The original version of this article was first published on the Digitalfire website.
Super-Refined Terra Sigillata
Note: This article is offered free to everyone, with the clear understanding that it cannot be copied for sale under any circumstance, and can be copied for distribution only with attribution to and written permission of the author.
Super-refined polishing slips were first used in the
1400 years later, 19th century German ceramic chemist Henrich Schumann was seeking low-cost coatings for sanitary sewer pipe. That doesnít sound very romantic, but it drove him to make the critical rediscovery. Schumann was fascinated by the surfaces on Roman pottery, and conducted cross section analysis of the shards. He determined that terra sigillata is a very thin coating of refined slip composed of only the very finest clay platelets. Through subsequent experimentation, he was able to reproduced high quality terra sig surfaces. Schumann's findings guided me, and experiments conducted with students Heather Shadron and Rob Williamson at the University of Massachusetts at Amherst helped to refine the specific techniques outlined below.
Why Does it Work?
To anyone new to terra sig, it seems magical when you polish the surface for the first time. Most experienced clay workers have discovered that almost any very-hard-leather-hard clay surface can be rubbed to create a soft gloss, or burnished to create a high shine, which involves a lot of work. Kaolinite crystals are thin and flat and are reflective on the surface, and the resulting platelets have the same reflective qualities. In normally-worked clay, the alignment of these flat particles is random on the surface of the ware, and thus there is little or no reflective quality. But if you do anything to cause the platelets to lay down flat, the amount of reflection increases. When you process terra sig according to the instructions below, you eliminate all but the finest fraction of particles, and those remaining ultra-fine platelets will naturally lay flat on the surface, producing some degree of reflection. Any compression through polishing with a piece of chamois or plastic film or even with the finger tip will increase the fraction of platelets laying flat on the surface, increasing the reflected light and our perception of a shiny surface.
What's the Difference Between Polishing and
Get an Appropriate Hydrometer
Selecting the Clay
Some clays are so coarse as to give a very low yield of sig. If you choose a particular clay for compelling reasons, and the yield of terra sig is especially low, the raw clay can be ball-milled before incorporating in to the initial mixture. See the section on ďAdding Coloring Oxides and Ceramic StainsĒ for more information on ball-milling.
Preparing the Deflocculant
The amount of deflocculant is figured as a percentage of the dry weight of clay to be used. To a small amount of hot water, add 0.25% (1/4 of 1%) soda ash (sodium carbonate) and 0.25% sodium silicate. A hand blender will make short work of dissolving the chemicals. The sodium silicate comes as a liquid, but just measure the weight as if it were a solid.
Measuring Specific Gravity with a Hydrometer
Immediately before measuring the specific gravity with a hydrometer, always be sure to mix your terra sigillata thoroughly with a hand blender or drill-mounted impeller mixer. The sig will settle, especially in the initial mixture, and you must make sure that all particles are in suspension in order to get an accurate hydrometer reading.
Determining the Initial Mixture
Add the dissolved deflocculant mix to cold water, starting with a quart of cold water for every two pounds of dry clay to be used. Add the dry clay, blending with a drill-mounted impeller mixer. When all the dry clay is mixed in, add more cold water, mixing well, and check frequently with the hydrometer until the specific gravity reading is 1.20 for any clays except ball clays, or 1.15 for ball clays. If you are mixing in a full five-gallon bucket, float the hydrometer right in the bucket. If mixing smaller quantities, you will need a graduated cylinder or other appropriate container as mentioned above.
Settling the Initial Mixture
Donít be careless about this. Do not move or agitate the container in any way during the 20-hour settling time, or during the siphoning process outlined below. If you agitate the mix, you ruin the terra sig. Consistency in method and timing is critically important in making good terra sig. If you accidentally agitate the mixture, or if you let the mixture stand for longer than 20 hours, thoroughly mix it again with a drill-mounted impeller mixer, note the time, and start the 20 hours settling period again.
Why 20 Hours?
The 20-hour settling time is a compromise based on average clay particle distribution in air-floated clays. For a claybody containing a finer distribution of particle sizes, such as a ball clay, a longer settling time might be more effective. For any raw clay, a longer settling time might produce an even finer terra sig, but would also produce a lower yield.
Decanting the Terra Sig: Get a Proper Racking Tube
An appropriate racking tube features a ďJĒ-shaped rigid plastic tube attached to a length of flexible plastic tubing. An important feature of a good racking tube is the diversion cap on the suction end of the rigid tube. This allows fluid to enter above the cap rather than below, and offers real advantages in siphoning terra sig. As you gently probe downwards with the rigid tube, the objective is to sense the increased density of heavier sediments before the siphon starts to pick up those sediments, thus avoiding contaminating the terra sig. The diversion cap makes this much more feasible.
