Jump to content

Does Magnesium Make Clay Gray? Talc Testing


curt

Recommended Posts

This topic emerged out of a very interesting discussion of clay body fluxes, alumina, and silica in another recent thread called "Flux Formula Limits for Porcelain."  That thread morphed into a discussion on magnesium and other contaminants in porcelain clay bodies based on some photos posted by Glaze Nerd.

 

I am continuing that discussion here, in the more general context of stoneware, and adding some information on oxidation vs reduction at the same time. 

 

Funny how just when you think your old tests have no more to give, they get dragged out again for another review on some new topic.  I was dreading trawling through my humble-jumble library of tests to look into the Mg question, but when I finally got started on it it was kind of fun, a bit of a trip down memory lane, reminding me of how far I had come.  And I rediscovered some old tests I had done that, while not designed to test the Mg question, addressed it nonetheless.

 

I have been working a lot over the last two years on developing stoneware bodies using local materials, for both oxidation and reduction work.  Part of this story is about the different fluxes necessary to vitrify stoneware, and talc (magnesium) is one such flux.  Probably a year ago I could not have answered this question with much confidence, but in this particular case I now happen to have lab chemistries for all of the materials in the body.

 

The photo below (taken on a somewhat cloudy day in the late afternoon) shows 5 very similar bodies in both oxidation and reduction, all fired to cone 10.  For the reduction, we generally err on the side of heavy, and in this stoneware body I was looking for some nice toasty oranges and browns.

 

Progressive Talc Additions in Stoneware Clay Body

These clay bodies substitute in increasing amounts of talc to replace silica from top to bottom, mainly to see how magnesium would perform as the main body flux.  Nothing else changes.  The bottom tile is 1.8% molar of magnesium.    The top tile is just over 4% molar magnesium.  Both Fe and Ti are virtually unchanged in these five bodies.

 

I think there is a slight but noticeable change in color from top to bottom in both the oxidation and reduction series.  It helps to back away and look at the whole photo rather scrutinize each bar. In the oxidation series I would call it a graying or a darkening of the yellow.  In the reduction series it is more like a dulling or clouding of the red. 

 

Do others see this?

 

Will keep looking through my tests to see if there is more on Mg...

 

I

Link to comment
Share on other sites

I see it more in oxidation, than in the reduction. Not sure if iron (Fe) being counted as a flux in reduction plays a role or not. I can see a subtle color shift, but more pronounced in oxidation. Will be more than interested to see reduction tests, the part missing in my own. Glad someone has some added data.

 

Mgo Test 2

 

Bar 12 Mgo 0.23   Fe 0.32  TiO2  0.71      Weight % / not molar

Bar 14 Mgo 0.21   Fe 0.35  TiO2  0.73   

 

Moved my latest test pic from the porcelain flux thread to here.

End of next week (9-15) or so, will have another set of extensive test for MgO in oxidation. The test bars will be out of the bisq kiln tonight. Decided to bisq duplicates to see the effect of glaze leaching. In this next round of testing: I kept Fe (iron) out of the equation.

 

Nerd

 

Curt: I know several "stoners" (stoneware throwers) are watching with intent. Yes, I stole the 80's terminology back from the hippies. You have to fire clay to be a real "stoner."

Link to comment
Share on other sites

Interesting notion going on here. Can you get some old NYTalc and try the same experiment? One difference between the old NYTalc and the new Texas AMTAL talc is its raw color - old NYTalc is a white powder while the new TX talc is grey, but boldly advertised as "white-firing." The cause is small amounts of carbon in the source geology. The carbon supposedly burns out with whatever other organics and other non-ceramic volatiles, but does it really leave behind no after-effects? Hence my suggestion of a control test using a non-carbonaceous talc if you can get some.

Link to comment
Share on other sites

Dick:

 

The tough part of in home testing: setting a test standard and controlling the variables. If I only had an extra 1/4 mill or so for equipment.

I calcined 5 lbs of Texas talc a couple of months ago: its all "off" white and perty now. (1900F).

 

The next round of test should confirm what MgO is doing. I used only NZ kaolin, with Imsil A25, and a blend of mahavir and nep sy flux: all in the attempt to keep iron, Mgo, and TiO2 at the lowest possible level. The clay body produced Mgo 0.03, Fe 0.15, and TiO2 0.03: cannot get it any cleaner. So now the Mgo is strictly talc additions for 4 bars; then 3 bars of increasing Mgo, with increasing of TiO2. We shall see: this test should confirm it. Fe was purposely left out to verify if it is Mgo, or Mgo in combination with TiO2.

