Catalyst-free synthesis of transparent, mesoporous diamond monoliths from periodic mesoporous carbon CMK-8 Li Zhanga, Paritosh Mohantyb, Neil Coombsc, Yingwei Feia, Ho-Kwang Maoa,1, and Kai Landskronb,1 a Geophysical Laboratory, Carnegie Institution of Washington, Washington, DC 20015; bDepartment of Chemistry, Lehigh University, Bethlehem, PA 18015; and cDepartment of Chemistry, University of Toronto, Toronto, ON, Canada M5S3H6
Contributed by Ho-Kwang Mao, June 16, 2010 (sent for review March 1, 2010)
high pressure ∣ nano
D
iamond is a material with superlative properties. It is the material with the greatest hardness and the highest thermal conductivity (1). It also has remarkable optical properties such as high refractive index and high dispersion, as well as interesting electrical properties, such as semiconductivity and superconductivity when doped with boron (2–4). Despite its great hardness, single-crystalline diamond has only a fair toughness due to the existence of well-defined crystal planes (1). This shortcoming can be overcome by the noncatalytic synthesis of nanopolycrystalline diamond from graphite and other carbon sources (5). Similarly to diamond single crystals, nanopolycrystalline diamond is obtained as a monolithic macrostructure. The monolithic macrostructure is a result of strong bonds at the grain boundaries between the nanocrystals. Like single-crystalline diamond, nanopolycrystalline diamond is optically transparent. In stark contrast to single-crystalline diamond, the toughness of nanopolycrystalline diamond is much greater due to the absence of well-defined crystal planes going through the monoliths. Moreover, a higher isotropy of hardness can be achieved for nanopolycrystalline diamond (5). Therefore, high-purity nanopolycrystalline diamond has great potential for industrial use as cutting tools and abrasion-resistant materials as well as use in diamond anvil cells. However, the synthesis of nanopolycrystalline diamond from graphite requires temperatures >2;000 °C at pressures >12 GPa, which is undesirable for practical applications due to the high energy consumption of the synthesis process and the small reactor sizes that must be used due to the limited heat-capacity of the multianvil assembly (6). Despite immense efforts to synthesize diamond from different types of carbon precursors, complete noncatalytic diamond formation at high pressure was so far only be achieved at 1,600 °C or above (7). Below www.pnas.org/cgi/doi/10.1073/pnas.1006938107
this temperature only partial conversion can be achieved even at long reaction times (1, 7, 8). Periodic mesoporous carbons are the only class of carbon materials that exhibit periodically ordered nanopore systems (9). The mesopores of periodic mesoporous carbons have typical diameters between 2 and 10 nm. The channel wall diameters are in the same size regime. In addition, they exhibit very high pore volumes of up to 1.3 cm3 g−1 and surface areas of ca. 1300 m2 g−1 . These structural features of periodic mesoporous carbons suggest high reactivity at high pressure and noncatalytic transformation into nanopolycrystalline diamond at mild reaction temperatures. Results and Discussion To test this hypothesis, we chose periodic mesoporous carbon CMK-8 with Ia-3d cubic symmetry (gyroid structure) as a precursor material (10). The high-pressure experiments were performed using an 8∕3 cell assembly in a multianvil apparatus. The samples wree compressed to 21 GPa and heated at temperatures of 1,100, 1,300 and 1,600 °C for 360, 180, and 60 min, respectively. The starting material was loaded into a MgO capsule and heated with a rhenium tube heater. Sample temperatures were measured with a W5%Re–W26%Re thermocouple. Although no correction was made for the effect of pressure on thermocouple electromotive force (emf), temperature uncertainties are expected to be within 25 °C. The temperature gradient over the length of the sample chamber (approximately 1.4 mm) is approximately 50 °C. Complete transformation of periodic mesoporous carbon CMK-8 into diamond was observed when periodic mesoporous carbon CMK-8 was heated to 1,300 °C. The diamond was obtained as a transparent monolithic material with a brown color (Fig. 1A). Analysis of the X-ray diffraction (XRD) data confirmed the presence of cubic diamond (Fig. 1B). The reflections can be indexed as 111, 220, and 311 reflections. The monolithic, transparent nature of the obtained product suggests that nanopolycrystalline diamond has been obtained. Apparently the activation energy for the nucleation of transparent diamond from periodic mesoporous carbon is considerably lowered. This behavior can be attributed to the highly ordered mesopore system, the high porosity (pore volume ¼ 1.338 cm3 g−1 ) and the high surface area (ca. 1250 m2 g−1 ) of the periodic mesoporous carbon, which facilitates the nucleation of diamond crystals and leads to complete transformation. The more ordered porosity and better defined surfaces of the periodic mesoporous carbons in comparison to traditional amorphous carbon sources leads to a more homogeneous (and thus complete) transformation pathway into diamond. Traditional amorphous carbons exhibit more heterogeneous and disordered surfaces and carbon domains with presumably a wider range of Author contributions: K.L. designed research; L.Z., P.M., and N.C. performed research; L.Z., P.M., N.C., Y.F., H.-K.M., and K.L. analyzed data; and L.Z. and K.L. wrote the paper. The authors declare no conflict of interest. 1
To whom correspondence may be addressed. E-mail:
[email protected] or
[email protected].