Siphoning off the Dilute Terra Sig
Start the siphon with the diversion cap just barely immersed in the settled mix. As mentioned above, do not agitate this container or the settled mix at all. Once the siphoning is started, gently lower the tip of the siphon tube into the mixture as the thin liquid is siphoned off, keeping the diversion tip just below the surface. The diversion cap will occasionally suck a bit of air, but that hurts nothing, and tells you that the tip is close to the surface. If it sucks too much air the siphoning action will cease, and youíll have to restart it. You will quickly get a feeling for this.
With some airfloated clays such as Redart, the heavier sediments will form a distinct layer with a fairly hard surface, and you will feel that layer with the tip of the diversion cap. As soon as this happens, lift the siphon tube and stop siphoning immediately. Donít risk picking up any of the heavier sediments.
With other clays, especially very fine ones like ball clays, you may never feel a distinctly heavier layer after 20 hours of settling. In that case, let the suction tube suck a little air every few seconds, and note the change when the siphon naturally begins to suck more air, indicating that the mixture is getting thicker. Once again, as soon as this happens, lift the siphon tube and stop siphoning immediately.
With most ball clays, the particle size is so consistent that you never will sense a heavier layer of sediment at all. In that case, just siphon off the top two thirds of the mixture and discard the rest, and you will get a good terra sig. That may seem inexact, but there doesnít seem to be any other choice, and that bottom third seems to accommodate the heavier particles. As instructed above, in making ball clay terra sig you should start with a thinner 1.15 specific gravity initial mixture, and that helps to allow the coarsest fraction of particles to settle out, so even if you are not able to identify thicker sediments in the lower portion of the vessel, you can be sure they are there. Donít try to siphon off more than two thirds. It just isnít worth risking contaminating the sig.
With most clays, the residue of heavier sediment left in the bottom of the container represents most of the weight of what you started out with, but I have never tried to do anything with that residue. Raw clay is cheap, while terra sig is very valuable based on your time invested. The residue already has the finest particles removed, and thus is relatively non-plastic. I just discard it, but you may find a good use for it.
Concentrating the Thin Sig by Evaporation
Concentrating the Thin Sig by Heat
Leave the lid of the kiln fully open for the duration of concentrating. On a kiln-sitter kiln, I leave the bottom and middle elements on low overnight or until the sig is adequately concentrated. On a programmable kiln, I just set the kiln on slow-bisque with a 24-hour preheat and intercept it before the end of the preheat. If it needs more time, I reset it for another 24-hour preheat. The atmospheric humidity will have a huge effect on concentration time. Again, leave the lid of the kiln fully open for the duration of concentrating.
As another option, use a crock pot set on medium heat with the lid off. In either case, the rate of drying will depend on the atmospheric humidity and the mount of heat applied. It will take some experimentation to learn the ideal conditions.
Using the Sig As Soon As It Has Concentrated
Storing the Sig in Dry Form
Determining the Specific Gravity for Application
When and How to Apply the Terra Sig
I apply terra sig to bone dry clay, and get the best adhesion when the clay has been sanded. Since most of my terra sig use is in conjunction with the Ancient Clay workshops I teach, the pieces are usually sanded, as is the tradition among Southwestern Native Americans and many other cultures past and present that polish or burnish their wares. I sand with 120-grit drywall sanding mesh, available from any building materials supplier or home-improvement center.
If you sand the wares, be sure to thoroughly wipe down all surfaces with a soft cloth such as tee-shirt material in order to remove all dust. If you neglect to do this, the paintbrush will carry the dust back to your sig container, contaminating the sig.
I apply terra sig with a wide, thick soft brush heavily loaded with sig, laying on flowing strokes, carefully avoiding drips. If the sig does drip down the surface, chase the drips with your brush. Donít let the drips harden on the surface, and donít let the piece sit in sig pooled at the base. I work my way around the piece, building up thin coats until I get an opaque surface that begins to conceal the sanded texture. Thatís still a very thin coat overall, but if you accumulate much more thickness, you are risking subsequent cracking and peeling of the sig coat.
If you wish to preserve some of the color and the surface texture of the base surface, you can get a good shine with an almost transparent coat, because the shine results from the ultra-fine clay platelets laying flat on the surface. Ultimately, itís a matter of experience with your particular terra sig and base claybody. Experiment with the thinnest buildup possible, and see if it meets your needs. Build up from there to see what you can get away with.
