Chromatography Forum

Hello,
How to give the acceptance criteria for no. of Theroritical Plates in GC??

Thanks,
Srinivas

First chose a peak that will be your reference, a standard for comparison.

Determine the retention time of an unretained peak.

Measure the half height peak width of your reference peak.

Report the number of plates at a fixed temperature for the reference peak with xx ng of reference on the column.

best wishes,

Rod

Dear Rod,

Question is not how to calculate but how to set an acceptance criteria. Actually in LC we give an acceptance criteria for No of theoritical plates Not Less than 2000. But in GC the no of Theoritical plates are arround 5,000 to 10,000. Is it anything to do with length of GC column (30m)???

Thanks,
Srinivas

Yes it does. The number of plates generally increases at a factor of 1.4 per 2X in length.

Different columns of the same length and ID and film and its thickness may have different actual number of plates.

I suggest you estimate how many plates will be needed to perform the separation you require and set the specification as a minimum number of plates using the specified type of column. As I stated before, also set the method of calculating the number of plates.

You might be surprised to learn that columns also may vary in length.

Don't get too specific and you and the people that follow you will be happier. Be conservative, not boastful.

best wishes,

Rod

Rod:

The number of plates generally increases at a factor of 1.4 per 2X in length

Hi Rod,

Either I misunderstood you or you made a mistake. Plate number (N) is proportional to a column length (L). In a capillary column with optimal or near optimal flow, N can be estimated as N = L/d where d is a column internal diameter. Thus, N of a 30m-long column having d=0.25mm can be close to 120,000.

lmb

We are discussing two different things, theoretical and actual.

There is the utopia world.

and there is the real world.

Just because you have a column of 30m length doesn't mean you will achieve an ACTUAL number of plates (resolving power). Every 30 meter column has the same number of plates (plate number) but

Generally, when you put two columns in series of the same type you get not twice the resolving power for a defined separation, but the square root of 2 increase. Likewise when you cut a column in half you don't get half the number of actual plates, but 1/1.414 the number of plates.

When defining a column for an application you have to set the number of plates needed to achieve the separation you require of 'a' from 'b'.

Otherwise ANY column of length L with a diameter d will work, right? (not in my universe!)

Rod

Rod,

I was writing ONLY about the REAL WORLD PLATE NUMBER as it is defined in IUPAC and in all know to me textbooks on chromatography. I stand by what I wrote.

If you like to switch the subject to "resolving power" or to something else, I might participate in that discussion too if it is useful and respectful. For now, I can only say that I do not understand what the “resolving power” in your posting exactly means, but I agree that meaningfully defined resolving power could be proportional to square root of a column length (L) and plate number (N) (doubling L, doubles N and increases the separation power by 1.4). I should add, however, that so defined separation power would not be the same as N. It would be different from the separation power in the phrase “number of plates (resolving power)” in your last posting.

lmb