Some questions about working on copper

Hi. Thanks for the kind words. I will respond within the text and in red
I’ve searched here on the topic, and also read the “rigid supports” document in Resources (which is a wonderful reference), but I have a few additional questions about working on copper. 
1. I’m using relatively thick copper (14-gauge etching plates) and working fairly small (maybe 11×14″ or so, down to 5×7″). Is it necessary to brace supports of this size, or would it be sufficient to put a lightweight (floating) backing board behind the plate in the frame–perhaps Gatorfoam, or a layer or two of museum board? 
I do not think that bracing would be important or even a good idea in that case. What you propose sounds fine and is exactly what I have done in the past. It may be a good idea to put an acid free interleaf between and board that may be acidic or become acidic to prevent corrosion of the reverse of the copper panel.
2. My process to prepare the surface of the plate: degrease with denatured alcohol, thoroughly abrade the surface with sandpaper or steel wool (with the aim of completely stripping the surface to expose fresh metal), vacuum off any copper dust, then degrease again, making sure that all dust and residue gets removed. Then allow to dry, and prime. Does that sound about right?
That sounds good.
3. How about where the copper is not covered by primer/paint? Like the back and sides? Can I just let that oxidize, or should I seal it with something? Renaissance Wax, maybe? 
I believe that most 17th century oil on copper panels were not coated on the back. There are even a few instances where the work was done on the reverse of an etched intaglio plate. However, if you wanted to avoid corrosion on the back you could spray it with an epoxy coating after degreasing the reverse. Do this before treating the front so that your degreasing and abrasion of the front would remove any accidental overspray that may have gotten onto the face of the panel.
4. I assume any oil-based primer will work? How about an alkyd primer, like Winsor & Newton’s Oil Primer? I have a few small test sheets on which I tested some straight lead carbonate in linseed oil (RGH), a lead painting primer that contains some titanium white and driers (Rublev), and the Winsor & Newton primer. All three dried very quickly (the W&N primer was touch dry in a matter of hours). I’m guessing that the copper is a drying catalyst? The Rublev primer turned very green upon drying; I don’t think that it was a matter of surface prep, because the RGH lead primer is right next to it on the same sheet of copper, and it didn’t change color at all. The Winsor & Newton primer took on a slight green tint, but it’s barely noticeable. Is this sort of thing common? Something to worry about? I don’t think I’d use the Rublev primer, since the color change in that case was considerable. 
For me, priming copper panels has had a bit of voodoo to it. I have primed a group of them at the same time and using the same prep and materials and had one of the bunch turn green as well. These should be cleaned of their ground and the process repeated as such immediate corrosion is a sign of trouble, which would likely only continue if left in place. Copper can catalyze drying oils but I am also guessing that the primers contain enough driers to create a rapidly setting ground. I am unsure why the primers all preformed so differently. Perhaps they employed oils of very different acid numbers. It makes sense that an oil ground would be formulated with an oil of a higher acid number as this tends to make a paint that contains a greater percentage of pigment but one that yellows a bit more strongly. The leaness would be useful in a ground and the yellowing would be far less of an issue than in a paint used for surface effect. Again, all of this is conjecture.
If you have the time, it would be useful to make a test of the available grounds and let them really oxidize for a while and then test for scratching before deciding on your preferred ground. I have always used lead white in linseed oil without a drier. There may be better modern materials for this but I have not personally tested them. I know that there has been a good bit of experimentation with various coatings on aluminum panels. I will send this along to a couple of our other moderators to see if they have anything to add.
5. I skipped the oft-recommended garlic step, just on the basis that I have been able to find a consistent or empirically supported reason for its use. Some sources say that it helps to “etch” the metal (though it is unclear how, since garlic is not acidic). Some sources say that it might serve as a wetting agent (which makes more sense, except for the fact that copper doesn’t seem to need a wetting agent–it takes oil paint really well, with no beading). Some sources say that it helps the paint bind to the surface chemically, rather than just mechanically, but I don’t think that’s correct. Don’t oil films exchange ions with copper? In any event, the idea of putting an aqueous paste between the metal and primer seems like a bad one to me, but perhaps there is a purpose for this step that I haven’t considered?
I do not think that garlic is necessary for a stable panel. Garlic is not very acidic but a cursory internet search shows that it has a pH of around 5.8, which is well below neutral (7). However, I do not believe that etching plays a major role here no matter what. The dried garlic juice does create a slightly pebbled or textured surface that may promote adhesion.
 As to whetting, I have personally seen thinned lead white ground bead on a copper panel that did not have such a layer. This may have just been an isolated occurrence or perhaps that panel was not sufficiently abraded.
I have discussed this idea with a fellow conservator who is more science savvy and they suggested that the garlic may also perform an additional effect; promoting a more stable form of corrosion. The copper is going to corrode to some degree at the interface between the panel and the oil primer (perhaps there are modern materials that would not do this) All cross-sections of paint taken from oil on copper paintings that I have seen exhibit a green layer at the bottom of the oil priming. The sulfur in the thiol groups contained in garlic may promote the creation of more stable corrosion products.
In closing, though, I do not believe that the garlic application is necessary nor does it absolutely make a more stable panel, I do believe that people did believe that it did, and that this was because of some of the aforementioned reasons.
Thanks!
 