This article contains supporting information online at www.pnas.org/lookup/suppl/ doi:10.1073/pnas.1006938107/-/DCSupplemental.
PNAS ∣ August 3, 2010 ∣ vol. 107 ∣ no. 31 ∣ 13593–13596
CHEMISTRY
We report on the synthesis of optically transparent, mesoporous, monolithic diamond from periodic mesoporous carbon CMK-8 at a pressure of 21 GPa. The phase transformation is already complete at a mild synthesis temperature of 1,300 °C without the need of a catalyst. Surprisingly, the diamond is obtained as a mesoporous material despite the extreme pressure. X-ray diffraction, SEM, transmission electron microscopy, selected area electron diffraction, high-resolution transmission electron microscopy, and Z-contrast experiments suggest that the mesoporous diamond is composed of interconnected diamond nanocrystals having diameters around 5–10 nm. The Brunauer Emmett Teller surface area was determined to be 33 m2 g−1 according Kr sorption data. The mesostructure is diminished yet still detectable when the diamond is produced from CMK-8 at 1,600 °C and 21 GPa. The temperature dependence of the porosity indicates that the mesoporous diamond exists metastable and withstands transformation into a dense form at a significant rate due to its high kinetic inertness at the mild synthesis temperature. The findings point toward ultrahard porous materials with potential as mechanically highly stable membranes.
Fig. 1. (A) The optical microscopy image of transparent nanopolycrystalline diamond synthesized at 1,300 and 1,600 °C. (B) Powder X-ray diffraction patterns of diamond synthesized at 21 GPa and 1,300 and 1,600 °C.
reactivity. Presumably, in traditional amorphous carbons the more reactive carbon domains transform directly into diamond whereas the less reactive ones graphitize first and resist further transformation into diamond even at long reaction times. Consequently, the transformation remains incomplete. To investigate the microstructure of the transparent diamond in greater detail, we performed transmission electron microscopy (TEM) and SEM. Surprisingly, TEM data revealed a mesostructure in the diamond material (Fig. 2A). Selected area electron diffraction of the mesostructured diamond particles exhibit concentric diffraction rings that are expected for nanopolycrystalline materials (Fig. 2B). This is consistent with the transparent monolithic nature of the product. The selected area electron diffraction (SAED) pattern can be indexed and reveal d-spacings of 0.2068, 0.1261, and 0.1074 nm that can be attributed to the cubic diamond lattice. The presence of diamond in the mesostructured material was further confirmed by high-resolution TEM (HRTEM). The image (Fig. 2C) show the presence of lattice fringes attributable to the diamond structure. The crystalline areas in the HR-TEM image have an approximate diameter of around 10 nm and occur in distances of around 10 nm, which corroborates the presence of a mesostructured diamond material. SEM data provides further evidence for the mesostructured nature of the diamond material (Fig. 2 D and E). The images reveal a nanogranular structure with grain sizes in the order of 10 nm. This observation is in accordance with the nanopolycrystalline nature of the diamond material. To provide further evidence for the presence of the mesostructure in the SEM-imaged particle, the same particle was subjected to Z-contrast (high angle annular dark 13594 ∣
www.pnas.org/cgi/doi/10.1073/pnas.1006938107
Fig. 2. TEM (A), SAED (B), HR-TEM (C), SEM (D and E), and Z-contrast image (F) of mesoporous diamond.
