Skip to content

Evolution of two-states system

(1) Stationary(time-independent)

Assume that \( \left|\psi_1 \right\rangle \) and \( \left|\psi_2 \right\rangle \) represent state 1 and state 2, respectively, and \( E_1 \) and \( E_2 \) represent the energy of state 1 and state 2, respectively.

If two electronic states are independent(without the coupling),then:

\[ \hat{H}_0\left|\psi_1 \right \rangle = E_1\left|\psi_1 \right \rangle \]
\[ \hat{H}_0\left|\psi_2 \right \rangle = E_2\left|\psi_2 \right \rangle \]

where $ H_0 $ is the Hamitonian of system, which without coupling.

If two electronic states are coupling,then the hamitonian changes into:

\[ \hat{H} = \hat{H}_0 + \hat{W} \]

In this condition, the time-independent Schrödinger equation can be written as:

$$ \hat{H} \left|\psi \right \rangle=E\left|\psi \right \rangle $$ We can linearly combine \( \left|\psi_1 \right\rangle \) and \( \left|\psi_2 \right\rangle \) to construct \( \left|\psi \right\rangle \):

\[ \left|\psi \right\rangle = c_1 \left|\psi_1 \right\rangle + c_2 \left|\psi_2 \right\rangle \]

where \( c_1 \) and \( c_2 \) are complex coefficients.

\[ \left|\psi \right \rangle = c_1 \left|\psi_1 \right \rangle + c_2\left|\psi_2 \right \rangle \]

Multiply the both sides of time-independent Schrödinger equation left by\(\left\langle\psi_1\right|\) and \(\left\langle\psi_2\right|\) respectively, then we obtain:

\[ \left\langle\psi_1\right|\hat{H} [c_1 \left|\psi_1 \right \rangle + c_2\left|\psi_2 \right \rangle]=\left\langle\psi_1\right|E[c_1 \left|\psi_1 \right \rangle + c_2\left|\psi_2 \right \rangle] \]
\[ \left\langle\psi_2\right|\hat{H} [c_1 \left|\psi_1 \right \rangle + c_2\left|\psi_2 \right \rangle]=\left\langle\psi_2\right|E[c_1 \left|\psi_1 \right \rangle + c_2\left|\psi_2 \right \rangle] \]

or simplely:

\[ c_1 \hat{H}_{11} + c_2 \hat{H}_{12} = c_1 E \]
\[ c_1 \hat{H}_{21} + c_2 \hat{H}_{22} = c_2 E \]

where \(H_{ij} = \left\langle\psi_i\right|\hat{H}\left|\psi_j \right \rangle\), and \(\left\langle\psi_i | \psi_j \right \rangle = \delta_{ij}\) is used.

In matrix representation, the equation above change into the form as follow:

\[\begin{align} \begin{bmatrix} H_{11} & H_{12} \\ H_{21} & H_{22} \end{bmatrix} \begin{bmatrix} c_1 \\ c_2 \end{bmatrix} = \begin{bmatrix} E_{11} & 0 \\ 0 & E_{22} \end{bmatrix} \begin{bmatrix} c_1 \\ c_2 \end{bmatrix} \end{align} \]

Now we introduce a common method for solving the equation above.

At first, we divide the total hamitonian matrix H into two parts:

\[ \begin{align} H &= \begin{bmatrix} \dfrac{1}{2}(H_{11} + H_{12}) & 0 \\ 0 & \dfrac{1}{2}(H_{11} + H_{12}) \end{bmatrix} \\ &\quad + \begin{bmatrix} \dfrac{1}{2}(H_{11} - H_{12}) & H_{12} \\ H_{21} & -\dfrac{1}{2}(H_{11} - H_{12}) \end{bmatrix} \end{align} \]

Then the the total hamitonian matrix H can be written as:

\[ H = \dfrac{1}{2}(H_{11} + H_{12}) I + \dfrac{1}{2}(H_{11} - H_{12}) K \]

where I is identity matirx, K is a matrix that:

\[ K = \begin{bmatrix} 1 & \dfrac{2H_{12}}{H_{11} - H_{12}} \\ \dfrac{2H_{21}}{H_{11} - H_{12}} & -1 \end{bmatrix} \]

