Centimeter-sized diamond composites with high electrical conductivity and hardness

Significance

Diamond is the hardest material for industrial use, but it faces poor conductivity and inherent brittleness. Improving the electrical and mechanical properties of diamond simultaneously, particularly in large (centimeter scale) forms, is a grand challenge. Here, we report the successful synthesis of centimeter-sized diamond composites by controlling the graphitization extent of nanodiamond precursors under moderate pressure and temperature conditions. The diamond composites, consisting of interconnected diamond nanograins and few-layer graphene units, exhibit the highest electrical conductivity ever reported and excellent hardness/toughness. Our findings pave the way for realizing conductive and superhard large-sized diamond materials under mild synthetic conditions, facilitating their practical utilization in related industrial applications.

Abstract

Achieving high-performance materials with superior mechanical properties and electrical conductivity, especially in large-sized bulk forms, has always been the goal. However, it remains a grand challenge due to the inherent trade-off between these properties. Herein, by employing nanodiamonds as precursors, centimeter-sized diamond/graphene composites were synthesized under moderate pressure and temperature conditions (12 GPa and 1,300 to 1,500 °C), and the composites consisted of ultrafine diamond grains and few-layer graphene domains interconnected through covalently bonded interfaces. The composites exhibit a remarkable electrical conductivity of 2.0 × 10⁴ S m⁻¹ at room temperature, a Vickers hardness of up to ~55.8 GPa, and a toughness of 10.8 to 19.8 MPa m^1/2. Theoretical calculations indicate that the transformation energy barrier for the graphitization of diamond surface is lower than that for diamond growth directly from conventional sp² carbon materials, allowing the synthesis of such diamond composites under mild conditions. The above results pave the way for realizing large-sized diamond-based materials with ultrahigh electrical conductivity and superior mechanical properties simultaneously under moderate synthesis conditions, which will facilitate their large-scale applications in a variety of fields.

Diamond is a widely utilized material due to its highest hardness, unmatched thermal conductivity, and chemical inertness (1, 2). Despite being the hardest natural material, diamond exhibits limitations in terms of electrical conductivity and toughness. Various strategies, such as nanotwinning and structural architecture (3, 4), have been developed to enhance diamond’s hardness/toughness, yet its electrical conductivity remains relatively low. The major reason is that the localized sp³ covalent bonds in diamond hinder efficient electron transport. While it is acknowledged that the electrical properties of diamond can be tuned by doping (5), practical challenges remain due to the low doping efficiency and the high activation energy of dopants in diamond. Therefore, developing novel diamond-based materials with customized microstructures that allow an exceptional balance of mechanical and electrical properties is highly desired for a broad range of applications.

In contrast to diamond, graphitic carbon exhibits high electrical conductivity and ductility due to the entirely sp²-bonded networks. The introduction of graphitic carbon phases into diamond matrix to create hybrid diamond composites could be a feasible approach that combines the advantages of both components. Various sp²–sp³ mixed hybridized carbon structures have been proposed theoretically (6–9), however, none of them has been experimentally synthesized. Recently, a new class of diamond–graphite nanocomposites (10–14) such as diaphite and Gradia structures with coherent interfaces have been recognized from natural impact meteorite or produced at high pressure and high temperature (HPHT), rather than traditional processing. The mixed sp²-bonded graphite/graphene and sp³-bonded diamond units in these structures contribute to their superior performance. Also, compressing glassy carbon at elevated temperatures (25 GPa, 400 to 1,200 °C) can transform into bulk amorphous carbons with lightweight, hard, and conductive properties (15). Very recently, ultrastrong and conductive carbon composites consisted of ultrafine nanodiamonds embedded in disordered multilayer graphene matrix with incoherent interfaces have been synthesized from glassy carbon at 25 GPa and temperatures from 1,050 to 1,150 °C (16). Despite these efforts, industrial application of these carbon composites is regrettably constrained by the limited sample size, dictated by the harsh synthetic conditions. These critical requirements are imposed by the high energy barriers involved in the transformation of conventional sp² carbon precursors into diamond phases. So far, the realization of highly conductive superhard diamond composites at a centimeter scale remains an elusive goal.

