GC-MS Volatile Fatty acids

[Peter Apps]

Hi Zia

In addition to the fibre, maize and organic waste in your samples you also have carbonate, so presumably an alkaline pH. This makes it even more important to acidify the samples, and even more important that you calibrate using the sample matrix. A calibration from a clean solution of the acids will almost certainly not be valid for real samples.

You should try running a real acidifed sample to see what the chromatogram looks like.

Peter

[MSCHemist]

I spent today doing some experiments with using isobutyl chloroformate to make isobutyl esters of the short chain fatty acids and I am very pleased with the results. the aqueous sample in .1n NaHCO3 is mixed with acetonitrile, pyridine, and isobutanol and ISOBCF 20% in chloroform is added, neutralized with 7N NaOH then added again. Finally Sat NaCl/NaHCO3 is added and the organic layer is analyzed containing the fatty acid isobutyl esters. I got nice sharp well resolved peaks for acetic, propionic, butyric, valeric, isovaleric, caproic, and caprylic acid isobutyl esters. I am testing to see if hexane can replace chloroform right now and am looking for a good internal standard (some branched acid). I'll see if I can post some chromatograms later my work computer doesn't like the image hosting sites.

the whole derivitization procedure took 20 minutes.

The only concern is that acetic acid isobutyl ester is just barely past the big pyridine peak (~.5 min) but it is baseline resolved. Email me at sschoenfeld[at]Innovaflavors.com if you want to talk some more about it.

[MSCHemist]

[MSCHemist]

big peak at 7.6 before acetic acid is Pyridine from rxn mix
8.184 Acetic acid [isobutyl ester]
10.695 Propionic acid
12.816 butyric acid
13.216 and 16.019 carbonic acid mono and di isobutyl ester (from neutralization with NaHCO3)
14.349 isovaleric acid
15.286 valeric acid
17.025 caproic acid
20.002 capylic acid

I glanced at the GC/MS before leaving I think hexanes works better becuase I believe acetonitrile partitions into the chloroform layer judging by the size of the organic layer. I saw a larger acetic acid peak using hexanes rather than chloroform for extraction and that is the most polar of the group.

[ziarehmann]

So,

First results. I did the experiments by acidifying the solution. In case of high concentration i have nice peaks. But for low concentration solution i have very bad peaks, peaks are not properly occuring. I am equilibrating the samples for exactly 15 minutes and the take the sample for GC Injection. Reproducability is still an issue and i think it will remain so, but from the triplicate min two results are within 10% range.
Now doing the analysis with with salt and acid addition and i reduce the equilibration temperature to 75°C.

One question i have regarding the quatification. I am using Xcalibur software. In order to develop a procurement method i am using TIC as trace in the processing setup. Should i use base peak? Because when i change the range from TIC to Base Peak in Qual Browser i have much better looking chromatogram.

[Peter Apps]

So,

First results. I did the experiments by acidifying the solution. In case of high concentration i have nice peaks. are peaks bigger with acid than without acid ? If bigger, how much bigger ? But for low concentration solution i have very bad peaks , peaks are not properly occuring what do you mean by "bad" - too small, too wide, too much tailing, interference form other peaks ?. I am equilibrating the samples for exactly 15 minutes and the take the sample for GC Injection what volume are you injecting ?. Reproducability is still an issue and i think it will remain so, but from the triplicate min two results are within 10% range.
Now doing the analysis with with salt and acid addition and i reduce the equilibration temperature to 75°C there is not much point reducing the temperature in small increments, if you want to improve the repeatbaility the syringe and sample have to be at hte same temperature. This means that the sample has to be at a temperature that you can hold in your hand for as long as it takes you to extract the headpsace sample and inject it to the GC.

One question i have regarding the quatification. I am using Xcalibur software. In order to develop a procurement method i am using TIC as trace in the processing setup. Should i use base peak? Because when i change the range from TIC to Base Peak in Qual Browser i have much better looking chromatogramthis is strange, taking only the base MS peak cannot change the shape of the chromatographic peak, but it will reduce the signal for interfering compounds that are eluting close to your acid peaks. Are you still working with standards ?, or have you moved over to samples ?.

I still waiting for yout to tell me what inlet liner you have in the GC (and when you last did inlet maintenance) and how you are injecting the 100 ul samples - especially how long it takes to push the plunger down.

