Neon-Cat on sat 3 apr 10
use shivering
Lithium glazes form products that have ultra-low or negative thermal
expansion coefficients. This is why glazes based on the
Li2O-Al2O3-SiO2 (LAS, from lithium aluminosilicate) system and
variations are prone to shivering problems. The answer has been around
longer than I've been alive and has been widely researched and written
about, especially in the field of glass-ceramics. One need only know
one's ceramic materials to understand what's happening.
Clayart experts write that lithium carbonate is especially prone to
causing shivering. They cite a high coefficient of diffusion for
lithium or wrongly proclaim that lithium carbonate is very soluble and
lay the blame there. In fact, shivering can occur in glaze systems
that use the insoluble lithium fluxes spodumene, petalite, and
lepidolite. Diffusion of lithium into the clay body is not the answer
to why lithium-based glazes sometimes cause shivering. Diffusion,
ionic or molecular, stops when the water in the glaze stops vessel
egress or evaporates and at any rate diffusion is random - a molecule
of lithium may just as speedily diffuse back on to the surface of a
pot or head in a direction perpendicular to the pore in which it has
found itself as it is to head off down the bore of a pore. It may be
carried to the surface during water evaporation. Pore size and the
nature of pore connections (open or closed porosity) will limit
diffusion. If a clay body is 10% porous after bisque, a glaze applied
to this surface will not flood the clay body no matter what glaze
ingredients are present. Besides, lithium carbonate, one flux from Lou
Nil's glaze "Crystalou Moss" recently under discussion ("lithium
playmates, Craig, Ivor and Ron", "For Craig and Ron", etc.) is one of
those exceptions to the rule - it is an inversely soluble compound.
Everyone seems to be familiar with the rule of thumb that a salt with
normal solubility becomes increasingly soluble when the temperature is
raised and this statement holds true for about 95% of the solutes we
might encounter. Then there are those salts with inverse solubility
that show the reverse trend and lithium carbonate is at the top of the
list. At room temperature very little lithium carbonate will dissolve
to form Li+ ions in the glaze slurry (1.31% at 20 C or 68 F). Lithium,
a reactive ion, will not be found free for very long - if nothing else
it will adsorb onto a clay particle much as sodium or hydrogen ions do
where it may help flocculate the glaze and so its diffusion into a
clay body is limited by the size of the clay particle it finds itself
tied to or the size of another compound it has formed. As a lithium
carbonate molecule, it competes with other molecules and particles for
a place within pores or on the pot's surface. Lithium diffusion into a
bisqued body or one being fired, if it occurs, will be minimal.
In a typical LAS glaze some very interesting lithium aluminosilicates
(=3DE2-eucryptite, =3DE2-spodumene, etc) are produced from quartz and its
polymorphs (beta-quartz, cristobalite, tridymite) through replacement
of some silicon ions (Si4+) by aluminum ions (Al3+) with charge being
balanced by lithium ions (Li+). It does not matter how we introduce
the lithium ions - our source can be lithium carbonate, spodumene,
lepidolite, petalite, or some other lithium compound. Beta-spodumene
is a more open, high temperature form of the familiar
(alpha)-spodumene. After quartz inversion beta-quartz solid solution
may transform directly to beta-spodumene when lithium ions are
available. This is typical in systems with structures for compositions
of LiAlSixO2X+2 where tetrahedral molecules AlO4 link to tetrahedral
molecules SiO4 and Li+ ions tuck away in SiO2 frameworks to satisfy
charge. Along with beta-spodumene (Li2O.Al2O3.4SiO2 or LiAlSi2O6),
beta-eucryptite (Li2O.Al2O3.2SiO2 or LiAlSiO4), petalite (LiAlSiO10),
and other lithium aluminum silicates may form or crystallize in a
LAS-type glaze. Notice the lithia to alumina ratio is constant for
beta-spodumene, beta-eucryptite, and petalite while the silica
contents varies in each of them. Beta-eucryptite has a negative
thermal expansion over a wide range from room temperature on up.
Beta-spodumene and other LAS compounds have ultra-low thermal
expansion coefficients (CTEs). Imagine that! No wonder potters see
their glazes shiver right off their pots sometimes. You'd never guess
this using the linear coefficient method so often promoted on-list.
Another flash -- beta-spodumene is anisotropic. You never read about
anisotropy on-list. In glaze chemistry it's important to understand
that many minerals and other substances can expand in not just one
direction but along other planes, too. Sometimes their anisotropic
behavior yields an increase in volume with a decrease in linear
expansion. Talk about a mismatch in clay body and glaze fit! Alpha
quartz, for example, shows highly anisotropic behavior and positive
volume expansion at all temperatures, while beta-quartz has near-zero
or even negative expansion coefficients. LAS glaze systems are an
exception to the general rule that gases, liquids or solids expand
when heated - some substances actually do contract when heated.
The development of different crystalline phase assemblages and their
corresponding thermal expansion coefficients in LAS-type glazes
depends on the initial glaze composition, the viscosity of the raw
glaze itself during application, the thickness of the glaze coat, the
firing schedule, hold times and temperatures, maximum firing
temperature, and the type and amount of nucleating agents, if present.
Other factors influence thermal expansion coefficients. For example, a
change in the distribution of the interstitial cations (Li+) may
critically influence the observed thermal expansion. Nothing is as it
seems with these systems; there's no norm where the coefficient of
linear expansion is estimated by an additive method involving the
relative proportions of all oxides in the system. Solid solutions of
the two compounds beta-spodumene and beta-eucryptite show negative
thermal expansion. You'd never guess this using the method advocated
on clayart but it's true (and now part of clayart yore).
While firing our glazes if we're using a mixed flux system like Lou's
glaze with its nepheline syenite, when molten, sodium and potassium
ions may substitute in lithium sites in the beta-spodumene structure
changing the picture (and CTE) again. Glazes are not definite chemical
compounds but complex mixtures that are sometimes described as
undercooled solutions, a mix of amorphous, semi-crystalline, and
crystalline phases. After firing Lou's glaze consists basically of two
portions - the portion that crystallized or devitrified to yield the
necessary low expansion phases, and the continuous vitreous matrix
that contains the low expansion phases. Although the CTE of the glaze
and the clay body need to be close, they are not usually identical.
The increase of Na2O and K2O in the glaze from the nepheline syenite
will increase the coefficient of thermal expansion of the glaze. With
regard to thermal expansion each oxide present in a glaze exerts its
own influence on the glaze, sometimes with exciting twists with regard
to products created during firing.
Marian (Neon-Cat)
www.neon-cat.com
| |
|
