An overview of glass-ceramics
TWI Bulletin, January/February 1993
Wendy Hanson joined TWI in 1992 and is a Senior Research Engineer in the Ceramics and Precision Processes Department. Before this, she studied Metallurgy and Engineering Materials at the University of Strathclyde, and performed research into the fabrication and properties of SiC ceramic matrix composites. Her TWI interests involve joining ceramics to metals by a novel technique based upon mechanically flexible interlayers, polymer derived ceramics and glass-ceramics.
John Fernie joined TWI in 1990. As leader of the Ceramics Section, he has responsibility for a number of projects connected with all aspects of ceramic joining technology, particularly brazing and diffusion bonding. His current research interests include microwave heating, glass-ceramics and polymer derived ceramics. Before working at TWI, John studied Metallurgy and Microstructural Engineering at Sheffield City Polytechnic, and went on to research oxynitride glasses and glass-ceramics at The Centre for Advanced Materials Technology at the University of Warwick.
Glass-ceramics are a family of materials produced by controlled crystallisation of a parent glass. They combine the ease of fabrication of a glass with the generic mechanical properties of a ceramic. They are truly engineered materials, capable of a wide range of microstructures and properties. Wendy Hanson and John Fernie examine their manufacture, their properties and some of their many applications.
Glass-ceramics are polycrystalline solids prepared by controlled crystallisation, or devitrification, of a parent glass. They are usually, but not necessarily, based on the oxides of silicon and/or boron.
There are two stages in production of a glass-ceramic. Initially, the constituents of the glass are melted and formed to shape. This is followed by a nucleation and crystallisation heat treatment. The product of crystallisation is a glass-ceramic which typically consists of fine grains ( <10µm) and has approximately 2% residual glass sited at the grain boundaries. The microstructure of a glass-ceramic can be tailored to meet a range of properties - most notably the coefficient of thermal expansion (CTE). They can also be designed to provide excellent thermal shock resistance, dielectric properties, chemical durability, high strength and toughness.
Numerous glass-ceramics have been developed for diverse applications such as substrates for electronic devices, radomes, mirror blanks and cooker tops (ceramic hobs). They are also used to make hermetic seals in many glass-ceramic/metal systems. Among the most common of the glass-ceramic systems are lithium-alumino-silicate (LAS), zinc-alumino-silicate (ZAS) and magnesium-alumino-silicate (MAS). These systems have been studied in detail and many formulations are commercially available.
Glasses
A glass can be defined as an inorganic product of fusion which has cooled to a rigid condition without crystallising.
Glass constituents
Most inorganic glass forming systems are based on a mixture of oxides. Three categories of oxide have been distinguished, each contributing differently to the glass structure:
- Network forming: These are oxides which are able to form glasses because of their ability to build up random, three-dimensional networks. The oxides of Si, B, P, As, Ge are in this category.
- Network modifiers: These oxides are incapable of building up a continuous network and effectively weaken a glass structure by introducing breaks in an otherwise continuous chain. Examples of modifying oxides are NaO, Li 2O, K 20, MgO, CaO, BaO.
- Intermediates: These are usually unable to form a glass themselves, but can be incorporated to some extent into an already existing network. This type of oxide includes Al 2O TiO 2, ZrO 2.
Fig. 1. A two-dimensional diagram of an ordered (crystalline) and disordered (glassy) SiO 2 network and a soda lime glass. The fourth bonds of the Si project upward or downward from the plane of the paper
Figure 1 compares the structure of crystalline SiO 2 ( e.g. quartz) with that of a 'glassy' material of similar composition. It also shows the effect of additions of other elements to the glass and in particular how modifying elements, in this example Na and Ca, weaken the structure. The breaks in the network caused by these oxides render the composition more susceptible to crystallisation.
Selection criteria
A parent glass must satisfy two important and conflicting requirements for a glass-ceramic to be used successfully:
- Melting and working: the parent glass must be capable of being melted and shaped by economic means, e.g. casting, drawing, injection moulding. Ease of processing requires low viscosity at working temperatures with only a small change in this viscosity with temperature. This requires a high proportion of modifiers and intermediates.
- Controlled crystallisation: this is the ability of a glass to crystallise in a controlled manner. Glasses which contain a high percentage of modifying oxides crystallise easily because these oxides weaken the glass network, but this gives poor crystallisation control. Therefore a lower proportion of modifiers should be present.
The conflict between these demands must be resolved by the ceramist for each individual case and application. In the presence of certain oxides, such as MgO, Al 20 3, CaO and ZnO, the chemical stability of the glass will be increased and this may also need to be considered.