 

Test bar bench mark in molar  Mgo 0.04, Fe 0.07, and TiO2 0.02. 

 

Nerd

Link to comment
Share on other sites

Round 2: MgO test/s in oxidation:

Clay formula: 50% NZ kaolin 20% Imsil A25, 20% mahavir, 10% Nep Sy. ( no macaloid, Tgum, or bentonite additions)

Base line in weight: MgO 0.03  Fe 0.15  TiO2 0.03  ( noted as bar 7 in the test series).

 

 

Bar 1  MgO 0.29  Fe  0.16 TiO2  0.03   Bar 2  MgO 0.56  Fe 0.16   TiO2  0.03    Bar 3  MgO 0.81  Fe 0.16  TiO2 0.03

Bar 4  MgO 1.07  Fe 0.17   TiO2 0.03   Bar 5  MgO 0.55  Fe 0.16   TiO2 1.06     Bar 6 MgO 0.80  Fe 0.42   TiO2 2.05

Bar 7 (control) MgO 0.03  Fe 0.15  TiO2 0.03  Test bar 3 (old test)  MgO 0.78  Fe 0.42  TiO2 1.16

 

Conclusion: What the he** is going on here?    Bar 2 has a very light gray cast, not detectable in the photo: so I am reasonably certain that MgO is producing the gray tone. Fe (iron) was intentionally left out of this series of test to confirm gray color was from Mgo, or Mgo with TiO2. Bar 2 confirmed that, however bar 3 & 4 did not show the gray color even though they had higher percentages of MgO. This leads me to question if there is a particular chemical reaction going on when exacting levels of MgO are present. Bar/s 3 & 4 should be a darker gray because of higher MgO levels, but they are not. Although, a dirty bag of clay used in the old test bar 3 is not out of the question either.

 

Bar/s 5 & 6 increased the TiO2 benchmark of 0.03 maintained in bars 1-4 to (5) 1.06, and (6) 2.05. Again, a rather puzzling result: bar 5 shows a almond hue with TiO2 of 1.06, and bar 6 does not even though it has 2.05 TiO2. Again it makes me wonder if at certain levels some kind of chemical reaction takes place when MgO and TiO2 are at certain levels. The results are rather odd, and even harder to explain:  they do make sense on many levels.

 

I did throw in some test pieces from commercial porcelains. One of the most glaring comparisons was test bar 5 Mgo 0.55, Fe 0.16, and TiO2 1.06 showed a almond/Lt. gray hue: but Laguna porcelain (10) was even darker.

 
Says something about contaminant levels in commercial clays. Getting to the point where I can tell what clays were used in commercial bodies by the color they produce when fired raw.
 
Nerd
 
Back to testing, will probably be awhile before I have any new info. Have to back track and substantiate some previous tests.
Link to comment
Share on other sites

Great tests, Nerd.  Thanks as always for sharing.

 

Looking at your results, I find your idea that it is the relative LEVELs of these three contaminants to each other which is important, to be the most compelling explanation overall.  This idea had crept into my brain as I stared hard at your first round of results and seemed even more viable as I was looking back at my own tests.

 

I am nobody's artist, and have no training in color theory (maybe someone here does?) but it has occurred to me that talc's gray cast may only be visible in a field of some other, NON-WHITE color.  Put another way, if your clay is too clean and white, then maybe the gray has nothing to show up against??  Or, - and let's really go out on a limb here - could it be that that gray cast from talc is not a color in itself, but rather only a modification of some other color (eg, iron), which means that if the actual color is not visible, neither is its modifier?  That sounds like (almost) total nonsense to me, and happy to have it ripped apart by someone who actually knows what they are talking about, but even if patently false, maybe it will trigger some lateral thinking.

 

The tests below could be interepreted to offer further support for this general idea, or just brand me as a complete loony.  Note I do not live in Canada.

 

3 Cone 10 Porcellanous Bodies

 
The top two bars are a fairly clean whiteware body in cone 10 oxidation (left) and cone 10 reduction (right) with Mg 2.60 Fe .07 Ti .02.  I have chemistry on all the ingredients except (crucially) the clay itself, but looking at the fired results I think it is something like NZ Kaolin, so I have borrowed that chemistry for it.  If anything, it is not as clean as NZ Kaolin, but I am really focused on the Mg in this body, so I am not too concerned.
 
