How to inlay metal powders and make them look (nearly) like real metal
I've been experimenting with the use of metal powders in my woodwork and wood turning for a while now, and feel there is not a lot of info on the topic online that goes beyond the absolute basics. I have discovered ways to improve the process and so wrote this article to share my findings.
This process covers all metal powders (copper, brass, aluminium, pewter, bronze), but is aimed principally at woodworkers.
There are two main techniques to bind the metal in place: epoxy and superglue/cyanacrylate. Superglue will give the closest appearance to real metal, if it is properly finished.
Health and Safety Warning!!
I have worked a lot with epoxy, and was stupid enough not to use gloves in the beginning. As a result, I have sensitised myself to epoxy. This is something that will happen to anyone who uses epoxy carelessly, and you will not un-sensitise yourself. Be warned! Also some superglues (cyanacrylate or CA) can give off serious fumes, so familiarise yourself with the different varieties discussed and work in well ventilated places when needed.
Choose your Metal
I strongly recommend you buy a range of metal powders and set them into a scrap piece of wood to test them out, rather than guessing. I have been surprised in the past by which looked the best against different woods. Additionally, if you need a specific colour, you can mix different metals together, as I did with the below pewter coloured pen kit. No need to consider grain size, pretty much all the metal powders on Amazon or eBay will be suitable.
- "Alumlite Gold Powder" is more expensive / oz but looks like gold!
- "Brass" and "Bronze" are alloys of various metals and the colour can vary between suppliers. Generally speaking, brass is copper and zinc whereas bronze is copper and tin. Also some suppliers might sell slightly oxidised products e.g. giving the bronze a slight green tint.
- I have not tried the stainless steel powder yet but it looks like a good option, potentially an improvement on aluminium if you need extra shinyness!
The most important thing is to maximise the metal content, and minimise epoxy. Epoxy resin comes in wide a variety of viscosities , and to make this work we want the epoxy to be as thin as possible. Naturally, it also needs to be clear. One obvious option that people often find are "penetrating epoxies", however for the most part these are naturally low viscosity epoxies that have been heavily thinned with alcohol, acetone or some other thinner (50%+ in some cases). Using this option you will be able to add even more metal, but the resulting matrix will not be as strong and is likely to penetrate the surrounding wood. Note there are a lot of "low viscosity" epoxies that cheat by adding thinner - so always check the Safety Data Sheet (SDS). A better alternative would be a product without thinners, such as below. You can always add acetone later if you like.
This epoxy has viscosity of 600 - 650 cP, which is as low as I can find. To put this into context, I have some coating resin from another company that has a viscocity of 12000 – 15000 cP. Huge difference! The thinner the epoxy, the more metal can be added before it becomes unmanageably thick. However also remember that thinner epoxies can more easily penetrate the skin, so more care must be taken when working with them. The basic process is to mix up the epoxy and then slowly add more and more metal powder, mixing with each addition until you have a thick paste. At this point you can literally smear and press it into the gaps you want to fill, and if they are vertical then duct tape over the top to stop the mix from slumping out (note that if you do this, you might be left with some voids that need re-filling). With very thin epoxies, you'll be surprised at how much metal you can add before it gets to this stage. Of course you could also stop adding metal before it becomes a paste and is still slightly runny – but again, our goal is to get as much metal in there as possible. Unlike superglues, epoxy without thinners is unlikely to significantly penetrate the wood, due to its much higher viscosity.
I prefer epoxies with longer working times. You can use basic cheap 5 minute epoxy, and quickly mix it with metal, but it wont look very nice. You'd be unlikely to get much if any shine, more likely it will just look like the epoxy was dyed a slightly muddy colour. Also when mixing epoxies with fine powders, the powders act a a catalyst which speeds up the curing process, so with 5 minute epoxies you'll probably only actually get a minute or two - if that. With resins like the above, I usually go with a cure time of a few hours - in this case that would mean the "slow" hardener.
Superglue / Cyanacrylate:
With superglues, again we need product that is very thin, but this time the reason is so that it soaks into the metal powder as deeply and as quickly as possible. Superglues vary a huge amount and we really need one that is suitable for our application - I have linked the product I recommend. The absolute thinnest superglues you can buy have a viscocity of 2-3 cP, which is slightly higher than water (0.9 cP), but less viscous than milk (3 cP)! In this case, the product I have linked is a little bit higher at 5cP, but is still way better because it is designed for infusion and, as such, sets a little slower. This allows it so penetrate deeper before setting hard.
