Stacked injections with Shimadzu or Agilent?

[chris.singleton]

Greetings,


I am developing an isocratic method to quantify an analyte with UV detection. Though there are other things in the sample, they do not absorb at the wavelength I am observing, and so I am basically seeing one peak in a chromatogram and quantifying it. Since there will be so much extra chromatographic space in the analysis, I would like to use stacked injections (inject every 15-30 seconds or so) with a fast method. Has anyone had experience with setting up stacked injections with a Shimadzu UFLC XR running LCSolution software, or using Chemstation running Agilent pumps and UV with a CTC autosampler?

When I called Shimadzu they said it was not trivial, I would have to take the autosampler off-line to do it. I'm sure I could do this, but I don't know if I could justify using half a day to do this when it wouldn't save that much time with the actual assay.


Thanks,

Chris

[bisnettrj2]

I don't have experience with what you're suggesting, but Michael D. Nelson and John W. Dolan wrote an article along similar lines. I'm not sure it will help, but it might give you some ideas.

http://chromatographyonline.findanalyti ... rticle.pdf

[macgyver]

Here's an idea...

Modify your current method to be only 1 minute long. Say your single peak comes out between 3 and 4 minutes, the first 3 datafiles won't have any peaks, the fourth datafile will have the peak from the first injection, the fifth datafile from the second injection etc.

Something like this:

Run 1: Inject Sample 1, No peaks
Run 2: Inject Sample 2, No peaks
Run 3: Inject Sample 3, No peaks
Run 4: Inject Sample 4, Detect Sample 1 Peak
Run 5: Inject Sample 5, Detect Sample 2 Peak

This example assumes there is no autosampler injection overhead, so you may have to fine tune the acquisition time in the method.

Could this work? (I might even have to try this myself!)

[chris.singleton]

Here's an idea...

Modify your current method to be only 1 minute long. Say your single peak comes out between 3 and 4 minutes, the first 3 datafiles won't have any peaks, the fourth datafile will have the peak from the first injection, the fifth datafile from the second injection etc.

Something like this:

Run 1: Inject Sample 1, No peaks
Run 2: Inject Sample 2, No peaks
Run 3: Inject Sample 3, No peaks
Run 4: Inject Sample 4, Detect Sample 1 Peak
Run 5: Inject Sample 5, Detect Sample 2 Peak

This example assumes there is no autosampler injection overhead, so you may have to fine tune the acquisition time in the method.

Could this work? (I might even have to try this myself!)

I like that idea, it's an easy way to get around doing stacked injections explicitly. I'll give it a try and see what I come up with....


Chris

[lmh]

oooh, clever idea. It even means you can use the instrument software to do all your integrations, and the peaks will be at constant retention times, just shifted to the wrong sample.