HPLC Products

Normal

                          

Problem                                                                Problem

When an HPLC has been allowed to stabilise and has reached Dynamic Equilibrium (everything is happening but nothing is changing) the injection of a sample should result in each component eluting in a constant volume of eluent. Assuming we have a constant flow rate, this should equate to constant retention times.  We tend to identify peaks by their retention time, so it is absolutely critical that the flow rate is constant, during a run, and from one day to the next.

We consider three scenarios here. But first, one question:

Does the void volume peaks have a constant retention time while the other peaks have unreproducible times, or does the void volume retention time vary with the others?

The void volume peak, sometimes called the solvent front, is a peak which arises when the solvent which was used to dissolve the sample arrives unretained at the detector flow cell. When using a UV detector, it represents a refractive index change, and hence its unusual and distinctive +/-/+ shape. Since the solvent is essentially unretained by the column, it takes the fastest way through the column, and its elution time is a measure of the volume between the injector and the detector, through the column, but excuding most if not all of the pores. This volume is fixed, unless the tubing, the loop or the column are changed. hence the void volume peak is not affected by solvent type or mixing ratio, or temperature, or pH, or buffer type or concentration, or anything at all except flow rate! So if the void volume peak remains constant, the flow rate must be constant. And if the void volume peak retention time changes, the flow rate has changed.

So going back to our question. Does the void volume peak retention time change when the other peaks retention times change?

If YES, we are looking for a reason why the flow rate is changing.

Things which can cause the flow rate to vary include:

• Check valve problems.
• Air bubbles
• Leaks
• Operating at the extremes of a pump’s range (ie very fast or very slow).

So each of these can cause the retention times to change, even if all the other parameters are kept constant.

The last of the above possibilities also applies especially when running two or more pumps in a high pressure gradient configuration. Even if used to pump an isocratic eluent, if one solvent is required at a very low percentage, the pump delivering it will be running so slowly the motor will hardly be turning, and the flow accuracy will be uncertain.

If the answer to our question was NO, the void volume peak does not change with the others, then we do not have a flow rate issue, and must look at factors which affect the distribution coefficient of each sample component between the stationary and mobile phases. Any factor which makes the sample components spend more time in the eluent will cause the peaks to elute faster, and vice versa.

The eluent volume required to elute a peak will change if any of the following parameters are changed:

• Solvent Temperature (eg when refilling solvent bottles in the winter)
• Column Temperature
• Eluent pH
• Buffer Salt
• Buffer Concentration
• Organic modifier in the eluent
• Eluent %B
• Column type or chemistry

So to achieve reproducible retention times, all of the above must be kept constant.

The effect of changing the above parameters is not the same for each component of a sample mixture. So if peaks elute faster as a result of the change, some peaks may show a greater change in retention time than others, thus affecting selectivity, and this can cause peaks to co-elute! It can of course improve a separation. All of the above parameters should be optimised when developing a method, and then controlled when running the method to avoid unexpected changes in retention time.

So now to our three scenarios:

Scenario 1 The retention times change a little each time an injection is made, sometimes earlier, sometimes later.

This is almost always a flow rate issue, which can be confirmed by observing the void volume peak. Suspect the check valves first. If the change in retention times is small, dirt check valves is almost always the answer. Remove the check valve cartridges, and clean in an ultrasonic bath for 20 minutes. Immerse the cartridges in a beaker containing methanol:water, 25:75, so that if they come apart during ultrasonication, you will not lose any components. Then reassemble and test the system. If the problem persists, replace the check valve cartridges with new ones.

If the retention time variation is greater, check the back pressure variation and suspect an air bubble. If seen during an overnight run, it would be unusual for the problem to have persisted all night, because eventually, air bubble trouble tends to clear itself. But if there is evidence of a fluctuating back pressure, purge the system to remove the air, before retesting.

Scenario 2. The retention times are gradually getting longer and longer.

If the void volume peak retention time is also getting longer, the flow rate is gradually reducing. Suspect a leak, or possibly an air bubble.

If the void volume peak retention time is not changing, then the eluent is becoming weaker, or the column is degenerating. Changes in the eluent which could cause it to become weaker and weaker include

• evporation of the organic solvent component,
• decreasing temperature in the column heater (the door not closed properly, or the heater has blown a fuse and is slow,ly cooling down)
• poorly mixed eluent which has been premixed into a bottle - maybe not shaken?
• a strong solvent is gradually being washed out of the column. This means that the system was not at dynamic equilibrium.
• bonded phase being stripped off. More and more silanol sites are being exposed, causing the retention of polar components to increase.

Scenario 3. The retention times are gradually getting shorter and shorter

Assuming the void volume peak is not changing, this means that the eluent strength is increasing. This is less common. Possible causes include

• the system is not yet at dynamic equilibrium
• the column heater is not yet fully warmed up. Just because the display is constant does not mean that the column is up to temperature
• organic solvent evaporating from a pre-mixed eluent in normal phase using an amino column with acetonitrile:water. Maybe the bottle is in the sunlight?

See also our Technical note "Retention Times change 'for no reason'!"