The given problem can be generalized as follows:
Given three positive integers k,m,n. Let k1,k2,…,km be any positive integer. Prove that there are infinitely many pair of positive integers (a,b) such that
k=1∏m(a+ki)∣b(b+k) but k=1∏m(a+ki)∤b and k=1∏m(a+ki)∤b+k.
Proof. Let k<p1<p2<…<pm be any m distinct primes and denote
M=k=1∏m(a+ki).
By Chinese remainder theorem, there are infinitely many positive integers a>n such that
a≡−ki(modpi),∀i=1,…,m.
Hence M≡0(modp1p2…pm).
Then we write
M=p1α1p2α2…pmαmq1β1q2β2…qsβs
with αi≥1,i=1,…,m;βj≥1,j=1,…,s and q1,q2,…,qs are s prime divisors of M which are different from p1,p2,…,pm.
By Chinese remainder theorem, there exist infinitely many positive integers b>m such that
{b≡0(modpmαmM)b≡−k(modpmαm)
This implies that
M∤b,M∤b+k and M∣b(b+k).
□