Large Prime Factors!

Algebra Level 4

Given that there are three prime factors to the following expression and the smallest prime factor is the only prime factor which is a 2 2 -digit number. Determine the smallest prime factor of

1 + 2 21 + 4 21 \large 1+2^{21}+4^{21}

Image Credit: Wikimedia Prime


The answer is 73.

This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try refreshing the page, (b) enabling javascript if it is disabled on your browser and, finally, (c) loading the non-javascript version of this page . We're sorry about the hassle.

9 solutions

Christopher Boo
Apr 13, 2014

The expression can be rewrite as

1 21 + 2 21 + 4 21 1^{21}+2^{21}+4^{21}

We use Vieta's Formula to construct a new equation with roots p , q , r p,q,r where p = 1 3 , q = 2 3 , r = 4 3 p=1^3, q=2^3, r=4^3 , we will have

x 3 + 73 x 2 + 584 x + 512 = 0 x^3+73x^2+584x+512=0

Similarly, we need to find the smallest prime factor of

P 7 = p 7 + q 7 + r 7 P_7=p^7+q^7+r^7

By Newton's Sum ,

P 1 73 = 0 P_1-73=0

P 2 73 P 1 + 1168 = 0 P_2-73P_1+1168=0

P 3 73 P 2 + 584 P 1 1536 = 0 P_3-73P_2+584P_1-1536=0

P 4 73 P 3 + 584 P 2 512 P 1 = 0 P_4-73P_3+584P_2-512P_1=0

P 5 73 P 4 + 584 P 3 512 P 2 = 0 P_5-73P_4+584P_3-512P_2=0

P 6 73 P 5 + 584 P 4 512 P 3 = 0 P_6-73P_5+584P_4-512P_3=0

P 7 73 P 6 + 584 P 5 512 P 4 = 0 P_7-73P_6+584P_5-512P_4=0

From the last equation, we know that 73 P 6 -73P_6 and 584 P 5 584P_5 are both divisible by 73 73 . The only left to investigate is 512 P 4 512P_4 .

From the forth and first equation, which states that P 1 P_1 is divisible by 73 73 , we know that P 4 P_4 is divisible by 73 73 .

Conclude everything above, we get that P 7 P_7 is divisible by 73 73 .

@Finn Hulse , any idea?

Christopher Boo - 7 years, 2 months ago

Log in to reply

Yes! I let x = 2 21 x=2^{21} and y = 1 y=1 and factored. I'm sure there's a better way to do it, but I pretty much just tried to find ways of manipulating the variables to allow for factorization. Yes, I did have to resort to Newton's Sums, but I got lucky. :D

Finn Hulse - 7 years, 2 months ago

Log in to reply

FYI, this problem was actually from an International Mathematics Competition and no participants (except SOME but that's a long story) can solve this.

Christopher Boo - 7 years, 2 months ago

Log in to reply

@Christopher Boo Woah. Hey, congrats on making the "Who to follow" page! I was kind of disappointed about not making it, but whatever.

Finn Hulse - 7 years, 2 months ago

@Finn Hulse OMG! I did the same thing. Just when I thought I would write a comment here, you were ahead of me!

Ameya Salankar - 7 years, 1 month ago

Log in to reply

@Ameya Salankar ;D

Finn Hulse - 7 years, 1 month ago

Per your instructions, I used the same method.

Trevor B. - 7 years, 2 months ago

Log in to reply

Same here.

Sharky Kesa - 7 years, 1 month ago
Adit Mohan
Apr 14, 2014

the expression can be expressed as 2 63 2^{63} -1 / 2 21 2^{21} -1.
2 9 2^{9} -1 is a factor of 2 63 2^{63} -1 but not 2 21 2^{21} -1 ( 2 n 2^{n} -1 is a factor of 2 a n 2^{an} -1).
2 9 2^{9} -1= 511 = 7(73).
7= 2 3 2^{3} -1 is a factor of 2 21 2^{21} -1 however, hence 73 is not a factor.
hence after division 73 is a factor of the expression.



I did it in the same way.

