A nitrogen-vacancy (NV) center is a substitutional nitrogen atom next to a missing carbon in diamond. Its ground state is a spin-1 triplet with a zero-field splitting D between the ms=0 and ms=±1 sublevels, driven and read out optically (ODMR):
D(T) = D₀ − a·(T − T₀)
D₀ = 2870.0 MHz, T₀ = 300 K, a = 0.0742 MHz/K
Optically-detected fluorescence vs microwave freq f:
PL(f) = PL₀·[1 − C·(L(f−f₋,Γ) + L(f−f₊,Γ))]
f± = D(T) ± γₑ·B_z , γₑ ≈ 28.02 MHz/mT (electron spin)
- True temperature — the hidden real temperature of the diamond; the sensor never sees this number directly, only the fluorescence dip.
- Bias field — a static Bz Zeeman-splits the single dip into a symmetric pair around D(T), visible as the two ms=±1 sublevels separating in the energy diagram; their midpoint still reports D(T) cleanly.
- Averaged sweeps N — each sweep is shot-noise limited; averaging N repeats shrinks the fluorescence noise by 1/√N. Watch the precision tracker panel: the scatter of single-shot temperature errors visibly tightens as N grows.
- The inferred temperature comes from a weighted-centroid fit of the dip(s) in the averaged spectrum, then inverting D(T) — exactly the estimator a real ODMR thermometry pipeline uses.
The lattice panel is a 2D schematic cross-section for orientation — drag to pan, scroll/pinch to zoom. The real defect sits in diamond's 3D cubic lattice; none of the thermometry physics above depends on that dimensionality, since D(T), the Zeeman splitting and the 1/√N noise scaling are properties of the electron spin and the optical readout, not of the crystal's geometry.
Real-world relevance: NV-diamond thermometers already map temperature inside living cells and on chip hotspots at sub-micron resolution and sub-degree precision — a distinct use of the same NV spin that magnetometers exploit via the Zeeman shift instead of this thermal shift of D.