Temperature ramp rate for HPLC column compartment ovens...

[cphelps@smatrix]

I am looking for temperature ramp-rate information on the Alliance 2690/2695 and Agilent 1100 column compartment devices. I'm not a chemist and don't have access to these devices directly, so unfortunately I can't just mess around until I determine a safe value.

I have combed the Waters PDF catalog for hours trying to find this information, but have been unable to find anything at all. Perhaps some of you with on-the-ground experience could help me out?

Thanks,
Chris

[Kostas Petritis]

I am not sure if I understand your question correctly, but if you are asking about assigning a rate at which an LC temperature column compartement device will go up, there is no such option (the rate in general is quite low).

As you are not chemist, I am just going to ask this... Do we agree that you are not trying to do gas chromatography and you know that in liquid chromatography you are not going to achieve your separation by increasing the temperature of your separation column over time?

[cphelps@smatrix]

Sorry, I guess I wasn't specific enough.

What I am doing is programatically creating a sample set with many injections through the Empower SDK. In addition to the injections, I often insert Equilibration runs to allow the software time to get everything to the correct levels (such as when running a Gradient). What I am trying to do right now is add time to the "Run Time" of a particular Equilibration run based on the # of degrees difference in Oven Temperature between the next run and the previous run. I need to know how long it takes the compartment oven to increase 1 degree Celsius so I can multiply it by the number of degrees I need to change.

If I don't add enough time here, the injection will fail because the Oven Temperature is supposed to be at something like 45 degrees, but it is currently at 25.

Is that better?

Chris

[Kostas Petritis]

I hope that you were not offended by my question, but I had to ask as you never now...

The reason that they do not add a specific ramp rate is that these modules in contranst to the gas chromatography are not very well isolated. As a result the time for going from 20 to 50 would be different from going from 50 to 80.

But I just did a small experiment for you with our Agilent 1100. After closing the column compartement and with an ambient temperature of 22.7 the ramp rate was about 6 sec per degree Celsius. This was gradually increasing and it was about 10 sec per degree Celsius when going from 50 to 60 degrees.

I guess this should be enough to help you with your programming. Keep in mind that other modules might have slightly different ramp rates...

Remember also that the temperature of the column from injection to injection should be the same during the equilibration of the column (for the same set of samples)...

[Mark Tracy]

Look through the command set for the instrument control program. There should be a "Wait ColumnOven.Ready" or something similar. There may even be "ColumnOven.EquilibrationTime" or "ColumnOven.ReadyTempDelta" parameters that you can set. (If you had Dionex equipment & software, those would be their names.) In your equilibration method, just set the parameters, and wait for it to come ready; you're done. The time is non-deterministic, but is that really important?

[Uwe Neue]

I hate to throw a wrench into the works:

The fact that the column oven reads some temperature does not mean at all that the column is at the same temperature. The column is a massive piece of steel...

[AdrianF]

I am curious as to how you are experimenting - when you say you do not have direct access to the instruments. Are you using Empower to work remotely?

Uwe has a very important point - the temperature readout doesn't tell you the temperature inside the column - I would suggest about 10 mins but the only way to check is experimentally ie do repeat injections until the retention times are stable. If you are considering temperatures above 40C you need to think about preheating the mobile phase.

[cphelps@smatrix]

I'm certainly not offended. Most of the folks around here have years of experience that I will never have. The reason I came here way to attempt to tap into all of that on the ground experience.

I really appreciate that you took the time to test out the Agilent system, this is exactly the kind of data I am looking for. If I were to take the SLOWEST ramp rate you saw, and then add in a minute or two for padding/stabilization, I could probably produce a pretty safe Equilibration time.

RE your final comment about the "temperature of the column from injection to injection should be the same during the equilibration of the column", I am not totally sure what that means...do you mean that I should set the column temperature to a specific value whenever I am conditioning the column with my mobile phase solutions?

Thanks,
Chris

I hope that you were not offended by my question, but I had to ask as you never now...

The reason that they do not add a specific ramp rate is that these modules in contranst to the gas chromatography are not very well isolated. As a result the time for going from 20 to 50 would be different from going from 50 to 80.

But I just did a small experiment for you with our Agilent 1100. After closing the column compartement and with an ambient temperature of 22.7 the ramp rate was about 6 sec per degree Celsius. This was gradually increasing and it was about 10 sec per degree Celsius when going from 50 to 60 degrees.

I guess this should be enough to help you with your programming. Keep in mind that other modules might have slightly different ramp rates...

