New positive siderotype (iron-based photography) process using iodine/starch as colorant

[David Knierim]

New Positive Siderotype Process using Iodine/Starch Colorant

Conventional cyanotype can be used for direct long-exposure photography, but the resulting negative print does not work well as a master for making positive prints. Scanning and digital reversal work, but aren’t in keeping with the idea of a simple analog process. Below is a new positive siderotype (iron-based) photographic process for use in long-exposure photography. Contact and shadow printing also work with this positive process.

The recipe below is an example. The process works over a fairly wide range of ingredient concentrations. Ideas for improvement are welcome, especially for reducing contrast, which tends to be on the high side.

Ingredients:

1. Unbuffered paper (no calcium carbonate) – any paper that works for cyanotype printing will work here. My favorite is UniversityProducts Perma/Dur Unbuffered Interleaving Paper (118 gsm).
2. Potassium Iodide (KI) crystals or powder.
3. Ferric Ammonium Oxalate (or Ferric Ammonium Citrate).
4. Ferric Chloride. (Corrosive, but relatively safe once mixed with the Ferric Ammonium Oxalate.)
5. Elmer’s Glue-All white glue. (Other brands would likely work too. Don’t use “school glue” nor “wood glue” as they are different formulas and don’t work very well.)
6. Corn starch (or tapioca or arrow root) powder.

Most of the ingredients can be purchased on EBay. The final two are not necessary if the paper contains starch. Starch is often used during paper manufacturing for dry strength and/or for sizing. The above UniversityProducts paper doesn’t have starch. Many old (1980’s and before) papers aren’t buffered and do contain starch.

Initial preparation:

1. Mix 10g starch with 30g water. Shake or stir to form a suspension. Let the starch soak for at least an hour. Shake or stir again immediately before use to reform the suspension.
2. Make the iron-salt solution: Mix 4.3g Ferric Ammonium Oxalate, 2.7g Ferric Chloride, and 18g water. Wear safety glasses, and gloves when handling the Ferric Chloride.
3. Dissolve potassium iodide crystals or power in water, 10g KI per 50g H20 (roughly 1 mole/liter). This should be plenty for developing a dozen 8” x 10” prints.

The starch solution can be kept for a week or so – until it gets moldy. The other two solutions should keep for months in a dark cool environment. Both the iron-salt and iodide solutions will stain absorbent surfaces.

To coat an 8.5” x 11” piece of paper, make 3.2g of sensitizer solution by mixing the following:

1. 0.8g iron-salt solution.
2. 0.5g white glue.
3. 1.9g starch suspension. Shake or stir the starch solution immediately before adding.

Mix thoroughly. If the paper contains starch, then the last two may be replaced with water. Leaving in the glue can improve maximum density, however, even if starch is present in the paper.

Pour mixed sensitizer across one end of the paper. Spread with a glass rod as is typically done with cyanotype sensitizer. After a few passes with the glass rod, push the remaining sensitizer off the edge of the paper. Don’t try to spread quite all the sensitizer, as spreading will become uneven once the glue starts to thicken. Finish while there is still a bead of sensitizer in front of the glass rod. Set the paper aside in a dark location to dry. Coated paper may be used immediately after drying or stored in the dark for a month or more.

Expose as with any cyanotype. For “quick” direct imaging, I use an old rear-projection TV lens that is optically fast, about F/1.4. This lens works well at 20 minutes exposure viewing a bright sunny scene. Slower lenses will require correspondingly long exposures. Photographing trees or other objects silhouetted against a bright sky (clear or light clouds) works with lower exposure times.

Developing uses glass rod spreading much like the initial sensitizer coating, but with only a single pass. Tape one edge of the exposed print to a sheet of plastic, such as a polypropylene “disposable” cutting board. Use a syringe or dropper to deposit a generous bead of potassium iodide solution along the top of the tape, 3g to 5g total solution. Using a single smooth motion, spread the potassium iodide down the print and off the trailing edge. Remove the tape. Wait a few seconds to a minute for the image to develop to the desired darkness, then rinse in a sequence of three pans of water. Take care to avoid touching the image surface until dry. The glue will be very soft and smear easily. Also avoid hard streams of water directly against the image surface. Agitating (horizontal shaking) of the water pans is effective.

