I hinted at this in my previous blog when I mentioned a brand new printing process developed by Photrio user @Raghu Kuvempunagar. Let’s have a closer look at this, because it’s quite striking in its own right and I suspect it’ll open up avenues for variations as well.
First off, let me be clear about this: all credits go to Raghu for developing and very transparently documenting this process. What I report here is just my first attempts at performing the process as he described it. This posts offers a few examples and details of my workflow.
What’s this ‘FerroBlend’ process? At its core, you could regard it as a combination of a classic cyanotype and a cuprotype, although I’m going out on a limb by referring to the cuprotype as such. The process basically consists of coating a ferricyande + ferric ammonium citrate (FAC) sensitizer and exposing that (which makes this a cyanotype process at its core), followed by a ‘development’ step that introduces a copper salt. This results in a print that consists in part of a cyan/blue-colored iron-based cyanotype (Prussian blue) image together with a pink-to yellow-colored copper-based image.
The process is quite simple to perform, but getting rid of all fog appears to be challenging. Under advantageous conditions, it appears that a very low-fog print is possible, in which case the process proves to provide interestingly split-toned, very long-scaled prints.
This is one of the first successful tests that I obtained, with is on Schut Salland paper, a quite beautiful, velvety-feeling 50% rag printing paper. Like most papers, this is alkaline (chalk) buffered, which can produce problems with alternative printing process as the alkalinity (high pH) interferes with processes, usually resulting in high fog and low maximum density. However, in this case, I have mostly managed to avoid these problems by taking a few measures I will discuss later.
First, let’s do a run-down of the process and materials.
Sensitizer
The sensitizer is a classic cyanotype sensitizer with ratios adjusted to meet the requirements of the process, notably an excess of potassium ferricyanide. The composition is as follows:
- Ferric ammonium citrate (FAC): 12.5% w/v or 125mg per 1ml
- Potassium ferricyanide: 10% w/v or 100mg per 1ml
With an affordable digital jeweler’s scale, these amounts can be fairly easily weighed out to mix the sensitizer as needed. The mixed sensitizer should be used up within an hour or so; it does not keep.
To allow for an easy workflow and long-term storage, two concentrated stock solutions can be mixed: solution A would be 25% w/v FAC, while solution B would be 20% w/v potassium ferricyanide. These separate solutions can be stored (in the dark) for a long time and can be mixed in a 1:1 ratio to make the sensitizer as needed. The FAC solution tends to grow mold over time; this can be filtered out and the solution can still be used.
Apply the sensitizer to a suitable paper with a brush. It’s a free-running liquid that can easily be brushed with good evenness. A little goes a long way; I can coat about an 8×10″ sheet (or even a little larger) with 1 ml of sensitizer. It depends a bit on the absorbancy of the paper. I don’t use any surfactant (Tween, Photoflo etc.); in my experience, doing so tends to do more harm than good, in general.
If you notice clear streaks or the sensitizer paper turning green or blue as it dries, try a different type of paper. The sensitizer should dry to a somewhat pale yellow-lime hue that’s not much different from what it looks like when you just brush it on. You can safely accelerate drying with a hairdryer at high heat.
Exposure
Exposure is under artificial UV or sunlight in direct contact with the negative in a contact printing frame. Expose until the shadow areas solarize into a silvery grey tones, just like with classic cyanotype.
You can tell by the example shown earlier (of the finished print) that the tonal scale of the process is quite long. Raghu recommends negatives with a scale of 1.8-2.2 logD, but I’d err on the long side, extending even to 2.5 and higher. This makes for a longer tonal scale than e.g. New Cyanotype, and more in the ballpark of salted paper. The exact tonal scale requirements are a little uncertain at this point given the mild fogging problems that can be expected, which make the process appear a little longer-scaled than it really needs to be.
I always recommend with processes like these to print with a masked border, so it’s easy to tell how much fogging you get of the sensitized but unexposed paper. This really helps in figuring out any process problems. You can see how the borders are neatly masked in the prints on this page; this is not just an aesthetic consideration!
Development
After exposure, ‘develop’ the print by brushing a developer onto the paper. This developer is composed as follows:
- Copper sulfate: 2% w/v or 20mg per 1ml
- Trisodium citrate: 8% w/v or 80mg per 1ml
- Optional sodium chloride (table salt): 10% w/v or 100mg per 1ml
A concentrate can be made at 5x the strength (10% w/v copper sulfate resp. 40% sodium citrate) and then diluted 1+4 before use. This developer is stable and should keep well for months or even years, either as working strength solution or a concentrate.
