit cells instead of two, while in the well-crystallised α-Y₂Si₂O₇, four Y sites can be clearly resolved in the ⁸⁹Y MAS-NMR spectrum. The ²⁹Si MAS-NMR spectrum suggests that the distinct Si sites in the polymorphs are 4, 1, 1, and 2 for α-, β-, γ-, and δ-Y₂Si₂O₇, respectively.

Figure 6 shows the crystal structures of the polymorphs of Y₂Si₂O₇, Y₂SiO₅, and Y_4·67(SiO₄)₃O. The corresponding atomic positions are tabulated in Table 4. Considering the well documented crystal structures of Y–Si–O oxides in the previous reviews, we have only briefly describe the recent reports on η-Y₂Si₂O₇, ζ-Y₂Si₂O₇, and Y_4·67(SiO₄)₃O.

Crystal structure of a y-Y₂Si₂O₇, b a-Y₂Si₂O₇, c b-Y₂Si₂O₇, d g-Y₂Si₂O₇, e d-Y₂Si₂O₇, f h-Y₂Si₂O₇, g X1-Y₂Si₂O₅, h X2-Y₂Si₂O₅ and i Y_4.67(SiO₄)₃O according to the parameters listed in Table 4

Table 4.

Crystal structure and atomic positions of Y–Si–O oxides

Formula	y-Y₂Si₂O₇ ¹¹⁶	Space group
Unit cell parameters	a = 5·0303 Å; b = 8·0689 Å; c = 7·3362 Å; β = 108·67; V = 282·1 Å³
Atomic positions	O1: (0·9440 0·2500 0·1949)	O5: (0·8352 0·0867 0·5992)
	O2: (0·2835 0·0874 0·0397)	Si1: (0·2109 0·2500 0·1290)
	O3: (0·4918 0·2500 0·3398)	Si2: (0·6435 0·2500 0·5706)
	O4: (0·4429 0·2500 0·6831)	Y1: (0·7305 0·0069 0·2516)

Formula	α-Y₂Si₂O₇ ¹¹⁷	Space group
Unit cell parameters	a = 6·5872 Å; b = 6·6387 Å; c = 12·032 Å; α = 94·50; β = 90·98; γ = 91·77; V = 524·16 Å³
Atomic positions	Y1: (0·4432 0·1627 0·8871)	Y2: (0·3946 0·4063 0·6411)
	Y3: (0·1276 0·2700 0·3716)	Y4: (0·1645 0·6816 0·8915)
	Si1: (0·3476 0·3541 0·1096)	Si2: (0·9950 0·2093 0·8473)
	Si3: (0·8972 0·2106 0·5940)	Si4: (0·3330 0·8714 0·6156)
	O1: (0·1480 0·0105 0·8751)	O2: (0·1139 0·3644 0·7879)
	O3: (0·7994 0·2037 0·9069)	O4: (0·9027 0·0651 0·6947)
	O5: (0·0871 0·3264 0·5519)	O6: (0·7260 0·4025 0·6233)
	O7: (0·2137 0·9269 0·5120)	O8: (0·2134 0·7175 0·6868)
	O9: (0·5400 0·7260 0·5754)	O10: (0·4277 0·0831 0·6889)
	O11: (0·2650 0·4161 0·9951)	O12: (0·1507 0·2870 0·1863)
	O13: (0·5007 0·1683 0·0809)	O14: (0·4955 0·5191 0·1859)

Formula	β-Y₂Si₂O₇ ⁵⁰	Space group
Unit cell parameters	a = 4·6916 Å; b = 10·8521 Å; c = 5·5872 Å; β = 96·04°
Atomic positions	Y: (0·8904 0·8495 0·0941)	Si: (0·6445 0·1136 0·3696)
	O1: (0·7933 0·0502 0·1358)	O2: (0·8670 0·1822 0·5371)
	O3: (0·5340−0·0106 0·4976)	O4: (0·3792 0·2017 0·2513)

Formula	γ-Y₂Si₂O₇ ³⁹	Space group
Unit cell parameters	a = 4·6916 Å; b = 10·8521 Å; c = 5·5872 Å; β = 96·04°
Atomic positions	Y: (0·8904 0·8495 0·0941)	Si: (0·6445 0·1136 0·3696)
	O1: (0·7933 0·0502 0·1358)	O2: (0·8670 0·1822 0·5371)
	O3: (0·5340−0·0106 0·4976)	O4: (0·3792 0·2017 0·2513)

Formula	δ-Y₂Si₂O₇ ¹¹⁸	Space group
Unit cell parameters	a = 13·655 Å; b = 5·016 Å; c = 8·139 Å; V = 557·88 Å³
Atomic positions	Y: (0·1255 0·3395−0·0096)	Si1: (0·3179 0·3711 0·25)
	Si2: (0·5410 0·6214 0·25)	O1: (0·2671 0·4875 0·0861)
	O2: (0·3476 0·0613 0·25)	O3: (0·4218 0·5588 0·25)
	O4: (0·5477 0·7994 0·0817)	O5: (0·6031 0·3386 0·25)

Formula	ζ-Y₂Si₂O₇ ²⁰	Space group
Unit cell parameters	a = 5·036 Å; b = 8·065 Å; c = 7·327 Å; β = 108·6°
Atomic positions	Y: (0·7306 0·0074 0·2518)	Si1: (0·2130 0. 25 0·1328)
	Si2: (0·6535 0·25 0·5733)	O1: (0·9274 0·25 0·1858)
	O2: (0·2840 0·0847 0·0347)	O3: (0·4887 0·25 0·3353)
	O4: (0·4308 0·25 0·6877)	O5: (0·8308 0·0805 0·5925)

Formula	X1-Y₂SiO₅ ²⁹	Space group
Unit cell parameters	a = 9·0139 Å; b = 6·9282 Å; c = 6·6427 Å; β = 106·68; V = 397·38 Å³
Atomic positions	Y1: (0·1198 0·1396 0·4268)	Y2: (0·5225 0·6273 0·2340)
	Si: (0·1994 0·5924 0·4800)	O1: (0·199 0·415 0·649)
	O2: (0·123 0·487 0·247)	O3: (0·372 0·644 0·504)
	O4: (0·090 0·777 0·457)	O5: (0·378 0·375 0·037)

Formula	X2-Y₂SiO₅ ⁵⁹	Space group
Unit cell parameters	a = 14·1883; b = 6·6224; c = 12·3040; β = 135·15°
Atomic positions	O1: (0·6927 0·2588 0·3547)	O2: (0·9485 0·4592 0·3268)
	O3: (0·4950 0·3147 0·4447)	O4: (0·1307 0·8468 0·8995)
	O5: (0·7203 0·1066 0·1734)	Si1: (0·6215 0·1607 0·1899)
	Y1: (0·5577 0·6199 0·1678)	Y2: (0·8243 0·4948 0·0381)

Crystal structure of ζ-Y₂Si₂O₇

In 2006, Hartenbach et al. ²⁰ occasionally obtained colourless, lath-shaped single-crystals of Y₂Si₂O₇ with the new ζ-type structure as a minor by-product during the preparation of yttrium oxotellurates (IV) using Y₂O₃ and TeO₂ as starting materials in YCl₃ flux after heating at 900°C for 8 days. This polymorph crystallises in space group P2₁/m with unit cell parameters of a = 5·036 Å, b = 8·065 Å, c = 7·327 Å, and β = 108·6°. The crystallographically unique Y³⁺ cation is coordinated by seven oxygen atoms arranged in the shape of a slightly distorted monocapped octahedron, and the Y–O bond lengths vary between 2·21–2·48 Å. The isolated oxodisilicate units [Si₂O₇]⁶⁻ consist of two Si⁴⁺ cations and seven O²⁻ anions, with Si–O bond lengths of 1·61–1·68 Å, Si–O–Si angle of 156°, and O–Si–O angles in the range 91–117°. These pyroanions exhibit an almost perfectly eclipsed conformation, consisting of a horseshoe-shaped backbone with two silicon atoms and three of the oxygen atoms situated on the mirror planes of the unit cell. The remaining four oxygen anions complete this [Si₂O₇]⁶⁻ entity of two vertex-sharing [SiO₄]⁴⁻ tetrahedra as terminal ligands for silicon.

Heward et al. ¹¹⁶ pointed out, however, that the structure of ζ-Y₂Si₂O₇ ²⁰ was similar to the formerly reported y-Y₂Si₂O₇ (ICSD card No. 28004), except that the monoclinic a- and c-axes were switched, which resulted in different β angles. Compared to the structure described in ICSD card No. 28004, where two non-equivalent Y positions were reported for y-Y₂Si₂O₇, the structure of Hartenbach et al. can fit the experimental XRD pattern of y-Y₂Si₂O₇ much better. There is an un-fittable low angle XRD peak around 16·5° in the work of Becerro et al. ^35,38,124 when the crystal structure data of ICSD card No.28004 was used, but it can be successfully fitted by employing the structure of ζ-Y₂Si₂O₇ to refine the XRD pattern of y-Y₂Si₂O₇. The crystal structure in Table 4 provided by Hartenbach et al. ²⁰ for ζ-Y₂Si₂O₇ also coincides well with that on y-Y₂Si₂O₇ provided by Heward et al. ¹¹⁶ Therefore, based on the work of Heward et al., it is concluded that the ‘ζ-Y₂Si₂O₇’ is actually ‘y-Y₂Si₂O₇’ instead of a new type of polymorph of yttrium disilicate.

Crystal structure of η-Y₂Si₂O₇

In 2007, Kahlenberg et al. ²¹ reported another new type of polymorph η-Y₂Si₂O₇, which was obtained as a by-product in the synthesis of NaYSi₂O₆ in a high-P/high-T experiment performed on the Na₂O–Y₂O₃–SiO₂ system and quenched from 6·0 GPa/1350°C. The crystallographic data are: space group P1¯, a = 6·6290 Å, b = 6·5840 Å, c = 35·916 Å, α = 91·1°, β = 94·5°, γ = 91·7°, V = 1561·6 Å³ and ρ = 4·415 g cm⁻³. This structure belongs to the group of mixed anion silicates containing isolated [SiO₄]-tetrahedra and [Si₃O₁₀]-trimers in the ratio of 1:1. The silicate anions are located in layers parallel to (11¯3). Linkage between the layers is provided by six- and eight-fold coordinated Y cations sandwiched between the sheets. The structure can be principally regarded as a two-fold superstructure of the L-type. The Si–Si–Si angles within the three trimers of the asymmetric unit vary between 145·75° and 156·11°. The Si–O distance and O–Si–O angles for SiO₄ tetrahedra are in the range of 1·579–1·739 Å and 95·8–119·5°, respectively, implying that the tetrahedral are considerably distorted. Figure 6f is the crystal structure of η-Y₂Si₂O₇ built from the data provided by Kahlenberg et al. ²¹

Crystal structure of Y_4·67(SiO₄)₃O

Y_4·67(SiO₄)₃O apatite is frequently detected in the preparation of Y₂SiO₅ and Y₂Si₂O₇ as a second phase. Single-phase Y_4·67(SiO₄)₃O cannot be obtained by conventional preparation methods, however, since it can only be stabilised by other cationic ions such as Fe³⁺, Al³⁺, etc.³⁴ Figure 6i presents the crystal structure of Y_4·67(SiO₄)₃O. Within the finite 1×1×3 supercell, two yttrium vacancies are formed in 4f (1/3, 2/3, 0) sites in the Wyckoff notation that feature three-fold symmetry. Owing to the difficulty in the synthesis of single-phase apatite, no further information on the atomic coordinates and the numbers of crystallographic equivalent Y and Si sites can be found to date. Recently, the apatite structure has been found to be promising in solid oxide fuel cells (SOFC) as a low-cost electrolyte,^127,128 which means that Y_4·67(SiO₄)₃O apatite may be used as a new generation SOFC material.

