The Number e and Its Series Expansion

The series for e and the bound on the remainder

The exponential series

The starting point of the proof is the series expansion of ee, obtained by evaluating the exponential series at 11:

e=∑k=0+∞1k!=1+1+12+16+124+1120+⋯e = \sum_{k=0}^{+\infty} \frac{1}{k!} = 1 + 1 + \frac{1}{2} + \frac{1}{6} + \frac{1}{24} + \frac{1}{120} + \cdots

Convergence is very fast. Let us denote the partial sums and the remainder:

Sn=∑k=0n1k!andRn=e−Sn=∑k=n+1+∞1k!S_n = \sum_{k=0}^{n} \frac{1}{k!} \qquad \text{and} \qquad R_n = e - S_n = \sum_{k=n+1}^{+\infty} \frac{1}{k!}

nn SnS_n e−Sne - S_n
22 52=2.5\dfrac{5}{2} = 2.5 ≈2.18×10−1\approx 2.18 \times 10^{-1}
44 6524≈2.70833\dfrac{65}{24} \approx 2.70833 ≈9.95×10−3\approx 9.95 \times 10^{-3}
66 1957720≈2.71806\dfrac{1957}{720} \approx 2.71806 ≈2.26×10−4\approx 2.26 \times 10^{-4}
1010 ≈2.7182818\approx 2.7182818 ≈2.73×10−8\approx 2.73 \times 10^{-8}

The decisive bound on the remainder

The whole proof rests on the following inequality, valid for n≥1n \geq 1:

0<Rn<1n⋅n!0 < R_n < \frac{1}{n \cdot n!}

Lower bound. All the terms of RnR_n are strictly positive, so Rn>0R_n > 0.

Upper bound. Factor out the first term:

Rn=1(n+1)!(1+1n+2+1(n+2)(n+3)+⋯ )R_n = \frac{1}{(n+1)!}\left(1 + \frac{1}{n+2} + \frac{1}{(n+2)(n+3)} + \cdots\right)

Each factor 1n+j\dfrac{1}{n+j} is strictly less than 1n+1\dfrac{1}{n+1} as soon as j≥2j \geq 2, so the parenthesis is bounded above by the geometric series with ratio 1n+1\dfrac{1}{n+1}:

Rn<1(n+1)!∑j=0+∞1(n+1)j=1(n+1)!⋅n+1n=1n⋅n!R_n < \frac{1}{(n+1)!} \sum_{j=0}^{+\infty} \frac{1}{(n+1)^j} = \frac{1}{(n+1)!} \cdot \frac{n+1}{n} = \frac{1}{n \cdot n!}

Why this bound is the right one

Multiply the inequality by n!n!:

0<n! Rn<1n0 < n!\,R_n < \frac{1}{n}

The quantity n! Rnn!\,R_n is therefore squeezed strictly between 00 and 1n\dfrac{1}{n}, that is, strictly between 00 and 11 as soon as n≥2n \geq 2. It will be enough to prove that this number is an integer to reach the contradiction: that is exactly the program of the next chapter.

Classic pitfall

Confusing 1n⋅n!\dfrac{1}{n \cdot n!} with 1(n+1)!\dfrac{1}{(n+1)!}, or settling for the crude bound Rn<1(n+1)!⋅(n+1)=1n!R_n < \dfrac{1}{(n+1)!} \cdot (n+1) = \dfrac{1}{n!}. The latter gives 0<n! Rn<10 < n!\,R_n < 1, which is in fact also enough for the contradiction; the version with 1n⋅n!\dfrac{1}{n \cdot n!} is simply sharper.