Tuesday, February 9, 2021

Bell's theorem

EPR-Bohm Experiment 

A pion decays at rest: $\pi^0\longrightarrow e^-+e^+$.

The pair electron-positron is maximally entangled in the singlet state (linear superposition):

\begin{eqnarray}|e^-e^+\rangle&=&\frac{1}{\sqrt{2}}(|+\rangle|-\rangle-|-\rangle|+\rangle)\nonumber\\&=&\frac{1}{\sqrt{2}}(|+\rangle_x|-\rangle_x-|-\rangle_x|+\rangle_x).\end{eqnarray}

 


Let us suppose that the measurement of the spin $S_z$ by Alice leads to the state:

\[|+\rangle|-\rangle=|+\rangle\frac{1}{\sqrt{2}}(|+\rangle_x-|-\rangle_x).\]

The measurement of the spin $S_z$ by Bob is unecesseary and Bob can instead measure the spin $S_x$. Bob can then determined the two spins $S_z$ and $S_x$ at the same time which is forbidden by Heisenberg's uncertainty principle since the operators $S_z$ and $S_x$ are incompatible.

This is the paradox and quantum mechanics is incomplete.

Classical Realism

$\lambda$: hidden variable. 

Probability density $\rho(\lambda)$ : $\rho(\lambda)>0$ , $\int \rho(\lambda)d\lambda=1$.

Local Causality (Free Will)

Measurement Alice  $\vec{a}$. 

Measurement Bob $\vec{b}$.

$\vec{a}$, $\vec{b}$ are freely and independently chosen.


Alice: $ S_a=\pm 1$....$f(\vec{a},\lambda)=\pm 1$.

Bob: $S_b=\pm 1$...$g(\vec{b},\lambda)=\pm 1$.

If $\vec{b}=\vec{a}$....$g(\vec{a},\lambda)=-f(\vec{a},\lambda)$.

Expected value of product of Alice and Bob measurements:

\begin{eqnarray}P(a,b)&=& \int f(\vec{a},\lambda)g(\vec{b},\lambda) \rho(\lambda)d\lambda\nonumber\\&=&-\int f(\vec{a},\lambda)f(\vec{b},\lambda) \rho(\lambda)d\lambda\nonumber\\\end{eqnarray}

 

\begin{eqnarray} P(a,b)-P(a,c)&=&-\int \bigg[f(\vec{a},\lambda)f(\vec{b},\lambda)-f(\vec{a},\lambda)f(\vec{c},\lambda)\bigg] \rho(\lambda)d\lambda\nonumber\\&=&-\int f(\vec{a},\lambda)f(\vec{b},\lambda)\bigg[1-f(\vec{b},\lambda)f(\vec{c},\lambda)\bigg] \rho(\lambda)d\lambda\nonumber\\\end{eqnarray}

 

\[-1\leq f(\vec{a},\lambda)f(\vec{b},\lambda)\leq 1\Rightarrow 0\leq 1-f(\vec{a},\lambda)f(\vec{b},\lambda)\leq 2\] 

Bell's inequality (hidden variables):

\begin{eqnarray} |P(a,b)-P(a,c)|&\leq &\int |f(\vec{a},\lambda)f(\vec{b},\lambda)|\bigg[1-f(\vec{b},\lambda)f(\vec{c},\lambda)\bigg] \rho(\lambda)d\lambda\nonumber\\&\leq &\int \bigg[1-f(\vec{b},\lambda)f(\vec{c},\lambda)\bigg] \rho(\lambda)d\lambda\nonumber\\&\leq & 1+P(b,c).\end{eqnarray}

Quantum Mechanics:

\begin{eqnarray}P(a,b)=-\vec{a}.\vec{b}.\end{eqnarray}

No comments:

Post a Comment

AI as the Hardworking and Perfect PhD Student

  Some people seem very defensive, and sometimes openly hostile, about the use of AI in scientific work. Others go even further and become j...