Defects in Y₂SiO₅

Pang et al. ⁵¹ studied the defects in Y₂SiO₅ crystals using γ-irradiation. The as-grown and annealed Y₂SiO₅ samples were irradiated by γ-rays at room temperature. The air-annealed sample and the as-grown sample remained colourless after irradiation, while the H₂-annealed sample became grey after irradiation. The absorption spectrum showed that the colour centre formed in the H₂-annealed sample after γ-ray irradiation was the F(V_o+2e) colour centre. When γ-photons pass through, they interact with the atoms to produce secondary electrons, and then the oxygen vacancies in the lattice trap one or two electrons and form F-type colour centres. During the air annealing, oxygen diffuses into the lattice and decreases the concentration of oxygen vacancies, so the concentration of F-type colour centres is low, and there is no obvious colour change before and after irradiation.

The defects in Ce³⁺ doped Y₂SiO₅ were reported by Aitasalo et al. ⁶⁴ They demonstrated the formation of oxygen vacancies in Y₂SiO₅ at reducing preparation conditions, and the formation of F-type colour centres (positive F⁺ and neutral F centres) by the interaction between oxygen vacancies and electrons. Usually, the lattice energies in oxides are very high, so the charge is imbalanced. The silicate host favours the creation of more F centres than F⁺ centres.

Recently, Liu et al investigated the formation mechanism of mono-vacancy and native point defects in Y₂SiO₅ and Y₂Si₂O₇ by first-principles calculations.^129,130 It was found that the formation of native point defects follows the same trend in Y₂SiO₅ and Y₂Si₂O₇: the Frenkel defect of O is predominant, accompanied by low concentrations of cation antisite defects and Schottky defects. When chemical potential is considered, O_i or V_O shows the lowest formation energy under O-rich or O-poor conditions, respectively. In addition, it is possible to control the concentrations of defects by tailoring the chemical potentials of O, Y, and Si. Nonstoichiometry in Y₂SiO₅ or Y₂Si₂O₇, such as Y₂O₃ or SiO₂ excess, brings out different mechanisms of native point defects or impurity phases. For Y₂SiO₅, Si_Y antisite, O_i interstitial, and V_Y vacancy defects are the main point defects, together with the formation of Y₂Si₂O₇ impurity phase when SiO₂ is in excess; while Y_Si antisites appear together with Y_i interstitial and/or V_O vacancy defects for excess Y₂O₃. For Y₂Si₂O₇, the formation of Si_Y antisites accompanied by O_i interstitial and/or V_Y vacancy defects are preferred when SiO₂ is excess; but the Y_Si antisite, V_O vacancy, and/or Y_i interstitial defects dominate the point defects, together with the formation of the Y₂SiO₅, when Y₂O₃ is excess.¹³⁰ The theoretical calculations also reveal that the self-diffusion of oxygen vacancies has a high-energy barrier in Y₂SiO₅, which is comparable to those in SiO₂ and Al₂O₃, demonstrating the low oxygen permeability in Y₂SiO₅.¹²⁹

Microstructure characteristics

MacLaren et al. ^45,131 observed the microstructure of a hot pressed γ-Y₂Si₂O₇ via high resolution transmission electron microscopy (HRTEM). Stacking faults were frequently identified on (010) planes with g = 040. Lin et al. ¹³² investigated the microstructure of γ-Y₂Si₂O₇ after indentation. The results revealed a stress-induced amorphisation process under severe mechanical conditions, where the crystalline to amorphous transformation is mediated by slip bands containing a high density of dislocations. The microstructure of Y₂SiO₅ after spherical indentation was observed by Sun et al.,⁵⁹ where they found that the grains in the deformation zone were commonly microcleaved into a number of well-orientated subgrains, suggesting that microcleavage is likely to be the major mechanism of energy dispersion upon mechanical damage. Abundant stacking faults and twins were further identified in the subgrains. It is interesting that the twins that are present in the subgrains are (100) twins along the [31¯1¯] shear direction, which have been determined to be the most common ones by far among the deformed grains of LaPO₄.¹³³

Properties of Y–Si–O oxides

Theoretical prediction of the properties of Y–Si–O oxides

Theoretical predictions of the mechanical properties of γ-Y₂Si₂O₇

The theoretical mechanical properties and atomistic shear deformation mechanisms of γ-Y₂Si₂O₇ were investigated using first-principles calculations by Wang et al.¹²³ Table 5 lists the calculated second-order elastic constants of γ-Y₂Si₂O₇, together with those of α-SiO₂ for comparison.¹²³ The elastic moduli representing stiffness against uniaxial strains, c ₁₁, c ₂₂, and c ₃₃ values of γ-Y₂Si₂O₇ are much higher than c ₄₄, c ₅₅ and c ₆₆ values, which correspond to the resistance to shear deformation. Table 5 also presents the calculated bulk modulus B, shear modulus G, and anisotropic Young’s moduli E of γ-Y₂Si₂O₇.¹²³ γ-Y₂Si₂O₇ shows 4·3 times and 2·1 times higher B (150 GPa) and G (77 GPa) than those of α-SiO₂, respectively. In addition, the Young’s moduli of γ-Y₂Si₂O₇ vary between 113 and 197 GPa, which are 1·3–2·2 times higher than those of α-SiO₂. Most interestingly, γ-Y₂Si₂O₇ exhibits low shear deformation resistance, which can be simply calculated from the ratio of shear to bulk modulus G/B.^{123,134–139} The ratio of G/B is 0·51 for γ-Y₂Si₂O₇, indicating that it is a damage tolerant ceramic.

Table 5.

Theoretical second-order elastic coefficients of γ-Y₂Si₂O₇ and α-SiO₂ ¹²³

Compound	c ₁₁	c ₃₃	c ₁₂	c ₁₃	c ₄₄	c ₆₆	c ₁₄	c ₂₂	c ₅₅	c ₂₃	c ₂₅	c ₄₆	c ₃₅
α-SiO₂	91	94	2	10	60	45	−12
γ-Y₂Si₂O₇	196	214	123	118	73	85	6	287	107	104	−19	−34	29

Compound	B/GPa	G/GPa	G/B	E/GPa
α-SiO₂	35	37	1·06	E _x = 88
				E_y = 88
				E_z = 92
γ-Y₂Si₂O₇	150	77	0·51	E _x = 113
				E _y = 197
				E _z = 131

The ductility of a ceramic is dominated by competition between cleavage and shear slip. If shear slip is activated more easily than cleavage, a ceramic may show intrinsic ductility.^{123,134–139} Wang et al. ¹²³ investigated the theoretical deformation modes by straining the unit cell along selected paths and studying the responses of chemical bonds to deformation. The ideal shear strength (12·8 GPa) of γ-Y₂Si₂O₇ is much lower than the ideal tensile strength (19·5 GPa). Based on their model, shear-induced deformation can be easily activated in γ-Y₂Si₂O₇ to generate shear slip, so that the concept of γ-Y₂Si₂O₇ ‘intrinsic ductility’ is thus endorsed. In the atomic deformation mechanism, the YO₆ octahedra are loosely bonded and readily accommodate strain on experiencing structural distortion, while the rigid Si₂O₇ pyrosilicates are resistant to applied shear strain. These weakly bonded YO₆ octahedra in the crystal structure of γ-Y₂Si₂O₇ line up to form a series of weakly bonded atomic interfaces and thus allow cleavage along certain atomic surfaces. The proposed shear deformation mechanism is similar to those of other ‘quasi-ductile’ ceramics such as LaPO₄ monazite,¹³⁵ layered ternary carbides, and nitrides (MAX-phases).^136–139

Theoretical predictions of optical and dielectric properties of γ-Y₂Si₂O₇ and Y₂SiO₅

The interband optical transitions with all dipole matrix elements for γ-Y₂Si₂O₇ and Y₂SiO₅ were calculated by Ching et al. ^121,122 The imaginary parts of the dielectric functions (ϵ₂(ω)) were calculated first and the real parts (ϵ₁(ω)) were obtained from the imaginary parts by the Kramers–Kronig conversion. The calculated optical dielectric constant ϵ₀ = ϵ₁ (ħω = 0) for Y₂Si₂O₇ and Y₂SiO₅ is 3·11 and 3·44, respectively. ϵ₀ can be related to the measured refractive index through n = √ϵ₀, and the refractive index for these two crystals is about 1·76 and 1·85, respectively. The low dielectric constants of yttrium silicates suggest promising applications of these materials in the semiconductor industry to replace SiO₂.

Mechanical properties of Y–Si–O oxides

Mechanical properties of γ-Y₂Si₂O₇

The mechanical properties of bulk γ-Y₂Si₂O₇ were measured by Sun et al.,⁵⁸ as summarised in Table 6, together with those of mica-glass ceramic (Macor),¹⁴⁰ LaPO₄ monazite,^141,142 and two end materials SiO₂ ¹⁴³ and Y₂O₃.¹⁴⁴ It is noteworthy that γ-Y₂Si₂O₇ has some close mechanical properties to Y₂O₃, but they are superior to those of machinable Macor and LaPO₄, and most importantly, are not just the weighted average of the two end members SiO₂ and Y₂O₃.

Table 6.

Mechanical properties of γ-Y₂Si₂O₇ and Y₂SiO₅ together with those of other selected oxide ceramics^58,59

	Y₂Si₂O₇	Y₂SiO₅	LaPO₄	Macor	SiO₂	Y₂O₃
Young’s modulus, E/GPa	155±3	124±2	134	66	73	169
Shear modulus, G/GPa	61	47	53	25	31	65
Poisson ratio, υ	0·27	0·31	0·275	0·29	0·17	0·304
Bulk modulus, B/GPa	112	108	99·3	52	41	143
Flexure strength, σ_b/MPa	135±4	116±3	101·82	94	55	145
Compressive strength, σ_c/MPa	650±20	620±13	/	345	1100	1100
Fracture toughness, K _Ic/MPa·m^1/2	2·12±0·05	1·85±0·17	1±0·1	1·4±0·5	0·79	2·3
Vickers hardness, H/GPa	6·2±0·1	5·3±0·1	4·86	3	7·0	6·8
Machinability	Good	Good	Good	Good	Poor	Poor

Sun et al. observed the presence of local energy dissipation capacity around the Vickers indentations, as well as the non-catastrophic compression fracture mode, and the ‘layered structure feature’ on the fracture surfaces of γ-Y₂Si₂O₇, as shown in Fig. 7a–b , which demonstrates that this ceramic possesses good damage tolerance, similar to those of LaPO₄ and Macor.⁵⁸ Figure 7c presents a high magnification micrograph of a typical fracture surface of γ-Y₂Si₂O₇ that features many layered steps or cleavage characteristics, which were caused by crack deflection during penetration inside the grains, proving the existence of ‘weak interfaces’ within the crystal structures. When the crack propagation direction is parallel to the weak interfaces, cleavage occurs along the weak planes, or it is deflected across the grains, and thus the energy is consumed. Figure 7d presents the residual three-point bending strength as a function of Vickers indentation load, in which the slope was less than −1/3, indicating that γ-Y₂Si₂O₇ is a damage tolerant ceramic.¹⁴⁵

a Microhardness versus indentation load and a SEM image of a crack propagation path (inset) for γ-Y₂Si₂O₇; b high magnification micrograph of a crack produced by microindentation at 30 N; c high magnification micrograph of typical cleavage surface feature of γ-Y₂Si₂O₇; d residual three-point bending strength as a function of Vickers indentation load; e half-surface (upper) and side section (lower) views of Hertzian contact damage for γ-Y₂Si₂O₇ at a load of 900 N or an indentation pressure of 3·6 GPa; f optical image of γ-Y₂Si₂O₇ samples after drilling by carbide tools; g–i transmission electron microscope (TEM) images different magnifications of γ-Y₂Si₂O₇ on the deformation zone under Hertz indentation, showing the distribution of amorphous bands in the inner volume in contact with the indenter, while the bottom-right inset of g is an high resolution TEM (HRTEM) image of the amorphous band parallel to (001), showing the crystal potential projected along¹⁰⁰ and separated by a band with a periodic signal about 6 nm wide (inset of g)^58,132

