Hello, I have repeatedly been told that it is a bad thing to mix a 100% organic solvent with 100% aqueous (i.e. 100% MeOH and 100% H2O)using the HPLC on-line mixer for isocratic analysis. The line of reasoning being that mixing the two solvents in the HPLC and the resulting temperature/volume changes in the mixing chamber and lines can lead to solvent "out-gassing" and baseline problems. If I am running an isocratic separation with an 80/20 MeOH/H2O MP, I have always pre-mixed the solvents, but a colleague recently suggested that we should use the on-line mixing to reduce MP preparation variability. I have looked for some literature to back up this "out-gassing" claim, but have thusfar come up empty. Does anyone know for a fact that on-line mixing of pure solvents is OK are not? We use Agilent 1100 HPLCs. If there is no problem, then the suggestion was to add this on-line mixing to the methods in the future, but I have never seen on-line mixing of pure solvents for isocratic separations in any method that I have ever run, so I am skeptical.

The HP1100's are equiped with inline vaccum degassers and can handle online mixing. In fact the HP OQ/PV tests utilize online mixing. Take a look at pump ripple, the spec is <1% and is should normally be <0.5%. If you are having any outgassing problems they will be evident in the % ripple.

The equipment can perform online mixing, however I still would not use it. Online and offline mixing are not chromatographically equivalent. I have even noticed differences between online mixing with the quaternary pump vs the binary pump.

Also, I like to use the organic fraction of my mobile phase as a biocide. If you use plain buffer or water in one channel you will have to be more careful to prevent the formation of a biofilm in your system(s). The numbers I remember are that 2% ACN or 10% MeOH are bacteriostatic. My feeling is that when all things are considered offline mixing still wins. Good Luck.

In contrast to the above opinion, I would always choose on-line mixing over pre-mixed mobil phases if possible. This assumes that my hardware can handle it (Waters Alliance and HP1100 binary and quart certanily can) and well as my detector (no RI). The reasons are simple, first off you always get a more consistant(day to day, week to week) mobil phase compositon. Chemist to chemist variations are reduced. As well small changes to the MP can be made quicky and easily, you need 49/51 instead of 50:50, no problem, and 1 less liter of un-needed MP down the drain. The ability to make these small changes can be very important during method development. Tom is quite correct in pointing out the possibility of bugs growing in the mobil phase, but proper filtering (through sub-micron filters) helps as well as providing fresh aq. solvent (or buffer) once in a while. And, just as a watch out to Tom, I have, in the past, seen microbes grow in 10% MeOH/Buffer solutions, not often mind you, but I have seen it.

Just another opinion.

Andy

I agree with Andy. You'll never get same MP consistency using offline mixing. The only time we use offline mixing is when using RI detector at high sensitivity, because baseline is lower there when using 1100 set up for 100% of a single channel.

Just to keep the score even, I agree with Tom. It doesn't take very long to mix an isocratic mobile phase. Methods will be more transferable because not every lab is using the latest equipment. In a previous life as a field service eng. I saw many problems associated with biofilm - some of these problems can be tricky to diagnose and troubleshoot - it is a good idea to do what you can to prevent its formation.

Jack

One can easily prevent biofilm by always using fresh aqueous phase and using clean glassware. I would not hesitate to use either method and have gotten good results going both ways, but for a system that is sparged with Helium, I prefer to let the gradient controller do the mixing due to the differing vapor pressures of components in a pre-mixed mobile phase. In cases when my mobile phase involves either very little of any one component, say less than 10%, I will often dilute that component, then set my gradient controller accordingly. For instance, if I want a gradient from 0-10% MeOH over 30 minutes, I'd dilute methanol in aqueous phase at 20% (v/v), then run a 0-50% gradientusing the gradient controller. With systems equipped with inline degassers, the differential vapor pressure problem pretty much goes away.

Merely another opinion in the mix!

Chris

My thanks to all who have contributed. It seems that there are valid reasons both methods. I have learned quite a bit. I think that it may be sufficient to say, "If it isn't broke, don't fix it." Thanks.

Hi Noname,
Pre-mix them unless the assay does not count for much.
Do not mix them with volumetric methods, use your scale to weigh them. Please send me an email if you can think of an consistency of concentration argument for weighed solvents.


Jim jimgorum43@cs.com

Guys, please make sure your proportioning valve is working on your equipment before you do anything else. I spent weeks trying to figure out why my results were not consistent. Slight changes in %MP can have a devastating effect on separations. I made the mistake of trusting when I should have been testing.