🏠 Home 📘 Track 1: Quantum Basics L23 — Reading the Results L24 — No-Cloning Theorem L25 — Everything Connects
L24 §4 · Build Something Real ~15 min

The No-Cloning Theorem —
You Can't Copy a Qubit

It is impossible to perfectly copy an unknown quantum state — not because hardware is imperfect, but because mathematics forbids it. Now that you've built circuits and manipulated qubits, you'll feel exactly why every strategy fails. And once you do, you'll see why this impossibility is the foundation of quantum cryptography.

✦ One Idea No quantum circuit can perfectly copy an unknown qubit state. Every approach — measuring first, using CNOT, designing a custom gate — either destroys the original, produces entanglement instead of duplication, or contradicts the linearity of quantum mechanics. The impossibility is a mathematical theorem, not a hardware limit.
no-cloning theorem linearity quantum cryptography BB84 eavesdropping Wootters–Zurek 1982 quantum information
Section 01
① Hook

The Impossible Copy

🔬
Think before you read — does CNOT copy a qubit?
You built the Bell pair circuit in L22. CNOT "spreads" a qubit's state to another. Does that make it a copier?

You want to copy qubit $|\psi\rangle = \alpha|0\rangle + \beta|1\rangle$ to a blank qubit in $|0\rangle$. You apply CNOT with $|\psi\rangle$ as control and $|0\rangle$ as target. Does this produce two independent copies — both in state $|\psi\rangle$?

In classical computing, copying information is trivial. A file can be duplicated in milliseconds. A bit can be copied with one transistor. Classical information obeys no law against duplication.

Quantum information is different. There is a theorem — proved by Wootters and Zurek, and independently by Dieks, in 1982 — that states: it is impossible to create a perfect copy of an arbitrary unknown quantum state. Not approximately impossible. Not impossible with today's hardware. Provably, mathematically impossible, for any physical device that obeys quantum mechanics.

This is the No-Cloning Theorem. It is one of the most important results in quantum information — and you now have every tool you need to understand exactly why it is true.

Section 02
② Intuition

Why You Feel This Already

Before the formal proof, consider what you already know from earlier lessons. Each piece of your knowledge points in the same direction.

💭 Your Prior Knowledge Already Implies No-Cloning
From L23 (measurement): When you measure $\alpha|0\rangle + \beta|1\rangle$, you get a single classical bit — and the superposition is destroyed. The amplitudes $\alpha$ and $\beta$ are gone. You cannot recover them from the one bit you measured.

From L22 (the Bell pair): When you apply CNOT with a superposed control to $|0\rangle$, you get an entangled state — not two independent copies. The Bell state $(|00\rangle + |11\rangle)/\sqrt{2}$ cannot be written as $|\psi\rangle \otimes |\psi\rangle$ for any single-qubit state $|\psi\rangle$.

Both together are the no-cloning theorem. Measurement destroys the state. Entanglement creates correlation, not duplication. There is no third path that threads between "destroy it" and "correlate it" to produce two genuine independent copies.

The informal reasoning is this: a quantum state $\alpha|0\rangle + \beta|1\rangle$ encodes more information than a classical bit — it contains complex amplitudes $\alpha$ and $\beta$ that a single measurement cannot reveal. To copy the state, you would need to first learn $\alpha$ and $\beta$. But learning them requires measuring the qubit many times — and each measurement collapses the original. By the time you know enough to recreate the state, the original has been destroyed many times over.

Key Insight
Cloning requires learning the state — and learning requires measurement — and measurement destroys superposition. You cannot learn and preserve simultaneously. This is the informal heart of no-cloning, and it follows directly from everything in Sections 2–4.
Section 03
③ Framework

Three Strategies That Fail — and Exactly Why

Here are the three most natural attempts at copying a qubit $|\psi\rangle = \alpha|0\rangle + \beta|1\rangle$ to a blank qubit $|0\rangle$ — and where each one breaks.

1
Measure first, then recreate
Measure $|\psi\rangle$ to learn its value, then set the blank qubit to match.
Measure $|\psi\rangle$: you get 0 (probability $|\alpha|^2$) or 1 (probability $|\beta|^2$). Set the blank qubit to the same classical result. Both qubits now show the same bit.

