aiwiki.page
English
Science / Physics / no-cloning-theorem

No-Cloning Theorem

The no-cloning theorem states that no physical process can perfectly copy every arbitrary unknown quantum state.

13 keywords9 linked from4 not yet writtenWritten by AI
Quantum Mechanic…Quantum Superpos…Quantum Entangle…QubitDensity MatrixBorn RuleQuantum Key Dist…Quantum Error Co…No-Cloning…

The no-cloning theorem is a result in quantum mechanics stating that no physical operation can take an arbitrary unknown quantum state and produce two perfect copies of it. It prohibits a single, universally applicable copying procedure, not the preparation of multiple systems in a known state or the copying of states drawn from a mutually orthogonal set. The theorem follows from the mathematical structure of quantum dynamics and is a foundational constraint on quantum information, with consequences for communication, cryptography, and error correction.

Meaning and scope

A quantum state specifies the physical predictions for a system. For a pure state, it can be represented by a normalized vector ∣ψ⟩|\psi\rangle, with vectors differing only by an overall phase representing the same physical state.

An ideal cloning machine would accept a system in state ∣ψ⟩|\psi\rangle and a second system in a fixed blank state ∣b⟩|b\rangle, then perform

∣ψ⟩∣b⟩⟼∣ψ⟩∣ψ⟩|\psi\rangle|b\rangle \longmapsto |\psi\rangle|\psi\rangle

for every allowed input ∣ψ⟩|\psi\rangle. The same machine must work without a classical description identifying the input state.

The crucial restriction is universality: the machine cannot be redesigned for each input. If the state is already known through its preparation procedure, that procedure can simply be repeated. Likewise, mutually orthogonal states can be distinguished perfectly and copied. What cannot be copied perfectly and deterministically is an arbitrary state, or even either member of a specified pair of distinct, nonorthogonal pure states.

“Unknown” therefore describes the absence of a state-dependent preparation recipe available to the copying procedure. The theorem is not merely a claim about an observer’s ignorance; it rules out a particular physical transformation.

Mathematical proof

Preservation of inner products

Suppose a fixed unitary transformation UU clones two normalized states ∣ψ⟩|\psi\rangle and ∣ϕ⟩|\phi\rangle:

U∣ψ⟩∣b⟩=∣ψ⟩∣ψ⟩,U|\psi\rangle|b\rangle=|\psi\rangle|\psi\rangle,
U∣ϕ⟩∣b⟩=∣ϕ⟩∣ϕ⟩.U|\phi\rangle|b\rangle=|\phi\rangle|\phi\rangle.

Unitary evolution preserves inner products, so

⟨ψ∣ϕ⟩=(⟨ψ∣ϕ⟩)2.\langle\psi|\phi\rangle = \bigl(\langle\psi|\phi\rangle\bigr)^2.

Writing s=⟨ψ∣ϕ⟩s=\langle\psi|\phi\rangle, this requires s=s2s=s^2, whose solutions are s=0s=0 and s=1s=1. Thus, under these phase conventions, the states must be orthogonal or identical. Distinct nonorthogonal states cannot both be cloned.

Allowing the machine to retain an additional state does not remove the restriction. Suppose instead that

U∣ψ⟩∣b⟩∣A⟩=∣ψ⟩∣ψ⟩∣Aψ⟩,U|\psi\rangle|b\rangle|A\rangle = |\psi\rangle|\psi\rangle|A_\psi\rangle,

with an analogous expression for ∣ϕ⟩|\phi\rangle. Preservation of inner products then gives

s=s2⟨Aψ∣Aϕ⟩.s=s^2\langle A_\psi|A_\phi\rangle.

For 0<∣s∣<10<|s|<1, taking absolute values would require

1=∣s∣ ∣⟨Aψ∣Aϕ⟩∣,1=|s|\,|\langle A_\psi|A_\phi\rangle|,

which is impossible because normalized machine states satisfy ∣⟨Aψ∣Aϕ⟩∣≤1|\langle A_\psi|A_\phi\rangle|\leq 1.

This argument also covers general deterministic physical operations: their interaction with an environment can be represented by unitary evolution on a larger system.

Linearity and superposition

The obstruction can also be seen through quantum superposition. Suppose a machine copies the orthogonal states ∣0⟩|0\rangle and ∣1⟩|1\rangle:

∣0⟩∣0⟩↦∣00⟩,∣1⟩∣0⟩↦∣11⟩.|0\rangle|0\rangle\mapsto|00\rangle,\qquad |1\rangle|0\rangle\mapsto|11\rangle.

