I am working on an HPLC method that separates our target protein from other excipients in the formulation. The goal of this method is to determine the concentration of protein in our sample. We do this by utilizing a standard of known protein concentration to determine the concentration of our unknown.
The standards are prepared (on the bench) at varying concentrations. 50 uL of each standard and sample are injected. The standard peak areas are used to create a linear standard curve against which the concentration of our sample is derived.
This method seems pretty straightforward and not uncommon. However, I am wondering why we cannot take our standard and inject different volumes (e.g. 10 uL, 30 uL, 50 uL, 70 uL, and 90 uL) and create a standard curve from this. It would save a good amount of time in running the assay.
I approached QC and QA about this, but their response was that no one does it this way because this type of method can't be validated. You can't be confident that all different instruments will inject the correct volume of standard for you. Any thoughts?
In my opinion, using different injection volumes could not get great dynamic range(10-90uL,
It is the mass on column that is relevant and not the injection volume. In addition, you need a 'pure' 99% standard. The method MUST comply with ICH Q2(r1).
In adedition to previous statements:
It would be most accurate to use full loop injection (a fixed volume) and do a serial dilution of standard.
The problem with auto-samplers is that partial and ul pickup can sometimes be inaccurate in a non-quantifiable sense. So if you are injecting just 1 ul of standard, there is potentially a large error introduction in comparison to 10 ul of standard.
Full loop injection is known to be most accurate but does waste a large amount of sample. Having said this, it's more reliable in the sense that any error associated with the injection mode will be systematic in nature.
- Badges
- Science topic
- Similar topics
- Instrumentation Engineering
- Instruments