The glass-ceramic process
The three key variables in preparing a glass ceramic are:
- glass composition;
- glass crystallisation - the glass-ceramic process;
- nature of the crystalline microstructure.
The parent glass should exhibit good workability and resistance to uncontrolled crystallisation during forming.
The glass-ceramic process converts the parent glass into a ceramic via a suitable heat treatment. The microstructure, i.e. types and proportion of crystalline phases, can often be predicted from the equilibrium phase diagram and is responsible for many physical and chemical properties including thermal expansion and hardness. However, many systems contain phases which are metastable and these are more difficult to predict accurately.
The nature or morphology of the crystalline microstructure is the key to many mechanical and optical properties including transparency/opacity, strength, fracture toughness and machinability.
Thermal treatment
The steps used in processing a glass-ceramic body are illustrated schematically by the temperature-time cycle shown in Fig.2. The material is melted and formed at elevated temperatures and then cooled to ambient (at which additional processing steps, or inspection may be carried out). The sample is then heated, at a rate limited by the need to avoid thermal shock, to a holding temperature at which nucleation of the major phases is effected. The sample is held at this nucleation temperature, typically for 1-2 hours. After nucleation is complete, the material is heated further to effect the growth of the major crystalline phases.
Fig. 2. The glass-ceramic process
This temperature and the subsequent holding time, which may be very brief, depend on the system and composition, as well as on the phases and properties desired in the final body.
A typical glass-ceramic has a very fine grained microstructure, with typical crystal sizes ranging from 0.1-l0µm, depending on the heat treatment and the presence of nucleating agents. The grains themselves usually have a random orientation, with small volumes of residual glass phase at the grain boundaries.
Nucleating agents
The presence of these agents in the parent glass is essential to promote the development of a high density of nucleation sites. It is necessary to have a large number of sites to obtain a fine-grained microstructure. This in turn is a requirement for good mechanical properties. Nucleation agents can also be used to promote certain phases and morphology within the structure.
Nucleating agents are added to the material during the initial glass formation. They are used in one of two forms - metals or oxides.
Metals - these form a uniform dispersion in the glass. Typical examples are Cu, Pt, Pd, Os. Glass-ceramics nucleated under these conditions are characterised by good mechanical and electrical properties.
Oxides - oxide nucleating agents induce a phase separation of colloidal particles of one phase, dispersed in the matrix of a second phase. Examples of this type of agent are TiO 2, ZrO 2, P 2O 5 and Cr 2O 3.
Glass-ceramic compositions
The most widely investigated glass-ceramic systems are those based upon Li
2O-Al
2O
3- SiO
2 (LAS) and MgO-Al
2O
3-SiO
2 (MAS).
LAS systems are notable for their very large range of thermal expansion coefficients. The type of nucleating agent used is critical in determining the final crystal sizes. For example, TiO 2 alone gives a grain size of 1-2µm, however, a mixture of TiO 2 and ZrO 2 gives highly efficient nucleation of β -Quartz and a correspondingly fine grain size of <100nm.
MAS systems are of increasing interest as monolithic materials for use as electronic substrates (where low thermal expansion and dielectric loss are particularly desirable). The low coefficient of thermal expansion also makes MAS attractive as matrices for ceramic matrix composites.
However, work is not restricted purely to the oxide svstems. A current trend is to investigate effects of additions of carbon or nitrogen and hence formation of oxycarbide or oxynitride glass-ceramics. Oxynitrides are of particular interest, since they play an important role as sintering aids in high technology engineering ceramics such as silicon nitrides and sialons.
Properties
The principal advantages of glass-ceramic materials are associated with the ease, economy and precision of the forming operations and the ability to tailor the microstructure to gain a particular suite of properties.
Achievement of desired properties, or a combination of properties, by a systematic variation of the chemistry, or by heat treatment and hence microstructure is probably the most essential feature of glass-ceramics. A systematic approach is made possible by the wide ranging variation in phase assemblages which can be obtained.
Probably the most versatile feature of glass-ceramics is their wide range of thermal expansions. This is shown in Fig.3, where the coefficients of thermal expansion (CTE) of four glass-ceramic systems are given. Taking the LAS system as an example, it is observed that there are two CTE ranges. At the low (near zero) end, the glass-ceramic CTE can be matched to that of tungsten. However, the real versatility of this particular system is then exemplified. When the composition is altered by decreasing the percentage of alumina, the CTE is raised and can be matched to that of nickel alloys and stainless steel. Similar versatility is also shown by other glass-ceramic systems.
Fig. 3. The range of thermal expansion coefficients available from a series of metal alumino-silicates and comparison with a selection of industrial metals and alloys.
