The instruction MOV will copy the whole number portion of a float to an INT or DINT (or even BYTE) type, as far as it can be represented. The fractional portion will be lost.
INT and HEX are different things entirely. INT is a data type, HEX is a numeric representation, specifically base 16. In any simple binary form (not encoded as Floating point, or BCD), the representation is purely for convenience. There is no difference at all between:
00001010b (binary)
0x0A (hex)
10 (decimal)
012 (octal)
or, for that matter, even
101 (base 3).
If you are communicating between devices, and need to maintain the 32 bit representation of the floating point number, you can COPy it into two consecutive integers, and send to a device. Assuming there are no endian problems.