Zsigmondy's Theorem

by KingSmasher3, Oct 10, 2013, 3:02 AM

Continuing my NT adventures...

ISL 2000 N4: Find all triplets of positive integers $ (a,m,n)$ such that $ a^m + 1 \mid (a + 1)^n$.
_______

From Zsigmondy's Theorem, $a^m+1$ has a prime factor that is not a factor of $a+1,$ unless $a=2$ and $m=3.$ Hence, we see that the given statement holds only if $(a,m,n)=(2,3,n)$ with $n\ge 2.$


ISL 1997 #14: Let $ b, m, n$ be positive integers such that $ b > 1$ and $ m \neq n.$ Prove that if $ b^m - 1$ and $ b^n - 1$ have the same prime divisors, then $ b + 1$ is a power of 2.
_______

Without loss of generality, let $n>m.$ By Zsigmondy's Theorem, $b^n-1$ has a prime factor that is not a factor of $b^m-1$ unless $b=2, n=6$ or $b+1$ is a power of $2$ and $n=2.$ We can check that the first case does not yield any possible value of $m,$ so we are done.


MOP 2001: Find all quadruples of positive integers $(x, r, p, n)$ such that $p$ is a prime number, $n, r > 1$ and $x^r-1 = p^n.$
_______

Obviously $x \neq 1.$

If $x=2,$ and $r$ is composite, then by Zsigmondy's theorem, $x^r-1$ cannot be a power of a prime. The only exception to this theorem occurs when $r=6,$ which clearly does not work. Therefore, $r$ must be prime. We have $p^n+1=2^r,$ implying that $p^n \equiv 3\mod 4.$ It follows that $n$ must be odd, so $p+1 | (p+1)^2=n.$ Thus, $p+1=2^b$ with $b<r.$ Now we know that $\text{ord}_p(2) | b$ and $\text{ord}_p(2) | r,$ so $r$ cannot be prime.

If $x\ge 3,$ by Zsigmondy's theorem, $x^r-1$ cannot be a power of a prime. The only exception occurs when $x+1$ is a power of $2$ and $r=2.$ Let $x=2^m-1.$ Then $(2^m-1)^2-1=p^n,$ or $2^{2m}-2{m+1}=p^n.$ It is clear that this only holds when $m=2, p=2, n=3.$

Hence, our only solution is $(x,r,p,n)=(3,2,2,3).$


ISL 2002 N3: Let $p_1,p_2,\ldots,p_n$ be distinct primes greater than $3$. Show that $2^{p_1p_2\cdots p_n}+1$ has at least $4^n$ divisors.
_______

Let $q$ be the product of a subset of $p_1, p_2, ..., p_n.$ Then clearly, $2^q+1 | 2^{p_1p_2\cdots p_n}+1.$ Now, let $N=2^{p_1p_2\cdots p_n}+1$ and $N'=2^{p_1p_2\cdots p_{n-1}}+1.$ Since there are $2^n$ possible values for $q,$ by Zsigmondy's Theorem, $N$ has at least $2^n$ distinct prime factors, which means at least $2^{2^n} \ge 4^n$ total factors. Note that the exceptional case $2^3+1$ does not occur because all primes $p_i$ are greater than $3.$

Alternate Solution: We induct on $n.$ Clearly when $n=1,$ $x=2^p+1$ has at least $4$ factors: $1, 3, x/3, x.$ Now given two integers $a, b$ with $\gcd(a,b)=1,$ we have

\[\gcd(2^a+1, 2^b+1) | \gcd(2^{2a}-1,2^{2b}-1) = 2^{\gcd(2a,2b)}-1 = 3.
\]
This can be seen by applying the Euclidean Algorithm. Furthermore, we can easily show that $2^y+1$ is a multiple of $3$ but not $9$ if and only if $3 \nmid y.$ We have $N=kN'$ and $2^{p_n}+1$ is a factor of $N$ and thus $k,$ so by the two facts stated above, $(2^{p_n}+1)/3$ is a factor of $k$ and is relatively prime to $N'.$ It follows that $Q=N'(2^{p_n}+1)/3$ has at least $2 \cdot 4^{n-1}$ factors. It can be seen that $N>Q^2,$ so $N$ has at least $4^n$ factors (for every divisor $d|Q,$ we have $d, \frac{N}{d} | N$) as desired.
This post has been edited 6 times. Last edited by KingSmasher3, Jan 5, 2014, 7:07 PM

IMO Shortlist NT

by KingSmasher3, Sep 6, 2013, 3:13 AM

1995 NT #1: Let $ k$ be a positive integer. Show that there are infinitely many perfect squares of the form $ n \cdot 2^k - 7$ where $ n$ is a positive integer.
_______

