PerspectiveMaterials Science

Designing Superhard Materials

See allHide authors and affiliations

Science  27 May 2005:
Vol. 308, Issue 5726, pp. 1268-1269
DOI: 10.1126/science.1109830

Ultrahard materials are used in many applications, from cutting and polishing tools to wear-resistant coatings. Diamond remains the hardest known material, despite years of synthetic (1, 2) and theoretical (3) efforts to improve upon it. However, even diamond has limitations. It is not effective for cutting ferrous metals, including steel, because of a chemical reaction that produces iron carbide. Cubic boron nitride—the second-hardest material, with a structure analogous to that of diamond—can be used to cut ferrous metals. However, it does not occur naturally and must be synthesized under conditions of extreme pressure and temperature, making it quite expensive. New superhard materials are thus not only of great scientific interest, but also could be very useful.

To design new superhard materials, we must understand what makes diamond special. In diamond, tetrahedrally bonded sp3 carbon atoms form a three-dimensional, covalent network of high symmetry. Other carbon-based materials have shorter and stronger carbon bonds, but not in three dimensions. For example, the trigonal sp2 bonds in graphite form sheets with shorter and stronger carbon-carbon bonds. But only weak van der Waals interactions hold the sheets together, allowing layers of graphite to cleave readily. A three-dimensional network composed of short, strong bonds is thus critical for hardness.

In thinking about new ultrahard materials, it is useful to consider the types of structural changes that a material can undergo under load. These changes can be divided into elastic (reversible) and plastic (irreversible) deformations.

A material is considered stiff if it is difficult to compress elastically. Such a material has a large bulk modulus (it is resistant to volume compression) and/or Young's modulus (it is resistant to linear compression). Elastic deformation in a direction different from that of the applied load results in shape rather than volume changes; these motions are measured by the shear modulus. In all elastic distortions, the basic relations between atoms do not change.

A material is considered hard if it resists plastic deformation. In contrast to elastic deformation, plastic deformation usually involves irreversible motion of the atoms with respect to each other, often via the creation and movement of dislocations.

Toward superhard materials.

By combining metals with a high density of valence electrons, such as osmium, iridium, or rhenium, with small, covalent bond-forming atoms such as boron, ultra-incompressible, hard materials may be created. Mixed metals, as shown in this orthorhombic structure predicted for (Os, Ir)B2, can act as barriers to the movement of dislocations. Osmium is shown in red, iridium in green, and boron in yellow.

It is a source of substantial confusion that high modulus and high hardness are often discussed together, even though the underlying deformations are fundamentally different. This grouping occurs because the processes can be correlated: If a material shows large elastic changes under small load (low modulus), it tends to respond to larger loads by deforming plastically (low hardness). This is particularly true for shear motions, which are required to scratch or indent a material; a good correlation has been found between shear modulus and hardness (13). Highly directional bonding is needed to withstand both elastic and plastic deformations. Purely covalent bonding (such as in diamond) is best, and some ionic character is acceptable. However, highly ionic or metallic bonding is the same in all directions and therefore poor at resisting either plastic or elastic shape deformations.

With these ideas in mind, efforts to design superhard materials can be divided into two main approaches. In the first, light elements, including boron, carbon, nitrogen, and/or oxygen, are combined to form short covalent bonds. In the second, elements with very high densities of valence electrons are included to ensure that the materials resist being squeezed together.

The first approach gained favor in the late 1980s, when calculations suggested that the hypothetical compound C3N4 may be even less compressible than diamond (4). However, after years of experiments, further calculations indicated that even for the least compressible C3N4 structure, the shear modulus would only be 60% of the diamond value (5). New forms of carbon, including fullerenes and nanotubes, generated great excitement in the 1990s, when high-pressure processing produced very hard substances (1). However, these substances, which fall under the rubric of diamondlike coatings, can approach but never reach the hardness of diamond (6); furthermore, squeezing fullerenes and nanotubes is unlikely to be an inexpensive, practical synthetic route to diamondlike carbon. Three-dimensional boron-rich compounds, including B4C, B6O, their solid solutions, and B/C/N phases, are very hard materials that deserve continued study. However, this approach is unlikely to produce materials with hardnesses exceeding those of boron nitride/diamond solid solutions, which are intermediate in hardness between diamond and cubic boron nitride (1, 2, 7).

In the second approach, transition metals that have a high bulk modulus but low hardness are combined with small, covalent bond-forming atoms such as boron, carbon, nitrogen, and/or oxygen. In this way, a material that can maintain both volume and shape can be created. This idea has led to highly incompressible phases such as RuO2 (8), WC, and Co6W6C (9). Unfortunately, these materials do not even approach the hardness of cubic boron nitride, owing to the partially ionic character of the Ru-O bond and the metallic nature of the W-W and Co-W interactions (3). Borides may be a better choice to achieve the required covalent bonding. Transition metal borides such as the tungsten borides WB4, WB2, and WB are promising (1, 2). Elements with a higher density of valence electrons (and thus high bulk modulus) such as rhenium, osmium, and iridium also have the potential to form very hard borides (10); mixed-metal borides could be even harder (see the figure).

Once the best combination of elements is found, hardness could be increased by controlling the underlying nanostructure. For example, if the motion of dislocations in a material is hindered, hardness can be increased. This phenomenon is well known to occur in an ultrafine-grained diamond called carbonado (11). More recently, nanoceramics with a grain size of ∼10 nm have exhibited the same phenomenon (12). Superlattices of TiN/AlN or carbon nitride/TiN with a periodicity of 6 to 8 nm also exhibit hardnesses two to three times as great as that of the bulk crystalline form of these materials (13, 14). In all these materials, the interfaces between the nanometer-scale components act as barriers to the movement of dislocations.

Despite all the research activity into synthesizing superhard materials, many opportunities remain unexplored. For example, the lightest element that could produce three-dimensional structures, beryllium, has been neglected, perhaps because it is toxic and may require specialized high-pressure equipment. Ternary phases of beryllium with other light elements—boron, carbon, nitrogen, and oxygen—could have exciting properties in their own right or in combination with high-valence electron density metals.

Despite their potential, new materials are unlikely to replace diamond altogether, because in addition to its hardness, diamond possesses many other amazing properties. It is the most incompressible material, has one of the highest indices of refraction, and has a room-temperature thermal conductivity five times as large as that of the best metals. The scientific challenge of finding a superhard material that surpasses diamond in any of these properties will keep the field energized for years to come. Combining high hardness with other properties, such as chemical inertness and low-cost synthesis, could quickly yield practical benefits, for example, by providing a replacement for cubic boron nitride for cutting and polishing steel.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
View Abstract

Stay Connected to Science

Navigate This Article