Quantification using internal standard... a protocol

[Chris Jones]

Hi Everybody!

It has been a long long time since I last visited the forum. But I have a fairly straight forward question concerning quantifying compounds in extracts.

I work on sandalwood plants; how and why they produce their essential oils. In my most recent experiment I have 6 time points, 4 biological replicates and two experimental conditions - treated and control. After the initiation of the experiment, I randomly select four plants where I grind various tissues up, freeze them and analyse them for all sorts of stuff - one of which is essential oil production.

Now, I have small bags of ground stem and root tissue which are air-dried. I want to know what the concentration of sandalwood oil components in the sample is. In the past I have weighed the sample, extracted in solvent with a known amount of internal standard, and run these on the GC. Here is where my problem comes in.

The determination of a detector response factor relies on being able to get a pure compound (the analyte) and seeing how it's signal relates to it's concentration relative to that of the internal standard. I have two problems - Sandalwood oil has well over 50 compounds in the mixture and standards don't exist for the vast majority of these.

I had one idea - how about taking pure distilled sandalwood oil and preparing an external standard curve. From the most dilute to the most concentrated, I would sum the area of the entire chromatogram and treat this as one analyte. Then I would run the extracted samples, determine the unknown oil concentration as per the standard curve, and work back to the final concentration based on the amount of material I extracted.

An alternative is to use an internal standard with a response factor calculated based on the sum of all components being one analyte 'peak'.

Which do you think is a better representation of the actual concentration of compounds in my tissues?

Cheers,
Chris

[Consumer Products Guy]

I think setting up external standard is more straightforward, with today's automation and autosamplers. But either should work fine, and think your approach is decent.

[Peter Apps]

Assuming that you are using an FID.

Although response factors might differ from compound to compound, they always remain the same for a given compound in different samples. So you can compare between your samples (in terms of palnt A produced X times as much oils as plant B) without standards or calibrations of any sort, provided of course that your extractions and injections are strictly repeatable.

In your protocol, an internal (as opposed to external) standard added before extraction corrects for solvent volume (before and after extraction), differences in how much you concentrate the extract (if you do), and injection volume. For this you do not have to have a standard with the same response factor as the analytes.

If you want to quantify separate components by comparing their peak areas with the peak area from a known quantity of a standard (either internal or external) you need to know the relative response factor. With the kind of compounds that are in essential oils, relative response factors are all very close to unity for compounds that have the same empirical formulae. You may be able to find an internal standard with the the same empirical formula as the compounds that you are interested in. And if the compounds of interest are all similar you can calculate the composition of the oil directly from their peak areas as a % of the total peak area (excluding solvent of course).

Given that you have biological material which is intrinsically varibale, and a multi-step sample preparation I doubt that response factor will be a significant contributor to variability in any case.

Peter