Clearly, for each $k,$ if there is one perfect square that is $-7\pmod{2^k}$ then there are infinitely many perfect squares. We proceed with induction on $k$. When $k=1,$ we have $1^2 \equiv -7\pmod{2^1}.$ Suppose that the result holds for some $k=j.$ Then we have $2^j | a^2+7$ for some $a.$ If $2^{j+1} \not{|} a^2+7,$ we try to find $b$ such that $2^{j+1} | (a+b)^2+7.$ It is clear that such a $b$ exists because we need for $2ab+b^2 \equiv 2^j \pmod{2^{j+1}},$ which has roots, such as $b=2^j.$ Our induction is complete. Therefore, the result holds for all $k.$

1999 NT #3: Prove that there exists two strictly increasing sequences $(a_{n})$ and $(b_{n})$ such that $a_{n}(a_{n}+1)$ divides $b^{2}_{n}+1$ for every natural n.
_______

Note that for each $a_n,$ if there exists a $b_n$ such that $a_{n}(a_{n}+1) | b^{2}_{n}+1,$ or equivalently, if $-1$ is a quadratic residue mod $a_{n}(a_{n}+1),$ then we can find a $b_n$ such that $b_n>b_{n-1}.$ Additionally, it is clear that if $-1$ is a quadratic residue of $a_n$ and $a_n+1$ then it is a quadratic residue of $a_n(a_n+1).$ It claim that this is true for all $a_n=5^{2k}, k\in \mathbb{Z}^+.$ In fact, we can generalize this to all primes $1\pmod{4},$ not just $5.$ Clearly, $-1$ is now a quadratic residue of $a_n+1.$ We now prove the stronger statement that $-1$ is a quadratic residue of all $a_n=5^k.$ We proceed by induction on $k.$ It is true for $k=1.$ Assume that it holds for some $k=j.$ It follows that $5^j | a^2+1$ for some $a.$ If $5^{j+1} \not{|} a^2+1,$ we find $b$ such that $5^{j+1} | (a+b)^2+1.$ Let $b=m5^j.$ Then we have $5^{j+1} | a^2+1+2ab+b^2,$ or $5 | (a^2+1)/5^j+2am.$ Since we can obviously find such an $m,$ our induction is complete, and we are done.

Main point(s): These problems demonstrate a trick in using induction to determine the existence of certain quadratic residues for powers.
This post has been edited 2 times. Last edited by KingSmasher3, Oct 10, 2013, 3:10 AM

USAMO 2006 Problem #2

by KingSmasher3, Apr 13, 2013, 3:54 AM

For a given positive integer $k$ find, in terms of $k$, the minimum value of $N$ for which there is a set of $2k + 1$ distinct positive integers that has sum greater than $N$ but every subset of size $k$ has sum at most $\frac{N}{2}.$
_______

Let the $2k+1$ distinct positive integers be $a_1, a_2, a_3, ..., a_{2k+1}$ in increasing order. Since the integers are distinct,

\[\frac{N}{2} \ge \sum_{i=k+2}^{2k+1} a_i \ge \sum_{i=1}^{k-1} a_i + k(k+1).\]
We are also given,

\[ a_{k+1} + \sum_{i=k+2}^{2k+1} a_i + \sum_{i=1}^{k-1} a_i > N.\]
These two inequalities tell us that $a_{k+1}>k(k+1),$ or $a_{k+1}\ge k(k+1)+1.$ Hence,

\[N \ge 2\sum_{i=k+2}^{2k+1} a_i \ge \sum_{i=1}^{k} (a_{k+1}+i) \ge 2k^3+3k^2+3k.\]
We can check that this minimal value does indeed hold for the set $\{k^2+1, k^2+2, k^2+3, ..., k^2+2k+1\}.$ We are done.

Main Point(s): The fact that the $2k+1$ integers are distinct is very useful in determining the bounds for $N.$ Compare max subset of $k$ and min subset of $k.$
This post has been edited 1 time. Last edited by KingSmasher3, Apr 13, 2013, 3:57 AM

USAMO 2010 Problem #5

by KingSmasher3, Apr 10, 2013, 11:08 PM

Let $q = \frac{3p-5}{2}$ where $p$ is an odd prime, and let\[
S_q = \frac{1}{2\cdot 3 \cdot 4} + \frac{1}{5\cdot 6 \cdot 7} + \cdots + \frac{1}{q(q+1)(q+2)}
\]Prove that if $\frac{1}{p}-2S_q = \frac{m}{n}$ for integers $m$ and $n$, then $m - n$ is divisible by $p$.
_______

