No-cloning, no-deleting, no-broadcasting
Suppose you could copy an unknown qubit. Share an entangled pair; Alice, far away, measures hers in the z or x basis to send a 0 or 1. Bob makes a thousand copies of his qubit and measures them all in z: if Alice used z, all copies agree; if she used x, they split half and half. A message, instantly, across any distance. Relativity would be broken. In 1982 William Wootters and Wojciech Zurek, and independently Dennis Dieks, proved in a few lines that quantum mechanics forbids the copying. Relativity is safe.
On the plate: how faithfully different copying machines reproduce a qubit, by where it sits on the Bloch sphere.
No-cloning, two proofs
From linearity. Suppose a unitary copies and . By linearity, , a Bell state, not . From unitarity. If and likewise for , taking inner products gives , so the overlap is 0 or 1. Only orthogonal states, which are effectively classical, can be copied. Established
No deleting, no broadcasting
The reverse fails too. Arun Pati and Samuel Braunstein proved in 2000 that no process can take two copies of an unknown state and delete one, leaving a blank, without the information turning up elsewhere: quantum information can be moved but not destroyed. For mixed states the natural generalization of copying is broadcasting: producing a two-part state each of whose halves equals the original. Howard Barnum and colleagues proved in 1996 that non-commuting mixed states cannot be broadcast. The three theorems together say quantum information behaves like a conserved, uncopyable quantity.
The best imperfect copier
If perfect copies are impossible, how good can approximate ones be? The CNOT “copier” copies and perfectly but turns into an entangled pair, each half with fidelity only 1/2. Vladimír Bužek and Mark Hillery found in 1996 a universal cloner whose two outputs each have fidelity 5/6 with the input, the same for every input state, and this is optimal. Choose: or , and set the input’s polar angle to 90°. Averaged over all inputs, the CNOT copier gets 2/3; the optimal cloner 5/6. Established
Why it matters
No-cloning is what makes quantum key distribution secure (Chapter 132): an eavesdropper cannot copy the qubits in transit to measure later. It is why quantum error correction cannot simply back up qubits and must instead spread information across entangled codes (Chapter 137). And it is why teleportation (next chapter) destroys the original: it moves a state, never copies it.
What you now have
Two proofs of no-cloning, no-deleting and no-broadcasting, the CNOT copier and the optimal 5/6 universal cloner, and the theorem’s role in cryptography, error correction and teleportation.
· · ·
Three questions to carry with you.
And one to work out. The CNOT copier’s fidelity for an input at polar angle is . What is it at ?
Sources: W. K. Wootters and W. H. Zurek, Nature 299, 802 (1982). D. Dieks, Phys. Lett. A 92, 271 (1982). V. Bužek and M. Hillery, Phys. Rev. A 54, 1844 (1996). H. Barnum et al., Phys. Rev. Lett. 76, 2818 (1996). A. K. Pati and S. L. Braunstein, Nature 404, 164 (2000). Checked by code: CNOT copier fidelity (1 + cos²θ)/2, average 2/3; Bužek–Hillery 5/6.