In this work, we present a strategy of surface graphitization of nanodiamonds to achieve centimeter-scale diamond composites using a home-designed sample chamber assembly under moderate pressure and temperature conditions (details in Materials and Methods and SI Appendix, Fig. S1). Through precise control of the graphitization extent of nanodiamonds, the resultant bulk material featured with nanodiamond and few-layer graphene interfaces achieves a remarkable combination of properties, including high Vickers hardness up to 55.8 GPa, fracture toughness of 19.8 MPa m^1/2, and electrical conductivity of 2.0 × 10⁴ S m⁻¹, which are significantly superior than nearly all diamond-based superhard materials reported before. The diamond nanograins and few-layered graphene sheets within the composite are mainly interconnected through covalent bonds, which contribute to the high electrical conductivity and mechanical properties of the composites. These results provide a practical strategy for achieving large-scale bulk diamond composites with simultaneously high hardness, toughness, and excellent electrical conductivity.

Results and Discussion

Theoretical Calculations.

To confirm the validity of the proposed scenario, we performed first-principles calculations on the designed diamond and diamond/graphene models (Fig. 1A, details in Materials and Methods). As shown in Fig. 1B, the diamond/graphene phase is energetically favored than diamond below 12 GPa, indicating that it could be easily synthesized at relatively low pressures. Furthermore, the activation enthalpy barriers from the initial state to the final state (termed H_b1) and that of the reversed process (termed H_b2) under pressure were calculated (Fig. 1C). Specifically, the calculated H_b1 increase monotonically from 0.033 to 0.176 eV/atom with increasing pressure from 0 to 30 GPa, whereas the H_b2 decreases in the same pressure range which agrees with previous studies (14, 17) for the graphite-to-diamond transition. It is worth noting that, below 12 GPa, the enthalpy barriers from diamond to diamond/graphene structure are significantly lower than those of the reversed process. However, above 12 GPa, situation changes, i.e., the formation of diamond is more favorable upon the reversed relative enthalpy. In contrast to diamond growth directly from other sp² carbon materials, these results strongly indicate that the diamond/graphene structure could be formed through the reverse strategy via partially graphitization of diamond at relatively moderate pressure and temperature conditions.

Fig. 1.

Energetics and kinetics for phase transition between diamond and diamond/graphene under pressure. (A) Structural models for diamond termed as initial state (IS), final state (FS) of diamond/graphene, and the transition state (TS). H_b1 and H_b2 represent the enthalpy barriers from the IS to FS and that of the reversed process, respectively. (B) Relative enthalpy per top-most atom for diamond/graphene structure versus pressure relative to cubic diamond. (C) Calculated H_b1 and H_b2 as a function of pressure.

Sample Synthesis and Characterization.

