Restek Blog Question for GC users - A Challenge

[GasMan]

JI2002,

You are correct that the pressure drop down the column is not uniform, I simplified this for the explanation.

Gasman

[jdezeeuw]

Hi Gasman,
Using the 30+10 column: I agree you are doing the actual separation under different pressure drop 30=>5 or 25=> 0.
I still believe its easier to understand/explain that it's the difference in gas velocity between inlet and outlet sections, that drives the difference in retention times (and K-factors)

Now we go a little extremer, do the same exercise, with only a 2 m coated column and 30m deactivated tubing. It was done yesterday.

The difference was exactly 42% increase in retention factor.. for the 2 meters Rtx-5ms when connected at the inlet side of 30m deactivated tubing..

There was another interesting observation: what do you guys think happened with the peak width/efficiency in last experiment when comparing the 2 situations?

[zokitano]

There was another interesting observation: what do you guys think happened with the peak width/efficiency in last experiment when comparing the 2 situations?

The peaks became broader and efficiency decreased when the coated column was connected in the inlet?

[chromatographer1]

Jaap,

Your answer and my answer are both aspects of the same situation. Both are correct and both have bearing on the explanation of the question.

Your primary attention to the flow rate is one part of the equation for the time of migration of a peak down the column.


My primary attention to the partition of the analyte into the stationary phase also is part of the equation.

As a thought experiment showing the different aspects of the problem, imagine a peak that was truly unretained being injected onto the column. Either way the column is installed the RT of the peak would be the same, whether the coated end of the tube was at the detector or if it were at the injector.

As different analytes are chosen as the test peak the difference in RT will increase showing the effects of the flow rate.

My comment about the nature of the analyte to perform the experiment still stands. For a volatile analyte the flow rate of the gas has a greater bearing on the difference of the RT. But for a more strongly retained analyte the partition effect of the greater pressure will have a greater effect or as great an effect as the flow rate.

The best separations and sharpest peaks are optimized when the retention is minimized by a liquid phase. That is why when comparing the same ID capillary column dimensions with thicker and thicker films the number of plate per meter goes down as the film increases.

It all boils down to the calculation of plate number depends on more than one physical aspect of the column.

My original comment about the nature of the analyte having an effect on the answer is correct. And I concede that given the choice of analyte the flow rate CAN have a larger effect on the RT of that analyte.

I think we can agree that both factors are important, the time of the partitioning of the analyte into the liquid phase, and the velocity of the analyte when present in the gas phase. These two factors determine ultimately the time required for an analyte to elute from a column.

best wishes,

Rod

[jdezeeuw]

Rod
do I understand you correctly:
Do you say that retention and (retention factor k) depends on the actual pressure we do the separation? So if I take a normal coated column and do an iso separation using 100kpa =>0 pressure drop; Now I add a little restriction at the end and I do the separation using 150 =>50 kPa, I can expect different kapacity factors because of the higher pressure?

I have not done exact measurements, but it sure does not apply when I compare atmospheric with vacuum outlet. I get the same K factors.

Some data to chew on:
The 30m-10m-guard situation columns as well as 20m-5m-guard both showed for all components increase in kapacity factor of 13 %

We did the same with 2m-30m-guard and all components showed 42% increase in retention factor..
No relation with polarity of the component

jaap

[chromatographer1]

Retention times will change. k' will change very little. With LITTLE restriction you will see LITTLE change.

Take a 30m 0.53mm ID column 1 micron column and put a 10m 0.2mm ID tube ahead and behind it and see the difference in retention times, plate number, and separation for yourself. You have to maintain the same flow rate EXACTLY.

You have to make changes so the physics can be manifest. But it does occur. And with temperature and analyte selection this difference may be minimized. I do realize that.

You disagree? You believe the partition between of any two gas/liquid phases is independent of pressure? If an analyte partitions the same amount of time under different pressures then the retention time of that analyte is determined only by the length of the tube and the flow rate, and your opinion is confirmed. But if an analyte spends more time in the liquid phase under a higher gas pressure the retention times will not be the same. Conduct the experiment under the right conditions and this change is seen. Conduct it under mild conditions and it can be within experimental error.