-Ben

​Hi –
I only have a few things to add to Brian’s comments.
But first I would want to ask why you are choosing copper in the first place, especially if willing to prime it – thus losing any of the metallic, reflective sheen that painters like Claude Lorrain were attracted to. And beyond that is the simple fact that copper remains a reactive metal and there are definite mechanism of degradation of oil on copper that are a real concern. See The degradation of oil painted copper surfaces by Lydia-Chara Pavlopoulou and David Watkinson for a good and current summary:
https://www.academia.edu/1031822/The_degradation_of_oil_painted_copper_surfaces
Certainly the analysis doesn’t mean that all paintings will fail – only that they are vulnerable given the potentially reactive nature of the oil/copper interface. It would see to me that if wanting something that is more stable and equally rigid and lightweight, that aluminum-composite panels – such as Dibond – might be a better option.
That said, I have certainly done my own share of oil painting on copper – although nearly always painting directly onto the copper – so know the allure. It was also a useful way for me to ‘recycle’ copper etching plates that students in printmaking departments would abandon at the end of the semester. 
As for the other points, I can share a quirky test we have done of soaking bits of copper into various oils (both linseed and safflower) to look at the ability of the oils to become infused with copper ions. Its mentioned in literature as one way to make a fast drying albeit ‘green’ oil. Curiously, only one of the oils turned green – and from what I recall it was the one with a higher acid value. But not sure. So it might be the same with the grounds you are using – one or more of them might have an oil more prone to react with the copper.
And one last thought – in terms of sealing the back – you might take a look at a product called Incralac, which you can find at conservation supply houses like Talas:
http://www.talasonline.com/search?keywords=incralac
It was formulated to protect copper and copper-alloys from corrosion, and while used principally on sculpture and objects, we have used it successfully as a form of clear primer on copper when priming with acrylics. A bit of an off-label use but we feel fairly safe. It might be a useful product to protect the backs of your plates and perhaps even as an isolating coat between the copper and the primers you are using? That latter I don’t know and you would need to test adhesion and make your own assessment.

Hope that helps.

Sarah
Good info and call on the Incralac. That slipped my mind but makes sense in this context. You are completely right about the probable superiority of other substrates. It is true about the allure of the visible copper surface but it is also true that a very high proportion of historical oil-on- copper panels had some sort of ground. This does bring us to one of the fallacies mentioned in the “Resources Section” just because some noted master of old used a material does not ALYWAYS mean that it is still the best tool for the job. As an aside, I always pounced my lead white ground with the palm of my hand (protected by wearing a nitrile glove) very thin so that the sheen of the panel could still be discerned through the thin ground layer.

​Thanks Brian. As for your aside – you clearly are one of those painters ready to pounce!
Ugh – I know.   🙂

There are a couple of comments here.
As to the sandpaper, yes, it is important to abrade the surface to provide tooth and the solvent is used to degrease the panel.
Ben, most copper corrosion products are in the green to bluish green color spectrum. These products range drastically in terms of reactivity, stability, and effect on the copper object. Copper chloride also known as bronze disease is at the bottom. This means that the form of corrosion that occurs between the copper and any oil ground or paint will have an effect on the stablity
Ben, your friend’s supposition about the sulphur promoting a specific chemical reaction which may be more stable, that was precisely what I was alluding to in my initial response. My first draft mentioned sulphur containing thiol groups and the preferential creation of more stable sulphur containing copper compounds but edited that out. Perhaps this limits the degree of the greenish corrosion but on the other hand, sulphur is one of the things known to promote darkening of lead white oil paint so…there are still questions here. Again, I do not believe that garlic juice is a necessary step but its use is interesting.

​Also I am posting this on behalf of conservator Laura Fuster: 

I thought the artist might find of interest the proceedings of the symposium ‘Copper painting and other metal plates: production, degradation and conservation issus’ held last January at the Polytechnical University of Valencia are available at : http://www.lalibreria.upv.es/portalEd/UpvGEStore/products/p_2102-3-1 ‘Copper painting and other metal plates: production, degradation and conservation issues’ ‘La pintura sobre cobre y otras planchas metálicas: producción, degradación y conservación’Edited by L. Fuster, I. Chulià, MF, Sarrió, ML. Vázquez, L. Carlyle, J. Wadum226 pagesFull color, 22 chapeters (18 in English, 3 in Spanish)ISBN: 978-84-1684-696-2

​Based on Ben’s comments, we created a test panel on copper substrate with the following grounds:
1. Rublev Colours Lead Oil Ground
2. Rublev Colours Lead Oil Paste Ground
3. Rublev Colours Lead Alkyd Ground
We also made two samples of lead white dispersed in linseed oil containing paint with free fatty acid (FFA) value of 12 (Sample 1), and another with FFA of 4 (Sample 2).

In addition half of the panel was rubbed with a fresh garlic clove and allowed to dry before applying the grounds and sample paints.

As the grounds dried, ground 1 exhibited a green tint on the area not coated with garlic juice, whereas on the area coated with garlic there was no change in color. Grounds 2 and 3 dried without any change in color on either the uncoated or garlic coated areas.

Paint sample 1 had a slight green tint, but not nearly as strong as the tint of ground 1 on the bare copper surface. Sample 2 exhibited less tinting than sample 1.

Interestingly, we measured the rate of drying of the paint films on the bare copper and the side coated with garlic, and noted that paint on the garlic dried at a much slower rate.

Grounds 2 and 3 have FFA values less than 1, whereas ground 1 has an FFA value greater than 4.

The simple test demonstrates that free fatty acids in linseed oil react with copper to produce a green tint. However, it does not explain the degree tinting in ground 1, since the FFA value of ground 1 is less than sample 1.

We are investigating this issue further to determine the differences, but in the meanwhile, we are recommending the use of ground 2 or 3 as a ground for copper.

As an side, we formulated Rublev Colours Lead Oil Ground for porous substrates, such as wood and canvas, but not for non-ferrous metal substrates such as copper or aluminum.

Thanks George. Very cool/ surprising info and results.. Thanks for the follow up and response.