Maths Olympiad Prep

Library / /3 of 10

, 2022

Number theory Difficulty 5.6 AIME, harder Prove it Switzerland

Problem:
Let nn be a positive integer and dd a positive divisor of nn. Show that if
d2+d+1n+1 \frac{d^{2}+d+1}{n+1}
is an integer, then it is equal to 11.

Solutions — 2

Solution 1

Solution:
Assume that the fraction is an integer, write
d2+d+1n+1=m \frac{d^{2}+d+1}{n+1}=m
Obviously mm is going to be positive, as both nn and dd are also positive. Since dd divides nn, we can write n=kdn=k d. Plugging it into the above equation we get
d2+d+1=(kd+1)md2+dkdm=m1 d^{2}+d+1=(k d+1) m \Leftrightarrow d^{2}+d-k d m=m-1
So dd also divides m1m-1, we can write m=ld+1m=l d+1 for lZ,l0l \in \mathbb{Z}, l \geq 0 (l1l \geq 1 if and only if m1>0m-1>0). Plugging that in again, gives us
d2+d+1=(kd+1)(ld+1)=kld2+d(k+l)+1d+1=kld+k+l d^{2}+d+1=(k d+1)(l d+1)=k l d^{2}+d(k+l)+1 \Leftrightarrow d+1=k l d+k+l
As kNk \in \mathbb{N}, we have k1k \geq 1. Now assume m1m \neq 1. From this it would follow that m1m-1 is strictly positive, leading to l1l \geq 1. Under this assumption we would get klddk l d \geq d, leading to
kdl+k+ld+k+ld+1+1>d+1=kdl+k+l k d l+k+l \geq d+k+l \geq d+1+1>d+1=k d l+k+l
which is impossible. Contradiction! So we must have l<1l=0l<1 \Rightarrow l=0 leading to m=0d+1=1m=0 \cdot d+1=1 as wanted.

Solution 2

Solution:
Write n=kdn=k d as before, and we obtain that (kd+1)(d2+d+1)(k d+1) \mid\left(d^{2}+d+1\right). As aba \mid b implies abaa \mid b-a we get
kd+1d2+dkd=d(d+1k) k d+1 \mid d^{2}+d-k d=d \cdot(d+1-k)
As gcd(kd+1,d)=1\gcd(k d+1, d)=1 by Euclid's Algorithm, we also get
kd+1d+1k k d+1 \mid d+1-k
This follows as abca \mid b c and gcd(a,b)=1\operatorname{gcd}(a, b)=1 implies aca \mid c. We do a case distinction on the size of kk
- If k<d+1k<d+1, we have d+1k>0d+1-k>0 which implies
kd+1d+1kd k d+1 \leq d+1-k \leq d
This is impossible
- If k>d+1k>d+1, we instead have kd1>0k-d-1>0. As aba \mid b implies aba \mid-b we also have kd+1kd1k d+1 \mid k-d-1. We get kd+1kd1kk d+1 \leq k-d-1 \leq k which is also impossible.
- The final case k=d+1k=d+1 gives us n=d(d+1)n=d(d+1). Then n+1=d(d+1)+1=d2+d+1n+1=d(d+1)+1=d^{2}+d+1, which means that the fraction is equal to one

Note that in the first two cases we used aba \mid b implies aba \leq b, if both a,b>0a, b>0. This finishes the proof.

Want a route through all this instead of an archive? The track puts 2,000 problems in a working order, from AMC 10 level to the IMO shortlist.

Source: MathNet, licensed CC-BY-4.0. Statement reproduced verbatim; metadata (topic, difficulty) added by this project.