Questionable lonsdaleite identification in ureilite meteorites
Questionable lonsdaleite identification in ureilite meteorites
Péter Németh and Laurence A. J. Garvie
PNAS, Letter to the editor, May 8, 2023, 120 (20) e2304890120
refers to original article: Sequential Lonsdaleite to Diamond Formation in Ureilite Meteorites via In Situ Chemical Fluid/Vapor Deposition / ORIGINAL LINK
“Lonsdaleite is the mineral name given to the hexagonal diamond component in the hard carbon grains from the Canyon Diablo iron meteorite (1). Although natural and synthetic materials with diffraction data matching the Canyon Diablo lonsdaleite have been widely reported, recent studies showed that the structure of the hard carbon grains from the Canyon Diablo meteorite corresponds to a nanocomposite of c/h stacking disordered diamond and diaphite consisting of graphene layers coherently bonded to the cubic diamond matrix (2, 3). These studies demonstrated that nanostructured elements are intimately intergrown giving rise to structural features erroneously associated with hexagonal diamonds (2, 3).
It was recently reported that many ureilites, which are primitive achondrite meteorites, contain abundant micron-sized lonsdaleite grains (4). The identification was based on selected-area electron diffraction (SAED) data and cathode luminescence (CL) measurements, with supplementary data based on electron energy-loss spectroscopy (EELS) and X-ray diffraction (4). However, the lonsdaleite identification is not supported by the data. Most importantly, all of the SAED data can be interpreted as either graphite or cubic diamond, both of which were shown to exist in the sample (4), or diaphite, a material consisting of crystallographically associated diamond and graphene units (5). The fitting of the SAED data with a commercial software package is qualitative and does not provide proof of the presence of lonsdaleite.
Within the experimental measurement error (±1 to 2%), all the SAED data are consistent with graphite, cubic diamond, and diaphite. A major concern is that the sample was tilted, but not so as to show the diagnostic <010> projection and the discrete h0l reflections from which lonsdaleite identification would be unquestionable. The EELS data provide information on the bonding, but the data are inappropriate to distinguish between lonsdaleite and cubic diamond and do not exclude the occurrence of some graphitic material, both of which are identified. The X-ray diffraction pattern is also inconclusive as all the peaks attributed to lonsdaleite overlap with cubic diamond, goethite, and olivine.
A further issue relates to the interpretation of the CL data. The paper associates the 2.317 eV CL signal with lonsdaleite and based on this assignment maps the distribution of lonsdaleite. However, the paper does not cite references to support the 2.317 eV-lonsdaleite association. Previous work associates the 2.317 eV (535 nm) CL signal to clusters of N-doped diamonds (6).
It is concerning that paper (4) does not provide diffraction evidence for lonsdaleite in ureilite meteorites and fails to demonstrate why a CL signal reported for N-containing diamond (6) should be associated with lonsdaleite. The proposed formation of “lonsdaleite” and diamond in ureilite reported in ref. 4 is questioned.”
Reply to Németh and Garvie: Evidence for lonsdaleite in ureilite meteorites
Andrew G. Tomkins, Nicholas C. Wilson, Colin MacRae, Alan Salek, Matthew Field, Helen E. A. Brand, Andrew D. Langendam, Natasha R. Stephen, Aaron Torpy, Zsanett Pintér, Lauren A. Jennings, and Dougal McCulloch
May 8, 2023, 120 (20) e2305559120
“In a recent letter to the editor, Nemeth and Garvie (1) questioned our experimental evidence for lonsdaleite in ureilite meteorites, which we suggested formed via in situ chemical fluid/vapor deposition (2). The hexagonal form of diamond, known as lonsdaleite, was first reported in the Canyon Diablo meteorite where it likely formed from graphite via high shock pressures (3). Nemeth and Garvie previously argued that the lonsdaleite in Canyon Diablo is instead a defective diamond (4) or combined graphite/diamond complexes coined, “diaphites” (5).
Their letter (1) suggested that within experimental error, our electron diffraction results (2) can be attributed to graphite, cubic diamond, or diaphites. In our work, polycrystalline Pt was used as a standard to ensure high calibration accuracy, better than 1%, for diffraction analysis. At this level of accuracy, it is easy to distinguish between key phases, including the hexagonal set of {110}/{100} diffraction spots from lonsdaleite 2.18 Å and graphite 2.23 Å (>2% difference) when viewed down <001> (SI Appendix, Fig. S5). The nearest diffraction spots from diamond [the {111} at 2.06 Å] can also be ruled out. Multiple diffraction patterns taken at different orientations from the same crystal (SI Appendix, Fig. S6) only match lonsdaleite.
Electron energy loss spectroscopy (EELS), particularly in combination with the diffraction analysis, provides conclusive evidence for lonsdaleite. In addition to detailed information on local bonding arrangements, EELS can measure density from the plasmon peak position (6). In carbon solids, this approach has been verified both experimentally and theoretically (7). We determined the density of our lonsdaleite crystals to be 3.5 g/cm3 (SI Appendix, Fig. S5E), which rules out the possibility that the <001> diffraction patterns (SI Appendix, Fig. S5B) were collected from graphite (much lower density of 2.26 g/cm3); the hexagonal symmetry is not compatible with diamond. This density also excludes the possibility that this region contained significant quantities of diaphite (expected density ~2.9 g/cm3). Recent hardness measurements (8) also show that the lonsdaleite in our samples has extreme hardness, not consistent with graphite or diaphite.
Our interpretation of cathode luminescence (CL) measurements was also questioned; they noted that N-doped diamonds have a broad luminescence band at 535 nm (9). This peak has a full-width half maximum (FWHM) of 450 meV, whereas the peak we observe (SI Appendix, Fig. S3) is much sharper (FWHM of 27 meV) and is accompanied by a set of phonon replicates with approximately 130 meV spacing. This indicates that the lonsdaleite we observed and N-doped diamond have different defect structures.
Lastly, the fast Fourier transform shown in SI Appendix, Fig. S8 matches the {100}/{110} reflections of lonsdaleite. This figure also shows the well-known orientation relationship between the graphite <001> and lonsdaleite <120> directions (4), which is consistent with the theoretical transformation pathways between graphite and lonsdaleite (10).
We suggest here that the diaphite described by Nemeth and Garvie in Canyon Diablo (4, 5) may be the precursor to lonsdaleite formation in ureilite meteorites. Intense shock is needed to form diaphite, and in the ureilites, intense shock immediately preceded the chemical vapor/fluid deposition process. Shock may have created the out-of-equilibrium structure that facilitated replacement by lonsdaleite.”
