Normalising the wavefunction - two answers

AI Thread Summary
Normalization of the wavefunction can yield multiple forms due to the arbitrary phase factor ##e^{i\phi}##, leading to an infinite number of acceptable solutions. The general form for normalization is given by ##A = \frac{e^{i\phi}}{\sqrt{L}}##, where ##\phi## is a real constant. While some may suggest only two solutions exist when fixing ##\phi##, this overlooks the fact that any phase shift corresponds to a different value of ##\phi##, thus maintaining an infinite set of solutions. The distinction between the arbitrary phase of the normalization constant and the spatially-dependent phase in wavefunctions is crucial and should not be conflated. In practice, it is common to omit the phase term for simplicity, allowing for a normalized wavefunction with ##\phi = 0##.
laser
Messages
104
Reaction score
17
Homework Statement
Notes only takes positive answer
Relevant Equations
See description
Hi, when normalising the wavefunction
Screenshot_2.png


I get two answers. Is this correct? My notes only has 1/sqrt(L) = A.

Screenshot_4.png
 
Physics news on Phys.org
Normalization of wavefunction is always up to an overall phase factor ##e^{i\phi}##. You can always multiply by such a factor without affecting normalization so not only are there two possibilities, there us an infinite number of possibilities.

Edit: To be more explicit: The condition ##|A|^2 L = 1## is solved by any ##A## on the form
$$
A = \frac{e^{i\phi}}{\sqrt L}
$$
with ##\phi## an arbitrary real constant.
 
  • Like
Likes MatinSAR, hutchphd, DeBangis21 and 1 other person
Orodruin said:
$$
A = \frac{e^{i\phi}}{\sqrt L}
$$
with ##\phi## an arbitrary real constant.

If ##\phi## is fixed, am I correct in saying that there are only two solutions in the form ##\frac{Be^{i\phi}}{\sqrt L}##, which is B = 1 and B = -1?
 
laser said:
If ##\phi## is fixed, am I correct in saying that there are only two solutions in the form ##\frac{Be^{i\phi}}{\sqrt L}##, which is B = 1 and B = -1?
No. There are an infinite number of solutions, which correspond to shifting ##\phi##. Once you have the form ##e^{i\phi}## that is the general solution already. There is no point in introducing yet another multiplicative constant. Multiplying by ##-1## specifically just shifts the value of ##\phi## by ##\pi##.
 
  • Like
Likes MatinSAR and laser
Orodruin said:
No. There are an infinite number of solutions, which correspond to shifting ##\phi##. Once you have the form ##e^{i\phi}## that is the general solution already. There is no point in introducing yet another multiplicative constant. Multiplying by ##-1## specifically just shifts the value of ##\phi## by ##\pi##.
I don't follow. By "shifting ##\phi##", you are changing the value of ##\phi##.
 
laser said:
I don't follow. By "shifting ##\phi##", you are changing the value of ##\phi##.
Yes, that's because ##\phi## is an arbitrary parameter that is not physically measurable. It can be set to ##\phi =0 \Rightarrow A=1## or ##\phi =\pi \Rightarrow A=-1## or to any other value whatsoever without changing the physical predictions that stem from the wavefunction. On the other hand, the phase difference between two quantum states is a physical observable.
 
renormalize said:
Yes, that's because ##\phi## is an arbitrary parameter that is not physically measurable. It can be set to ##\phi =0 \Rightarrow A=1## or ##\phi =\pi \Rightarrow A=-1## or to any other value whatsoever without changing the physical predictions that stem from the wavefunction. On the other hand, the phase difference between two quantum states is a physical observable.
I understand, but for a wavefunction of definite momentum, ##\phi=\frac{px}{\hbar}##.

Edit: Sorry if I am not making any sense, it seems like I have a misconception here which I don't see yet.

To help clarify, you guys are saying that if I have a -1/sqrt(L), it is equivalent to using the 1/sqrt(L) form given that I shift the angle ##\phi## by pi. I am fine with that.

But I claim that -1/sqrt(L) is acceptable without shifting the angle by pi.
 
laser said:
I understand, but for a wavefunction of definite momentum, ##\phi=\frac{px}{\hbar}##.
No, that's not right. In the notation of your original post:$$\psi_{p}\left(x\right)=Ae^{ipx/\hbar}=\left|A\right|e^{i\phi}e^{ipx/\hbar}=\left|A\right|e^{i\left(px/\hbar+\phi\right)}$$So ##\phi## is the (arbitrary) constant phase of the complex normalization-constant ##A##, whereas ##px/\hbar## is the spatially-dependent phase of the complex wavefunction. Don't confuse your phases!
 
  • Like
Likes MatinSAR and laser
renormalize said:
No, that's not right. In the notation of your original post:$$\psi_{p}\left(x\right)=Ae^{ipx/\hbar}=\left|A\right|e^{i\phi}e^{ipx/\hbar}=\left|A\right|e^{i\left(px/\hbar+\phi\right)}$$So ##\phi## is the (arbitrary) constant phase of the complex normalization-constant ##A##, whereas ##px/\hbar## is the spatially-dependent phase of the complex wavefunction. Don't confuse your phases!
Thank you, that makes sense
 
  • #10
laser said:
But I claim that -1/sqrt(L) is acceptable without shifting the angle by pi.
This is wrong. Using -1 instead of 1 is changing the phase by ##\pi##. You could also use ##i##, ##-i## or any other phase shift (multiply by a complex number of norm 1).

Edit: The general solution to ##|B| = 1## is ##B = e^{i\phi}## with ##\phi \in \mathbb R## (or in ##[0,2\pi)## if you want to avoid duplicates). So the general solution is not ##A = \pm 1/\sqrt L##, it is ##A= e^{i\phi}/\sqrt L## with ##\phi## an arbitrary constant real number.
 
  • #11
Orodruin said:
This is wrong. Using -1 instead of 1 is changing the phase by ##\pi##. You could also use ##i##, ##-i## or any other phase shift (multiply by a complex number of norm 1).

Edit: The general solution to ##|B| = 1## is ##B = e^{i\phi}## with ##\phi \in \mathbb R## (or in ##[0,2\pi)## if you want to avoid duplicates). So the general solution is not ##A = \pm 1/\sqrt L##, it is ##A= e^{i\phi}/\sqrt L## with ##\phi## an arbitrary constant real number.
Yes sorry I misread your original comment, and thought that the angle was px/hbar, my bad! It makes sense now :) - I'm just used to seeing |A| only in real numbers, in which case then A is only +-A.
 
  • #12
Orodruin said:
This is wrong. Using -1 instead of 1 is changing the phase by ##\pi##. You could also use ##i##, ##-i## or any other phase shift (multiply by a complex number of norm 1).

Edit: The general solution to ##|B| = 1## is ##B = e^{i\phi}## with ##\phi \in \mathbb R## (or in ##[0,2\pi)## if you want to avoid duplicates). So the general solution is not ##A = \pm 1/\sqrt L##, it is ##A= e^{i\phi}/\sqrt L## with ##\phi## an arbitrary constant real number.
Is it standard in physics to omit this e^(iangle) term? As in the original post, it was just 1/sqrt(L).
 
  • #13
laser said:
Is it standard in physics to omit this e^(iangle) term? As in the original post, it was just 1/sqrt(L).
If you just want a normalised wave function, you can pick any value for the angle. Choosing ##\phi = 0## is then perfectly permissible.
 
  • Like
Likes PhDeezNutz and laser
Back
Top