A permanent impression after Hertzian indentation (Fig. 7e ),⁵⁸ which is confined in a well-defined hemispherical contact without brittle conical cracks, was left at the contact site. Within the impression, a series of shear faults were observed. In the cross-sectional image, the damage pattern resembles the continuous plastic zones in the subsurface of slip-line fields beneath indentations. The permanent plastic deformation and shear-dominated faults demonstrates the ‘quasi-plastic’ nature of γ-Y₂Si₂O₇.¹⁴⁶ Lin et al. observed the deformation zone of γ-Y₂Si₂O₇ after Hertzian indentation via HRTEM.¹³² The volume directly beneath the indent comprises nanometre-sized grains delaminated by an amorphous phase layer, while dislocations dominate in the periphery, either as dense slip bands in the border of the indent or, further away, as individual dislocations. As shown in Fig. 7g–i , the deformation centre zone consists of regular lozenge-shaped crystalline cells coexisting with slightly elongated ones, and all these crystalline cells are surrounded by amorphous bands ∼6 nm in thickness that lie along the (001), (010), , and planes, as well as near the planes. The presence and the organisation of the amorphous layers contribute to the exceptional damage tolerance of γ-Y₂Si₂O₇.¹³² Owing to the good damage tolerance, low shear deformation resistance, and low hardness, γ-Y₂Si₂O₇ can be machined by conventional carbide tools (Fig. 7f ). From a practical point of view, the machinability is a big advantage of γ-Y₂Si₂O₇, as it means that complex shape components can be machined by conventional metal-working facilities. Moreover, low shear resistance implies the potential of using γ-Y₂Si₂O₇ as weak-bonding interface material in fibre-reinforced ceramic-matrix composites.

Mechanical properties of X2-Y₂SiO₅

Although a number of experiments have been carried out on coating X2-Y₂SiO₅ on C/C composites,^82–99 systematic work has seldom been conducted on the thermal and mechanical properties of Y₂SiO₅.

The mechanical properties of X2-Y₂SiO₅ are listed in Table 6. X2-Y₂SiO₅ shows close mechanical properties to those of γ-Y₂Si₂O₇ and LaPO₄. Y₂SiO₅ has lower hardness and a lower shear modulus, however, than γ-Y₂Si₂O₇ and LaPO₄, but higher fracture toughness than LaPO₄. The resemblance of Y₂SiO₅ to γ-Y₂Si₂O₇ and LaPO₄ with respect to mechanical properties suggests that Y₂SiO₅ also possesses lower shear deformation resistance and good machinability.^58,59 It should also be noted that Y₂SiO₅ has low Young’s modulus (124 GPa), but not as low as 20 GPa as reported by Kondo et al. ⁵⁶

According to its crystal structure, the weaker Y–O bonds in Y₂SiO₅ respond to mechanical perturbations more significantly than the strong Si–O bonds. When damage occurs, bond breaking mainly takes place at Y–O bonds along some particular weak planes, while the rigid SiO₄ tetrahedra tend to release the energy by rotation.⁵⁹ Figure 8a reveals that the grains in the deformation zone are commonly microcleaved into a number of well-orientated subgrains, suggesting that microcleavage is the major mechanism of energy dispersion upon mechanical damage. The HRTEM image (Fig. 8b ) shows that stacking faults and twins exist in the subgrains.⁵⁹ The upper-right HRTEM fringes in Fig. 8b presents a (100) twin along the shear direction, as indicated by the corresponding electron diffraction pattern at the lower-right. Hay et al.^133,146 studied the deformation twinning modes in LaPO₄, and found that (100) twins were the most common ones in the deformed grains. Similar conclusions were also reached for Er₂Si₂O₇ and Nb₂Si₂O₇.^147,148 Based on the similar crystal structures, the low-energy and weakly bonded (100) atomic planes in Y₂SiO₅ respond to applied mechanical perturbations by microcleavage, slippage, and twinning.

a Grains in the deformation zone of Y₂SiO₅ after Hertz indentation, and b high resolution transmission electron microscopy (HRTEM) image of the deformed grains; the upper-right HRTEM pattern presents a (100) twin along the shear direction, as indicated by the corresponding electron diffraction pattern at the lower-right⁵⁹

Thermal properties of Y–Si–O oxides

Recently, some papers have focussed on the application of yttrium silicates as EBCs, and reports on them for high-κ thin films were published.^{93,96,100–102} The knowledge of some key thermal properties of Y–Si–O ceramics, however, was lacking for a long time, because of the difficulties in the synthesis and sintering of single-phase Y–Si–O ceramics. The TECs of Y₂Si₂O₇ and Y₂SiO₅ vary over a wide range in literature [(2·2–4·6)×10⁻⁶ K⁻¹ for γ-Y₂Si₂O₇ and (3·1–19)×10⁻⁶ K⁻¹ for Y₂SiO₅].^{41,60,61,83,109,113,114,127,149,150} The misuse of TECs for coating design resulted in severe cracking and spalling of the coatings during tests.^41,82–84

Thermal properties of Y₂Si₂O₇

Thermal expansion of Y₂Si₂O₇

Table 7 lists the TEC of Y₂Si₂O₇, obtained by monitoring the variation of the unit cell as a function of temperature via high-temperature XRD and by directly measuring the length change with temperature via a dilatometer.^60,3,109,117 Dolan et al. ¹¹⁷ reported the TECs of some Y₂Si₂O₇ polymorphs (α, β, γ, and δ-Y₂Si₂O₇) obtained by the XRD powder diffraction method. The TECs are only around 4·0×10⁻⁶ K⁻¹ for β- and γ-Y₂Si₂O₇, while those for α- and δ-Y₂Si₂O₇ are around 8·0×10⁻⁶ K⁻¹. The reported linear TEC of γ-Y₂Si₂O₇ by Fukuda and Matsubara¹⁰⁹ was (2·2–3·8)×10⁻⁶ K⁻¹ in the range of 504–1473 K. The volume TEC of γ-Y₂Si₂O₇ reported by Sun et al. ⁶⁰ was 6·68×10⁻⁶ K⁻¹. Figure 9a shows the temperature-dependent XRD patterns of γ-Y₂Si₂O₇, and Fig. 9b presents the change in the unit cell parameters with temperature from 300 to 1373 K.³⁹ No phase transformation appeared in the heating process.

a Temperature-dependent X-ray diffraction (XRD) patterns of γ-Y₂Si₂O₇, and b change of the unit cell parameters with temperature in the temperature range from 300 to 1373 K; c linear thermal expansion behaviour of polycrystalline γ-Y₂Si₂O₇ measured by push-rod dilatometer; and d temperature dependence of the heat capacity for γ-Y₂Si₂O₇ ⁶⁰

Table 7.

Reported thermal expansion coefficients (TECs) for Y₂Si₂O₇ and Y₂SiO₅

Material	Temperature/K	TEC (×10⁻⁶/K)	References
Polycrystalline Y₂Si₂O₇ bulk material (54·5% of TD^§)	300–1273	α = 4·6	Aparicio and Duran⁸³
γ-Y₂Si₂O₇ powders	504–1473	α = 2·2–3·8	Fukuda and Matsubara¹⁰⁹
α-Y₂Si₂O₇ powders	293–1473	α = 8·0	Dolan et al. ¹¹⁷
β-Y₂Si₂O₇ powders	293–1673	α = 4·1
γ-Y₂Si₂O₇ powders	293–1473	α = 3·9
δ-Y₂Si₂O₇ powders	293–1673	α = 8·1
γ-Y₂Si₂O₇ powders	300–1573	β = 6·68	Sun et al. ⁶⁰
Polycrystalline γ-Y₂Si₂O₇ bulk material (97% of TD)	300–1573	α = 3·9	Sun et al. ⁶⁰
Single-crystal Y₂SiO₅	298–798	α_a = 0·6	O’Bryan et al. ¹⁴⁹
		α_b = 7·5
		α_c = 11·4

Y₂SiO₅ powders	293–1123	α_a = 5·2	Nowok et al. ¹¹³
		α_b = 7·5
		α_c = 10·7
	1123–1673	α_a = 19·5
		α_b = 12·5
		α_c = 25
X1-Y₂SiO₅ powders	394–1273	α = 5·0–8·7	Fukuda and Matsubara¹¹⁴
X2-Y₂SiO₅ powders	394–1473	α = 5·0–7·7	Fukuda and Matsubara¹¹⁴
Polycrystalline bulk material (Y₂SiO₅+Y_4·67(SiO₄)₃O)	300–1623	α = 5·3	Ogura et al. ⁴¹
Polycrystalline bulk material (58% of TD) (Y₂SiO₅+Y₂O₃)	300–1273	α = 6·9	Aparicio and Duran⁸³
Polycrystalline bulk Y₂SiO₅material	300–1173	α = 3·05	Wagner et al. ¹⁵⁰
Polycrystalline bulk Y₂SiO₅material	1173–1673	α = 4·09	Wagner et al. ¹⁵⁰
Single-phase polycrystalline Y₂SiO₅ bulk material (98% of TD)	300–1573	α = 8·36	Sun et al. ⁶¹

TD: theoretical density

The TEC of polycrystalline Y₂Si₂O₇ was reported by Aparicio and Duran⁸³ and Sun et al. ⁶⁰ using a high-temperature dilatometer. The sample that was used by Aparicio and Duran was a polycrystalline bulk Y₂Si₂O₇ with a relative density of 54·5% (2·2 g cm⁻³). The reported linear TEC of the porous γ-Y₂Si₂O₇ by Aparicio and Duran was 4·6×10⁻⁶ K⁻¹.⁸³ Sun et al. reported the linear thermal expansion behaviour of polycrystalline γ-Y₂Si₂O₇ up to 1300°C, measured by a push-rod dilatometer on a single-phase γ-Y₂Si₂O₇ bulk sample with a relative density of 97% (Fig. 9c ).⁶⁰ The average linear TEC is (3·90±0·4)×10⁻⁶ K⁻¹, which is quite close to that reported by Dolan et al. ¹¹⁷

Heat capacity of Y₂Si₂O₇

Cordfunke et al. ¹⁵¹ and Sun et al. ⁶⁰ measured the heat capacity of Y₂Si₂O₇ by the drop-calorimetry and laser-flash methods, respectively (Fig. 9d ). The temperature-dependent heat capacity of Y₂Si₂O₇ is C _p = 250·5+24·6×10⁻³ T−72·64×10⁵ T ⁻².⁶⁰ The experimental heat capacity data^60,151 agreed well with the heat capacity derived from the summation of the heat capacities of Y₂O₃+2SiO₂.¹⁵² The calculated temperature-independent heat capacity of Y₂Si₂O₇ is about 275 J mol⁻¹ K⁻¹

Thermal conductivity of Y₂Si₂O₇

The temperature-dependent thermal conductivity of Y₂Si₂O₇ is κ = 1·039+1162T ⁻¹.⁶⁰ Clarke and Phillpot^153,154 proposed a method to calculate the minimum thermal conductivity of a material and used it to choose suitable materials for ETBC application, as shown in Table 8. It is clearly shown that γ-Y₂Si₂O₇ has a low minimum thermal conductivity of 1·35 W m⁻¹ K⁻¹, suggesting that it is a promising candidate for EBC/thermal barrier coating (TBC) applications.

