All incredited work mentioned in this note took place at IPI.
Initial questions were raised in this discussion. I had brought up some issues based on an old research paper that I'd found, and on dodgy memories of some other things I'd read a long time ago. But I couldn't answer a lot of the questions being raised. Happily, someone put me in touch with Douglas W. Nishimura, who's a Senior Research Scientist with the Image Permanence Institute at Rochester Institute of Technology. He may know more about this issue than anybody. This is an email that he sent to me, reproduced here with permission.
The reader's digest version is this: selenium toning is not very effective at protecting silver prints from deterioration, unless toning is carried out to an extreme degree. Gold toning is better, working at much lower doses. Sulfide toning is better still. Most surprising of all, incomplete washing of the prints, leading to trace amounts of thiosulfate in the paper, may actually improve permanence. But there are many caveats and twists and turns. It's worth reading the full text if it doesn't make your eyes glaze over. Paul R.
I suspect that in part, your question is in regard to work that we did her at the Image Permanence Institute back in the late 1980's and early 1990's. We were working on a project funded by the National Endowment for the Humanities to look at the use of selenium toner to improve the stability of silver-gelatin microfilm. Historically, there were a number of classic papers written on the topic starting in the the early 1960s with regard to gold toning (there is a connect that we'll get to later.)
1965. Henn, R. W.; Wiest, D. G.; Mack, B. D. "Microscopic Spots in Processed Microfilm: the Effect of Iodide." Photographic Science and Engineering. 9(3):167-173.
While iodide is not toning, it is an important article with regard to the science of toning and permanence.
1965. Henn, R. W.; Mack, B. C. "A Gold Protective Treatment for Microfilm." Photographic Science and Engineering. 9(6): 378-384.
1966. Henn, R. W.; Wiest, D. G. "Properties of Gold-Treated Microfilm Images." Photographic Science and Engineering. 10(1): 15-22.
1984. Lee, W. E.; Wood, B.; Drago, F. J. "Toner treatments for photographic Images to Enghance Image Stability." Journal of Imaging Technology, 10(3): 119-126.
1984. Brandt, E. S. "Mechanistic studies of Silver Image Stability. 1: Redox Chemistry of Oxygen and Hydrogen Peroxide at Clean and at Adsorbate-Covered Silver Electrodes." Photographic Science and Engineering, 28(1): 1-12.
1984. Brandt, E. S. "Mechanistic Studies of Silver Image Stability. 2: Iodide Adsorption on Silver in the Presence of Thiosulfate and the Influence of Adsorbed Iodide on the Catalytic Properties of Silver Toward Hydrogen Peroxide." Photographic Science and Engineering, 28(1):13-19.
1987. Brandt, E. S. "Mechanistic Studies of Image Stability. 3. "Oxidation of Silver from the Vantage Point of Corrosion Theory." Journal of Imaging Science 31(5): 199-207.
The first problem is the question of what the primary cause of instability in silver. Because many photographers were intentionally not washing well or were messing with the fixing bath when processing albumen prints, everyone hears that bad processing is the root of all deterioration. For the albumen people, they initially got a very neutral looking photograph because the sulfur ions adsorbed to the colloidal silver image particles drove the absorption peak for the colloidal silver from the violet region down to the green part of the spectrum. The effect is pleasant for a short time and if you only have albumen (or other POP process from which to make prints) and you really want a neutral black and white image (such as we got with the modern silver-gelatin DOP) then adding sulfur to the colloidal silver image did the trick. However, the large excess of sulfur present had consequences and as the colloidal silver turned into colloidal silver sulfide, the absorption peak dropped in to the UV region (causing the image to turn to a pale yellow). The drop in optical density combined with the eye's general lack of sensitivity to yellow caused albumen prints to all but disappear completely. The interesting thing is the fading committee report included problems such as air pollutants, bad adhesives, and high humidity (all of which result in the oxidation of silver rather than sulfiding.) All of this issues were ignored and people only focussed on the poor processing. Hence, every photographer hears that bad processing is problem. You will find that finding examples of poor processing are not so easy to find. In fact, the major cause of deterioration is oxidation. As Fuji described it in 