I almost always have effects in buses. This way I control of all parameters of the effect. Including sending it to another bus for further processing.
Pretty much I only use EQ on channels.
As described by others: put the reverb in a bus
Add a send from the vocal track to the bus
Adjust volumes to taste
Create a new track, name it so the wav file is properly named
Select the vocal track
Bounce to tracks
Select the new track as destination
Select buses as the source
Select only the reverb bus
Select Stereo
Select NO dither
Make it so.
With the reverb now on a bus you can use it later. It essentially uses no CPU/Ram by just lurking. Disable it if you feel the need.
Some have reported that reverb tails sound better when recorded in real time instead of Fast Bounce.