The bottom shard (1/3 the thickness of the bars above it) is an ultraclean porcelain not unlike Nerd's control bar, fired in cone 10 oxidation with Mg .02, Fe .06 and Ti .02.  (and its complete chemistry is known).
 
The top bars contain 130 times (no typo) as much Mg as the bottom shard.  But while my eyes seem some difference, it is not 130 times different.  No, in fact what catches my eye is that the reduced top bar (right side) is using its now blued iron tinge to trick my eye into thinking that it is whiter than it actually is.  The oxidized bar on the left (same exact chemistry!) looks filthy by comparison and is what make me confident that the clay in it contains more iron than NZ kaolin.   So lets revise that Fe .07 to Fe, say, .21.  Anywhooooo....
 
What these samples suggest to me is that the Mg effect needs some iron or other contaminant for it to be really noticeable.  The upper right bar has massive amounts of Mg, but looks almost as good as the control shard. The oxidized bar (upper left) is showing the iron, and quite possibly the Mg is what makes it look so shabby, since there is really no Ti in any of these bars.
 
Also makes me think that kiln atmosphere, not surprisingly, plays a big role in what colors show up.   
Link to comment
Share on other sites

Curt:

 

I think your premise in the original "flux limits for porcelian" is correct: which you state again above. I left the Fe (iron) out on purpose in this series: checking in part the theory you put forth. There is a light gray cast in 2 pieces, absent the Fe: so that sorta confirms to me that it is MgO producing that color. However, by increasing the TiO2; it did not increase the hue of the gray color: so that would exclue TiO2 as the color modifier. That would only leave the Fe as being the culprit in conjunction with MgO. What confirms it for me is these bars:

 

Vertical Bar 5 is the same as bar 5 above: it does have a slight gray hue, although the almond/buff color of the TiO2 over shadows it. However, when Fe (iron) gets into the picture: the gray hue becomes:

Mgo Test 2

 

Bar 12 Mgo 0.23   Fe 0.32  TiO2  0.71      Weight % / not molar

Bar 14 Mgo 0.21   Fe 0.35  TiO2  0.73
 
So I cannot rule out MgO as a contaminant that produces color: because it does-however slight that may be. However, perhaps classed more as a color modifier if used in conjunction with Fe (iron.) As you point out, if high purity clays/fluxes, and silica are being used to produce ultra white porcelain: then it is magnified all the more. I will run one final round of just bars 1-4, with increasing amounts of Fe to confirm that: but I fully expect to see the original colors of bar 2-3.  For now, back to fhe final flux limits testing to confirm my proposed limits. Now that I can "see" the results up close: all the more comprehensive. From there, body formulation in relation to specific use: in lieu of one size fits all.
 
**Testing note: the talc used was highly calcined to produce a nearly pure white color before blending.
 
I still believe formula limits have to be in place for MgO; in addition to Fe and TiO2. From testing, MgO will produce a strong gray cast in the presence of Fe. From bars 12-14: it takes very little of each to produce that color.
 
Nerd
Link to comment
Share on other sites

I managed to convince myself that bar 5 was yellow rather than grey. Which would make sense as blue and

yellow are opponent colours. Just as bluish reduced bodies look whiter, a yellowish body might look darker.

 

So I took averages of colours of a 50x50 pixel samples from the picture of the bars, and expressed it in 

CMYK (cyan, magenta, yellow, black) colour coordinates.

 

Renaming  "Test bar 3 (old test)" as  "Bar 3'", this gives:

 

Bar 1  MgO 0.29  Fe 0.16  TiO2 0.03   CMYK  7%  5%   0%  29%
Bar 2  MgO 0.56  Fe 0.16  TiO2 0.03   CMYK  8%  5%   0%  25%
Bar 3  MgO 0.81  Fe 0.16  TiO2 0.03   CMYK  8%  6%   0%  35%
Bar 4  MgO 1.07  Fe 0.17  TiO2 0.03   CMYK  7%  5%   0%  22%
Bar 5  MgO 0.55  Fe 0.16  TiO2 1.06   CMYK  0%  4% 14%  28%
Bar 6  MgO 0.80  Fe 0.42  TiO2 2.05   CMYK  0%  2%   3%  37%
Bar 7  MgO 0.03  Fe 0.15  TiO2 0.03   CMYK  3%  3%   0%  26%
Bar 3' MgO 0.78  Fe 0.42  TiO2 1.16   CMYK  3%  0%   1%  31%

 

Bar 5 certainly seems to be a lot yellow-er than the others.

