Proving [\hat{x}, \hat{k}]=i in Operator Notation

  • Thread starter Thread starter latentcorpse
  • Start date Start date
  • Tags Tags
    Notation Operator
Click For Summary

Homework Help Overview

The discussion revolves around proving the commutation relation [\hat{x}, \hat{k}] = i in the context of operator notation, where \hat{x} and \hat{k} are defined in terms of integrals over position and momentum states, respectively. Participants are exploring the implications of these definitions and the mathematical manipulations involved in deriving the commutation relation.

Discussion Character

  • Exploratory, Mathematical reasoning, Assumption checking

Approaches and Questions Raised

  • Participants discuss the application of the operators \hat{x} and \hat{k} on various states and question the assumptions made regarding the derivatives and the order of operations. There are attempts to derive the commutation relation through integration and manipulation of the operator expressions.

Discussion Status

The discussion is active, with participants providing insights and corrections to each other's reasoning. Some participants have suggested alternative approaches and questioned the validity of certain steps, indicating a collaborative exploration of the problem. There is no explicit consensus yet on the final form of the commutation relation.

Contextual Notes

Participants are navigating the complexities of operator algebra in quantum mechanics, including the treatment of basis states and the implications of applying differential operators. There are indications of potential sign errors and misunderstandings regarding the manipulation of integrals and operators.

latentcorpse
Messages
1,411
Reaction score
0
Ok so i need to prove that [itex][ \hat{x} , \hat{k} ] = i[/itex] in this operator notation where
[itex]\hat{x}=\int dx | x \rangle \langle x | x[/itex]
[itex]\hat{k}=\int dk | k \rangle \langle k | k[/itex]
and [itex]\langle x | k \rangle = \frac{1}{\sqrt{2 \pi}}e^{ikx}[/itex]

so i have worked out
[itex]\hat{x} \hat{k} = \int dx \hat{k} | x \rangle \langle x | x[/itex]
but [itex]\hat{k} | x \rangle = i \frac{\partial}{\partial x} | x \rangle[/itex]
so [itex]\hat{x} \hat{k} = \int dx i \frac{\partial x}{\partial x} | x \rangle \langle x|[/itex]
[itex]=i \int dx | x \rangle \langle x | = i \hat{1}=i[/itex] where [itex]\hat{1}[/itex] is the unit/identity operator

similarly i find that [itex]\hat{k} \hat{x}=-i[/itex]
and so [itex][\hat{x},\hat{k}]=i-(-i)=2i[/itex]
i can't see where i was supposed to get rid of that factor of 2 though?

thanks.
 
Physics news on Phys.org
i just have one comment (in line 8), why did you assume that the partial operator will just be applied on x? I am curious since I am taking the same course now ..
 
i think it's because the bra and the ket are basis states so their derivative would be zero. It says on this article
http://en.wikipedia.org/wiki/Matrix_mechanics
that
[itex]D \int \psi(x) | x \rangle dx = \int \psi'(x) | x \rangle dx[/itex]
where D is a derivative operator.
 
one other question .. What is the effect if we are applying x operator on k state? .. I want to try working out this problem maybe i would be helpful ..
 
it's as follows:

[itex]\hat{x}|k \rangle = \int x | x \rangle \langle x | k \rangle dx = \int \frac{x}{\sqrt{2 \pi}} e^{-ikx} | x \rangle dx = \frac{dx}{\sqrt{2 \pi}} i \frac{\partial}{\partial k} e^{-ikx} | x \rangle = \int dx i \frac{\partial}{\partial k} |x \rangle \langle x | k \rangle = i \frac{\partial}{\partial k} | k \rangle[/itex]
 
latentcorpse said:
[itex]\hat{x} \hat{k} = \int dx\, \hat{k} | x \rangle \langle x | x[/itex]
You seem to have mistakenly commuted the operators. The order of x-hat and k-hat are different on the LHS and RHS.
 
vela said:
You seem to have mistakenly commuted the operators. The order of x-hat and k-hat are different on the LHS and RHS.

so the one I've actually shown the working for is [itex]\hat{k} \hat{x}[/itex]. This gives me the i that i want so i just have to show [itex]\hat{x} \hat{k}=0[/itex]

i have [itex]\hat{x} \hat{k} = \int dx x | x \rangle \langle x | \hat{k}[/itex]

i don't know what to do now. should i sub for the k operator? then i'll have an integral over both x and k and i'll still have something operating to the right when there's nothing on the right to operate on.
 
latentcorpse said:
so [itex]\hat{k} \hat{x} = \int dx \, i \frac{\partial x}{\partial x} | x \rangle \langle x|x[/itex]
This isn't quite right either (even after correcting the order of x and k). It might be clearer to see what's going on if you consider

[tex]\hat{k} \hat{x}|\psi(x)\rangle = \int dx \, i \frac{\partial}{\partial x} | x \rangle \langle x|x|\psi(x)\rangle[/tex]
 
vela said:
This isn't quite right either (even after correcting the order of x and k). It might be clearer to see what's going on if you consider

[tex]\hat{k} \hat{x}|\psi(x)\rangle = \int dx \, i \frac{\partial}{\partial x} | x \rangle \langle x|x|\psi(x)\rangle[/tex]

not following this. how did you know to change [itex]\hat{k}[/itex] to [itex]i \frac{\partial}{\partial x}[/itex] without doing the lines i did above to prove it?

anyway i get following it through
[itex]\int dx i \partial_x | x \rangle \langle x | \hat{x} | \psi \rangle[/itex]
[itex]=i \int dx \int dx' \partial_x |x \rangle \langle x | x' | x' \rangle \langle x' | \psi \rangle[/itex]
[itex]=i \int dx \partial_x | x \rangle \langle x | x \psi(x)[/itex]
[itex]=i \partial_x (x \psi(x)) = i ( \psi(x) + x \psi'(x))[/itex]
which looks a bit...erm...off!
 