1982, the deterioration process is rather cyclic. The image silver particle is exposed to oxidant which causes silver ions to break away from the image particle. These silver ions are mobile and migrate through the emulsion with spherical symmetry (no preferred direction of movement.) The ions may run into something like a sulfur source or halide source and precipitate in the emulsion as an insoluble, immobile particle of a silver compound (a dead end event removing that silver ion from the rest of the system.) The silver ion may find an electron source and reduce back into a particle of colloidal silver. The colloidal silver may appear as silver mirroring (if it happens to reduce back to metallic silver at the surface of the emulsion), as orange, red, or yellow colloidal silver (most obvious in the midtones and highlights), or coallesce as a redox blemish (also known as measles, red spots, or microspots.) When I worked for Klaus Hendriks, we could see that no matter how cross-sections were cut in deteriorated photographs, the result was always a circular halo of colloidal silver particles around the former image particle, thus demonstrating that silver migration has no preferred direction and is spherically symmetrical. In relation to the Fuji work, is another surprise for most people. Fuji found that a small amount of residual fixer in their samples actually made them more resistant to oxidative attack. They weren't the first people to discover this. George Eaton told us that back in the 1960s, they found the same thing at Kodak, but didn't know how to tell people to wash well, but not too well. As a result, they only told people about washing well (since failure to do so would directly result in damage to the photograph; whereas "over" washing would only cause trouble if the silver was exposed to oxidants (which really were everywhere anyway.) In hind-sight I agree with their decision. We observed the same effect in 1987 as has Ilford and Agfa. So the primary purpose of toning for permanence is to prevent oxidation.
There is nothing magic about microfilm silver (except that it tends to be very fine grained), but the chemistry that governs what happens to microfilm silver, also applies to paper (and we'll get into that later.)
Microfilm was the first photographic material considered to have enough value that its deterioration warranted reseach funding and interest. One thing that was observed with microfilm was that older fixer (so called, "seasoned" fixer) resulted in more stable film than film processed in new fresh fixer. It was determined the iodide in the fixer (that was coming out of the fixed film) was producing the higher stability so microfilm fixers were sold with potassium iodide added. Why the iodide was helping the stability of the film was not answered until Steve Brandt's papers in the 1980s. He pointed out that hydrogen peroxide (a very common atmospheric oxidant found in storage and display environments) was catalytically decomposed on the surface of silver. During the process, the peroxide acts as both an oxidizing agent and reducing agent causing the silver to oxidize and later reduce back to metallic silver. (Migrating in between.) Brandt found that adsorbed iodide ions on the silver surface poisoned the surface as a catalyst and prevented the decomposition of the peroxide. Thermodynamically, iodide and silver have a stronger drive towards forming silver iodide than pure silver does to forming silver ion. (The Gibbs free energy change for the have cell reactions are -14.69 kJ per K per mole and +77.16 kJ per K per mole. Thermodynamics says that the Gibbs free energy must decrease for spontaneous reactions (the change must be negative.) However, although the oxidation of silver and iodide to silver iodide is thermodynamically preferred, it is kinetically very very slow-- slow enough that chemists call the system stable. Thermodynamics tells us which direction reactions go (and why we can readily change high octane gasoline and oxygen quite readily into carbon dioxide and water, but can't convert carbon dioxide and water into high octane gasoline and oxygen very easily.) Given enough time, silver with adsorbed iodide ions will change into silver iodide, but kinetics tells us that it will be a very long and slow process. In theory, sulfide ions should do the same thing, but Brandt had a feeling that sulfide ions were less stable as ions than iodide ions were so he recommended iodide over sulfide. Selenium and sulfur are both in the same family and should have similar chemistries, although selenium is one period lower on the table and it has d-orbitals that allow it a wider range of possible oxidation states.)
