How can factorization prove the compositeness of a given expression?

In summary, we can prove that for every positive integer $a$, the integer $5^{4a}+5^{3a}+5^{2a}+5^a+1$ is composite. To do this, we can use the fact that if $a$ is not a multiple of $5$, then the numbers $a$, $2a$, $3a$, $4a$ are congruent $\pmod5$ to $1$, $2$, $3$, $4$, in some order. From this, we can work mod $11$ and $71$ to show that $5^{4a}+5^{3a}+5^{2a}+5^a+
  • #1
anemone
Gold Member
MHB
POTW Director
3,883
115
Prove that for every positive integer $a$, the integer $5^{4a}+5^{3a}+5^{2a}+5^a+1$ is composite.
 
Mathematics news on Phys.org
  • #2
Hint:

Try to think along the line of the following expression:

$x^8+x^6+x^4+x^2+1$
 
  • #3
I generally do not provide a solution leaving out a few cases but in this case I could cover all the cases with one exception

we have the polynomial
$P(a)=\dfrac{5^{5a}-1}{5^a-1}$

There are 3 cases
a is 1 so P(1) = 781 = 11 * 71 not a prime

a is even say 2b

we get
$P(x)=\dfrac{5^{10b}-1}{5^{2b}-1} = \dfrac{5^{5b}-1}{5^{b}-1} \dfrac{5^{5b}+1}{5^b+1}$
$= \dfrac{(5^{4b}+5^{3b} + 5^{2b}+5^b+1)(5^{5b}+1)}{5^b+1}$

now we see that $5^b+1$ must divide the numerator and it is not same as any term of the numerator as both terms are bigger so we shall be left with 2 terms in numerator and it is not a prime

Now we need to consider when a is odd and > 1

then
numerator can be expressed as product in 2 different ways 1) $(5^a-1)(1+5^a+5^{2a}+5^{3a} + 5^{4a})$2) $(5^5-1)(1+5^5+5^{2*5}+5^{3*5} + 5^{5*(a-1)})$

now the denominator cannot be cancelling one of the terms in numerator in both the cases and hence this is not a prime

but with one exception that is a = 5.

this is left out
 
  • #4
kaliprasad said:
I generally do not provide a solution leaving out a few cases but in this case I could cover all the cases with one exception

we have the polynomial
$P(a)=\dfrac{5^{5a}-1}{5^a-1}$

There are 3 cases
a is 1 so P(1) = 781 = 11 * 71 not a prime

a is even say 2b

we get
$P(x)=\dfrac{5^{10b}-1}{5^{2b}-1} = \dfrac{5^{5b}-1}{5^{b}-1} \dfrac{5^{5b}+1}{5^b+1}$
$= \dfrac{(5^{4b}+5^{3b} + 5^{2b}+5^b+1)(5^{5b}+1)}{5^b+1}$

now we see that $5^b+1$ must divide the numerator and it is not same as any term of the numerator as both terms are bigger so we shall be left with 2 terms in numerator and it is not a prime

Now we need to consider when a is odd and > 1

then
numerator can be expressed as product in 2 different ways 1) $(5^a-1)(1+5^a+5^{2a}+5^{3a} + 5^{4a})$2) $(5^5-1)(1+5^5+5^{2*5}+5^{3*5} + 5^{5*(a-1)})$

now the denominator cannot be cancelling one of the terms in numerator in both the cases and hence this is not a prime

but with one exception that is a = 5.

this is left out

Can you explain how $5^{5a} - 1 = (5^5-1)(1+5^5+5^{2*5}+5^{3*5} + 5^{5*(a-1)})$? Also, I am not sure your proof works for $a$ power of 5 (not just $a = 5$ but every power of 5) can you verify.
 
  • #5
anemone said:
Prove that for every positive integer $a$, the integer $5^{4a}+5^{3a}+5^{2a}+5^a+1$ is composite.
Another partial solution.
[sp]If $a$ is not a multiple of $5$ then the numbers $a$, $2a$, $3a$, $4a$ are congruent $\pmod5$ to $1$, $2$, $3$, $4$, in some order. Now working mod $11$, you find that $5^a$, $5^{2a}$, $5^{3a}$, $5^{4a}$ are congruent $\pmod{11}$ to $5$, $3$, $4$, $9$. It follows that $5^{4a}+5^{3a}+5^{2a}+5^a+1 \equiv 9+4+3+5+1 = 22 \equiv0 \pmod{11}$.

The same result applies if you work mod $71$ instead of mod $11$, because both $p=11$ and $p=71$ have the property that $5^5\equiv1\pmod p$.

It follows that if $a$ is not a multiple of $5$ then $5^{4a}+5^{3a}+5^{2a}+5^a+1$ is divisible by $11\times 71 = 781$. But I could not find any way to prove the result when $a$ is a multiple of $5$.[/sp]
 
  • #6
Bacterius said:
Can you explain how $5^{5a} - 1 = (5^5-1)(1+5^5+5^{2*5}+5^{3*5} + 5^{5*(a-1)})$? Also, I am not sure your proof works for $a$ power of 5 (not just $a = 5$ but every power of 5) can you verify.

