search  current discussion  categories  glazes - chemistry 

eutectics (long)

updated tue 14 jul 98 on mon 13 jul 98

Ron provides a great lead-in to something I've been thinking about
putting together, and since my last few posts have been kind of
negative, I hope I can provide some positive input to this list.

This is aimed more at those of you interested in more than the
most useful part of the information for potters that a eutectic
mixture is the one with the lowest melting temperature. That's
all well and good but there are some neat other properties involved.
I'm going to have to assume you can visualize the classic eutectic
between at least two "endmember" phases as a graph of composition
on the horizontal axis and temperature where crystalization (or
solidification) begins on the vertical axis. The line curves
down from each side to a eutectic at some composition between.
Got it? Ok.

To clip a bit from Ron's message (following his calculation of a
eutectic mixture):

>All that CaO is going to lead to some of it coming out of the melt on
>cooling I suppose so fast cooling would be best.

I learned about eutectics from the perspective of a geochemist --
in other words with time on our side and an interest in what
crystalizes out rather than in quenching the melt to a pure glass.
A magma cooling in the earth's crust can have oodles of time to form
crystals and the way that happens is really interesting. I think
it has some relevance when you are talking about glazes since crystals
can and do form even if you don't have millions of years to soak.

I'll hit the punchline right away: at a eutectic point each of the
stable phases will form at a rate that keeps the melt composition
constant! You won't crystallize your calcium phase (I would assume
a calcium silicate like wollastonite) without crytallizing the other
phases (like silica, I suppose). If you think about it, it makes
intuitive sense. Say you are at the minimum melting point composition
and then cool it down until one phase starts to crystallize out.
That would move the composition of the melt away from the eutectic
-- it then would have a *higher* melting point and would have to
solidify instantly. Can't happen. So all the crystal phases form
together keeping the composition and the temperature(!) constant
until either the whole melt is turned to crystals or the system
is crash cooled forming glass instead of crystals.

So what happens if you slow-cool a melt that isn't at the eutectic?
Think of a composition half-way between one of the pure endmembers
and the eutectic. You have melted the stuff so now your point on
the graph is at a temperature up above the curve. As the melt
cools the temperature first drops without changing the compostion of
the melt. If you drop the temperature really fast -- boom you get
a glass of the same composition with no crystals. But if you drop
the temperature more slowly when you get to the curve (the liquidus
for those of you who like fancy words) crystals start to form. But
you only form crystals of the stuff on that side of the graph.
If you are on the calcium rich side of the graph you will form those
wollastonite (or whatever the phase really is) crystals. In this case
however, taking out calcium makes the melt richer in silica, but
that's ok because it moves the melt towards a lower melting
composition. Going that direction keeps the melt liquid and life
is good. So the temperature drops a little more and you remove
some more calcium silicate and so on. What happens is that the
composition and temperature slides down that liquidus curve
towards the eutectic. If you were to quench the melt at any
time until you reach the eutectic you would
only have one kind of crystal (assuming only two phases here
because it is somewhat more complicated for three or more). If you
don't quench the melt then the composition will eventually shift
until it reaches the eutectic where both phases will crystallize
until there is no melt left.

If you have three phases like you would show on a ternary diagram,
first one phase crystallizes out shifting the composition until
a second phase becomes stable. Then both of those crystallize
and you slide down the boundary between the two until you
reach the eutectic between all three. Then all three crystallize
keeping the melt composition constant until it all solidifies.

Ron then says:

>It would seem to me that being near a eutectic would be the best way to get
>a clear glaze - I am also concerned that this would result in a glaze with
>a very short firing range - can anyone confirm those two statements?

I guess the first statement is more or less correct. It seems to me the best
way to get a clear glaze is to cool it quickly enough that no crystals form.
I suppose that if you have composition where you are on the side of the
eutectic of a phase where the crystals form easily then the further from
the eutectic point the higher the temperature where the crystals first
start. Then you have to cool a lot farther before the crystalization is
quenched. If you start near the eutectic
then the initial crystalization is at a lower temperature and you don't
have as far to go before the melt gums up into a glass.

I don't know all the factors that go into the firing range. I think it has
a lot to do with viscosity which probably isn't strictly related to how close
you are to the eutectic. But being near a eutectic will probably make the
glaze fickle.

Firstly, any tiny variation in composition will change the
melting temperature a lot. Remember that those curves get steeper towards
the eutectic. Also tiny composition changes would send you off to places
where different phases would crystallize first so you may get unexpected

Well, I'm tired and my dog needs her allergy shots before bed. Hope this
was somewhat understandable. Good night.

-- Evan in W. Richland WA -- just back from Seattle where I managed to
restrain myself to buying only 4 pots.