How to calculate Partiction coefficent ( K ) ?

[mlivio]

Hi to all

this is my first time on this forum but it's quite a lot of time since i started reading and follow you.
I'm an expert HPLC analyst and method development.

The company where i work have recently purchased a new GC- FID, Thermo Scientific TRACE 1300, and we are going to implement and validate several methods specific for the pharmaceutical industry.

One of the analysis that we have to implement is the analysis for residual solvent in static headspace, both on raw and finished product.

Since is the first time that i approach HeadSpace methods i have performed a little research on the theoretical basis of this technique.

As i know, the critical point of it is the liquid-gas equilibrium that it forms in the vial.
The area of the peaks produced is proportional to :
Area = Concentration of the analyte in the gas phase = ( initial concentration ) / ( K + Beta )
where, K is the partition coefficient and Beta is the Ratio of volume in the vial.

Now my problem is how to calculate the partition coefficient and once i have it, i plan to use it to create all my samples and standards, right ?

Anyone can explain me how to do so ?

Or can any one help me out and explain to me how to perform such analysis ?

I already have tried out some simple theory such as Antoine's law, Raoult law, Heanry law and so on, i think that the calculation involves thermodinamics but i can't find any help on how to do so.

Thanks in advance for any help.

[Peter Apps]

Welcome to the forum.

You are making it more complicated than it needs to be. You do not need to use the partition co-efficient in headspace analysis. Make up your standards in a matrix that is a close as possible the same as your samples, run them and plot peak area vs concentration to get the calibration line, just as you would with HPLC. The run samples under the same conditions and use the calibration equation to calculate from peak area to concentration.

Peter

[mlivio]

Welcome to the forum.

You are making it more complicated than it needs to be. You do not need to use the partition co-efficient in headspace analysis. Make up your standards in a matrix that is a close as possible the same as your samples, run them and plot peak area vs concentration to get the calibration line, just as you would with HPLC. The run samples under the same conditions and use the calibration equation to calculate from peak area to concentration.

Peter

Dear Peter,

thanks for the suggestion,
in fact in the practical use i already used this approach, by doing as you said.

In the validation process i used a five level calibration curve and we are using that.

I just want to know if there's a way to perform that calculation for evaluate if the curve that i obtain experimentally is fitting the theoretical one.

thanks again

Marco

[GOM]

As Peter said - you are over thinking this.

I appreciate that you may wish to understand it from an academic approach but in practise in a sealed static headspace it is highly dependant on the matrix, every component of which will have an effect on the partition, coupled with the temperature, which is why we would normally use a single or multiple standard addition procedure of the analyte in question in analysis. edit If you are using SPME there is partition into the fibre to consider.

You may be able to do it assuming a 1 or 2 component system under normal temperature and pressure but that is beyond my "simple bear" maths." I am a bear of very little brain, and long words bother me"

I can predict the winner of any horse race based on assuming that all horses are perfectly spherical and travelling in a vacuum


Regards

Ralph

[mlivio]

As Peter said - you are over thinking this.

I appreciate that you may wish to understand it from an academic approach but in practise in a sealed static headspace it is highly dependant on the matrix, every component of which will have an effect on the partition, coupled with the temperature, which is why we would normally use a single or multiple standard addition procedure of the analyte in question in analysis. edit If you are using SPME there is partition into the fibre to consider.

You may be able to do it assuming a 1 or 2 component system under normal temperature and pressure but that is beyond my "simple bear" maths." I am a bear of very little brain, and long words bother me"

I can predict the winner of any horse race based on assuming that all horses are perfectly spherical and travelling in a vacuum


Regards

Ralph

Thanks a lot Ralph.

I'll assume that the partition coefficence are the same, working with the same matrix.

Doing so i think i'll be able to avoid all the thermodinamic calculus

[GOM]

Hi

As a practical example

You would disperse or dissolve say 1g of your sample in 5mls of hexane and measure the area of your target analyte

Then you would add 5mls of hexane containing different levels of your target analyte to 1g of your sample and measure the areas, This is multiple standard addition.

From back extrapolation of the graph of the resultant areas you can determine the original concentration.

If you know the levels and only need a pass level you may only need to do a single dosed addition

Regards

Ralph

[mlivio]

you have described the method that i've used since the beginning of the analysis.

in 4 vials i've placed:
- 250 mg of my sample powder + 5 ml of solvent
- 250 mg of my sample powder + 5 ml of standard at level 1 ( 50% of target concentration )
- 250 mg of my sample + 5 ml of standard at level 2 ( 100 % of target concentration )
- 250 mg of my sample + 5 ml of standard at level 3 ( 150 % of target concentration )

by doing that i've created a calibration curve and by using y=mx+q and inverse formulas i've obtained the concentration of the analyte in my sample.

comparing what you have described with my method I'd say all in all ,i've not reasoned that bad: D

thanks again