I have applied very thin coats of terra sig to bisqueware with adequate results, but it never produces the level of shine or durability of surface achieved with a sig applied to bone dry clay. Terra sig applied too thickly on bisqueware will almost always crack and peel.
Polishing and/or Burnishing the Sig Surface
Thereís no reason to polish between coats. When you do polish, if you are not satisfied with the shine, you can simply paint on more terra sig. The new coat disrupts the previous shine, achieving a good bond.
It is possible to burnish any re-moistened bone-dry clay surface, but usually some sort of refined slip is applied to the base clay. A properly prepared terra sig makes an excellent burnishing slip, but if you rely entirely on the moisture introduced by the sig, burnishing must be completed in one sitting while the coating is still damp. Otherwise the surface will begin to flake and crack. To prevent this, especially with larger forms, many potters apply the sig, allow the water to absorb, apply a thin smear-coat of lard or glycerin, and let that absorb before burnishing. The grease retards drying, giving you time to finish burnishing the piece.
Even though terra sig makes a good burnishing slip, it is not the ideal use of this medium. The whole point of this elaborate preparation process is to produce a thin slip medium that polishes beautifully and gives the unique terra sig shine, but doesnít involve the labor-intensity of burnishing. After proper application of terra sigillata, the actual polishing process on a medium-sized pot using grocery bag plastic takes ten or fifteen minutes. A thorough burnishing job on the same pot takes at least two or three hours.
The maximum temperature at which a terra sig retains any integrity to the original material depends on the clay used. At cone 012, Redart terra sig gives a bright brick-red-orange color. At cone 08 it gives a classic brick red, and cone 02 it gives a red brown. At midrange and highfire temperatures, a Redart (or other earthenware clay) terra sig will fuse and become a glaze with no resemblance or connection to terra sig.
Stoneware clays like Goldart and most fire clays give an off-white color at low temperatures, and at midrange and high-fire temperatures will produce a tan in oxidation and gray in reduction.
Ball clays generally give white at lowfire temperatures, and an off-white or gray at high-fire.
Most kaolins produce white at all temperatures, but some will give an grayed eggshell white in reduction firing.
Red stoneware clays like Carbondale, Lizella, or Neuman red often tend to be quite coarse, but can still give a good brick-red terra sig that will survive to high-fire temperatures without fusing into a glaze. Some red stoneware clays may need to be ball-milled before incorporation into the initial mixture.
Slip clays like Barnard or Blackbird are often too coarse to make a terra sig, but might respond well to ball milling.
For the purposes of blackware firing, a properly-made Redart sig produces a very dense coating, and gives a range of colors from brick red to jet-black, and any re-oxidized spots will go back to brick red. Ball clay terra sig is slightly more porous, and in a blackware bonfiring can give intense jet-black, and any re-oxidized spots will go to gray or white.
Adding Coloring Oxides or Ceramic Stains
Using Terra Sig to Coat Feet and Lid Seats
Slake down scrap claybody clay overnight in ample water and mix with a drill-mounted impeller mixer, adding more water to achieve the 1.2 specific gravity initial mixture. Donít worry about sand or grog, because those will be the first components to settle out in the initial mix.
If you do not use the base claybody, select a clay appropriate to the base claybody and the intended firing temperature. Remember, an earthenware clay slip will become a glaze at midrange and high-fire temperatures. Terra sigs made from stoneware clays like Goldart or from a ball clay or kaolin will work fine up to high-fire temperatures, and even then will still give a silky smooth surface.
Using Terra Sig
Under a Glaze
Anyone who has done considerable slip decoration has experienced the effect where a very thin application of slip simply disappears, having been absorbed into the glaze. That's what happens with terra sig. However, like any liquid medium applied to dry greenware or to bisque-ware, terra sig will settle more thickly in recesses and textural details, and thinly on high spots and flat areas. In the thicker areas, the glaze will not dissolve the terra sig coating, and it will have a localized effect on glaze color. This could be an advantage with almost any glaze at any temperature, but it is especially useful in electric-fired lowfire and midrange work, where the exposed claybody often looks pallid and lifeless, and where the glaze often benefits from more surface activation. A terra sig can bring interesting color and surface to unglazed areas, even at midrange and highfire temperatures, and it will give more variegation of color and visual texture in the glazed areas. If you are going to coat the foot and/or lid seats of the pot, be sure to use a terra sig appropriate for the intended firing temperature (see the section above on "Firing Ranges and Natural Clay Colors Achieved with Terra Sig"). That doesn't preclude using a Redart terra sig on cone 10 wares - it simply requires that you use wax resist or sponge off the foot and/or lid seats.