field) imaging. This experiment resulted in the observation of strong electronic contrast at the mesoscale showing that TEM and SEM observations are in accordance (Fig. 2F). The Brunauer Emmett Teller (BET) surface area has been calculated from Kr sorption data of a 0.6 mg sample that has been produced from three high-pressure experiments. The calculation revealed a surface area of 33 m2 g−1 . This surface area is fairly low yet significant. Similar surface areas have been observed also for other mesoporous materials with crystalline channel walls (11, 12). The relatively low surface area can be explained by the lower surface roughness and surface energy of porous materials with crystalline channel walls. Overall, the data corroborates that a porous diamond material has been obtained. The data suggests that diamond nanograins are interconnected to form a mesoporous nanopolycrystalline diamond material. The possibility that the mesostructured product is a mesostructured composite of a sp2 hybridized form of carbon and diamond is very unlikely because of the very strong electronic contrast observed in the TEM images and the optical transparency of the product. The presence of another crystalline phase (e.g., graphite) in addition to diamond is excluded by XRD, HR-TEM, and SAED images that clearly show diamond as the only crystalline phase. To confirm that no other element but C is present in the product material we performed EDX experiments that showed that only carbon is present. Moreover, the significant surface area of the diamond material suggests porosity. The formation of mesoporous diamond at an extreme pressure of 21 GPa is surprising because high pressure has a very strong Zhang et al.
Fig. 3. TEM (A) and SAED (B) of diamond product obtained at 21 GPa and 1,600 °C.
Zhang et al.
The diamond is obtained as monolithic optically transparent material, which is in accordance with the nanopolycrystalline microstructure. In contrast to the product obtained at 1,300 ° C, which was brown, the diamond material exhibited only a light-yellow color (Fig. 1A). The lighter color of the diamond at increased temperature may reflect a smaller nitrogen content in the diamond material (18). A smaller nitrogen content at higher temperature is plausible due to the higher nitrogen fugacity associated with higher temperature. It has been reported that the low-pressure/high-temperature annealing in a hydrogen environment can significantly enhance the optical properties of chemical vapor deposition (CVD) single-crystal diamond (19). We suggest that the similar treatment might apply to nanopolycrystalline diamond. This may provide monolithic colorless diamond materials that can be synthesized at unprecedented low temperatures. To probe for the effect of pressure on the phase transformation temperature, we performed an additional experiment at 14 GPa and 1,300 °C. The experiment yielded graphite and diamond coexisting in two spatially separate, opposite areas inside the capsule, likely due to a small temperature gradient (approximately 50 °C) according to light microscopy and X-ray powder diffraction of each area (Fig. S1). The observation of spatially clearly separate phases across the small temperature gradient in the capsule further supports that a sharp phase transformation point exists for the transformation of periodic mesoporous carbons into diamond. Furthermore, the result indicates that the phase transition temperature of periodic mesoporous carbon is only weakly pressure-dependent in the pressure range between 14 and 21 GPa. The recovered diamond exhibits snow-white color but is not optically transparent. The snow-white color indicates low nitrogen content compared to the diamond synthesized at 21 GPa and 1,300 °C. This can be explained by a higher nitrogen fugacity at lower pressure. The nontransparency suggests a significantly different microstructure. Indeed, SEM of the product revealed that the diamond crystallizes in a platelet like morphology. The platelets appear to be loosely packed and have diameters of several hundered nanometers and thicknesses in the order of tens of nanometers (Fig. S2). Diamond platelets have been reported for diamond synthesis by CVD (20–22). However, the growth of platelet-shaped diamond crystals at high pressure has been so far only observed when a periodic mesostructured silica/carbon composite was used as the reactant (23). This suggests that the presence of a mesostructure in a carbon-based material can significantly alter the crystal morphology of the diamond material obtained at high pressure. Conclusion In conclusion, transparent, monolithic, mesoporous diamond can be directly prepared from periodic mesoporous carbon CMK-8 at 21 GPa in multianvil high-pressure apparatuses. The diamond material is comprised of diamond nanograins that are interconnected to form a mesoporous structure. The observation of the formation of a porous