Adding a little surfactant (polysorbate 20) to the potassium iodide developer can improve spreading, but also increases the speed of developing. This reduces control, as the print cannot be moved to rinse trays fast enough if it gets too dark. If such control isn’t needed (exposure is accurate), then surfactant is helpful in the developer.

If anyone tries this process, I’d love to hear about your results, and about any variations or improvements.

A few more random notes that didn’t warrant inclusion above:

• Ferric Ammonium Citrate results in lower speed (longer exposure times by 2x to 3x) and lower contrast.
• Lower amounts of Ferric Ammonium Oxalate have a somewhat similar effect, somewhat lower speed and lower contrast. The above recipe has the same molarities for the two iron compounds. Dropping oxalate by 50% hits a cliff where white is no longer attainable. Higher amounts of oxalate increase speed and contrast. Double the amount of oxalate is a good choice for a high-speed high-contrast process.
• Using less of both iron salts (less of the iron-salt solution in the sensitizer) increases speed and reduces maximum optical density (DMax). Lower iron salt concentrations in the sensitizer can be paired with higher potassium iodide concentration in the developer to get back to roughly the same image characteristics.
• Actually, my favorite paper is an old 1980’s bond paper that is unbuffered and has plenty of starch. I didn’t list such above because it isn’t available now. The UniversityProducts paper is good except for missing starch, and works well for conventional cyanotype too.
• Siderotype processes are sensitive from roughly 430nm down into UV. 395nm LEDs work well for exposing contact prints. White LED bulbs are generally fine for “darkroom” illumination.
• Archival life of starch/iodine images is yet to be determined. I’ve had prints lying around for a couple months now without obvious degradation.
• The normal starch/iodine color is purple, resulting from amylose starch. Amylopectin starch forms a brown image with starch, but is less stable. Using glue without starch (on non-starch paper) forms a light brown image, so the glue may include some amylopectin starch. Sweet rice starch is also mostly amylopectin, so can be used for a more brown image. The brown images sometimes age to a darker purple color as the iodine finds more stable homes in any bits of amylose starch present.
• Cooked starch may be used, replacing the glue and suspension of starch powder. When cooked, the starch concentration must be reduced to 3% to keep the sensitizer mixture thin enough for uniform spreading. Less starch reduces DMax, which is why I first tried a glue and starch powder suspension. If anyone has ideas for generating a uniform coating of a high starch concentration, I’d love to hear about it.

 

[NedL]

Elemental iodine is nasty to handle, but I wonder if you could get a more black or very dark purple positive with Lugol's iodine? Very interesting process, thanks for telling us about it!

 

[David Knierim]

There actually isn't any elemental iodine in this process. Development is with potassium iodide solution, which is clear. The remaining ferric ions (ironIII) in the dark image areas oxidize some of the iodide ions to iodine, which immediately complexes with other iodide ions to form (I3)- ions. This mixture of iodine and potassium iodide is exactly what makes up Lugol's Solution (10% potassium iodide and 5% iodine in water). It is K(I3), or K+ ions and (I3)- ions in solution. It is the (I3)- ions that complex with starch to form the purple image.

 

[NedL]

Ah, got it, thanks. So here's another question. Do you think the image formation is because the ferric makes the (I3)- ions available at a particular location to complex with the starch there, or because the ferrous formed with light prevents the complex by altering the starch ( or maybe the glue! ) and/or by making it less permeable? I might play around with the AFC version of this....

 

[David Knierim]

I'm reasonably sure that the image forms as follows: Initially there are no (I3)- ions at all. The developer is just KI, which is K+ ions and I- ions. The remaining Fe+++ ions in unexposed portions of the image oxidize some of the I- ions to I. Every two of those I atoms combine with an I- ion to form an (I3)- complex ion. The (I3)- ions then complex with starch, primarily the amalose (linear coiled) starch, taking up residence in the center of the starch coils, forming the purple colorant.

The oxidation of I- by the reduction of Fe+++ to Fe++ has a 0.2V potential (0.2 electron-volts per atom). With this relatively-weak driving force, it requires reasonably unbound Fe+++ ions. That is why the FeCl3 is necessary. If using just oxalate or citrate, the ferric ions are too tightly complexed to oxidize I-. With oxalate (ferric ammonium oxalate), the process can work, but the developer must be very acidic (3 to 6% hydrochloric acid in the KI), which reduces the strength of the oxalate complexing. In such acidic conditions, however, the I- will oxidize by dissolved oxygen, which makes that version of the process marginally usable.