Marius Munteanu - 7 years, 1 month ago

yeah....me too

Rutvik Paikine - 7 years, 1 month ago

that's a pretty good method i must say :)

Sachin Mourya - 7 years, 1 month ago
Michael Mendrin
Apr 13, 2014

This is not an infallible or efficient method for finding such prime factors, but this one worked for me. Let a = 2 3 { a=2 }^{ 3 } so that we have 1 + a 7 + a 14 1+{ a }^{ 7 }+{ a }^{ 14 } . We note that 1 + a + a 2 = 73 1+{ a }+{ a }^{ 2 }=73 , a prime number, so we hope that 1 + a + a 2 1+{ a }+{ a }^{ 2 } divides into 1 + a 7 + a 14 1+{ a }^{ 7 }+{ a }^{ 14 } . After some foodling around, we see that it does, i..e ( 1 + a + a 2 ) ( ( 1 + a 3 + a 6 + a 9 + a 12 ) ( a + a 4 + a 8 + a 11 ) ) = ( 1 + a 7 + a 14 ) (1+a+{ a }^{ 2 })((1+{ a }^{ 3 }+{ a }^{ 6 }+{ a }^{ 9 }+{ a }^{ 12 })-(a+{ a }^{ 4 }+{ a }^{ 8 }+{ a }^{ 11 }))=(1+{ a }^{ 7 }+{ a }^{ 14 }) .
And so, I entered "73" as the solution to see what happens.

It's one thing to take a stab at a solution, quite another to prove beyond a doubt that it is indeed the correct one.

You don't have to try real hard to find out if an expression is divisible by a 2 + a + 1 a^2+a+1 . Here's a cool trick.

Let 1 + a 7 + a 14 1+a^7+a^{14} be equal to f ( a ) f(a) . Notice that f ( ω f(\omega ) and f ( ω 2 ) f({\omega}^2) are both equal to zero [ ω \omega and ω 2 \omega ^2 are the cubic roots of 1 1 that have non-zero imaginary part]. That means,

f ( a ) = g ( a ) ( a ω ) ( a ω 2 ) = g ( a ) ( a 2 + a + 1 ) f(a)=g(a)(a-\omega)(a-\omega^2)=g(a)(a^2+a+1) .

Mursalin Habib - 7 years, 2 months ago

Log in to reply

Not all 1 + a n + a 2 n 1+{ a }^{ n }+{ a }^{ 2n } is divisible by 1 + a + a 2 1+{ a }+{ a }^{ 2 } , so I don't quite understand how to determine which integer n it is, using your method. But the subject of factorizing 1 + a n + a 2 n 1+{ a }^{ n }+{ a }^{ 2n } is an interesting one, and not as straightforward as factorizing 1 + a n 1+{ a }^{ n } , possible only for odd n.

Michael Mendrin - 7 years, 2 months ago

Log in to reply

Mursalin's idea is to use the Remainder Factor Theorem.

Observe that ( x ω ) ( x ω 2 ) = x 2 + x + 1 ( x- \omega) ( x - \omega^2) = x^2 + x + 1 . Consider f n ( x ) = x 2 n + x n + 1 f_n (x) = x^{2n} + x^n + 1 . Suppose that x 2 + x + 1 f n ( x ) x^2 + x + 1 \mid f_n (x) , then this means that f n ( x ) = ( x ω ) ( x ω 2 ) g n ( k ) f_n(x) = (x - \omega) ( x - \omega^2) g_n(k) , hence f n ( ω ) = 0 , f ( ω 2 ) = 0 f_n(\omega) = 0, f( \omega^2 ) = 0 .

Checking f n ( ω ) = ω 2 n + ω n + 1 f_n (\omega ) = \omega^{2n} + \omega^n + 1 , this value is 0 n 1 , 2 ( m o d 3 ) n \equiv 1, 2 \pmod{3} and is equal to 3 if n 0 ( m o d 3 ) n \equiv 0 \pmod{3} . The same holds for f n ( ω 2 ) f_n(\omega^2) . Hence, f n ( x ) f_n (x) is divisible by x 2 + x + 1 x^2 + x + 1 if and only if n n is not a multiple of 3.

Calvin Lin Staff - 7 years, 2 months ago

Log in to reply

@Calvin Lin In this vein, then, given that the problem states there is one prime factor with just 2 digits, wouldn't this be an easier way to solve the problem? I think Boo made the problem easier by saying that one of the prime factors is just 2 digits long. We could quickly rule out the single digit prime factors 2, 3, 5, 7.

