The variable temperature (VT) unit on our machine uses a thermocouple wire to measure the temperature of “the sample.” However, because the thermocouple wire does not physically touch the sample (as a matter of fact, it is exterior to the NMR coils), it is possible for there to be a difference between the displayed temperature and the actual temperature of the sample.
The thermocouple temperature can be calibrated by using the temperature dependence of the 1H NMR signals of ethylene glycol (most useful at high temperatures) or methanol (most useful at low temperatures) as a function of temperature. In this example, the difference in chemical shift between the -OH and -CH2- protons (D) in neat ethylene glycol was recorded, and the actual temperature of the sample was calculated according to the equation:
T = (4.637-D) / 0.009967
In this sample spectrum, taken at room temperature, D = d(OH) – d(CH2) = 1.665. This corresponds to a temperature of 298.2 K. The absolute chemical shift tends to drift slightly because a sample containing neat ethylene glycol has no significant deuterium so the spectrometer cannot be locked. However, the difference between the chemical shifts is not affected by spectrometer drift.
Plotted against the recorded thermocouple temperature, we can see that the thermocouple is usually within 1 K of the actual sample temperature. For precision variable temperature work, we suggest calibrating the temperature based on the measured temperature, rather than the thermocouple temperature, and carefully making sure that all samples are at thermal equilibrium prior to starting work. For routine variable temperature work, including 2-dimensional NMR, the thermocouple temperature is sufficiently accurate.
If you’d like to learn more about NMR, variable temperature NMR, or 2-D NMR, please contact Emery Pharma.