Take the 2-minute tour ×
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.

How do I prove that the exponential integral $$f(x)=\int \frac{e^x}{x}\mathrm dx$$ is not an elementary function?

Also, what are the general methods and tricks to prove that an integral or solution to an equation is not an elementary function?

share|improve this question
3  
The general method is the Differential Galois Theory or the Galois theory of differential equations. There is an extremely nice exposition of the theory in a book by Andy Majid published by the AMS---at the end, IIRC, there is a complete argument that $\int e^{x^2}$ is not elementary, as an example. –  Mariano Suárez-Alvarez Oct 28 '11 at 11:21
    
Risch's services would be of great assistance. –  J. M. Oct 28 '11 at 11:21
    
en.wikipedia.org/wiki/Risch_algorithm –  pedja Oct 28 '11 at 11:23
add comment

1 Answer

All we need for this is a theorem of Liouville (1835): Suppose that $f$ and $g$ are rational functions with $f\neq 0$ and $g$ non-constant. Then $$\int f(x) e^{g(x)} dx $$ is an elementary function if and only if there exists a rational function $r$ such that $ f=r'+g'r.$

Here we have $g(x)=x$ and $f(x) = 1/x.$ Assume there exists a rational function $r$ such that $ 1/x = r' + r \text{ } $ (1). Denote the multiplicity of the pole at $0$ by $m$ (where $m\geq 1$ so that both sides of (1) agree when $x\to 0$), so that $ \displaystyle r(x) = \frac{p(x)}{x^m Q(x)} $ where $p, Q$ are polynomials with no common factors and $Q$ is not divisible by $x.$

Substituting this form into (1) and multiplying both sides by $x^m$ yields $$ x^{m-1} = \frac{p(x)+ p'(x)}{Q(x)} - \frac{Q'(x) p(x) }{Q^2(x)} - \frac{m}{x Q(x)} .$$ Taking limits of both sides as $x\to 0$ illustrates the lunacy in our assumption that such a rational function exists, and hence $\displaystyle \int \frac{e^x}{x} dx$ is non-elementary.

share|improve this answer
7  
The fun is then of course hidden in the proof of Liouville's theorem. We could also say that «we only need the theorem that $\int e^x/x$ cannot be expressed in elementary terms»! :) –  Mariano Suárez-Alvarez Oct 28 '11 at 13:56
add comment

Your Answer

 
discard

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.