Is it Possible to Make a Terra Sig
that will Develop a Gloss with No Polishing?
After this man contacted me I thought about
my own terra sig process, and realized that I already have definitive proof that this will work
an ultra-refined slip with much finer average platelet size will give a greater
spontaneous shine without polishing. I
usually concentrate the siphoned suspension with heat, in which case I can
completely dry out five gallons of sig in 24 to 48 hours.
On occasion, when there is no rush, I'll let a batch dry out on its own
in a big stainless steel bowl sitting in a quiet corner of my studio.
That takes a lot longer, usually at least a month, and it is a matter of
days before the level begins to drop noticeably.
A coat of terra sig adheres to the sides of the bowl as the liquid level
drops, and the quality of that sig coating changes distinctly in bands from top
to bottom. Characteristically, when I come back to harvest the dry terra
sig, there is a very narrow band of satin-finish sig around the very top, and
immediately below that band the sig gets very glossy, with a band
of shiny sig at least several inches wide. The surface is not as shiny as
the best Greek Attic pots, but it's a very good gloss.
A few inches above the bottom of the bowl, the layer starts to get
thicker, and also more matt. In the bottom of the bowl is a
half-inch-thick layer of cracked "mud flats" of dry terra sig, and the surface of that
material is completely matt.
So what's happening to produce that
spontaneous high gloss? When the
bowl of suspension is initially left to concentrate, water is already
evaporating before any appreciable amount of sig particles have settled, since
the particles are so fine and are deflocculated.
Thus, the very narrow band deposited at the upper edge of the liquid
contains an average range of the platelet sizes in the sig.
But after a few more days there has been noticeable settling, leaving
much finer particles at the top. It
might take ten days for the level to drop an inch, and by then the platelets
remaining in suspension will be only the very finest that are present.
That will be the case as the level drops a few more inches, with only the
finest particles in the coating. But as the level drops to the lower
levels where there is still a broader mix of particles in the drying sig, the coating thickens and the
finish dulls, and when you get to the bottom, you've got a concentration of
coarser particles and a matt surface.
It might seem like the obvious solution would be to simply allow a longer settling time. It would give you a smaller yield, but it would develop a greater degree of spontaneous shine, and would probably be easier to polish further. Any disadvantages to this? Yes, at least with the way most people use terra sig. I generally apply sig to sanded surfaces, or to textured clay where I want to soften the surface texture a bit, and this requires a buildup of layers. With a finer distribution of particles in the terra sig, the shrinkage would be greater, especially in firing, and there would certainly be a considerable increase in cracking and flaking when the sig is applied in multiple layers. A sig with a much finer distribution of particles would tolerate only very thin application.
It makes perfect sense that an ultra-refined sig such as the Greeks probably used would give two results very advantageous to them. First, it would give very good color coverage and opacity with an extremely thin layer (with my recipe, when the first few coats of sig still seem transparent, it's because we're just seeing through the spaces between particles in the layer of sig, so a finer sig would have far smaller spaces and would be more opaque). Second, it would give a high gloss with little or no polishing. The Greeks no doubt prepared the surface in advance to give the degree of smoothness desired before applying the sig, and thus a thin layer sufficed for all their needs - opacity, uniformity, color, and shine.
there is good evidence that such an ultra-refined sig would have a lower
sintering temperature, and thus would vitrify enough to lock in the black color
resulting from reduction atmosphere at the end of the firing, while a slightly
less-refined terra sig would remain porous at the same maturation temperature
and would thus re-oxidize to red, buff, or white, depending on the clays used in
its formulation. This is likely the key to the ancient Greek firing
process that produced those amazing black-figure, red-figure, and white-ground
My recipe was devised to satisfy the
particular needs of bonfired or sagger-fired pottery, similar to a number of
different ancient and tribal traditions, but derivative of no single one.
So it is necessarily a compromise. The
20-hour settling time gives the combination of performance features needed.
In the finished product, the particle
range is still coarse enough to allow a build-up of layers to eliminate
the sanded texture in a claybody that must contain a lot of tempering material
(in this case, very fine grog) to give the thermal-shock-resistance necessary
for the open bonfiring process, and fine enough to allow a high polish with relatively little
effort (as compared to burnishing).
But if the objective is to make an ultra-refined Greek-style terra sig, it is conceivable that successive repeated settlings and siphonings simply produces a superior product as compared to a single longer settling. This is as area for some exciting research.
A well-prepared terra sigillata is truly a remarkable
medium. It has become a popular
option in contemporary ceramics throughout the