How to ab-stain? Dichromate stain issues in carbon printing

No, I haven’t given up on the color carbon project. Yet! But there are challenges, and they can be, well, challenging. For one, I’m running into trouble with hue and chroma of the color layers and I’ve been having a hard time figuring out what happens. One probable cause is actually dichromate staining. Let me exstain. Err, explain.

It all starts with something like this:

Different shades of magenta

Alright, not the best illustration, because a lot of the difference in these three test prints displayed side by side can be explained by differences in overall density. But compensating for this, I still ran into the odd phenomenon that some of my test strips looked higher in chroma/saturation than others, and I also think I’ve been seeing hue differences between test prints from the same tissue batches. It seems to happen with all colors/pigments, and I’ve had a very hard time spotting any variables that could explain the differences. I also tried to quantify the difference and took some Lab* measurements on two pairs of cyan and magenta test strips that didn’t quite look the same:

Sample #1Sample #2difference
Cyan (PB15:3)(53, -10, -48)(52, -18, -34)(+1, -8, +14)
Magenta (PR122)(51, 68, -15)(50, 60, -4)(-1, -8, +11)
Lab* measurements on patches of similar Lightness for cyan and magenta tissues

 Visually, I judged (rather subjectively) that something wasn’t right with sample #2 and sample #1 looked as it should. I determined this mostly based on the apparent purity and saturation of the hues and the difference in hue and chroma between the processed test strips and fresh, unsensitized tissue.

If I translate the differences between sets #1 and sets #2 in words, it would be that the sample #2 patches are both more green and more yellow. Or, conversely, the sample #1 patches are more red and more blue than the #2 patches. Picture instead of 1000 words? Here you go:

Left to right: sample #1 cyan, sample #2 cyan, sample #1 magenta, sample #2 magenta

The visual impression of the difference I would describe as the #2 patches being less pure in hue and less saturated. But…why?

I was kind of stuck for an answer, so I reached out to Calvin Grier to see if he had any wisdom to share. Sure enough, he was so kind as to actually respond to my emails. He was under no obligation whatsoever to think along, but he did so anyway! The very first thing he offered was that my problems were related to dichromate stain.

In all honesty, I was critical of this. The reason was that while there appeared to be a correlation between the sensitizer strength I used and the degree to which the strips were affected with the chroma-loss problem, there were very clear outliers. I had a couple of strips with very low dichromate concentrations that yet looked affected. Moreover, the effect appeared to be quite abrupt: I trialed 4 distinct sensitizer concentrations and only the strips with the highest concentration seemed to suffer, and all others didn’t – except the couple of outliers indicated before.

There was one more thing that didn’t sit well with me. Calvin advised me to soak an affected print in water overnight; any dichromate stain should clear (or at least diminish drastically). Which I tried, and it had no effect. I also soaked some affected strips in a weak solution of sodium metabisulfite (a popular ‘clearing’ agent for dichromate stain; I’ll come back to this later) – no change. Also no change in strips I soaked for some time in carbonate (high pH) and acetic acid (low pH) solutions to see if pH had any effect.

Well, Calvin was actually so accommodating to subtly nudge me onto a discovery track. Frankly, I respect him for choosing this approach: instead of outright telling me I was wrong, he suggested an experiment to me so I could find out for myself. No doubt he’s an effective educator! The experiment was simple enough: give it a try with some unpigmented tissues. That’s right: make a couple of prints with carbon tissue, but no pigment at all. This makes it easy to see the effect of any sensitizer stain. Alright, I can do that!

So I poured some tissues consisting of 8% gelatin, 3% sugar and around 0.5% glycerin, which is my default glop recipe at the moment. After drying, I took a couple and sensitized them with 1% and 16% dichromate concentrations. I use 0.5ml of sensitizer solution for this size tissue, diluted with a few ml’s of ethanol, and apply with a foam roller. The tissue size is 12x15cm (0.018 m2) which translates to a dichromate load of ca. 275mg dichromate / m2 for the 1% sensitized tissue and nearly 4500mg / m2 for the 16% concentration. Or, perhaps more relevant: a tissue takes around 18ml of glop, so that makes nearly 1.5g of gelatin, which results in around 3.3mg dichromate per gram of gelatin for the 1% sensitized tissue and nearly 55mg dichromate per gram gelatin for the 16% tissue. If my math checks out…

Well, the differences in sensitizer strength show up pretty much right away, which is obvious if you just look at a bottle of 1% dichromate next to a 16% bottle. But it’s clear on the tissue as well. Let’s have a look at some freshly sensitized tissues:

Sensitized and dried gelatin-only tissues (unpigmented). Left: 1% dichromate, top right: 16% dichromate, bottom right: processed test print using 16% concentration.

I admit I had never actually tried making prints with just gelatin and no pigment, but I can really recommend doing so. It’s an interesting experience. For one thing, it’s interesting to see how the gelatin emulsion actually soaks up the sensitizer and with strong sensitizer concentrations it’s even possible to see (well, a bit, at least) the evenness of the sensitization. Also interesting to see is the actual dichromate print-out image. I’m used to seeing this as it stands out pretty clearly on color tissues, but it’s quite fascinating to see it without any underlying pigment, so just the dichromate image itself:

Print-out dichromate images, after exposure and before sliding the prints into the mating bath; left: 1% sensitizer, right: 16% sensitizer. Exposure adjusted to give roughly similar gelatin density/thickness in the processed print. Note the darkening of the exposed dichromate.

