We are using a reverse phase gradient system (waters Resolve C-18) and are having issues with our retention time shifting by as much as 20 min after running approximately 500 samples. We are running mostly extracts from biological tissues such as serum and livers. Our pressure remains about the same; probably an increase in 200-400 psi so I assume not too much is building up. We have tried cleaning the column with THF and it does not solve the problem. We are thinking that the mobile phases maybe the problem and interacting with the column and destroying it. Our solvent A: acetonitrile/water (95:5) containing ammonium acetate (10mM) and triethylamine (0.1%)and Solvent B: acetonitrile:methanol:dichloroethane (85:10:5) containing ammonium acetate (10 mM) and triethlyamine (0.1%). We do leave the column overnight in solvent A because we are using an autosampler and letting it run overnight (although it does not run the whole night). Does anyone have any suggestions as to why our columns are being destroyed?

In which direction is the RT shift (peaks now earlier or later)? Is your shift in RT gradual (20 minute shift over the course of the run) or rather abrupt? Do you control the temp of your column? If so, what is the temp?

You could also try reading the literature that came with the column (or check out Waters' website) and see if any technical notes have anything to offer (such as solvents to avoid, sample matrices to avoid, expected column lifetime, etc).

Do I read this right? Your gradient goes from 95% acetonitrile with 5% water to acetonitrile/methanol/dichloroethane? And both solvents contain ammonium acetate salt and triethylamine?

From "shifing retention times": The retention times are shifting gradually shorter by 20 min over the course of about 500 runs. You can see the retention times getting shorter day by day. Both solvents contain triethylamine (0.1%) and ammonium acetate (10mM).

From "shifting retention times": Yes our solvents are going from A: 95:5 (Acetonitrile: water) to B Acetonitrile: Methanol: Dichloroethane (85:10:5)using the following gradient: 100% A for the first 3 minutes; from 3-10 min, 100% A to 100% B; and to end 100 B for 20 min. Both solvents contain ammonium acetate and triethylamine.

If I may go off on a tangent:

I have never been able to get even 10mM ammonium acetate to dissolve into such high acetonitrile concentrations (85-95%). Even if I dissolve it in the aqueous portion first, then add the ACN, I get phase separation that appears to be two immiscible liquids. Yet the popular conception seems to be that it is possible to get high ammonium acetate concentrations into acetonitrile by dissolving it in a small amount of water first. I have found this to be possible with methanol, but not acetonitrile. I would love to hear others' comments on this.

[In the case of "shifting retention times", triethylamine is also an additive, which could change the solubility of ammonium acetate in that system.]

Perhaps part of the problem is that, according to your gradient program, you never re-equilibrate with A before beginning your next injection. I would try re-equilibrating with A for about 10 minutes before starting the next injection.

None of your solvents seems harsh enough to actually 'destroy' the column, which is your basic C18 column.

From "shifting retention times": Thanks for the feedback. Here is a response from comments thus far: the ammonium acetate is in solution. For solvent A, we use a sonicating bath (sonicate a few seconds) to dissolve the ammonium acetate in the water before adding the acetonitrile. It does not fall out of solution and we do not have 2 immiscible solutions. For solvent B, add the methanol to the ammonium acetate and then we sonicate (a few seconds) and then add acetonitrile. we do not have problems with miscibility or the ammonium acetate falling out of solution in solvent B either. FOR THE GRADIENT, after the run, we do bring system back to 100% A over 2 min (from B to A). Then equilibrate it the system on 100% A for 3 min. We have tested the equilibration time and have validated it to be long enough (our retention time does not change whether system is equilabrated for 20 min or 3 min within a given day). Our gradient system is run on 2 ml/min.

I see two significant issues with your system. First, you are using a highly alkaline eluent system (the only saving grace being you have a low water content which might be reducing the rate of stationary phase attack). But still, it's reasonable to expect that this mobile phase should be "corrosive" to your column. You would be better off using one of the phases designed specifically for use under high pH conditions if you really need to be using such extreme conditions.

Second, triethylamine can be expected to react with dichloroethane to produce quaternary byproducts. The initial product would containing a residual Cl in the beta position relative to the quaternary nitrogen but given enough time a diquat can also be formed in this eluent. These compounds may not be readily removed by your mobile phase. It would be best to avoid leaving triethylamine in contact with dichloroethane in order to avoid the possibility of this reaction or at least the eluent needs to be freshly prepared. Removal of the byproducts from the column will probably work better with a higher water content mobile phase.

In retrospect, I was thinking about your system and there's a third even more daunting possibility that is consistent with this system: the ability of ammonia to react with dichloroethane to produce a condensation polymer. Your buffer contains essentially equimolar amounts of triethyl amine and ammonium acetate so considering the relative basicity of the two amines, your buffer system is really better described as triethylammonium acetate and ammonia. I have a fair amount of experience producing condensation polymers of this sort and would described typical eluent conditions as not favoring the condensation reaction but it does seem likely that a small percentage of the condensation polymer could be formed even under ambient temperature conditions. The polymer should bind tightly to the column and be difficult to remove, but your best hope would be to use a fairly highly aqueous eluent with significant ionic strength.

You have a very strange method. Please tell us about the compound(s) that you are trying to analyze! One would expect that the mechanism is not reversed-phase chromatography, because there are very few things that are retained on a packing by RP under your conditions. Also, you are working at alklaine pH, which is guaranteed to strip of the coating on the column that you are using. How is the retention of your compound(s) changing? Increasing or decreasing? This may help us understand the mechanism.

a guess at what you are analyzing are lipids. that is the only time I can think of using non-polar solvents in RP for biological samples. My guess is you still have glycolipids and other "lipo-biostuff" material tangled in the C18 chains. There polar substituents are facing the mobile phase and thus reducing RP selectivity slowly as they build-up.

Secondly, could the salts of your samples be depositing in the column?

sound plausible?

If so, two solutions:

1) can you use a C4 column instead?

2) wash the C18 in a very strong solvent. My suggestion would be 2-propanol, hexanes, 2-propanol, solvent B. See if this regenerates it? (and be careful of ignition hazard)

for salts, a higher aqueous wash should do, someone can recommend a good de-salting mobile phase.

To A. Mouse and anonymous:
From "shifting retention times"
Thanks for your feedback. We are analyzing lipid soluble components such as carotenoids and retinoids; both of which are very unstable. We are using hexane, ethyl acetate, dichorolmethane extracts from tissues such as serum and liver for analysis. While this method works for these lipid soluble components it also separates the isomers of carotenoids which is tricky (for example, lutein vs zeazanthin). I will try washing with suggested solvents. Do you have a "desalting" suggestion?

For reproducible retention of your compounds, you do not need the triethylamine to make the separation work. I see no good reason why it is in your mobile... maybe for supressing interferences. But it will attack your stationary phase, and I have some doubts that you will be able to recover the column. The amine will remove the C18 from the surface.
Honestly, I think your column is dead and you need to buy a new one.