Why it fails: Measurement collapses the original. If $|\psi\rangle = \frac{1}{\sqrt{2}}(|0\rangle + |1\rangle)$ and you measure and get 0, the original is now $|0\rangle$ — not $(|0\rangle + |1\rangle)/\sqrt{2}$. You copied the classical outcome, not the quantum state. The superposition is gone permanently.
✗ Destroys the original superposition. Copies the measurement outcome, not the quantum state. The amplitudes α and β are lost.
2
Use a CNOT gate
$|\psi\rangle$ as control, $|0\rangle$ as target — spreads the control's state to the target.
Apply CNOT to $|\psi\rangle \otimes |0\rangle$. Expand by linearity:

$\text{CNOT}\bigl[(\alpha|0\rangle + \beta|1\rangle)|0\rangle\bigr] = \alpha|00\rangle + \beta|11\rangle$

Why it fails: Two perfect copies would give $|\psi\rangle \otimes |\psi\rangle = \alpha^2|00\rangle + \alpha\beta|01\rangle + \beta\alpha|10\rangle + \beta^2|11\rangle$. The CNOT output is missing $|01\rangle$ and $|10\rangle$ entirely. The result $\alpha|00\rangle + \beta|11\rangle$ is the Bell state (when $\alpha = \beta = 1/\sqrt{2}$) — it is the L22 entanglement circuit. Entanglement, not duplication.
✗ Output is entangled state α|00⟩+β|11⟩ — not two copies |ψ⟩⊗|ψ⟩. Missing the |01⟩ and |10⟩ cross terms that two independent copies would have.
3
Design a custom cloning unitary U
Find some gate $U$ such that $U|\psi\rangle|0\rangle = |\psi\rangle|\psi\rangle$ for every state $|\psi\rangle$.
Suppose $U$ clones $|0\rangle$: $U|00\rangle = |00\rangle$. And $U$ clones $|1\rangle$: $U|10\rangle = |11\rangle$. Now test it on $|\psi\rangle = \frac{1}{\sqrt{2}}(|0\rangle + |1\rangle)$. By linearity of $U$:

$U|\psi\rangle|0\rangle = \frac{1}{\sqrt{2}}(U|00\rangle + U|10\rangle) = \frac{1}{\sqrt{2}}(|00\rangle + |11\rangle)$

But two perfect copies require: $|\psi\rangle|\psi\rangle = \frac{1}{2}(|00\rangle + |01\rangle + |10\rangle + |11\rangle)$.

These are different states. Linearity forces the entangled result. No unitary $U$ can satisfy both "works for basis states" and "works for all superpositions."
✗ Linearity of quantum mechanics makes a universal cloning unitary mathematically impossible. Working on |0⟩ and |1⟩ forces incorrect output on all superpositions.
🚫
CNOT copies basis states — but that is not useful cloning
CNOT maps $|0\rangle|0\rangle \to |0\rangle|0\rangle$ and $|1\rangle|0\rangle \to |1\rangle|1\rangle$ — so it "copies" $|0\rangle$ and $|1\rangle$ individually. The problem is it cannot simultaneously copy all superpositions. No-cloning is about copying an unknown state — one you have not been told. If you already know the state is $|0\rangle$ or $|1\rangle$, you can copy it trivially. The challenge — and the theorem — is about states whose amplitudes are unknown to you.
Section 04
∑ Mathematics

The Proof — Short, Elegant, Final

The formal no-cloning proof is one of the shortest in quantum information theory. It requires only the linearity of quantum mechanics and the definition of what "copying" means. No advanced mathematics. No complex analysis. Just logic applied to what you already know.