We set:

\[ tan \theta = \dfrac{2|H_{12}|}{H_{11} - H_{12}} \]
\[ |H_{12}| = |H_{21}| \]
\[ H_{12} = |H_{12}| e^{-i\varphi } \]
\[ H_{21} = |H_{21}| e^{i\varphi } \]

Then matrix K becomes:

\[ K = \begin{bmatrix} 1 & tan \theta e^{-i\varphi} \\ tan \theta e^{i\varphi} & -1 \end{bmatrix} \]

Because matirx H satisfies \(\hat{H} \left|\psi \right \rangle=E\left|\psi \right \rangle\), then matirx K also satisfies:

\[ \hat{K} \left|\psi \right \rangle=k\left|\psi \right \rangle \]

or in matrix representation:

\[ \begin{aligned} \begin{bmatrix} 1 - k & \tan \theta \, e^{-i\varphi} \\ \tan \theta \, e^{i\varphi} & -1 - k \end{bmatrix} \begin{bmatrix} c_1 \\ c_2 \end{bmatrix} &= 0 \end{aligned} \]

If there has a non-trivial solution of the equation above, it must satisfies:

\[ \begin{aligned} \begin{bmatrix} 1 - k & \tan \theta \, e^{-i\varphi} \\ \tan \theta \, e^{i\varphi} & -1 - k \end{bmatrix} \begin{bmatrix} c_1 \\ c_2 \end{bmatrix} &= 0 \end{aligned} \]

Solve for:

\[ k_+ = \dfrac{1}{cos \theta} \]
\[ k_- = -\dfrac{1}{cos \theta} \]

So we obtain the expression of E:

\[ E_+ = \dfrac{1}{2}(H_{11} + H_{12}) + \sqrt{\dfrac{1}{4} * (H_{11} - H_{22})^2 + |H_{12}|^2} \]
\[ E_- = \dfrac{1}{2}(H_{11} + H_{12}) - \sqrt{\dfrac{1}{4} * (H_{11} - H_{22})^2 + |H_{12}|^2} \]

Finally, we can solve the expression of constant \(c_1\) and \(c_2\). With the example of \(k_+\), we obtain:

\[ (1-k)c_1 + tan \theta e^{-i\varphi}c_2 = 0 \\tan \theta e^{i\varphi}c_1 -(1+k)c_2 = 0 \]

Solve for:

\[ c_1 = cos\frac{\theta}{2} * e^{-i \dfrac{\varphi}{2}}\\ c_2 = sin\frac{\theta}{2} * e^{i \dfrac{\varphi}{2}}\\ \]

That is the constant of \(\left|\psi_+ \right \rangle\), whose eigenvalue is \(E_+\).

Similarly, we can know the constant of \(\left|\psi_- \right \rangle\), whose eigenvalue is \(E_-\):

\[ c_1 = -sin\frac{\theta}{2} * e^{-i \dfrac{\varphi}{2}}\\ c_2 = cos\frac{\theta}{2} * e^{i \dfrac{\varphi}{2}}\\ \]

where normalization condition \(|c_1|^2 + |c_2|^2 = 1\) is used.

(2) Evolution of Two-States(time-dependent)

The evolution of two-states system can be descibed by time-dependent Schrödinger equation:

\[ i\hbar \dfrac{\partial }{\partial t}\left|\Psi(t) \right \rangle = \hat{H}\left|\Psi(t) \right \rangle \]

\(\left|\Psi(t) \right \rangle\) can be written as:

\[ \left|\Psi(t) \right \rangle = a_1(t)\left|\psi_1 \right \rangle + a_2(t)\left|\psi_2 \right \rangle \]

Similarly, we multiply the both sides of time-dependent Schrödinger equation left by\(\left\langle\psi_1\right|\) and \(\left\langle\psi_2\right|\) respectively, then we obtain:

\[ \left\langle\psi_1\right|i\hbar \dfrac{\partial }{\partial t}[a_1(t)\left|\psi_1 \right \rangle + a_2(t)\left|\psi_2 \right \rangle] = \left\langle\psi_1\right|\hat{H}[a_1(t)\left|\psi_1 \right \rangle + a_2(t)\left|\psi_2 \right \rangle] \]
\[ \left\langle\psi_2\right|i\hbar \dfrac{\partial }{\partial t}[a_1(t)\left|\psi_1 \right \rangle + a_2(t)\left|\psi_2 \right \rangle] = \left\langle\psi_2\right|\hat{H}[a_1(t)\left|\psi_1 \right \rangle + a_2(t)\left|\psi_2 \right \rangle] \]

Then:

\[ i\hbar \dfrac{\partial }{\partial t}a_1(t) = a_1(t)H_{11} + a_2(t)H_{12} \]
\[ i\hbar \dfrac{\partial }{\partial t}a_2(t) = a_1(t)H_{21} + a_2(t)H_{22} \]

We can choose:

\[ a_i(t) = exp(\frac{-iE_it}{\hbar}) \]

Then the total wavefunction becomes:

\[ \left|\Psi(t) \right \rangle = exp(\frac{-iE_+t}{\hbar})\left|\psi_+ \right \rangle + exp(\frac{-iE_-t}{\hbar})\left|\psi_- \right \rangle \]

We choose \(\left|\psi_1 \right \rangle\) as initial state, then:

\[ \left|\Psi(0) \right \rangle = \left|\psi_1 \right \rangle \]

and \(\left|\psi_1 \right \rangle\) is given by:

\[ \left|\psi_1 \right \rangle = [cos\dfrac{\theta}{2} \left|\psi_+ \right \rangle - sin\dfrac{\theta}{2} \left|\psi_- \right \rangle] * exp(i \dfrac{\varphi}{2}) \]

The Probability of state 2 is defined as:

\[ P_2 = |\left\langle\psi_2|\Psi(t)\right \rangle|^2 \]

and \(a_2(t)\):

\[ \left\langle\psi_2|\Psi(t)\right \rangle = exp(i \varphi)*sin\dfrac{\theta}{2}cos\dfrac{\theta}{2}*[exp(\frac{-iE_+t}{\hbar}) - exp(\frac{-iE_-t}{\hbar})] \]

Thus:

\[ P_2 = \dfrac{1}{4}sin\theta * [2-exp(-i\frac{E_+-E_-}{\hbar}t) - exp(-i\frac{E_--E_+}{\hbar}t)] \]

by Euler's formula:

\[ e^{i\theta} = isin\theta + cos\theta \]

\(P_2\) changes into:

\[ P_2 = sin^2\theta * sin^2\frac{E_+-E_-}{2\hbar}t \]

Finally, \(P_2\) becomes:

\[ P_2 = \frac{4|H_{12}|^2}{4|H_{12}|^2 + (H_{11} - H_{22})^2} * sin^2\sqrt{\frac{(H_{11} - H_{22})^2 + 4|H_{12}|^2}{2\hbar}t} \]

That is so_called Rabi equaition.

(3) Rabi frequency

According to:

\[ P_2 = \frac{4|H_{12}|^2}{4|H_{12}|^2 + (H_{11} - H_{22})^2} * sin^2\sqrt{\frac{(H_{11} - H_{22})^2 + 4|H_{12}|^2}{2\hbar}t} \]

we obtain the f, or so-called Rabi frequency:

\[ f = \sqrt{\frac{(H_{11} - H_{22})^2 + 4|H_{12}|^2}{2\hbar}} \]

The period of \(P_2\) satisfies:

\[ \frac{(H_{11} - H_{22})^2 + 4|H_{12}|^2}{2\hbar}t = (2k+1)\pi \]
\[ t = \frac{2(2k+1)\pi \hbar}{(H_{11} - H_{22})^2 + 4|H_{12}|^2} \]

Comments