There’s some very interesting chemistry going on between a copper pan and its tin lining.
Tin has been used to line copper pans for centuries for a number of reasons, perhaps first and foremost because tin sticks so readily to copper. Take a look at this video from Mauviel showing how they apply tin to a brand-new copper pot.
They make it look easy: hot pan, molten tin, swirl and wipe, quench. Of course it’s not easy at all — Mauviel’s tinsmiths are highly skilled at what can be a very hazardous process — but my point is that the tin sticks quickly and firmly to the sides of the pan.
But the thing is, it’s quite difficult to get that tin off. Tinners can’t just heat the pan up again and pour out the tin like it’s water — there’s a base layer that has to be scraped away. You can see that base layer on antique copper pans that have been cleaned the old-fashioned way (prior to the invention of dish soap in the 1940s) by scouring them with sand. Here are three such pans before I sent them off to be retinned, and you can see that the base layer is still there despite what looks like very harsh scouring.
If you want to get this last layer of tin off, it takes hard physical abrasion. Fast-forward to 3:09 in the retinning video below from Sherwood Tinning (apparently not known for being gentle with old copper) and you’ll see that they use a buffing wheel.
What is this stuff? It’s not shiny like tin, it doesn’t melt like tin, and it’s harder and more resilient than tin. What is it and how did it get there?
The intermetallic layer
The short answer is that the base layer is an intermetallic compound made up of both copper and tin. A metallurgist explained it to me:
When tin is coated on copper, an intermediate layer is created called an intermetallic that is sort of the “glue” between the two metals, and is really a different material altogether. The intermetallic is harder and more brittle, and would not be easily removed.
At the threshold between the copper (Cu) and tin (Sn), the atoms bond to form a type of alloy that behaves more like a ceramic than a metal. Intermetallics always form in two variants (called phases) with atoms in ratios of 3:1 or 6:5, and in the case of copper and tin, this creates Cu3Sn and Cu6Sn5. The pure tin portion of the lining rests on top and is held in place by the intermetallic “glue,” as the metallurgist describes it. This is why overheated tin doesn’t just slosh down the pan — the intermetallic layer holds the tin in place until some stronger force (like your spatula) disturbs it.
It’s important to note that this is not bronze (CuSn), the metallic alloy of tin and copper. According to the metallurgist,
The atomic bonding of intermetallic materials include both metallic and ionic bonds, thus they are not actually metallic alloys like bronze. This makes them hard and brittle, as opposed to bronze which is quite ductile… You would not be able to create a bronze alloy without melting down and mixing the constituents.
In other words, the copper and tin atoms aren’t forming the bonds that make bronze, but instead form different bonds that lock the atoms into a rigid lattice. That rigidity makes it more like a ceramic — it’s smooth, hard, and brittle. That’s why the base layer of a tin lining is so resilient and difficult to separate from the copper underneath.
The metallurgist also explained the purpose of flux during the tinning process. I’ve heard tinners describe flux as a cleaner but I thought this was just to get dirt or ash off the surface. According to the metallurgist, the purpose of flux is to clear the way, chemically speaking, for the right bonds to form between copper and tin.
Flux is basically a way to clean off the surface where you are attempting to create a metal-to-metal bond. It usually consists of an acidic element that concentrates and activates with heat. Since most metals react with the atmosphere, typically forming an oxide, the flux is needed to eat away the oxide so that a true metal to metal bond can occur, which creates the intermetallic.
Think of the metalsmiths over the millennia who figured this out by trial and error!
If you were to slice through your copper pan and look at the cross-section of the tin you’d be able to see the intermetallic layer. Fortunately you don’t have to do this — below is an electron scanning micrograph of the interface between pure copper and pure tin. Imagine that the dark Cu at the bottom is the surface of your pan; above it is a layer of Cu3Sn, a layer of Cu6Sn5, and then finally the pure tin (Sn) at the very top.
The intermetallics start forming the moment the copper and melted tin come into contact. How quickly the layers build up depends on time and temperature; this particular sample was heated to 150°C (302°F) for 1,000 hours, and as you can see in the diagram above, the layer of combined Cu3Sn and Cu6Sn5 is 18.74 micrometers (μm) or 0.018 millimeters thick.
But here’s the thing: the reaction keeps going, even at room temperature. Tin atoms diffuse quite well into copper and with a little heat energy will readily bond to form molecules of Cu6Sn5. (In fact, a good deal of recent industrial chemistry is about preventing the formation of copper-tin intermetallic, as it gets in the way of the function of soldered circuit boards.) Given more heat (or time, or both), the Cu6Sn5 converts to Cu3Sn. (The Cu3Sn phase is more chemically stable, so that’s where the atoms end up.) This happens pretty slowly at room temperature but speeds up when you use your pan to cook over heat.
The diagram below illustrates this. Think of each stack as a cross-section of your tin lining over time: on the left is your freshly-tinned pan with a layer of pure tin sitting on top of a narrow layer of intermetallic. Over time, your pan looks like the middle column: more pure tin has converted to Cu6Sn5 and then to Cu3Sn. Eventually your pan’s tin looks like the stack on the right, where there’s pretty much no free tin left and the intermetallic is all converted to Cu3Sn. The transformation happens slowly at room temperature and faster as the temperature increases — for example, when you cook.