Based on the theoretical calculations above, we performed the HPHT experiments on the nanodiamond precursors (details in Materials and Methods). Fig. 2A shows optical images of six centimeter-sized, black, and opaque samples produced from initial nanodiamonds recovered from 7 and 12 GPa at various temperatures. These well-sintered bulk samples are 13.4 to 14.0 mm in diameter and 1.4 to 2.0 mm in thickness. X-ray diffraction (XRD) patterns of the samples synthesized at different HPHT conditions are shown in Fig. 2B and SI Appendix, Fig. S2. At ambient conditions, the initial nanodiamonds show three main diffraction peaks at ~43.9°, ~75.4°, and ~91.4°, which can be attributed to the (111), (220), and (311) planes of cubic diamond. Besides these peaks, the samples treated at pressures from 7 to 12 GPa and temperatures from 1,100 to 1,500 °C exhibit another peak centered at ~27°. This peak comes from the incomplete structural transition of the nanodiamonds, corresponding to the interlayer spacing of graphene sheets formed within the diamond composites. The slightly reduced interlayer distance (3.30 Å) compared to the bulk graphite indicates the compressed graphene layers in the composite (12, 14). Such XRD features most closely resemble those recently observed in the nanodiamond/disordered multilayer graphene (ND/DMG) composites (16). As the temperature increases under identical pressure, the diffraction peak intensity of the graphene layers is gradually enhanced as opposite to nanodiamond domains, suggesting the increased graphene content at higher temperatures. This significantly differs from the previous sp² carbon precursors where the content of diamond increases with increasing temperatures (14, 16), indicating a gradual graphitization process of the nanodiamonds. When heated to 1,600 °C at 15 GPa, pure diamond is formed since no diffraction sign of graphitic carbons is observed. The relatively low pressure and temperature for the partial graphitization of the nanodiamonds undoubtedly indicate lower energy barriers than those for diamond formation directly from conventional graphitic carbon precursors at HPHT conditions. This further verifies our theoretical results (Fig. 1). Hereafter, for simplicity, samples recovered from 12 GPa and temperatures of 1,300 °C, 1,400 °C, and 1,500 °C are referred to as diamond/graphene (Diaphene) composites and are labeled as Diaphene-1, Diaphene-2, and Diaphene-3, respectively. By the Rietveld refinement method, the percentages of diamond in each diaphene sample are obtained as 62.9%, 54.2%, and 37.4%, respectively (SI Appendix, Fig. S3).

Fig. 2.

Optical photographs, XRD patterns and Raman spectra of the samples recovered from different HPHT conditions. (A) Optical images of the samples synthesized at different HPHT conditions. The scale bar of a grid is 2 mm. (B) XRD patterns and (C) Raman spectra (325 nm excitation) of the samples recovered from various HPHT conditions, compared with the nanodiamond precursor. D and G represents the diffraction peaks of diamond and few-layered graphene, respectively.

As shown in Fig. 2C, the ultraviolet (UV) Raman spectra of the diaphene samples are dominated by two very broad, asymmetric bands between 1,200 and 1,800 cm⁻¹, distinctly different from that of the nanodiamond precursor but similar to those of the diaphite structure (18), indicating a structural transformation of nanodiamonds. These Raman spectra are fitted and deconvoluted into several bands. Two characteristic peaks centered at 1,382 and 1,586 cm⁻¹ are the D and G bands, arising from the breathing mode of hexagonal rings and stretching vibration of sp² atoms in graphene layers (19), respectively. The peak between 900 and 1,300 cm⁻¹ is attributed to T-band, a characteristic signature of sp³-bonded carbons (20). The F band at 1,450 and 1,470 cm⁻¹ corresponds to the vibrational modes of pentagonal aromatic rings and the D’ band is related to the features of double-resonant, defect-related graphene (20, 21). The disappearance of the zone-center phonon line at 1,319.6 cm⁻¹ of nanodiamonds is attributed to the complete relaxation of the k = 0 selection rule of Raman scattering caused by the small size of the diamond crystallites (22). The intensity of G band increases with increasing temperature at certain pressure, while D band gradually decreases, as evidenced by the reduced I_D/I_G intensity ratio for recovered samples (SI Appendix, Fig. S2B). This indicates the gradual transformation of sp³ diamond into sp² carbon units, which is consistent with XRD results and further supported by the X-ray photoelectron spectroscopy (XPS) analysis (SI Appendix, Fig. S2C). For the sample quenched from 15 GPa and 1,600 °C, the diffraction sign of graphitic carbon is not observed, and a characteristic Raman peak at 1,332.4 cm⁻¹ appears, suggesting the formation of cubic diamond.

Microstructure.

