§1 — Ruby fluorescence
Ruby is corundum (Al₂O₃) doped with Cr³⁺ ions (~3000–5000 ppm). Under laser excitation, Cr³⁺ ions emit two sharp fluorescence lines: R1 (694.25 nm at 300 K, P = 0) and R2 (692.86 nm), from the ²E → ⁴A₂ transition of chromium. R1 is brighter and sharper — the primary pressure indicator. BETSA ruby beads are annealed microspheres of 3–5 µm with narrow, reproducible linewidths [Klotz et al. 2009].

§2 — Excitation laser
Klotz et al. (2009) used an argon-ion laser at 514.5 nm, noting that "at 300 K, the effect of laser radiation has no detectable influence on the temperature of the illuminated ruby." Other commonly used wavelengths: 488 nm (argon-ion, blue), 532 nm (Nd:YAG doubled), 405 nm (violet diode). Statistical precision on the R1 position: ±0.1 cm⁻¹, corresponding to ±0.015 GPa [Klotz et al. 2009].

§3 — Why a pressure gauge?
Under compression, the Cr³⁺ crystal field strengthens, shifting R1 and R2 to longer wavelengths. This shift is reversible, monotonic and reproducible up to ~150 GPa. The measurement is fast, non-destructive, and requires only a simple optical setup.

§4 — Calibration equations
The power-law calibrations (Mao, Dewaele) use: P = (A/B)·[(λ/λ₀)^B − 1]
· Mao et al. (1986): A = 1904 GPa, B = 7.665 — original quasi-hydrostatic scale in Ar [Klotz ref. 9]
· Dorogokupets & Oganov (2007) — polynomial form: P = 1884·(Δλ/λ₀)·[1 + 5.5·(Δλ/λ₀)], self-consistent EOS recalibration (non-hydrostatic, to 150 GPa); PRB 75:024115
· Dewaele et al. (2008): A = 1920 GPa, B = 9.61 — transition metals (Al, Co, Cu, Mo, Ag, Ta, W, Pt, Au) as primary standards in He medium; PRB 78:104102
· Ruby2020 (Shen et al. 2020) — polynomial form, different from the power-law above: P [GPa] = 1870·(Δλ/λ₀)·[1 + 5.63·(Δλ/λ₀)] Calibrated against MgO, Mo, Cu and diamond EOS to 150 GPa; endorsed by AIRAPT June 2020. Recommended standard for all publications from 2020 onwards.

§5 — Validity range
Ruby fluorescence works best up to ~150 GPa. Beyond this, the signal weakens as the Cr³⁺ electronic transition approaches the corundum band gap under compression. The diamond Raman gauge takes over in the multimegabar range: Akahama & Kawamura (2006): A = 547 GPa, B = 3.75, calibrated to 310 GPa, then extended to 410 GPa with a separate non-linear relation (Akahama & Kawamura 2010); Eremets et al. (2023): extended to ~500 GPa.

§6 — Thermal correction
Temperature shifts R1 independently of pressure. Two corrections are implemented in Wave Fit:
· GSECARS: corrects both λ₀(T) and A(T) simultaneously — for laser-heated DAC experiments
· Datchi et al. (2007): λ_eff = λ₀ + 7.28×10⁻³·ΔT + 1.006×10⁻⁶·ΔT² (nm/K) — recommended for most experiments below 600 K

§7 — Pressure-transmitting media (Klotz et al. 2009)
Klotz et al. (2009) compared 11 media using the standard deviation σ of pressures from 5–10 ruby beads distributed across the chamber. This method is more sensitive and reliable than R1 linewidth or R1–R2 splitting on a single bead — "in several cases (argon, nitrogen, neon), R1–R2 can be misleading."

Medium	Solidification	Hydrostatic limit (σ)	Notes
Helium	12.1 GPa	σ < 0.15 GPa at 40 GPa	Best medium. Recrystallizes under pressure, releasing stress. σ < 0.5% at 40 GPa. Gas-loading required.
Neon	4.8 GPa	σ < 0.5 GPa at 50 GPa	First non-hydrostaticity at 15 GPa. Excellent beyond solidification. Gas-loading required.
Nitrogen N₂	2.4 GPa	~10 GPa	Competitive with neon below 10 GPa. Gradients 3–4% at 25 GPa (0.6–0.8 GPa). Cost-effective neon alternative.
Argon	1.4 GPa	~2 GPa	Bell & Mao (1981) reported ~9 GPa via R1 linewidth. Klotz (2009) measured first gradients at 2 GPa via σ. The ×5 discrepancy illustrates the low sensitivity of R1–R2 as an indicator.
MeOH:EtOH 4:1	~10.5 GPa	~10.5 GPa	Glass transition at 10.5 GPa confirmed by ultrasonics. Adding water (16:3:1 deuterated) has no measurable benefit [Klotz 2009].
Iso-n-pentane 1:1	7.4 GPa	~7.4 GPa	Gradients ~10% at 15 GPa. Easy to load at ambient conditions.
Daphne 7474	3.7 GPa (20°C)	~3.7 GPa (20°C) ~6.5 GPa (100°C)	Popular for clamp cells and low-temperature magnetic measurements. Hydrostatic range doubles at 100°C.
Fluorinert FC84/FC87 1:1	< 2 GPa	~2.3 GPa	No hydrogen → neutron scattering. Very limited hydrostatic range. Gradients reach 0.5 GPa at 10 GPa.
No medium	—	0 GPa	Direct compression: strongly non-hydrostatic from the start. EOS measurements unreliable above a few GPa.

§8 — Detecting non-hydrostatic stress
In helium, Dewaele & Loubeyre (2007) measured by X-ray diffraction that macroscopic non-hydrostatic stress reaches 0.3–0.5 GPa at 150 GPa — negligible except for gold (overestimates K'₀ by ~2). The hydrostatic R1–R2 relation in He is [Dewaele & Loubeyre 2007, eq. 1]:
R1 − R2 = 1.37 + 0.0011·P   (nm, P in GPa) A measured splitting larger than this prediction indicates either PTM solidification or direct ruby bridging between the anvils. The most sensitive indicator remains σ across multiple beads [Klotz 2009] — R1 linewidth or R1–R2 on a single bead can be misleading, especially for noble gases.