Michael Mendrin - 7 years, 2 months ago

Log in to reply

@Michael Mendrin Yes. I believe this approach (or something similar), is what the competition organizers were expecting. The facts given in the question are hard to know, and the problem would be hard to solve without them.

Calvin Lin Staff - 7 years, 2 months ago
Patrick Corn
Apr 15, 2014

The number is 2 63 1 2 21 1 \dfrac{2^{63}-1}{2^{21}-1} . Since x n 1 = d n Φ d ( x ) x^n - 1 = \prod_{d|n} \Phi_d(x) where Φ d \Phi_d is the d d th cyclotomic polynomial, we get that the number is Φ 9 ( 2 ) Φ 63 ( 2 ) \Phi_9(2)\Phi_{63}(2) .

Now Φ 9 ( x ) = x 6 + x 3 + 1 \Phi_9(x) = x^6+x^3+1 , so Φ 9 ( 2 ) = 73 \Phi_9(2) = \fbox{73} . The stipulations of the problem allow us to stop there and be done.

Sameer Arora
Apr 14, 2014

the following expression 1 + 2^21 + 4^21 can be restated as:

1^3 + (2^3)^7 + (4^3)^7 which is divisible by (1^3 + 2^3 + 4^3 = 73)

thus, the answer is 73

Jack D'Aurizio
Apr 14, 2014

Since n = 1 + 2 21 + 4 21 n=1+2^{21}+4^{21} is a number of the form x 2 + x + 1 x^2+x+1 , 3 -3 is a quadratic residue for every prime divisor of n n , hence all the prime divisors of n n are numbers of the form 3 k + 1 3k+1 , and we have only to check divisibility by the elements of the following set: { 13 , 19 , 31 , 37 , 43 , 61 , 67 , 73 , 79 , 97 } \{13,19,31,37,43,61,67,73,79,97\} . We notice that p n p\mid n implies p x 3 1 = 2 63 1 p\mid x^3-1 = 2^{63}-1 , or 2 63 1 ( m o d p ) 2^{63}\equiv 1\pmod{p} . We rule out 13 13 since 2 63 2 3 ≢ 1 ( m o d 13 ) 2^{63}\equiv 2^3\not\equiv 1\pmod{13} ; we rule out 19 19 since 2 63 1 2 9 1 ≢ 1 ( m o d 19 ) 2^{63}-1\equiv 2^9-1\not\equiv 1\pmod{19} and so on, till 73 73 . The order of 2 2 mod 73 73 is just 9 9 , so n 1 + 2 3 + 2 6 0 ( m o d 73 ) n\equiv 1+2^3+2^6\equiv 0\pmod{73} .

Shyan Akmal
Apr 17, 2014

Let Φ n ( x ) \Phi_n(x) denote the n n th cyclotomic polynomial.

Then number in question is Φ 3 ( 2 21 ) = Φ 9 ( 2 7 ) = Φ 9 ( 2 ) Φ 63 ( 2 ) . \Phi_3\left(2^{21}\right)=\Phi_9\left(2^7\right)=\Phi_9(2)\Phi_{63}(2).

Thus, Φ 9 ( 2 ) = 2 6 + 2 3 + 1 = 73 \Phi_9(2)=2^6+2^3+1=73 is a prime divisor. Since it has two digits, we're done.

Shane Arnold
Apr 17, 2014

I used Mersenne numbers as my guide. Others below used the same idea if not the name. The number given is equivalent to M63 / M21. M63 is divisible by both M7 and M9. M21 is divisible by both M7 and M3. So the number M9 / M3 should divide evenly into M63 / M21. M9 / M3 = 2^6 + 2 ^3 +1 = 73. So 73 is a factor of the original number given.

Phil Peters
Apr 13, 2014

I know this isn't a solution or a hint, but what I did was factorized (difference of 2 cubes), then got one expression over another. That should get you somewhere...

I'm not sure if this can lead us to the solution, can you explain more?

Christopher Boo - 7 years, 2 months ago

Log in to reply

Sorry, I really don't know, I was just throwing in an idea (possible one).

Phil Peters - 7 years, 2 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...