Is it possible to convert a symbolic scalar into an LTI model so that I can obtain its poles and zeros?

For example,I implement the code F = ztrans(2^n) where n is declared as syms. The resulting symbolic scalar expression is: z/(z - 2). Although I know I can obtain the zeros from the function solve(F), this does not get me the poles? Is there any other way I can do this? Thank you very much!