 

 

It might be interesting to look at bar 5 though a microscope to see if the colour is uniform at a finer scale,

or consists of discrete speckle of spots.

 

 

... oops, lost half the text as I tried to post. Will try again later

 

Link to comment
Share on other sites

... as I was saying.

 

I cannot think of any obvious yellow pigment that might be involved, although I do feel that

iron might well be involved somewhere.

 

Looking at the differences between bar 4 and bar 5.

MgO down 1.-7 -> 0.55, Fe stable 0.17 -> 0.16, TiO2 up 0.03 -> 1.06

Which sort of argues for Ti involvement. 

Well TiFe2O5 (pseudobrookite) is certainly yellow enough, as seen in "marbled terra sigillata"

http://tinyurl.com/jc284bj

... I would love to hear of any other slip or glaze that uses pseudobrookite as a colourant, and

the conditions under which it can be reliably produced.

 

It would be nice if the colourant was an Mg compound. Magnesium ferrite (MgFe2O4) is a potential

candidate, but I've usually seen it mentioned as a reddish brown colour:

- as an oil colour

http://www.danielsmith.com/Item--i-284-300-040

- as an iron-red glaze

https://www.jstage.jst.go.jp/article/jcersj/113/1314/113_1314_161/_pdf

https://www.jstage.jst.go.jp/article/jcersj/113/1315/113_1315_232/_pdf

Although the colour is described as "presenting a yellow-orange color" here:

http://link.springer.com/article/10.1007/s10973-006-7744-6

 

Anybody have any thoughts on candidate yellow pigments, and why it only appears in one of the test bars?

 

 

Added: This paper suggests that getting pseudobrookite to work as a yellow ceramic pigment isn't that easy,

with a tendency to produce browns.

http://tinyurl.com/jjcq57j

 

 

Link to comment
Share on other sites

I think Peter is on the right track with brookite for bar number 5. 

 

I just ran across this in Hamer and Hamer under Titanium Oxide:

 

"Titanium Oxide acts in three ways on colors.  In small amounts it intensifies the colour making it brighter, eg iron oxide can be induced to give bright yellows and oranges.  Moderate amounts of titanium oxide (2% to 6%) break up the colouring effect by a mottle. Larger amounts tend to subdue colours but give interesting sparkles to the surface."

 

This is interesting for a few reasons. 

 

First, it is a higly likely answer to Peter's question about why this yellow color only appears on Bar 5: namely that Bar 5 has just the right amount of iron and titanium (and maybe Mg?) to produce this color.  Similarly, it suggests, Nerd, that your bar #5 MIGHT be pointing the way to a reliable blend of Mg, Fe and Ti to produce that brookite effect Peter was after.

 

Following on, it also raises new questions about your test bars (at least for me).  Now it would seem that Bar 2 and Bar 5, with identical amounts of Mg and Fe, but vastly different Ti (35x different), are the interesting ones to compare.  One is white and one is orange!  Bar 6 also zooms into the foreground, with more of everything, but no mostly white and no orange color to speak of.  But with that much Ti, does Bar 6 have a sparkly or mottled surface a-la Hamer and Hamer?

 

For me this explains a lot, and tends to reinforce the idea that their are certain LEVELS of groups of materials like Mg, Fe and Ti which matter a lot to color. Above or below these levels, little or no color modification takes place.  But get in the right zone and big color changes can occur. 

 

If time and money were no object I would be exploring several line blends of porcelain, each with its own stable amount of Fe, but then adding in Ti in increasing amounts to each specimen...

 

Not sure where this leaves the whole Mg discussion, but to my mind it does answer some questions about what these tests are showing

Link to comment
Share on other sites

Curt:

Think I have come to the same conclusion: been reading "Ceramic Science for the Potter by Lawrence & West" looking for additional info. There are very specific combinations that develop colors: including graying from MGO. The only time I see a gray color is when MGO levels rise: but the other levels have to rise as well.

 

Bar 5 is what I would call buff/ light almond perhaps: but a yellow hue to it. Bar 6 has the beginnings of mottling: random patches here and there. The only real explanation for the gray color in the original bar 3 post is a dirty bag of clay, with much higher MGO levels than programmed in the calculators. Could be a fluke: but I will be watching for color shifts with clay orders ahead. Hopefully this does not turn out to be a long term problem like Custer.

 

Speaking of Hamer & Hamer; been looking for my copy for 3 weeks. In the midst of my nerdiness, I do not remember where I put it. Probably glazed and fired it: been rather absent minded lately.

 

Nerd

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use.