|Monthly Tech-Tip |
How to create a blend of materials to chemically substitute for another (Cornwall Stone is used as an example).
Cornwall Stone is used worldwide in ceramics as a source of fluxing oxides in both bodies and glazes. As a glaze material, Cornwall stone is akin to feldspar. It is sourced from granite rock and is, therefore, a combination of a number of minerals. Actually, there are several varieties depending on the state of alteration. The blue and purple colored types (less altered) contain more fluxes, while the whiter varieties contain less K2O, Na2O and almost no fluorine. For people who have used it in glazes like I have, it is hard to know whether the multiple personalities of the material are a result of grade confusion in the distribution channel or actual variability in the product. However, for many of us, it is the high cost of the material (to have it shipped half way across the world) that makes it well worth while to duplicate, that is, replace with a chemically equivalent mix of cheaper, more consistent, and easier to obtain materials. Really, it is somewhat like making the muffins your grandmother in England used to make. You will have to substitute a local flour (or mix of flours) and possibly even make adjustments to the rest of the recipe to duplicate those muffins in your locale.
Amazingly, many people have tried to simply replace Cornwall gram-for-gram with a feldspar. But this will not work, Cornwall Stone is too different from a feldspar. For example, consider Custer Feldspar: the amount of flux is much greater and it has a much lower silica content. Nepheline syenite also has an even higher flux content, much lower SiO2 and much higher alumina.
|Nepheline Syenite Unimin||0.35||0.03||4.80||10.60||23.50||60.20||0.08|
The first real challenge is to decide what analysis to duplicate. Since the whiter varieties are what I am accustomed to and they are the most pure, and contain almost no fluorspar mineral, I feel most comfortable zeroing in on this type. The first step was to locate all available analyses from data sheets and textbooks.
|100 NUMBER 1||640.93||0.21||0.06||0.29||0.43||0.01||0.01||0.97||7.72|
|101 NUMBER 2||682.90||0.23||0.03||0.34||0.37||0.01||0.02||1.02||8.27|
|102 NUMBER 3||680.57||0.29||0.02||0.27||0.40||0.00||0.02||1.06||8.19|
|103 NUMBER 4||705.21||0.27||0.02||0.32||0.37||0.01||0.02||1.10||8.47|
|104 NUMBER 5||717.95||0.22||0.04||0.32||0.40||0.01||0.02||1.09||8.73|
|105 NUMBER 6||696.23||0.22||0.02||0.16||0.59||0.00||0.01||1.11||8.45|
|106 NUMBER 7||718.73||0.27||0.04||0.31||0.36||0.01||0.01||1.13||8.66|
|107 NUMBER 8||736.85||0.28||0.04||0.31||0.35||0.01||0.02||1.12||8.97|
|108 NUMBER 9||1118.35||0.18||-||0.56||0.22||0.01||0.03||1.85||13.68|
|109 NUMBER 10||963.42||0.15||-||0.39||0.42||0.01||0.02||1.61||11.70|
|110 NUMBER 11||685.63||0.26||0.01||0.48||0.25||0.01||-||1.13||8.10|
|111 NUMBER 12||738.13||0.18||0.02||0.35||0.43||0.01||0.01||1.15||8.82|
Although each analysis contains some fluorine (not shown), I decided to attribute the amounts to other oxides. To get an average overall analysis, you can calculate the average of each column. I did it in Digitalfire desktop INSIGHT by entering each analysis (with LOI) as a recipe, converting to a formula and then inserting each into the MDT materials database. Then I keyed amounts of 500 of each of these into a new recipe (although I could have weighted it toward certain analyses that might be judged more reliable). Below is the resulting calculation to yield an average Cornwall stone analysis.
|DETAIL PRINT - CORNWALL STONE AVERAGE|
|WEIGHT OF EACH OXIDE||56.10||40.30||94.20||62.00||160.00||79.70||102.00||60.00|
|PER CENT BY WEIGHT||1.81||.14||4.31||3.30||.20||.15||16.33||73.76|
This sounds like a lot of trouble to go through since this collection of analyses are remarkably consistent. But for many materials it is necessary, one isolated analysis is unlikely to be representative. Many material substitution efforts have been less than successful merely because the original material, toward which formulation efforts were aimed, was not clearly defined.
I had excellent results working out recipes that substitute other materials for Cornwall stone. INSIGHT's cost calculation abilities make it easy to compare unit costs and it was a surprise to see the cost difference between the original glazes and the newly developed recipes (although I have not shown it here).
