Carbon-Film-Transfer in Saggar-Fired Porcelains

by Dick Lehman

Reprinted with permission from
the March 2000 issue of Ceramics Monthly Magazine,
PO Box 6102, Westerville OH 43086-6102, USA;

Many year ago I discovered, in an accident of American Raku firing, that fresh leaves and grasses can, under the right conditions, create vegetation images on hot pots: During an all-day firing marathon, a storm blew up unexpectedly. The tornado-force-winds arrived so quickly that I didn't have time to stop the firing which was in process. I had just pulled a large bottle out of the kiln, had flamed it in sawdust, and covered it with a metal garbage can. Fierce gusts of wind unexpectedly blew the trash-can half a block away, and the still-thousand-degree bottle rolled down a dirt embankment, through some grasses and came to rest against a fence post. To say that the retrieval of both the cover and the pot was "spirited", would be to understate the duress of the moment. However when the resultant cooling had taken place, I discovered to my amazement that the pot retained images on the copper-stained surface from the grasses through which it had rolled.

Not wanting to abandon this fortuitous event, I continued to fire in the waning wind, hoping to possibly replicate the "accident" by placing the very next fresh-from-the-kiln copper-stained pot on its side, atop a mound of sawdust covered with freshly picked sumac leaves. Twenty minutes later I had my first successful follow-up piece: a lovely soft likeness of the sumac appeared as an image in the copper stain, on the side of the pot. And subsequent firings yielded only successes as I pursued this method.

Several years later I abandoned my pursuit of American Raku as I had known it -- disillusioned by the fading/re-oxidizing surfaces of the copper stain. The once- brilliant colors, after several years, became drab and muted, much to my disappointment and to that of my customers. Not having found a solution to the problem, I discontinued marketing these pieces. It was a hard decision: I loved the vivid images of vegetation against the serendipitous variation of the Raku coloring.

I began to look for some other method of firing which might capture, on the same piece, the magic blend of spontaneity and explicit detail. Somehow I intuitively moved toward saggar firing.

Looking back, I conclude that it was an unlikely leap of logic which caused me to assume that saggar firing would be the answer: that saggar firing would somehow capture the explicit detail of fresh vegetation pressed against pots, while still offering unpredictable spontaneous surfaces.

And it was likely one of those acts of grace or good fortune, which occasionally enter each of our lives, which caused the startlingly successful results in my first 'veggie-saggar' attempt: wonderfully explicit images of fresh vegetation dancing in and through the seemingly 'celestial/astronomical' patterning which blessed the rest of the pots' surfaces.

I say, "good fortune/grace", because for the next several hundred pieces I fired, there were no successes at all! Oh yes, the pieces had some successful saggar-fired markings, but no explicit imagery from the vegetation -- at all.

Perhaps it was a 'healthy' state of denial which caused me to keep trying, in the face of relentless failure ("Surely it will happen again if I just work hard enough."); perhaps it was my life-long interest in landscape photography -- my passion for images which were close up, which revealed explicit and intimate detail (like those made by Paul Caponigro and Arthur Lazar) that spurred me on; perhaps it was the too-easy transition from tornado-accident to regular successes in the 'veggie-raku' realm that caused me to remain stubbornly optimistic.

It was an odd and curious experience to have these several (original) very beautiful veggie-saggar pieces (which I had made) in front of me on my desk each day -- pots with delightfully-delicate imagery, pots with such explicit detail that I could see the veins and tears and worm holes in many of the leaves; pots which were the ceramic counterparts to the contact prints which I routinely made from my 4x5 negatives in the darkroom; pots which had seemed so easy to produce in my very first attempt -- and at the same time to be continually, perpetually, interminably failing in each attempt to produce any more of them.

I continued to investigate this mysterious process, but I was not able to explore it in a full-time manner. (I was producing and supervising in a full-time production studio while this pursuit was in motion.) So the loss of several hundred pots covered a period of about a year and a half. Each unsuccessful unloading had me scratching my head, writing more notes about procedural decisions I had made in that particular firing, and comparing my results with the pieces on the desk.

If ever a character defect was an artistic advantage, perhaps my stubbornness was, in this case. Unable to repeat the accomplishment of the first firing, but also unable to believe that I couldn't do it again, I kept trying. And eventually after a year and a half, I did finally begin to see at least occasional results which lived up to my best hopes.

