Method sensitivity issue- Please help

[rgonsal]

I'm developing a method for bisphosphonates using Ion-pair RP (tetrahexyl salt). I'm already limited with the sample concentration of 0.05mg/ml. I need sensitivity @ 50 nanograms/ml. The problem i'm facing is due to the tailing of this bisphosphonate, the peak area is good enough (but the height is small). Injection volume is 200-250 microliter on 250 x 3.0 mm, 5micron column. detection = 215 (210 is max). How can I inject a larger volume without dispersing the sample band or increase the signal?
Suggestions are very appreciated.
Thanks!

[tom jupille]

To get "on-column concentration", the diluent should be much weaker than your mobile phase. You didn't specify either one, but 250 microliters on a 3mm ID column is already getting fairly large (the equivalent of over 500 microliters on a 4.6 mm ID column), so I don't know how much more you could get.

LOD and LOQ are ultimately limited by signal/noise ratio. If you can decrease your baseline noise by 10x, that drops your LOQ by 10X. So:

- if you move down toward the 210 nm absorbance maximum, does your signal increase more than your noise?
- how much noise do you have (and what is the manufacturer's spec for noise on your system)?
- have you tried to optimize detector settings (things like bandpass, time constant, etc.)? If you're using a PDA, you can play with these ex post facto to see what happens.
- is your pump operating at it's best (degassing, check valve operation, etc.)?

Every little bit helps.

[rgonsal]

Hi Tom,
The diluent is the mobile phase. The IP reagent itself has absorbance @ 215/210 nm (changing to 210nm didn't help). The detector used is a PDA , but seems like nothing works. I have to explore the variable VWD (we have Agilent's LC in our lab) in the hopes of having a lower detector noise. I have been using Optima grade ACN as well.
How can I possibly eliminate the tailing of bisphosphonate inorder to get a sharp peak?

[tom jupille]

If your diluent is the mobile phase, then 250 microliters is a tremendous volume overload on a 3mm ID column, and I suspect that's the genesis of your peak shape problem. What *is* your mobile phase? and can you use something weaker (i.e., less organic solvent) as the diluent?

The detector used is a PDA , but seems like nothing works.

What is the noise spec for your detector?
How much noise are you actually seeing?
What, exactly, have you tried that hasn't worked?

[rgonsal]

Hi Tom,
I don't know the exact noise spec for the detector. The mobile phase is phosphate buffer with EDTA pH=7.70 (900ml) combined with 100ml of ACN.

[tom jupille]

You might gain a bit by using water or buffer as your diluent (your mobile phase is 10% ACN).

I don't know the exact noise spec for the detector

You should be able to find it in the manual or on the manufacturer's web site. If you are operating in a regulated environment (GLP or cGMP), you should have documentation on the performance qualification (PQ) tests run on your system. That qualification should have included a noise measurement.

Once you know what the detector *should* do, you can measure what it *is* doing to see if improvement is possible.