Natural diamonds crystallize in Earth’s mantle at high temperatures and pressures. The most popular method for making synthetic diamonds mimics those conditions by using anvil presses to achieve extreme pressures. Another approach, chemical vapor deposition (CVD), works at low pressures and is used to grow diamonds layer by layer on a substrate in a vacuum, a process more akin to the diamond growth that occurs in interstellar gas clouds.
Now Rodney Ruoff (at the Ulsan National Institute of Science and Technology and the Institute for Basic Science Center for Multidimensional Carbon Materials in South Korea) and colleagues have unveiled another route to diamond growth that works at ambient pressure. The approach, which involves methane and hydrogen gas and a catalyst of molten metals, requires less energy and less-advanced equipment than the two leading diamond synthesis methods.1
Below pressures of about 20 000 atm, graphite is the stable form of pure carbon. Diamond is kinetically stable at those lower pressures but not thermodynamically favored. CVD succeeds at growing diamond rather than graphite at low pressures by the inclusion of free hydrogen, which continually reacts with and etches away at any graphite that forms, while diamond is allowed to grow.2 Many energy-intensive techniques—such as hot filaments, microwaves, or electric arcs—are used to ionize hydrogen gas and produce those free hydrogen atoms.
The new liquid-metal catalysis method differs from CVD in that it does not use any high-energy-density methods to ionize hydrogen gas, and the diamond is grown within a liquid-metal mixture, as shown in figure 1, rather than being deposited onto a solid surface from a vapor. And unlike many CVD approaches, the new technique does not rely on seed crystals. Ruoff says liquid-metal catalysis’s lower energy requirement, seedless growth, and ability to spread liquid metal over large surface areas could make it an attractive alternative to the CVD and high-pressure, high-temperature (HPHT) synthesis methods, at least for some applications.
A diamond film (light gray) can be seen breaking through the surface of a metal mixture (dark-gray rectangle) used to precipitate the crystals. (Courtesy of Yan Gong.)
A diamond film (light gray) can be seen breaking through the surface of a metal mixture (dark-gray rectangle) used to precipitate the crystals. (Courtesy of Yan Gong.)
Seed of an idea
Ruoff was inspired to explore liquid-metal catalysis after learning of experiments in which scientists had grown single sheets of graphite from methane using liquid gallium as a catalyst.3 Breaking the strong covalent bonds between the carbon and hydrogen atoms in methane usually requires a high activation energy, but gallium had significantly lowered the reaction barrier. And the catalysis had occurred at temperatures as low as 50 °C. (Pure gallium, with a melting point just under 30 °C, can turn to liquid from just the warmth of a hand.) The finding made Ruoff think that gallium’s strength as a catalyst could lead to lower-energy diamond synthesis.
In early trials, Ruoff and his colleagues poured molten metals over diamond flecks in a chamber with carbon-bearing gases. They performed the experiments on silicon dioxide wafers, an unreactive base. The flecks didn’t grow—until the metal spilled over the wafer during one experiment and touched pure silicon at its edge. Soon after, Yan Gong, a PhD student in Ruoff’s lab, saw evidence of new pyramid-shaped crystal growth on the diamond seeds.
When the researchers looked closer, they found that new diamond had grown on the seed crystals and that the pyramids were topped by single crystals of silicon carbide. Their presence indicated that silicon from the wafer had dissolved into the metals and somehow aided the growth of diamond.
The researchers then pivoted to exploring the role of silicon, using a graphite crucible instead of a silicon wafer as a base. They experimented with varied concentrations of silicon in the mix and found that any concentration above 1% would grow silicon carbide. At smaller concentrations, they began to see diamond growth, but not in the way they expected. Graphite was growing on the seed crystals, but diamond was growing on the base of the graphite crucible. Again, the lab pivoted, this time to seedless growth.
In their new paper, Gong and colleagues now show how a mixture of gallium, iron, nickel, and silicon can form diamonds: When heated to 1025 °C in a chamber with methane and hydrogen gas, the mixture nucleated diamond nanocrystals at the base of the graphite crucible after 15 minutes and formed a continuous diamond film after 150 minutes (see figure 2). Although silicon’s role in the nucleation process is not fully understood, it seemed to be crucial to the experiment’s success. Silicon was incorporated throughout the lattice of the diamond, and slight variations of the silicon concentration produced different crystal sizes and densities. And when it was depleted from the mix, diamond growth came to a halt.
Diamond nanocrystals, captured in scanning electron microscope images after growing at ambient pressure for (a) 30 minutes and (b) 150 minutes, at which point they’ve formed a continuous film. Silicon aids the nucleation of the carbon crystals from a liquid-metal mixture. (Adapted from ref. 1.)
Diamond nanocrystals, captured in scanning electron microscope images after growing at ambient pressure for (a) 30 minutes and (b) 150 minutes, at which point they’ve formed a continuous film. Silicon aids the nucleation of the carbon crystals from a liquid-metal mixture. (Adapted from ref. 1.)
Carat and stick
The polycrystalline diamond film grown with the new method is far from the gem-quality single crystals that the jewelry industry prizes. As the method is further developed, it may get better at making larger individual crystals. But diamond’s many valuable qualities—including extreme hardness, high thermal conductivity, and semiconducting electron transport—make it a useful material for a range of applications, not just jewelry. For example, diamond films are used as wear- and corrosion-resistant surfaces and as heat sinks on integrated circuits.
“Diamond is essential for everything in our daily life,” says Soumen Mandal, a postdoctoral research associate at Cardiff University in the UK who specializes in nanocrystalline diamond growth. “For example, in your car there’s almost a kilometer of wire, all drawn with diamond tools,” he says. (For more on the emerging applications of diamond, see Physics Today, March 2022, page 22.)
CVD and HPHT are both commonly used to make gem-quality diamonds, for which color, clarity, and size are important. HPHT is used to make nearly 99% of synthetic diamonds and is the preferred method for industrial applications such as diamond drill bits and grit, for which the diamonds can be smaller and less pure than those used for jewelry. Diamond nanocrystals have also been made via ultrasonic cavitation on graphite and by explosions confined to metal cylinders to produce what’s known as detonation nanodiamonds.
For research and technological applications, CVD is the preferred method, since its layer-by-layer growth allows for the greatest degree of control. To measure magnetic fields (see Physics Today, May 2024, page 12) or build qubits, scientists can aim to create defects from specific atoms, such as nitrogen or silicon. In those highly sensitive applications, even the faintest bit of metal contamination will interfere with the intended use.
Mandal is somewhat doubtful that the liquid-metal catalysis approach will offer enough advantages to get widespread uptake. He says that both CVD and HPHT synthesis work well and have the benefit of several decades of research investment. Billions of carats of synthetic diamond are already produced every year, mostly via the HPHT method.
Ruoff plans to continue exploring liquid-metal catalysis for synthesizing not only diamonds but also materials made of other elements, such as boron and nitrogen. He and his colleagues are still unraveling the mechanisms that led to diamond growth in their experiment. Mixing a metal that doesn’t dissolve carbon, such as gallium, with others that do, such as nickel and iron, may be an important factor. (Iron dissolves diamond so effectively that diamond is not used to cut it.) Hydrogen’s role in the process also remains unclear.
Ruoff says that he hopes the result will inspire others to do basic research in the same vein. “I think that by exploring liquid metals broadly and with a lot of different configurations, new science will be done, even if people don’t make diamond.”