Peter

[ziarehmann]

1. Comparison of peaks:
Acetic Acid Area
Original 3984687
Acidified 12263385
Acidified and salt 14918052
Diluted (1:10) 305344
Dil. Acidified 98527
Dil. Acidified and salt 170721


Butyric Area
Original 4597502
Acidified 29393867
Acidified and salt 34916015
Diluted (1:10) 1046104
Dil. Acidified 185701
Dil. Acidified and salt 906268




2. Peaks for a diluted sample:
The chromatogram with acidified samples and plot type TIC and Base peak:
http://oi60.tinypic.com/2vc9duo.jpg

The chromatogram with acidified and salt samples and plot type TIC and Base peak:
http://oi59.tinypic.com/eiscao.jpg

3. The amount of injection is 100uL, i am planning to increase the volume to 200uL.

4. Once the method is in place i will try to reduce the equilibration temperature.

5. TIC chromatogram v/s Base peak (See attached pics)
Yes i am still working with standards because they are lot easier to handle and i have symmetry.

6. Manual Injection:
http://oi61.tinypic.com/28s4nn.jpg
To push the plunger down takes max 1-2 seconds and then immediately after injectiong i start my analysis.
we have a straight tube liner with max volume of 1 ml. Did the type of liner play an important role?
The maintenance was performed before starting the analysis i.e. at the begining of march.

[ziarehman]

Any suggestions??

[MSCHemist]

I am still developing my isobutyl chloroformate method I ordered 4-methyl valeric as an ITSD
Here is a nice article on using SPME to quantititate SCFA's in milk (C4-C12 (not sure why you can't do C2 and 3 as well))
http://pubs.acs.org/doi/pdf/10.1021/jf010108d

[ziarehmann]

Hi Everybody,

After a long time and lots of analysis i am back with a question. A short summary:

I am trying to analyze volatile fatty acids (acetic, butyric, iso butyric, propionic....) in biogas digester using headspace. At the moment i am using standard solution with different concentrations and an internal standard. I am doing manual injection (400uL), equilibrating at 70°C for 10 minutes and then using a heated needle injecting on GC. So far the repeatibility is not the best but under 10-15%.

Now the thing is the concentration calculations. I mean the area which i got from GC is the concentration of related acid in the gas phase. how can i relate this area with the concentration in liquid phase. From theory i foun area from headspace is propotional to gas phase concentration.
A=Cg=Cl/(Kh+Vg/Vl)
Cg= concentration in gas phase
Cl= concentration in liquid phase
Kh= partition co-effecient
Vg= gas volume (injection volume or headspace volume??)
Vl= liquid volume

but the partition coeffecient of VFA'S is very high so the expression becomes:
A=Cg=Cl/Kh

But how can i calculate the partion coeffecient of acids.

Secondly in order to do calibration what concentrations should i write in Xcalibur.
Hope to hear from your side soon. Have a nice day.
Regards!

[Peter Apps]

Hi Zia

You need to construct a calibration where you plot peak area against the known concentration of the acid in the liquid phase. Do not worry about the concentration of acid in the gas phase. How do you know the concentration of acid in the liquid phase ? - from the amount of acid that you added when you made up the standard.

You find the best straight line fit to your area vs concentration data and use the equation of that line to calculate from the peak area for a sample with an unknown concentration to the concentration that you are trying to measure.

Since you have a complex sample matrix whose composition will affect the partition of acid into the headspace it is essential that you make up your standards in the same matrix as the samples. You will not get accurate results for samples if your standards are made up in water.

Peter

[MSCHemist]

Agreed if you can't get a blank matrix you may need to do standard addition.

Your calibation says you have a certain concentration of analyte in the liquid and the headspace dynamics should cancel out if they are the same in the samples and calibration (IE if you put n ug of analyte into a 22ml headspace vial or whatever) and the Int standard should cancel out variation in vial size, temperature, injection volumn and possibly some or all of the matrix effects on headspace concentration).

It is just like when you inject liquids into the GC. I don't care if the autosampler is very accurate I care whether it is precise and delivers the same volume of calibrations and sample though even if it is not very precise the int standard takes care of that.

[Guigngnm]

Zia, -

have you figured out your headspace calibration by now? Do you still need help? If you tell me what country you are in, I can possibly have an app chemist contact you locally.

Alex.

[MSCHemist]

Anyone still looking at this should see this article. It is illustrated in Bruno Kolb's Static headspace book as well. It aparently works for all carboxylic acids even di's and tri's like citric.

Headspace insitu methylation
http://www.sciencedirect.com/science?_o ... xcerpt.pdf

http://www.umb.no/statisk/nordost/ulf.pdf

aqueous sample + saturated NaHSO4 or 1N H2SO4 +alcohol you wish to esterfy with headspace sampling yeilds esters of the acids that can be done via GC/MS or FID.