1
$\begingroup$

I'm studying a discrete state coupled to a continuum of states. The Hamiltonian of this interaction, which i called Friedrichs Hamiltonian, is given by: $$H_{F}=E_{e}|e\rangle\langle e|+\int U|\chi_{U}\rangle\langle\chi_{U}|dU+\int\left(\xi_{U}|\chi_{U}\rangle\langle e|+\xi_{U}^{*}|e\rangle\langle\chi_{U}|\right)dU.$$ $E_{e}$ is the energy of the discrete state, $|e\rangle$, $U$ is the energy of the continuum of states, $|\chi_{U}\rangle$ and $\xi_{U}$ is the coupling coefficient of the discrete state to the continuum.

After some calculations, i got that the amplitude of probability is given by: $$\langle e\vert e^{-iHt}\vert e\rangle=\int\vert A(E)\vert^{2}e^{-iEt}dE,$$ so, the time evolution of the probability is: $$P=\vert\langle e\vert e^{-iHt}\vert e\rangle\vert^{2}=\int\vert A(E)\vert^{2}e^{-iEt}dE\int\vert A(E^{\prime})\vert^{2}e^{iE^{\prime}t}dE^{\prime}.$$ Where $E$ and $E'$ are energies of the continuum. $\vert A(E)\vert$ is the coeficient of the wave function associated with the discrete state. The value of $\vert A(E)\vert^{2}$ was also calculated and given by: $$\Rightarrow\vert A(E)\vert^{2}=\frac{\left|\xi_{E}\right|^{2}}{\left[E-E_{e}-F(E)\right]^{2}+\left|\xi_{E}\right|^{4}\pi^{2}}.$$ So, from this, i can suppose that the bandwith of the continuous spectrum is given by $\Gamma=2\pi\left|\xi_{E}\right|^{2}$ and the decayment time by $\tau=\Gamma^{-1}=\frac{1}{2\pi\left|\xi_{E}\right|^{2}}$.

I'm trying to calculate the transition rate from the discrete state to the continuum, Fermi's Golden Rule.

Can someone help me calculating for this case, please? I'm having some difficulties.

$\endgroup$

1 Answer 1

1
$\begingroup$

Lowest-order Fermi's golden rule for a single state is \begin{equation} \Gamma_f = \frac{2\pi}{\hbar} |\langle f|V|i\rangle|^2 \delta(E_i-E_f) \end{equation} and taking $|i\rangle = |e\rangle$, $|\chi_U\rangle = |f\rangle$, the matrix element of your $V$ is \begin{equation} \langle \chi_U| \int dU' \left (\xi_{U'}|\chi_{U'}\rangle \langle e| +\xi_{U'}^* |e\rangle\langle \chi_{U'}|\right )|e\rangle = \xi_U \,. \end{equation} Since you don't care which continuum state you transition to, the total transition rate, assuming the normalization \begin{equation} 1 = |e\rangle\langle e| + \int dU |\chi_U\rangle \langle \chi_U| \,, \end{equation} gives \begin{equation} \Gamma = \int dU \Gamma_U = \int dU \frac{2\pi}{\hbar} |\xi_U|^2 \delta(U-E) = \frac{2\pi}{\hbar} |\xi_E|^2 \,. \end{equation} Taking units with $\hbar = 1$ gives the same result you found.

$\endgroup$
3
  • $\begingroup$ So the bandwith of the continuous spectrum is the same as the transition rate from the discrete state to the continuum? $\endgroup$
    – bdzh
    Commented Jul 1, 2022 at 3:02
  • $\begingroup$ One way to derive Fermi's golden rule is is to calculate the rate as the time derivative of the transition probability. $\frac{d}{dt} |\langle f|e^{-iHt}|i\rangle|^2$. The probability of staying in the initial state is 1 minus the sum of all transition probabilities. In lowest order this is an exponential whose Fourier transform gives a Lorentzian. So the exponential factor which gives the total transition rate gives the Lorentzian bandwidth. $\endgroup$
    – user200143
    Commented Jul 1, 2022 at 19:05
  • $\begingroup$ Yes, but that $\Lambda$ i found before was the bandwith of the continuous spectrum, and the result you found for the the total transition rate gives the same thing. Is there any relation? $\endgroup$
    – bdzh
    Commented Jul 1, 2022 at 19:30

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.