A question regarding Goldbach's conjecture and an equivalent way of writing it

70 Views Asked by At

I was thinking about how a even number, $2n$, can be represented as the sum of two natural numbers in $n$ ways, and that $\forall a \in \Bbb N \Rightarrow \exists b \in \Bbb N:a+b=2n$, where $a,b<2n$. I thought, that if we consider that $Q(n)$ represents the amount of ways in which you can write $2n$ as the sum of two composite numbers, and we consider that for $a + b$ to be equal to $2n$, $2n-a=b$, such that, if both $a$ and $b$ are composite, $a$ has to be a divisor of $2n$, so that $b$ is also a composite number, and now if we consider the prime factorization of $2n$ as: $$2n=2(p_1^{a_1}p_2^{a_2}\cdots p_k^{a_k})$$

We can now conclude, that we can write $Q(n)$ as a sum of the amount of distinct multiples up to $n$ of every distinct prime factor of $2n$, except for themselves, plus a term that represents weather $2n-1$ is composite or not, since we are not going to be able to reach that representation in another way, so, let $a(x)$ represent weather $x$ is prime or not, such that: $$a(x)= \begin{cases} 1, & \text{if $x$ prime} \\ 0, & \text{if otherwise} \end{cases}$$

Thus, $Q(n)$ can be written as: $$Q(n)= \sum_{i=1}^k (\lfloor\frac{n}{p_i}\rfloor-1)-\sum_{i=1}^k\sum_{j=1\;i \neq j}^k \lfloor\frac{n}{p_ip_j}\rfloor+1-a(2n-1)$$ Now, if we assume, that $2n$ cannot be expressed as the sum of two prime numbers, we conclude that every way of writing $2n$ as the sum of two numbers has to be in one of three forms only: $$ 2n= \begin{cases} prime + composite \\ composite + prime \\ composite + composite \end{cases} $$ And since $\forall a \in \Bbb N \Rightarrow \exists b \in \Bbb N:a+b=2n$, where $a,b<2n$, iff $2n$ is not expressible as the sum of two primes, we have that: $$Q(n)+\pi(2n)=n$$

And as we know, we can rewrite this as: $$\sum_{i=1}^k (\lfloor\frac{n}{p_i}\rfloor-1)-\sum_{i=1}^k\sum_{j=1\;i \neq j}^k \lfloor\frac{n}{p_ip_j}\rfloor+1-a(2n-1)=n-\pi(2n)$$ Thus, we can conclude, that Goldbach's conjecture is the same as: $$\sum_{i=1}^k (\lfloor\frac{n}{p_i}\rfloor-1)-\sum_{i=1}^k\sum_{j=1\;i \neq j}^k \lfloor\frac{n}{p_ip_j}\rfloor+1-a(2n-1) \neq n-\pi(2n)$$ Now, to extend this to a proof of Goldbach's conjecture I do not know how, so, if you can answer this, it'd be amazing, I'm just looking for an answer. And also, I'm sorry if the title isn't very accurate.

1

There are 1 best solutions below

0
On BEST ANSWER

One of your earlier statements is

... we consider that for $a + b$ to be equal to $2n$, $2n-a=b$, such that, if both $a$ and $b$ are composite, $a$ has to be a divisor of $2n$, so that $b$ is also a composite number ...

However, it's not true that if both $a$ and $b$ are composite that $a$ has to be a divisor of $2n$. For example, if $n = 17$, then note $a = 25$ and $b = 9$ gives $25 + 9 = 2(17) = 34$, but neither $a$ or $b$ are a divisor of $2n$.

With what you stated further on being based on this, it means the rest of your statements afterwards will not be accurate.