Chromatography Forum

Can someone pls explain the importance of the detector temperature when using a refractive index detector? I understand that it is critical to keep the tremperature very constant but what difference would it make if the temperature is set at 50C rather than 40C? Does this temperature affect the actual response, so that one would observe different responses at different temperatures?

Thanks,
Omar

Hi Omar,

As the angle at which the light, traversing the sample solution, gets refracted, is temperature dependent, the beam could be directed closer or further away from the detection unit. This temperature dependent directional variation would result in varying amounts of light hitting the detection unit (typically a PMT) which in turn would manifest its self as higher or lower signals (peaks).

Best Regards

What you're actually measuring is the difference in RI between your mobile phase and your mobile phase + analyte. While temperature *stability* important, the actual temperature is less so, except insofar as it affects the ability to maintain constant temperature.

In short, no, the temperature itself does not have much effect.

In an absolute sense one can imagine that solutions may have a significant rf dependance on temperature. One shouldn´t forget that HPLC is a relative method, it is calibrated. It is a result of this calibration that temperature is irrelevant here, if one stays within stable conditions.

Different compounds exhibit different temperature dependencies (see some examples below) and note how little the refractive index of water decreases with temperature increase – compared to that of acetone, for instance.

Temperature dependence of refractive index for selected substances:

Substance n (at 15°C) n (at 20°C) n (at 25°C)

Isopropanol 1.3802 1.3772 1.3749
Acetone 1.3616 1.3588 1.3560
Ethyl Acetate 1.3747 1.3742 1.3700
Water 1.3334 1.3330 1.3325

So, actually one can – with some patience – optimize peak heights only by finding the “rightâ€

One thing to check is whether your RI cell can chill as well as heat.
If it only heats then you should try to keep the cell hotter than
your column temperature. That can be tough for people running
ion exclusion which is very common with RID.

mardexis, why should one "keep the cell hotter"?

HW: You want the cell hotter because it will be easier to control the temperature in the RI detector. If the solvent coming in is higher than the set point and the detector can't cool then the detector isn't controlling the temperature any more.

You do want to make sure that the detector is a differential RI detector and not just a RI detector because they are less sensitive to temperature variations both in sensitivity and noise.

Also there is some dependancy on the dn/dc (change in refractive index vs. change in concentration) for most compounds so most of the time there will be an optimal temperature for a given analyte. This is typically not very significant.

Also the sensitivity of an RI detector is dependant on the differences in RI between the analyte and the mobile phase so the temperature has an affect on the RI of the mobile phase as well. This is typically not very significant.

Also higher temperatures does increase dispersion by increasing difussion which can reduce peak height which then reduces sensitivity. Once again
this is not typically significant.

The major reason for running higher temperatures (higher than required for stable base lines) in the RI is so the sample will remain in solution and solvent viscosity is reduced so as to reduce backpressure contributed by the RI detector.

I don´t understand most of this.
For instance, the detectors used (or still have?) heat sinks to compensate for slight mobile phase temp. variations.
dn/dc variations, etc., used to be taken care of during calibration (never saw it).
What dispersion is increased by diffusion? .....

In addition to HW Mueller’s remarks/questions, I’m intrigued by the priority of the arguments. Most of the reasons for controlling the temperature (being higher in the flow-cell, than in the column) are assigned to minor significance while major role is assigned to the need for securing the solubility of the analytes in the flow-cell. But if the analytes aren’t perfectly soluble in the mobile phase, then how would they reach the flow-cell in the first place? Aren’t they going to get stuck in the column or even before that?
Also, the pressure (due to mobile phase viscosity or whatever) is highest before the column and in it, not after it and thus in the flow-cell.

Best Regards

My point may not have been very clear so let me just try to state it in another way. I think we all agree and understand that the most important aspect of temperature control in RI is to maintain constant temperature in the RI cell to avoid baseline noise. My point was that some RID's provide temperature control by heating and cooling the flow cell. Other cells control temperature in the flow cell only by heating. If yours is of type that heats only you will need to set the temperature above the temperature of the column or your RI will not be controlling the temperature in the cell, ie, the flow cell will already be above the set point due to heat input from the effluent so no temp control can be provided by heating.

It can be controled by not heating.

Temperature variations can be dampened by heat sinks and counter current heat exchangers but can not be controlled. If you are not actively controlling the temperature in the detector you will have temperature variations that are controlled by enviromental conditions. Mainly this will result in baseline drift instead of noise. i.e. Long term variations vs short term variations

In terms of sample solubility, which is mainly an issue in GPC analysis with polymers, the sample is dissolved in an autosampler with temperature control and run through a column with temperature control.

The B term in the Van Deemter equations relates the effect of diffusion on longitudinal dispersion which will increase peak width and reduce sensitivity. It's really only significant at low flow rates.

Dn/dc is typically taken care of during calibration but it still can vary if the temperature in the detector varies. Hence keep the detector at a constant temperature.

Yes, the pressure is highest before the column but pressure drop happens every where in the system including the detector and running it at a higher temperature will reduce overall system pressure. May not be significant depending on your conditions.

van Deemter in a flow-cell context

Best Regards

The longitudinal dispersion is present everywhere in the system. Obviously molecules diffuse even when they aren't in a column and that diffusion increases with temperature. I referenced the B term to help people relate to the effect even though it doesn't directly correlate to the diffusion in a tube or flow cell. This would be described more accurately with Ficks' laws.

Please forgive me if your comment was in jest.