First note that for any $k \in \mathbb{Z}^+,$ we have

\[\frac{2}{(k-1)(k)(k+1)}=\frac{1}{k-1}-\frac{2}{k}+\frac{1}{k+1}=\frac{1}{k-1}+\frac{1}{k}+\frac{1}{k+1}-\frac{3}{k}.
\]
Applying this to the problem,

\[2S_q=\sum_{i=2}^{\frac{3p-1}{2}} \frac{1}{i} - \sum_{i=1}^{\frac{p-1}{2}} \frac{1}{i}=\sum_{i=\frac{p+1}{2}}^{\frac{3p-1}{2}}-1.\]
It follows that,

\[\frac{1}{p}-2S_q= 1+\frac{1}{p}-\sum_{i=\frac{p+1}{2}}^{\frac{3p-1}{2}}.\]
Consequently,

\[\frac{m-n}{n}=\frac{1}{p}-2S_q-1= \frac{1}{p}-\sum_{i=\frac{p+1}{2}}^{\frac{3p-1}{2}}
\] \[=-\left(\frac{1}{\frac{p+1}{2}}+\frac{1}{\frac{p+3}{2}} + \cdots + \frac{1}{p-1}+\frac{1}{p+1}+ \cdots \frac{1}{\frac{3p-1}{2}}\right).\]

Now, the numerator of (the absolute value of) this sum is

\[\sum_{i=\frac{p+1}{2}}^{p-1} \frac{P}{i} + \sum_{i=p+1}^{\frac{3p-1}{2}} \frac{P}{i} = \sum_{j=1}^{\frac{p-1}{2}} \frac{P}{p-j} + \frac{p+j},\]
where $P=\left(\frac{p+1}{2}\right)\left(\frac{p+3}{2}\right) \cdots (p-1)(p+1) \cdots \left(\frac{3p-1}{2}\right).$ This sum is clearly $0\pmod{p},$ so $m-n$ is a multiple of $p$ as desired.

Main Point(s): When given a product in the denominator such as this one, consider using partial fraction decomposition to greatly simplify the problem.

[img]http://s03.flagcounter.com/count/OsVD/bg_E8F2FF/txt_000000/border_CCCCCC/columns_2/maxflags_18/viewers_0/labels_1/pageviews_0/flags_1/[/img][/url]

avatar

KingSmasher3
Shouts
Submit
  • orz blog!!!

    by KevinChen_Yay, Dec 29, 2024, 1:39 AM

  • 1 year bump lol

    by Yiyj1, Mar 1, 2024, 1:55 AM

  • 2 year bump rip

    by cinnamon_e, Mar 25, 2023, 5:46 PM

  • Bumpity bump

    by mathboy282, Dec 13, 2020, 9:38 PM

  • NT God Please Return!

    by Pluto1708, Mar 20, 2019, 2:25 PM

  • dude,your blog is awesome.Please don't stop and continue your posts!! :)

    by Jiminhio 10, Jan 16, 2014, 2:11 PM

  • thanks haha

    by KingSmasher3, Sep 6, 2013, 3:15 AM

  • happy birthday

    by cire_il, Sep 3, 2013, 9:10 PM

  • btw im totally not trolling

    dude problem 2s are so hard
    what is this madness
    what is going on
    hehe
    we are not spamming up your blog like it might seem at first
    lalalalalalalalala

    by applepi2000, Jun 25, 2013, 12:31 AM

  • Hmm USAMO so hard

    by antimonyarsenide, Jun 25, 2013, 12:28 AM

  • dude you do problem 2s?
    dude so legit man
    i am not trolling

    by applepi2000, Jun 25, 2013, 12:28 AM

  • Hmm USAMO so hard

    by antimonyarsenide, Jun 25, 2013, 12:28 AM

  • dude you do problem 2s?
    dude so legit man
    i am not trolling

    by applepi2000, Jun 25, 2013, 12:27 AM

  • Hey look an excellent problem blog.

    It contains a bunch of USAMO problems that are familiar to me because I did them a couple months ago.

    by yugrey, Apr 5, 2013, 1:55 AM

  • Are you my mommy?

    by meisepic, Mar 26, 2013, 6:47 AM

  • too much math

    by cire_il, Mar 26, 2013, 3:15 AM

  • Yay you made a blog!

    by dinoboy, Mar 19, 2013, 4:29 AM

17 shouts
Tags
About Owner
  • Posts: 1399
  • Joined: Jul 16, 2009
Blog Stats
  • Blog created: Nov 11, 2012
  • Total entries: 26
  • Total visits: 14840
  • Total comments: 14
Search Blog
a