Notes on how -not- to test ferric oxalate purity.
Originally Posted by sanking
Kevin Sullivan here, from Bostick & Sullivan. I generally try to stay above the fray and refrain from comments. But this thread bugs me because of the general lack of understanding about ferric oxalate as a chemical, and the false info being propagated.
This pertains particularly to Jeffrey Mathias' well intentioned, but completely off base testing of ferric oxalate purity. His test is significantly flawed. His methodology is permanently crippled. Yet it looks sophisticated enough to fool some folks. This is very bad science.
Ferric oxalate is a non-stoichiometric compound, meaning that the molecular structure is undefined. Think of ferric oxalate as a three armed octopus with a suction cup at the end of each arm. When you make a batch of FO, these octopi arms can stick together in any combination and size. They can stick to each other, they can stick to other FO molecules, or they can be free floating. This will effect the size and shape of the molecules, but notice that it is all still 100% FO no matter how jumbled and dense their structure becomes.
This is what Dr Mike Ware means when he says FO is an ill defined compound, it has no regular shape and size to its structure, it is entirely random.
Stoichiometric compounds are things we are commonly familiar with that have a fixed atomic structure and relationship to themselves. Sodium chloride, potassium dichromate, and most familiar chemicals are stoichiometric. They just don't have multiple sticky arms that can link in an infinite variety of ways. So their density stays stable because they can only interact in limited, known ways. If you took 10g pure sodium chloride and dissolved it in 100ml water, the specific gravity would be readily predictable, and any extra contamination or impurities would shift this specific gravity measurement.
Unfortunately, we don't know at any given time how the FO has decided to array itself.
For a simplistic analogy, picture each FO basic unit as a strip of velcro. We fill up two buckets with velcro strips. I reach into one bucket and pull out a single strip of velcro and it weighs 50g. Jeffrey Mathias reaches into a second bucket and pulls out a giant wad of velcro strips all stuck together in a ball. His lump weighs 500g. Jeffrey then ratios the weight of my velcro to the weight of his velcro and determines "Kevin's velcro is only 10% velcro, where as my lump is true 100% velcro." When, to any objective observer both pieces are 100% pure velcro, just different sized clumps.
The math that Mathias relies upon only works with stoichiometric (fixed) compounds. If you were doing this test with a dichromate, then it would be a valid measure of purity. His testing is meaningless to non-stoichiometric materials.
To jump into Sandy King's world, let's look at gelatin for a second. Gelatin is also capable of chaining together into bigger or smaller molecules, often measured as bloom (hardness). So, what's purer, 100% pure 75 bloom gelatin or 100% pure 250 bloom gelatin? If you use Jeffrey's methodology you will calculate out that 75 bloom gelatin is only ~25% pure/ 75% contamination, if based on the density of 250 bloom gelatin. We can immediately see how phoney that is.
The major flaw is that specific gravity is not an indicator of purity for these materials. It just won't work that way. You can't even use it as a relative measure, i.e. "this FO is purer that that FO based on specific gravity...."
So we have a guy (JM) who has a limited understanding of what he's measuring running around making claims like "B&S FO is only 91% pure" and confusing a couple of folks with his elaborate charts and graphs and math. Clearly he's confused himself. He makes statements like 'All weight measured to within 0.01 grams." in an attempt to bolster his claims of precision. Sorry, but no matter how accurately you weigh your samples, you have an inherently flawed result which proves nothing about purity.
Folks make much ado about Dick Stevens' measurement of specific gravity from the Kallitype book, and use it as a baseline comparison for measuring other FOs. But the specific gravity of Dick Stevens' FO is again a red herring. It tells us nothing about the purity, even if we assume Stevens made absolutely perfect 100% pure FO, his specific gravity is not a useful number for comparing against other FOs as a determination of purity. Again, this would work with a stoichiometric compound but not FO.
Mathias has several other tests in his chapter, all equally as flawed. His determination of purity based on dissolving time for the two powders is so obviously wrong I'll only mention it in passing. His other major test is an "optimized" printing eyeball observation test, this test is just plain silly. I could write several more pages about this but suffice it to say they have holes in the methodology so large I could park the goodyear blimp inside the empty spaces.
I hope that this straightens out some of the misperceptions about FO and its purity.
Also, Eric Neilsen casually throws out the idea that John Rudiak and David Michael Kennedy somehow struggled to improve the quality of B&S ferric oxalate, all to no avail. This is because DMK and Rudiak also misunderstood the very nature of FO. They confused the use of oxalic acid in manufacturing the FO, with the addition of oxalic acid to pure FO when mixing it for use in printing.
As many people have noted, adding oxalic acid to FO solution turns the solution from a dusty yellow brown color to a bright translucent green. Rudiak and DMK believed that they were neutralizing excess ferric nitrate with the addition of oxalic acid, and that "pure" FO would be a vivid lime green color. They were wrong. Pure FO powder is the familiar yellow brown dust. Adding oxalic acid to the FO powder starts to stoichiometricize the material, and it becomes greener with the addition of extra "impurities", in this case oxalic acid. We told them add as much oxalic acid as they like if it seems to help their particular printing style, but we weren't going to make "bad" FO just because they didn't understand the chemistry. Rudiak eventually got it, as he had a degree in chemistry from RISD. I have no idea what Kennedy thinks, he was never very technical and Rudiak was the brains in this situation. But I don't think it excuses Eric Neilsen from posting third-hand misinformation which slights B&S without telling the whole truth.
In closing, I'd like to say that I have spoken with Jeffrey Mathias by telephone and explained to him the flaws of his study. At times he was speechless and could offer -no- valid defense of his techniques (especially his "optimized" print test). He has not corrected the major problems with his study even though they are quite obvious, once understood. So Jeffrey, for all his bluster about science, is not a scientist. A true scientist would acknowledge the flaws of his study and post information about the evidence that might disprove his research. Instead he hides behind a wall of complicated but useless mathemetics, hoping to fool a few more unsuspecting chumps into drinking his particular flavor of poison kool-aid.
Jeffrey lists on his website several empirical tests for determining the purity of FO (using silver nitrate and pot. ferricyanide, for instance.) He readily admitted to me that B&S FO powder will easily pass all of those standard tests showing a high level of purity. When asked how, then, that the B&S powder passes all easily observable traditional tests of purity, yet we end up only ~90% pure in his crackpot specific gravity test, Jeffrey claimed that Bostick & Sullivan must have an unknown, untraceable impurity which we can mask so that it doesn't show up in his tests. Yes, seriously, according to Mathias, B&S has a special "stealth" chemical we use to cut the FO that is untraceable by modern science.
B&S has spent 25+ years refining our ferric oxalate manufacturing technique, including quite a few unique, unpublished, methods for guaranteeing absolute purity. We are quite happy with the results and our customers get great prints. I'm sure that the materials provided by Artcraft and PhotoFormulary are also quite acceptable and make fine prints. But I'm not going to sit around and let a bunch of bad scientists with bad technique tell me that our fine product is 10% contaminated. These folks are either grossly uninformed or intentionally misleading people.
Sorry for the long rant. Thanks for your time.
Bostick & Sullivan, Inc.