Iron is a pretty economically attractive material: extremely abundant (5.6% of Earth’s crust, 32% of Earth), moderately refractory (doesn’t melt until 1600°), can form carbides called “cementite” that make it into a very hard and strong “alloy” called steel (arguably really a cermet), and can be heat-treated to increase its hardness further. But could we make “steels” based on other similar metals, if we had enough of them? I think we could.
Steel is a very complex system, and I am doing it a bit of an injustice by simplifying its behavior to just “derives its strength from cementite”.
Regular iron cementite (Fe₃C) has a “hardness” of 7–11 GPa, which I take to mean that its ultimate tensile strength is 7–11 GPa. Jiang and Srivilliputhur calculate ideal tensile strengths for cementite between 15 and 30 GPa in different directions, though I haven’t really read their paper. When steel changes phase from austenite to ferrite, the solubility of carbon in the iron phase drops greatly, precipitating submicron-thickness layers of cementite alternating with ferrite in a structure called “pearlite”. (I don’t think they’re thin enough to be below the flaw-tolerant critical size, which is estimated at 30 nm for goethite fibers such as those found in limpet teeth, but which exists for any material; nacre actually uses 200–500-nm-thick crystals, similar to the thickness of cementite layers in some pearlite. The cementite layers in bainite might be thin enough to be flaw-tolerant.) As a consequence, pearlite wires can reach tensile strengths over 6 GPa. But cementite is unstable above 723°, so steels become soft and malleable when they transition to the austenitic phase.
Mushet steel, the nascent form of high-speed steel, includes 1.5–2.5% carbon, 4–12% tungsten, which also forms a hard carbide with carbon, and 2–4% manganese. As I understand it, the manganese makes it austenitic at room temperature, allowing it to be air-hardened without quenching and to have much greater toughness than earlier hardened steels. I think that most of the carbon in Mushet steel and similar high-speed steels ends up in tungsten carbide rather than carbides of iron or manganese, and that nearly all the tungsten does.
Tungsten carbide’s tensile strength is normally cited as being only around 0.4 GPa, but because it’s a brittle ceramic, I suspect that number is dominated by flaw-sensitivity. It doesn’t decompose until 2800°, and I think this is why Mushet steel and the modern high-speed steels that are based on tungsten remain hard at high temperatures.
I think you could probably make a “steel” consisting of tungsten and a little carbon. The solubility of small amounts of carbon in tungsten rises up to 2715°, so you could probably get some kind of interspersed pearlite-like microstructure by quenching tungsten down to a lower temperature. It wouldn’t have to be anywhere near room temperature; quenching it to 1091° in molten magnesium or 907° in molten zinc would be just fine. (There’s a second tungsten carbide, but it only becomes important at higher concentrations.) But at room temperature tungsten is kind of brittle, and of course tungsten itself is a very rare element (0.17 ppm of Earth by weight, 1.25 ppm of the crust), which sometimes matters.
Tungsten is a group-6 transition element, along with chromium, molybdenum, and the wildly radioactive and presently irrelevant seaborgium.
Chromium (4700 ppm of Earth, 100 ppm of Earth’s crust) is a bit more refractory than iron (melting at 1907°), is the base of some stainless steels, and is commonly plated on top of steel to make it harder, shinier, and more corrosion-resistant. In many ways it’s reasonable to think of it as a sort of half-assed tungsten. So what about its carbides?
Well, chromium has three carbides, which are indeed refractory (1895°, so less refractory than metallic chromium), very hard, and corrosion-resistant, and they’re commonly used to improve the wear and corrosion resistance of metals. So far so good. I should look up the chromium–carbon phase diagram.
Molybdenum is damn near as refractory as tungsten itself, melting at 2623°, but also damn near as rare: 1.7 ppm of Earth, 1.2 ppm of its crust. Its oxide is a lot more volatile than tungsten’s, limiting its refractory usefulness in applications exposed to air, but it also has a very hard refractory (2687°) carbide. I should look up the relevant phase diagram.
So far, we’ve looked at carbides of iron, tungsten, chromium, and molybdenum, all of which are very hard and refractory due in part to the somewhat covalent character of their bonding. But there are other elements that can play a similar role: boron, oxygen, nitrogen, and sulfur, as well as oxoanions like phosphate. Some of these are counterproductive in iron itself: the oxides, nitrides, and sulfide of iron are all weaker and less refractory than iron itself. But iron boride is somewhat of a hit (Vickers hardness of 15–22 GPa, melts at 1389°, already in use as a steel ingredient for hardness and used for surface hardening), and iron tetraboride is superhard, and, in combination with other metals, some of these elements produce very interesting ceramics.
Sticking to just the metals so far mentioned, tungsten borides have Vickers hardnesses of 20–30 GPa and WB₄ is described as “an inexpensive superhard material” because you can make it with just arc melting from the elements; chromium borides are also very hard and strong and can be made by SHS, especially if you add aluminum; and molybdenum borides are predicted to be superhard but apparently nobody has managed to produce them in volume yet.
As for the oxides, I’ve mentioned the iron oxide goethite above (not usually thought of as superhard, but the limpets manage); tungsten trioxide has 5–7 GPa hardness at 800°; chromium oxides include chromia (viridian) which melts at 2435° and has Mohs hardness 8 as the mineral eskolaite; and, though it melts at only 802°, molybdenum trioxide has a hardness of 18.7 GPa, though as a mineral it’s only Mohs 3–4.
Nitriding, carbonitriding, and nitrocarburizing are commonly used as a surface hardening process for steel, chromium, and molybdenum, and nitriding has been used to harden iron since antiquity, with urine, leather, and hooves being preferred case-hardening ingredients. Tungsten nitride is also hard, but it decomposes in water, limiting its use in air-contact applications.
Many elements also have interesting oxynitrides, oxyborides, borocarbides, boronitrides, borocarbonitrides, carbonitrides, and oxycarbonitrides. Oxycarboborides and oxyboronitrides seem to be either neglected or too difficult to make, and although some “oxycarbides” are reported (including a molybdenum oxycarbide), many more are just carbonyls or oxalates, which are neither hard nor refractory.
What about nickel, cobalt, vanadium, manganese, titanium, zirconium, hafnium, silicon, niobium, and tantalum? They also form carbides! That makes 110 more candidate ceramics to investigate!