Chromatography Forum

Hi all, I recently synthesised a new compound and I ve already run the column. However, I found two major peaks on the graph after performing analytical HPLC. They are major peaks and quite close to each other. I am thinking a way to separate them apart. Re-synthesise this compound is the last thing I want to do( I wonder if there is contaninant or it is even the catalyst.) I am not familiar with preparative HPLC, to be exact, I ve never done it before. Can anyone give a clue on how this is done? It will be much appreciated. THanks.

by the way, it is under 254nm.

Its still just chromatography, typically on a larger scale. The only real difference is that you will want to collect the peaks as they elute from the column. Usually one would use a fraction collector for this but there is no reason you cant stand there with the outlet line in your hand and collect into a couple of beakers, you can use your current column and method to do this but it might take a while. If you want to isolate more (in less time), you will need to use a bigger column, (larger diameter, same length, same particle size) and scale up your current seperation (check the web, there is a lot of info how to do this). Of course, you must consider the flow capacity of you current pump if you plan to use it. Just as an example, a typical flow rate for a 19 mm prep column is about 25 mL/min, well out of the range for analytical scale pumps.

To elaborate a bit on AA's response:

The approach depends a bit on how much you need to collect, and how much time versus money you are willing to spend.

If you only need a few micrograms and you don't (or can't invest the $$ to optimize a prep separation), what you need to do is the following:

1. Do a loading study to find out how much you can recover in a single run. Take your existing method and do successive runs doubling the load (mass-on-column) each time until the peak shapes deteriorate to the point where the two peaks just barely touch. That will tell you how much you can process in a single run. As AA suggested, you can switch collector vessels either on a timed basis or manually between the peaks.

2. Repeat until you have collected what you need.

A better approach (but one which requires more up-front time) is to re-optimize your separation, shooting for the best alpha value with k's down around 2. Then follow the two steps given above. The advantage here is that the time per run will likely be shorter, so you can recover more material (nore runs) in a given time.

If the above will be insufficient or time-prohibitive, then you will need to scale up the column. Once you have determined a loading per the above, you can scale to larger columns. You can generally increase the mass-on-column in direct proportion to the column volume. The flow should increase in proportion to the square of the diameter. If you do that, the run time and retention times will be proportional to the column length.

You can get even better throughput by overloading and collection three fractions (A, A+B, B) and then re-processing the middle fraction, but that's beyond the scope of a forum post!

just to think on:

are your peaks really two different substances or are they just artefacts, produced by inproper sample solvent or equilibrium effects of two isomers or so on?