Following is a side-by-side report from INSIGHT. On the left is a cone 10R fatty matte glaze recipe. I am only using this recipe for demonstration since it has a high Cornwall Stone content (but it crazes on porcelain and has virtually no clay content so it will not suspend as a slurry nor will it have good application properties). The first column in the chemistry is the formula, the second the percentage analysis, the last the molar percent.
I began by entering this as recipe 1 in INSIGHT. Next, I deleted the Cornwall and recalculated. It was immediately evident that sodium oxide, potassium oxide, Al2O3 , and SiO2 were all lost with the Cornwall stone. Since INSIGHT calculates a formula so quickly, it only takes a minute to juggle the materials to supply the needed oxides. Below right is the glaze recipe I came up with. For me it calculates to a cost less than one third of the original.
Cornwall Stone.............. 65.00 Dolomite.................... 20.71 Dolomite.................... 20.00 Silica...................... 31.47 Silica...................... 10.00 Custer Feldspar............. 19.80 Bentonite................... 3.00 Nepheline Syenite........... 16.91 ========= Whiting..................... 1.52 98.00 Kaolin...................... 9.59 ========= CaO 0.42* 8.29 9.20 100.00 MgO 0.36* 5.14 7.95 K2O 0.10* 3.20 2.12 CaO 0.42* 8.32 9.23 Na2O 0.12* 2.53 2.55 MgO 0.36* 5.16 7.96 TiO2 0.00 0.11 0.09 K2O 0.10* 3.19 2.11 Al2O3 0.36 12.72 7.77 Na2O 0.12* 2.54 2.55 SiO2 3.22 67.74 70.23 Al2O3 0.36 12.76 7.78 Fe2O3 0.00 0.26 0.10 SiO2 3.22 67.98 70.36 Fe2O3 0.00 0.05 0.02 Calculated LOI: 10.02 Imposed LOI: Calculated LOI: 11.75 Si:Al: 9.04 Imposed LOI: SiB:Al: 9.04 Si:Al: 9.05 Thermal Expansion: 6.62 SiB:Al: 9.05 Formula Weight: 285.63 Thermal Expansion: 6.60 Formula Weight: 284.71
There is a second method. Let's consider an example to demonstrate the concept. Here is my recipe for the Cornwall stone substitute.
Dolomite.................... 0.37 Custer Feldspar............. 30.43 Wollastonite................ 3.30 Kaolin...................... 12.85 Silica...................... 29.29 Nepheline Syenite........... 23.75 ========= 100.00
This is the glaze recipe into which I want to substitute Cornwall stone, I derived it the same way as the recipe above. It is a transparent that works well with chrome-tin pink stains at cone 4-6.
FERRO FRIT 3134..... 55.80
CORNWALL STONE...... 20.30
There are two ways to mix up a glaze using our substitute. First, you could weigh up a batch of the Cornwall substitute, mix it thoroughly, then treat it as a raw material. However, there is a problem with this. Powder mixing is very difficult and industry must use specially designed equipment to achieve a reasonable mix. A shaken batch in a plastic bag might look mixed, but it isn't.
A second method is to calculate a new recipe that is a combination of the two. INSIGHT can combine two recipes, or, more precisely add one to another. The two are merged so that common materials combine their amounts. In this case you could do it as follows:
We have demonstrated an effective approach to duplicating any material: Generate a final recipe for a substitute and save it. Any time the old material is needed, just perform the magic shown here, and you have a new recipe that is likely to work. Note that this recipe provided the basis for a long project I undertook (#1213) to produce one quality, documented, adjustable recipe for Orton oxidation cone 3-8. You will have different materials to source the oxides lost by removing Cornwall stone from a recipe and you may go about the substitution process a little differently, but the key to success is still the same: View things from the formula viewpoint.
These flow tests demonstrate how similar the substitute recipe (left) is to the real material (right). 20% Frit 3134 has been added to each to enable better melting at cone 5 (they do not flow even at cone 11 without the frit). Links below provide the recipe for the substitute and outline the method of how it was derived using Digitalfire Insight software. This substitute is chemically equivalent to what we feel is the best average for the chemistry of Cornwall Stone.
This is a cone 11 oxidation melt flow test. Shown (left to right) are the new shipment of Cornwall Stone 2011, the L3617 calculated equivalent (a recipe, see link), the older Cornwall shipment we have been using and the H&G substitute 2011 (far right, mislabelled on the picture). These do not flow well here, a small frit addition is needed to better compare them. However they have melted enough to see some differences in whiteness and degree of melt. Notice the L3617 is more like the old Cornwall than the new Cornwall is.