This is based on expansion of

$a^{mn} - 1= (a^m)^n - 1= (a^m-1)(1 + a^m +\cdots + a^{m(n-1)})$
 
  • #7
kaliprasad said:
This is based on expansion of

$a^{mn} - 1= (a^m)^n - 1= (a^m-1)(1 + a^m +\cdots + a^{m(n-1)})$

Ah, I see, thanks, an ellipsis was missing. Though I think that still is not strong enough to solve the problem for $a$ power of 5, since in that case $5^a - 1$ may very well be composite and automatically is a multiple of $5^5 - 1$, leaving open whether $1 + 5^5 + 5^{2 \cdot 5} + \cdots$ is prime or not, unless I misunderstood.

Here is another partial solution slightly stronger than Opalg's:

Suppose $5^{4a} + 5^{3a} + 5^{2a} + 5^a + 1$ is composite for some $a \geq 1$. Then for every prime $p$ dividing that expression we have:
$$5^{4a} + 5^{3a} + 5^{2a} + 5^a + 1 \equiv 0 \pmod{p}$$
That is:
$$\frac{5^{5a} - 1}{5^a - 1} \equiv 0 \pmod{p}$$
Which implies that:
$$5^{5a} \equiv 1 \pmod{p} ~ ~ ~ \text{and} ~ ~ ~ 5^a \not \equiv 1 \pmod{p}$$
Observe that these equations imply $5^a$ has order 5 modulo $p$, since 5 is prime. Now let $k$ be an integer not multiple of 5, then we have:
$$\left ( 5^{5a} \right )^k \equiv 5^{5ak} \equiv 1 \pmod{p}$$
And since $k$ is not a multiple of 5 and $5^a$ has order 5 modulo $p$, it must follow that:
$$\left ( 5^{a} \right )^k \equiv 5^{ak} \not \equiv 1 \pmod{p}$$
So we conclude that:
$$\frac{5^{5ak} - 1}{5^{ak} - 1} \equiv 0 \pmod{p}$$
Such that:
$$5^{4ak} + 5^{3ak} + 5^{2ak} + 5^{ak} + 1 \equiv 0 \pmod{p}$$
And so we conclude that $f(ak)$ is a multiple of $f(a)$ and so is also composite. With the above theorem it suffices to show that the expression is composite for powers of 5. In particular, it is composite at $a = 1 = 5^0$, hence is composite at all $a$ not multiple of 5, and it can be verified that it is also composite at $a = 5$, hence it is composite at all $a$ not multiple of 25. We can keep going and check that the expression is composite for $a = 25$ as well, showing that the expression is composite for all $a$ not multiple of 125, but I don't know how to prove that it must be composite for all powers of 5, which is required to show it is composite for all integer $a$.​
 
  • #8
Thanks to kaliprasad, Opalg, and Bacterius for trying and partially solving this challenge!:)

I have a solution at hand that merely used the factorization technique to prove the given expression is a composite, and I will share it here now:

Observe the following representations:

$x^8+x^6+x^4+x^2+1=(x^4+x^3+x^2+x+1)(x^4-x^3+x^2-x+1)$---(1)

and

$x^4+x^3+x^2+x+1=(x^2+3x+1)^2-5x(x+1)^2$---(2)

When $n=2a$ is even, we can substitute $x=5^a$ into equation (1) to get a factorization.

When $n=2a-1$ is odd, we can substitute $x=5^{2a-1}$ into equation (2) to get a difference of squares, which can then be factored.
 

FAQ: How can factorization prove the compositeness of a given expression?

1. How can I prove that a sum is composite?

To prove that a sum is composite, you must show that it can be written as a product of two or more smaller numbers. This can be done by factoring the sum into its prime factors. If the sum can be written as a product of at least two different numbers, then it is composite.

2. What is the definition of a composite number?

A composite number is a positive integer that can be divided evenly by at least one number other than itself and 1. In other words, it has more than two factors. Examples of composite numbers include 4, 9, and 15.

3. Can a sum of two prime numbers be composite?

No, a sum of two prime numbers cannot be composite. This is because prime numbers can only be divided evenly by 1 and themselves, so the sum of two prime numbers will only have two factors and is therefore not composite.

4. Is 1 considered a composite number?

No, 1 is not considered a composite number. This is because it only has one factor, itself, and does not meet the definition of a composite number.

5. How do I prove that a sum is not composite?

If a sum cannot be written as a product of two or more smaller numbers, then it is not composite. In other words, if you cannot factor the sum into its prime factors, then it is not composite.

Similar threads

Replies
1
Views
916
Replies
6
Views
2K
Replies
2
Views
986
Replies
3
Views
1K
Replies
5
Views
1K
Back
Top