Strange behavior from Atlantis T3 column

[shaun78]

I have recently started using Atlantis T3 columns for various reasons, but the primary would be that the phase is available in both a HPLC and UPLC format. Validated methods are being migrated from a Synergi Hydro to the Atlantis T3 HPLC columns and eventually down to a UPLC T3 column.

All in all I am happy with the results that I have been getting with the Atlantis column. However, I have noticed something that I find a little odd: this column seems to be going through something very similar to traditional column dewetting, but it is happening under 100% organic conditions.

Some of the methods that I am migrating to the Atlantis column have a 100% ACN wash at the end of the run. I have noticed that I have to let a 4.6 mm x 150 mm column equilibrate for about 20 minuets at a flow of 1.6 mL/min in order for me not to experience analyte retention time shifts in subsequent injections. Clearly, this retention time is excessive given flow and column dimensions.

I have found that if I change the wash step to 75:25 ACN:Water, then normal column equilibration times apply.

Initial mobile phase is 50:12:38 Water:MeOH:ACN.

I am curious to know if anyone else has observed this or, more importantly, if anyone can tell me what is causing this? Seems I have the problem sorted, but I am very curious as to the "why".

[pharmaguy]

Monitored this post for a couple days to see if anyone else answered... I'll stick my neck out. T3 seems to be relatively high carbon so it could be the same process that "dewets" standard C18 columns, but in this case the water can't displace the ACN. Do you immediately switch back after gradient completion to 100% water or is it a slower transition over a few minutes? Going to 100% water still requires overcoming the pressure/contact angle equation to wet the pore/phase.

[Uwe Neue]

The T3 column has a lower carbon content and has been designed specifically to be run in 100% water. While we have a large experience with the T3 column, we have not observed the problem that you have described, nor do I have a rational explanation. Since you seem to be successful with a 75% wash, the simplest path forward seems to me to use this in the future work.

[HW Mueller]

Seems a simple case of ACN precipitating something out, "coating" the column, while the H2O/ACN washes it out (salt, protein?).

[shaun78]

Thank you all for your insightful replies.

I noticed this happening on both standards and samples. So, if something is precipitating, I would not expect to see that from a standard. As far as potential buffers/salts go, there are none in the mobile phase. Initial conditions are literally 50:12:38 water:MeOH:ACN.

The gradient runs linearly to 35:12:58 water:MeOH:ACN over 15 min. and steps to 100% ACN and holds that for 5 min. The acetonitrile step is being used to quickly elute some excipients that we do not want to assay. The mobile phase then returns to initial conditions over 2 min. The change to 75:25 for the wash does not significantly impact the length of the wash step either.

I would be inclined to think that pharmaguy is correct. This is what I initially thought as well concerning displacement of the ACN. I have just never really observed that before and thought it was really only really occurred with water dewetting.

Thank you for your answer as well Uwe. I have searched around Waters and was not able to find anything relating to this either. It is interesting to know that you are unfamiliar with this as well and can’t come up with a rational explanation for it. I will definitely keep the wash at 75:25; I really am more interested in this out of personal curiosity at this point.

Again, thank you all for your input.

[HW Mueller]

Now I am even more convinced that my suspician holds, instead of removing the excipients with ACN you are "cementing" a component(s) in there. Any fatty acid salts or other detergents in the mix? If a series of standard injections did this, with the column never seeing excipients, then your St gets partially permanently retained and almost fixed there by ACN.
Your theory on ACN doing dewetting, or whatever, can be discounted out of hand, unless you re-invent chemistry, thermodynamics and kinetics.

[shaun78]

I don't think I fully understand where you are coming from HW Mueller.

Samples and standards are dissolved in 30:70 water:MeOH. Only three components are placed into the standard, two parabens and the api. As I have mentioned, this seemlingly long equilibration will be required even if I am only running a long sequence of standard injections.

In light of that, how could anything in the standard be getting fixed onto the column by the ACN wash (I put three analytes in and I see three peaks in the chromatograms that definately correspond to the analytes as shown from injections of each individual analyte)? I noticed this happening before the column ever saw a sample injection as well.

However, your ideas of excipient permantely getting fixed on the column by a wash just gave me wonderful insight into the problems of some of our other methods that are from before my time ...

Thanks for the clarification on possible ACN dewetting as well. I'll move on from that one after that --- thought it could happen but did not consider the physical reasons for why it could not.

[HW Mueller]

Ah, did I understand this correctly now, that
"this seemlingly long equilibration will be required even if I am only running a long sequence of standard injections"
to mean that if you inject one standard on a new column, then wash with ACN there is no long equilibration?
That would be clear proof that something accumulates on the column. Here is one direction from which I am coming: Can your API be positively charged? If you have cations stick on silanoates you will have a hard time of washing them out with ACN.
Does the API have detergent properties?
On fixing via ACN: A long time ago I got very good evidence that MeOH washes can fix proteins to columns. If proteins do this one can imagine that there might be some crystals that do this also.

[shaun78]

If I take a new column and inject one standard I still need the long equilibration time after washing the column with ACN.

The api is most certanly positively charged in solution, though it does not have any detergent properties that I am aware of. Would it still be possible that some of the cations of the api end up binding to this column, for which a pure ACN wash serves no purpose with respect to removing these cations? As a result, I need to requilibrate the column for excessive amounts of time with initial mobile phase in order to free up the active sites of the column again. However, when I wash the column with a mixture of ACN and water, the water portion is enough to remove the cations from the active sites of the column, making the initial equilibration with mobile phase quicker (back to what would normally be expected for the column size and flow).

You make a very interesting point here; one I certantly would have never thought of. This certantly seems to fit the reasoning as to why I was slowly and steadily droping retention time when I was using 100% ACN as the column wash.

Thank you much for your time and patience!

[HW Mueller]

Without an acidic buffer you will have a good chance of + charged species being strongly shifted toward Sio-M+. Depending on the nature of M+ the reversal will be quite different. Some compounds, like detergents, appear to have the property to built up on the initially bound M+ (like crystal formation or micelle formation). To be honest, I have only cicumstantial evidence for that.

[shaun78]

Again, thank you for your input.

Your theory seems to be about the only idea that fits. I thought what I was observing was a little odd and an indication that something else was at play, hence the post here.

Again, thank you for your time and input.