I came across this question that asks weather there are infinitely many primes of the form $an+d$ with $\gcd$($a, d$) $=$ $1$. To show there are infinitely many primes of the form(s) $4k-1$ and $6k-1$ can be proven with Euclid's Theorem:

For $4k-1$:

Suppose there are finitely many primes (in general), they are $2$, $p$, $p_2$,... $p_n$. Now let $q$ be the product of all these primes. $2q$ is still divisible by all the primes $p_n$ on our list. $2q-1$, However is $-1$ $\pmod {p_n}$ for all primes $p_n$ on the list of primes first created. Therefor, $2q-1$ must be divisible by a prime $n$ that was not on our list. Since $2$ was on our list of primes, and the product $2q$ is divisible by $4$, $2q-1$ is $3$ $\pmod 4$, and therefore must have a prime factor $3$ $\pmod 4$ which was not on our list of primes.

The $6k-1$ proof is similar by first starting with $2, 3, p, p_2$, $p_n$. instead. Given the product $q, q-1$ has a prime factor $n$ $=$ $5$ $\pmod 6$, but $n$ is not on our list.

Is there a proof like these for the $+1$ sides or any other progressions? Thanks.


In general it's true that whenever $(a,d) = 1$ then the arithmetic sequence $ax + d$ has infinitely many primes. This is true due to the Dirichlet's Theorem. I doubt that there's an easy proof of this.

  • 1
    $\begingroup$ This is kind of orthogonal to the question. $\endgroup$ – Erick Wong Feb 9 '18 at 6:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.