The No-Cloning Theorem

Classical computers copy bits constantly — your CPU shuttles data between registers and cache billions of times per second with perfect fidelity. It seems natural to expect the same for qubits. The no-cloning theorem says this is impossible: there is no operation that takes an unknown quantum state $|\psi\rangle$ and produces two identical copies.

The argument from linearity

Suppose we had a hypothetical cloning machine — a unitary $U$ that copies any qubit onto a blank target $|0\rangle$:

$U\bigl(|\psi\rangle \otimes |0\rangle\bigr) = |\psi\rangle \otimes |\psi\rangle.$

Test it on the two basis states. Cloning $|0\rangle$ gives $|0\rangle|0\rangle$ and cloning $|1\rangle$ gives $|1\rangle|1\rangle$ — so far no contradiction. Now try the equal superposition

$|{+}\rangle = \tfrac{1}{\sqrt{2}}\bigl(|0\rangle + |1\rangle\bigr).$

Because $U$ is linear (every quantum operation must be), it distributes over the superposition:

$U\bigl(|{+}\rangle \otimes |0\rangle\bigr) = \tfrac{1}{\sqrt{2}}\bigl(U(|0\rangle|0\rangle) + U(|1\rangle|0\rangle)\bigr) = \tfrac{1}{\sqrt{2}}\bigl(|0\rangle|0\rangle + |1\rangle|1\rangle\bigr).$

But the result we wanted — two copies of $|+\rangle$ — is

$|{+}\rangle \otimes |{+}\rangle = \tfrac{1}{2}\bigl(|0\rangle|0\rangle + |0\rangle|1\rangle + |1\rangle|0\rangle + |1\rangle|1\rangle\bigr).$

These two states are not equal, so no single linear $U$ can clone both $|0\rangle$ and $|+\rangle$. The same reasoning applies whenever the desired clone of a superposition $\alpha|0\rangle + \beta|1\rangle$ (with $\alpha,\beta \neq 0$) differs from what linearity actually produces, which happens for every non-trivial superposition. A complementary inner-product argument — if $U$ cloned both $|\phi\rangle$ and $|\psi\rangle$, unitarity would force $\langle\phi|\psi\rangle = \langle\phi|\psi\rangle^{2}$, leaving only $\langle\phi|\psi\rangle = 0$ or $1$ as solutions — shows that only orthogonal or identical states could ever be cloned simultaneously, confirming the prohibition is completely general [nielsen-chuang].

Why it matters for teleportation and superdense coding

No-cloning is not a nuisance — it is a foundational feature. Because quantum information cannot be duplicated, transmitting a qubit's state from one party to another requires actually moving or destroying the original. Quantum teleportation gets around this not by copying but by consuming an entangled Bell pair and sending two classical bits, with the original state being destroyed in the process. Superdense coding, conversely, uses entanglement to pack two classical bits of information into a single qubit transmission. Both protocols only make sense in a world where no-cloning holds, and both are the subject of the lessons that follow in this module.

No-cloning also underlies quantum key distribution: an eavesdropper cannot intercept and re-transmit a qubit without disturbing it, because doing so would require cloning an unknown state.

Sign in on the full site to ask questions and join the discussion.