sample dilution vs change of injection volume

[dmaa]

if I have a method set up that uses a 10ul injection volume, and I needed to make a 10 x dilution would the better practice be to make a 1:10 dilution into the autosampler vial or could I just change the injection volume for that sample to 1ul. Ive seen people make a curve by having one 100ppm dilution with a method setup for a 10ul injection. Rather than making multiple dilutions they just changed the injection volume to 2, 4, 6, 8 to represent 20, 40, 60, 80 ppm.

Im just wondering what is considered the best practice, and what the possible errors that one could encounter.

[DR]

If only you could try both approaches and run a linear regression to determine whether your pipetting skills are more accurate than your autosampler's...


With all due respect to any analytical prowess you may have acquired throughout your career, my $ is on the autosampler, especially if you plan on doing this more than once or twice. I can beat an autosampler, but not daily.

[Peter Apps]

It depends very much on the design of the auto-injector - and whether or not the volume of the injection affects peak area in more ways than simply the amount of analyte it contains. Since you presumably want to calibrate amount of analyte against peak area for a method with a fixed injection volume, then varying the amount of analyte by dilution should be the better approach.

[itspip]

The best practice is to treat the standards the same as a sample, which would mean the same injection volume. I have made standards both ways but usually use the auto-sampler curve as a way to check the auto-sampler.

That being said, if your industry/regulators are okay with it, and you have the data to fend off any internal complaints, you could run it either way and be fine.

[Multidimensional]

"dmaa" : Following Good Chromatography Fundamentals is very important. We do not make the sample or standard Injection Volume a variable in an established HPLC analysis. The injection VOLUME should be the same for all standards and the actual sample(s). Changes to sample injection volume may create changes in peak area, peak shape or fall outside the proper injection volume range of the actual injector (Details of the specific injector, method and flow path would be needed to determine if any of the proposed volumes are appropriate).

Additionally, any proposed changes would first need to be formally evaluated for repeatability, accuracy and ruggedness via SOP to evaluate changing the volume of your standards (to create the different calibration levels) when analyzing samples. This can be done and qualified if needed, but why? Using established methods and procedures where we do not change the volume used requires no additional work or verification. Doing it differently means having to explain and document the change for everyone that questions it.

Note: In special LC methods which require IN-LINE sample preparation steps as part of the method (e.g. Amino Acid analysis using pre or post-column derivatization), you most certainly can use the A/I to act as a sample prep station (in fact, that is exactly what it is!), adding desired volumes from one vial to another, mixing contents, creating dilutions and preparing solutions for injection, all within the sampler. Programming the A/I to perform these complex tasks is rather easy, but the final injection volume used will still be static, not changing, so it matches up with all samples and standards. - - In your example: You could even prepare your 'concentrated" std in one vial, then use the programmed functions of the A/I to create new vials with the diluted solutions in them. These could be used for the fixed volume injections. However, doing so is often far more complicated and adds more documentation requirements, plus limitations in how much std solution a 2 mL vial can hold to be useful.

[Consumer Products Guy]

The injection VOLUME should be the same for all standards and the actual sample(s).

We also always used the same injection volume for all samples and standards.