Take a fast Ge singlet for an uncooled LWIR core. Germanium's thermo-optic coefficient is enormous — dn/dT ≈ +396 × 10⁻⁶/K — so as the lens warms from 20 °C to 60 °C its focal length shortens enough to blur a 12 µm-pixel sensor several depths of focus. A screw-drive refocus mechanism fixes it, and adds cost, weight, current draw, and one more thing that breaks. Passive athermalization fixes it with algebra.
Chalcogenide glass transmission — the low-dn/dT workhorse of athermal LWIR design
| The Athermal Toolbox — dn/dT (10⁻⁶/K) | |
|---|---|
| Germanium | +396 · the problem |
| Silicon | +160 |
| ZnSe | +61 |
| ZnS | +38.7 |
| Chalcogenide (Ge-As-Se) | +30 – 60 · the diluter |
| CaF₂ | −10.6 · the canceller |
| Aluminum housing (CTE) | +23 · the free compensator |
The three-term budget
Thermal defocus is the sum of three effects: the optic's dn/dT (index change), the optic's CTE (surfaces change shape), and the housing's CTE (the sensor physically moves). Athermalization means choosing materials so those terms cancel over your temperature range. You are solving one equation: total defocus = 0.
The moves that work
Move one — dilute the Ge. Chalcogenide glasses (dn/dT +30–60 × 10⁻⁶/K) carry most of the optical power; a weaker Ge element supplies color correction. The system's net thermal defocus drops nearly an order of magnitude before you touch the housing.
Move two — use a negative element. CaF₂ is the classic tool with dn/dT = −10.6 × 10⁻⁶/K: paired against positive-dn/dT elements it actively cancels rather than merely dilutes. Watch its CTE (18.85 × 10⁻⁶/K) and thermal-shock sensitivity in the mount design.
Move three — let the barrel work. An aluminum housing (CTE ~23 × 10⁻⁶/K) grows with temperature; designed deliberately, that growth repositions the lens group by exactly the residual defocus. Free compensation, if — and only if — it is designed in from the first layout.
What to send your optics vendor
The operating and survival temperature ranges, the housing material (or the freedom to recommend one), the detector's depth of focus, and whether a molded-chalcogenide production path interests you at volume. With those four facts, an athermal design is engineering; without them, it is archaeology performed later, at expedite prices.