I repeat, the size of the effect is also dependent upon the analyte chosen to demonstrate the physics.

best wishes,

Rod

[DSP007]

Sorry, time to go home, why use a translator.
If it does not overload the column on a thin film - the retention times should not change greatly.
But efficiency in the isothermal will suffer greatly, and it must be considered.

[jdezeeuw]

Take a 30m 0.53mm ID column 1 micron column and put a 10m 0.2mm ID tube ahead and behind it and see the difference in retention times, plate number, and separation for yourself. You have to maintain the same flow rate EXACTLY.


Ron, Its an interesting experiment:
But can you coach me how can I measure retention times only for the the 30/0.53 having a 10m 0.2mm column on each end? this system is effectively 50 meters long..
I need to compare this with the 30m x 0.53 in normal flow(low-pressure) mode, but need secure reference points.

Now If there is a difference, a short restrictions (say 10mm 25 um) on the outlet may be easier to do a 1:1 comparison.. (do not understand why do you need a restriction at the inlet.. the pressure should be built up by the restriction in the back..).
We still talk about the impact of higher pressure on retention over a liquid film, right?

Next situation: If you are correct, shouldn't we also find higher K values for longer columns as they are also tested at higher pressures.. I will look at some data..

jaap

[JI2002]

The relative time (retention factor) an anlyte partitions in the two phases only depends on thermodynamic factors (temperature and phase ratio). Retention factor is the same at the inlet end and detector end of the column, but the linear velocity is different, so an analyte would travel at different speed at different part of the column. Although it seems retention factor has changed in this particular case, in reality it doesn't because the chemistry is the same.

[jdezeeuw]

Thanks JI 2002
what's the outcome? if I put the uncoated section at the end: Longer, less or the same retention..

[JI2002]

With guard section in the back, retention time is longer. But I wouldn't say the retention factor has increased because in my opinion retention factor should be measured without guard column (in the front or back). With guard column (in the front or back), you are measuring retention factor in a different column.

[jdezeeuw]

you are correct JI2002. The retention time increases. We can still calculate a Retention factor on the integrated guard and integrated "gap" columns, but like you said, it remains difficult for any sandwich column what we are really measuring. We measure this to test reproducibility of our integra-solutions.

Next step is to coat a column with a gradient film. Thick film in front for loadability (where we need it) and thin at the end.. (where the peak is broader).. how would we define retention factor for such a system?

[jdezeeuw]

Hi, Getting back to Ron's indication that retention depends on actual pressure inside the column. Promised to check some data:

30 and 60m 0.25mm Rxi-5ms columns are coated with same amount of polymer; columns are iso-tested. Pressures 110-220 kPa hydrogen.

Retention factors for 2000 columns: 6,49 for 30m and 6,45 for 60m. If there is any effect of increased retention due to higher pressure, in the range we work in, we do not measure this. Based on this, I would not expect the pressure to have any effect on the retention for the 30+10 combination we started this discussion with..

Ron, is my statement correct or do I overlook something?

rgds
jaap

[jdezeeuw]

It remembers be to an experiment I did some time ago to check impact of High pressure on separations. A 10m/0.32 Rtx-5 column was connected with restriction at the end, so I could run it at much higher pressures and see how the van deemter curve changed.

Tested this column from 20 upto 80 psi isothermally. Of course the plates were dropping fast when doing high pressure separations. It left 7200 plates at 30.9 cm/s @ 80 psi(with restriction) while this column produces 30.000 plates @ 4psi., 32.6 cm/s (no restriction).

I checked this data also on the retention factor discussion we had above.
Also here retention factor measured at 4psi was the same as measured at 20 or 80 psi. So again no indication that pressure impacted the retention factor, but now we have data up to 80 psi. If the effect is there, it must be very small and not measurable under the conditions tested.

[Peter Apps]

Hi Jaap

As you increased the pressure, did you decrease the flow rate to stay at the minimum on the van Deemter curve ? - the maximum plates that you can get from a column is supposed to be independent of carrier gas, but you have to wait longer to get them with more viscous gasses.

Peter