Table 8.

Calculated minimum thermal conductivity κ_min for γ-Y₂Si₂O₇, Y₂SiO₅, and some commonly used thermal barrier coating (TBC) materials^{60,61,153,154}

Compound	κ_min/W m⁻¹ K⁻¹	Compound	κ_min/W m⁻¹ K⁻¹	Compound	κ_min/W m⁻¹ K⁻¹
SiC	3·00	MgAl₂O₄	2·34	γ-Y₂Si₂O₇	1·35
Al₂O₃	2·89	Mg₂SiO₄	2·00	La₂Zr₂O₇	1·20
MgO	2·56	Mullite	1·68	LaPO₄	1·13
TiO₂	2·07	LaMgAl₁₁O₁₉	1·48	Y₂SiO₅	1·13

Thermal shock resistance of Y₂Si₂O₇

Sun et al. ⁶⁰ investigated the thermal shock resistance of γ-Y₂Si₂O₇ by the water quenching method and calculated the theoretical thermal shock resistance parameters R, R' and R''' according to Refs. 155 and 156 (Table 9). The experimental R value, the critical temperature difference (ΔT _c) in water quenching, was 300 K for γ-Y₂Si₂O₇.⁶⁰ Based on Table 9, γ-Y₂Si₂O₇ has a good thermal shock resistance among the various oxide ceramics, according to either the calculated thermal shock resistance parameters or the experimental R value.

Table 9.

Thermal shock resistances R, R' and R''' for γ-Y₂Si₂O₇ and some selected ceramics⁶⁰

Ceramic	σ_f/MPa	E/GPa	μ	α (×10⁻⁶/K)	κ/W m⁻¹ K⁻¹	R/K		R' (×10³ W m W⁻¹)	R'''/m³ MJ⁻¹
Ceramic	σ_f/MPa	E/GPa	μ	α (×10⁻⁶/K)	κ/W m⁻¹ K⁻¹	Cal.	Exp.	R' (×10³ W m W⁻¹)	R'''/m³ MJ⁻¹
HPSN	750	310	0·27	3·2	17	550	400–800	9·5	0·8
RBSN	240	220	0·27	3·2	15	250	200–600	3·7	5
RBSC	500	410	0·24	4·3	84	215	400	18	2
Al₂O₃	500	390	0·22	9·0	8	110	200–225	0·9	2
BeO	200	400	0·34	8·5	63	40	300	2·5	15
PSZ	634	205	0·22	9·0	8·7	270	300–375	2·4	0·7
γ-Y₂Si₂O₇	135	155	0·27	3·9	4·9	165	300	0·8	11
MgAl₂O₄	176	260	0·24	5·4	15	90	/	1·4	11

Thermal properties of Y₂SiO₅

Thermal expansion coefficient of Y₂SiO₅

The measured TECs of Y₂SiO₅ show a wide deviation from author to author, as shown in Table 7. O’Bryan et al. ¹⁴⁹ measured the thermal expansion of single-crystal Y₂SiO₅ in the principal crystallographic direction and the two orthogonal directions on the (010) plane in the temperature range of 298–973 K using a dilatometer. Anisotropic TECs ranged from 0·6×10⁻⁶ K⁻¹ in the [100] direction to 11·4×10⁻⁶ K⁻¹ in the [001], and the volume TEC of 17×10⁻⁶ K⁻¹ were given. The anisotropic thermal expansion of Y₂SiO₅ was also reported by Nowok et al. ¹¹³ and Fukuda and Matsubara¹¹⁴ using high-temperature XRD. During the measurement, a phase transition from X1- to X2-Y₂SiO₅ at 1123 K was observed by Nowok et al.,¹¹³ and the measured TEC of X2-Y₂SiO₅ is much larger than that of X1-Y₂SiO₅, as shown in Table 7. The TECs along the b- and c-axes reported by Nowok et al. and O’Bryan et al. are in excellent agreement, however, the TEC of the a-axis shows a large deviation, which may have resulted because the [100] and [001] directions are not parallel to the principal axes of the expansion ellipsoid. Fukuda and Matsubara¹¹⁴ examined the change in the unit cell dimensions with temperature up to 1273 K for X1-Y₂SiO₅ and 1473 K for X2-Y₂SiO₅. The mean linear TEC of X1-Y₂SiO₅ is higher than that of X2-Y₂SiO₅. The TEC increased from 5×10⁻⁶ K⁻¹ (394 K) to 8·7×10⁻⁶ K⁻¹ (1273 K) for X1-Y₂SiO₅, and from 5×10⁻⁶ K⁻¹ (394 K) to 7·77×10⁻⁶ K⁻¹ (1473 K) for X2-Y₂SiO₅. The mean linear TEC of bulk Y₂SiO₅ was also reported by Ogura et al.,⁴¹ Aparicio and Duran,⁸³ Wagner et al. ¹⁵⁰ and Sun et al. ⁶¹ Possibly because of the presence of impurities or the low density of some samples, the reported TEC data varied from 3·05×10⁻⁶ K⁻¹ to 6·9×10⁻⁶ K⁻¹.^41,83,150 Sun et al. measured the TEC on a single-phase polycrystalline X2-Y₂SiO₅ sample with a relative density of 98%. The TEC of 8·36×10⁻⁶ K⁻¹ ⁶¹ is much larger than the previously reported data. Obviously, the big difference in the TECs of the Y₂SiO₅ and the SiC substrate resulted in a large thermal mismatch and finally, the failure of the Y₂SiO₅ coatings.

Heat capacity, thermal diffusivity and thermal conductivity of Y₂SiO₅

The temperature dependence of the heat capacity, thermal diffusivity, and thermal conductivity of Y₂SiO₅ was reported by Sun et al. ⁶¹ The temperature-dependent heat capacity of X2-Y₂SiO₅ obeys the relationship C _p = 177·48+29·68×10⁻³ T−59·2×10⁵ T ⁻². The molar heat capacities extrapolated from this relationship are 120 J mol⁻¹ K⁻¹ at room temperature (300 K) and 215 J mol⁻¹ K⁻¹ at high temperature (1400 K). Thermal diffusivity of Y₂SiO₅ can be fitted by a second-order polynomial equation, as D _th = 0·01124−1·314×10⁻⁵ T+6·331×10⁻⁹ T ². The thermal conductivity of Y₂SiO₅ has an inverse proportional relationship with temperature: κ = 1·138−216·2T ⁻¹. The extrapolated thermal conductivity of Y₂SiO₅ is 1·86 and 1·29 W m⁻¹ K⁻¹, respectively, at 300 K and 1400 K.

Therefore, Y₂SiO₅ has a much lower thermal conductivity than the most commonly used TBC materials.^140,160,161 Figure 10a displays the thermal conductivities of some TBC materials as a function of temperature.^60,61 The thermal conductivity of a dense Y₂SiO₅ is only slightly higher than those of an air plasma-sprayed 7 mol.-% yttria stabilised zirconia (APS 7YSZ) coating and monoclinic zirconia.¹⁴⁰ Figure 10b presents the TECs of some coating materials and nickel based superalloys versus thermal conductivity.⁶¹ According to Fig. 10b , Y₂SiO₅ can be selected as a coating material for nickel based superalloys, if only the thermal expansion and thermal conductivity are considered. Moreover, Y₂SiO₅ is also an excellent environmental barrier coating on top of YSZ TBC layer.

10.

a Thermal conductivities of some selected thermal barrier coating (TBC) materials as a function of temperature, and b thermal expansion coefficient (TEC) as a function of thermal conductivity for some coating materials and nickel based superalloys⁶¹

Tribological properties of Y–Si–O oxides

Tribological properties of γ-Y₂Si₂O₇

Ceramics are candidates for wear-resistance applications especially under severe environmental conditions.¹⁶² Thus, wear characteristics in different environments need to be well understood. Sun et al. ¹⁶³ investigated reciprocating ball-on-flat dry sliding friction and wear behaviour of γ-Y₂Si₂O₇ ceramic flats in contact with AISI 52100 bearing steel and Si₃N₄ balls at 5–15 N normal loads in the ambient environment. The kinetic friction coefficients (COFs) of γ-Y₂Si₂O₇ varied between 0·53 and 0·63 against AISI 52100 steel, and between 0·51 and 0·56 against Si₃N₄. The wear rates of γ-Y₂Si₂O₇ were on the order of 10⁻⁴ mm³ (N⁻¹ m⁻¹). Wear occurred predominantly during the running-in period and almost ceased at the steady friction stage. The wear debris played an important role in the steady friction stage. When against AISI 52100 bearing steel, a thick transfer layer was left both in the worn tracks of the flats and in the contact surfaces of the counterpart balls, owning to the high chemical affinity between γ-Y₂Si₂O₇ and the AISI 52100 balls. In contrast, the worn surfaces rubbed by Si₃N₄ were much smoother, the asperities on the surface were removed, and the worn tracks were all covered with a continuous dense debris layer. The brittle microfracture of γ-Y₂Si₂O₇ grains at the running-in stage is the major mechanism. The adhesion wear and tribochemical wear are the major mechanisms responsible at the steady friction stage. The severe running-in wear indicates that the strengthening of γ-Y₂Si₂O₇ is beneficial for wear applications.

Tribological properties of γ-Y₂Si₂O₇/ZrO₂ composites

To improve the mechanical and tribological properties of γ-Y₂Si₂O₇, Sun et al. ¹⁶⁴ developed a series of Y₂Si₂O₇/ZrO₂ composites, which are promising as protective coatings against thermal loads and chemical attack.¹⁶⁵ Compared to γ-Y₂Si₂O₇, γ-Y₂Si₂O₇/ZrO₂ composites showed improved mechanical properties, e.g. the Young’s modulus was enhanced from 155 to 188 GPa (121%), and the flexural strength from 135 to 254 MPa (181%), when 50 vol.-% ZrO₂ was incorporated.

The COFs of the Y₂Si₂O₇/ZrO₂ composites against Si₃N₄ balls vary in the range of 0·49–0·59 under 5 N but vary in a narrow range of 0·69–0·76 under 15 N. The wear rates of the composites against Si₃N₄ balls were independent of ZrO₂ content under low-load conditions; but decreased from 7·53±0·25 mm³ N⁻¹ m⁻¹ for the Y₂Si₂O₇–10 vol.-% ZrO₂ composite to (0·64±0·21)×10⁻⁴ mm³ N⁻¹ m⁻¹ for the Y₂Si₂O₇–50 vol.-% ZrO₂ composite under 15N.

The composition–mechanical properties–tribology relationships of Y₂Si₂O₇/ZrO₂ composites were elucidated by Sun et al. (Fig. 11).¹⁶⁵ The wear resistance of the composites was not only influenced by the applied load, hardness, strength, toughness, and rigidity, but also by the micromechanical stability of the microstructures. To achieve high wear resistance, a compromise between the flexural strength and the fracture toughness should be reached.

11.

