From statistical mechanics I've found $ \left( c,\alpha \in \mathbb{R}^+ \right) $:
$ lnZ = c\int_0^\infty dx\ x^2ln\left( 1 + e^{-\alpha x^2}\right) $
what I performed a integration by parts to get:
$ lnZ = \dfrac{2\alpha}{3}c\int_0^\infty dx\dfrac{x^4}{1 + e^{\alpha x^2}} $.
Every exercise in quantum gases (fermions/bosons) lead me to this integration in classical limit. I try to solve it via Taylor's polynomium around $x=0$ because it's a integration limit value. The first term which will not vanish (get 0 value in $x=0$) is in $\frac{d^4f}{dx^4}\left( x\right) = \frac{4!}{24!}x^4\ $ what take me a divergence in the integration.
Other way I've tried is apply Taylor's to $ ln\left( 1+e^{-\alpha x^2}\right) $ at $x=0 $ what is the same to span $ ln\left(1+x\right) $ around $ x=1 $:
$ \int ln\left(1+x\right) dx = ln2 +\sum_{k\in\mathbb{N}}\left( -1 \right)^{k+1}\dfrac{\left( x-1 \right)^{k}}{k\left( k+1 \right)} $.
Using test of $\ \ lim_{k\rightarrow\infty}\left| \frac{A_k}{A_{k+1}} \right| < 1 \ \ $ I conclude this converges on $ x\in\left( 0,2\right) $.
My questions:
- The integration was solved and the result is valid just for $ x\in\left( 0,2\right) \Rightarrow e^{-\alpha x^2}\in\left( e^{-4\alpha}, 1\right) $. Is this conclusion right?
- I feel Like is something going wrong in my solution. Else way to solve this integration or workaround it?
$\newcommand{\bbx}[1]{\,\bbox[15px,border:1px groove navy]{\displaystyle{#1}}\,} \newcommand{\braces}[1]{\left\lbrace\,{#1}\,\right\rbrace} \newcommand{\bracks}[1]{\left\lbrack\,{#1}\,\right\rbrack} \newcommand{\dd}{\mathrm{d}} \newcommand{\ds}[1]{\displaystyle{#1}} \newcommand{\expo}[1]{\,\mathrm{e}^{#1}\,} \newcommand{\ic}{\mathrm{i}} \newcommand{\mc}[1]{\mathcal{#1}} \newcommand{\mrm}[1]{\mathrm{#1}} \newcommand{\pars}[1]{\left(\,{#1}\,\right)} \newcommand{\partiald}[3][]{\frac{\partial^{#1} #2}{\partial #3^{#1}}} \newcommand{\root}[2][]{\,\sqrt[#1]{\,{#2}\,}\,} \newcommand{\totald}[3][]{\frac{\mathrm{d}^{#1} #2}{\mathrm{d} #3^{#1}}} \newcommand{\verts}[1]{\left\vert\,{#1}\,\right\vert}$
\begin{align} \ln\pars{Z} & = {2\alpha \over 3}\,c\int_{0}^{\infty}\dd x\,{x^{4} \over 1 + \expo{\alpha x^{2}}} \,\,\,\stackrel{\alpha x^{2}\ \mapsto\ x}{=}\,\,\, {c \over 3\alpha^{3/2}}\int_{0}^{\infty} {x^{3/2} \over \expo{x} + 1}\,\dd x \\[5mm] & = {c \over 3\alpha^{3/2}}\int_{0}^{\infty} x^{3/2}\pars{{1 \over \expo{x} + 1} - {1 \over \expo{x} - 1}} \,\dd x + {c \over 3\alpha^{3/2}}\int_{0}^{\infty} {x^{3/2} \over \expo{x} - 1}\,\dd x \\[5mm] & = -\,{2c \over 3\alpha^{3/2}}\int_{0}^{\infty} {x^{3/2} \over \expo{2x} - 1} \,\dd x + {c \over 3\alpha^{3/2}}\int_{0}^{\infty} {x^{3/2} \over \expo{x} - 1}\,\dd x \\[5mm] & = -\,{2c \over 3\alpha^{3/2}}\bracks{2^{-5/2}\int_{0}^{\infty} {x^{3/2} \over \expo{x} - 1} \,\dd x} + {c \over 3\alpha^{3/2}}\int_{0}^{\infty} {x^{3/2} \over \expo{x} - 1}\,\dd x \\[5mm] & = {\pars{1 - 2^{-3/2}}c \over 3\alpha^{3/2}}\ \underbrace{\int_{0}^{\infty} {x^{3/2} \over \expo{x} - 1}\,\dd x} _{\ds{\underbrace{\Gamma\pars{5 \over 2}}_{\ds{3\root{\pi}/4}}\zeta\pars{5 \over 2}}}\ = \bbx{{1 - 2^{-3/2} \over 4}\root{\pi}\zeta\pars{5 \over 2} \,{c \over \alpha^{3/2}}} \end{align}