What this means is that over many years of use, the tin lining on your pan is building up this hard and resilient base layer. The base layer anchors the tin to the copper and helps hold the pure loose tin in place. This process was jump-started by the hot tinning process but it’s going on right now, even as your pan is sitting on your shelf. Eventually — over decades — all of the tin will be converted to Cu3Sn.
You will most likely have retinned your pan before all the tin was consumed. But what would happen if you didn’t? As luck would have it, I can show you.
Take a look at this 31cm skillet with a strange-looking lining. This pan has seen a lot of use and has gotten pretty hot, as often happens with skillets. The lining looks like a dull pewter with tiny lumps and smears of tin all over it.
The closeup photo below shows that the surface doesn’t look like a normal tin lining. There are areas of mottled matte gray with spots of lumpy metal on top. (There’s also some brown seasoning color on there — this is a skillet that I use frequently and it’s picked up some polymerized oils.) I showed this photo to the metallurgist, who said that the mottled gray patches are exposed intermetallic and the lumps on top are the pure tin that remains.
What’s interesting to me is that this is one of my very favorite pans to use. It cooks beautifully — in particular it seems to withstand high heat quite well. I use all my tinned skillets frequently, and while the others have picked up a few lumps and smears here and there, this one’s surface is unchanged. This makes sense according to the metallurgist: “Over time the intermetallic will grow, consuming any free tin, so the smearing would happen less often as the free tin becomes thinner, eventually leaving only the intermetallic layer.”
It’s also quite low-stick, more so in my experience than my Mauviel skillets with their shiny newer tin. It doesn’t quite have the non-stick performance of a PFTE coating like Teflon, but the intermetallic areas feel similar under my fingertips, smooth and hard like an eggshell. Intermetallics in general have lower coefficients of friction than pure metals and they’re often used industrially as coatings for metal bearings. This suggests to me that the copper-tin intermetallic on this skillet’s surface could be performing like a low-stick cooking surface, the same as (if not better than) the pure tin it has replaced. (Look back at the photo above — the polymerized oils don’t seem to be adhering to the intermetallic areas!)
The properties of copper-tin intermetallic suggest to me some possible explanations for three phenomena I’ve observed with my tinned copper pans.
Overheated tin tends to stay in place because the intermetallic layer stabilizes it. Tin melts at 450°F (232°C), but copper-tin intermetallic has a much higher melting point: Cu6Sn5 melts at 779°F (415°C) and Cu3Sn at a much higher 1,248°F (676°C). This means that the intermetallic layer stays solid and keeps stabilizing the lining at temperatures well above tin’s melting point. It also explains why smearing a tin lining takes some physical action, like scraping or stirring — the melted tin will stay put until some external force is applied.
An older tin layer resists smearing better than a new one. As the intermetallic layer grows and the layer of pure tin thins, more of those pure tin atoms are in proximity to and stabilized by the intermetallic layer.
An older tin layer resists scraping through to the copper. One vulnerability of a tin lining relative to harder materials like stainless steel is that harsh scraping or scratching could gouge through the soft tin layer to the copper underneath, hastening the need for a new lining. Tin has a Vickers hardness of less than HV 80, but according to the Copper Industry Association, copper-tin intermetallic has a hardness around HV 300, almost as hard as copper. (This is why it’s possible to scrub tin off by accident with a scouring sponge, but tinners have to scrape copper-tin intermetallic off with a grinding wheel.) This means that a robust ceramic-like intermetallic layer is more likely to withstand abrasive pressure, deflecting a kitchen utensil or scrub sponge before it reaches the copper.
This suggests to me a tantalizing possibility: tin linings get better with age and use. Maybe the best thing you can do for your tinned pans is to cook with them — a lot! — so that the lining matures and becomes better for cooking. What do you think?
If you are a scientist — and particularly a metallurgist! — I would love to hear from you, even if — especially if! — you disagree with or can disprove my ideas. Please leave a comment or email me at vfc at vintagefrenchcopper dot com. Thank you!
In addition to input from my metallurgist friend, I drew on a few sources for information.
Intermetallic cross section: Yang, Ping-Feng & Lai, Yi-Shao & Jian, Sheng-Rui & Chen, Jiunn & Chen, Rong-Sheng. (2008). Nanoindentation identifications of mechanical properties Of Cu6Sn5, Cu3Sn, and Ni3Sn4 intermetallic compounds derived by diffusion couples. Materials Science and Engineering: A. 485. 305–310. 10.1016/j.msea.2007.07.093.
Diagram of intermetallic layer growth:
(Note that the diagram has mis-labeled Cu6Sn5 as Cu3Sn5 in the white area of each cross-section.)
Brinell hardness to Vickers hardness conversion:
An excellent discussion of hardness as a concept and the differences between Rockwell, Brinell, and Vickers hardness measurements:
Hardness values for Cu and Sn:
Vickers hardness for copper-tin intermetallic:
Also, thanks to reader Phil for a correction from grinding wheel to buffing wheel! I appreciate it!