top of page

The paper was published in Diamond and Related Materials (2022)

Many carbon phases exhibit attractive properties, which further motivates researchers to continue the search for new stable compounds. Some experimental findings suggest the possible existence of an intermediate carbon phase (ICP) between diamond and graphite. Intermediate structures were observed in the process of diamond formation, where products, obtained by high-temperature shock compression of wood charcoal besides nanocrystalline diamond contained an amorphous carbon phase having a density intermediate between the densities of the original charcoal and diamond. In the presence of water, other mechanisms of the diamond-graphite transition with the formation of other intermediate states are possible with a formation of metastable phase of a linear hydrocarbon.

In the present work we report the successful synthesis and investigation of physical properties of the intermediate carbon phase. For the first time, this phase was obtained by the treatment of diamond in the planetary mill. Under conditions of cyclic stresses near the graphite-diamond boundary at the carbon phase diagram at low temperatures the following processes occur: plastic deformation by the mechanical twinning; martensitic phase transition of diamond to ICP; transition of ICP to onions with a radius below ~5 nm.

Figure (a) shows a particle containing not only a diamond but also a fragment of ICP crystal lattice with an interplanar distance of 0.255 nm which forms an angle of 13º with planes (002) of the diamond. This data proves that the diamond was not broken completely but transformed into a new structure and therefore in the process of transformation of the plane (111) of diamond into the plane (001) of graphite or vice versa, the intermediate carbon phase structure is formed.

a) Inverse Fourier transformation. The diamond structure is broken, but it does not change into graphite. Planes with an interplanar spacing of 0.255 nm appear, which form an angle of 13º with planes (002) of the diamond b) side view of the proposed model for ICP with sp3-hybridized atoms concentration 1.6 %. The unit cell is marked by the dashed line. In the inset the enlarged region of crosslink is shown. Vacancy surrounding atoms and interstitial atoms (crosslinks) are marked by green and orange colors, respectively; c) Energy vs volume for the graphite, diamond and ICP with various sp3-hybridized atoms concentration; d) Enthalpy differences of ICP and diamond relative to graphite vs. pressure. In (c,d) data for graphite and diamond are marked by black hexagons and diamonds, respectively, ICP with sp3-hybridized atoms concentration of 1.6, 2.0, 2.8, 4.0, 6.3, 11.1 and 43.8 % are marked by red, orange, yellow, green, cyan, blue, and purple circles, respectively


We proposed that the intermediate interlayer distance between graphite and diamond originates from the crosslinking of sp2-hybridized layers by self-interstitial (Frenkel) defects (Figure (b)). In the proposed model the intimate Frenkel pair defects are located on top of each other throughout the structure.

We consider the energy–volume calculations and the thermodynamic criterion of equal free energies to study the stability of the ICP phase comparable with diamond and graphite. These phases are equilibrated at several volumes and their energy–volume relations are fit to the third-order Birch–Murnaghan equation of state. We estimated bulk moduli values using this equation. It was found that all ICP phases display similar B0 values between 240 and 300 GPa, close to graphite value 258 GPa. The computed total energy as a function of volume is presented in Figure (c).

It was found that the total energy is lower than diamond for the five concentrations of sp3-hybridized atoms until 6.3%. The structures at these concentrations are relatively sparse, with crosslinks separated from each other by 1.97, 1.73, 1,48, 1.25 and 0.99 nm, respectively. Meanwhile, the interlayer distance in the vicinity of crosslinks was found to be 0.26 nm which corresponds well to experimental data.

The covalent bonding of such concentration prevents the layers from separation and does not distort the structure too much. These structures are the most consistent candidates for describing the phase obtained in the experiment. For the lower concentration (less than 1.6 %), the van der Waals repulsion transforms ICP into a graphite structure whereas a higher concentration of crosslinks buckles the layers with inducing the mechanical strain and rises the overall strain energy.

The enthalpy analysis allowed us to conclude that under pressure most favorable ICP phase (sp3-hybridized atoms concentration 1.6 %) become stable with respect to graphite beyond 8 GPa (Figure (d)). We can therefore assume that this phase may be a side product of some of the pressure-induced metastable phase transformations of graphite. On the other hand, the high-density structure (sp3-hybridized atoms concentration 43.8%) becomes more stable only after 70 GPa.

The covalently bound crosslink enhances the out-of-plane stiffness of the structure. We found that at a crosslink concentration range from 1.6 to 6.3 % C33 constant slightly varies about 200 GPa whereas the denser distribution of crosslinks yields C33 value of 358 GPa (x = 44 %). These values are located between graphite and diamond values 49 GPa and 1079 GPa, respectively. Another important property of the phase is the appearance of out-of-plane conductivity. The density of electronic states of all considered ICP structures displays finite states on the Fermi energy which justifies the metallic properties consistent with the observed drop in resistance of irradiated graphite systems

The paper was published in Science 374, 1616-1620 (2021)


This work was done in collaboration with a number of foreign institutes, with the main experiment carried out at NIMS (Tsukuba, Japan) by Prof. D.M.Tang. Using in situ transmission electron microscopy, we applied heating and mechanical strain to alter the local chirality and thereby control the electronic properties of individual single-wall carbon nanotubes. A transition trend toward a larger chiral angle region was clearly observed (Fig. (c)).


(a) Schematics of a CNT intramolecular transistor with local chirality altered by mechanical strain and Joule heating. (b) TEM image of a SWCNT intramolecular transistor with a channel length of ~2.8 nm. (c) Changes of the chiral angles revealing a converging trend toward large angles. (d) Atomic structure of a (10,7) SWCNT with the basic vectors b1 and b2, the chiral angle χ and misorientation angle α for a 5|8|5 defect. (e) Schematics and formation energies of a pair of dislocations, including the sublimation of a carbon dimer and bond rotation steps. (f) Predicted changes of nanotube chiral angles along chirality transitions.


Generally, plastic deformation and chirality transformations of CNTs have been attributed to the dislocation activities. In theory, two mechanisms have been discussed for the nucleation of dislocations in CNTs, i.e. bond flip driven by stress and thermally activated carbon dimer evaporation. The chirality changes due to gliding of dislocations nucleated from bond flip and associated Stone–Thrower–Wales (STW) defects have been studied in detail to predict a decreasing trend for the chiral angles to near zigzag type chirality due to the energetically favorable (0,1) dislocations, which is opposite to the experimentally observed increasing trend for the chiral angles in this work. As mentioned, a “near critical condition” was used in our experiments so that the CNT was exposed at a high temperature from Joule heating and accompanied with a slow elongation in a quasi-static process. Therefore, we considered the orientation dependent formation energy of the dislocations generated from evaporation of carbon dimers (C2) and associated formation of 5|8|5 defects. The 5|8|5 defects can yield 5|7 dislocation cores responsible for the chirality transitions (Fig. (e)). Calculated formation energy of 5|8|5 defects indicated that at a low chiral angle, the (1,0) dislocation was energetically favored. Therefore, nanotubes with originally small chiral angles tend to increase the chiral angle during the transitions. In a high angle region, the formation energies of 5|8|5 defects associated with (1,0) and (0,1) dislocations become close. This explains the fluctuation behaviors of the chiral angles in the high angle region. The ratio of probabilities of forming (1,0) and (0,1) dislocations was calculated to predict the chirality dynamics (Fig. (f)), showing a converging trend toward larger angles, consistent with the experimental observations. Thus, we have studied a method of local chirality modification that allows to realize metal-semiconductor-metal contact in single-walled carbon nanotubes, i.e., to create an intramolecular nanotube-based transistor.

These four days of the "New Carbon Nanomaterials: Ultrathin Diamond Films" conference were wonderful. You have made fascinating talks that arouse interest, curiosity, and motivation for further research. Our modest meeting allowed the participants to exchange opinions, ideas and maybe even to create collaborations. Therefore, we humbly wish that the conference has helped to develop the field of 2D diamonds. Some participants suggested that maybe we may make our conference a regular event. Good idea! We hope that after some time we may meet again to discuss the development of this field of science that interests us all.

bottom of page