Chromatography Forum

Note: this is not my thought experiment, and I do not know the answer. I just thought this was an interesting thought experiment to pose to the Forum. I would assume this came up on accident (someone installed a column with an integrated guard backward and noticed weird retention times). In any case, here's the text of the blog post/question.

http://blog.restek.com/?p=2316

"Chromatography challenge: Does a GC-capillary column produce different retention times when installed in the opposite direction? Share your opinion!"

April 3rd, 2011 by Jaap de Zeeuw
A little challenge for our regular blog-visitors..

Capillary chromatography has become quite an adult science, but we still discover new and better ways to do better chromatography and generate more accurate data. Besides that GC columns get very inert, have low bleed and are usable up to high temperatures, also the design of columns has changed.

With “design” I mean how columns are physically made. By using “segment” coating we can deposit stationary phase in different positions of the capillary. Instead of having a long, uniform coated capillary, we can make now an Integrated Guards or Integrated Retention-gaps, all in one piece of capillary. So, new generation columns often have a “front” and a “back” side.. Now an interesting discussion starts: “what can I expect to happen if I install my column in the “wrong” way?”..

The challenge is here (see the link for the associated graphics):

We have a 40m x 0.25mm Rxi-5Sil MS capillary, where we have an integrated guard column of 10 meter. The coated analytical column is effectively 30 meter long. The 10m guard section is only “deactivated” and the retention will be minimal.

The same column, used in 2 different directions: A: with guard section in FRONT; B: with Guard section at the BACK
We position the column in our GC first in position A, see figure. Here the guard is direct connected at the inlet. The column is heated to test temperature and we check the linear gas velocity by injecting methane; Then we run a test mixture and we measure the retention times for all components.

Now we change column position into position B. The Guard section is now directly connected to the detector. We use exactly the same column and do exactly the same exercise as above.

The Question is: What will happen with the retention times for all components?

Note:

We use the same column, so the length, ID and amount of stationary phase is exactly the same;
We test at the same temperature and use the same inlet pressure;

Our Options:

1: The components in situation “B” elute at exactly the same retention times a situation “A”;

2: The components in situation ”B” elute FASTER, so it seems the column has less retention;

3: The components in situation “B” elute LATER. So it seems the column has more retention;"


So, GC people - what do you think? Is this a slam dunk for the experienced GC operator, or is this a bit trickier than that?

Hint:

If the coated section of the column is homogeneous (and sometimes a column may not be) then:

The answer is not 1.

best wishes,

Rod

I think that this a special case of two columns of different selectivities joined in series - which at one time was a reasonably popular way of tuning selectivity by changing temperature programmes, in some ways a precursor to GCxGC.

The critical parameter is whether the column temperature was programmed or not - since nearly all separations these days are programmed, let's assume that it was.

With the guard column at the beginning, the analytes are in the stationary phase when the oven is marginally hotter, so they are moving faster and eluting earlier. Vice versa with the column backwards.

There is probably also some complicated effect of pressure and flow gradients.

Peter

I would say that the answer is 3. In the situation where the guard column section is before the column section, the exit of the column section is at atmospheric pressure ( for the sake of this argument) if you are using a FID , for example. If the guard column section comes after the column, then the exit of the column section is no longer at atmospheric pressure, you have the pressure drop across the guard column section.

From an instrument point of view, you have the same head pressure and can measure the same flow at the exit of the combined column, but the section of the column that does the separation is not working under the same pneumatic conditions.

Gasman

Gasman is correct.

Any analyte that is capable of being retained by the column phase will be retained longer in the stationary phase when under higher pressure, thus having a greater retention time. On average the coated column has a greater pressure over its length when the guard column is behind the coated column, not ahead of it.

The difference will be minimal for very volatile components, and more noticeable for less volatile components.

Happy Friday everyone.

Rod

Hi

Well how do you interpret "Segment coating technology allows the retention gap or guard column to be built directly in the same piece of tubing " ?

Sounds to me as it is the same dimension for retantion gpa/guard column part as for the analytical part or is it not? In fact does not the question state that is is a 40mx0,25mm column where the first 10m is deactivated?

The 40 meter column is prepared by drawing a measured amount of solvent (10 meters amount) nto a tube followed by the 30 meters of phase coating solution.

The tube is evacuated with the initial 10 meters of the tube remaining uncoated while the remaining 30 meter length of the tube has a layer of phase deposited. The phase is then bonded to the tube.

Most users have no idea how a capillary column is made, or only has a meager idea of the techniques used.

It is not trivial but it is interesting, especially when one learns how to plug a tube so the solvent can be evaporated out of the remaining end. Watching it happen with FSOT tubing is quite interesting initially, but then becomes as boring as watching grass grow.

Most manufacturers won't allow guests to observe the process.

best wishes,

Rod

Hi Guys,
tanks for the discussion. You almost got there.

It has nothing to do with pressure. TIP: Think about the velocity..

regards
jaap

Rod
a different way of making them is: fully coat them and rinse out one side with a plug of solvent before polymerization..

jaap

My answer is 3. With guard column in the front, the average linear velocity of the carrier gas in the coated section is faster, hence retention time is shorter. With guard column in the back, the average linear velocity is slower in the coated section, the retention time is longer.

I didn't see Japp's post before I had my answer.

Krikos,

Take the following. You have your 40 meter column with 10 meters without a coating. Let us say that you set a pressure of 30 psi to get the required flow that you need. It does not matter which way you connect the column, if you put 30 psi on your column you will get the same flow. Assume that the over the 10 meter length of uncoated column there is a pressure drop of 5 psi. This would mean that when the uncoated end is connected to your 30 psi source, the coated length of column is running with the 'inlet' end at 25 psi and the outlet at 0 psi. If you reverse your column connection, the inlet end of the coated column is now working between 30 psi and 5 psi.

I hope this makes it a bit clearer for you.

Gasman

Jaap,

Yes, but the reason that the velocity is different is that the pressures are different.

Gasman

Dear JAAP

There is more than one way to skin a cat. Never said there wssn't !



Rod

Assume that the over the 10 meter length of uncoated column there is a pressure drop of 5 psi. This would mean that when the uncoated end is connected to your 30 psi source, the coated length of column is running with the 'inlet' end at 25 psi and the outlet at 0 psi. If you reverse your column connection, the inlet end of the coated column is now working between 30 psi and 5 psi.

Gasman,

The pressure drop over the same length of the column is different across the column. You may have 5 psi pressure drop in the first 10 m of the column(the injector end), but you may have only 2 psi pressure drop in the last 10 m(the detector end).

Krikos,

Take the following. You have your 40 meter column with 10 meters without a coating. Let us say that you set a pressure of 30 psi to get the required flow that you need. It does not matter which way you connect the column, if you put 30 psi on your column you will get the same flow. Assume that the over the 10 meter length of uncoated column there is a pressure drop of 5 psi. This would mean that when the uncoated end is connected to your 30 psi source, the coated length of column is running with the 'inlet' end at 25 psi and the outlet at 0 psi. If you reverse your column connection, the inlet end of the coated column is now working between 30 psi and 5 psi.

I hope this makes it a bit clearer for you.

Gasman

Apologies, Gasman/Rod, I thought you guys misinterpred the question ie that the guard/retention part had a different dimension than the analytical part lol

Need some sleep I guess