The developer is brushed onto the print right after exposure. I find you only need a small amount of developer; just enough to cover the exposed area of the print. I find that the appearance of the print changes almost immediately, with orange and reddish tones appearing in the midtones and highlights.
As I understand it, the appearance of the orange hues is actually due to the copper in the developer reacting with white/colorless Prussian white pigment that is formed during the exposure. By the copper binding to this, a warm-colored pigment forms, which creates the second image on top of the cyanotype that’s already largely visible.
Interestingly, the copper-augmented image extends far deeper into the highlights than the cyanotype image itself. It seems to me that very scarcely exposed ferric salts couple effectively with the copper compound, gaining significant density, and lodging themselves firmly in the mesh of the paper. In a classic cyanotype, this entire part of the tonal scale remains invisible or even washes out of the paper. With the FerroBlend process, these delicate midtones and highlights are preserved by their combination into an insoluble copper-iron pigment, effectively extending the tonal scale as well as creating a duotone image.
With the materials I’ve used so far, development seems to be complete in a matter of minutes; somewhere between 1 and 5 minutes. Raghu mentions longer development times of up to 30 minutes. I find that if development is extended, no image-wise copper image is gained, and only fog is added to the highlights.
After development, rinse the print in a weak solution of citric acid (e.g. 1%, although I use a little less generally) followed by a wash. It seems that the prints withstand a thorough wash of half an hour or so without permanent density loss, and this may in fact help minimize fog.
Just like a cyanotype, the colors take one-two days to develop fully. In particular, the Prussian blue of the cyanotype part of the image needs to fully oxidize over time. I’ve done a brief test by adding a tiny amount of hydrogen peroxide to the final wash water to immediately intensify the blue image. This seems to work OK without affecting the copper image, but I only did a single informal test. Proceed at your own risk.
Dealing with the fog
The main problem with this printing process appears to be its propensity to fogging of the copper image. This can be seen in the example above. So far, I’ve found a few ways that seem to help in minimizing this fog – although I’ve not yet managed to fully eliminate it.
Fog can show up at two stages of the process, I’ve observed so far. Evidently, it can manifest at the end of the process as the print is being washed. At this point, you will notice how much of the sensitized, unexposed ares will effectively wash away and fade to paper white. However, at an earlier stage, bad cases of fogging can also manifest itself while applying the developer. This happens through rapid precipitation of an almost black residue that won’t stay in solution and tends to be brushed around the paper:
My first day of testing with this process resulted in prints that all failed in this way. I tried everything (including many things on the list below). For some reason, probably a combination of the factors outlined below, results improved dramatically in further testing.
- Reducing exposure to non-image light. It’s kind of obvious, but since classic cyanotype is relatively immune to fogging, I was being rather sloppy at first in practicing this new FerroBlend process. As shown above, the process has a long tonal scale, which means it’s quite sensitive to even small amounts of stray light that contains UV. So better work at night under lamplight, or with properly shuttered windows.
- Raghu proposes to add ammonium chloride to the sensitizer end/or the developer. I’ve tried this and it seems to work some, but I also suspect that the ammonium may in fact do more harm than good. Instead, I’ve had good results by adding sodium chloride to the developer – which is why I list it in the ingredients above, contrary to the original recipe by Raghu. I’ve not yet attempt to add any sodium chloride to the sensitizer. Raghu alerted me to a quote from Mike Ware (I don’t know where it appeared, unfortunately) that mentions a possible mechanism through which small anions like chloride might reduce fogging:
…but many printers still resort to pre-treating alkaline buffered fine art papers with dilute acids, such as hydrochloric or sulphamic, in order to destroy the chalk before coating with sensitizer. It is possible that this ‘decalcification’ treatment also brings an added hidden benefit: the small anions such as chloride or sulphamate from the acid will tend to neutralise the cationic polyelectrolytes that are present as retention aids, unacknowledged, and thus may proof the treated sheets against blotching.
- Adding a few drops of 10% ammonia to the developer instead of sodium or ammonium chloride may help, too. I’ve done some very cursory testing, and adding 3 drops of ammonia to 1.5ml of developer (enough for two 5×7″ prints) seemed to produce significantly lower fog than when adding sodium chloride to the developer.
