Maximum volume that can be injected for semi-prep

[zvi]

I would like to use my C-18 analytic columns (3 x 150 mm), Symmetry and XBridge, for example, for semi-prep purification. Assuming the mass of the analyte is not a problem as its solubility is relatively low, what is the maximum volume that I would be able to inject for such work? What is the formula or rule-of-thumb for such purposes?
Many Thanks.

[tom jupille]

If your sample diluent = mobile phase, then about 15% of the volume of your narrowest peak. If your diluent is stronger than the mobile phase, the injection volume will have to be reduced, but I prefer to experiment rather than rely on a rule of thumb in that case. If the diluent is weaker than the mobile phase, then you can inject a larger volume; again, rules of thumb are not terribly useful and experimentation is a better approach.

[zvi]

Thanks Tom for your super-fast reply.
My sample diluent is DMSO and mobile phase is methanol/water/phosphoric acid.
15% for a peak of width of about 1 min, flow-rate of 0.8 mL/min is 0.12 mL or only 120 microliters. That's it?! Ouch.
Wouldn't the injected volume depend on the volume or void volume of the column?

[tom jupille]

The "15%" rule represents the point at which the injection volume will begin to make a noticeable contribution to the peak width. The column void volume is subsumed into the peak volume (a larger column will be operated at a higher flow rate so the peak volume will be greater)

If you have "excess" resolution between your product and the closest impurity/contaminant, then you can increase the injection volume until the resolution drops below the "adequate" level (whatever that is). I'll repeat what I said at the end of my previous post: rules of thumb are not terribly useful (for this sort of thing) and experimentation is a better approach.

[Uwe Neue]

If your sample is still reasonably soluble in a solvent that contains (much) more water than the mobile phase, you can potentially inject a "ton" of sample. For example, if you can dissolve the sample in a solvent that contains about 20% more water, you can inject about 10 times more sample volume for equal peak distortion (this is a rule of thumb, the actual value depends on the sample). If you have a second pump, you can inject the sample as is, but dilute it just before the column with water or buffer delivered by the second pump. This technique is called at-column dilution, and you can find some information in the literature or at the Waters website. This again permits the injection of a much larger sample volume. How much depends on the composition of your mobile phase etc.

[zvi]

Thanks Tom & Uwe. You've really been tremendously helpful.

My compound of interest is quite hydrophobic and the best solvent that I found for it is DMSO (any added water will only lower its solubility). Now, I do not have an additional pump, which as Uwe calls it, would certainly be helpful. I completely understand Tom that experimentation, and not formulas or rules, is the only way to really see what's going on. But, for a first "guess" or "attempt" may I ask the following: If indeed the closest component (an impurity) is, say, 3 minutes away (before and after), how much volume would I then probably be able to inject?
Many Thanks.

[tom jupille]

Given that your solvent (DMSO) is signficantly stronger than your mobile phase, your guess would be as good as mine.