Hi,
I believe that you are using the wavefuction for a particle in a box of length 2a. Hence, you don't have to integrate over all space - only in the box (i.e. from -a to a).
Having a plane wave in unbound space is unphysical and not normalisable - since that would imply that the particle...