Annealing copper

This is a question for the material scientists in the Forum.
When annealing copper, the internet recommends recommends the heated wire (about 1/2 the melting point) be quench in water to reduce the oxidation. For those that may not be as familiar with this part of material science, the oxidation is the dark color that flakes off as dust when bending the wire, as it is apply it to a tree. This dust is likely not something you want to expose you lungs to for an extended time. We need annealed wire so it will be soft enough to apply and hold the branch in the desired position but rapid cooling is routinely used to harden metals (tempering steel). So slow cooling would equate with softer wire but more oxidation, fast cooling results in less oxidation but harder wire.
Does tempering of copper have to be so fast that the rate of change by quenching in water is too slow to cause it to harden?
Why then do we occasionally cut a piece of wire to use and discover that it’s as stiff as a piece of dead wood but the rest of the coil may be fine?
What if the wire is cooled slower in a oxygen reduction environment as you wound a piece of Raku pottery?
I’m just thinking of what to do with that old 100,000 BTU burner that I used to fry turkeys with before the low fat diet happened.
Oops, I mean “Cheers”


How a metal behaves during quenching depends upon the type of phase change that it undergoes (it any) during cooling. In the case of the fairly pure copper that we use for wiring our trees it does not undergo a phase change - it is a face-centered-cubic (FCC) metal at high temperature and an FCC metal at room temperature. As a result, quenching the copper has no impact on its stiffness, but the quick change in size (it shrinks more than the oxide layer) causes the oxide to flake off. The key to annealing is to get it hot enough, long enough so that new more perfect grains (crystals) form and grow to an optimal size. Too small and the wire is stiff because there are too many grain boundaries. Too large and the grains span the wire and the properties are inconsistent along the length. I have not researched it, but my educated guess is that the optimal time/temperature will depend upon the wire size, particularly in comparing 4 to 18 gauge.

Steel (an iron-carbon alloy) undergoes a fairly dramatic phase change upon cooling. At high temperature iron is a body-centered-cubic (BCC) metal while it is FCC at room temperature (there is an additional very high temperature FCC phase that is not part of this discussion). The carbon will fit into the gaps in the BCC structure, but not the FCC structure. During slow cooling (traditional annealing) the carbon can move around and form iron-carbide (Fe3C) and let the iron forms its FCC structure. During quenching the carbon cannot move around and the BCC structure distorts into a body-centered-tetragonal (BCT) structure which is very hard and brittle. The steel is then tempered at a moderate temperature to allow the carbon to form very small bits of Fe3C with small FCC iron grains to give a hard, but strong metal.

Now to make things more complicated we have the aluminum alloys that contain a few percent of different alloying elements. At high temperature they form an FCC solid solution that can be quenched with no phase change. They are then reheated to a moderate temperature (tempered) to cause the alloying elements and aluminum to form compounds just like we saw for steel with Fe3C, but these compounds are a bit softer.

All of this effort is to control the motion of dislocations (imperfections) in the metal structure. These dislocations are like putting a bump into the edge of a large rug and pushing the bump across the rug which is far easier than trying to pull the entire rug across the floor at once. It is the same in deforming a metal - the dislocation is far easier to move than trying to move two pieces of the metal past each other at one time. During bending or other cold working (like drawing the wire to diameter) lots of dislocations are put into the metal. This is like having lots of cars on the road in the city center with the stop lights out - grid lock develops and it becomes harder and harder for the cars to move - same with the dislocations. In copper the dislocations become tightly entangled fairly quickly so the metal work hardens quickly and becomes much stiffer. In aluminum they do not entangle as quickly and can even self adjust a bit at room temperature due to the lower melting point of aluminum. As a result, aluminum does not work harden as much or as quickly. The high temperature annealing allows the formation and growth of new, more perfect grains without the dislocations. A bit like waiting until 9 pm to go home to avoid the gridlock.


Brilliant examples !
Material science is so cool.
So, how do we apply this to lower oxidation on the wire? I recently had to use some 4 gauge and it literally looked like it was smoking as I worked it. Your guess about time and temp makes sense. Lower temp for a Longer time for consistent crystal formation?

