admin
Admin
Admin
| Posts: 138 |  |
|
Loss of resolution through heating. - 2005/11/28 19:02
Heating the mobile phase could result under some circumstances in the loss of resolution (compared to the unheated mobile phase):
Friction leads to higher temperature in the inner area of the column than in the outer area and to an acceleration of mobile phase (and parts of the deluted sample).
Let me simplify the mechanism: We have an outer (stationary) phase which is cooler (cooled from outside) than the inner (stationary) phase (heated up by friction) due to the different proximities to the wall. As the solvent passes through the column the outer mobile phase slows down while the inner mobile phase continues at fast speed. The outer mobile phase passes through in a round-about way while the inner mobile phase passes directly. At the entrance friction remains constantly high. The higher the pressure the higher the friction. The higher the friction the higher the heat. The higher the heat the lower the viscosity. The lower the viscosity the faster the speed.
This results in peak broadening.
Of course, this happens also at ambient temperature and rises with increased pressure and flow rates.
If the mobile phase has a lower temperatur than the column the higher viscosity slows down the inner mobile phase and peak broadening will be reduced.
Also in some cases it is not possible to transfer temperature sensitive methods from a hplc system with heating blocks (like the agilent systems) to systems with air heating, because the selected and the resulting real temperature of one hplc system vary too much and interlab reproducibility (with different hplc) is even worse (as much as 10°C difference!).
Best way to reproduce heating a column is to use a circulating water bath and extra temperature control (± 0.1°C). Normally this is not realistic in practice.
It should be good lab practice - however not a common one - to report the data of the thermostat used in an hplc system.
Always give the column enough time to adapt to non-ambient temperature. The wider the column id the longer the waiting time.
On the other hand, most methods I've seen produce the same results at ambient temperature as at 30°C.
And even if the built-in thermostat of a hplc system shows some diviation between selected und resulting real temperature these effects would not be as high as the ambient temperature changes inside some labs (especially if in the same room one or more gc systems are running as well).
So in most cases using a thermostat is not a mistake.
Peak broadening could be reduced by increasing the column temperature (and so reducing back pressure and the viscosity of the mobile phase).
Also, it could be reduced by changing the dimensions of the column; for example instead of using the typical column (250 x 4.6 mm) I suggest unsing a shorter column with wider id (150 x 6 mm). Or you can reduce the lenght and flow rate by half if seperation allows this.
Even when using a column with 150 x 6 mm and 3 µm particles instead of 250 x 4.6 mm and 5 µm particles, back pressure will decrease and often results in smaller and higher peaks.
However, be carefully, higher peaks could disturb the linearity.
Bye, Andreas Neumaier
|
Chorma Forum 
