Hello,
PLease help me to make a right choice.

I have seen someone perfomed HPLC method validation for linearity as outline below 1 and 2.

HPLC method condition:
Injection volume = 5microL of 0.1mg/mL

1. Prepare a stock concentration of 0.2mg/mL and injected as 5, 2.5, and 1 microL (1, 0.5, and 0.2microgram)

2. Prepare a solution of 0.2, 0.1, and 0.04mg/mL and injected 5microL (1, 0.5, and 0.2microgram) for all solutions as stated in the method.

My question is: Since both have the same answer, which one should I pick?

Thanks

I prefer numer 2, because there is a fault of your injector, and when you do some sample preparation, you can weigh out al little bit more or less, but you would never inject different volumes.

A second reason: Inj vol can have an influence on peakshape, so you can't validate you method this way.

And when you stock solution is wrong, the whole linearity is wrong. But in the second manner, you make different solutions, so if one is wrong, you can actually see it and figure it out how this is possible.

Use choice #2. For pharmaceutical validation, usually five points are required to demonstrate linearity, not three.

I think it would probably be a good idea to make sure injector linearity is part of your instrument qualification if you are going to use 1.

I would also validate the method like I was going to use it. It is obviously faster to construct a calibration curve using the first method. If your test method is going to be constructed using different injection volumes then validate it that way.

Check your data system(s), some make using the first method so cumbersome that it is not worth it. You may use excel or a statistical package in validation but the method users will have to use their data system.

I disagree with A.Nonymous that you "can't" validate your method this way. There is a little more to check, but if your standard has multiple components and/or is very time consuming to prepare validating and calibrating using mass on column with variable injection volume can result in a more user friendly method.

You're right Tom, in some cases its possible, but I wouldn't do it. Even if injector linearity is part of our instrument qualification.

It is also important to consider the way to prepare your different solutions. If you do a dilution of your stock solution you will keep the error of weighing in your several dilutions. So it would be better to prepare your 5 concentrations independently. Depending on the analyst and his/her experience it will somtimes be more accurate to let the autosampler do your dilutions. Many modern autosamplers can do that easily.

There ia third possibility:
Prepare one stock solution and dilute to the actual levels.
The First choice would be, as Tom mentioned to stay as close to the method as possible. When the method doesn't use different level then you have the choice between the three ways:

1.) Injecting different volumes of one stock solution can caus problems with peakshape etc. and autosampler linearity

2.) preparing different solutions at different levels might rise problems with weighting/sample prep. errors and needs more standard (only sometimes an issue)

3.) preparing a stock solution and diluting to levels: success is based on the ability to carry out dilutions accuratly.

If you use the autosampler, the autosample dilution method should be checked (or maybe even validated).
If I have the choice I prefer 3.)

Thank You to ALL of You for your time to address my question. You guys have helped me to make a better choice.
Regards,.