Perturbation theory, why are the assumptions of the method satisfied?

496 Views Asked by At

I am a undergrad student interested in math taking quantum mechanics.

Yesterday I was introduced to what physicists call perturbation theory, non-degenerate case. According to authors Griffiths, Shankar and Sakuraii, they all state the procedure approximately in the same way. In the following, I will quote Griffiths over the others, for no particular reason. He states that this is the procedure:

The hamiltonian $H = H^0 + \lambda H'$, where $H'$ is the perturbation and the superscript $0$ always identifies the unperturbed quantity. For the moment we'll take $\lambda$ to be a small number, later we will crank it up to $1$ to get the true Hamiltonian. Now, we write $\psi_n$ and $E_n$ as power series in $\lambda$: $$ \psi_n = \psi_n^0 + \lambda \psi_n^1 + \lambda^2\psi_n^2 + \cdots $$ $$ E_n = E_n^0 + \lambda E_n^1 + \lambda^2E_n^2 + \cdots $$

The term $E_n^k$ is called the $k$-th order correction to the $n$-th energy eigenvalue and the term $\psi_n^k$ is called the $k$-th order correction to the $n$-th energy eigenfunction. I dont get any further, because at this point, and don't understand what is going on. So here are some questions to begin with:

  1. Regarding the two power series in $\lambda$, how do we know that the solution to $H$ has this form? It may well have any other form. Why this?

  2. How do we know that the assumptions on which the method is derived still hold when we increase $\lambda$ to $1$? In particular, how do we know that the two power series still converge for $\lambda = 1$? Why did we even have to to let $\lambda$ be small, when we might as well have let it be $1$ anyway?

1

There are 1 best solutions below

0
On

For some operators $H^0, H'$ it can be shown that the eigenvalues and eigenvectors are analytic functions of $\lambda$ so we can express them in a Taylor series. In other cases, the series may not converge, but for small $ \lambda $ the first few terms may still be a good approximation for physics purposes. This is a whole branch of mathematical physics. The classic text is Kato's Perturbation Theory for Linear Operators (This link downloads the entire book from University of Edinburgh).