you can reverse the temp and apply heat to the extraction vessel and build up your desired pressure.
If you filled the vessel completely with liquid, there would be no room for expansion and the pressure would spike to the burst disk failure pressure very quickly.
Just wanted to re-visit this, With CO2 and open blast its more of a soak time in a supercritical or subcritical state.. I agree If you just ran it like butane you would need hundreds of pounds but if you heat it to a desired temp then let it soak for a good amount of time between 3 or 4 different runs you should get a nice yield from liquid CO2 or Dry ice.
Why do you believe more THC will dissolve with time? Especially at supercritical pressures, saturation (the maximum amount that will dissolve in a given volume of CO2) occurs virtually instantly. This is why supercritical machines pump massive volumes of CO2 through the sample. Letting it soak will do nothing.
One reference I have gives 0.2g/kg CO2 at 2kpsi. Solubility drops with pressure. Even if you could maintain 2 kpsi you'd need 5kg of CO2 per gram of THC extracted.
Another critical issue is the actual separation of THC from the CO2. If you simply open blast, as soon as the pressure drops in the chamber the dissolved solute will start precipitating Inside the extraction chamber. You need to move the entire volume of gas to another chamber and de-pressurize it there otherwise your yield will be very close to zero.
RB