Poor semi-prep peptide recovery

[chichibabin]

Dear All,
I knew semi-prep HPLC was fairly inefficient in terms of recovering peptide but I was very surprised how inefficient it was compared to analytical HPLC. Using the same column in analytical and semi-prep sizes I can recover 500% more peptide with the analytical column. Is this normal? My peptides are precious and I can only prepare small amounts (a few mg) so I have recently started purifying by doing multiple runs and loading the analytical column high as I can get away with.
Is there anything I can do to improve peptide recovery with semi-prep columns as doing multiple runs is very tedious.
Thanks in advance,
Chich

[tom jupille]

Offhand, I can't think of any reasons why recovery from a semi-prep column run under equivalent conditions should be different compared to an analytical column packed with the same material. So we need to establish whether you are really running equivalent conditions.

More detail would certainly help: packing (chemistry and particle size), column dimensions, flow rate, injection volumes, gradient conditions, etc.

[chichibabin]

Thanks. That's encouraging to hear recovery should be the same.

I am using:
Mobile Phase A H2O/0.1% TFA
Mobile Phase B MeCN/0.1% TFA

10-40% B over 30 minutes

Analytical
C4 5 um 300 Angstrom 4.6 mm x 250 mm
flow rate 1ml/min
sample volume 20 uL

Semi-prep
C4 5 um 300 Angstrom 10 mm x 250 mm
flow rate 3 mL/min
sample volume 100 uL
With the above 2 conditions I recover the same amount of peptide with each. This is based on the peak area at 214 and 280 nm.
The gradients are not quite the same as I would need to be running the semi-prep close to 5 ml/min but that shouldn't make too much difference.
Sat

[chichibabin]

Any more suggestions?
I should reiterate that I only recover the same amount of peptide when loading 20 uL on the analytical column as 100 uL on the semi-prep.
Cheers,
Sat

[HW Mueller]

You are using the same detector and setting for both columns?

[tom jupille]

You are correct that you should be running at a flow rate of about 5 mL/min in order to have "equivalent" conditions on the two columns (actually, 4.7 mL/min, since the flow should scale to the square of the column diameter). The difference between 4.7 mL/min and 3 mL/min can have a significant effect on peak spacing, especially when you are dealing with peptides. If you are collecting fractions based on time, then you may have gotten only part of the actual peak.

If you are collecting fractions based on detector response, then I don't know what to tell you. As I said earlier, I can't think of a reason your recovery should be that much lower.

[Uwe Neue]

Check on delay volumes in your prep setup! Are you collecting post detection, or parallel to detection?

[chichibabin]

Check on delay volumes in your prep setup! Are you collecting post detection, or parallel to detection?

Thanks but I am quantifying by peak area so errors in collection are not to blame.
Sat

[HW Mueller]

We still don´t know whether you used the identical detector (flow cell) and whether you calibrated everything, for instance did you account for differerences in flow rate?

[AdrianF]

Perhaps the analytical system has a 10mm detector cell and the semiprep system a 2mm cell.

[chichibabin]

We still don´t know whether you used the identical detector (flow cell) and whether you calibrated everything, for instance did you account for differerences in flow rate?

Hi, So do some detectors switch to a larger flow cell based on flow rate? I used the same system for analytical and semiprep runs and the only different parameter was flow rate. I am using a Varian Prostar unit.

How would one account for flow rate when peak integrating? This is something I have not done as I was assuming peak areas were directly comparable regardless of flow rate. Am I in error?
Thanks,
Sat

[Uwe Neue]

Sorry, I did not read your setup carefully enough. You are mistaken in the assumption that you will get a 5 times larger peak area when you change the column volume and the flow rate.

The ratio in column volumes is close to 5x. Your change in the injection volume is 5x. Your peak area would remain identical, if you would have changed the flow rate also by 5x. Since you change the flow rate by 3x only, your peak area will increase by 5/3.