search  current discussion  categories  materials - misc 

analysing chemicals

updated tue 22 jun 99

 

Paul Lewing on sat 19 jun 99

During glazing for my last firing, I was trying to adjust some glazes
from an old analysis of Gerstley Borate to a newer one (thanks, Tom
Buck, although it still didn't work very well). But I was thinking
about these chemical analyses that we rely on so much, and realized I
had a serious gap in my understanding of the process.

How do they do an analysis of a mineral? What is the tool or machine
that is used and how does it work?

And, in the unity part of a formula there are parts of a molecule listed
for each oxide, e.g. .6 NaO, .3 K2O, .1 CaO, adding up to one. How do
they determine how much of each there is? We know that this must be an
average composition across a large (how large?) sample, since each
individual molecule must have a complete CaO or K2O, not a part of a
molecule.

A molecular analysis is actually counting molecules, not weighing them.
Do they do a weight analysis first and then convert it using atomic
weights to a unity formula, or is it the other way around? Do they take
many samples from different parts of the mine and average them, or do
they just do one? Is it different for more consistent materials than it
is for inconsistent ones?

And how accurate are they? Can they really tell what's there out to
four decimal places, as we sometimes see, or is that just the
mathematical result of averaging?

I'm convinced that someone on this list knows each fact there is to be
known, so I'm sure we'll get a complete answer here, the more detailed
the better.

Thanks,
Paul Lewing, Seattle

Tom Buck on sun 20 jun 99

Paul Lewing wants to know how good is the analysis of Gerstley Borate. The
answer: Not very good. But it is better than zilch.
My understanding is this (which may have changed): US Borax mines
once every two-three years the small Gerstley Borate deposit (alongside
the huge borax deposit). The GB, as three-inch pebble, is stockpiled in a
spot in the mine complex where the two main dealers each come with a truck
every few weeks to pick up a load. The truck proceeds to a mill where the
GB is crushed to powder, "air-classified" (ie, "screened automatically")
and the desired particle range is then bagged, and shipped. Sometime
during this preparation some samples for analysis are removed from the
stream going to the bagging machine.
The analysis is on a weight per cent basis for specified
components, viz., CaO, Na2O, B2O3, Al2O3, SiO2, perhaps MgO, and sometimes
the moisture and bound water are also determined. Sometimes a test is done
to determine the Loss on Ignition, and that data reported; otherwise one
gets a value by difference (not very reliable). There are several ways of
doing this analysis automatically; I do not know the specific one used by
Laguna or by Hammond & Gillespie. Not all components (eg, shale) are
determined so the analysis never adds up to 100.
How good are the numbers? I personally consider two significant
numbers only as valid, but usually carry a third to allow for appropriate
rounding.
Does the analysis track with time, ie, stay even for a year or so.
Not likely. You should always run line-tests when you get a new bag of GB.
For Raku potters, the variation in GB's analysis really doesn't bother
them since an 80GB/20spar mix won't change enough to fail to respond
to post-firing reduction, and all firings are by eyeball, not a cone.
So, if you have a cirtical glaze with GB in fair quantity, and the
mix needs just this amount of B2O3, you are advised to use a borate frit
instead of Gerstley Borate.
Good tiles, Paul. Til later. Peace. Tom.

Tom Buck ) tel: 905-389-2339 (westend Lake Ontario,
province of Ontario, Canada). mailing address: 373 East 43rd Street,
Hamilton ON L8T 3E1 Canada

Kathy Stecker on sun 20 jun 99

You do realize that you have a whole lot of questions, don't you? My best
guess after working in an analytical lab 15 years ago follows. X-ray
Diffraction can give some wonderful results on crystal structures. Different
crystalline minerals give different spike patterns, as the x-rays bounce off
the crystal faces. It used to be quite a puzzle figuring samples out as they
had many compounds in them . I used to love doing them the hard way looking
through all the books of tables. Now of course it's all computerized.
Relative percentages could be found based on the results of prepared
standards. X-Ray Fluorescence can be used to determine elemental analysis.
Atomic absorbtion can also, but the sample needs to be dissolved . Now with
the speed everything was advancing-scientific analysis may have other tricks
up it's sleeve at this point.

I would hope that anyone doing analysis at the mines would be doing combined
sampling and individual samples to seek some quality control . I'm sure
someone else will fill us in. There are some mighty smart chemistry people
lurking around here. Can't help but be amused , as I struggle to put my
chemistry degree back to thinking mode after all these years. I keep thinking
some people's minds must be spinning like the potter's wheel with all the
science on here.

So who else has some up to date accurate info on these questions?

Kathy Stecker who just saw a Loggerhead turtle laying her eggs on the beach
at Canaveral National Seashore last night. Scratching at all the no-see-um
bites today- but it was so cool.

David Hendley on mon 21 jun 99

Paul, to answer part of your question,
here's how it works with my supplier, Trinity Ceramic Supply:
They have on hand a 'typical analysis' from the mine, which is,
I suppose, kind of an average over the last few years.
Then, with each new car-load shipment of material, they get an
'actual analysis', that was calculated on the day the material in
that shipment was bagged.

My experience has been that the two are so similar as to be
insignificantly different, but I would use the 'actual analysis'
for that shipment if I had to choose the most accurate.
(These comments are not directed to Gerstley Borate. I think Tom
is correct in that they only mine it once every few years).

David Hendley
Maydelle, Texas
hendley@tyler.net
http://www.farmpots.com




At 06:57 PM 6/19/99 EDT, you wrote:
>----------------------------Original message----------------------------
>During glazing for my last firing, I was trying to adjust some glazes
>from an old analysis of Gerstley Borate to a newer one (thanks, Tom
>Buck, although it still didn't work very well). But I was thinking
>about these chemical analyses that we rely on so much, and realized I
>had a serious gap in my understanding of the process.
>
>How do they do an analysis of a mineral? What is the tool or machine
>that is used and how does it work?
>
>And, in the unity part of a formula there are parts of a molecule listed
>for each oxide, e.g. .6 NaO, .3 K2O, .1 CaO, adding up to one. How do
>they determine how much of each there is? We know that this must be an
>average composition across a large (how large?) sample, since each
>individual molecule must have a complete CaO or K2O, not a part of a
>molecule.
>
>A molecular analysis is actually counting molecules, not weighing them.
>Do they do a weight analysis first and then convert it using atomic
>weights to a unity formula, or is it the other way around? Do they take
>many samples from different parts of the mine and average them, or do
>they just do one? Is it different for more consistent materials than it
>is for inconsistent ones?
>
>And how accurate are they? Can they really tell what's there out to
>four decimal places, as we sometimes see, or is that just the
>mathematical result of averaging?
>
>I'm convinced that someone on this list knows each fact there is to be
>known, so I'm sure we'll get a complete answer here, the more detailed
>the better.
>
>Thanks,
>Paul Lewing, Seattle
>