Column Size: recommended me

[usuchrom]

Hello, I need your help about which column size I should to use to purify 500mg of an organic compund that has farmacology interest. And to inject 1 grame?

Thanks in advanced

[tom jupille]

No simple answer to that one. It depends on what the impurities are and how close to the product they elute. And do you want to do one injection or multiple injections (the latter will often give you better throughput).

One approach is to begin with an analytical-scale column and get your conditions worked out (best alpha value with low k'), then do a loading study (increasing the mass-on-column by successive multiples of 2 or 3) until the product peak just touches the closest impurity. You can then choose a larger column, scaling the mass injected to the column volume.

[usuchrom]

My suplier offer to me a 200(L)mmx50(id)mm in the case of the 1grame (in 1 injection). I could be in a lab where they use a 180x50mm for a grame (in 1 injection). But I want to know your experience. But above all, I need to know the size of a column to purify amounts about 500mg per injection.
Thanks

[tom jupille]

I'll repeat what I said earlier:

No simple answer to that one. It depends on what the impurities are and how close to the product they elute.

[Gerhard Kratz]

In addition to Tom`s posts you need at first the results on an analytical column. Increase the sample concentration and/or sample volume until your main compound is no longer baseline separated. Than I would do the separation on a 20mmID semiprep column with larger particles. If the separation is optimized on that dimension you can do the calculation for up scaling. It all depends on the loading capacity of the packing material due to your compound.
The most straightforward way to scale up a column is to maintain the same column length and to increase the cross-sectional area to maintain a constant ratio of unknown to column volume.
To recreate identical operating conditions on the scaled column as the unscaled one, the linear flow rate must be kept constant.
Good luck

[Uwe Neue]

As Tom said, there is no easy answer. First of all, do experiments on an analytical column to see how much you can load. Then you know what column size you will need to purify 500 mg.

Tom mentioned that part of the loadability depends on how closely your impurities are eluting. Another fact is that the loadability depends by a factor of 10 up to 50 on the charge of the compound. If it is uncharged, loadability increases a lot, and to load a compound under uncharged conditions can save you a ton of money. If this applies, look for suitable packings at the Waters website.

Of course, all of this does apply to small molecules such as pharmaceuticals and is of less value for peptides proteins etc.

[usuchrom]

Thanks for answering me.
Tom, I know that it depends of each sample.
Gerhard and Uwe, After 3 years using and standard way of purification of dirty and closer impurities, I know that I have to make some probes before to purify. I use an analytical column (L 100x4.6 id) to study what is the best method to separate each product, and a preparative column (100x21.2) to purify 50mg per injection.
But now, I need to find a new size column to increase de amount of product+impurities, to purify.
I wish any of you have the answer to my question.
Thanks

[tom jupille]

We *gave* you the answer: once you have worked out a separation and found the maximum load on an analytical column, the loading scales approximately to the column volume (i.e., a function of column length and column diameter squared).

twice the diameter = 4x the load
twice the length = 2x the load.

So, doubling the length *and* the diameter should give you 8x the load. If the maximum load on a 100 x 21.2 mm column is 50 mg, then a 300 x 38 mm column should give you around 500 mg (as should a 200 x 47 mm).

[carls]

250x50mm is a standard dimension. Based on the information you provided and the discussion so far this dimension should give you the needed capacity. As I'm sure you are aware these columns are not cheap so you may want to double check the calculations. Also, a 50mm column will require ~4.7x the flow rate to maintain the same linear velocity as the 21.2mm column.

[Chris0000]

Hi

If you can isolate just 50mg on a 20mm I.D. column, a 50mm I.D. column will probably not be able to isolate 500mg under the same conditions. On the other hand side, for larger column diameters, you will need a matching prep system, espacially the pumpmhas to deliver higher flowrates.

Best regards Chris

[usuchrom]

Tom: thank you very much for your information. I will have in mind.
Carls: thank you too for the information.
Chris0000: I have the equipment necessary to operate at high flow rates. Up to 200mlxmin.
I leave you some information that I found on the website of Phenomenex and I think that may be useful for all: The Scaling Factor
When only the column ID is changing:
SF= (d2/d1)2
When both ID and length are changing:
SF= (d2/d1)2*(L2/L1)
Where
SF= Scaling factor
d1= diameter of starting column L1= length of starting column
d2= diameter you are scaling to L2= length you are scaling to

I hope you find it useful.

[danko]

Actually I don’t see the need for employing a longer column in up-scaling context, unless (very importantly) the particle size of the new column is enlarged.
Especially useless is to talk about column volume without specifying the deciding factor for the volume increase (i.e. the length, or the diameter, or both?).

Best Regards

[mbreslav]

My pragmatic approach would be to use the column that was offered to you. But try to maximize loadability by using both Uwe's innovations: at-column dilution and charge neutralization. Then overload the column and take an advantage of the displacement effect. I found that investments in detailed preliminary analytical investigations usually do not pay off, unless you intend to develop a large-scale production method.

[ATL Chrom]

I would simply recommend an 8g or 12g SuperFlash column. This is a prepacked flash cartridge with 40-75um Premium Silica. Call me if you would like to try free samples.