To examine the microstructure of the diaphene composites, scanning transmission electron microscopy (STEM) experiments have been conducted. The integrated differential phase contrast STEM (iDPC-STEM) image (Fig. 3A and SI Appendix, Fig. S4) shows that the recovered sample is predominately composed of nanodiamond particles and few-layered graphene. The curved few-layered graphene units are covalently bonded to the ultrafine nanodiamond grains, forming an intertwined architecture within the nanodiamond matrix. The grain size of nanodiamonds ranges from 5 to 10 nm compared to the raw nanodiamonds with an average size of 5.8 nm (SI Appendix, Fig. S5). The high-angle annular dark field-STEM (HAADF-STEM) observations of the diaphene sample (Fig. 3B and SI Appendix, Fig. S6) further reveal that the diamond nanograins are partially surrounded by curve few-layered graphene domains, with an average layer number of 5. This provides strong pieces of evidence that the graphene sheets are generated from the diamond surface. These results show that the synthetic samples consist of diamond and graphene nanoscale units connected with each other, consistent with the XRD and Raman results, forming a bicontinuous phase network. This hybrid structure at the interface produced by the diamond-to-graphene transition is in strong contrast to those of previously proposed carbon materials, such as type 2 diaphite in natural meteorolite (11), the diamond-graphite composite from compressed multi-walled carbon nanotube fibers (12), and Gradia from the graphite-to-diamond transition (14). Note that it resembles the recently reported incoherent interface in a ND/DMG composite obtained from the glass carbon-to-diamond transition. Such unique interfacial covalent bonding may enable the diaphene to achieve an excellent combination of high toughness and hardness (16). To further demonstrate the bonding states of the diaphene at the interface, the carbon K-edge electron energy loss spectroscopy (EELS) line scan was acquired along a linear path across the nanodiamond and graphene regions (Fig. 3B). For the nanodiamond, the averaged EELS spectra exhibit diamond features with a dominating σ* peak at 292 eV and a dip at 302 eV (23). In the case of the few-layer graphene, the peak at ~284 eV representing the 1 s⟶π* orbital transition is associated with sp² graphitic carbon. From nanodiamond to few-layer graphene domain, the relative intensity of the π* peak is obviously increased (Fig. 3C). This indicates that the nanodiamond domains are randomly connected by few-layered graphene units at the interface through mixed sp²–sp³ bonding in the composite. Fig. 3D is the high-resolution HAADF image of a nanoparticle from a focused ion beam (FIB) sample. The distinctive dumbbell atomic structure indicates that the particle is a nanodiamond, captured along the [110] zone axis. A visual representation of the diamond’s dumbbell model when viewed from the [110] direction is shown in Fig. 3E. In Fig. 3F, a corresponding simulated HAADF image is depicted, which is consistent with the features observed in Fig. 3D.

Fig. 3.

Microstructures of the samples recovered from 12 GPa and 1,300 °C. (A) The integrated differential phase contrast scanning transmission electron microscopy (iDPC-STEM) image of the sample, where the contrast of diamond (“D”) and few-layered graphene sheets (“G”) can be observed. (B) High-resolution high-angle annular dark field STEM (HAADF-STEM) image, showing the nanodiamond particles are surrounded by curved few-layer graphene. (C) The extracted C-K edge EELS spectra from line scan area (the white dashed arrow in B). The red, orange, and black lines are the EELS signals acquired from nanodiamond domains, interfacial area, and few-layered graphene, respectively. (D) High-resolution HAADF image of a typical nanodiamond particle, where the dumbbell structure of diamond along the [110] zone axis can be seen. (E and F) The atomic model and the corresponding simulated HAADF-STEM image of the diamond from the [110] direction, matching well with the experimental image in (D).

