We prove that the entangling capacity of a two-qubit unitary operator without local ancillas, both with and without the restriction to initial product states, as quantified by the maximum attainable concurrence, is directly related to the distinguishability of a closely related pair of two-qubit unitary operators. These operators are the original operator transformed into its canonical form and the adjoint of this canonical form. The distinguishability of these operators is quantified by the minimum overlap of the output states over all possible input probe states. The entangling capacity of the original unitary operator is therefore directly related to the degree of non-Hermiticity of its canonical form, as quantified in an operationally satisfactory manner in terms of the extent to which it can be distinguished, by measurement, from its adjoint. Furthermore, the maximum entropy of entanglement, again without local ancillas, that a given two-qubit unitary operator can generate is found to be closely related to the classical capacities of certain quantum channels.
We analyze the class of single-qubit channels with the environment modeled by a one-qubit mixed state. The set of affine transformations for this class of channels is computed analytically, employing the canonical form for the two-qubit unitary operator. We demonstrate that
83 of the generalized depolarizing channels can be simulated by the one-qubit mixed-state environment by explicitly obtaining the shape of the volume occupied by this class of channels within the tetrahedron representing the generalized depolarizing channels. Further, as a special case, we show that the two-Pauli channel cannot be simulated by a one-qubit mixed-state environment.
Entanglement-changing power of an arbitrary two-qubit operation, including increasing and decreasing power, is investigated in this paper. We consider the maximal entanglement
Cmax and the minimal entanglement
Cmin of the states obtained by a given two-qubit unitary operation
Ud acting on arbitrary pure states with fixed entanglement
C0. We give the condition that the maximal entanglement
Cmax of the obtained states can be
1 and the minimal entanglement
Cmin can be
0. When the maximal entanglement
Cmax cannot be
1, we give the maximal value it can reach. When the minimal entanglement
Cmin cannot be
0, we give the minimal value it can reach. We think
Cmax and
Cmin represent the entanglement-changing power of two-qubit unitary operations.