Question 11.36: The driven harmonic oscillator. Suppose the one-dimensional ...

The driven harmonic oscillator. Suppose the one-dimensional harmonic oscillator (mass m, frequency ω) is subjected to a driving force of the form F(t) = m ω² f(t) , where f(t) is some specified function. (I have factored out m ω² for notational convenience; f(t) has the dimensions of length.) The Hamiltonian is

H(t)=-\frac{\hbar^{2}}{2 m} \frac{\partial^{2}}{\partial x^{2}}+\frac{1}{2} m \omega^{2} x^{2}-m \omega^{2} x f(t)             (11.140).

Assume that the force was first turned on at time t=0: f(t)=0 \text { for } t \leq 0 .

This system can be solved exactly, both in classical mechanics and in quantum mechanics.

(a) Determine the classical position of the oscillator, assuming it started from rest at the origin \left(x_{c}(0)=\dot{x}_{c}(0)=0\right) . Answer:

x_{c}(t)=\omega \int_{0}^{t} f\left(t^{\prime}\right) \sin \left[\omega\left(t-t^{\prime}\right)\right] d t^{\prime} .                (11.141).

(b) Show that the solution to the (time-dependent) Schrödinger equation for this oscillator, assuming it started out in the nth state of the undriven oscillator \left(\Psi(x, 0)=\psi_{n}(x) \text { where } \psi_{n}(x)\right. is given by Equation 2.62), can be written as

\psi_{n}(x)=A_{n}\left(\hat{a}_{+}\right)^{n} \psi_{0}(x), \quad \text { with } \quad E_{n}=\left(n+\frac{1}{2}\right) \hbar \omega                (2.62).

\Psi(x, t)=\psi_{n}\left(x-x_{c}\right) e^{\frac{i}{\hbar}\left[-\left(n+\frac{1}{2}\right) \hbar \omega t+m \dot{x}_{c}\left(x-\frac{x_{c}}{2}\right)+\frac{m \omega^{2}}{2} \int_{0}^{t} f\left(t^{\prime}\right) x_{c}\left(t^{\prime}\right) d t^{\prime}\right]}           (11.142).

(c) Show that the eigenfunctions and eigenvalues of H(t) are

\psi_{n}(x, t)=\psi_{n}(x-f) ; \quad E_{n}(t)=\left(n+\frac{1}{2}\right) \hbar \omega-\frac{1}{2} m \omega^{2} f^{2}                 (11.143).

(d) Show that in the adiabatic approximation the classical position (Equation 11.141) reduces to x_{c}(t) \approx f(t) . State the precise criterion for adiabaticity, in this context, as a constraint on the time derivative of f. Hint: Write \sin \left[\omega\left(t-t^{\prime}\right)\right] \text { as }(1 / \omega)\left(d / d t^{\prime}\right) \cos \left[\omega\left(t-t^{\prime}\right)\right] and use integration by parts.

(e) Confirm the adiabatic theorem for this example, by using the results in (c) and (d) to show that

\Psi(x, t) \approx \psi_{n}(x, t) e^{i \theta_{n}(t)} e^{i \gamma_{n}(t)}           (11.144).

Check that the dynamic phase has the correct form (Equation 11.92). Is the geometric phase what you would expect?

e^{i \theta_{n}(t)}, \quad \text { where } \theta_{n}(t) \equiv-\frac{1}{\hbar} \int_{0}^{t} E_{n}\left(t^{\prime}\right) d t^{\prime}               (11.92).

The Blue Check Mark means that this solution has been answered and checked by an expert. This guarantees that the final answer is accurate.
Learn more on how we answer questions.

(a) Check the answer given: x_{c}=\omega \int_{0}^{t} f\left(t^{\prime}\right) \sin \left[\omega\left(t-t^{\prime}\right)\right] d t^{\prime} \Longrightarrow x_{c}(0)=0 .

\dot{x}_{c}=\omega f(t) \sin [\omega(t-t)]+\omega^{2} \int_{0}^{t} f\left(t^{\prime}\right) \cos \left[\omega\left(t-t^{\prime}\right)\right] d t^{\prime}=\omega^{2} \int_{0}^{t} f\left(t^{\prime}\right) \cos \left[\omega\left(t-t^{\prime}\right)\right] d t^{\prime} \Rightarrow \dot{x}_{c}(0)=0 .

\ddot{x}_{c}=\omega^{2} f(t) \cos [\omega(t-t)]-\omega^{3} \int_{0}^{t} f\left(t^{\prime}\right) \sin \left[\omega\left(t-t^{\prime}\right)\right] d t^{\prime}=\omega^{2} f(t)-\omega^{2} x_{c} .

Now the classical equation of motion is m\left(d^{2} x / d t^{2}\right)=-m \omega^{2} x+m \omega^{2} f . For the proposed solution, m\left(d^{2} x_{c} / d t^{2}\right)=m \omega^{2} f-m \omega^{2} x_{c} , so it does satisfy the equation of motion, with the appropriate boundary conditions.

(b) Let  z \equiv x-x_{c} \quad\left(\text { so } \psi_{n}\left(x-x_{c}\right)=\psi_{n}(z)\right. , and z depends on t as well as x).

\frac{\partial \Psi}{\partial t}=\frac{d \psi_{n}}{d z}\left(-\dot{x}_{c}\right) e^{i\{\}}+\psi_{n} e^{i\{\}} \frac{i}{\hbar}\left[-\left(n+\frac{1}{2}\right) \hbar \omega+m \ddot{x}_{c}\left(x-\frac{x_{c}}{2}\right)-\frac{m}{2} \dot{x}_{c}^{2}+\frac{m \omega^{2}}{2} f x_{c}\right] .

[]=-\left(n+\frac{1}{2}\right) \hbar \omega+\frac{m \omega^{2}}{2}\left[2 x\left(f-x_{c}\right)+x_{c}^{2}-\frac{\dot{x}_{c}^{2}}{\omega^{2}}\right] .

\frac{\partial \Psi}{\partial t}=-\dot{x}_{c} \frac{d \psi_{n}}{d z} e^{i\{\}}+i \Psi\left\{-\left(n+\frac{1}{2}\right) \omega+\frac{m \omega^{2}}{2 \hbar}\left[2 x\left(f-x_{c}\right)+x_{c}^{2}-\frac{\dot{x}_{c}^{2}}{\omega^{2}}\right]\right\} .

\frac{\partial \Psi}{\partial x}=\frac{d \psi_{n}}{d z} e^{i\{\}}+\psi_{n} e^{i\{\}} \frac{i}{\hbar}\left(m \dot{x}_{c}\right) ; \quad \frac{\partial^{2} \Psi}{\partial x^{2}}=\frac{d^{2} \psi_{n}}{d z^{2}} e^{i\{\}}+2 \frac{d \psi_{n}}{d z} e^{i\{\}} \frac{i}{\hbar}\left(m \dot{x}_{c}\right)-\left(\frac{m \dot{x}_{c}}{\hbar}\right)^{2} \psi_{n} e^{i\{\}} .

H \Psi=-\frac{\hbar^{2}}{2 m} \frac{\partial^{2} \Psi}{\partial x^{2}}+\frac{1}{2} m \omega^{2} x^{2} \Psi-m \omega^{2} f x \Psi .

=-\frac{\hbar^{2}}{2 m} \frac{d^{2} \psi_{n}}{d z^{2}} e^{i\{\}}-\frac{\hbar^{2}}{2 m} 2 \frac{d \psi_{n}}{d z} e^{i\{\}} \frac{i m \dot{x}_{c}}{\hbar}+\frac{\hbar^{2}}{2 m}\left(\frac{m \dot{x}_{c}}{\hbar}\right)^{2} \Psi+\frac{1}{2} m \omega^{2} x^{2} \Psi-m \omega^{2} f x \Psi .

But -\frac{\hbar^{2}}{2 m} \frac{d^{2} \psi_{n}}{d z^{2}}+\frac{1}{2} m \omega^{2} z^{2} \psi_{n}=\left(n+\frac{1}{2}\right) \hbar \omega \psi_{n} ,  so

H \Psi=\cancel {\left(n+\frac{1}{2}\right) \hbar \omega \Psi}-\frac{1}{2} m \omega^{2} z^{2} \Psi-\cancel {i \hbar \dot{x}_{c} \frac{d \psi n}{d z} e^{i\{\}}}+\frac{m}{2} \dot{x}_{c}^{2} \Psi+\frac{1}{2} m \omega^{2} x^{2} \Psi-m \omega^{2} f x \Psi .

\stackrel{?}{=} i \hbar \frac{\partial \Psi}{\partial t}=\cancel {-i \hbar \dot{x}_{c} \frac{d  ψ_{n}}{d z} e^{i\{\}}}-\hbar \Psi\left[-\cancel {\left(n+\frac{1}{2}\right) \omega}+\frac{m \omega^{2}}{2 \hbar}\left(2 x f-2 x x_{c}+x_{c}^{2}-\frac{1}{\omega^{2}} \dot{x}_{c}^{2}\right)\right] .

-\frac{1}{2} m \omega^{2} z^{2}+\cancel {\frac{m}{ 2 } {x_{c}}^{2}}+\frac{1}{2} m \omega^{2} x^{2}-\cancel {m ω^2 f x} \stackrel{?}{=}-\frac{m \omega^{2}}{2}\left(\cancel {2 x f}-2 x x_{c}+x_{c}^{2}-\cancel {\frac{1}{ω^{2}} \dot{x}_{c}^{2}}\right) .

z^{2}-x^{2} \stackrel{?}{=}-2 x x_{c}+x_{c}^{2} ; \quad z^{2} \stackrel{?}{=}\left(x^{2}-2 x x_{c}+x_{c}^{2}\right)=\left(x-x_{c}\right)^{2} .

(c)

Eq. 11.140 ⇒ H=-\frac{\hbar^{2}}{2 m} \frac{\partial^{2}}{\partial x^{2}}+\frac{1}{2} m \omega^{2}\left(x^{2}-2 x f+f^{2}\right)-\frac{1}{2} m \omega^{2} f^{2} . \quad \text { Shift origin: } \quad u \equiv x-f .

H=\left[-\frac{\hbar^{2}}{2 m} \frac{\partial^{2}}{\partial u^{2}}+\frac{1}{2} m \omega^{2} u^{2}\right]-\left[\frac{1}{2} m \omega^{2} f^{2}\right] .

The first term is a simple harmonic oscillator in the variable u; the second is a constant (with respect to position). So the eigenfunctions are \psi_{n}(u) , and the eigenvalues are harmonic oscillator ones, ( n + \left.\frac{1}{2}\right) \hbar \omega , E_{n}=\left(n+\frac{1}{2}\right) \hbar \omega-\frac{1}{2} m \omega^{2} f^{2} .

(d) Note that \sin \left[\omega\left(t-t^{\prime}\right)\right]=\frac{1}{\omega} \frac{d}{d t^{\prime}} \cos \left[\omega\left(t-t^{\prime}\right)\right], \quad \text { so } \quad x_{c}(t)=\int_{0}^{t} f\left(t^{\prime}\right) \frac{d}{d t^{\prime}} \cos \left[\omega\left(t-t^{\prime}\right)\right] d t^{\prime}, ,   or

x_{c}(t)=\left.f\left(t^{\prime}\right) \cos \left[\omega\left(t-t^{\prime}\right)\right]\right|_{0} ^{t}-\int_{0}^{t}\left(\frac{d f}{d t^{\prime}}\right) \cos \left[\omega\left(t-t^{\prime}\right)\right] d t^{\prime}=f(t)-\int_{0}^{t}\left(\frac{d f}{d t^{\prime}}\right) \cos \left[\omega\left(t-t^{\prime}\right)\right] d t^{\prime} .

(since f(0) = 0). Now, for an adiabatic process we want df/dt very small; specifically: \frac{d f}{d t^{\prime}} \ll \omega f(t) .

\left(0<t^{\prime} \leq t\right) .  Then the integral is negligible compared to f(t), and we have x_{c}(t) \approx f(t) . (Physically, this says that if you pull on the spring very gently, no fancy oscillations will occur; the mass just moves along as though attached to a string of fixed length.)

\text { (e) Put } x_{c} \approx f \text { into Eq. 11.142, using Eq. 11.143: }

\Psi(x, t)=\psi_{n}(x, t) e^{\frac{1}{\hbar}\left[-\left(n+\frac{1}{2}\right) \hbar \omega t+m \dot{f}(x-f / 2)+\frac{m \omega^{2}}{2} \int_{0}^{t} f^{2}\left(t^{\prime}\right) d t^{\prime}\right]} .

The dynamic phase (Eq. 11.92) is

e^{i \theta_{n}(t)}, \quad \text { where } \theta_{n}(t) \equiv-\frac{1}{\hbar} \int_{0}^{t} E_{n}\left(t^{\prime}\right) d t^{\prime}              (11.92).

\theta_{n}(t)=-\frac{1}{\hbar} \int_{0}^{t} E_{n}\left(t^{\prime}\right) d t^{\prime}=-\left(n+\frac{1}{2}\right) \omega t+\frac{m \omega^{2}}{2 \hbar} \int_{0}^{t} f^{2}\left(t^{\prime}\right) d t^{\prime}, \quad \text { so } \quad \Psi(x, t)=\psi_{n}(x, t) e^{i \theta_{n}(t)} e^{i \gamma_{n}(t)} .

confirming Eq. 11.144, with the geometric phase given (ostensibly) by \gamma_{n}(t)=\frac{m}{\hbar} \dot{f}(x-f / 2) . But the eigenfunctions here are real, and this means the geometric phase should be zero. The point is that (in the adiabatic approximation) \dot{f} is extremely small (see above), and hence in this limit \frac{m}{\hbar} \dot{f}(x-f / 2) \approx 0 (at least, in the only region of x where \psi_{n}(x, t) is nonzero).

Related Answered Questions