  • #10
latentcorpse said:
not following this. how did you know to change [itex]\hat{k}[/itex] to [itex]i \frac{\partial}{\partial x}[/itex] without doing the lines i did above to prove it?
I was just picking up from the middle of what you did. You'd still want to show the steps you did earlier to get to this point.
anyway i get following it through
[itex]\int dx i \partial_x | x \rangle \langle x | \hat{x} | \psi \rangle[/itex]
[itex]=i \int dx \int dx' \partial_x |x \rangle \langle x | x' | x' \rangle \langle x' | \psi \rangle[/itex]
[itex]=i \int dx \partial_x | x \rangle \langle x | x \psi(x)[/itex]
[itex]=i \partial_x (x \psi(x)) = i ( \psi(x) + x \psi'(x))[/itex]
which looks a bit...erm...off!
Actually, this is correct. So you can see that [itex]\hat{k}\hat{x}[/itex] is the operator represented by [itex]i(1+x\partial_x)[/itex], and if you look at that last term, you should see it's the representation of [itex]\hat{x}\hat{k}[/itex]. Hence, the commutator is...?
 
  • #11
vela said:
I was just picking up from the middle of what you did. You'd still want to show the steps you did earlier to get to this point.

Actually, this is correct. So you can see that [itex]\hat{k}\hat{x}[/itex] is the operator represented by [itex]i(1+x\partial_x)[/itex], and if you look at that last term, you should see it's the representation of [itex]\hat{x}\hat{k}[/itex]. Hence, the commutator is...?

well it will be i.
this however, relies on me being able to show that [itex]\hat{x} \hat{k} = x \partial_x[/itex]

[itex]\hat{x} \hat{k} | \psi \rangle = \int dx x | x \rangle \langle x| \hat{k} \psi \rangle[/itex]
[itex]= \int dx \int dk x | x \rangle \langle | k \rangle \langle k | \psi \rangle k[/itex]
[itex]=\int dx \int dk x | x \rangle (-i \partial_x) \langle x | k \rangle \langle k | \psi \rangle[/itex]
[itex]=-i \int dx x | x \rangle \partial_x \langle x | \psi \rangle[/itex]
[itex]=-i x \partial_x | \psi \rangle[/itex]

which cancels the other term leaving just the i

also the term in the previous post should have been in terms of the ket [itex]|psi \rangle[/itex] instead of [itex]\psi(x)[/itex] if we're being pedantic about the algebra, shouldn't it?
 
  • #12
latentcorpse said:
also the term in the previous post should have been in terms of the ket [itex]|psi \rangle[/itex] instead of [itex]\psi(x)[/itex] if we're being pedantic about the algebra, shouldn't it?
Yes, you're right. I was just looking at the derivative part. You got rid of one of the integrals when you shouldn't have. You had:

[tex]\hat{k}\hat{x}|\psi\rangle = i \int dx \int dx'\, \partial_x |x\rangle\langle x| x' | x'\rangle \langle x'|\psi\rangle = i \int dx \int dx' \, \partial_x |x\rangle \langle x|x' \rangle x'\psi(x')[/tex]

You can then do one integral because of the <x|x'> delta function to get:

[tex]\hat{k}\hat{x}|\psi\rangle = i \int dx \, |x\rangle \partial_x (x\psi(x))[/tex]

P.S. It looks like there might be a sign error somewhere. I'm not sure what the proper signs are, but I figure you can straighten all that out.
 
Last edited:
  • #13
latentcorpse said:
[itex]=-i \int dx x | x \rangle \partial_x \langle x | \psi \rangle[/itex]
[itex]=-i x \partial_x | \psi \rangle[/itex]

im just wondering, in the last step where you carried on the integral, is it really fine to ignore the fact you have an operator on the state x, and take out the identity operator?? .. I just have a feeling that there is something missing here? ..
 
  • #14
No, it's not. You should end up with

[tex]-i \int dx\,|x\rangle x\partial_x\psi(x)[/tex]
 
Last edited:
  • #15
vela said:
No, it's not. You should end up with

[tex]-i \int dx\,|x\rangle x\partial_x\psi(x)[/tex]

surely the final answer in post 9 should have an integral in it as well then?
 
  • #16
Yes, that's what I said in post 12.
 

Similar threads

  • · Replies 10 ·
Replies
10
Views
3K
  • · Replies 0 ·
Replies
0
Views
2K
Replies
3
Views
2K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 0 ·
Replies
0
Views
2K
  • · Replies 4 ·
Replies
4
Views
3K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 6 ·
Replies
6
Views
3K
Replies
10
Views
2K
  • · Replies 6 ·
Replies
6
Views
4K