Plots of correlations of tribological properties of the composites with the basic mechanical parameters: a friction coefficients (COF) vs. flexural strength, b wear rate vs. flexural strength; c COF vs. fracture toughness, d wear rate vs. fracture toughness; e the ratio of the hardness to Young’s modulus (H/E) as a function of ZrO₂ content for the γ-Y₂Si₂O₇/ZrO₂ composites, and f variation of COF and wear rate against H/E ¹⁶⁴

Environmental durability of Y–Si–O oxides

Oxygen permeability in Y₂SiO₅

Ogura et al. ⁷⁴ measured the oxygen permeability constant through an Y₂SiO₅ wafer from 1973 to 2033 K. The oxygen permeability was measured in a two-chamber apparatus, in which the upper chamber contained a mixture of argon and oxygen, and the bottom chamber contained high-purity argon. The two chambers imposed an oxygen potential gradient across the test wafer. The detailed theoretical basis of the permeability testing can be found in Ref 74. At 1973 K, the oxygen permeability constant of Y₂SiO₅ is 10⁻¹⁰ kg m⁻¹ s⁻¹. The activation energy of diffusion, E _d, varied between 227 and 349 kJ mol⁻¹, which is close to the energy of vacancy diffusion and much higher than the energy of interstitial diffusion of molecular oxygen. Vacancy diffusion was considered dominant in oxygen transport through Y₂SiO₅ below 1913 K. On the contrary, interstitial diffusion is dominant above 1913 K.

Argirusis et al ⁹² also studied the oxygen self-diffusion in Y₂SiO₅ by the isotope exchange depth profiling method. The single-crystal sample was put into a furnace and equilibrated at the diffusion temperature (1150–1530°C) in a ¹⁶O₂ atmosphere. Then, the atmosphere was rapidly changed from ¹⁶O₂ to ¹⁸O₂. At 1505°C, the oxygen self-diffusion coefficient was determined to be 1·5×10⁻¹² cm² s⁻¹, and the activation energy was 3·5 eV (∼338 kJ mol⁻¹). This activation energy coincides well with the value reported by Ogura et al. ⁷⁴ Although the oxygen diffusion coefficient in single-crystal Y₂SiO₅ is two orders of magnitude higher than that of single-crystal mullite, the oxygen diffusivity of Y₂SiO₅ is still low enough to render it a good oxidation protection material for SiC–C/C composites.⁹²

Hot corrosion resistance of Y₂Si₂O₇

Hot corrosion behaviour of γ-Y₂Si₂O₇ in strong basic Na₂CO₃ molten salt environment

Sun et al. ⁶² studied the hot corrosion behaviour of γ-Y₂Si₂O₇ in Na₂CO₃ molten salt at various temperatures (850–1000°C). Compared with Na₂SO₄, Na₂CO₃ decomposes more readily to Na₂O and CO₂, so it is usually used to study the corrosion under strong basic conditions. Almost all silicon-based ceramics are severely corroded in Na₂CO₃ molten salt due to the dissolution of passive SiO₂ thin film in strong basic molten salt.^{94,95,166–171} In the hot corrosion of Y₂Si₂O₇ in Na₂O melt (since Na₂CO₃ decomposes into Na₂O and CO₂ at the test temperatures), it was found that Na₂O melt attacked the grain boundaries first and then penetrated into the inner part of the sample to surround the grains and then react with Y₂Si₂O₇. Subsequently, the YO₂ ⁻ ions that were formed in the melting product would diffuse outwards due to the basicity (or Na₂O) gradient between the oxide/salt interface and the salt/gas interface,^170,171 and then precipitate as a layer of non-protective, loose Y₂O₃ particles at the salt/gas interface. On the other hand, SiO₃ ²⁻ could not re-precipitate from the melt, since the solubility of SiO₂ is independent of Na₂O activity under basic conditions,¹⁷¹ so it is left as Na₂O.xSiO₂ at the oxide/salt interfaces. Once x = 3·65 was approached in the Na₂O.xSiO₂ melt, a corrosion-protective silica layer would then precipitate beneath the silicate melt and eventually seal off the inner grains from further corrosion. In general, γ-Y₂Si₂O₇ showed severe corrosion in the extremely strong basic Na₂CO₃ molten salt. The thicknesses of the corrosion scales were less than 90 μm, however, after 20 h exposure at 1000°C, which is attributable to the formation of a thin layer of protective SiO₂ under the Na₂O.xSiO₂ melt that protects the inner material from further corrosion.

Hot corrosion behaviour of γ-Y₂Si₂O₇ in weak acidic Na₂SO₄ molten salt environment

According to the hot corrosion theory proposed by Rapp,¹⁷¹ the oxyanion sodium sulphate melt can be described by an acid-base chemistry in a manner analogous to the pH of aqueous solutions. For pure Na₂SO₄, . Thus, can be defined as a quantitative measure of the melt basicity, or alternatively, can be defined as the melt acidity. Sun et al. ⁶³ studied the hot corrosion behaviour of γ-Y₂Si₂O₇ in Na₂SO₄ molten salt in the temperature range of 850–1000°C. A low-temperature Na₂O.3SiO₂ liquid phase formed at 776°C at first by the rapid reaction between sodium sulphate and grain boundary silica. Then, the Na₂SO₄ and Na₂O.3SiO₂ liquid melts penetrated into the grain boundaries and reacted with the yttrium disilicate grains to form a Na–Y–Si–O melt. The weight losses during hot corrosion were caused by the evaporation of Na₂SO₄ molten salt and the evolution of SO₂+O₂ gases, while the hot corrosion reaction obeyed a paralinear law, ,⁶³ the linear part is for the linear evaporation of Na₂SO₄ molten salt, and the parabolic part for the parabolic mode of the hot corrosion reaction that contributes to the release of SO₂+O₂. The parabolic mode for the hot corrosion reaction between Na₂SO₄ and Y₂Si₂O₇ substrate, with an activation energy of 255 kJ mol⁻¹, was caused by the silica barrier that was formed under the hot corrosion scale. The silica film formed under the corrosion scale separated the direct interaction between sodium molten salts and the Y₂Si₂O₇ grains, and thus protected the inner material from further etching.

Water vapour corrosion of Y₂Si₂O₇ and Y₂SiO₅

The corrosion behaviour of Y₂Si₂O₇ containing Y, Yb, and Lu was investigated by Maier et al. ¹⁷² in a gas stream containing water at 1500°C. They found that the facilities for water vapour corrosion testing had a strong influence on the experiments: silica or silica-forming tubing causes high internal P _Si(OH)4, which should artificially slow down corrosion rates; and alumina tubing causes alumina contamination via P _Al(OH)3, which led to alumina transfer to the test samples.¹⁷² In a similar way to silica, the water vapour corrosion of rare-earth silicates also results in the formation of volatile silicon hydroxide as: Re₂Si₂O₇+2H₂O→Re₂SiO₅+Si(OH)₄ (g).¹⁷² Courcot et al. ¹⁷³ also studied the water vapour induced corrosion of Y₂SiO₅ and Y₂Si₂O₇ at 1400°C for 300 h under 50 kPa H₂O partial pressure. They observed that the volatilisation of both Y(OH)₃ and Si(OH)₄ owing to the reaction with water during corrosion. The monitored mass change showed that Y₂SiO₅ had a weight gain, whereas Y₂Si₂O₇ exhibited a weight loss. The difference in weight change was mainly caused by the reaction of Al(OH)₃ from the sample holder with Y₂SiO₅ and Y₂Si₂O₇ to form yttrium aluminium garnet (YAG). Since Y₂Si₂O₇ had a lower reactivity with Al(OH)₃ than Y₂SiO₅, less YAG was detected at the end of the corrosion test and resulted in a weight loss. It was also reported that the preparation methods have an influence on the corrosion resistance. The material synthesised by the sol-gel route was less resistant against corrosion.

Therefore, it seems that the water vapour corrosion on Y₂Si₂O₇ and Y₂SiO₅ is influenced by many factors, such as the preparation method, heating history, impurity, morphologies, etc., and further investigations are still needed. Moreover, a rapid reaction between yttrium silicates and alumina-containing compounds was observed, indicating that the water vapour corrosion testing system has to be improved, and an environment containing the aluminium element should be avoided for the high-temperature application of yttrium silicates.

Dielectric properties of Y₂Si₂O₇

Cao et al. ^174,175 studied the dielectric properties of γ-Y₂Si₂O₇ in the X-band from room temperature to 1400°C. In the measured frequencies (8·0, 9·1, 10·3 and 11·8 GHz), γ-Y₂Si₂O₇ exhibits a low dielectric loss, which is on the order of 10⁻³ in magnitude. The permittivity maximum and the loss tangent maximum are 5·71 and 8·3×10⁻³, respectively, in the tested temperature range. There is very weak frequency dependence, but strong temperature dependence of its permittivity. The permittivity increases monotonically with temperature at all investigated frequencies. The strong temperature dependence of the permittivity indicates that γ-Y₂Si₂O₇ may present special polarisation in a microwave field. According to Cao et al,^174,175 this corresponds to a non-Debye relaxation process with a dielectric relaxation time of 10⁻¹¹ s at 800°C, which is attributed to the movement of thermal-excitation structural defects that are responsible for structural relaxation polarisation.

Structure-property relationships of Y–Si–O oxides

Wang et al. ¹²³ calculated the electronic structure and predicted the mechanical properties of γ-Y₂Si₂O₇. The low shear to bulk modulus ratio and the low ideal shear to tensile strength ratio indicate its low shear deformation resistance and damage tolerance. The Y–O bond is weaker and readily stretches and shrinks, while the Si–O bond is strong and more rigid. The relatively soft YO₆ octachedron positively accommodates shear deformation by structural distortion, while the Si₂O₇ pyrosilicate unit is more resistant to deformation. The unique crystal structure and nature of the chemical bonding of γ-Y₂Si₂O₇ thus lead to the low shear deformation resistance. This kind of deformation mechanism was later proved by experimental mechanical properties evaluation and microstructure characterisation. Figure 12 presents the possible weakly bonded atomic planes in γ-Y₂Si₂O₇ (a–d), the valence electron density isosurface of γ-Y₂Si₂O₇ projected on the (001) plane (e), and the microstructure of γ-Y₂Si₂O₇ after deformation (f). It clearly shows that the inhomogeneous distribution of the chemical bonding in some special planes contributes to the low shear deformation resistance of the mechanical response mode.

12.

a–d Possible weakly bonded atomic planes in γ-Y₂Si₂O₇, e valence electron density isosurface of γ-Y₂Si₂O₇ projected on the (001) plane, and f microstructure of γ-Y₂Si₂O₇ after deformation¹³²

Cao et al. ^174,175 observed a non-Debye relaxation behaviour different from that of SiO₂ when measuring the dielectric properties of γ-Y₂Si₂O₇, which is actually also associated with the crystal structure of γ-Y₂Si₂O₇. With increasing temperature, the Y–O bonds become much weaker, which results in the elongation or rupture of some Y–O bonds, so that the rigid Si₂O₇ pyrosilicates and Y ions can rotate or move more easily at high temperatures. Since the (100) and (010) planes are the most common weak interfaces, the thermal-excitation atomic groups (rigid Si₂O₇ pyrosilicates and Y ions) would move easily along the direction of the electric field when either plane is perpendicular to the electric field. The local motions of the thermal-excitation atomic groups, which are actually attributed to the movement of structural defects, would create structural relaxation polarisation in electromagnetic fields.

