§1 — Why a diamond Raman gauge?
Ruby fluorescence weakens progressively above ~120 GPa and becomes unreliable above ~150 GPa. Beyond this range the only practical laboratory optical pressure gauge is the high-frequency edge of the first-order Raman band of the diamond anvils themselves. This edge is always present in any DAC Raman measurement — no extra pressure marker needs to be loaded in the chamber, a decisive advantage when chamber volume is tiny (culets of 10–30 µm at megabar pressures).

§2 — Physical origin of the signal
Diamond has a single intense first-order Raman line at ω₀ = 1332.5 cm⁻¹ at ambient conditions, from the triply-degenerate F₂g zone-centre optical phonon (LTO mode). Under compression the stiffening C–C bonds shift this phonon to higher frequency. In a loaded DAC the stressed anvil shows a broad band with a well-defined high-frequency cutoff: this cutoff — the diamond Raman edge — corresponds to the phonon at the culet face, where the stress (and pressure) is maximum and equal to the sample pressure. The band extends to lower wavenumbers because deeper regions of the anvil are less stressed. The edge is determined in practice as the minimum of the differential spectrum dI/dω (Akahama & Kawamura 2010).

§3 — Practical measurement with the BETSA BROT optical head
BETSA has developed the BROT, a dedicated Raman optical head for DAC diamond-edge measurements. It records the complete diamond Raman band in backscatter geometry, allowing simultaneous determination of the diamond edge and any sample Raman signal in a single acquisition, without repositioning the optics. It connects to standard spectrometers and is optimised for the 1300–2000 cm⁻¹ range critical for multimegabar work.

§4 — Calibration history
The method was first proposed by Hanfland & Syassen (J. Appl. Phys. 57:2752, 1985), who reported a linear pressure dependence of the diamond Raman edge up to 30 GPa. Three calibrations are implemented in BETSA Wave Fit, in order of increasing range. Note an important distinction: Occelli (§5) calibrated the hydrostatic diamond phonon, whereas Akahama and Eremets (§6–§7) calibrated the stressed anvil edge — the quantity actually measured on a loaded anvil. The two stressed-edge scales are therefore the recommended ones for the megabar range.

§5 — Occelli et al. (2003) — hydrostatic phonon, ≤140 GPa
Occelli, Loubeyre & Letoullec (Nature Materials 2:151, 2003) measured the LTO phonon of a natural diamond single crystal under hydrostatic conditions in helium up to 140 GPa, with a 488 nm laser. Their phonon calibration is: ω_LTO = 1333.0 + 2.83·P − 3.65×10⁻³·P²  (cm⁻¹, P in GPa) The pressure is obtained by inverting this relation (the tool solves the quadratic for P). Occelli et al. also showed that diamond is more compressible than previously assumed (K₀ = 446 GPa, K₀′ = 3.0) and behaves as a Grüneisen solid to 140 GPa. Because it describes hydrostatic diamond rather than the stressed culet, this scale is best used as a low-pressure cross-check, or for a diamond chip placed in the chamber.

§6 — Akahama & Kawamura (2010) — stressed edge, ≤410 GPa
Akahama & Kawamura (J. Phys.: Conf. Ser. 215:012195, 2010) extended the stressed-edge calibration to 410 GPa using bevelled anvils and a He-Ne laser (632.8 nm) in backscatter, against the Pt EOS. The relation between edge frequency and culet stress is formally that of an isotropic elastic body with K₀ = 547 GPa and K₀′ = 3.75 (Akahama & Kawamura 2006): P = K₀·(Δω/ω₀)·[1 + ½(K₀′−1)·(Δω/ω₀)]  (K₀=547 GPa, K₀′=3.75, ω₀=1333 cm⁻¹) This holds well below ~200 GPa. Above 200 GPa the data deviate and the authors give a separate equation for the multimegabar range: P = 3141.3 − 4.157·ω + 1.429×10⁻³·ω²  (ω in cm⁻¹, P in GPa) The tool switches to this second equation above 200 GPa (the edge reached 1907 cm⁻¹ ≈ 408 GPa in their highest run).

§7 — Eremets et al. (2023) — universal scale, ≤477 GPa
Eremets et al. (Nature Communications 14:907, 2023) established a universal stressed-edge scale against the gold EOS, using 660 nm and 532 nm lasers (switching wavelength avoids parasitic luminescence from synthetic anvils above 300 GPa). With ω₀ = 1332.5 cm⁻¹: P = A·(Δω/ω₀) + B·(Δω/ω₀)²  (A = 517±5 GPa, B = 764±14 GPa) This is the recommended scale above 200 GPa, validated to ~477 GPa (record edge shift 2026 cm⁻¹). The authors showed a linear law is insufficient and that the non-linear term matters above 200 GPa; their scale gives pressures ~20% lower than extrapolated older scales near 0.5 TPa, with consequences for the hydrogen phase diagram.

§8 — Important practical notes
· The edge shape depends on anvil geometry, gasket and optical alignment — a consistent edge-definition protocol (minimum of dI/dω) is essential, and the calibration depends on culet shape (Baer, Chang & Evans 2008).
· The diamond Raman signal of the sample (if any) is hidden by the anvil signal below ~10 GPa; the two separate above ~10 GPa (Occelli et al. 2003).
· Low-fluorescence Type Ia diamonds reduce background. Above ~300 GPa some synthetic anvils luminesce over the Raman band — switching laser wavelength (660 ↔ 532 nm) solves this (Eremets et al. 2023).

Key references
Hanfland & Syassen (1985) J. Appl. Phys. 57:2752 — First diamond Raman edge gauge (linear to 30 GPa)
Occelli, Loubeyre & Letoullec (2003) Nat. Mater. 2:151 — Hydrostatic diamond phonon to 140 GPa; K₀=446 GPa, K₀′=3.0; 488 nm
Akahama & Kawamura (2006) J. Appl. Phys. 100:043516 — Stressed-edge scale; K₀=547 GPa, K₀′=3.75; Pt EOS
Akahama & Kawamura (2010) J. Phys.: Conf. Ser. 215:012195 — Extension to 410 GPa; multimegabar equation
Baer, Chang & Evans (2008) J. Appl. Phys. 104:034504 — Edge calibration and culet-geometry dependence
Eremets et al. (2023) Nat. Commun. 14:907 — Universal scale to 477 GPa; A=517 GPa, B=764 GPa; Au EOS; 660+532 nm