Figure 4 shows an example of a lithium-alumino-silicate. This is a substrate of Fe-Cr coated with LAS glass-ceramic.
Fig. 4. An Fe-Cr alloy coated with a matched expansion LAS glass-ceramic
Some glass-ceramic properties are given in the Table, which compares the four major systems: LAS, MAS, ZAS and CAS alongside an oxynitride Y-SiAlON.
Selected properties of glass-ceramics.
| System | Density,
g/cm 3 | CTE,
x 10 6-/°C | Bending strength,
MPa | Softening temperature,
°C |
| LAS 1 | 2.42-2.57 | 0-2 | 60-180 | 1000-1200 |
| MAS 2 | 2.49-2.68 | 2-6 | 150-300 | 1200-1400 |
| ZAS 3 | 2.99-3.13 | 3-20 | 60-130 | 700-1200 |
| CAS 4 | 2.46-2.80 | 6-9 | 70-130 | 900-1200 |
| Y-SiAlON | 3.82-4.00 | 3.5-7 | 170-190 | 800-925 |
1Li 2O-Al 2O 3-SiO 2 (high Al 2O 3)
2MgO-Al 2O 3-SiO 2 | 3ZnO-Al 2O 3-SiO 2
4CaO-Al 2O 3-SiO 2 |
Properties such as tensile strength are unreliable, but in general, the glass-ceramic is approximately three times as strong as the parent glass,
i.e. 50-200 MN/m
2 (for glass).
As with all glass containing ceramic materials, the final service temperatures are dictated by the amount and composition of the residual glass phase.
Applications
As mentioned above, the properties of glass-ceramics lead them to be used in several industrial sectors including aerospace, nuclear, medical and electronics. They are also used as bonding media and for joining ceramics and domestically as ceramic hobs.
Aerospace
The aerospace industry is investigating use of glass-ceramics as matrix materials for ceramic composites, where their high temperature capability, coupled with ease of fabrication and controlled microstructure, make them potentially ideal matrices. This is also a function of the ability of glass-ceramics to have their CTE matched to that of, say, the reinforcing SiC fibres of ceramic matrix composites.
Nuclear
In the nuclear industry, control rods, seals and radioactive waste disposal are all areas in which glass-ceramic materials are used. This is because of the ability to incorporate suitable neutron-absorbing elements into the glass-ceramic and their stability for use as a solid means of waste disposal.
Medical
In the medical feld, glass-ceramics have been used in such demanding environments as artificial teeth and hip joints where a high resistance to wear and compatibility with the surrounding tissue are prerequisites.
Electronics
Glass-ceramics have a number of functions in the electronics industry. These include microelectronic substrates, capacitors and glass-ceramic/metal seals. Figure 5 shows a glass-ceramic/metal seal used for electronic packaging. The reasons for selection are their dimensional stability at high temperatures; their ability to form strong glass-ceramic/metal bonds; the facility to be produced as thin-films; good thermal conductivity; CTE match to substrate materials (in particular silicon) and low dielectric constant (and losses).
Fig. 5. A glass-ceramic/metal seal
Bonding media
The low residual glass content of glass-ceramics allows them to withstand extremes of temperature, making them ideal as sealing and bonding media in high temperature and/or corrosive environments. Also, the nature of the system and its ability to have a tailored microstructure lend it to applications where the CTE between two substrates needs to be accommodated through the joint or interlayer.
Ceramic hobs
Ceramic hobs have proved extremely popular in recent years because of an unusual combination of optical and thermal properties.
Only materials where the CTE is <0.5 X 10 -6/°C are capable of resisting large temperature differences without breaking due to thermal stresses. This is an essential requirement for cooker tops, as is the necessity for a high thermal shock resistance.
The glass-ceramic itself is transparent at infrared wavelengths and therefore appears dark and opaque in normal viewing conditions. However, transmission of infrared allows the heating elements positioned below the hob to become visible. These optical properties are a function of the composition and microstructure of the material.
One overriding requirement of glass-ceramic hobs is the thermal conductivity of the material. It should be high enough to allow a sufficiently rapid heat flow from the heat source to the food; however, it is also necessary that this energy remains confined to the cooking zone to minimise the danger of being burned.
Another benefit is the ease of cleaning because of the non-porous, smooth surface of the panel. The size of the crystals is very small (<50nm) and therefore the glassceramic is practically as smooth as glass itself.
In conclusion
Glass-ceramics have many attractive properties, based on the ease with which their microstructures can be tailored. This combines with an ease of processing which is not available to other ceramics. Although they have a number of potential applications, to date their most widespread uses are as hermetic seals and ceramic hobs; however, continued research will ultimately lead to many further applications in the electronic and aerospace industries.
