After studying the Raman spectra of single crystal CVD diamond, some of the explanations for these new peaks that are given in the literature can be ruled out. For example, the possibility that these peaks result from the interface between diamond crystals and their substrates seems very unlikely. It is more likely that these peaks are a result of individual defects in the diamond lattice, especially given the spatially discrete nature of the peaks that show up in the surface maps and depth profiles. Silicon for example, is known to form defect centers in a diamond lattice, much like nitrogen.13, 14 Silicon carbide gives rise to Raman peaks at 777 cm-1 and 967 cm-1 , which are within the range of the peaks seen in this study.12 Since the CVD single crystal used in this study is of very high purity, it is unlikely that there can be many chemical dopants present in the lattice. With an estimated depth probed by the laser of 1 µm, the volume measured during one spatial map is approximately 440 µm3 . Since most of the peaks observed occur only once or twice within a given map, the approximate surface concentration of a specific defect is 0.0023 – 0.0045 /µm3 . This value may be different within the bulk. Individual vacancies themselves, or other defects such as dislocations, could be one possibility for the source of these peaks. Migration of defects and vacancies during heat 28 treatment could have led to clustering, which could be another possible source of these peaks. Clearly more work is required in order to determine peak assignments.

Thank you very much for your comment.

What is Framesite? If you have a transparent mineral with a 1332 cm-1 line, then it is a diamond. If you have an opaque mineral with a 1332 line, then it can be polycrystalline diamond (carbonado, framesite, etc.). Or do you want to distinguish them by Raman spectra ONLY? It is necessary to report the laser wavelength since diamond is very rich in luminescence

Dear E. Vasil'ev,

Thank you very much for your remarks.

At first, I use the 532 nm laser. Second, the mineral is transparent. Third, I see only the typical graphite lines at 1360 and 1580 cm-1, which are dominant. The 1332 cm-1 position is a shoulder. All other lines are weak.

Best regards,

Rainer Thomas

Dear Rainer Thomas, show the spectrum of your sample, please.

Dear E. Vasil'ev,

Attached are the figures to the problem.

Best regards,

Rainer Thomas

Dear Rainer Thomas. In this spectrum, it is impossible to reliably distinguish the diamond line. As I understand it, this spectrum has very strong luminescence initially. And after the baseline subtracting, bands have appeared (or just were revealed) that may not be relevant to the features of the sample. You should record the Raman spectrum with other lasers.

Raman line of diamond is very strong: it is clearly seen in all varieties - carbonado, ballas, and (I didnt try) in Framesite. Only in impact lonsdaleite-containing diamond the line is broadened and weak but in most (not all) cases it is distinguishable with any excitation.

Dear E. Vasil'ev,

Not all diamonds show the sharp 1332 cm-1 line. See for this: Alexander M. Zaitsev (2001) Optical Properties of Diamonds. A Data Handbook. Chapter 4 Scattering, 4.2 Raman Scattering, pages 69-121

I will check the sample with another laser wavelength and try to obtain TEM data.

Dear Rainer Thomas,

anyway it is very interesting!

As mentioned above, your best here is to look with other wavelength lasers. The advantage of this is that non diamond carbon (sp, sp2, etc) related bands are dispersive and thus their peak position will move with wavelength. The diamond peak at 1332cm-1 will not. What you show could easily be a D-band as its very broad.

This is really well covered by the papers of Ferrari and Roberton, especially Phys. Rev B 61 (2000) 14095 and Phys Rev B 63 (2001) 121405

Oh I forgot to mention, ta-C (a form of DLC) can be pretty transparent

Some other useful plots from Robertson, Materials Science and Engineering R37 (2002) 129

The third plot shows the shift in peak position with laser wavelength

Dear Oliver A. Williams,

Thank you very much for your comments, reference, and hints.

The problem is, however, the good transperency of the mineral in question.

Best regard,

Rainer

See my previous comment on optical adsorption. ta-C is pretty transparent in the but its not diamond.

could you define "good" transparency? i.e. what does the adsorption / wavelength spectra look like?

Another obvious point, the image you show looks like there are dark inclusions in an otherwise transparent (-ish) material. sp2 carbon has a huge Raman cross section so this would explain why you see large 1330-1370 and 1580cm-1 bands, they could be microcrystalline graphite inclusions / traces of glassy carbon . The other bands could be the other material. Do you have any other characterisation or idea of its origin? Knowing the main elements would help, i.e. even a low resolution EDX would give an idea of the elemental composition. Then you can look for likely candidates from there, do XRD etc :).

Dear Oliver A. Williams,

Thank you very much for your comments. The surrounding mineral is phlogopite (F-phlogopite). The transparency is similar to feldspar. I have only (at moment) this small sample and have to save this for TEM studies.

Regards,

Rainer