Yes, the citrate version works well. I've focused more on oxalate due to Mike Ware's recommendations that ferric ammonium oxalate is a more consistent (better characterized) starting material, and that it is more light-sensitive (shorter exposures). The latter I've verified.

Thank you for sharing your interest, especially in the chemistry mechanisms. I'm much more of a scientist than an artist.

David Knierim

 

[NedL]

Thank you very much!

 

[AgX]

David, welcome to Apug!

Could you state in a few words the pros-and-cons of your new process against the classical cyanotype?

 

[David Knierim]

Great question! There is really only one advantage: It's a positive process. That is why I developed it. Making images of the real world directly needs a positive process. Shadows come out dark and bright objects come out light. The image in my original post is not a copy. It is a picture taken using this process. I used a "camera" consisting of a large surplus rear-projection TV lens and a holder for a 120mm x 150mm piece of this paper. For the image above, the "camera" was pointed at a pair of lawn chairs in my back yard on a sunny day for 20 minutes. I have made some similar images with a smaller "camera" using both this process and cyanotype, so will include a comparison below. (Cyanotype negative images do not work well as masters to make cyanotype positives by contact printing, as cyanotypes don't absorb well in the 400nm violet/UV range where cyanotype is sensitive.)

Advantages:
1) It's a positive process (makes positive images). This makes it useful directly as camera film for long-exposure photography. It could presumably be used for contact printing of positive prints, but I have not tried such.
I'll mention a couple other "advantages" just for fun, but none of these are really of significance:
2) If you prefer purple images to cyan ones, this is a process for you It can also directly make brown images, but these aren't as stable. Use non-starch paper and add just white glue (makes a nice brown, but not very dark) or add starch that is mostly amalopectin such as sweet (sticky) rice flour or waxy corn maize.
3) Doesn't contain any cyanide. The cyanide in cyanotype is so tightly bound to iron that it is not dangerous, so this matters only if you have an emotional aversion to the thought of cyanide.
4) Coated paper easily keeps for a month in the dark.

Disadvantages:
1) Developing is more complex, not just a water rinse. It can take a bit of practice to pull a uniform bead of potassium iodide solution across the print in a single smooth motion.
2) Works best with paper that already contains starch, which further limits paper selection over cyanotype. It does work with non-starch papers if starch is added to the sensitizer. This makes the sensitizer thicker, and therefore more tricky to spread evenly.
3) Contrast is higher than would be ideal for photography.
4) DMax (darkness of the darkest areas) is not as high as with a good cyanotype formula, such as Mike Ware's "new cyanotype".
5) Archival life is unknown. I've had some images laying around my house for a couple months that still look fine. I'm aging a few images against the inside of a window that gets morning sunlight. These show some fading after four weeks.
6) It's a new process, so other problems may show up. Cyanotype is very old and much better understood.

Here are direct photographs of lawn chairs using this positive process and cyanotype. The positive ones used non-starch paper (UniversityProducts Perma/Dur Unbuffered) with added white glue and arrow root (left) and corn starch (center). Right image is with Mike Ware's "new cyanotype" formula on the same paper. They were all taken near mid-day, but not at exactly the same time, so shadow angles vary. There is a black drape behind the chairs in these images. (The image in my original post uses a starch-containing paper - some ancient bond paper before buffering became standard - and a newer larger lens - and a different location and angle and backdrop.)

 

[AgX]

Thank You!

 

[David Knierim]

Here's another image, taken indoors with UV (395nm) LED illumination:

The "camera" here consisted of a 3.75 diopter (267mm focal length) 35mm diameter reading-glass lens and a poster-board box. Required 3-hour exposure using a total of about 25 watts of 395nm illumination (65 watts of electrical power input and efficient UV LEDs, EPILeds brand). Contrast is still too high, but I'm out of ideas for improving (reducing) it. I think I'm on to other non-photographic endeavors now, after this summer's obsession with positive siderotype.

 

[David Knierim]

Here's a lower exposure, 2 hours, so not as over-exposed. Under-exposed tends to look worse than over, as too little light quickly makes white areas dark (steep response curve at the white end).

 

[Guugyjus]

Hey it's been a while since anyone has commented here, but if there's anyone still there, could you elaborate on the camera you built with a rear projection TV lens?