If we pour tea into a cup, then the cup heats up (cooling the tea somewhat). If we then put the cup on a desk then some heat will also be transferred to the desk surface, the amount depending on the thermal conductivity and specific heat capacity of both the cup and the desk material. So the rate at which the tea cools will be affected by the desk. For the worst (fastest cooling) case, put the tea in a metal beaker and stand it on a metal bench… For slow cooling tea, use china and a wooden desk.
If we were to attach temperature sensors to the cup and the desk and put a thermometer in the tea, they would show the cup and the desk to heat up, although not by the same amount, and not at the same rate. At some point the cup would probably indicate the same temperature as the tea. Left (for longer than you think), all three would end up at the same (room) temperature.
Finally, we could re-heat our tea by placing it on a hotplate whose temperature has been set to that of the tea at the start. But the temperature reached will probably be lower, as heat is now lost continually through the cup and by imperfect heat transfer from the surface of the hotplate. Our metal cup will now do a lot better than the china one although it may end up too hot to touch.
Finally, the hotplate may have temperature variations across its surface so our result may depend on where we put the cup. It’s surprisingly difficult to heat a block of metal to the same temperature throughout, and even more difficult to measure it.
All these effects occur when we do a thermal threshold test, using either a hotplate or a dedicated thermal probe. This makes it very difficult to compare data sets, even though each may be repeatable, reproducible (between sites) and show treatment effects.
Although the average temperature of the hotplate may be indicated on the display, it is difficult to know for sure what is happening at the surface, especially when it is changing quite quickly (several degrees/sec). Similarly, in a small thermal probe, heat is travelling from the heater in several directions at once: towards the tissue, towards the sensor and into the air.
Temperature sensors all respond differently:
Thermocouples, being welded, are usually spherical so most of the junction may be in air rather than in contact with the surface being measured. Glueing the thermocouple with a thermally conductive epoxy helps, but such adhesives are still poor conductors of heat compared to metals. Glueing it into a hole is better.
The connection pins of an electronic sensor may be physically attached to the surface. The main thermal path is usually via the legs, but the rest must then be well insulated to prevent heat flow back out, and the response is usually slower than a thermocouple (because the mass is larger)
The commercial sensor in the picture is well suited to the measurement of liquids but hopeless for a surface as, like the thermocouple junction, most of it is in the air.
Caution is needed when comparing Thermal Threshold techniques
Whatever the method, latency or ramped, hotplate or probe, heat is flowing in many directions, at different rates, through materials of different heat capacity and conductivity, and being measured by different sensors.
Topcat Metrology has spent 20 years characterising the performance of its thermal threshold probes, to provide continuity to its database of threshold data on many species through several generations of probes. The work has resulted in continually smaller and lighter probes with a greatly decreased chance of tissue damage. Topcat use both integrated sensors and thermocouples in the probes and are cautious of believing either of them implicitly. Calibration is carried out against commercial sensors and back checked against a mercury thermometer.
The end result is a suite of thermal nociceptive testing systems which are repeatable and reproducible. Topcat’s wireless thermal threshold systems are known worldwide to produce tight baseline data and to demonstrate analgesic effect and we are the only company to produce such systems commercially.
We believe that the thresholds produced by MouseMet Hot are in line with the rest of our range. Validation data from the University of Queensland shows baseline thresholds of approximately 50C on mice with heat allodynia being detected at about 43C. Click here for the paper.