The selection of nanodiamonds as precursors plays a key role in the preparation of diaphene samples under moderate HPHT conditions. Previous studies have demonstrated that, under ambient conditions, diamond is energetically less stable compared to graphite (24). Thus, the transformation from diamond to graphite requires a lower energy barrier than the formation of diamond from sp² carbon precursors (17), as confirmed by our theoretical calculations (Fig. 1C). In contrast with high synthetic conditions of ND/DMG composites from sp² glass carbon when heated at 25 GPa, starting from nanodiamond precursors results in the facile synthesis of the bulk diamond composites. A possible transition mechanism from nanodiamond to diaphene through diamond surface graphitization is proposed. The graphitization process initiates at the surface layer of ND particles under HPHT conditions. Initially, the breaking of some sp³ bonds between adjacent layers leads to the nucleation of graphitic seeds. Subsequently, the graphitization spreads throughout the whole surface layer, simultaneously extending into the inner core by breaking nearest-adjacent bonds to form curved graphene fragments, reducing surface energy (25). These fragments between neighboring nanodiamonds merge to prevent the formation of dangling bonds. During this transformation, a diamond/graphene interface is formed, which is consistent with our HAADF observations (Fig. 3B and SI Appendix, Fig. S6). As the synthesis temperature increases, the graphitic phase penetrates deeper in a layer-by-layer fashion from the surface to the inner core, resulting in an increased number of graphene layers, as evidenced by the enhanced diffraction peak intensity (Fig. 2B). Some disordered amorphous carbon fragments may exist from small nanodiamond particles within the diamond matrix for the diaphene samples. At higher pressure and temperature (15 GPa, 1,600 °C), the nanodiamonds are transformed back to pure diamond, consistent with the synthesis conditions of diamond from various carbon precursors (26, 27).

Mechanical and Electrical Properties.

To investigate the mechanical properties of the diaphene bulk samples, Vickers hardness measurements were performed on the pre-polished surfaces of these samples. Fig. 4A shows the Vickers hardness (H_V) of Diaphene-1, Diaphene-2, and Diaphene-3 as a function of applied loads, which gives asymptotic H_V values of 55.8 ± 3.8, 46.7 ± 2.2 and 42.5 ± 2.7 GPa, respectively, at 9.8 N load. These hardness values are higher than that of cubic boron nitride (cBN, 30 to 43 GPa) single crystal (28), and the hardness for Diaphene-1 is comparable to that of natural diamond (111) plane (62 GPa). Note that the sample quenched at 15 GPa and 1,600 °C achieves a hardness of 116.8 ± 3.6 GPa, which are approximately comparable with that of natural diamond (110) plane (3) (SI Appendix, Fig. S7). For comparison, the Knoop hardness (H_K) values of the diaphene samples at a load of 4.9 N are 49.5 ± 2.2, 41.5 ± 1.8, and 40.2 ± 1.1 GPa, respectively (SI Appendix, Fig. S8). The slightly smaller H_K values compared to the H_V values can be attributed to the elastic recovery around the Knoop indentation especially for hard materials (29). The densities of our diaphene composites are calculated to be 2.7, 2.9, and 3.0 g cm⁻³, respectively, according to the mass and cylindrical sample volume. These density values are lower than that of diamond (3.52 g cm⁻³), which is consistent with the volume percentages of each composition from the XRD patterns.

Fig. 4.

Mechanical and electrical properties of the diaphene bulk samples. (A) Vickers hardness (Hv) of the diaphene composites as functions of applied loads. Error bars show the SD. Inset shows optical image of the indentation of Diaphene-1 at 9.8 N. The dashed line represents lowest Hv threshold (40 GPa) for superhard materials. (B) Hardness versus electrical conductivity at room temperature of the diaphene composites and other materials such as cBN, ND/DMG composites (16) and amorphous carbon phases including compressed glassy carbon (Com. GC) (15), amorphous carbon (a-C) (30, 31), and tetrahedral amorphous carbon films (ta-C) (32). The Knoop hardness of the ND/DMG composite is used here.