1 Like

I have only the practical experience of trial and error to draw upon, but I find the opposite to be true. The longer the copper sits at temp The thicker the layer of oxidization becomes. It gets flakey and produces a rough surface that is terrible on softer bark. Also, the more difficult it is to strike off with quenching. The “smoking” effect is probably the result of wire soaking at temp for too long.
Additionally, I find there to be a range of color tones produced on the surface of different spools of wire that are predictive of stiffness. I assume this is due to different compounds or impurities present in the metal and unfortunately does not seem to be discernible until the wire is annealed. Copper that produces purple tones tends to be stiffer than copper that produces redish tones. Orangish tones seem to be the softest.
Here’s my process. I bring the oven up to 950 and then load the wire, I try to keep the wire hot for as little time as possible to still achieve maximum softness. This takes about 15 min for everything up to 4 gauge. I give that an extra couple of min.
Then I use a hook to pull the spools out and drop them as fast as possible into a full bucket of water. At the end the water is black and the copper is pretty clean.
This is just what I have found to work. I’m sure there are more precise ways and would love to hear about those if anyone has some tips.


Not anything I’m going to be able to try in my Num Noms Mini Baking Oven ™. :grin:

I agree with @ryan.marin that shorter time at higher temperature will produce less surface oxidation and that the little extra time at temperature for 4 gauge makes sense. It will also depend upon the mass of the wire and how close the wires are together. For example, my coil of 18 gauge is about 2 cm in diameter and it probably takes quite a while for the wire in the center to heat up since the air gaps between all of the different wires serve as thermal insulation. I am also curious to know if @ryan.marin is at 950 Celsius or 950 Fahrenheit. 950C is about as hot as I would go in an at home system since the melting point of pure copper is 1085C (memorized during my MS work 30 years ago). The most common grade of electical copper wire (the most likely source of our wire) is electrolytic-tough pitch (ETP) copper which is 99.90% copper (Wikipedia) so the melting point will be close to 1085C.

Copper(II) oxide or cupric oxide is black and has the formula CuO. Copper(I) oxide or cuprous oxide has the formula Cu2O and is generally red, but fine crystals are yellow (Wikipedia again). At atmospheric oxygen concentration, black CuO is the stable oxide at all temperatures so the surface will be black after a long time at temperature. However, the red/yellow Cu2O will be the first to form since the reaction will be driven by oxygen diffusion into the surface of the copper. In addition to the oxygen forming oxides on the surface it will diffuse into the copper (very small amounts) as the wire is held at high temperature (I had to look up the Cu-O phase diagram).

Based upon this information, my guess is that what @ryan.marin is seeing is that purple wire has been annealed for a long time at temperature (red & black is purplish) to produce a mix of CuO and Cu2O on the surface plus a fair bit of oxygen diffused into the bulk copper which will act to impede dislocation motion making the copper harder. The red surface is mostly Cu2O from a lower time-temperature combination that results in larger grains of Cu2O which are red and some oxygen diffusion into the wire. Finally, the orange wire saw the lowest combination of time-temperature so the Cu2O grains on the surface are finer (more yellow) and there is even less oxygen in the wire.

What does it all mean? Best solution is to anneal to get the optimal grain size (my guess is 10-20% of the wire diameter) in a very low oxygen environment, ideally less than about 1 pPa (about 10^-13 of atmospheric oxygen content and very hard to achieve). However, the comment about using a Raku kiln for the reducing atmosphere will help quite a bit. Soft copper tubing is typically annealed in a reducing atmosphere (still contains some free oxygen) at 704C (1300F) so that seems like a good temperature for the copper to reach. If in air, we would heat in a high temperature oven it for as short a period as possible to avoid formation of the CuO and diffusion of oxygen into the copper and then quench.

It would be fun to do some analysis of stiff (purplish), medium (red), and soft (orange) wire. I could easily look at the grain structure which would tell us a lot about how much time-temperature the wire saw assuming the process to make the wire was the same (an unannealed same from the same batch would be good). Doing the chemical analysis is beyond my current capabilities. I would need about 1 cm of a heavy gauge (4 or 6) to do the metallography.

1 Like

i do not use Oven, i use a MeatPit, with vegetal coal and wood, put the coils at 1 finger thick max under the ember until they turn all red/vine, work in the dark because the UV or light prevents you to see the real colour and you can cook the wire in place to heat it up. after i put it in water and and then in a floor cleanning acid+soap liquid for 5min. after this brush the coils with a wire brush.

here is the images.


Your Insta is private

i only have the images on insta, but i changed to public for you can see.

