# Functions with operator valued arguments acting on eigenstates

by qtm912
Tags: acting, arguments, eigenstates, functions, operator, valued
 P: 29 This question concerns the outcome when operator valued functions act on an energy eigenstate. Given an eigenstate at t =0, say |Ej > , I have seen or inferred in some of the literature that the following applies : exp(-iHt/h) |Ej > = exp(- iEj t/h) |Ej > Where h = h-bar Ej is energy eigenstate j H is the Hamiltonian I am unclear in particular why we can say that if we apply a function with an operator as the argument to this energy eigenstate, it will return the function with the eigenvalue as the argument times the energy eigenstate. (assuming I have inferred what is in the notes correctly) Would this then be a general result for all functions or do only certain functions satisfy this relationship. (So for example would it be true for sin(H) instead of exp(-iHt/h) )?
 P: 1,937 Functions of operators are defined from the power series for those functions. So, because H|E>=E|E>, and H^2|E>=H(E|E>)=E(H|E>)=E^2|E>, etc., as long as these power series converge, then you will have f(H)|E>=f(E)|E>.
 P: 29 Thank you matterwave for the clear explanation
P: 640

## Functions with operator valued arguments acting on eigenstates

 Quote by Matterwave Functions of operators are defined from the power series for those functions. So, because H|E>=E|E>, and H^2|E>=H(E|E>)=E(H|E>)=E^2|E>, etc., as long as these power series converge, then you will have f(H)|E>=f(E)|E>.
Thus, if the operator has eigenvalues that exceed the convergence radius of the expansion, problems will arise.
 P: 29 Yes, makes sense, thanks for pointing it out torquil.

 Related Discussions Advanced Physics Homework 1 Advanced Physics Homework 6 Quantum Physics 0 Quantum Physics 2 General Math 2