In the paper "Quantum Langevin Equation" by G. W. Ford, J. T. Lewis, and R. F. O’Connell I found the following statement \begin{equation} \frac{1}{\pi} \int_{0}^{\infty} \omega \cos[\omega t] \coth[\alpha \omega/2] \text{d} \omega = \frac{1}{\alpha} \frac{\text{d}}{\text{d}t} \coth[\pi t/\alpha]. \end{equation} It is Equation (2.11) in the paper I linked.

I would like to understand how to obtain this result. I have noticed that the left hand side can be written as \begin{equation} \frac{\text{d}}{\text{d}t}\frac{1}{\pi} \int_{0}^{\infty} \sin[\omega t] \coth[\alpha \omega/2] \text{d} \omega, \end{equation} and then tried using Wolframalpha to evaluate the integral \begin{equation} \int_{0}^{\infty} \sin[\omega] \coth[\omega] \text{d} \omega, \end{equation} but the result is that this integral diverges.

Furthermore I know that \begin{equation} \frac{1}{\pi} \int_{0}^{\infty} \cos[\omega t] \text{d} \omega = \delta(t), \end{equation} I thought maybe I could use this in the following way, starting at the l.h.s. \begin{align} \pi\frac{\text{d}}{\text{d}t} \coth[\pi t/ \alpha] &= \pi\frac{\text{d}}{\text{d}t} \int_{R} \text{d} \mu \delta(t-\mu) \coth[\pi \mu/\alpha] \\ &= \frac{\text{d}}{\text{d}t}\int_{R} \text{d}\mu \int_{0}^{\infty} \text{d} \omega \cos[\omega(t-\mu)] \coth[\pi \mu/\alpha] \\ &= \frac{\text{d}}{\text{d}t} \int_{R} \text{d} \mu \int_{0}^{\infty} \text{d} \omega \cos[\omega \mu] \coth[\pi(\mu + t) /\alpha] \\ &= \int_{0}^{\infty} \text{d} \omega \int_{R} \text{d} \mu \cos[\omega \mu] \frac{\text{d}}{\text{d}\mu} \coth[\pi(\mu+ t)/\alpha], \end{align} and from there I don't really know what to do... One could try partial integration, but I don't really see that working out either.

Any help to figure this out would be greatly appreciated!


1 Answer 1


I suggest that either they use some other sense of integration (see edit below), or there is an error.

As far as I can see, the integral (forgetting different scaling constants) $$ \int_0^{+\infty}w\cos(w)\coth(w)\,dw $$ is divergent. One argument, that could be done more rigorous, is that $$ \coth(w)\approx 1 $$ where the $\approx$ is really exponentially fast as $w$ increases. Practically, $\coth w=1$ for $w>10$.

Since the integral $$ \int_0^{+\infty} w\cos (w)\,dw $$ is divergent, the integral $$ \int_0^{+\infty} w\cos(w)\coth(w)\,dw $$ is also divergent.

One way out could maybe be to consider complex constants $t$ and $\alpha$?


What they most likely do is that they do a Fourier cosine transform, considering $w\coth(w)$ as a distribution. Indeed, then $$ \mathcal F_{\cos}(w\coth w)(s)=-\frac{\pi^2}{2}\bigl(\text{csch}\,(\pi s/2)\bigr)^2. $$ Shamelessly, I suggest you to read this answer of mine to see how one can proceed in calculating such transforms.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .