Response factor for fatty acid esters

[leo1352]

Hi,

I am kind of stuck in some calculations. I have analyzed fatty acids of butyric, myristic and palmitic esters, (short to long chain) with GC-FID.
I wonder how to take into account the theoratical responce factor for each of above compounds? e.g. in papers theoratical factors mentioned for butyric is 1.4, myristic 1.04 and palmitic 1.09 . I wonder by multiplying these factors by the area of their corresponding fatty acids I have got from GC-FID I would get to the corrected area and hence concentration.

Thanks

[Don_Hilton]

Maybe

This depends on the quality of the measurement of the response factors and the condition of your chromatographic system.

Assuming that the paper is exactly right on the ratio of ions counted in the FID on a specific instrument, you then need to be sure that you have accounted for all sources of discrimination in your instrument. This includes factors in the inlet, column and even the detetor.

For perspective: some of us will run calibration checks every day -- even after cleaining the inlet and restoring it to the condition that it was on the day before. Some of us will not even bother with a calibration check, but simply run an new curve at the start of the day. This to be sure that we have good ratios of response for compunds of interst on the same instrument in the same condition and with the very same instrumental conditions -- as close as we can possibly get them. Why? Because things change.

[chromatographer1]

First,

The responses you have quoted don't seem to be extremely accurate, so in any case I would not hold too much credence to their application in any calculations. Given that statement.....

Response factors can be based on mass, or they can based on mole. You have to discover the basis of the relation of any number given.

For example, in a solution of a volume of 1L, 250 grams of myristic acid methyl ester and 250 grams of palmitic acid methyl ester per liter may give approximately the same response for each ester, but based on a mole response the number given will be different by about 10%.

Multiplying areas by a response factor based on mass will give you the relative (not true) mass, but not concentration, unless you have an internal standard of known response to each of the fatty acids, and of known concentration in your solution.

Homework can be a bear, no?

Rod

[Peter Apps]

In addition to al that Rod and Don have said .....

Nearly all of response factors reported in the literature are derived from injections onto gas chromatographs, and so they include the inlet discrimination and losses on column as well as the FID response. They are specific to the particular instrument and analytical conditions under which they were measured, which means that they are not likely to apply to your analysis. If you want to use response factors you need to measure them for your particular conditions by running standards of known composition. If you are going to do that you may as well just do a conventional multi-point calibration.

Peter

[leo1352]

Thanks people for all the good clarification. The reason why I am thinking about using response factor for my comounds is that I have esterified thriglyceride standards of Butyric, Myristic and Palmitic and I am wanting to calculate the esterification yield by measuring the concentration of each of those fatty acid esters. Having to say that I have made a five point calibration with dillutions of C23:0 -Me external standard and have internal standard of C23:0 spiked to my triglycerides before estrerification as well. The thing is, except butyric acid ester, I am getting about 109% yield for palmitic and myristic ester. I am beggining to think that some miss interpretation has been made by FID signals . could any of you help me through this?

thanks
Ed

[Peter Apps]

There is a long list of potential reasons for a result being 9% higher than expected.

How accurate is the expected yield - what is it based on ?.

How accurate and repeatable is you addition of internal standard and your dilutions of standards and samples ?

How repeatable are the peak areas if you make multiple injections from one vial ?

How pure is you C23 Me standard ?

How linear is your five point calibration (and I am assuming that it spans the concenctrations that you get from your samples) ?

etc

Peter

[leo1352]

There is a long list of potential reasons for a result being 9% higher than expected.

How accurate is the expected yield - what is it based on ?.

How accurate and repeatable is you addition of internal standard and your dilutions of standards and samples ?

How repeatable are the peak areas if you make multiple injections from one vial ?

How pure is you C23 Me standard ?

How linear is your five point calibration (and I am assuming that it spans the concenctrations that you get from your samples) ?

etc

Peter

The yield is based on the equation that one mole of triglyceride will give three moles of esters. I add and weighted a suitable concentration of C23:0-Me which is form of powder to one vial and dilluted that to 5 diferent dillutions for calibration. To the other five kimax tubes I added my C23:0-Me internal standards and transfered my triglyceride to it and esterified the triglycerides. after esterification I added 2ml of hexane (with precision syringe) and vortexed and extracted them with this organic solvent.

[chromatographer1]

9% is about the amount of weight added to the free acid by the methanol to form the ester.

Why don't you show us your equation?

best wishes,

Rod

[Peter Apps]

Can you explain how you are using the same substance (C23 Methyl ester) as both external and internal standard ? Usually an internal standard is added to the sample and standards at the same concentration, and the calibration is made by adding different concentrations of the target analytes to the standards, and then plotting analyte area/IS area vs analyte concentration.

Peter

[leo1352]

Can you explain how you are using the same substance (C23 Methyl ester) as both external and internal standard ? Usually an internal standard is added to the sample and standards at the same concentration, and the calibration is made by adding different concentrations of the target analytes to the standards, and then plotting analyte area/IS area vs analyte concentration.

Peter

Well, I did use C23-Me dilutions for making the calibration curve. That allows me to measure the amount of spiked C23-Me in my samples and thereby calculating the recovery of spiked C23-Me during transesterification.

[leo1352]

9% is about the amount of weight added to the free acid by the methanol to form the ester.

Why don't you show us your equation?

best wishes,

Rod

Hi there, thanks for asking. the equation is: Yield= (amount of ester/ 3*amount of triglycerides) * 100

Cheers
Ed

[Don_Hilton]

And what kind of quantity? Stoichiometry suggests moles...

[Peter Apps]

Can you explain how you are using the same substance (C23 Methyl ester) as both external and internal standard ? Usually an internal standard is added to the sample and standards at the same concentration, and the calibration is made by adding different concentrations of the target analytes to the standards, and then plotting analyte area/IS area vs analyte concentration.

Peter

Well, I did use C23-Me dilutions for making the calibration curve. That allows me to measure the amount of spiked C23-Me in my samples and thereby calculating the recovery of spiked C23-Me during transesterification.

And how do you get from the recovery of the C23-Me to the quantity of the other esters ?

NB - you will find that this gets to a useful conclusion much quicker if you provide us with a full description of what you aare doing rather than us having to wring it out of you one line at a time .

Peter

[leo1352]

Thanks for your help, here are the data you might help me with:
C23:0 dillutions has been injected twice for each dillution

Sample Name Ret.Time Area Height Amount (without correction factor)
min pA*min pA mg/ml Plates(EP) correction factor
C 23:0 C 23:0 C 23:0 C 23:0 C 23:0

C 23:0 2.03mg 34.352 12.5684 126.824 2.3218 780842 0.87
C 23:0 2.03mg 34.359 12.5901 126.4878 2.3258 772715 0.87
C 23:0 4.06mg 34.387 22.2237 197.9915 4.1055 601788 0.99
C 23:0 4.06mg 34.4 22.308 201.5622 4.121 613968 0.99
C 23:0 6.06mg 34.44 30.2191 250.1527 5.5825 516133 1.09
C 23:0 6.06mg 34.422 30.4183 251.9165 5.6193 512550 1.08
C 23:0 7.98mg 34.489 43.1355 318.3762 7.9686 404315 1.00
C 23:0 7.98mg 34.473 43.6013 322.7215 8.0546 408176 0.99
C 23:0 10.05mg 34.503 55.643 375.1049 10.2791 330158 0.98
C 23:0 10.05mg 34.503 55.0123 378.5389 10.1626 350521 0.99

correctionfactor=Ratio of the added ammount/determined ammount

FA = Fatty Acids
(internal standard) C23:0 in each vial(mg/2ml) pA*min
FA_vial_1 2.01 7.1385
FA_vial_2 2.13 7.1738
FA_vial_3 2.12 7.0168
FA_vial_4 2.09 6.9852
FA_vial_5 2.09 6.9852




Area pA*min(from GC)
vial_1 _palmitic 36.6696
vial_2 _palmitic 41.4675
vial_3 _palmitic 41.3238
vial_4 _palmitic 39.0559
vial_5 _palmitic 40.5046


Tripalmitin weight added to each vial

Tripalmitin mg/ml
Vial_1 11.3
Vial_2 11.5
Vial_3 11.3
Vial_4 11.4
Vial_5 11.1

1 mol of Triglyceride/1 mol of ester= ammount of Triglyceride added/ x ammount of ester produced

[Peter Apps]

One last try at being absolutely clear:

We need a description of what you do, not the raw results that come out at the end of it. I have plenty of raw data of my own to analyse.

Peter