# Proof of no-cloning

I was reading a proof of No-cloning theorem, there are a couple of steps that are not clear to me, but the book does not give explanation for them. So here it is: Theorem: It is impossible to create an identical copy of an arbitrary unknown quantum state. Proof (by contradiction): Suppose that there exists a unitary $$C$$ (that copies an arbitrary unknown q-state). Then: .

This is the first thing I have trouble with, why do we take only one state ($$\left| 0 \right>$$ state) to prove that it is impossible to copy any arbitrary state into $$any$$ qubit? The proof goes on stating:

Here I don't understand how (116) is the same as (117) (don't mind explaining the identity matrix in the middle). I will appreciate it if you make these 2 steps clearer to me.

• Hi, Archil. Please use MathJax to write mathematical expressions. Images are not text-searchable (c.f. physics.meta.stackexchange.com/a/10573/199113). Dec 13 '18 at 15:54
• I also give an intuitive proof of the no-cloning theorem in this answer. Dec 13 '18 at 19:48
• Kindly mention the source from where you read this no cloning theorem? I am also interested Aug 27 at 9:34
• @A B, I probably read it on Vazirani's course on EdX, couldn't find it exactly but here is something very similar inst.eecs.berkeley.edu/~cs191/fa05/lectures/lecture6_fa05.pdf Aug 27 at 16:57

$$$$(\langle\psi_1|\otimes\langle0|)C^\dagger C(|\psi_2\rangle\otimes|0\rangle) = (\langle\psi_1|\otimes\langle0|)(|\psi_2\rangle\otimes|0\rangle) = \langle\psi_1|\psi_2\rangle\otimes\langle0|0\rangle=\langle\psi_1|\psi_2\rangle\langle0|0\rangle,$$$$ where in the second step I used the property of tensor products that $$(A\otimes B)(C\otimes D)=AC\otimes BD$$, where obviously the dimensions of the multiplying matrices should match; and for the third step I used the fact that $$\langle\psi_1|\psi_2\rangle$$ and $$\langle0|0\rangle$$ are scalars, implying that their tensor product is just a multiplication between them.
For your first question, state $$|0\rangle$$ is used because it is the typical ancillary system used for quantum algorithms, that is, the zero state is the one used a auxilliary variable for doing the corresponding operations. However, note that the proof is not state dependent, meaning that you can take an arbitrary ancillary qubit, name it $$|\rho\rangle$$, and prove the theorem. Here I leave you such proof: $$$$C(|\psi_1\rangle\otimes|\rho\rangle) = |\psi_1\rangle\otimes|\psi_1\rangle\\ C(|\psi_2\rangle\otimes|\rho\rangle) = |\psi_2\rangle\otimes|\psi_2\rangle$$$$ And so $$C$$ would be the unitary we are looking for, so now the proof: $$$$\langle\psi_1|\psi_2\rangle=\langle\psi_1|\psi_2\rangle\langle\rho|\rho\rangle =(\langle\psi_1|\otimes\langle\rho|)C^\dagger C(|\psi_2\rangle\otimes|\rho\rangle) = (\langle\psi_1|\otimes\langle\psi_1|)(|\psi_2\rangle\otimes|\psi_2\rangle)=(\langle\psi_1|\psi_2\rangle)^2.$$$$ And so the same contradiction as before is obtained, and the theorem is proved.