That's not guaranteed to work.
If min+max is a negative signed int, then the result of right shifting is implementation dependent.
Instead, you could do
int middle = (unsigned)(min + max) >> 1;
Still, this relies on the original indices falling in the non-negative range of signed ints.
This will work for n < MAX_INT/2. For larger n, middle must be calculated as min + ((max - min) >> 1), which is one subtraction per iteration more.
Making middle an unsigned int should make (min + max) >> 1 work for all int values.