I did a few tests on this, using the linked Starbond infiltrant cyanoacrylate and comparing to a standard 2-3cp super thin cryano. The results were really interesting, lets take a look at them:
The three different techniques showed in this image were:
- 1a and 1b: The holes were filled entirely with brass powder and then the surface flooded with superglue. Image show pure penetration depth, which for 1a was around 2mm or 5/64", for 1b it was more like 1.5mm or just 1/16. However, although more difficult to see the thickness of 1b was less consistent, with the depth being slightly less behind where the cut was taken (i.e. what is being seen here is a slight overhang)
- 1b and 2b: Holes were roughly half filled, and then flooded so that there was a small amount of excess fluid above the metal. More metal was then sprinkled on top to soak up the excess and then extra to create a thin dry layer. Superglue was then applied again in the same manner, with enough to soak in and create excess, and more powder added on top, repeating the process until the top of the hole was reached.
- 1c and 2c: Holes were roughly half filled, then flooded with lots of superglue, roughly enough to reach the top of the hole. Large amount of metal powder was then dumped in, in one go, so that it reached the top.
This was a really interesting experiment! A few observations:
- The biggest surprise was the difference in reflectivity. The image above is actually two separate images, showing each side of test piece of wood, each angled in such a way as to show maximum reflectivity of the metal. As you can see, the infusion resin has way more shine to it, indicating a higher metal density. If you look at 1a and 2a, then 1a is a little more shiny but not a great deal in it. Similarly if you look at 2b and 2c, at the bottom of the metal you can see where the initial penetration took place, again it is almost as shiny as the infusion side. However the higher sections of 2b and 2c are a mess, I can only guess this is because the cyano set so fast the metal didnt have time to settle. The small black grains are therefore voids. Conversely, because the infusion cyano took just a little longer to set, the metal was able to full settle. This result is not too surprising on reflection, although the fact that infusion cyano produces slightly more shine even in pure penetration of a layer of metal (1a and 2a) is unexpected.
- This different in set times also explains the large voids in 2b and especially 2c. I suspect the metal was forced to set in clumps, which could not settle. These voids are also seen in 1c and to a lesser degree in 1b, though they are less severe, with the voids in 1c being large and well defined - if you were to cut down to this level when sanding, they would be easy to fill and set again.
- Penetration depth is better for the infusion cyano, but it is still relatively shallow. For filling deeper holes/cracks, where there is the possibility of having the cut down a significant way e.g. with sanding, the best technique is as described in 1b and 2b above.
Next, lets look at the differences with aluminium and copper:
More really interesting results with this test. Here, holes were filled entirely with metal power, and flooded above on the surface with our two different varieties of thin cyano (same as 1a and 2a above). I used aluminium powder, and copper powder. The differences are striking:
- With aluminium, I was shocked that I had to keep adding cyano, and it kept soaking in. It is possible I could have added more, or done so slightly faster, and gotten full penetration depth for both holes. I suspect this is because the aluminium is lighter, so the cyano is able to more easily wick between the grains. One thing to note here - I actually cut the test piece open about 5-10 minutes after I had finished applying the cyano, and found that the inside of the larger aluminium infusion cyano hole was still wet. You can see in the picture I had to come back later, after it had set, and cut it back again - this is probably why there are small black voids near the bottom of each hole. So if you are using aluminium and doing deep fills, I recommend leaving as long as possible before cutting down into it.
- Penetration depth with infusion impregnated copper was around 1.5mm or 1/16" on the wider 8mm or 1/3" hole, and closer to 1mm on the smaller hole - about 1/32" to 3/64". Copper depth on the standard super thin cyano side was just 0.75mm and 0.5mm, on the larger and smaller holes respectively, as measured with a vernier calliper. This is a bit over 1/64" and a bit under 1/32" respectively. So in this test, when impregnating copper, the infusion cyano went twice as deep as the normal super thin.
- Unlike the brass image above, these pictures are "as cut" - without any sanding or polishing. However, the infusion side is once again slightly brighter than the normal cyano side.
Honestly, I was a bit shocked by the pathetic penetration depth of copper on the right. This again really emphasises that the best cyanoacrylate to use, is one that has been designed for impregnation. As an aside, the slower set time is due to it having a different chemical composition, rather than having agents added to slow the process down. This is also why it is low/no odour. Again, I have linked the Starbond cyano above - I am not aware of any other companies selling this type of cyanoacrylate, at least not at a reasonable price.
