The HPLC systems in house are normally used for analytical purposes which means a small diameter columns like 4.6mm are connected. If I wanted to use the system for prepare some pure samples, what is max. column size which can be compatiable with current system? What is max. injection volume I could use?

Thank you very much.

Depends on the flow rate capability of your system. For equal linear velocity, flow rate increases with the square of the column diameter. For example, if you go from a 4.6-mm id to 7.8-mm id (a common semi-prep size), your flow rate should be increased by a factor of (7.8/4.6)^2 = 2.9.

Watch the backpressure on your flow cell, too. I would suggest measuring the backpressure with no column or detector connected, then measure it with your detector connected to determine whether you have any limitation there. Leave yourself a good cushion, too, as you really don't want to break your flow cell. With regard to injection volume, that will typically increase with the same relationship to the column diameter as does the flow rate, though you do have more flexibility there if you're very careful. If your sample dissolved in the mobile phase, if it's properly filtered, and if you have suitable delay at the start of your gradient (if any), you can inject through the pump as long as your sample isn't shear sensitive. You must be REALLY careful when you do this, because if your sample precipitates once on the LC, you'll have bought yourself a pump rebuild at best. I've used a Waters 600E with 100uL heads running full tilt (20 mL/min) through a 20mm (dia) column, injecting 20mL through the pump. The mobile phase was MeCN based and the regular analytical flow cell on my 996 was used with no ill-effects. This was a one-off project, though. I wouldn't do this every day with my analytical equipment. If I was to do routine scale-up, I'd buy an appropriate system for that.

Hope this helps.

Chris