Sun et al. ⁵⁹ also observed the same relationship between the crystal structure and mechanical properties in investigating the mechanical properties of X2-Y₂SiO₅. The crystal structure of Y₂SiO₅ suggests that the low-energy weakly bonded atomic planes are crossed only by the easily broken Y–O bonds. The weakly bonded Y–O atomic planes as well as the rotationally rigid SiO₄ tetrahedra, are the origins of the low shear deformation, good damage tolerance, and easy machinability of Y₂SiO₅. Transmission electron microscopy observations (Fig. 9) also demonstrate a large number of microcleavages, stacking faults, and twins along these weakly bonded atomic planes, which disperse the strain energy and thus allow ‘ductility’ of Y₂SiO₅.

Besides the intrinsic crystal structure features, the grain boundaries also exhibit strong influences on the material properties. Up to now, no systematic work has been carried out on the grain boundaries of Y–Si–O oxides. MacLaren et al. ⁴⁵ carried out TEM observations on hot-pressed Y₂Si₂O₇ ceramic with 3% LiF fluxing additive. They found that the microstructure consisted of large grains of γ-Y₂Si₂O₇ containing rounded glassy Y-doped SiO₂ inclusions. At the grain boundaries, excess glassy SiO₂-rich material is dominant. The formation of silicon-rich grain boundaries in Y₂Si₂O₇ has a dramatic influence on the performance of this ceramic. Figure 13 shows the Young’s modulus (E) as a function of temperatures in γ-Y₂Si₂O₇, where silicon-rich grain boundaries were detected when excess silica was used for the synthesis.¹⁷⁶ As the temperature increases, the variation of E clearly presents three stages. From room temperature to 600°C (stage I in Fig. 13b ), the elastic modulus is linearly damped owing to the lattice expansion induced change in the atomic bonding force and swelling of the material. From 600 to 800°C (stage II), an abnormal increase in the elastic modulus was observed, which might be caused by the crystallisation of grain boundary silica. When the temperature is above 800°C (stage III), the elastic modulus of Y₂Si₂O₇ dramatically decreases. The fast degradation of stiffness at this stage is likely to be because of the softening of silica grain boundaries. Therefore, the high-temperature mechanical properties of Y₂Si₂O₇ are strongly influenced by the grain boundary silica.

13.

a Transmission electron microscopy (TEM) images of γ-Y₂Si₂O₇ with the inset corresponding selected area electron diffraction (SAED) pattern showing the existence of grain boundary glass phase, and b variation of the Young’s modulus of γ-Y₂Si₂O₇ as a function of temperature¹⁷⁶

During the environmental durability testing of γ-Y₂Si₂O₇,^62,63 the preferential etching of grain boundaries by molten salt was also observed. Without question, the glassy grain boundaries contribute to the rapid corrosion of γ-Y₂Si₂O₇ at the initial stage. In the water vapour corrosion tests, the grain boundaries may also affect the performance, but no related grain boundary characterisation was performed. The optimisation of the grain boundary structure/chemistry of Y–Si–O ceramics may further increase their corrosion resistance.

Potential applications of Y–Si–O oxides

Y–Si–O oxides are refractory with melting points higher than 1775°C. From the mechanical perspective, γ-Y₂Si₂O₇ and Y₂SiO₅ are readily machinable and can be shaped into complex parts by conventional cemented carbide tools. The low shear deformation resistance implies that γ-Y₂Si₂O₇ and Y₂SiO₅ are promising materials for complex shape components and weak interfacial phases in fibre-reinforced ceramic-matrix composites. γ-Y₂Si₂O₇ also has good resistance to strong basic Na₂CO₃ molten salt and weakly acidic Na₂SO₄ molten salt at temperatures up to 1000°C. Moreover, the critical temperature difference ΔT _c of γ-Y₂Si₂O₇ is 300 K, ensuring its tolerance to rapid temperature variations. Recently, Boakye et al. ¹⁷⁷ proved the feasibility of using γ-Y₂Si₂O₇ for weakly bonded interfacial phases in SiC_f/SiC composites. In their study, Y₂Si₂O₇ was coated on α-SiC and SiC fibre using SiO₂/Y(OH)₃ sol-gel precursors. The coatings were heat treated at 1400°C for 8 h in argon with 500 ppm oxygen. A mixed γ-Y₂Si₂O₇/β-Y₂Si₂O₇ coating was formed on the SiC fibres demonstrating that it is feasible to coat SiC fibres with oxidation resistant fibre matrix interphases in SiC_f/SiC composites (Fig. 14a–d ). Thereafter SiC fibre-reinforced Y₂Si₂O₇ matrix composites were fabricated, and the fibre push-out behaviour was studied.¹⁷⁸ Preliminary results showed that SCS-0 fibres in dense β, γ-Y₂Si₂O₇ matrices debonded and pushed-out with sliding stresses of 15–60 MPa, within the range reported for C, BN, and LaPO₄ coatings (Fig. 14e ). Fibres in the α-Y₂Si₂O₇ matrix showed poor and erratic push-out behaviour because of the radial compressive stresses that developed on the fibres as result of the differences in TECs (fibre ∼4 and matrix ∼8).

14.

Scanning electron micrographs of RE(OH)₃/SiO₂ coatings on SCS-0 SiC fibres heat treated in argon/500 ppm oxygen at 1400°C/8 h: a, b Y(HO)₃/SiO₂ 10 coats; c, d Ho(HO)₃/SiO₂ 20 coats; and e optical micrographs of pushed SCS-0 fibres in the β-Y₂Si₂O₇matrix^177,178

γ-Y₂Si₂O₇ and X2-Y₂SiO₅ have complementary linear TECs (∼3·9×10⁻⁶ K for γ-Y₂Si₂O₇ and 8·36×10⁻⁶/K for Y₂SiO₅) and low thermal conductivity (<3 W m⁻¹ K⁻¹ above 300°C for γ-Y₂Si₂O₇ and ∼1·6 W m⁻¹ K⁻¹ for Y₂SiO₅).^60,61 The unique thermal properties imply the high competitiveness of γ-Y₂Si₂O₇ and Y₂SiO₅ used together or separately as candidate materials for environmental/TBC. In particular, γ-Y₂Si₂O₇ can be coated on silicon-based ceramics, owing to its close linear TECs with silicon carbide or silicon nitride. γ-Y₂Si₂O₇ ceramic is also a promising communication window material for the re-entry space shuttles according to Cao et al. ^174,175 Therefore, Y–Si–O oxides have promising applications as high-temperature structural ceramics, machinable ceramics, environmental/TBC, and weakly bonded interphases for fibre-reinforced ceramic-matrix composites. Moreover, the defect formation and migration behaviours of Y–Si–O ternary oxides imply the functional potential of yttrium silicates as luminescent materials and solid-state ionic conductors. Y–Si–O oxide with apatite structure might be an especially promising low-cost electrolyte candidate for SOFC.

Conclusions

In this review, the crystal structure, preparation methods, mechanical properties, dielectric properties, and thermal properties, as well as the environmental durability of Y–Si–O ceramics are reviewed from both theoretical and experimental points of view. Some of the polymorphs of Y–Si–O oxides, such as γ-Y₂Si₂O₇ and X2-Y₂SiO₅, feature high damage tolerance, low shear deformation resistance, good machinability, tunable tribological properties, and excellent environmental durability in hazardous atmospheres. The unique multi-scale structure features ensure promising applications of this family of materials as high-temperature structural components, machinable ceramics, oxidation/thermal/EBCs, and weakly bonded interphases in fibre-reinforced composites, as well as luminescent materials, and solid-state ionic conductors.

Footnotes

Acknowledgements

This work was funded by the National Natural Science Foundation of China (NSFC) under grant Nos. 50802097, 50832008 and the IMR Innovation Research Foundation.

References

Hidden

and Mackintosh

: Amer. J. Sci., 1889, 38, 474–486.

Hata

: Sci. Paper Inst. Phys. Chem. Res., 1938, 34, 1018–1023.

Protoseko

: Zap. Vses. Mineral. Obschch., 1962, 91, 260–270.

Lazarev

TeNishnva

and Bondar

: Izv. Akad. Nauk, SSSR, Neorg. Mat., 1965, 1, 1207–1209.

Nishimuna

and Unea

: J. Geol. Soc. Jap., 1954, 60, 131–137.

Fenra

: Vunta Invest. Ultramar Portugal, Eslud. Ensaios Doc., 1964, 112, 33–34.

Warshaw

and Roy

: ‘Crystal chemistry of rare earth sesquioxides, aluminates and silicates,’ in ‘Progress in science and technology of the rare earths’, (ed. Leroy Eyring), Vol. 1, 215–221; 1964, New York, Pergamon Press.

Sidorenko

: Geol. Mestorozhd. Redk. Flem., Vses. Nach.-Issled. Inst. Mineral., 1966, 26, 125–129.

Bondar

and Toropov

: Mater. Res. Bull., 1967, 2, 479–489.

10.

Liddell

and Thompson

: Br. Ceram. Trans. J., 1986, 85, 17–22.

11.

Ito

and Johnson

: Am. Mineral., 1968, 53, 1940–1952.

12.

Feslche

: Struct. Bonding, 1973, 13, 99–199.

13.

Becerro

and Escudero

: Phase Transit., 2004, 77, 1093–1102.

14.

Lee

Drummond

Hilmas

and Kumar

: J. Am. Ceram. Soc., 1990, 73, 3575–3579.

15.

Lange

Singhal

and Kuznicki

: J. Am. Ceram. Soc., 1977, 60, 249–252.

16.

Gauckler

Hohnke

and Tien

: J. Am. Ceram. Soc., 1980, 63, 35–37.

17.

Ziegler

Idrobo

Cinibulk

Kisielowski

Browning

and Ritchie

: Science, 2004, 306, 1768–1770.

18.

Hong

Yoshida

Ikuhara

Nishimura

and Mitomo

: J. Eur. Ceram. Soc., 2002, 22, 527–534.

19.

Zhou

Hirao

Watari

Yamauchi

and Kanzaki

: J. Eur. Ceram. Soc., 2004, 24, 265–270.

20.

Hartenbach

Meier

and Schleid

: Z. Naturforsch., 2006, 61B, 1054–1060.

21.

Kahlenberg

Kaindl

and Konzett

: Solid State Sci., 2007, 9, 542–550.

22.

Harris

and Finch

: Am. Mineral., 1965, 50, 1493–1495.

23.

Levin

Robbins

and McMurdie

: ‘Phase diagrams for ceramists – 1969 Supplement’, 76; 1969, Fig. 2388, Columbus, OH, The American Ceramic Society Inc.

24.

Nikl

Nižňanský

Ruzicka

Cannas

and Yanagida

Takayuki

: ‘Silicate glass-based nanocomposite scintillators,’ in ‘Advances in nanocomposite technology’, (ed. A. Hashim), 201–226; 2011, InTech, Rijeka, Croatia.

25.

Giesche

and Matijević

: J. Mater. Res., 1994, 9, 436–450.

26.

Díaz

Garcia-Cano

Mello-Castanho

Moya

and Rodríguez

: J. Non-Cryst. Solids, 2001, 289, 151–154.

27.

Kahlenberg

Wertl

Tobbens

Kaindl

Schuster

and Schottenberger

: Z. Anorg. Allg. Chem., 2008, 634, 1166–1172.

28.

Parmentier

Bodart

Audoin

Massouras

Thompson

Harris

Goursat

and Bessons

: J. Solid State Chem., 2000, 149, 16–20.

29.

Wang

Tian

Liao

and Jing

: Mater. Res. Bull., 2001, 36, 1855–1861.

30.