high-pressure phase at high-pressure conditions is counterintuitive, but becomes plausible considering elastic strain effects, crystallization-induced volume shrinkage, the high kinetic inertness of diamond at high temperature, and the possible formation of superhard graphite-like intermediates. The scalability of multianvil apparatuses is routinely used for the production of diamond and should also allow for the production of larger quantities of the reported materials. The reduced synthesis temperature suggests reduced energy cost and further facilitation of scale-up. The monolithic, nanopolycrystalline structure suggests super hardness and super toughness in a porous material. The weak mechanical properties of porous materials limit many of their applications. For example, membrane applications require monolithic structures of high and uniform porosity. Their performance is limited by the back-pressure that can be applied, PNAS ∣ August 3, 2010 ∣
vol. 107 ∣
no. 31 ∣
13595
CHEMISTRY
tendency to eliminate porosity. Even at ambient pressure the synthesis of a porous material is challenging and special conditions, most importantly templating strategies, must be applied to produce a porous material (13). Possibly, a superhard graphite-like intermediate prevents the full collapse of the mesoporous carbon during pressurization and subsequent heating, and diamond is able to crystallize as a porous material. Superhard forms of graphite are known to form from both carbon nanotubes and polycrystalline graphite at high pressure (14, 15). The volume shrinkage associated with the phase transformation into diamond may further support the formation of a porous structure. Elastic strain in a partially or fully collapsed intermediate structure can serve as an additional explanation for the formation of the mesoporous diamond. Upon crystallization (phase transformation) the elastic modulus of the elastically collapsed material is expected to increase dramatically, which may cause deformed or collapsed mesopores to reform. Elastic, reversible deformation of mesopores at high pressure is a demonstrated effect in periodic mesoporous silicas at high pressure (12, 16). It is reasonable to assume that this effect can extend to mesoporous carbons. The mild reaction temperature of 1,300 °C can explain why the mesoporous diamond is kinetically inert enough to withstand transformation into a completely dense form at a significant rate. To test whether a complete transformation of periodic mesoporous carbon into diamond can take place at temperatures lower than 1,300 °C we have performed an experiment at 1,100 °C and 21 GPa. However, under these conditions only amorphous carbon was observed according to XRD. This indicates that a sharp phase transition temperature exists at which complete transformation occurs. The absence of product crystallinity at 1,100 °C and 21 GPa suggests that the diamond formation at 1,300 °C might occur without intermediate graphitization of the carbon phase. This type of transformation mechanism is typical for nongraphitic carbons, such as quasiamorphous soot (17). A further experiment was performed at 21 GPa and 1,600 °C to explore whether the mesopores would still form at higher temperatures. XRD of the product shows that crystalline diamond has formed (Fig. 1B). TEM investigations demonstrate again the presence of a mesostructure. However, the porosity appears to be diminished in comparison to the diamond obtained at 1,300 °C (Fig. 3). This indicates that a transformation into a completely dense structure is accelerated due to the higher synthesis temperature. SAED investigations show well-resolved diffraction rings that can all be attributed to diamond (Fig. 3). The SAED pattern demonstrates that diamond is the only crystalline phase present. The observation of rings rather than individual diffraction spots provides additional evidence for the nanopolycrystalline nature of the diamond material. The combined TEM and SAED data suggests that the electronic contrast observed in TEM stems from mesoporosity rather than the presence of a second phase. The absence of a second phase is furthermore supported by EDX that shows C as the practically only present element.
