Injecting Very Large Volumes in HPLC

[adam]

We have an application where analytes are present at very low levels - like 0.05 ug/mL. We could use something like solid phase extraction to concentrate things, but we prefer to keep things simple. And since its pharmaceutical QC it is preferrable to stick with conventional UV detection.

Since this is a gradient method, and the analytes of interest are late eluters, I thought a good approach would be to try injecting large volumes. The HPLC software allows you to program such that you can inject, for example, a 40 uL volume 10 times, with 30 or 60 seconds delay in between each injection. We have just started to investigate this.

Does anyone have any experience with this kind of tecnique? It seems like a simple way to get more sensitivity from standard equipment. And something that would be of interest for situations like this or for cleaning validation, etc.[/img]

[tom jupille]

You can get away with suprisingly large injection volumes (my personal record is 10 mL) if your sample is dissolved in something much weaker than your mobile phase which elutes your analytes. In the case of a gradient, you could probably get away with a large volume if your diluent is 100% aqueous. Early eluters will probably be a mess, but the late eluters should look OK.

You will probably have to start the gradient at a lower %B than you might otherwise in order to give the messed-up early eluters a chance to clear the system, and there is no good way to judge a priori how large an injection you can get away with (you'll just have to try increasing volumes until things fall apart).

You may also want to look closely at the economics. Simpler is indeed better, but injecting a large volume will very probably result in a significantly longer run time, and the hit to throughput may negate some of the savings from avoiding sample workup (your mileage may vary! ).

[DR]

In addition to what Tom said, I would suggest that you look into whether your autosampler can accomodate larger sample loops (see if programming supports it). Some vendors instruments can take 2 or even 5mL loops (some require a syringe swap too).

[HW Mueller]

Your intermittant injection may cause problems if you have mobile phase flowing inbetween. My record injection amount (routinely) is around 8 mL on a normal 4.6x250mm RP column, but that was in one bolus (in one go on a large loop). The start of the chromatography is then just delayed by the ratio ml inj./flowrate (provided that the sample solvent fits the criteria mentioned by Tom).

[Bruce Hamilton]

My experience with large volume by multiple injections using an Agilent 1050 injector was that retention times varied much more than for single injections, especially for long sequences, which I attributed to a lot of junk eluting before the peaks of interest.

The reason we didn't pursue the method was that the pressure was varying noticably on the samples, going high and then reducing as the early-eluters washed off the column. If it was for fraction purification, it would have been OK, but for a routine analytical method, it seemed a bit uncontrollable.

If it works for you, that's OK, but sometimes simple solutions aren't.

Please keep having fun,

Bruce Hamilton

[SIELC_Tech]

If your compounds are ionic and hydrophobic in nature you can use a guard column (mixed mode guard) to trap you compound on the guard. Starting with lower concentrations of your gradient should not move your compounds (no ions in mobile phase). We had a project in which we used Primesep 100 guard to trap hydrophobic basic compound with initial point of gradient ACN/water-5/95.

[adam]

Sielc-Tech

Interesting. But what did you use to elute the material from the guard.

If the guard column retains more strongly than the analytical column, it will help focus the analytes, BUT there will then be a band broadening process as the analytes transfer from the guard column onto the analytical column. Unless, you somehow changed the mobile phase to make the guard column less retentive (after the initial trapping).

Please let us know what, exactly, you did. It sounds like you may be on to a very clever trick.

Adam

[SIELC_Tech]

Unfortunately, I cannot reveal you exactly what we did for this particular customer. But I can give you a similar example.
Lets say that you have hydrophobic basic molecule which retains on regular C18 column. pKa of you basic group is around 9-10.5 (most of the amines). If you use mixed-mode cation exchange guard (Primesep 100) and start with no ions in the mobile phase (ACN/water) your hydrophobic basic compound will be trapped on the guard by combination of ion-exchange and reverse phase mechanisms. In mixed mode chromatography 1+1 is never 2, because you have synergy of interactions. Then after you pass all "injection" you can increase buffer strength to facilitate ion-exchange elution. While you injecting you can either pass your "injection" through the column or divert it to waste similar to this approach.
http://www.sielc.com/pdf/SIELC_September_2004.pdf

In you case you are trapping and releasing your compound into column and not diverting late eluters as in the newsletter. If you compound is hydrophobic and acidic then you need anion-exchange mixed mode guard (Primesep D, Primesep SB).
For hydrophobic basic compounds (Cetylpyridinium or dextromethorphan) you can use Primesep 100 guard and Primesep D column. The latest will give you good shape for basic analytes.
http://www.sielc.com/application_151.html
http://www.sielc.com/application_065.html
http://www.sielc.com/application_066.html

You need to play with conditions and watch for ions in your sample. If you have a lot of salts in the sample and you are injecting large volumes this will not work, as ions in you injection will be eluting your compound from the guard.

Regards,

Vlad

[mcbart]

If you need to automate your large volume injection you can consider HT300LV

[url]http://www.hta-it.com/content/view/23/37/lang,en/[/url]