Retention time shift

[chhubert]

During these few months, I find the retention times of analytes increase gradually, though very little but obvious. Seems not the problems on HPLC column and mobile phase, is there any part in the HPLC hardware that causes this problem? I am using Agilent HPLC 1100 bin pump, with the mixer replaced by a 0.010" id peek tubing in order to reduce system delay volume.

[danko]

Hi chhubert

Check the column temperature provided it is controlled – lower temperature will result in longer retention time. If it isn’t controlled, it might be a good idea to do so.
Check whether the pump delivers the desired flow. At last but not least check for leakages in the flow path. The latter should result in lower back presser.
I hope it gives you some ideas.

Good luck.

Danko

[Bruce Hamilton]

Good suggestions above, I'd check pump plunger seals, and also check whether your samples are clean. It just may be that the column is aging and some activity is appearing. If there is light crud accumulating throughout the column it may be retaining the peaks ( and you could see more tailing ), or the column may have been used with a mobile phase containing something like ion-pair reagents, and they are only slowly going away.

You could repeat the initial column qualification test, and see if peak times are the same. If peaks are delayed, wash the column with a more aggressive solvent that is able to dissolve any insoluble rubbish.

Bruce Hamilton

[chhubert]

Thanks for all reply.
Bruce, why you'd check pump plunger seals? Is it for testing any leak inside the system? But I didn't find any liquid leaking from the pump. I did change a new column, freshly prepared mobile phase but it's still the same - retention time shifted backward. Strangely enough, I tested using other analytes and columns and found the same phenomenon. Any clues?

[ym3142]

I found Prior suggestions were very good.
how long was the time span you were talking about?
hours? days? weeks?
was column equilibrated enough?
what kind of column? and mp?
do you know if they match well?

good luck.

[Bruce Hamilton]

Thanks for all reply.
Bruce, why you'd check pump plunger seals? Is it for testing any leak inside the system? But I didn't find any liquid leaking from the pump. I did change a new column, freshly prepared mobile phase but it's still the same - retention time shifted backward. Strangely enough, I tested using other analytes and columns and found the same phenomenon. Any clues?

The reason I suggested plunger seals is because I've had a very small leak that never dripped, just evaporated, and was eventually detected by increased retention times during a sequence. It depends on your solvent volatility and how much of a leak is required to increase the RT.
I assume that the system and analysis used to work fine, and has only just started to show this problem?.

Anyway the fact that it occurs on other solvents and columns precludes my crud theory ( unless you have really dodgy grade of a common solvent - like water ).

I assume your area is temperature controlled ( all solvents are at RT ), but also juat check to ensure none of the instrument fans are partially obstructed, allowing a solvent line/reservoir to warm up as the sequence proceeds.

Assuming the temperature is OK, I'd be looking very closely at all your hardware, starting with the inlet sinters in your solvents ( make sure they are all clean and free-flowing ), and the solvent mixing system eg does the problem occur on isocratic runs when you've premixed the solvents?, if you change the solvents over A <=> B, does the problem still occur? - accepting that Agilent may recommend aqueous buffers in one solvent position, I'd still try the other way around.

Are then any other solvent lines in reservoirs that could leak in?, is the degasser working properly?, are you using a plunger seal wash - if so, make sure nothing is leaking past the seals in either direction, the possibility of a wash solvent contaminating the injector ( unlikely ), or the injector rotor valve leaking across the seal ( more likely ), and then check the column oven ensuring that there are no leaks and that the temperature is as recorded.

The important aspect of trouble shooting is to estimate what volume of solvent blend, or individual solvents, is necessary to induce the problem. Then you can change procedures to systematically eliminate possible escape routes or contaminants.

Good luck,

Bruce Hamilton

[mbreslav]

If you are using Agilent 1100, then you may want to check both flow rate and pressure possible changes with time recorded and kept in the data file. Just load UV signal together with signal details.

[chhubert]

Thanks for all reply in particular, Bruce for his comprehensive advice. I'd try and let you know the progress in due course.

[chhubert]

Hi all,

The problem has been resolved by changing the outlet ball valve, though it didn't show any leak. Apparently the ball valve was partially stuck with residue, leading to unproportional flow of organic mobile phase. Thus analytes shifted backward.
Again, thanks all for your suggestions.

[tx]

What and where is the "outlet ball valve". I had the same problem with 1100 before.

[DR]

It is the ball/seat combination in the outlet check valve (one per pump). Ruby ball/sapphire seat combinations tend to get weird when fed 100% ACN. It has happened to almost everyone at some point...

There are generally ceramic alternatives to ruby/sapphire if you really need to pump pure ACN.

[tx]

Thanks DR

[chhubert]

Hi DR,

Normally in binary pump configuration, one channel is fed with 100% aqueous and another 100% organic, say ACN. In order to prevent the ball/seat configuration from getting "weird", do you recommend not feeding 100% organic in any channel by initally pre-mixing some aqueous phase into the organic one?

[DR]

For actually running the method in a routine setting, it is a good idea to have at least some aqueous content to the strong (ACN) phase.

For development work however, it is highly convenient to be able to run straight ACN, especially if you do a lot of development work. For those who do lots of development, or those who truly require 100% ACN from time to time, I would suggest the ceramic components.