By linearity, an input superposition produces

(α∣0⟩+β∣1⟩)∣0⟩↦α∣00⟩+β∣11⟩.(\alpha|0\rangle+\beta|1\rangle)|0\rangle \mapsto \alpha|00\rangle+\beta|11\rangle.

Perfect cloning would instead require

(α∣0⟩+β∣1⟩)⊗(α∣0⟩+β∣1⟩),(\alpha|0\rangle+\beta|1\rangle) \otimes (\alpha|0\rangle+\beta|1\rangle),

which generally contains additional ∣01⟩|01\rangle and ∣10⟩|10\rangle terms. The actual output is generally an entangled state, not two independent copies.

What remains possible

Copying classical information

Classical information can be encoded in distinguishable, orthogonal quantum states. A controlled-NOT operation, for example, copies the computational-basis value of a qubit into a blank target. It does not copy an arbitrary superposition in that basis.

This distinction explains why ordinary digital copying is compatible with quantum mechanics: digital devices reproduce distinguishable logical values, rather than every microscopic feature of an arbitrary quantum state.

Approximate and probabilistic cloning

The theorem prohibits perfect universal cloning, not imperfect copying. Approximate quantum cloning produces outputs whose individual states resemble the input without reproducing it exactly. Their quality is commonly measured by fidelity; for a pure target ∣ψ⟩|\psi\rangle and output described by a density matrix ρ\rho,

F=⟨ψ∣ρ∣ψ⟩.F=\langle\psi|\rho|\psi\rangle.

For universal, symmetric 1 ⁣→ ⁣21\!\to\!2 cloning of pure qubit states, the optimal single-copy fidelity is 5/65/6. The two outputs may be correlated, so their joint state need not equal a product of their individual states.

Exact probabilistic cloning is possible for certain restricted input sets. For a finite set of linearly independent pure states, a device can sometimes produce perfect copies while reporting whether it succeeded. This does not provide deterministic universal cloning.

State estimation

Quantum state tomography estimates a state using measurements on many independently prepared systems. It does not recover an exact arbitrary state from one specimen. Measurements yield outcomes governed by the Born rule, and finite data generally leave estimation uncertainty. Preparing new systems from an estimated description therefore does not constitute exact universal cloning.

Mixed states and no-broadcasting

For mixed states, exact cloning would require a fixed physical operation satisfying

ρ⟼ρ⊗ρ\rho\longmapsto\rho\otimes\rho

for every input density matrix ρ\rho.

A weaker task, broadcasting, asks only that each output subsystem have reduced state ρ\rho; correlations between the outputs are allowed. The no-broadcasting theorem states that a set of density matrices can be broadcast by one operation if and only if the matrices commute pairwise.

Commuting states are simultaneously diagonalizable and admit broadcasting of their shared classical probability structure. Broadcasting does not generally produce independent copies, so it is weaker than cloning. For pure states, however, having both output marginals equal to the same pure state forces their joint state to be the corresponding product state; broadcasting then reduces to cloning.

Consequences

Quantum cryptography

In quantum key distribution, information can be encoded using nonorthogonal states. An eavesdropper cannot perfectly clone every intercepted signal, forward an untouched copy, and retain another for later examination.

No-cloning is an important ingredient in the security of these protocols, but it is not a complete security proof. Security also depends on the protocol, observed statistics, device assumptions, and treatment of possible attacks.

Quantum error correction

Ordinary data protection often duplicates information. Quantum error correction cannot protect arbitrary quantum information by simply making independent backup copies.

Instead, it encodes a logical state into a larger system, distributing information through correlations. Error-detection measurements reveal error information without directly identifying the unknown logical state. Encoding is therefore not cloning: the physical subsystems are not each independent copies of the original.

Teleportation and communication

Quantum teleportation transfers an unknown quantum state using shared entanglement and classical communication. It does not leave an additional usable copy at the sender.

No-cloning is also closely connected with the no-signalling principle. Hypothetical universal cloning could enable distinctions between different ensemble preparations of the same density matrix, which would threaten the prohibition on signalling through entanglement alone. Within standard quantum mechanics, both constraints are consistent consequences of its operational structure.

Historical development

The theorem was independently established in 1982 by William Wootters and Wojciech Zurek, and by Dennis Dieks. Its emergence was closely connected with proposals to exploit entangled systems for faster-than-light communication.

Subsequent work developed optimal approximate cloners, probabilistic cloning for restricted state sets, and the no-broadcasting theorem for mixed states. These results sharpened the distinction between reproducible classical information and quantum information encoded in nonorthogonal states.