Solution:
1. We shall use the following well-known facts.
LEMMA 1. For any integer x and any positive integer t the number t! divides (x+1)(x+2)…(x+t).
Proof. The statement is obvious for x∈{0,−1,−2,…,−t}. We have to prove it for x>0. Let p be a prime divisor of t. Then the power of p in the prime factorization of t! is [pt]+[p2t]+⋯. This power does not exceed the power of p in the prime factorization of (x+1)(x+2)…(x+t)=x!(x+t)! since the last power equals [px+t]+[p2x+t]+⋯−[px]−[p2x]−⋯=[px+t]−[px]+[p2x+t]−[p2x]−⋯ and [a+b]≥[a]+∗∗.
LEMMA 2. For g(x)∈R[x] with degg=n one has that g(Z)⊂Z if and only if g(x)=∑i=0nbi(ix), where b0,b1,…,bn∈Z.
Proof. Obviously, if g has the above form, then g(Z)⊂Z.
Conversely, let g(Z)⊂Z. Set g(i)=αi∈Z, i=0,1,…,n and apply the Lagrange interpolation formula with knots 0,1,…,n. Then
g(x)=i=1∑ni(i−1)…(i−(i−1))(i−(i+1))…(i−n)x(x−1)…(x−i+1)(x−i−1)…(x−n)αi=i=1∑n(−1)n+ix−iαi(n+1)(n+1x)
It remains to use that any polynomial
x−i(n+1)(n+1x)=n!x(x−1)…(x−i+1)(x−i+1)…(x−n)
can be written in the given form (compare the respective coefficients). The lemma is proved.
Let now m be a divisor of n! and consider the polynomial f(x)=(x+1)(x+2)…(x+n). It follows by Lemma 1 that m divides f(k) for any k∈Z. Moreover, f is a monic polynomial (i.e., the leading coefficient of f equals 1) and hence gcd(a0,a1,…,an,m)=1. So all divisors of n! are solutions of the problem.
Assume that m does not divide n! and m is a solution of the problem. Set r=(m,n!)m. It is clear that r>1 is an integer and (r,n!)=1. Let f(x)=a0+a1x+⋯+anxn, an=0, be a polynomial with the desired properties. Then for g(x)=rf(x) one has that g(Z)⊂Z and, by Lemma 2, g(x)=∑i=0nbi(ix), where b0,b1,…,bn∈Z.
Hence f(x)=∑i=0nmbi(ix) and using that (r,i!)=1 for i=0,1,…,n, we get that r divides all the coefficients of f, a contradiction.