- It’s very well possible that alkaline buffers play a role in fogging with this process as it does with other processes. Pre-soaking buffered papers in a dilute solution of sulfamic or citric acid is supposed to help break down these alkaline buffers. I’ve tried this and it seemed to help a little with the Schut Salland paper. It made no appreciable difference with Schut Simili Japon. Because of the limited effects and the extra step it adds to the workflow, it’s not my preferred approach, but it’s worth a try if you want to eliminate the fog.
- Raghu proposes development times of 5-30 minutes. In my tests, development seemed to be complete by the 5 minute mark. I think it’s probably best to stop development once it gets to the point where you like the result. Prolonged development just gives the copper-based developer to throw down more fog.
- I’ve not tested this, but I have the impression that a more concentrated developer (e.g. 1+1 or 1+2 from the concentrate, instead of 1+4) results in less fog. A possible mechanism might be the higher relative amount of citrate in the working strength developer. A stronger developer will also allow for shorter development times, which may help some (see above).
- It’s possible that some batches of FAC work better than others. I’ve corresponded some with Raghu about this and he suggested this might be part of the cause of surprisingly heavy fog in my prints under conditions that produced perfectly fine prints for him. FAC is apparently an ill-characterized, variable substance and perhaps some batches work better than others. Raghu is working on alternative ways to perform the process that avoid FAC altogether, essentially based on Mike Ware’s Simple Cyanotype. Unfortunately, I did not have the required iron nitrate at hand, so I could not test this route.
Some measurements
Finally, for the ultra-nerds among us, I’ve done some measurements on one of the cleaner, fog-free prints I’ve made so far. This makes some of the aspects that are also apparent on visual inspection more tangible/objective. Firstly, the color measurements allow for a digital reconstruction of the printed step wedge you can see in the prints above, illustrating the palette obtained (at least by me) with this process:
The hue transition essentially shows a short-scaled cyanotype on the left and a long-scaled cuprotype overlayed on top of it, extending all the way to the densest steps of the step tablet. If we plot a* vs b* from Lab*, this also becomes apparent:
The dark hues are in the bottom left, in the cyan corner. The first steps take big strides towards the upper right corner where orange is, and then inch back towards neutral as density decreases in the lightest values. Ideally, without any fog, the plot should finish somewhere close to the origin – although paper is usually not perfectly white.
Visual density follows the following pattern:
Note the gradual tonal scale of the highlights – i.e. the warm-toned copper image. The plot gets a lot steeper towards the shoulder, where the cyanotype part adds the most density. If we plot not just overall density, but also cyan, magenta and yellow separately, the transition from cuprotype highlights to cyanotype shadows becomes very visible in the steepness of the cyan curve, and the odd kink in the yellow curve:
And that’s about as far as I’ve come in two days of testing! The process is (as far as I know) brand new, which means that we’re only scratching the surface as of yet. Interesting avenues that come to mind, would include:
- Obviously reducing or even eliminating the fog. Clean whites would be a boon.
- Selective development is an obvious step forward; since the developer is brushed on, doing this selectively is something that’s more or less guaranteed to work.
- Somehow manipulating the transition point between the cyan/blue cyanotype and the pink/orange cuprotype. I feel this should be possible by leveraging the Simple Cyanotype concept. Raghu proposes an alternative way of developing the FerroBlend by adding the required excess potassium ferricyanide to the developer instead of the sensitizer, which should make it possible in principle to use either Classic or Simple Cyanotype. New Cyanotype likely won’t work as a basis since it will produce a solid precipitate (much like in my failed attempts).
- Selectively toning the cyanotype and/or the cuprotype with suitable toners, to manipulate the hues away from their natural colors. I’ve done one very quick (and very dirty) test involving tannic acid; in principle, this sort of works, but I’m not particularly adept at toning cyanotypes and I’m sure in more able hands, very interesting results should be possible. For instance, tannic acid toning of the cyanype and iron toning of the cuprotype might be an interesting way to produce a more neutral split-toned print.
…and probably many more. Let’s leave this to the more imaginative printers – I’m sure they’ll come up with gorgeous, fascinating adaptations of the process.
Finally, one more time – all credits of this development go to Raghu. Give him a big round of applause!