1 Like

Nice. Where do you get your copper from? I’ve considered annealing my own, but figured that by the time I source the copper and go through the work that I’m better off just buying it. :man_shrugging:t4:

You can buy unannealed copper from most hardware stores. It’s usually in the electrical section.

There are secrets… if I tell, they have to kill me…:roll_eyes:
I guess I have time now to throw pots. A kiln can be used as an annealing oven… the price is high. Cheaper than the real thing. Still cheaper to buy from the pros.

OK folk,
I’ve tentatively got some time to fiddle this weekend so I’m going to pick up some generic copper ground wire from the HD and see what I can discover. I’ll be using a 200,000 BTU propane burner and the largest gauge bare copper that I can get from HD. The variables that I’m considering are:
time on the burner
gross temp assessment by color
quenching in water
quenching in a reduced O2 environment
I don’t want to introduce too many variables but I’m very open to suggestions.

here in Brazil is really cheap, i buy all gauges from 16 to 4 in a suplyer of coper wire to eletric motors. they come with a resin that evaporates on the pit. the quality of the final product is very close to the kaneshin wire.

in comaprison here in Brazil kaneshin wire coasts US$80,00 1kg. My wire US$6,00 1Kg paying wire, coal, and wood to fire, and i sell the excess of my use for US$10,00 1Kg.

1 Like

there are very good vídeos on youtube explaining the diference between pit, oven and torch. the propane torch heats very rapid, and can cause uneven heating, be awere because excessive heat can cause hardiness and break the wire during the use. i watch youtube several vídeos of annealing coper from jewalery. using the pit you can cover the wire and heat very even and low O2 contact, that promotes rapid annealial and almost nothing of oxidazing black dirt


Just be aware, the copper oxides are hazardeous, especially lungs. If you are doing a lot of this work, wear a mask, gloves, and wash your work clothes… Don’t take your work home…
Keep a log of good and bad results.


This is just an update on annealing study. The hypothesis was that cooling in a O2 deprived environment would result in less oxidation on the annealed wire. Three samples of gauge 4 wire were tested. They were all cut to an equal length (~ 12 inches). The sample that was the control was a piece of wire from a bonsai vendor who does annealing. One of the other samples was to be quenched in a reduction environment (metal trashcan with wadded up newspaper) and one to be quenched in water. All three samples were heated at the same time over an open flame propane burner (very hot).
I did not have way to determine how hot the samples were so I watched for the color changes that have been described earlier in this string. I got nothing. If there was ever a orange color, I completely missed it. both of the samples that were to be quenched heated and turned black very quickly. After about 5 min, I removed one sample and placed it in the O2 reduced environment covering the can with the lid to prevent / reduce O2 entry and immediately placed the other in the water.

the results are:
right - commercial purchased
middle - water quenched
left - O2 reduction environment

The O2 reduced environment had less oxidation but was not uniform in the color. The best portion matched the commercial piece and the worst matched the water quenched.
The water quenched was uniformly black.

Stiffness test
The samples were supported by a vise over a fixed surface with parallel to the floor. A 1Kg hook weight was placed on the wire at its exit from the vise and then serially moved away from the vice, watching for sustained deformation of the wire with the endpoint being contact of the weight to the fixed height surface. When this deformation occurred, the distance from the vice was measured.
water quenched deformed at 125 mm
O2 reduced deformed at 125 mm
commercial deformed at 165 mm

In this limited sample, there was no difference in the “softness” between the water quenched and the O2 reduced samples. The commercial sample was the stiffest tested
The water quenched had a more uniform heavily oxidized surface
The O2 reduced was mottled in appearance and had less oxidation present
The commercial looked the best with the least oxidation and was the stiffest.

Subjectively, after the test, I manipulated the samples like I was going to wire a tree. Those that I annealed were like working with butter and hardened with manipulation nicely.

The next question is how to get accurate temps and how to increase the sample size (usable coils) and have consistent heating. I’m also considering if core fiber would be a good choice for the reduction environment to promote less mottling.

1 Like

Didn’t Troy do a short feature on annealing? I can no longer find that

I’d be surprised if they did. Mirai is super secretive about their annealing process.


When you prepared the commercially annealed wire for testing did you first straighten it out? If so this would cause a certain amount of hardening to occur and could account for the increased stiffness.