Chromatography Forum

I do analytical work for RD department.
Now this is a molecular(MW. >500) containing one COOH group, two OH groups and one -OTBS group. But there is no absorbance of it by HPLC above 200nm and peak shape is strange on Rtx-1 and HP-5 column by GC. could you give me some suggestion about this molecular by GC, HPLC or other methods?
Thanks.

I am not sure what question you want answered. If it is how to determine the structure, you have not given much information. The information that resulted in the identfication of carboxyl, hydroxyl and -OTBS groups present would, most likely give you more information.

For any of us to answer your question. Please expand on what it is you want to kknow. For us to give an answer, please tell more about the sample, and the information you already have.

The strange peak shape - could be an number of things, including decomposition of the compund on the column. That happens with smaller molecules as well.

We have know the structure of this molecular and have been checked by HNMR. We want to develop a GC method for purity.
Thanks.

I think that GC is the wrong route to follow:

Very large molecules need very (relatively) high temperatures for elution unless you use very short, very thin film columns. You then have limited separating power. The high temperatures are likely to cause thermal breakdown of both the target analyte and anything else that is in the sample, leading to erroneous results. Unless you have a specialized high temperature GC setup I am surprised that you even see a peak.

Acids and hydroxyl groups do not GC well, a total of three OHs on a molecule calls for derivatization, which will increase the molecular weight even more, and introduce a lot of extraneous impurities that were not part of the original sample.

The strange peak shape that you see (if indeed the peak that you are looking at is the compound that you want to analyse) is probably due to a combination of thermal decomposition and adsorption. Just for interest, what were the GC conditions ?

Peter

The bad peak shape is probably a function of the polarity mismatch between column stationary phase and analyte. You have three acidic groups on the molecule and are running on a non-polar column. I am assuming that your peak is very broad, and rises gradually to a maximum then drops rapidly to baseline. That is the classic peak shape for an organic acid on a non-polar column, and the peak also tends to get broader as well as taller with increasing concentration with a polarity mismatch.

I agree with Peter, GC would not be my first choice of instrument for this assay. LC would be better, but you probably need to go do a different detector, maybe an ELSD.

I'd dissolve in DMF or pyridine, and mix with BSTFA reagent to derivatize the carboxylic acid and hydroxy groups and try by GC. Yes, this will additional mass, but greatly decrease the polarity and increase molecular stability. For molecules this large, I'd use a high-temperature column designed for large molecules, like a metal capillary with thin non-polar phase (e.g. Restek, Quadrex, or others). I don't know what OTBS stands for.

I'm very glad to receice your reply. Thank you very much for your suggestion!
My GC condition:
column: HP-5 30m * 0.32mm 0.25um
injection temp.: 280C dectector temp.: 300C
oven temp.: Initial Temp.(C) Hold time(min) Increasing rate(C/min)
50 0 20
280 20
column flow: 7ml/min split ratio: 10:1

About HPLC, we have only VWD and DAD detector, no ELSD.

-OTBS: (CH3)3C(CH3)2Si-O-

So next: (1) I want to try to derivatize;
(2) try FFAP column with thin film
Do you think these are reasonable?

I also want to ask a question about solvent for derivatization for GC.
Pyridine is a common solvent. About other solvent: CH3CN, EA, Methanol and so on?

Pyridine is a common solvent. About other solvent: CH3CN, EA, Methanol and so on?

I've used mostly DMF and pyridine for 30 years, but also have used ethyl acetate a few times. Methanol or any alcohol definitely will NOT work as solvent, it will react with the derivatization chemical preferentially and cause a lot of heat

I've used mostly DMF and pyridine for 30 years, but also have used ethyl acetate a few times. Methanol or any alcohol definitely will NOT work as solvent, it will react with the derivatization chemical preferentially and cause a lot of heat[/quote]

Thank you very much!

Given the GC conditions I would be surprised if the peak that you are seeing is the target compound, even with the very high flow rate. What do you see if you hold the column at max temperature for an hour ?

Peter

For the big stuff, on the specialty high temperature column described above (not HP-5), we would routinely use temperatures 370C or 400C.

Given the GC conditions I would be surprised if the peak that you are seeing is the target compound, even with the very high flow rate. What do you see if you hold the column at max temperature for an hour ?

Peter

I'm also not sure. I can't do GC-MS with this high column flow rate. But there is no peak after 20min. Total run time of the method above is 41.5min.

Thanks Consumer Products Guy, but we don't have this high-temperature metal capillary column currently.

If you have a shorter narrower column, you could run a similar chromatogram with a sufficiently low flow rate to use mass spectral detection. A thinner stationary phase will retain the compund less, allowing it to elute a bit faster - without the extreme linear velocity. Using a 0.25 mm ID colulmn or a 0.18 mm id column wih the same phase ratio with a thinner stationary phase. I have chromatographed C35 to C40 hydrocarbons through a 10 m x 0.18 mm id x 0.2 micron RTx-5ms column at a normal flow rates and at the column maximum (350 degrees).

Having said that, I suspect that the compound eluteing from a 30 m DB-5 column at 280 degrees, will prove to be a degredation product. And, if you need to detect and quatify impurities that are simiar in size to the desired compund, they may not survive the chromatograpic conditions. Even if you continue to persue the GC analysis, it might be a good idea to explore another technique as well - even if it involves asking for another LC detector.

If I read your GC conditions correctly you are only going to a column temperature of 280C, while the column has a maximum temperture of 340C. Why don't you go to a higher final temperature and see if you get out any more peaks at the higher temp.

In regard to using a FFAP column, with the low maximum operating temperature of this phase it is very unlikely to work for compounds this heavy.

What will it cost to pursue this approach with the development time and expense of finding the proper derivitive, column, and method, compared to purchasing an LC detector and using LC which is more suited to this analysis?