Thermostatted control of the eluent temperature before it reaches the cell reduces the effects of ambient temperature change on electrochemical detection. Furthermore, temperature control of the column protects the separation (peak retention and width) from the effects of temperature change, and when the LC-23C cartridge column heater is used with the cross-flow cell package, the distance between the heated column and cell inlet is quite small and there will be sufficient temperature control to minimize effects on detection. However, in extreme environments (e.g., lab temperature changes due to automated heating/cooling shutdown after business hours, or proximity to vents, ducts, or drafts) or at high detector gain, small fluctuations in temperature at the detector cell can still produce marked deviations in the detector baseline. In these cases, both a column heater and a cell preheater may be needed. Figure 3.1 shows plots of EC detector background current versus temperature. The slope of the curve (di/dT) is significant, typically 0.5-1.5 nA per °C. Hence it is easy to see how a small change in eluent temperature (e.g., 0.1 °C) could still cause appreciable shifts in the baseline.
Figure 3.1. Plot of electrochemical background current versus mobile phase temperature on a glassy carbon electrode operated at 650 and 800 mV applied potential.
What phenomena are responsible for this dependence? The background current in electrochemical detection derives from several contributions, the majority component being the oxidation or reduction of the solvent itself.
For water, the reaction is sluggish at moderate potentials; this is due to poor heterogeneous electron-transfer kinetics at the electrode surface. Elevations in temperature increase the heterogeneous rate constants, and the background current (the measure of the rate of electron transfer) correspondingly increases. From a noise standpoint, if we must operate at elevated temperature, we must do so precisely. In many cases, only a small rise over ambient is a good compromise. In doing so, one gains control of environmentally induced baseline drift without fighting large temperature coefficients.
Elevated temperatures similarly affect the magnitude of the peak current. It is not unusual to increase peak currents 50-70% (over ambient) by elevated-temperature operation. Although the temperature dependence of diffusion coefficients alone cannot explain this, it is probable that the diffusion layer thickness decreases as the viscosity drops. The concentration gradient at the electrode surface is accentuated, and the end effect is larger peak currents.
Taken separately, the trends in both background current and peak current versus temperature are inadequate in predicting the effect, if any, on detection limits. When the pertinent data are properly expressed in terms of the signal-to-noise ratio, the improvement is not so dramatic. For example, operation at 55 °C requires more vigorous temperature precision than at 35 °C. Thus, a 1.8-times increase in peak current is counterbalanced by a 23-times increase in baseline noise. A small increase in temperature (35 °C vs. ambient) makes the most sense in terms of signal/noise (Table 3.1).
Table 3.1. Signal and Noise vs. Temperature Setpoint on CC-5 Preheater Module. Conditions: 2.2 mL/min, +750 mV/GC/RE-4 reference, reversed-phase ion-pair separation, norepinephrine is test solute.