Boyer

and Derby

: J. Am. Ceram. Soc., 2003, 86, 1592–1594.

31.

Cannas

Mainas

Musinu

Piccaluga

Speghini

and Bettinelli

: Opt. Mater., 2005, 27, 1506–1510.

32.

Zhou

Yang

and Gao

: J. Lumin., 2007, 124, 241–244.

33.

Lin

Wang

Kong

Liu

Wang

and Lin

: Eur. J. Inorg. Chem., 2006, 2006, 3667–3675.

34.

Parmentier

Liddell

Thompson

Lemercier

Schneider

Hampshire

Bodart

and Harris

: Solid State Sci., 2001, 3, 495–502.

35.

Becerro

Naranjo

Perdigón

and Trillo

: J. Am. Ceram. Soc., 2003, 86, 1592–1594.

36.

Díaz

Pecharromán

del Monte

Iglesias

Moya

Yamagata

and Mello-Castanho

: Chem. Mater., 2005, 17, 1774–1782.

37.

Trusty

Chan

and Ponton

: J. Mater. Res., 1998, 13, 3135–3143.

38.

Becerro

Naranjo

Alba

and Trillo

: J. Mater. Chem., 2003, 13, 1835–1842.

39.

Christensen

Hazell

and Hewat

: Acta Chem. Scand., 1997, 51, 37–43.

40.

Seifert

Wagner

Fabrichnaya

Lukas

Aldinger

Ullmann

Schmucker

and Schneider

: J. Am. Ceram. Soc., 2005, 88, 424–430.

41.

Ogura

Kondo

Morimoto

Takeda

and Notomi

: Mater. Trans., 2001, 42, 1124–1130 (in Japanese).

42.

Kumar

and Drummond

: J. Mater. Res., 1992, 7, 997–1003.

43.

Leskelä

and Jyrkäs

: J. Am. Ceram. Soc., 2005, 70, C160–C161.

44.

Aparicio

Moreno

and Duran

: J. Eur. Ceram. Soc., 1999, 19, 1717–1724.

45.

MacLaren

Schierholz

Trusty

and Ponton

: J. Am. Ceram. Soc., 2007, 90, 3307–3310.

46.

Sun

Zhou

and Li

: J. Mater. Res., 2006, 21, 1433–1450.

47.

Sun

Zhou

and Li

: J. Mater. Res., 2008, 23, 732–736.

48.

Leonyuk

Belokoneva

Bocelli

Righi

Shvanskii

Henrykhson

Kulman

and Kozhbakhteeva

: Cryst. Res. Technol., 1999, 34, 1175–1182.

49.

Zhang

Wang

Shen

Wang

Zhong

Zhang

and Xu

: J. Cryst. Growth, 1999, 197, 901–904.

50.

Günther

and Georg

: Acta. Cryst., 2003, C59, i103–i106.

51.

Pang

Zhao

Jie

and He

: Mater. Lett., 2005, 59, 3539–3542.

52.

Gordon

Becker

Hausmann

and Suh

: US patent, No. US 7507848 B2.

53.

Sun

and Kwok

: Appl. Phys. A, 1999, 69, S39–S43.

54.

Hanada

Shinoda

Tanabe

and Soga

: J. Am. Ceram. Soc., 1995, 78, 1383–1386.

55.

Kaya

Thomas

and Boccaccini

: Surf. Coat. Tech., 2005, 191, 303–310.

56.

Kondo

Ogura

and Morimoto

Mater. Trans., JIM, 1998, 39, 1146–1151.

57.

Chambers

and Parsons

Appl. Phys. Lett., 2000, 77, 2385–2387.

58.

Sun

Zhou

Wang

and Li

: J. Am. Ceram. Soc., 2007, 90, 2535–2541.

59.

Sun

Wang

and Zhou

: J. Eur. Ceram. Soc., 2008, 28, 2895–2901.

60.

Sun

Zhou

Wang

and Li

: J. Am. Ceram. Soc., 2008, 91, 2623–2629.

61.

Sun

and Zhou

: J. Eur. Ceram. Soc., 2009, 29, 551–557.

62.

Sun

and Zhou

: J. Eur. Ceram. Soc., 2008, 28, 259–265.

63.

Sun

and Zhou

: J. Am. Ceram. Soc., 2008, 91, 2236–2242.

64.

Aitasalo

Hölsä

Lastusaari

Niittykoski

and Pellé

: Phys. Stat. Solid. C, 2007, 2, 272–275.

65.

Pang

Zhao

Jie

and Xu

: Cryst. Res. Technol., 2006, 41, 255–258.

66.

Bondar

and Koroleva

: ACerS-NIST phase equilibria diagrams, CD-ROM database, V3·0·1, 06, Fig. 06576.

67.

Zou

and Petric

: J. Phys. Chem. Solids, 1994, 55, 493–499.

68.

Matovic

Rixecker

and Aldinger

: J. Am. Ceram. Soc., 2004, 87, 546–549.

69.

Bosze

Mckittrick

and Hirata

: Mater. Sci. Eng. B, 2003, 104, 47–54.

70.

Tzvetkov

and Minkova

: J. Mater. Sci. Lett., 2001, 20, 1273–1275.

71.

Huang

Deng

Cao

and Ge

: J. Syn. Cryst., 2007, 36, 464–466 (in Chinese).

72.

Argirusis

Chr

Damjanovic

and Schneider

: J. Mater. Sci., 2006, 41, 8059–8067.

73.

Kang

Lee

Kim

Lee

and Cho

: J. Eur. Ceram. Soc., 2004, 24, 1031–1040.

74.

Ogura

Kondo

Morimoto

Notomi

and Sekigawa

: Mater. Trans., 2001, 42, 1124–1130.

75.

Lange

: J. Am. Ceram. Soc., 1989, 72, 3–15.

76.

Lewis

: J. Am. Ceram. Soc., 2000, 83, 2341–2359.

77.

King

: ‘Ceramic technology and processing’, 175; 2002, Now York, Noyes Publications.

78.

Sun

Zhu

Zhou

and Sakka

: J. Am. Ceram. Soc., 2008, 91, 2521–2529.

79.

Sun

Zhu

Zhou

and Sakka

: J. Am. Ceram. Soc., 2009, 92, 54–61.

80.

Fitzer

and Huttner

: J. Phys. D: Appl. Phys., 1981, 14, 347–371.

81.

Naslain

: Compos. Sci. Tech., 2004, 64, 155–170.

82.

Webster

Westwood

Hayes

Day

Taylor

Duran

Aparicio

Rebstock

and Vogel

: J. Eur. Ceram. Soc., 1998, 18, 2345–2350.

83.

Aparicio

and Duran

: J. Am. Ceram. Soc., 2000, 83, 1351–1355.

84.

Aparicio

and Duran

: Ceram. Inter., 2005, 31, 631–639.

85.

Huang

Zeng

Xiong

and Huang

: J. Mater. Sci., 2004, 39, 7383–7385.

86.

Huang

Zeng

Deng

Xiong

and Li

: Ceram. Inter., 2007, 33, 887–890.

87.

Smeacetto

Ferraris

Salvo

Ellacott

Ahmed

Rawlings

and Boccaccini

: Ceram. Inter., 2008, 34, 1297–1301.

88.

Zheng

X.H.

Y.G.

Xiao

J.Y.

Zhang

W.J.

and Zhang

L.C.

Appl. Surf. Sci., 2009, 255, 4250–4254.

89.

Coetsee

Terblans

and Swart

: Phys. Stat. Solid, 2008, 2, 634–637.

90.

Ogura

Kondo

Morimoto

Notomi

and Sekigawa

: J. Jpn. Inst. Metals, 2001, 65, 13–20 (In Japanese).

91.

Huang

Zeng

and Li

: Ceram. Inter., 2006, 32, 417–421.

92.

Argirusis

Chr

Damjanovic

and Borchardt

: J. Eur. Ceram. Soc., 2007, 27, 1303–1306.

93.

Lee

Fox

and Bansal

: J. Eur. Ceram. Soc., 2005, 25, 1705–1715.

94.

Jacobson

: J. Am. Ceram. Soc., 1993, 76, 3–28.

95.

Smialek

Robinson

Opila

Fox

and Jacobson

: Adv. Compos. Mater., 1999, 8, 33–45.

96.

Lee

: Surf. Coat. Tech., 2000, 1–7, 133–134.

97.

Lee

: Trans. ASME, 2000, 122, 632–636.

98.

Lee

Fox

Eldridge

Zhu

Robinson

Bansal

and Miller

: J. Am. Ceram. Soc., 2003, 86, 1299–1306. (NASA Report: NASA/TM-2002-211372).

99.

Ueno

Jayaseelan

and Ohji

: Inter. J. Appl. Ceram. Tech., 2005, 1, 362–373.

100.

Chambers

Busch

Schulte

Gustafsson

Garfunkel

Wang

Maher

Klein

and Parsons

: Appl. Surf. Sci., 2001, 181, 78–93.

101.

Niu

Ashcraft

Kelly

Chambers

Klein

and Parsons

: J. Appl. Phys., 2002, 91, 6173–6180.

102.

Chambers

and Parsons

: J. Appl. Phys., 2001, 90, 918–933.

103.

Gordon

Berker

Hausmann

and Suh

: US patent, No. US 6969539 B2.

104.

Ronda

Kynast

Dingen

WPM

and van Hal

HAM

: J. Alloy. Comp., 1993, 192, 55–56.

105.

Bondar

: Mater. Sci. Eng. B, 2000, 69–70, 505–509.

106.

Dupree

Lewis

and Smith

: J. Am. Chem. Soc., 1988, 110, 1083–1087.

107.

Vomacka

and Babushkin

: J. Eur. Ceram. Soc., 1995, 15, 921–928.

108.

Maier

Rixecker

and Nickel

: J. Solid State Chem., 2006, 179, 1630–1635.

109.

Fukuda

and Matsubara

: J. Am. Ceram. Soc., 2004, 87, 89–92.

110.

Dinger

Rai

and Thimas

: J. Am. Ceram. Soc., 1988, 71, 236–244.

111.

Maclaren

Trusty

and Ponton

: Acta Mater., 1999, 47, 779–791.

112.

Batalieva

and Pyatenko

: Zhurnal Strukturnoi Khimii, 1968, 9, 921–922., or J. Struct. Chem., 1968, 9, 820.

113.

Nowok

Kay

and Kulas

: J. Mater. Res., 2001, 16, 2251–2255.

114.

Fukuda

and Matsubara

: J. Mater. Res., 2003, 18, 1715–1722.

115.

Batalieva

and Pyatenko

: Sov. Phys. Crystallogr., 1972, 16, 786–789.

116.

Heward

Sarrafi-Nour

and Gao

: J. Am. Ceram. Soc., 2009, 92, 511–516.

117.

Dolan

Harlan

White

Misture

Bancheri

and Bewlay

: Powder Diffr., 2008, 23, 20–25.

118.

Dias

Glasser

Gunwardane

and Howie

: Z. für Kristallogr., 1990, 191, 117–123.

119.

Maksimov

Kharitonov

Ilyukhin

and Belov

: Dokl. Akad. Nauk. SSSR, 1968, 183, 1072–1075.

120.

‘International tables for crystallography’, (2006). 1^st online edn. Chester: International Union of Crystallography. (ebook, available at http://onlinelibrary.wiley.com/book/10.1107/97809553602060000001).

121.

Ching

Ouyang

and Xu

: Phys. Rev. B, 2003, 67, 245108.

122.

Ching

: J. Am. Ceram. Soc., 2004, 87, 1996–2013.

