If you use a complex FFT, you will get an array of complex numbers for your spectrum. You could modify the array, then convert it back to time data using the inverse complex FFT.
I have doubts that you are going to achieve what you want, though. If you are just trying to shift all the frequencies to an audible range, there is an easier way to do it. Use a high sampling rate when you record the signal (200 kHz), then play it back at a slower rate (50 kHz). All of the frequencies will drop to 1/4 of their original frequency.
Bruce
Bruce Ammons
Ammons Engineering