We also measured the fracture toughness (K_Ic) from the radial cracks at three loads of 9.8, 19.6, and 29.4 N, respectively. At 9.8 N, no crack was observed in all three diaphene composites (SI Appendix, Fig. S9 A–C), suggesting a high fracture toughness. The calculated K_Ic values for the three samples at high load of 19.6 N ranged from 10.8 to 19.8 MPa m^1/2 (SI Appendix, Fig. S9 D–F and Table S1), which is almost sevenfold that of single crystal cBN (2.8 MPa m^1/2) (33), more than four times that of natural and synthetic diamond (3.4 to 5.0 MPa m^1/2) (34), slightly higher than that of nt-diamond (9.7 to 14.8 MPa m^1/2) (3), and comparable to that of nt-diamond composite (26.6 MPa m^1/2) (4). As we know, diamond has a low fracture resistance because of crack propagation along cleavage planes. The enhanced fracture toughness of our diaphene is intimately attributed to the presence of graphene layers connected with the nanodiamond matrix, which can absorb or deflect incipient crack formation. The achieved hardness and toughness are summarized to compare with existing tough alloys and common tooling materials (4, 35) (SI Appendix, Fig. S10), indicating the unique combination of high hardness and improved toughness in our diaphene composite.

The electrical transport properties of the diaphene composites were investigated via a standard four-probe method in the temperature range of 2 to 300 K. The temperature-dependent electrical resistivities of the diaphene show a typical semiconducting characteristic (SI Appendix, Fig. S11). Notably, the electrical conductivity of our diaphene samples at room temperature is in the range of 18,248 to 20,450 S m⁻¹. Moreover, the temperature dependence of the charge carriers and mobilities for the diamphene-1 is discussed (SI Appendix, Fig. S12). Fig. 4B compares the hardness and room temperature electrical conductivity of the diaphene composites with other superhard materials. Remarkably, the diaphene simultaneously possesses exceptionally high electrical conductivity and impressive hardness. Its electrical conductivity far surpasses that of cBN (28) and various amorphous carbon materials reported (15, 30–32) and is nearly one or two orders of magnitude higher than the ND/DMG composites (670 to 1,240 S m⁻¹) (16). Furthermore, boron-doped diamond (BDD) single crystals display good electrical conductivity, dependent on the concentration of boron dopants in the diamond structure. However, the conductivities of BDD single crystals (13.9 to 4,545 S m⁻¹) are lower than those of our diaphene composites (36). To the best of our knowledge, these centimeter-sized diaphene composites exhibit a unique combination of electrical conductivity and hardness that outperforms the known carbon composites and BDD single crystals. As discussed above, the diaphene sample consists of few-layered graphene domains embedded with the nanodiamond matrix. The two components are mainly connected through random covalent bonds. As the nanodiamond is acted as an insulator, the remarkable electrical conductivity observed in our diaphene composites can be attributed to the presence of few-layer graphene domains that create conducting channels within the insulating diamond matrix (10). Due to van der Waals interactions between multilayer graphene, the efficiency of charge transfer across multilayer graphene is lower than that within the few-layer graphene. In the case of diaphene, the few-layer graphene plays a crucial role in enhancing both electrical conductivity and fracture toughness, while the high hardness inherits raw nanodiamonds.

In addition, the thermal stability of the diaphene composites was determined in air by thermalgravimetric (TG) and differential scanning calorimetry (DSC) measurements, giving an onset oxidation temperature of 1,005 K. This temperature value is slightly lower than that of natural diamond (~1,040 K) (3) but much higher than those reported for nanodiamonds, diamond-like carbon (DLC) films (37), chemical vapor deposition (CVD) diamond (38), nano-polycrystalline diamond (NPD) (39), and paracrystalline diamond (p-D) (39) (SI Appendix, Fig. S13). The enhanced thermal stability should be related to the diamond nanograins, delaying the graphitization of diamond (3). These excellent combinations of high hardness, fracture toughness, electrical conductivity, and oxidation resistance endow the large-sized bulk diaphene with potential applications in various fields.

Conclusion

In conclusion, centimeter-sized diamond composites (one order of magnitude larger than those reported previously) have been synthesized by the diamond-to-graphite transition under moderate pressure and temperature conditions. The resultant diaphene materials consist of an ultrafine nanodiamond matrix connected with few-layer graphene. By controlling the graphitization extent of diamond surface varying the pressure and temperature, the diaphene exhibits an excellent combination of high Vickers hardness (42.5 to 55.8 GPa), ultrahigh toughness (10.8 to 19.8 MPa m^1/2), and highest electrical conductivity (2.0 × 10⁴ S m⁻¹) among nearly all superhard carbon materials. This work provides a feasible strategy to prepare high-performance diamond composites at the centimeter scale under industrial synthetic conditions, presenting great potential for large-scale applications in the future.