I did not test with medium or even thick varieties of cyanoacrylate, from the above pictures I think you can guess why! I really do not advise using them unless for some reason you have no choice e.g. cannot allow any wicking into the surrounding wood, and even then you should prepare for a penetration depth in the micron range.
Wrapping up on this section, two other things to consider regarding viscosity. First, in general, the thinner the superglue the faster it will set – beware!! And when it sets, it will give off heat. Dab it up with tissue paper, or flood sawdust with superglue, and it will set super fast, creating enough heat to produce little tendrils of extremely nasty white smoke. Now I don't know how dangerous this is, but when superglue is burned it gives off cyanide...and the smoke is caused by overheating as it sets so... do the math! Secondly, thin superglues may be capable of penetrating the wood itself. Of course this depends on the wood you are using – none of them will get inside of dense, oily woods like African Blackwood. However with lighter or even spalted timbers, superglue penetration is something that must be considered. Some people oil or wax the surface around the spot being inlayed first, however I prefer to work with the superglue. If we accept that it can be a part of the finish, then we can get even better results – see section below on finishing.
Many people are surprised by how strong superglue can be. Cyanoacrylates are actually one of the favourite finishes for pen turners – just superglue, usually low or medium viscosity and sometimes used with accelerators so that multiple layers can be put down rapidly one after another. It is very tough and hard wearing in and of itself, and can be easily polished to a glassy finish. Metal powders flooded with superglue are similarly strong but in extreme cases could become brittle. In general I see no reason to only limit the use of superglues and metal to small inlays - however! If the wood is not properly seasoned, then at some point in the future it might move enough to crack the metal and cyano section. I have not seen this myself but I have read others claiming this can happen. Epoxy and metal will not produce as nice a finish, but I would expect it to be less likely to develop cracks in such cases.
The most important part. Finishing up to 1200 sandpaper is a great start, and if done right you can get a degree of shine. However for best results you'll need to use the right product. Below I'll look at three different options and how to avoid any problems that might arise:
- Best option - Micro-Mesh. These are soft and very cleverly made pads that go up the way up to a crazy 12,000 grit. They will leave the metal looking nice and shiny and don't suffer from any of the issues of burnishing pastes etc. I absolutely recommend these as the best choice, although they are a bit pricey - check the link for current price on Amazon. Fortunately because of the way they are made, they will last longer than you might expect.
- Metal burnishing pastes or liquids will absolutely do a great job of bringing the metal power to a shine, but are also likely to turn black on contact with some metals and stain the surrounding wood. Not great. Test to see the effects on the metal you are using. It might be OK with aluminium, I don't know, but with brass it is a real problem. Also not a problem if you are using a super dense and oily wood, but otherwise you'll need a workaround. The solution to this is to flood the surface with cyanoacrylate, and then cut (sand) back to the metal in just that area. You can then burnish it either by hand or using a mop, for larger areas. When the paste turns black it wont be able to penetrate the surrounding wood, and can just be wiped off without staining. Of course, this wont work too well if you want to apply a natural finish like wax or oil - in that case, micro mesh or cerium oxide are better options. Aside, I love Mirkas Polarshine products, and use them all the time to quickly bring a shine to artificial finishes e.g. cyanoacrylate finished pens. It seems expensive, but 1 litre lasts a long time.
Bonus Idea - Rot Hardener?
Another option I have not experimented with would be to use a wet rot wood hardener like the below:
Rot hardeners are designed to impregnate rotten timber and set hard after a couple of hours. They have a high acetone content of over 50% (acetone viscosity 0.4 cP) and so potentially might be even thinner than water - unfortunately I cannot find any data on this. Of course, the long set time and low viscosity would cause the hardener to penetrate any surrounding wood (if it is not incredibly dense like lignum vitae), or with oily woods it might even start dissolving into the oil (potentially sticky...). Also I have no idea how strong the resulting metal/hardener mix would be. The benefit of this method would be an extreme penetration depth. And if you are using a softer wood, you could just accept the hardener penetration and roll with it! After all, it can only produce harder and denser wood that absolutely will never rot! Who knows, you might even be able to sand and polish the surfaces straight to a matt or gloss condition without adding any additional finishes. Worth playing with I think.
OK that's all! If you have any comments or suggestions, please leave them below or get in touch with the contact form. I hope you have found this article useful!