Consider going back in history and read the techniques used for filling glass capillary columns in gas chromatography around the 60-70ties. Maybe you can even find an old GC oven to mount the capillary in.

Succes

Alessandro,

The comment of Jos Wielders will guide you to several options.

You can place the column in a thermostate (or in GC oven - the old one are the best ;) connect one end with a funnel or other liquid reservior and the gravitation will do the rest. otherwise you can force the flow through the column using a pump or vacuum.

regards,

G

Note, when you fill a capillary, you get Poiseuille flow, which means the stuff at the surfaces of the capillary don't move up into the capillary only the center does. So if you want a good sample, you need to overfill the capillary by several capillary volumes (≥ 5) to ensure uniformity. Other than that, as per the other answers, heat the capillary to 70 C, at those dimensions the contents will transfer heat quickly to the walls and equilibrate to that temperature. If the contents are at a different, higher, temperature and you are using the capillary to cool the contents, then glass is not the best choice since it will be path of most heat transfer resistance. A higher thermal conductance material would be better, as would a liquid bath on the outside.

Thank you all for the answers, but I think my question was misinterpreted.

I need to fill small quartz capillaries (see link)

https://www.charles-supper.com/en/page/product.cfm?idProduct=17

with a binary mixture that is homogeneous at 70°C but demixes at lower temperature. To be sure that the composition is the right one, I need to do the whole process at 70°C

Yes, the whole process will need to be done at 70C.