123.

Wang

Zhou

and Lin

: Acta Mater., 2007, 55, 6019–6026.

124.

Becerro

Escudero

Florian

Massiot

and Alba

: J. Solid State Chem., 2004, 177, 2783–2789.

125.

Becerro

and Escudero

: Chem. Mater., 2005, 17, 112–117.

126.

Becerro

and Escudero

: J. Solid State Chem., 2005, 178, 1–7.

127.

Liao

Sasaki

Yoshinichi

and Sun

: J. Mater. Chem., 2011, 21, 3234–3242.

128.

Liao

Sasaki

and Sun

: Phys. Chem. Chem. Phys., 2013, 15, 17553–17559.

129.

Liu

Wang

and Zhou

: J. Am. Ceram. Soc., 2012, 95, 1093–1099.

130.

Liu

Wang

Sun

Wang

and Zhou

: J. Am. Ceram. Soc., 2013, 96, 3304–3311.

131.

MacLaren

and Richter

: Phil. Mag., 2009, 89, 169–181.

132.

Lin

Zhuo

Sun

Zhou

and Veyssiere

: Acta Mater., 2009, 57, 2851–2857.

133.

Hay

: Phil. Mag., 2005, 85, 373–386.

134.

Liao

Wang

and Zhou

: J. Phys.: Condensed Matter, 2006, 18, L527–L533.

135.

Y Wang

Zhou

and Lin

: Appl. Phys. Lett., 2005, 87, 051902.

136.

Liao

Wang

and Zhou

: Phys. Rev. B, 2006, 74, 174112.

137.

Liao

Wan

and Zhou

: J. Mater. Res., 2009, 24, 556–564.

138.

Liao

Wang

and Zhou

: J. Phys.: Condensed Matter, 2006, 18, 6183–6192.

139.

Liao

Wang

and Zhou

: Phys. Rev. B, 2006, 73, 214109.

140.

NIST material properties database on the American Ceramics Society web site: http://www.ceramics.org/publications/ceramic_properties_databases.aspx

141.

Morgan

PED

and Marshall

: J. Am. Ceram. Soc., 1995, 78, 1553–1563.

142.

Wang

Pan

Jian

Fang

Cao

and Luo

: Mater. Chem. Phys., 2003, 79, 30–36.

143.

http://www.accuratus.com/fused.html

144.

Desmaison-Brut

and Montintin

: J. Am. Ceram. Soc., 1995, 78, 716–722.

145.

Cook

Lawn

and Fairbanks

: J. Am. Ceram. Soc., 1985, 68, 604–615.

146.

Hay

and Marshall

: Acta Mater., 2003, 51, 5235–5254.

147.

Choudhry

and Islam

: J. Mater. Sci., 2005, 40, 3745–3747.

148.

Choudhry

Javed

and Bano

: Mater. Chem. Phys., 2007, 10, 251–254.

149.

O’Bryan

Gallagher

and Berkstresser

: J. Am. Ceram. Soc., 1988, 71, C42–C43.

150.

Wagner

Seifert

and Aldinger

: Proc. 10th International Conf. on ‘Advances in materials & manufacturing processes’ (ICAMMP-2002), 71; 2002, New Delhi, Tata McGraw-Hill.

151.

Cordfunke

EHP

Booij

and Van der Laan

: J. Chem. Thermodyn., 1998, 30, 199–205.

152.

Brain

: ‘Thermochemical data of pure substances’, 1995, Weinheim, Germany, Wiley-VCH Verlag GmbH.

153.

Clarke

and Phillpot

: Mater. Today, 2005, 8, 22–29.

154.

Clarke

: Surf. Coat. Tech., 2003, 163–164, 67–74.

155.

Hasselman

DPH

: Ceram. Bull., 1970, 49, 1033–1037.

156.

Aksel

and Warren

: J. Eur. Ceram. Soc., 2003, 23, 301–308.

157.

Hecht

Graves

McCullum

Berens

Goodrich

Wolf

Hoenigman

Yaney

Grant

and Hilton

: Oak Ridge National Laboratory (ORNL) Report Number: ORNL/Sub/84-00221/2, 1990, 1-111.

158.

Hampshire

: ‘Engineering properties of nitrides,’ in ‘Engineered materials handbook’, (ed. S. J. Schneider, Jr.), Vol. 4, 812–820; 1991, ASM International, Ohio, USA.

159.

Krohn

Larson

and Hasselman

DPH

: J. Am. Ceram. Soc., 1973, 56, 490–491.

160.

‘Dynamic-Ceramic technical information series,’ Crewe, UK, Dynamic-Ceramic Ltd, available at http://www.dynacer.com/data_sheets.htm

161.

Vassen

Cao

Tietz

Basu

and Stöver

: J. Am. Ceram. Soc., 2000, 83, 2023–2028.

162.

Wen

: ‘Theory of tribology’, Beijing, Tsinghua University Press; 1990 (In Chinese).

163.

Sun

and Zhou

: Wear, 2009, 266, 960–967.

164.

Sun

and Zhou

: J. Am. Ceram. Soc., 2013, 96, 3228–3238.

165.

Mechnich

‘Ceramic material’, US patent 7348287.

166.

Fox and Jacobson

: J. Am. Ceram. Soc., 1988, 71, 128–138.

167.

Hirschfeld

and Brown

: J. Mater. Sci., 1997, 32, 4455–4461.

168.

Pickrell

Sun

and Brown

Jr : Fuel Proc. Tech., 1995, 44, 213–236.

169.

Rapp

: Corros. Sci., 2002, 44, 209–221.

170.

Zhang

: J. Chinese Soc. Corros. Protect., 1992, 12, 1–10 (in Chinese).

171.

Shi

and Rapp

: J. Electrochem. Soc., 1986, 133, 849–850.

172.

Maier

Nickel

and Rixecker

: J. Eur. Ceram. Soc., 2007, 27, 2705–2713.

173.

Courcot

Rebillat

and Louchet-Pouillerie

: Mater. Sci. Forum, 2008, 595–598, 923–931.

174.

Cao

Zhou

Jian

Xiong

and Shi

: J. Appl. Phys., 2009, 105, 106102.

175.

Hou

Cao

Yuan

and Song

: Chinese Phys. B, 2010, 19, 017702.

176.

Sun

and Zhou

: unpublished work.

177.

Boakye

Mogilevsky

Hay

and Cinibulk

: J. Am. Ceram. Soc., 2011, 94, 1716–1724.

178.

Hay

Cinibulk

Boakye

Keller

Mogilevsky

Parthasarathy

and Ahrens

: ‘Processing and testing Re₂Si₂O₇ matrix composites’, Air Force Research Laboratory Report, AFRL-RX-WP-TP-2012-0376, 2012.

Recent progress on synthesis,multi-scale structure,and properties of Y–Si–O oxides

Abstract

Introduction

Preparation of Y–Si–O oxides

Synthesis of powders and sintering of bulk ceramics

Sol-gel synthesis of Y2Si2O7 and Y2SiO5 25–34

Hydrothermal method

Solid-state reaction method

Solid-liquid reaction method

Molten salt flux growth method (Czochralski technique)

High-pressure/high-temperature (high-P/high-T) synthesis

Pressureless sintering of bulk Y–Si–O ceramics

Hot pressing of bulk Y–Si–O ceramics

Colloidal processing of bulk Y–Si–O ceramics

Colloidal processing of Y2SiO5

Colloidal processing of γ-Y2Si2O7

Preparation of Y–Si–O coatings/films

Oxidation resistant coatings on SiC–C/C composites

Slurry dip coating

Plasma spray coating

Electrophoretic deposition

Environmental barrier coatings on SiC–C/C substrate

High-κ gate dielectric thin films

Sputtering of amorphous yttrium silicate thin film

CVD or ALD of yttrium silicates

Deposition of Y2SiO5:Ce phosphoric thin film

Polymorphs, crystal structure, defects and microstructure of Y–Si–O oxides

Polymorphs and phase transformations

Crystal structure and chemical bonding

Crystal structure of ζ-Y2Si2O7

Crystal structure of η-Y2Si2O7

Crystal structure of Y4·67(SiO4)3O

Defects in Y2SiO5

Microstructure characteristics

Properties of Y–Si–O oxides

Theoretical prediction of the properties of Y–Si–O oxides

Theoretical predictions of the mechanical properties of γ-Y2Si2O7

Theoretical predictions of optical and dielectric properties of γ-Y2Si2O7 and Y2SiO5

Mechanical properties of Y–Si–O oxides

Mechanical properties of γ-Y2Si2O7

Mechanical properties of X2-Y2SiO5

Thermal properties of Y–Si–O oxides

Thermal properties of Y2Si2O7

Thermal expansion of Y2Si2O7

Heat capacity of Y2Si2O7

Thermal conductivity of Y2Si2O7

Thermal shock resistance of Y2Si2O7

Thermal properties of Y2SiO5

Thermal expansion coefficient of Y2SiO5

Heat capacity, thermal diffusivity and thermal conductivity of Y2SiO5

Tribological properties of Y–Si–O oxides

Tribological properties of γ-Y2Si2O7

Tribological properties of γ-Y2Si2O7/ZrO2 composites

Environmental durability of Y–Si–O oxides

Oxygen permeability in Y2SiO5

Hot corrosion resistance of Y2Si2O7

Hot corrosion behaviour of γ-Y2Si2O7 in strong basic Na2CO3 molten salt environment

Hot corrosion behaviour of γ-Y2Si2O7 in weak acidic Na2SO4 molten salt environment

Water vapour corrosion of Y2Si2O7 and Y2SiO5

Dielectric properties of Y2Si2O7

Structure-property relationships of Y–Si–O oxides

Potential applications of Y–Si–O oxides

Conclusions

Footnotes

Acknowledgements

References

Sol-gel synthesis of Y₂Si₂O₇ and Y₂SiO₅ ^25–34

Colloidal processing of Y₂SiO₅

Colloidal processing of γ-Y₂Si₂O₇

Deposition of Y₂SiO₅:Ce phosphoric thin film

Crystal structure of ζ-Y₂Si₂O₇

Crystal structure of η-Y₂Si₂O₇

Crystal structure of Y_4·67(SiO₄)₃O

Defects in Y₂SiO₅

Theoretical predictions of the mechanical properties of γ-Y₂Si₂O₇

Theoretical predictions of optical and dielectric properties of γ-Y₂Si₂O₇ and Y₂SiO₅

Mechanical properties of γ-Y₂Si₂O₇

Mechanical properties of X2-Y₂SiO₅

Thermal properties of Y₂Si₂O₇

Thermal expansion of Y₂Si₂O₇

Heat capacity of Y₂Si₂O₇

Thermal conductivity of Y₂Si₂O₇

Thermal shock resistance of Y₂Si₂O₇

Thermal properties of Y₂SiO₅

Thermal expansion coefficient of Y₂SiO₅

Heat capacity, thermal diffusivity and thermal conductivity of Y₂SiO₅

Tribological properties of γ-Y₂Si₂O₇

Tribological properties of γ-Y₂Si₂O₇/ZrO₂ composites

Oxygen permeability in Y₂SiO₅

Hot corrosion resistance of Y₂Si₂O₇

Hot corrosion behaviour of γ-Y₂Si₂O₇ in strong basic Na₂CO₃ molten salt environment

Hot corrosion behaviour of γ-Y₂Si₂O₇ in weak acidic Na₂SO₄ molten salt environment

Water vapour corrosion of Y₂Si₂O₇ and Y₂SiO₅

Dielectric properties of Y₂Si₂O₇