Normalization constant: Difference between revisions

From PhyWiki
Jump to navigation Jump to search
No edit summary
No edit summary
 
(One intermediate revision by the same user not shown)
Line 8: Line 8:


<math> + \sum_{k=1}^{+\infty}\lambda ^{k+1} \left ( \sum_{m_{1}}'...\sum_{m_{k}}' \langle n|H'|m_ {1} \rangle \frac {1}{E_{n}- \epsilon _{m_{1}}}\langle m_{1}|H'|m_{2} \rangle.\frac {1}{E_{n}- \epsilon _{m_{2}}}\langle m_{2}|H'|m_{3} \rangle...\frac {1}{E_{n}- \epsilon _{m_{k-1}}}\langle m_{k-1}|H'|m_{k} \rangle.\frac {1}{E_{n}- \epsilon _{m_{k}}}\langle m_{k}|H'|n \rangle \right ) </math>
<math> + \sum_{k=1}^{+\infty}\lambda ^{k+1} \left ( \sum_{m_{1}}'...\sum_{m_{k}}' \langle n|H'|m_ {1} \rangle \frac {1}{E_{n}- \epsilon _{m_{1}}}\langle m_{1}|H'|m_{2} \rangle.\frac {1}{E_{n}- \epsilon _{m_{2}}}\langle m_{2}|H'|m_{3} \rangle...\frac {1}{E_{n}- \epsilon _{m_{k-1}}}\langle m_{k-1}|H'|m_{k} \rangle.\frac {1}{E_{n}- \epsilon _{m_{k}}}\langle m_{k}|H'|n \rangle \right ) </math>
where <math>\sum '</math>does not allow the running indexes equal to n.


Taking the derivative of <math>E_{n}</math> with respect <math>\bold \epsilon _{n}</math> to, using the chain rule ,we get:
Taking the derivative of <math>E_{n}</math> with respect <math>\bold \epsilon _{n}</math> to, using the chain rule ,we get:
Line 40: Line 42:


<math> \bold . \langle m_{1}|H'|m'_{2} \rangle ...\frac {1}{E_{n}- \epsilon _{m'_{q-1}}} \langle m'_{q-1}|H'|m'_{q} \rangle \frac {1}{E_{n}- \epsilon _{m'_{q}}} \langle m'_{q}|H'|n \rangle </math>
<math> \bold . \langle m_{1}|H'|m'_{2} \rangle ...\frac {1}{E_{n}- \epsilon _{m'_{q-1}}} \langle m'_{q-1}|H'|m'_{q} \rangle \frac {1}{E_{n}- \epsilon _{m'_{q}}} \langle m'_{q}|H'|n \rangle </math>
Let's define <math>k=p+q-1</math> and exchange the indexes as follows:
<math>m_{p} \rightarrow m_{1}; \; m_{p-1} \rightarrow m_{2}; \; ... ; m_{1} \rightarrow m_{p} </math>
<math>m'_{2} \rightarrow m_{p+1}; \; m'_{3} \rightarrow m_{p+2}; \; ... ; m'_{q} \rightarrow m_{p+q-1}=m_{k} </math>
Doing so we can see that <math>Z^{-1}</math> exactly equals to <math>(\frac {\partial E_{n}}{\partial \epsilon_{n}})^{-1}</math> given in (2). Therefore:
<math>Z=\langle N|N \rangle ^{-1} = \frac {\partial E_{n}}{\partial \epsilon_{n}}</math>

Latest revision as of 16:54, 18 April 2009

Using Brillouin-Wigner perturbation theory we will proof that

In this theory, the exact state and exact energy can be written as follows:

where does not allow the running indexes equal to n.

Taking the derivative of with respect to, using the chain rule ,we get:

From this we can solve for

Now let's evaluate from

We have , therefore the summing over is equivalent to setting . We get:

Let's define and exchange the indexes as follows:

Doing so we can see that exactly equals to given in (2). Therefore: