In the document https://www.hep.anl.gov/czachos/aaa.pdf, they derive the Wigner functions, $f_n$ for the harmonic oscillator. However, I have some problems understanding some of the steps. On page 37 they write the equation: $$\tag{1} \left(\left(x+\frac{i \hbar}{2} \partial_{p}\right)^{2}+\left(p-\frac{i \hbar}{2} \partial_{x}\right)^{2}-2 E\right) f(x, p)=0 $$ Defining $z\equiv \frac{4}{\hbar}H= \frac{2}{\hbar}\left(x^{2}+p^{2}\right)$ they say that the real part of eq. $(1)$ can be written: $$\tag{2} \left(\frac{z}{4}-z \partial_{z}^{2}-\partial_{z}-\frac{E}{\hbar}\right) f(z)=0$$ and by setting $f(z)=\exp (-z / 2) L(z)$ we get Laguerre's equation: $$\tag{3} \left(z \partial_{z}^{2}+(1-z) \partial_{z}+\frac{E}{\hbar}-\frac{1}{2}\right) L(z)=0$$ Solved by Laguerre polynomials: $$\tag{4} L_n=\sum_{k=0}^{n}\left(\begin{array}{l} n \\ k \end{array}\right) \frac{(-z)^{k}}{k !}$$ and the Wigner functions are then: $$\tag{5} f_{n}=\frac{(-1)^{n}}{\pi \hbar} e^{-z/2} L_{n}\left(z\right)$$

There are 3 things that I do not get:

  1. We can write the real part of eq. $(1)$ as $\left(x^{2}-\frac{\hbar^{2}}{4} \partial_{p}^{2}+p^{2}-\frac{\hbar^{2}}{4} \partial_{x}^{2}-2 E\right) f=0$. How is eq. $(2)$ obtained from this? (I don't get the $z \partial_{z}^{2}-\partial_{z}$ term)
  2. How is eq. $(3)$ obtained from eq. $(2)$?
  3. where does the $\frac{(-1)^{n}}{\pi \hbar}$ pre-factor in eq. $(5)$ come from?

Product support.

  1. Recall, crucially, f was shown to be a function of z only, f(z), so, acting on it, $$\partial_x = \frac{\partial z}{\partial x } \partial_z \leadsto \\ \partial_x^2 = \left (\partial_x \frac{\partial z}{\partial x } \right )\partial_z + \left ( \frac{\partial z}{\partial x }\right )^2 \partial_z^2 ~ , $$
    and similarly for y, so that $\partial_x^2 + \partial_y^2 = 8(\partial_z + z\partial_z^2)/\hbar$.

  2. You've been there before with the integrating factor of the oscillator equation to Hermite's in Hilbert space. Analogously, $$ \left(\frac{z}{4}-z \partial_{z}^{2}-\partial_{z}-\frac{E}{\hbar}\right) ( e^{-z / 2}~L(z) )=0, ~~~~~\leadsto\\ e^{-z / 2} \left(z \partial_{z}^{2}+(1-z) \partial_{z}+\frac{E}{\hbar}-\frac{1}{2}\right) L(z)=0.$$ But the exponential can never be zero, and can thus be dropped.

  3. Any multiple of these polynomials will solve this linear equation. However, it is practical/convenient to simplify their Rodrigues formula $$L_n(z)=\frac{e^z}{n!}\partial_z^n \left(e^{-z} z^n\right) =\frac{1}{n!} \left( \partial_z -1 \right)^n z^n,$$ and Sheffer sequence recursive formula, $$ \tag{7} \partial_z L_n = \left ( \partial_z - 1 \right ) L_{n-1},$$ generating function, etc, as you probably learned from your Hydrogen atom. So they are all unity at the origin. Recall, from the text, these are all ingredients of Wigner functions f normalized to 1, whence the common normalization; trivially checkable for n=0, $$ 1=\int\!\! dxdp ~f(z)= \frac{\pi \hbar }{2}\int_0^\infty\!\! dz ~e^{-z/2} L_n (z)\frac{(-)^n}{\pi \hbar} ~. $$ But from n=0 and the Sheffer sequence recursion (7), you may readily check the normalization for n=1, through integration by parts to be a mere change of sign. So, recursively, for all n, you show the alternating sign normalizations.

  • $\begingroup$ I get point 2 and 3 now. But I am still a bit lost on point 1. Is the following argument correct: $\partial f=\partial z \frac{\partial f}{\partial z} \Rightarrow \frac{\partial f}{\partial x}=\frac{\partial z}{\partial x} \frac{\partial f}{\partial z}$? And I do not see how you get the expression for $\partial_x^2$. As far as I know, the square of the product of two operators is the product of the squares, so I would assume $\partial_{x}=\frac{\partial z}{\partial x} \partial_{z} \Rightarrow \partial_{x}^{2}=\left(\frac{\partial z}{\partial x}\right)^{2} \partial_{z}^{2}$ $\endgroup$ – Physics101 Apr 24 at 18:00
  • $\begingroup$ Notice in my first equation in the comment I use notation $\frac{\partial}{\partial x}=\partial_{x}$, so my denominator is not the same as the one you wrote above. $\endgroup$ – Physics101 Apr 24 at 18:04
  • 2
    $\begingroup$ @Physics101 You have to apply the product rule, check this math.stackexchange.com/questions/1313764/… and apply it here. $$\frac{\partial}{\partial x}\frac{\partial}{\partial x} f= \frac{\partial}{\partial x}\left(\frac{\partial z}{\partial x}\frac{\partial f}{\partial z}\right)\\ =\left(\frac{\partial^2 z}{\partial x^2}\frac{\partial f}{\partial z}\right) +\left(\frac{\partial z}{\partial x}\right)^2 \left(\frac{\partial^2 f}{\partial z^2}\right)\\ $$ $\endgroup$ – Hans Wurst Apr 24 at 18:37
  • 2
    $\begingroup$ @Physics101 $$\frac{\partial^2 }{\partial x^2}\rightarrow \frac{\partial^2 z }{\partial x^2}\frac{\partial}{\partial z} + \left(\frac{\partial z}{\partial x}\right)^2 \frac{\partial^2 }{\partial z^2} $$ $\endgroup$ – Hans Wurst Apr 24 at 18:38
  • $\begingroup$ Thanks @Hans , I couldn't have said it better. I corrected my TeX-typos $\endgroup$ – Cosmas Zachos Apr 24 at 18:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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