Step 1 — Define what a cloning machine would need to do
$$\text{Suppose a unitary } U \text{ exists such that for ALL states } |\psi\rangle:$$ $$U\bigl(|\psi\rangle \otimes |0\rangle\bigr) = |\psi\rangle \otimes |\psi\rangle$$
$|\psi\rangle$ is the unknown input qubit we want to copy. $|0\rangle$ is the blank "target" qubit that will hold the copy. $U$ is the hypothetical cloning gate. $\otimes$ is the tensor product combining two qubits into a joint two-qubit state. We assume $U$ exists and derive a contradiction. This is a proof by contradiction.
Step 2 — U must work on both basis states
$$U\bigl(|0\rangle \otimes |0\rangle\bigr) = |0\rangle \otimes |0\rangle$$ $$U\bigl(|1\rangle \otimes |0\rangle\bigr) = |1\rangle \otimes |1\rangle$$
If $U$ is a universal cloner, it must in particular clone $|0\rangle$ and $|1\rangle$ correctly. These two equations are what "cloning works on the basis states" means in mathematical notation.
Step 3 — Test on a superposition; contradiction appears
$$\text{Let } |\psi\rangle = \tfrac{1}{\sqrt{2}}(|0\rangle + |1\rangle). \text{ Cloning rule demands:}$$ $$U\bigl(|\psi\rangle \otimes |0\rangle\bigr) \;\overset{?}{=}\; |\psi\rangle \otimes |\psi\rangle = \tfrac{1}{2}\bigl(|00\rangle + |01\rangle + |10\rangle + |11\rangle\bigr) \quad\cdots(1)$$ $$\text{But linearity of } U \text{ forces:}$$ $$U\bigl(|\psi\rangle \otimes |0\rangle\bigr) = \tfrac{1}{\sqrt{2}}\bigl(U|00\rangle + U|10\rangle\bigr) = \tfrac{1}{\sqrt{2}}\bigl(|00\rangle + |11\rangle\bigr) \quad\cdots(2)$$ $$\text{(1) }\neq\text{ (2). The contradiction is complete.} \quad\blacksquare$$
Equation (1) is what a perfect cloner must output — two independent copies of $|\psi\rangle$, which when expanded gives four equal-amplitude terms. Equation (2) is what linearity of $U$ forces — the Bell state, with only two terms and different amplitudes ($1/\sqrt{2}$ vs $1/2$). These states are provably different. Therefore no such $U$ can exist. The proof is complete in three steps. Source: Wootters & Zurek, Nature 299, 802–803 (1982).

This proof generalises immediately to any superposition $\alpha|0\rangle + \beta|1\rangle$ with $\alpha \neq 0$ and $\beta \neq 0$. No quantum gate, no quantum circuit, no physical device consistent with quantum mechanics can clone an unknown qubit.

💡
What the proof actually uses
Only two facts: (1) quantum gates are linear — $U(\alpha|\phi\rangle + \beta|\psi\rangle) = \alpha U|\phi\rangle + \beta U|\psi\rangle$ — and (2) the definition of a perfect clone — $U|\psi\rangle|0\rangle = |\psi\rangle|\psi\rangle$. No other quantum mechanics is needed. No-cloning is a direct consequence of linearity — the most fundamental property of quantum evolution.
Section 05
⑤ Interactive

Clone Attempt Lab

Pick a state to clone and a method to try. The lab shows exactly what each approach produces — and why it is not a perfect copy. Try every combination. The key experiment: apply every method to $|{+}\rangle$ and see no-cloning made concrete.

🔬 Clone Attempt Lab
Pick state → pick method → attempt clone → see exactly why it fails
INTERACTIVE
1 — State to clone
|ψ⟩ =
2 — Cloning strategy
Output after cloning attempt
Original qubit
—
Press the button above to run.
Target qubit
—
—
Choose a state and a method, then press 🔬 Attempt to Clone to see exactly what the method produces and why it is or is not a perfect copy.

The key pattern across all experiments: basis states ($|0\rangle$, $|1\rangle$) can be copied trivially by several methods — but that is because they are already classical. Superpositions ($|{+}\rangle$, $|{-}\rangle$, $\alpha|0\rangle+\beta|1\rangle$) fail with every strategy. No-cloning is specifically a statement about unknown quantum states — classical information is copyable; quantum superposition is not.

Section 06
③ Framework

Why This Is Good News — The Cryptography Connection

No-cloning sounds like a restriction. It is. But restrictions in physics have often turned out to be foundations of technology. The speed of light cannot be exceeded — and that gives us the causal structure of the universe. Energy cannot be created from nothing — and that gives us thermodynamics. Quantum states cannot be cloned — and that gives us provably secure cryptography.

Why no-cloning defeats eavesdroppers

In classical cryptography, an eavesdropper can intercept a message, copy every bit, and forward the original — completely silently, leaving no trace. This is why classical security relies on computational hardness: problems that are believed to be difficult, but not provably impossible to break.

🔐 No-Cloning Makes Eavesdropping Detectable
In the BB84 quantum key distribution protocol, the sender (Alice) transmits a secret key encoded as a sequence of qubit states. An eavesdropper (Eve) on the quantum channel faces an impossible choice:

Option 1 — Measure and forward: Eve measures the qubits to learn their values, then sends replacement qubits to the receiver. But measurement collapses the original superpositions. The replacements are in the wrong basis half the time. The receiver (Bob) and Alice compare a sample of their data — the error rate is elevated above the expected level for a noise-free channel. They detect the eavesdropper.