Lee, Wood, and Drago published the first paper dealing with the stability properties of a variety of toned images (including selenium.) They found that selenium acted very well as a protective treatment and in theory we would've (and did) guess that that would be the case. If sulfur works, so should selenium. When we studied selenium toning for the preservation of microfilm, we found that selenium worked pretty well for high density areas (shadows), but failed in the highlights and mid-tones (where we tend to see colloidal silver formation.) It apparently just doesn't convert the mid-tones and highlights all that well. When we brought this to Kodak, they tried to dig back in their records for formulations and chemical sources. They had run into something similar with regard to stabilization processed photographs. Prints produced by users in the field had unusually stable prints (quite resistant to oxidants), but when they tested the processors in the lab, they didn't find any high level of resistance to oxidants. It drove them crazy for some time before they realized that they weren't using the same chemicals in each test. If they used the chemicals that consumers were using, then they found the high level of stability. The difference was that consumers were using processing chemicals made from what a chemist would call practical grade chemicals (Kodak referred to "tank car quality" while experiments in the lab were being done using analytical reagent grade chemicals (very high purity.) It turned out that the sodium thiocyanate made from practical grade chemicals was contaminated with a number of active sulfur compounds while the high purity analytical grade chemicals weren't. Photographic chemists distinguish sulfur compounds that contain sulfur with different oxidation numbers. "Active" sulfur compounds contain sulfur atoms with an oxidation number of -2. All other sulfur compounds are inactive. For us, the two main rules are that oxygen always has an oxidation number of -2 and one must maintain the charge on the entire radical species. So sulfite ion has a total charge of -2. It has 1 sulfur and 3 oxygens. Each oxygen has an oxidation number of -2 (for a total of -6). Since the entire sulfite radical (or ion) has a charge of -2, then the sulfur must have an oxidation number of +4. Similarly, with sulfate ion, the sulfur has an oxidation number of +6. So neither sulfite nor sulfate are active species. Soaking film or prints in either sulfite or sulfate does not impart any degree of stability, nor if left long enough, will the silver image sulfide. If we follow the rules for oxidation number with thiourea, we find that the lone sulfur in this compound has an oxidation number of -2 so thiourea is active and we found that thiourea solutions could sulfide silver and impart a certain degree of stability to silver images. The sulfur in thiocyanate is not -2 so thiocyanate is not active. Selenium toner can be made by combining selenium with sodium sulfide to make sodium selenosulfide. The other way to do it is to combine selenium with sodium sulfite to produce sodium slenosulfate. Effectively sulfur combined with sodium sulfite makes sodium thiosulfate which is why fixer is stabilized by additing sodium sulfite. The excess sulfite tends to minimize the decomposition of thiosulfate to sulfite and sulfur. It was possible that selenium toner tested by Bard et al was an old bottle possibly contaminated with active sulfur compounds. However, we obtained a bottle of selenium toner from the same approximate time and tested it and got no better results. Kodak tried similar tests out and also found that the selenium wasn't working as well as it apparently did in the early 1980s and no one understands why.
However, we did find it interesting that the long-known solution of gold chloride and sodium thiocyanate (known as Kodak GP-1) worked pretty well. It lays down a pretty even amount of gold all over and if you remove all of the remaining silver, you're left with a very low density image made of gold. Henn and Mack, in the 1960s worked on a better formula that was to become known as GP-2. It consisted primarily of gold chloride and thiourea with a few other salts added. This formula really worked well on films and prints. Henn and Mack observed that as they increased the gold content of the solution, they observed no increase in the "protectiveness" of the solution, but as they increased the thiourea content, the treated film stability went way up. They didn't pursue the problem and simply decided to use five times as much thiourea as gold. So here we have pretty strong evidence that the thiourea complexing agent (that kept gold from falling out of solution) was contributing to the stabilizing effect of the toner. Gold solutions not containing active sulfur (such as GP-1) had no such effect if we varied the concentration of the complexing agent. Interesting.
Combine this observation with the observation that a small amount of residual thiosulfate (hypo) is good for the stability of silver photographs and you're forced to draw some conclusions about sulfur. Thiosulfate has two sulfur atoms, one with an oxidation number of +6 and the other with an oxidation number of -2 so thiosulfate is an active sulfur compound. Just a tiny amount of sulfur "dusted" on the surface of silver image particles can make the silver fairly stable against hydrogen peroxide oxidation, but for things like ozone and nitrogen oxides, actual conversion to silver sulfide is required. Selenium will work, but a heavier dose of toner is required in order to ensure that the mid-tones and high-lights are adequately protected.
How about papers? Back in 1992 we had a Swiss Graduate student here doing his MFA here. The title of his thesis was On Black-And-White Paper Image-Stability Enhancement: Effectivenss of Toning Treatments on Silver Gelatin Prints Determined by the Hydrogen Peroxide Fuming Test. This 411-page thesis reached the same conclusions that we had regarding toners with film. If you search for the thesis through the library, the author is Christopher Gmuender.