Cold finger
A thermally-conductive rod or plate inside a cryostat or test fixture that conducts heat from the device under test to a cooling reservoir. The standard thermal interface for low-temperature laser characterization.
A cold finger is the thermally-conductive extension that physically supports the device under test inside an evacuated cryostat or environmental chamber, transferring heat from the device to a remote cooling reservoir (liquid nitrogen, liquid helium, cryocooler cold head, or large TEC).
Construction. Typical geometry: a copper or oxygen-free high-conductivity (OFHC) copper rod, polished and gold-plated on the device-mounting surface to ensure good thermal contact and low surface oxidation over time. Length ranges from 30 mm (Stirling cryocoolers, small dewars) to mm (large LN continuous-flow cryostats).
Standard package interfaces:
| Cold finger type | Cooling | Achievable temperature |
|---|---|---|
| Single-stage TEC + Cu finger | TEC | to °C |
| Multi-stage TEC stack | Cascaded TEC | to °C (sub-room ambient required) |
| LN continuous-flow | Liquid nitrogen | 77 to 300 K |
| Closed-cycle Stirling cryocooler | Mechanical | 30 to 300 K |
| LHe continuous-flow | Liquid helium | 4 to 300 K |
| Pulse-tube cryocooler | Mechanical, no consumables | 4 to 300 K |
| Dilution refrigerator (research) | Mixed-gas | mK |
Thermal performance. Two parameters dominate:
- Thermal conductance between device and base (limited by the contact between submount and cold finger; typically 0.1 – 10 W/K depending on area and pressure)
- Heat load on the cold finger (radiative absorption from warm chamber walls + conductive load via electrical leads + parasitic gas conduction)
For laser characterization at moderate cryogenic temperatures (e.g., 77 K), thermal links need to handle 0.1 – 1 W of dissipation from the laser itself, plus another 0.5 – 2 W of parasitic load. Below 20 K, parasitic load typically dominates over device dissipation; thin-walled vacuum chambers, radiation shields at intermediate temperatures, and resistive leads with low thermal conductivity are required.
Common failure modes:
- Gas leak compromising vacuum raises gas conduction parasitically; device temperature drifts upward, often presenting as anomalously low characteristic temperature in temperature-dependent measurements
- Loose mechanical mount between submount and cold finger raises thermal resistance and causes irreproducible measurements between mounting cycles
- Oxidation of cold finger surface over many mount-unmount cycles; standard maintenance is periodic re-polish or re-plating
Electrical lead heat sinking. Wire leads connecting the device to room-temperature feedthroughs act as thermal shorts that bring heat into the cold space. Best practice is to thermally anchor leads at intermediate temperature stages (e.g., 77 K shield on a 4 K cryostat) using bobbins or copper braids before they reach the device. Twisted-pair wires with poly-imide insulation are the standard choice for sub-pA dark-current measurements.
The cold finger geometry is the most thermally critical part of any cryogenic optical measurement. Small variations in thermal contact between the device submount and the finger surface produce 1–10 K offsets between the reported (thermometer) temperature and the actual device junction temperature, which can dominate parameter extraction errors.