On the cryptographic potential of single-qubit rotations