What “hard” means here
In materials science, hardness is resistance to permanent surface deformation, usually measured by pressing a small, very hard tip into a surface and looking at the mark it leaves. A related everyday idea is scratch resistance. Neither is the same as strength or toughness, and that difference matters throughout this topic.
“Superhard” is a convention rather than a precise physical category. The usual cutoff is a Vickers hardness of about 40 gigapascals, a level that only a small group of solids reaches. Diamond sits at the top of that group, and cubic boron nitride is generally next. Many other materials are described as hard, such as tungsten carbide or alumina, but fall below the superhard line.
What these materials have in common
Looking at the superhard materials side by side, a few bonding features keep appearing.
Strong covalent bonds. In a covalent bond, neighboring atoms share electrons. These bonds are strong and have a definite direction in space. Metallic bonds, by contrast, let electrons move freely and let layers of atoms slide past each other, which is why most metals are soft compared with ceramics.
Short bonds between small atoms. The elements involved are mostly light ones near carbon on the periodic table: boron, carbon and nitrogen. Small atoms sit close together, so the bonds are short and the atoms are packed densely. A high density of strong bonds means a lot of bonding must be disturbed to make a dent.
A rigid three-dimensional network. Diamond is the clearest case. Each carbon atom bonds to four others, and the pattern repeats in all directions with no weak plane running through it. Graphite, also pure carbon, has strong bonds within each layer but weak attraction between layers, so it is soft and slippery. The atoms are the same; the bonding geometry is what differs. The details of the diamond crystal structure explain why no such weak plane exists, apart from the cleavage directions noted later.
Resistance to dislocation motion. In a metal, plastic deformation mostly happens as line defects called dislocations glide through the crystal. In directional covalent solids, those defects move with great difficulty, so the material tends to resist deforming and fails by cracking instead.
The main members of the family
Diamond is the benchmark. Cubic boron nitride has a similar structure with alternating boron and nitrogen atoms, and it is the main alternative where diamond is unsuitable. Boron carbide and silicon carbide are very hard ceramics with more complicated crystal structures, though both fall short of the superhard threshold in most measurements, depending on the method. Cemented carbide is a composite: hard tungsten carbide grains held in a metal binder.
Some researchers have also reported superhard behavior from nanostructured materials, in which extremely fine grains or twins block deformation. Results like these are harder to reproduce and are not yet common in industrial use, so treat individual claims cautiously.
Intrinsic versus engineered hardness
A useful way to separate cases is to ask where the hardness comes from. Intrinsic hardness comes from the bonding of a single perfect crystal, as in diamond. Engineered hardness comes from microstructure: grain size, added phases, residual stress and heat treatment. Steels and cemented carbides rely heavily on the second kind. In many ceramics both contribute, and the exact balance is one reason sintering process choices have such a large effect on the final part.
Hardness is not the whole story
The same features that give a covalent network its hardness make it unforgiving. With dislocations locked in place, a crack has little way of relieving stress and can run through the material. This is why the hardest materials are often the least tolerant of flaws, a trade-off captured by the idea of fracture toughness. A part that is the best possible choice against abrasion may chip under impact.
Chemistry sets further limits. Diamond is carbon and can react with iron at high temperature, which is part of the reason for the machining problems seen in ferrous metal machining. Heat in air can also oxidize several of these materials, which sets oxidation limits on their use.
Why the numbers differ between sources
Published hardness values for the same material often vary. Results depend on the test load, the tip shape, crystal orientation, sample purity and whether the sample is a single crystal or a sintered body. That is why a careful reader should treat any single figure as approximate, and why indentation testing deserves its own explanation.
Common misconceptions
“Hardest means strongest.” Hardness, strength and toughness are separate properties. A superhard crystal can be easy to chip along a cleavage plane, and a modest-looking steel can absorb far more impact.
“Superhard means pure carbon.” Diamond is carbon, but cubic boron nitride is not, and the hard phase in many tools is a carbide. What matters is the bonding pattern, not a particular element.
“Hard materials last forever.” Wear depends on the pairing. A material that resists abrasion may still be attacked chemically or thermally by the workpiece, which is why the right abrasive depends on what is being cut.
Where superhard materials are used
Because they wear slowly, these materials appear wherever abrasion is the main problem: grinding wheels, cutting tools, drill bits, wire-drawing dies, abrasive blasting nozzles and protective plates. They are rarely used as bulk structural parts. More often a small amount of the hard phase is bonded to a tougher body, and the choice of bond types or binder phase matters as much as the abrasive itself.