search  current discussion  categories  safety - toxicity 

ph and leaching (long, again)

updated mon 31 mar 97

 

Gavin Stairs on sun 30 mar 97

At 08:57 AM 29/03/97 EST, Max Richens wrote:
....
>Acid solutions (ignore HF for the moment) attack the fluxes in the glass
>dissolving them and opening up the silica skeleton so you get a dull
>finish where attack takes place. (used for producing pure Vitreous
>silica sheaths for heavy duty heaters)...

This shows that fluxes in a glaze retain, to some extent, their individual
character, and do not remain simply enmeshed in a silica complex. Whether
this happens as a result of the dynamic evolution of the structure while
under attack by solvents, or is fundamental to the bonding structure, I do
not know.

Conjecturally, the flux molecules probably bond loosely to the silica,
distorting and breaking bonds in the silica matrix, but not completely
dissociating it. Fluxes are generally highly ionic, resulting in a liquid
flux phase full of polarized molecules with a strongly ionic cation. This
is pretty much what water solution bases are. So the dissolution of silica
in a strong base is no surprise. On the other hand, the fluxes are all
basic, so the dissolution of the flux in acid is likewise expected. I guess
the miracle is that the whole thing doesn't just collapse in a soggy mess at
the first hint of water, which has both characteristics in equal measure. I
think the basis of this miracle must be that the fluxes are absorbed into
the silica matrix, and incorporated to a greater or lesser degree into its
structure. This greatly stresses the silica matrix, lowering its melting
point, which is the fluxing action.

In order for chemical attack to proceed, this intimate association must be
disassembled. At the atomic level, there is no such thing as stability.
What God or man has put together WILL come asunder. The only question is,
will it go back together the way it was, or will it reform in a new
configuration. The higher the temperature, the more likely the latter.
These rates of action depend on absolute temperature. Room temperature is
about 300K, while glazes melt between, say, 1000K and 1500K. By a linear
approximation, at room temperature the glazes are 1/3 to 1/5 of the way
toward melting. This is enough for appreciable activity. At water boiling,
about 373K, there is just a bit extra energy around to make things happen.
Some activity, but not rapid.

Also the more self similar the material, and the more tightly constrained
the box in which the atoms or ions are contained (atomically speaking), the
less activity is likely. This is the explanation for the durability of the
silica structure, along with the quasi covalent (non-ionic) nature of the
silica bonds. My guess is that the silicate web essentially unravels by its
own thermal activity, and the chemical agents are simply around to pick up
the pieces. Sort of like the garbage removal. They contribute just enough
energy to tip the balance between going back to sleep (the original, "solid"
configuration) and getting up (the new, "dissolved" configuration). But
once freed from the silicate trap, the strongly ionic fluxes would be little
inclined to slip back into the silicate, where there are essentially foreign.

I hope this mish-mosh of psuedoscience and animism isn't too offputing. I'm
really thinking out loud, trying to form a thought-picture of what actually
goes on in a glaze under attack.

In the end, it seems to me that what happens in the acid case is that the
attack begins at the surface, where ionic species are easily dislodged.
Then, where this removal places other ionic species near the new surface,
that proceeds as well. By this process, if the ionic species are
topologically connected, which may be preferred by reason of their ionic
nature, there will be formed ionic sized microchannels into a somewhat
relaxed and resistant silica framework. By means of these, attack can
proceed, by mutual diffusion of solvent and solute up and down the channels.
This attack would slow as time went on, by reason of the length of time
required to traverse the channels.

However, since water also has the character of a base, it can chip away at
the silica structure where there is a weakness, as for example in the
microchannels, or beside ionic inclusions. This would have the effect of
widening the channels, and thereby increasing the rate of deep attack. I
would guess that both acid and pure water attack proceed by this sort of
mechanism.

In the case of attack by a strong base, the silica itself is attacked and
the whole structure is dissolved from the surface. Attack should proceed at
a more or less uniform rate. It should be facilitated by the ionic
inclusion, in the same way that the melting point is diminished.

If this stuff makes sense, it actually predicts that high temperature glazes
(all else being equal, especially fit and homogeneity) should be more
resistant to acid/base/water attack than low temperature glazes. Not
exactly earth shattering in the way of predictions, since it is more or less
intuitive, one might say, but there you are. It also predicts that there is
no way that you can formulate a slica based glaze to make it impervious to
attack, since the attack includes the silica. However, if the silica were
replaced with a less soluable species (alumina? zirconia?), then the deep
attack by acids and water might be significantly inhibited. That, I
suppose, is what makes zirconia encapsulated glazes work. They must still
be suceptible to surface attack, but that may, indeed, be negligible.

Sorry to go on like this. I hope this is useful to somebody. It certainly
is to me. Helps me to get my own mental image in tune.

Gavin

Gavin Stairs Hi Dannon
http://isis.physics.utoronto.ca/

Gavin Stairs Hi Dannon
http://isis.physics.utoronto.ca/