which is limited by the mechanical strength of the membrane. Our findings therefore suggest high potential for the development of mechanically ultrastable membranes. Materials and Methods The high-pressure experiments were conducted in a multianvil apparatus using the 8∕3 assembly at the Geophysical Laboratory. The sample was loaded into an MgO capsule of 0.8 mm inner diameter and 1.3 mm length. The sample was heated with a tube Re heater. The assembly was placed into a hole in the center of a Cr2 O3 doped MgO octahedron with an edge length of 8 mm. Each of the eight cubes has a truncated corner (truncated edge length 3 mm) that rest against the face of an octahedron. Pressure was increased first with a rate of 2 GPa∕hour and then temperature was increased to the peak value with a rate of 150 °C∕ min. The maximum temperature in each run was maintained up to 180 min and was then quenched by turning off the electric power supply. The pressure was released with a rate of 2 GPa∕hour. Sample temperatures were measured with a W5%Re–W26%Re thermocouple. Although no correction was made for the effect of pressure on thermo1. Sumiya H, Irifune T (2008) Microstructure and mechanical properties of ultra-hard nano-polycrystalline diamond. SEI Tekunikaru Rebyu 172:82–88. 2. Tshepe T, Kasl C, Prins J, Hoch M (2004) Metal-insulator transition in boron-ionimplanted diamond. Phys Rev B 70:245107/1–245107/7. 3. Klein T, et al. (2007) Metal-insulator transition and superconductivity in boron-doped diamond. Phys Rev B 75:165313/1–165313/7. 4. Ekimov E, et al. (2004) Superconductivity in diamond. Nature 428:542–545. 5. Irifune T, Kurio A, Sakamoto S, Inoue T, Sumiya H (2003) Materials: Ultrahard polycrystalline diamond from graphite. Nature 421:599–600. 6. Irifune T, Sumiya H (2004) Nature of polycrystalline diamond synthesized by direct conversion of graphite using Kawai-type multianvil apparatus. New Diam Front C Tec 14:313–327. 7. Le Guillou C, Brunet F, Irifune T, Ohfuji H, Rouzaud J (2007) Nanodiamond nucleation below 2273 K at 15 GPa from carbons with different structural organizations. Carbon 45:636–648. 8. Wentorf R, Jr (1965) The behavior of some carbonaceous materials at very high pressures and high temperatures. J Phys Chem 69:3063–3069. 9. Ryoo R, Joo S, Jun S (1999) Synthesis of Highly ordered carbon molecular sieves via remplate-mediated structural transformation. J Phys Chem B 103:7743–7746. 10. Kleitz F, Choi S, Ryoo R (2003) Cubic Ia3d large mesoporous silica: Synthesis and replication to platinum nanowires, carbon nanorods, and carbon nanotubes. Chem Commun 2136–2137. 11. Mohanty P, Fei Y, Landskron K (2009) Synthesis of periodic mesoporous coesite. J Am Chem Soc 131:9638–9639. 12. Mohanty P, et al. (2010) Direct formation of mesoporous coesite single crystals at extreme pressure. Angew Chem Int Edit 49:4301–4305.
13596 ∣
www.pnas.org/cgi/doi/10.1073/pnas.1006938107
couple emf, temperature uncertainties are expected to be within 25 °C. The temperature gradient over the length of the sample chamber (1.3 mm) was expected to be within 50 °C. The recovered samples were characterized with an optical microscope, a scanning electron microscope and a microfocus X-ray diffractometer. The SEM images were taken on a JEOL JSM-6500F Field Emission Scanning Electron Microscope operated at 15.0 kV. X-ray diffraction data of the recovered samples were collected using a Rigaku microdiffraction system with an imaging plate area detector (D/MAX-RAPID), using a Mo Kα radiation source (λ ¼ 0.071073 nm). The system was calibrated using the National Bureau of Standards Al2 O3 standard (SRM-675) for unit cell parameter determination. ACKNOWLEDGMENTS. We thank Bjorn O. Mysen, Alexander Goncharov, and Angele Ricolleau for technical assistance. We thank the Carnegie Institution of Washington and Lehigh University for financial support of this work. Li Zhang’s fellowship was supported by National Science Foundation Grants EAR-0911492, EAR-0810255, and DOE-DE-SC0001057. 13. Kresge C (1996) Hierarchical inorganic materials. Adv Mater 8:181–182. 14. Wang Z, et al. (2004) A quenchable superhard carbon phase synthesized by cold compression of carbon nanotubes. Proc Natl Acad Sci USA 101:13699–13702. 15. Mao W, et al. (2003) Bonding changes in compressed superhard graphite. Science 302:425–427. 16. Wu J, Zhao L, Chronister E, Tolbert S (2002) Elasticity through nanoscale distortions in periodic surfactant-templated porous silica under high pressure. J Phys Chem B 106:5613–5621. 17. Hirano S, Shimono K, Naka S (1982) Diamond formation from glassy carbon under high pressure and temperature conditions. J Mater Sci 17:1856–1862. 18. Liang Z, Liang J, Jia X (2009) Effects of NaN3 added in Fe-C system on inclusion and impurity of diamond synthesized at high pressure and high temperature. Chinese Phys Lett 26:038104/1–038104/3. 19. Meng Y, et al. (2008) Enhanced optical properties of chemical vapor deposited single crystal diamond by low-pressure/high-temperature annealing. Proc Natl Acad Sci USA 105:17620–17625. 20. Chen H-G, Chang L (2004) Characterization of diamond nanoplatelets. Diam Relat Mater 13:590–594. 21. Chen H-G, Chang L, Cho S-Y, Yan J-K, Lu C-A (2008) Growth of diamond nanoplatelets by CVD. Chem Vapor Depos 14:247–255. 22. Chen H-G, Chang L (2005) Structural investigation of diamond nanoplatelets grown by microwave plasma-enhanced chemical vapor deposition. J Mater Res 20:703–711. 23. Mohanty P, Fei Y, Landskron K (2009) On the high-pressure behavior of periodic mesoscale SBA-16 silica/carbon composites: studies at 10 GPa between 25 and 1800 °C. High Pressure Res 29:754–763.
Zhang et al.
