Difference between revisions of "2008 IMO Problems/Problem 3"
Line 33: | Line 33: | ||
Now armed with <math>a+b-1>\sqrt{p}</math> and (2), we get | Now armed with <math>a+b-1>\sqrt{p}</math> and (2), we get | ||
<cmath> | <cmath> | ||
− | + | 4(n^2+1) \le p( a^2+b^2 - 2(a+b-1) ) < \\ | |
− | 4(n^2+1) | + | < p( p-2\sqrt{p} ) < \\ |
− | + | < u^2(u-1)^2, | |
− | |||
− | |||
</cmath> | </cmath> | ||
where <math>u=\sqrt{p}.</math> | where <math>u=\sqrt{p}.</math> |
Revision as of 23:33, 3 September 2008
Problem
Prove that there are infinitely many positive integers such that
has a prime divisor greater than
.
Solution
The main idea is to take a gaussian prime and multiply it by a "twice as small"
to get
. The rest is just making up the little details.
For each sufficiently large prime of the form
, we shall find a corresponding
satisfying the required condition with the prime number in question being
. Since there exist infinitely many such primes and, for each of them,
, we will have found infinitely many distinct
satisfying the problem.
Take a prime of the form
and consider its "sum-of-two squares" representation
, which we know to exist for all such primes. As
, assume without loss of generality that
. If
, then
is what we are looking for, and
as long as
(and hence
) is large enough.
Assume from now on that
.
Since and
are (obviously) co-prime, there must exist integers
and
such that
In fact, if
and
are such numbers, then
and
work as well for any integer
, so we can assume that
.
Define and let's see why this was a good choice. For starters, notice that
.
If , then from (1), we see that
must divide
and hence
. In turn,
and
. Therefore,
and so
, from where
. Finally,
and the case
is cleared.
We can safely assume now that
As
implies
, we have
so
Before we proceed, we would like to show first that . Observe that the function
over
reaches its minima on the ends, so
given
is minimized for
, where it equals
. So we want to show that
which obviously holds for large
.
Now armed with and (2), we get
where
Finally,