Materials and Methods

Sample Synthesis.

Nanodiamond powders (99%) with average sizes of ~5.8 nm purchased from Shanghai Aladdin Biochemical Technology Co., Ltd. were annealed at 450 °C in air for 1 h to remove any amorphous and graphitic sp²-bonded carbons. High-pressure and high-temperature experiments were performed using either a China-type cubic large-volume press (TB750) with a newly designed sample assembly or a Walker-type multi anvils press with a press load of 1,000 tons. Based on the domestic cubic press, a second-stage high-pressure apparatus, designated as the 6-2 type, has been developed to generate high pressure up to 15 GPa, along with a spacious sample cavity exceeding 10 mm. The components of the sample assembly block include pyrophyllite tube, ceramic mold, graphite tube, salt tube, MgO cup, steel cap, and second-stage anvil, etc. In the assembly, the second-stage anvils comprise a pair of conical-cylinder indenters constructed from polycrystalline diamond, each featuring a bottom edge length of 30 mm. The sample chamber has a diameter of 15.5 mm and a height of 4.0 mm. Temperature was monitored using a type-C thermocouple, while pressures were estimated from previously calibration curves. The starting materials were compressed to the target pressure and then heated to the target temperature at a rate of 100 °C/min and held for 10 to 20 min. Afterward, the sample was cooled down to room temperature with a speed of 60 °C/min and then quenched slowly at a rate of 2 GPa/h. The recovered samples were about 13.4 to 14.0 mm in diameter and 1.4 to 2.0 mm in height. For the sample synthesized at 15 GPa and 1,600 °C, the standard 10/5 sample assembly with chromium-doped MgO octahedra (10 mm edge length of octahedron and 5 mm truncated edge length of tungsten carbide anvils) in a Walker-type press was used as the pressure medium and the rhenium capsule were used as the heater. The pressure was calibrated in advance with Bi, ZnTe, ZnS, and GaAs and the temperature was monitored using a type-C thermocouple.

X-ray Diffraction, Raman, and X-ray Photoelectron Spectroscopy.

X-ray diffraction patterns were required by X-ray diffractometer (Rigaku, Smartlab) with Cu Kα radiation source (λ = 1.5406 Å). The volume percentages of diamond were determined from the XRD data via Rietveld refinement method. The ultraviolet (UV) Raman spectra were measured at room temperature using a Raman spectrometer (LabRAM HR Evolution, HORIBA Jobin Yvon) equipped with a 325 nm laser as the excitation source. The chemical composite of the sample surface was characterized by X-ray photoelectron spectroscopy (Thermo Fischer ESCALAB 250Xi) with a monochromatized Al Kα radiation. Survey spectra were acquired at 20 eV pass energy with 0.1 eV step.

Transmission Electron Microscopy.

The TEM samples were prepared via focused ion beam (FIB, Helios G4, FEI). The scanning TEM (STEM) measurements were performed by a spherical aberration-corrected scanning transmission electron microscope (FEI Themis Z) operated with an accelerating voltage of 300 keV. The HAADF-STEM images were acquired with a convergence angle of 25 mrad and collection angle from 37 to 94 mrad. With the same convergence angle, the iDPC images were acquired with the collection angle from 5 to 24 mrad. Electron energy loss spectroscopy (EELS) spectra were acquired to determine the bonding sates in the samples. All EELS data processing, including calibration of the zero-loss-peak position, background subtraction, and normalization was handled by Gatan Microscopy Suite software.

Mechanical Measurement.

