Chromatography Forum

zokitano: "Yes, the recommended velocity of 30 cm/s for helium is optimal average velocity of helium in any column"

lmb: “optimal average velocity of helium in 1m×0.1mm column in GC-MS is NOT 30 cm/s, but close to 180 cm/sâ€

Dear lmb

I agree completely that;

"Theory predicts and practical reality confirms that a factor of two departure of flow from Fopt in either direction reduced resolution by only about 10 %. As a result, formula (*) works reasonably well for isothermal and temperature-programmed conditions. In some cases, separation of a critical pair can be improved by experimental fine-tuning to find conditions that are the best for that particular pair. From that point of view, the flow rate found from formula (*) is just the best starting flow rate."

.... or in other words that changes in gas flow within reasonable limits have only very limited impact on resolution.

I am sure that we can also agree that changes in gas flow do have a very marked effect on retention time, which is the main reason why they need to be set to specific values for a given method.

Perhaps in saying "practical reality" I did not choose my words carefully enough. Theory and experimental reality do indeed agree; by "practical reality" I meant the demands of setting up and specifying methods so that they can be reliably run and reproduced in practice, with a workably useful level of robustness to operator error and instrument malfunction. To simply set a volume flow on an instrument makes the method vulnerable to malfunction of the flow controller, and (depending on the plumbing) to leaks at the inlet. Measuring the elution time for an unretained peak (in other words the linear gas velocity) helps to verify these aspects of instrument performance (as well as other things such as that the detector is actually functioning !).

The 50, 30 and 15 cm/s velocities of urban legend are puzzling, in that they are slower than expected, and I would like to draw on your expertise - does this make separations run at these flows more robust to column deterioration and poor inlet performance ?

Regards
Peter

Dear Peter,

I agree that there are two issues in our discussion.

One is how one should choose a column pneumatic conditions (pressure, flow, velocity, etc. of a carrier gas) when one begins to develop a new method, and another is how to maintain robustness of and existing method.

My previous posting addressed only the first issue, suggesting that the starting point should be selection of gas flow rate rather than its linear velocity. The flow rate could be selected according to formula (*) in my previous posting.

I would like to add the following to your comments on maintaining method integrity.

Of course, selection of the flow rate is not the end of the story. It is a good idea to compare expected and actual values of pressure and hold-up time. This might help to verify integrity of the system and to establish the control points for the system maintenance.

An important control parameter for verification of system integrity is hold-up time. For the system integrity, hold-up time is far more important than, for example, gas flow rate or linear velocity. For example, in some applications, a column is trimmed from time to time. To have the same retention times of all peaks after the trimming as they were before the trimming, one only needs to restore the hold-up time by reducing the column pressure. It is important to stress here that restoration of average velocity of a carrier gas would not restore the retention times.

You probably noticed that here I go again agitating against practical usefulness of average velocity. To set up pneumatic conditions of a carrier gas, use its flow rate (in mL/min). Once the method development is complete and system integrity is verified, measure hold-up time and keep it constant. It can be concluded that in both cases – method development and its maintenance – there is always something better than carrier average velocity. I really believe that it would be very healthy for practice of GC method development to pretend that the concept of average velocity does not exist.

[quote]“does this make separations run at these flows [50, 30 and 15 cm/s] more robust to column deterioration and poor inlet performanceâ€

Dear Peter,

I agree that there are two issues in our discussion.

One is how one should choose a column pneumatic conditions (pressure, flow, velocity, etc. of a carrier gas) when one begins to develop a new method, and another is how to maintain robustness of and existing method.

My previous posting addressed only the first issue, suggesting that the starting point should be selection of gas flow rate rather than its linear velocity. The flow rate could be selected according to formula (*) in my previous posting.

I would like to add the following to your comments on maintaining method integrity.

Of course, selection of the flow rate is not the end of the story. It is a good idea to compare expected and actual values of pressure and hold-up time. This might help to verify integrity of the system and to establish the control points for the system maintenance.

An important control parameter for verification of system integrity is hold-up time. For the system integrity, hold-up time is far more important than, for example, gas flow rate or linear velocity. For example, in some applications, a column is trimmed from time to time. To have the same retention times of all peaks after the trimming as they were before the trimming, one only needs to restore the hold-up time by reducing the column pressure. It is important to stress here that restoration of average velocity of a carrier gas would not restore the retention times.

You probably noticed that here I go again agitating against practical usefulness of average velocity. To set up pneumatic conditions of a carrier gas, use its flow rate (in mL/min). Once the method development is complete and system integrity is verified, measure hold-up time and keep it constant. It can be concluded that in both cases – method development and its maintenance – there is always something better than carrier average velocity. I really believe that it would be very healthy for practice of GC method development to pretend that the concept of average velocity does not exist.

[quote]“does this make separations run at these flows [50, 30 and 15 cm/s] more robust to column deterioration and poor inlet performanceâ€