• Support PF! Buy your school textbooks, materials and every day products via PF Here!

BCH formula example

Problem Statement
I want to calculate [tex]D(\beta)bD(-\beta)[/tex]
where b is ladder operator in problem of LHO and ##\beta## is complex number.
[tex]D(\beta)=exp(\beta b^{\dagger}-\beta^*b)[/tex]
Relevant Equations
Commutator [tex][b,b^{\dagger}]=I[/tex]
BCH formula:
[tex]e^Ae^B=exp(A+B+\frac{1}{2}[A,B]+\frac{1}{12}[A,[A,B]]+\frac{1}{12}[B,[A,B]]+...)[/tex]
From
[tex][b,b^{\dagger}]=I[/tex] it is easy to see that
[tex][b,[b,b^{\dagger}]]=[b^{\dagger},[b,b^{\dagger}]]=0. [/tex]
And from that
[tex]e^be^{b^{\dagger}}=e^{b+b^{\dagger}}e^{\frac{1}{2}I}[/tex]
Using that I get
[tex]exp(\beta b^{\dagger}-\beta^*b)=e^{\beta b^{\dagger}}e^{-\beta^*b}e^{-\frac{1}{2}|\beta|^2I}[/tex].
So
[tex]D(\beta)bD(-\beta)=e^{\beta b^{\dagger}}e^{-\beta^*b}e^{-\frac{1}{2}|\beta|^2I}be^{-\beta b^{\dagger}}e^{\beta^*b}e^{-\frac{1}{2}|\beta|^2I}[/tex]
and I do not see what to do from here.
 

fresh_42

Mentor
Insights Author
2018 Award
10,730
7,350
For my taste these notations are far too sloppy, esp. the fact that you do not distinguish ##b## and ##\mathfrak{ad}b##. So assuming you made no mistake, you have
Correction:
\begin{align*}
D(\beta)bD(-\beta) &= (\exp(\beta b^\dagger)(\exp(-\beta^*b)b(\exp(\beta b)(\exp(-\beta b^\dagger)\cdot e^{-|\beta|^2}\\ &= e^{-|\beta|^2} \cdot \operatorname{Ad}(\exp(\beta b^\dagger)\cdot (\exp(-\beta^*b) ) (b)
\end{align*}
since ##e^I## commutes with everything and the rest is the conjugation by ##\exp( \beta b^\dagger) ## and ## \exp (- \beta^* b) ##.
 
Last edited:
What is ##Ad()##?
 

fresh_42

Mentor
Insights Author
2018 Award
10,730
7,350
What is ##Ad()##?
The adjoint representation of the Lie group on its tangent space. It is a conjugation: ##\exp(-\beta b^*) \stackrel{ \operatorname{Ad}}{\longmapsto} \left( b \longmapsto \exp(-\beta b^*)b\exp(\beta b^*)\right)##
 
Last edited:
Correct result should be
[tex]D(\beta)bD(-\beta)=b-\beta I[/tex]. And I am still not sure how to get this.
 

fresh_42

Mentor
Insights Author
2018 Award
10,730
7,350
Correct result should be
[tex]D(\beta)bD(-\beta)=b-\beta I[/tex]. And I am still not sure how to get this.
Why not calculate it directly without BCH?
\begin{align*}
D(\beta)bD(-\beta)&= \exp(\beta b^\dagger - \beta^* b)b \exp(-(\beta b^\dagger - \beta^* b)) \\
&= \operatorname{Ad}(\exp(\beta b^\dagger - \beta^* b))(b) \\
&= \exp(\operatorname{ad}(\beta b^\dagger - \beta^* b))(b) \\
&= \left( 1 + \operatorname{ad}(\beta b^\dagger - \beta^* b) + \dfrac{1}{2!} \operatorname{ad}^2(\beta b^\dagger - \beta^* b)+\ldots \right)(b)\\
&= b+ [\beta b^\dagger - \beta^* b,b] + \dfrac{1}{2} [\beta b^\dagger - \beta^* b,[\beta b^\dagger - \beta^* b,b]] + \dfrac{1}{3!} [\beta b^\dagger - \beta^* bv,[\beta b^\dagger - \beta^* b,[\beta b^\dagger - \beta^* b,b]]]+\ldots
\end{align*}
 
Thanks. You will obviously get good result in this way. However I am not familiar with adjoint representation. Is it possible to solve it without it?
 

fresh_42

Mentor
Insights Author
2018 Award
10,730
7,350
Thanks. You will obviously get good result in this way. However I am not familiar with adjoint representation. Is it possible to solve it without it?
Without it in physics means nothing else than drop the notation. Physicists tend to use the same words regardless whether it takes places on the manifold (Lie group) or its tangent spaces (Lie algebra). If you're lucky they call the tangents generators. So without it means nothing else as to jump from line 1 directly to line 5.

Here is what formally happens (at the example of ##SU(2)##):

Btw. you can still use BCH. Just use it for the last line.
 

George Jones

Staff Emeritus
Science Advisor
Gold Member
7,190
737
Problem Statement: I want to calculate [tex]D(\beta)bD(-\beta)[/tex]
where b is ladder operator in problem of LHO and ##\beta## is complex number.
[tex]D(\beta)=exp(\beta b^{\dagger}-\beta^*b)[/tex]
Relevant Equations: Commutator [tex][b,b^{\dagger}]=I[/tex]
Writing ##A = \beta b^\dagger - \beta^* b## gives
$$\begin{align}
D\left(\beta\right)bD\left(-\beta\right) &= e^A b e^{-A} \nonumber \\
&= \left( \left[ e^A , b \right] + b e^A \right) e^{-A} \nonumber
\end{align} $$
Now, Calculate ##\left[ e^A , b \right]## using
$$\begin{align}
\left[ A^n , b \right] &= \left[ A A^{n-1} , b \right] \nonumber \\
&= A \left[ A^{n-1} , b \right] + \left[ A , b \right] A^{n-1} \nonumber
\end{align} $$
and mathematical induction.
 

Want to reply to this thread?

"BCH formula example" You must log in or register to reply here.

Physics Forums Values

We Value Quality
• Topics based on mainstream science
• Proper English grammar and spelling
We Value Civility
• Positive and compassionate attitudes
• Patience while debating
We Value Productivity
• Disciplined to remain on-topic
• Recognition of own weaknesses
• Solo and co-op problem solving
Top