GC temperature programming

[badrul]

Hai everybody,

How do you establish a temperature programming condition in gas chromatography? For example, if I want to separate mix pesticide standard solution containing 20 analytes, what approach should i take in order to separate all 20 pesticides into 20 peaks respectively in the chromatogram?

Currently i just refer the GC conditions from the literatures or other people works. Is there a proper way of establishing temperature programming condition? Let say I want to start from scratch, not just refer to somebody else work and follow its GC conditions from A to Z.

I really love to hear all your opinions about this topic since I'm still new in Chromatography techniques, and sometimes I really amazed by how patient some researchers find out the best GC parameters in order to separate 20, 30, and sometimes more than 100 peaks.

Really love to hear your take in this and appreciate all replies regarding this matter.

Thank you..

[Jumpshooter]

Fortunately, other analysts have already "invented the wheel" for you so all you have to do is steer in the direction you want to go. However, I am surmising from your post that you want to start from scratch and work up a method on your own. Welcome to the process of what I call like to call "Enlightened Trial & Error".

To establish the GC conditions for chromatographing your analytes prep a diluted STD containing 1 of each of the analytes of interest. So, that you can i.d. its retention time (Tr) and build a library of Tr's.

Do some library research to dentify their boiling points/vaporizations and chemical structures to ascertain relative polarity. Select an analytical column (preferably capillary column) of a phase that is about mid-polarity (e.g. Rtx 50), then for the temp program (assuming you do not know where to begin, start with 50C and ramp at 5C per minute until you see a peak. Notate the temperature and time that the peak appeared. Then inject the 2nd single analyte STD and notate the characteristics. So forth and so on for the other 18 analytes.

[Don_Hilton]

I suspect that most of us do not start from scratch - even if we do not look up a method to copy. After you have done this a few times, you have a feel for what goes on in a GC and you have a few favorite operating conditions.

In simple cases, I look for my favorite column (whatever was left in the GC from the last analysis) and if it is reasonable for the separation I use it. (This saves multiple column changes if the instrument is shared between methods.) As long as the solvent is compatible with the column, the injection is below the boilign point of the solvent, Wait a bit for sovent focusing to take place and ramp the temperature quickly up to a temperature at which the analytes may be expected to elute. Depending on the number of analytes and boiling range, the temperature may be held as isothermal beyond that point or ramped at a reasonable rate to get the last componds off in reasonable time. If after a few tweaks of the temperature program, I am having trouble with coelutions, I will most likely get a column with a different stationary phase and try again.

For 20 pesticides, I would look for retention indicies on various stationary phases to see what column will separate the compunds well - although to be honest, I will open a catalog and look for pictures of chromatograms. One will have the pesticides of interst listed - and chances are it will be a specialty phase. If you have the right phase, finding good tempeature conditions simply requries a mixture of the analytes and a few injections to make the peaks spread out nicely. If you have the wrong phase - the separation will not happen. And the way you know how to select the column is by looking at retention information. So it not truly from scratch.


And when you find that paper showing 100 analytes, well separated and quantified. Understand that the researcher probably started with a list of 130 analytes and these are the ones that work. It is not uncommon to have to make injections onto two different columns - or even make two different sample preparations to analyze everythign you want to see.

[lmb]

Hi badrul

Please do not read my reply as a recommendation to start method development from scratch. There are to many unknowns to do so. However ...

I interpret your starting from scratch idea simply as another way to ask: Are there some basics? What are they? Yes, there are some. Not all of them are equally easy to express. Here are two most basic basics:

If the column and the carrier gas, say helium, are already chosen then

1. Use flow rate of 0.8 mL/min per each 0.1 mm of column internal diameter (id). For example, if id = 0.25 mm then flow rate of helium should be 2 mL/min (please check my previous postings for flow rates of other gases)

2. Start with a single-ramp temperature program at the heating rate of 10ºC per hold-up time. Thus, if the hold-up time is 2 min then the heating rate of the first choice (default heating rate) should be 5ºC/min

These default conditions do not guarantee that all 20 peaks would be resolved (no default conditions generally guarantee resolution of all peaks even if their number is relative small). But the conditions do guarantee statistically largest number of resolved peaks in the shortest time.

IMPORTANT! The default heating rate is not 10ºC/min, but 10ºC per hold-up time. This form of expressing the default heating rate makes it equally useful for columns of all dimensions and for all carrier gases.

There are many more basics. Unfortunately, they are a bit more nuanced.

Hope this might be useful

[badrul]

My bad. I really shouldn't said 'from scratch'.

Yes you are right lmb, what I meant to say is what are the basics (or proper ways) of method development. How to start a new method development without a "jumpstart" by referring other articles or journals. Is there any simple guidelines in doing so?

But thanks to all for sharing your valuable experiences in this subject. For a start, they really gave me some insights regarding the matter.

Hope to hear more valuable experience from all of you.

Thanks..

[cody84]

Hahaha I could have used this thread last month! Very good info, thanks everyone. I tried to copy USP method for IPA purity. Everything was set up accordingly and yet the column was completely overloaded by the injection volume and the impurites had (still have) poor resolution.

I adjusted temp, inj. volume and flow rate to the maximum limits from USP and still 4 compounds are not fully resolved and the retentions times are completely different from the possible ones suggested in the USP assay.

[Ron]

Cody84, I wish I could say I am surprised, but that is typical for USP methods. Try the exact column that was used in the method, "identical" columns from different manufacturers have different polymers, and the selectivity may be different.

[cody84]

Thanks Ron, I will do that for the next USP method I try. In the meantime, QA has accepted this method as it still conforms to the allowable limits set by USP for chromotography.