I am trying to understand the solution to the infinite square well centered at zero in Principles of Quantum Mechanics by Shankar. Here is how it goes:

Inside the well (region II - Outside left is I and outside right is III) the solution is that of the free particle:


$k$ has the obvious value. Applying the boundary conditions, we require that

$$\psi_{I}\left(\frac L2\right)=\psi_{II}\left(\frac L2\right)=0$$

$$\psi_{III}\left(-\frac L2\right)=\psi_{II}\left(-\frac L2\right)=0$$

Setting the determinant of the augmented matrix corresponding to this system to zero (otherwise only the trivial solution holds):






$$B=-e^{in\pi}A = -A(-1)^n$$

We find that there are two families of solutions; those for odd n and those for even n. For the odd case:

$$\psi(x)=Ae^{ikx}+Ae^{ikx}=2A\cos\left(\frac{n\pi x}{L}\right)$$

But for the even case:

$$\psi(x)=Ae^{ikx}-Ae^{ikx}=2Ai\sin\left(\frac{n\pi x}{L}\right)$$

These intermediary steps are not gone over in the book. I am wondering if it is valid to multiply the even solutions by $i$. It is probably of fundamental importance. I know I can multiply certain things without messing up the physics but will this? Anyway, if I have made an error somewhere please point it out. I am a beginner.

Addendum: I suppose all that physically matters is whether the functions are eigenfunctions of the hamiltonian representing the system and obey the boundary and normalization conditions.

  • 1
    $\begingroup$ What is $A$ here? $\endgroup$
    – jacob1729
    Aug 3, 2020 at 0:19
  • $\begingroup$ A and B are just arbitrary coefficients determined by boundary conditions and normalization.@jacob1729 $\endgroup$ Aug 3, 2020 at 0:58
  • $\begingroup$ What happens if you apply the substitution $A \rightarrow iA$? $\endgroup$
    – Sandejo
    Aug 3, 2020 at 1:53
  • $\begingroup$ Related: physics.stackexchange.com/q/77894/2451 and links therein. $\endgroup$
    – Qmechanic
    Aug 3, 2020 at 3:59
  • $\begingroup$ Also worth keeping in mind that the probability density function ψψ* is what we interpret physically and that will *always be real. $\endgroup$ Aug 3, 2020 at 22:52

1 Answer 1


Yes, you can multiply the even solutions by $i$. In general, you can multiply a wavefunction by any constant number of unit modulus (any number $c$ such that $|c|^2=1$) without changing the physical properties of the wavefunction. Multiplication by a number of unit modulus is called changing the phase of the wavefunction. The only time the phase of a wavefunction is relevant is when you are comparing the phases of two wavefunctions. In this case, a relative phase difference can cause interference. You can still multiply by a number of unit modulus, but you have to be consistent and multiply all wavefunctions by the same factor so that the relative phases remain the same.

  • $\begingroup$ Thank you. This is exactly what I was looking for. Given what I already knew I could have deduced that this is possible as e^ipi/2 is i and I knew I could do that. but the underlying principle is that it doesn't change the physics. $\endgroup$ Aug 4, 2020 at 1:50

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.