Remember also that the temperature of the column from injection to injection should be the same during the equilibration of the column (for the same set of samples)...

[cphelps@smatrix]

I'm not really 'experimenting' at all. My company produces a software platform for HPLC Method Development and Method Validation experimentation using DoE methodology. After you have a design with your input variables, we can create a sequence/sample set and associated methods in several data systems. The idea is to be able to run multi-column, multi-solvent, multi-variable designs in a walk-away mode.
We also import results data and do automated analysis, reporting and optimization.

I'm just the guy who is (recently) in charge of handling data exchange with the various instrument control programs like Galaxie, Empower, TotalChrom, etc. In the coming weeks I will be implementing connections with EZ Chrom and Chromeleon as well.

You can find out more here: www.smatrix.com

Chris

I am curious as to how you are experimenting - when you say you do not have direct access to the instruments. Are you using Empower to work remotely?

Uwe has a very important point - the temperature readout doesn't tell you the temperature inside the column - I would suggest about 10 mins but the only way to check is experimentally ie do repeat injections until the retention times are stable. If you are considering temperatures above 40C you need to think about preheating the mobile phase.

[cphelps@smatrix]

Dionex methods clearly have this advantage over all other data systems.

Galaxie has some automatic wait built in for this, but TotalChrom and Empower have nothing of the sort, thus my dilema. I do appreciate the pointer to the specific Chromeleon functionality that I will need to implement all to soon, that will save me some digging.

Chris

Look through the command set for the instrument control program. There should be a "Wait ColumnOven.Ready" or something similar. There may even be "ColumnOven.EquilibrationTime" or "ColumnOven.ReadyTempDelta" parameters that you can set. (If you had Dionex equipment & software, those would be their names.) In your equilibration method, just set the parameters, and wait for it to come ready; you're done. The time is non-deterministic, but is that really important?

[DR]

Rule of thumb, fairly safe approximation for standard 4.6mm ID jacket columns - 10min per 1°C increase, 15 min per 1°C decreases.

Overall thickness of the column will be the primary driver of how long it takes to equilibrate in a given column compartment.

eg a 3.9mm ID Waters column - like a NovaPak - will take longer to heat or cool than a Mac-Mod 4.6mm ID column because the tubing wall is a lot thicker.

[Kostas Petritis]

Chris,

What I meant is that the column temperature must be maintained the same during the whole set of samples that you want to run. I said that in reply to your comment which was:

"What I am trying to do right now is add time to the "Run Time" of a particular Equilibration run based on the # of degrees difference in Oven Temperature between the next run and the previous run."

Anyways, people make some good points here. The column it will take some time to be at the same temperature as your temperature control module. I didn't mention anything as you said that you inserting equilibration runs (which for me were blank runs before any sample injection to let the system fully equilibrate).

I personally think that DR's rule of thumps are overconservative (for example if you needed to go from 20 to 80 C you would have needed 10 hours of equilibration).

[cphelps@smatrix]

I guess this is mostly an Empower issue, as the other data systems we're dealing with right now seem to be able to handle the column temperature adjustment automatically before performing an injection.

I appreciate all of the information, it has been very helpful. Also good to hear that the 10 min / deg equilibration time is conservative, as that would be almost unusable.

Chris

Chris,

What I meant is that the column temperature must be maintained the same during the whole set of samples that you want to run. I said that in reply to your comment which was:

"What I am trying to do right now is add time to the "Run Time" of a particular Equilibration run based on the # of degrees difference in Oven Temperature between the next run and the previous run."

Anyways, people make some good points here. The column it will take some time to be at the same temperature as your temperature control module. I didn't mention anything as you said that you inserting equilibration runs (which for me were blank runs before any sample injection to let the system fully equilibrate).

I personally think that DR's rule of thumps are overconservative (for example if you needed to go from 20 to 80 C you would have needed 10 hours of equilibration).

[Mark Tracy]

I think a chat with the engineers at each company would be a useful exercise. For example, I know that the Dionex ovens have warmup and cooldown specifications that are run in factory testing, but these are not part of the normal published specs.

If you can't get the info from the vendor, you might have to hire a couple laboratories to gather data for you. Analytical labs will do almost anything if you pay them for their time.

[danko]

I guess this is mostly an Empower issue, as the other data systems we're dealing with right now seem to be able to handle the column temperature adjustment automatically before performing an injection.

Empower can easily handle this functionality but only with newer instruments. I can’t remember the exact models, but it works fine with the Acquity system (UPLC) – for sure!

Best Regards