help choose a prep column

[jiang295]

struggle with particle size. 5micron or 10micron. I want to use prep column to separate very complex sample(>500peaks). I do not know which peak is my analyte yet. and i already do my best for sample prep.
for particle size, i know if you increase your sample size to some point, plate count will very less rely on particle size. I just do not know how much sample I inject that can get to this point.
any rule of thumb?

[tom jupille]

any rule of thumb?

Do a loading study on your analytical scale column. Then scale your loading up in proportion to the square of the column diameter (assuming column length is constant).

[jiang295]

yes, so you think the best way to do clear my confusion would be two analytical columns, one with 5micron and one with 10micron with the same packing material.
then see at which point, these two column give very similar resolution.
I know it is the most reliable way. do you have any theoretical prediction?

any rule of thumb?

Do a loading study on your analytical scale column. Then scale your loading up in proportion to the square of the column diameter (assuming column length is constant).

[Uwe Neue]

Not really... A 5 micron column will always give you more resolution than a 10 micron column to start with, but, when you overload the column, the resolution is determined by the load, and not any more by the column performance.

[jiang295]

yes, so my problem is that how can I know at which point(how much is the sample load) the resolution will be determined by sample load rather than intrinsic plate count?

Not really... A 5 micron column will always give you more resolution than a 10 micron column to start with, but, when you overload the column, the resolution is determined by the load, and not any more by the column performance.

[Uwe Neue]

You make me smile, because you continue to give me opportunities to tell you about my book (HPLC Columns):
The information that you are after can be found in Chapter 15, Preparative Chromatography, pages 276-295. A good paper on the real-life loadability of real columns with real samples can be found in Journal of Chromatography, Volume 1030, Issues 1-2, 19 March 2004, Pages 123-134; title: Differences in preparative loadability between the charged and uncharged forms of ionizable compounds; authors: Uwe Dieter Neue, Thomas E. Wheat, Jeffrey R. Mazzeo, Cecilia B. Mazza, Jie Y. Cavanaugh, Fang Xia, Diane M. Diehl

Since you are running gradients, you can load up to about 5 mg per mL of packed bed, if all your compounds are not charged. If some or most of your compounds are charged, expect to load 20 times less. Also, expect to load much less if you are doing isocratic chromatography, in both cases. The statement includes a slight overload.

Going back to your continued question about particle size: your peaks on the larger particle size are wider to start. You will have overloaded your higher performing column to a much higher degree before you end up with the same inferior plate count as the column with the larger particle size. Look at the pictures in my book to understand why this is the case. The lesson is that you will get better results at equal loss of resolution from the column with the smaller particle size. Only when you completely overload the column, will it not make any difference any more how good the column was to start with.

[jiang295]

yes, that is really good book and nice paper!

when you said completely overload column , did you mean saturated column?

You make me smile, because you continue to give me opportunities to tell you about my book (HPLC Columns):
The information that you are after can be found in Chapter 15, Preparative Chromatography, pages 276-295. A good paper on the real-life loadability of real columns with real samples can be found in Journal of Chromatography, Volume 1030, Issues 1-2, 19 March 2004, Pages 123-134; title: Differences in preparative loadability between the charged and uncharged forms of ionizable compounds; authors: Uwe Dieter Neue, Thomas E. Wheat, Jeffrey R. Mazzeo, Cecilia B. Mazza, Jie Y. Cavanaugh, Fang Xia, Diane M. Diehl

Since you are running gradients, you can load up to about 5 mg per mL of packed bed, if all your compounds are not charged. If some or most of your compounds are charged, expect to load 20 times less. Also, expect to load much less if you are doing isocratic chromatography, in both cases. The statement includes a slight overload.

Going back to your continued question about particle size: your peaks on the larger particle size are wider to start. You will have overloaded your higher performing column to a much higher degree before you end up with the same inferior plate count as the column with the larger particle size. Look at the pictures in my book to understand why this is the case. The lesson is that you will get better results at equal loss of resolution from the column with the smaller particle size. Only when you completely overload the column, will it not make any difference any more how good the column was to start with.

[Uwe Neue]

No. Saturation is different. Saturation means that you load the column with a continuous stream of analyte(s) until you see breakthrough. I was thinking about "normal" gradient chromatography, with an "appreciable" overload of the column, until your peaks get maybe 2 to 3 times wider than they should be with small mass load.