The standard Vickers/Knoop hardness and fracture toughness (K_IC) were measured on the polished sample surface using a microhardness tester (KB5-BVZ) with a diamond indenter for a dwell time of 10 s. Vickers hardness (Hv) was determined from the equation: H_v = 1854.4F/L², where F (N) is the applied load and L (µm) is the average diagonal length of the Vickers indentation. Knoop hardness (H_K) was determined from H_K = 14,229 F/d², where F (N) is the applied load and d (µm) is the longer diagonal length of the Knoop indentation. Five indentations were obtained at each load to calculate an average hardness value. K_IC was determined from K_IC = 0.016(E/H_v)^0.5F/C^1.5, where F (N) is the applied load, C (μm) is the average length of the radial cracks measured from the indent center, and E (1,140 GPa) is Young’s modulus of diamond. At 9.8 N, no crack was observed in the diaphene samples. Five data points were used to determine the average K_IC value at 19.6 N.

Density Testing.

The densities of the diaphene samples were measured directly form the mass and cylindrical sample volume.

Thermal Stability.

The thermal stability of the recovered samples was measured by thermogravimetry (TG) and differential scanning calorimetry (DSC) in air using NETZSCH STA 449F3 with a heating rate of 5 K/min in the temperature range of 300 to 1,373 K.

Electrical Measurement.

The temperature dependence of electrical resistivity (ρ) of the diaphene samples was measured in the temperature range between 2 and 300 K using the standard four-probe method with the Physical Property Measurement System (PPMS, Quantum Design). The Hall effect was also measured in a magnetic field of 9 T between 2 and 300 K by a four-probe method. Four platinum wires (Φ 20 μm) were attached to the surface of the polished sample with silver paste. The electrical conductivity σ was calculated from the resistivity according to σ = 1/ρ. The Hall coefficient (R_H), carrier concentration (n), and mobility (μ_H) were obtained from the relationships: R_H = ρ_xy/H, R_H = −1/(ne), and μ_H = 1/(ρne), respectively. The activation energy was calculated by fitting the Arrhenius relationship of ρ = A exp (E_a/K_BT), where E_a is the activation energy, K_B is the Boltzmann constant, and T is the temperature.

Computational Methods.

The first-principles calculations were based on the density functional theory (DFT) (40) as implemented in the Vienna ab initio simulation package (VASP) (41, 42) with the projector augmented wave (PAW) method (43, 44) and revised-Perdew−Burke−Ernzerhof (RPBE) scheme (45) for the exchange−correlation functionals. The wave functions were employed in a plane wave basis set with an energy cutoff of 400 eV. The zero damping DFT-D3 method of Grimme (46) was adopted in describing van der Waals interactions. The diamond (111) surface was simulated by a six-layer slab model with a vacuum of around 15 Å between the slab images, where each C atom in the bottom-most layer was passivated by one H atom to maintain the sp³ nature of the nanodiamond. Specifically, the initial state (IS) structure is simulated by a 2 × 2 diamond (111) slab supercell model. The final state (FS) structure is based on the first one, nevertheless with the top-most carbon atoms are cleaved to form a flat graphene over layer. The k-space integration was done in a 2 × 2 supercell with a 5 × 5 × 1 k-point mesh. The convergence criterion for the self-consistency process was set to 10⁻⁵ eV between two ionic steps. All atoms except the bottom two layers were allowed to relax along the calculated forces until all the residual force components were less than 0.01 eV/Å. The well-recognized climbing image nudged elastic band (CI-NEB) method (47) was used to identify the minimum energy paths (MEPs) and transition state. The enthalpy change was calculated as: H = U + PV, where U is the total energy obtained directly from DFT calculations at zero temperature; P represents the constant external pressure applied on the surface atoms of the slab, as implemented in the improved VASP code developed by our group (48–50) particularly for low-dimensional and finite system where there is no periodic boundary condition, and V is the volume of the simulated system, respectively.

Supplementary Material

Appendix 01 (PDF)

Data, Materials, and Software Availability

The data that support the findings of this study are available within the paper and SI Appendix.