Option 2 — Clone and forward: Eve tries to keep a copy and forward the original. No-cloning forbids this — any copy attempt disturbs the original, again producing detectable errors.

There is no Option 3. The no-cloning theorem closes all gaps. The security is not computational ("hard to break") — it is physical ("breaking it violates quantum mechanics"). Provably secure, not probably secure.

No-cloning and quantum error correction

Classical error correction works by making redundant copies of a bit and using majority voting to detect and fix errors. No-cloning prevents the direct quantum analogue. Quantum error correction therefore uses a different approach: it spreads a single qubit's information across multiple entangled qubits in a way that allows errors to be detected and corrected without ever measuring the encoded state directly. This is more subtle, more powerful, and it is made necessary by no-cloning.

📜
Two groups, same proof, same week — 1982
William Wootters and Wojciech Zurek published their proof in Nature on 28 October 1982. Dennis Dieks published an independent, essentially identical proof in Physics Letters A the same month — neither group knew about the other's work. The theorem had been informally understood for some time, but the clean proof had not been published. It is now a cornerstone of quantum information theory, and its implications for cryptography were recognised almost immediately after publication.
The Flip
No-cloning is not a limitation to work around. It is a feature to exploit. Quantum cryptography is secure because of it. Quantum error correction is designed around it. Quantum teleportation works through it — teleporting a state without cloning it, by destroying the original at the source the instant it appears at the destination. The "impossibility" is the foundation, not the obstacle.
Lesson Summary

What You Now Know About No-Cloning

  • 🚫
    No quantum circuit can perfectly copy an unknown quantum state
    This is a mathematical theorem, not a hardware limitation. Every approach — measuring first, using CNOT, designing a custom unitary — fails on superpositions. The theorem was proved in 1982 by Wootters & Zurek and independently by Dieks. It cannot be circumvented by more clever engineering.
  • 📐
    The proof uses only linearity — and takes three steps
    Assume $U|\psi\rangle|0\rangle = |\psi\rangle|\psi\rangle$ for all $|\psi\rangle$. Apply $U$ to $|0\rangle$ and $|1\rangle$ correctly. Now apply linearity to $(|0\rangle+|1\rangle)/\sqrt{2}$: you get $(|00\rangle+|11\rangle)/\sqrt{2}$ (Bell state), not $(|0\rangle+|1\rangle)^{\otimes2}/2$ (two copies). These differ. Contradiction. $\blacksquare$
  • 🔬
    CNOT creates entanglement, not duplication
    CNOT maps $(\alpha|0\rangle+\beta|1\rangle)|0\rangle \to \alpha|00\rangle+\beta|11\rangle$. A perfect copy would be $|\psi\rangle\otimes|\psi\rangle = \alpha^2|00\rangle+\alpha\beta|01\rangle+\alpha\beta|10\rangle+\beta^2|11\rangle$. The $|01\rangle$ and $|10\rangle$ cross-terms are missing. This is why the L22 Bell pair circuit creates entanglement — it is mathematically impossible for it to create a copy.
  • 🔐
    No-cloning is the foundation of quantum cryptography
    An eavesdropper on a quantum channel cannot copy quantum states without disturbing them. Measurement collapses superpositions; any replacement qubits show elevated error rates detectable by the legitimate parties. BB84 quantum key distribution is provably secure — not "computationally hard to break" but "physically impossible to break without detection" — because of no-cloning.
  • ✨
    The "impossibility" is a feature, not a bug
    No-cloning shapes the entire landscape of quantum information: it forces quantum error correction to use entanglement instead of redundancy, makes quantum teleportation possible (teleport without clone), and gives quantum cryptography unconditional security. Understanding why something is impossible is often more powerful than knowing what is possible.
Quick Check
How clearly does the no-cloning theorem click for you?

You have built circuits, measured results, and proved
that quantum states cannot be copied.
Now: bring it all together.
Section 4 synthesis — everything connected in one lesson.

→ Everything Connects — L25
Sources & Further Reading
← Previous
Reading the Results
L23 — Statistics, histograms & classical bits
← → navigate lessons · Space = mark complete