Info on HPLC please

[Torsten]

I finally have a handle on GC/MS, but in doing so I realised that maybe I should (also) be looking more closely at HPLC. Only for detection, not preparation!!

Anyway, I kinda know how they work now, but can't work out which detector is the best to use for the applications mentioned before (alkaloids, essential oils, kavalactones). It appears that most tests are run around the 200-400nm, so I presume this is a UV detector. What are the advantages and disadvantages of the other detectors (as applicable to the above compounds).

Any help would be muchly appreciated.

[bark]

I belive you can get combination detectors for HPLC. Perhaps UV and electrochemical would be appropriate.

From memory electrochemical detectors can pick up very small concentrations. Not sure how well it would work for ess. oils but should give excellent results with amines.

Big advantage with going for HPLC is that you can keep the components of the extract for further analysis by MS IR NMR etc. Essential for identifying unknown components. GC/MS of course destroys the sample.

Disadvantage is that if you are not going to do further analysis you will need standards for identification.

There are also less problems with damage to the column as there is a continuous flow of solvent which your sample will be soluble in to varying amounts.

Chromatography is a very interesting (and huge) topic. Even such simple techniques such as a glass column packed with a silica stationary phase with some pressure (or suction) can separate complex mixtures with some fine tuning of the solvent and (sometimes a lot of) patience.

[Torsten]

Originally posted by bark:

Disadvantage is that if you are not going to do further analysis you will need standards for identification.

so this system does not work on a database?? I thought there is a database that can match peaks to what other people have analysed.

Also, does HPLC indicate the ratio of the constituent (ie the actual amount of each)?

There are also less problems with damage to the column as there is a continuous flow of solvent which your sample will be soluble in to varying amounts.

that appears to be the main advantage.

Chromatography is a very interesting (and huge) topic. Even such simple techniques such as a glass column packed with a silica stationary phase with some pressure (or suction) can separate complex mixtures with some fine tuning of the solvent and (sometimes a lot of) patience.

Not so concerned about separating things except for the purpose of analysing. Don't really mind the destructiveness of GC/MS either. Although I won't be using the machine myself much, I would at least like to be able to every now and then. I am no chemist and the only way it is going to be useful is if I can 'stick something in and the machine tells me what it is'. This means it needs to run off a database and has to be able to match the results with stored data. The GC/MS does this. It appears that the HPLC does this too, but I am not sure and would like confirmation.

Thanks for your help!!!

[squiresk]

Also, does HPLC indicate the ratio of the constituent (ie the actual amount of each)?

My limited knowledge says no.

Even if there is a ratio, of one isomer campared to another for example, this doesn't not translate into anything like actual compositions of you sample. Unless of course there is a very limited number of subtances present. ie <3. From what I understand, even if you had a 'pure' mixture

of say DMT and serotonin then it will fragment into hundreds of peaks anyway.

l.k.

[bark]

For a given column and solvent system there should be a fixed retention time (time it takes to hit the detector) for various chemicals which have no doubt been analysed before, however the columns performance may change with time so a standard is usually run or used to 'spike' the extract by adding a small amount of standard to it (extract and extract-standard run separately). Not sure about databases.

I suspect that the analysis on the GC/MS is a result of the MS. This would make analysis independent of the gasses run through the column and the column performance.

Area of the peak is proportional to concentration. A standard of known concentration is usually run to find out a ratio of area-concentration (hope this is accurate as the only machine I have used was a dinosaur).

[bark]

For a given column and solvent system there should be a fixed retention time (time it takes to hit the detector) for various chemicals which have no doubt been analysed before, however the columns performance may change with time so a standard is usually run or used to 'spike' the extract by adding a small amount of standard to it (extract and extract-standard run separately). Not sure about databases.

I suspect that the analysis on the GC/MS is a result of the MS. This would make analysis independent of the gasses run through the column and the column performance.

Area of the peak is proportional to concentration. A standard of known concentration is usually run to find out a ratio of area-concentration (hope this is accurate as the only machine I have used was a dinosaur).

ooops

[This message has been edited by bark (edited 30 July 2002).]

[bark]

Originally posted by Torsten:

Also, does HPLC indicate the ratio of the constituent (ie the actual amount of each)?

The area of peaks are proportional to concentration but not to each other as each component of the extract reacts differently to the detector.

[Torsten]

Thank you for all that. I'm getting a pretty good picture of it now.

Originally posted by bark:

For a given column and solvent system there should be a fixed retention time

so this is essentially what is measured and what identifies the 'arriving' compound? So the detector doesn't actually identify the compound (like a MS does)..... it merely tells me that a compound has arrived and the time it took to get there. This data is then compared to a reference sample or to a database.

[bark]

Not sure of the available detectors. Perhaps some are coupled with IR which would be the tool of my dreams, but no doubt the price tag of my nightmares.

I think databases based on retention time would be unreliable so they may not exist.