I now know that the devil is in the details. After ten years of practicing this firing method, I have concluded that these are the variables to which I must pay attention, and which ultimately make a difference in the successes (or failures) of the firings: a) sawdust particle size, b) the amount of sawdust in the saggars, c) the type, thickness, substance and placement of the vegetation, d) the firing temperature, e) the kind of kiln in which the saggar is fired, f) the kind of saggar used, g) the manner in which the saggar is sealed, h) the length of time the saggar is cooled, and i) the manner in which all of these variables work together. It was only after paying close attention to these variables, and their interrelatedness, that I discovered what I had so blithely (and fortunately) stumbled upon in that first successful firing. However even now, more than ten years later, the best results are sparingly attained and seem elusive. With the success rate for the best works at far less than 20%, I've needed to develop a tolerance for more failures than successes.

Here is a general explanation of how I am now firing for the results which are shown in the photos which accompany this article: A saggar is partially filled with 5 inches of fine sawdust. The pot is laid sideways onto the sawdust and pressed down to create a "nest" in the sawdust. The pot is removed, and fresh vegetation is placed into the nest, and the pot is put back in place (on its side) on top of the vegetation.

Next, more vegetation is placed onto the exposed top side of the pot. It is then covered by an additional 5 inches of fine sawdust. The saggar is then lidded.

During firing, an anaerobic atmosphere develops inside the lidded saggar. The vegetation turns into "activated charcoal" and in the process, releases a film of carbon. The bisqued porcelain, being porous, absorbs the film of carbon, capturing the image released from the vegetation.

This speeded-up process mimics, according to the paleontologists to whom I have spoken, the much-slower fossil-formation process called "carbonization" or "carbon-film-transfer" (which, in nature, causes volatile materials such as nitrogen and oxygen to be squeezed out of vegetation, and chemical action changes the tissues of the vegetation into a thin film of carbon. What remains is a residue forming an outline of a portion of the previously living leaves. If thick accumulations of plants derived from swampy coastal lagoons or deltas are carbonized more completely, coal deposits may develop).

Because the saggar maintains a relatively anaerobic atmosphere (no oxygen), the vegetation does not burn (and instead becomes activated charcoal), and is still present after the firing is completed (albeit in a black, shrunken form of charcoal). The anaerobic atmosphere also explains why the carbon image on the pot is not burned away: red-heat temperatures would "perceive" the carbon in the image as a fuel, and burn it off the pot if oxygen were present to complete the combustion process.

Very slight air leaks in the lid of the saggar may cause white areas on the surface of the pot to occur (in these cases, carbon is burned away). Careful control of the particle size and depth of the sawdust can lend a bit of mastery over the light and dark areas -- allowing one to dictate whether the carbon images read as "negatives" or "positives", and adding to the sense of the 'miraculous' which sometimes attends these surfaces.

A final word about the "miraculous": the paleontologists with whom I have spoken have referred (when speaking about fossils) to the "miracle" of preservation, due to the vulnerability of organisms to decay and destruction after death (and the resulting potential loss of record of many 'individuals' or even entire species). Recently a 'combustion chemist' visited my studio and expressed a similar perspective regarding the unlikelihood of consistently creating these "fast fossils" (due in large part to the near-endless number of variables at work in 'simple' cellulose combustion or partial-combustion).

Something about the unlikelihood of this process, and its elusiveness are an attraction to me. It is an arena in which there is much to learn: one could choose to take a paleontological or combustion chemistry approach in an attempt to discover what is actually happening in the process. And such learnings would, no doubt, offer some keys to additional methods of control (and perhaps successes) in what is otherwise a rather capricious process.

And there is much to master simply through the continued practice of the firing process (regardless of what one understands of the paleontological or combustion-chemistry implications of the method). Careful observation and documentation of one's decisions will continue to yield tools which may offer a little more control.

However the inherent lack of control which will always accompany this approach to firing and surface articulation is also what contributes to some of it's most amazing successes. I believe that the process produces results which are far superior to anything which I might make, if I did control most or all of the variables. In that respect, this approach, while full of loss and disappointment, is also full of wonder and surprise -- an approach which transforms me into as much a "receiver" of these works, as a "maker" of them. It is, in part, this mystery which propels me to continue on with this approach.

Return to Contents

Return to Home Page