It’s also a kind of weird experiment to develop a near-transparent carbon print, by the way. Have a look at this one, made with a 1% sensitizer concentration:

Transparent carbon transfer on Yupo, still wet. No pigment, only a little dichromate stain.

So how about the results? Here are some processed test prints resulting from this experiment:

Dried test prints with no pigment

Well, that already shows something, but let’s take a closer look. The image below shows a small section of each of the prints above, in the same order (left – right becomes top – bottom):

Dichromated gelatin tissues without pigment. Top to bottom: (1) 1% sensitizer, no wash or clearing (2) 16% sensitizer, no wash or clearing, (3) 16% sensitizer, no clearing, washed/soaked for >2 hours, (4) 16% sensitizer, ‘cleared’ with bisulfite, but not washed (long).

You can see the results above. I took a small part of each test print (a 4×5″ step density pattern) and put them side by side for comparison. First, I made two prints with a 1% and a 16% sensitizer strength; the extremes of the concentrations I use. These were exposed to give roughly similar gelatin thicknesses (‘density’ – albeit there’s no pigment to build optical density) and transferred to plain Yupo. Apart from a brief rinse of one or two minutes, neither of these were washed or otherwise treated. As you can see, the 16% sensitizer strip shows severe dichromate staining. I sample a Lab* value on a dark patch of ca. (87, -0.5, 25) – so quite yellow. By comparison, the 1% sensitizer patch of the same ‘density’ (gelatin thickness) measures around (98, -1, 3) and paper/Yupo white rounds out at (100, 0, 0).

So I then decided to try and apply the common methods for clearing this stain: simply soaking in plain water and soaking in a weak sodium metabisulfite solution. I did not test the combination of both measures, but apart from each other, they achieve a comparable effect, but with a different final hue. The extended (> 2 hours) soak in plain water shows a persistent yellowish stain. The Lab* value I measure on one of the darker patches is around (94, -1, 11), so distinctly yellow, maybe a fraction of green and quite a bit lighter than the untreated patch of the same concentration, but not quite as pure as the 1% dichromate patch of the same gelatin thickness. Perhaps this remaining stain drops further by soaking even longer, but I consider a process that requires such extremely long wash times to work reliable undesirable, so I didn’t bother trying it.

The bisulfite treatment shows a remaining stain that’s near neutral in hue and comparable in lightness to the one that soaked in water for a couple of hours. The Lab* value of the reference patch is (93, -1, 3): most of the yellow is gone, but some density remains. Overall this gives a similar final outcome as the extended soak, but it’s much quicker. There’s a catch here, though. A treatment with bisulfite (or vitamin C as suggested in the Carbon Printing book by King et al.) is often called ‘clearing’, while technically, this is incorrect. What happens is that the chromium (VI) ions that are responsible for the yellow/ochre dichromate stain are reduced to chromium (III) which happens to be (1) less toxic, (2) more neutral to slightly greenish in color, but also (3) far less soluble. So whatever stain is left after ‘clearing’ (more accurately, reducing) a dichromate stain will be very persistent and virtually impossible to wash out.

This leaves me with a somewhat unsatisfactory conclusion, namely that dichromate stain is very real, but also very persistent and there appears to be no quick way to get rid of it. I have never worked with DAS sensitizer, but understand it also leaves a very pronounced stain – however, I also understand this can be cleared with permanganate, and I wonder if this clearing is perhaps more of an actual clearing and not so much a conversion with its own distinct downsides as is the case with dichromate.

Now, there’s still a loose end I’d like to pick up on. In the table with the Lab* values for the cyan and magenta patches above, I also calculated the difference between the ‘good’ and the ‘bad’ patches (subjective assessment). These were (+1, -8, +14) for cyan and (-1, -8, +11) for magenta. Ignoring the L* value for a bit, this leaves a distinct green shift (-8) and an even stronger shift towards yellow (+14 resp. +11). By the way, this allows me to make a final illustration: a visualization of the hue impurity of the cyan and magenta patches I measured, together with the hues of the dichromate stain I produced with the clear gelatin tissues.

Impurity and stain hues. From left to right: (1) hue difference between the cyan ‘good’ and ‘bad’ patches, (2) same for magenta, (3) 16% dichromate stain, untreated, (4) 16% dichromate stain, extended wash, (5) 16% dichromate stain, bisulfite reduced

The yellow shift seems to match up fairly well with a dichromate stain, which can even be stronger as illustrated by the unpigmented gelatine experiment. However, the green shift in the cyan and magenta patches remains unexplained. The unpigmented gelatin experiment showed no significant shift in this direction, after all. Perhaps this is a visual interaction between the dichromate stain and the pigment – if so, it’s beyond my understanding and hence just a wild guess. But it doesn’t sit entirely well with me, this loose end. It suggests that there’s more going on here that I haven’t figured out yet.

Leave a Reply

Your email address will not be published. Required fields are marked *