Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Today I had an exam and I mixed up the integration by parts formula. The question was to integrate $$ \int\nolimits_0^\infty \frac{e^{-at} - e^{-bt}}{t} \text{d}t $$

I will try solve this again with the right formula when I arrive home. I would appreciate if somebody could tell me the solution so I can double check and maybe give a hint to another way of solving this instead of integration by parts (if possible).

share|cite|improve this question
Hint: Write the integrand as an integral over [a,b], then switch the order of integration. – Ragib Zaman Sep 6 '11 at 10:21
I actually asked this question a couple of weeks ago:… – Gerben Sep 6 '11 at 12:45
I guess this is a duplicate then, but @Ragib's answer is different (from robjohn's) and also nice. – Srivatsan Sep 6 '11 at 12:46

I may as well give an answer:

Note $$ \frac{ e^{-at} - e^{-bt} }{t} = \int^b_a e^{-xt} dx $$ so our integral is

$$ \int^{\infty}_0 \int^b_a e^{-xt} dx dt = \int^b_a \int^{\infty}_0 e^{-xt} dt dx $$ $$ = \int^b_a \frac{1}{x} dx = \log(b/a) $$

This is a general method, and often this whole process is compressed into a well known integral called Frullani's Integral.

share|cite|improve this answer
nice solution you gave! – user 1618033 Aug 5 '12 at 14:45

By differentiating under the integral sign. Fix $a$, and let

$$ g(b) = \int_{0}^\infty \frac{e^{-at} - e^{-bt}}{t} dt. $$

Differentiating w.r.t. $b$, we get $$ g'(b) = \frac{d}{db}\int_{0}^\infty \frac{e^{-at} - e^{-bt}}{t} dt = \int_{0}^{\infty} \frac{\partial}{\partial b} \frac{e^{-at} - e^{-bt}}{t} dt. $$ Doing the differentiation, $$ g'(b) = \int_{0}^{\infty} e^{-bt} dt = - \left.\frac{e^{-bt}}{b} \right|_{0}^{\infty} = \frac{1}{b}. $$ Thus we must have $g(b) = \ln b + C$ for some constant $C$. To determine $C$, plug in $a$, and use the fact that $g(a) = 0$.

Note. I am yet to convince myself that all the steps of the proof are rigorous. I will edit my answer later if additional argument is necessary.

share|cite|improve this answer
@Ragib Sure. I will delete my comments now (and this one in a little while). – Srivatsan Oct 3 '11 at 17:22

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.