Let S(n,k)=∑i=0n(in)k.
We are to find all positive integers k such that n+1S(n,k) is an integer for all n≥1.
First, consider small values of k.
**Case k=1:**
S(n,1)=∑i=0n(in)=2n.
So n+12n must be integer for all n. But for n=1, 221=1 (integer), but for n=2, 322=34 (not integer). So k=1 does not work.
**Case k=2:**
S(n,2)=∑i=0n(in)2.
Recall that ∑i=0n(in)2=(n2n).
So n+1(n2n) must be integer for all n.
But (n2n)=n!n!(2n)!.
It is known that n+1(n2n)=Cn, the n-th Catalan number, which is always integer.
So k=2 works.
**Case k=3:**
S(n,3)=∑i=0n(in)3.
For n=1, (01)3+(11)3=1+1=2, 22=1 (integer).
For n=2, (02)3+(12)3+(22)3=1+8+1=10, 310=310 (not integer).
So k=3 does not work.
**Case k=4:**
For n=1, (01)4+(11)4=1+1=2, 22=1.
For n=2, (02)4+(12)4+(22)4=1+16+1=18, 318=6.
For n=3, (03)4+(13)4+(23)4+(33)4=1+81+81+1=164, 4164=41.
So for small n, it is integer. Try n=4:
(04)4=1, (14)4=256, (24)4=1296, (34)4=256, (44)4=1.
Sum: 1+256+1296+256+1=1810, 51810=362.
Still integer. Try n=5:
(05)4=1, (15)4=625, (25)4=10000, (35)4=10000, (45)4=625, (55)4=1.
Sum: 1+625+10000+10000+625+1=21252, 621252=3542.
Still integer. But this is not a proof for all n.
But for k=2, we have a closed formula: ∑i=0n(in)2=(n2n), and n+1(n2n) is always integer.
For k>2, there is no such closed formula, and for k=1, it fails for n=2.
Now, let's try to prove that k=2 is the only solution.
Suppose k>2. For large n, (⌊n/2⌋n)k dominates the sum, and n+1 does not generally divide S(n,k).
Alternatively, consider n=p−1 for a prime p.
Then n+1=p.
(in) for 0<i<n is divisible by p except for i=0 and i=n.
So (ip−1)≡0(modp) for 1≤i≤p−2.
Thus, (ip−1)k≡0(modp) for 1≤i≤p−2.
So S(p−1,k)≡(0p−1)k+(p−1p−1)k≡1+1=2(modp).
Therefore, pS(p−1,k) is integer if and only if p∣2.
But for p>2, p does not divide 2, so pS(p−1,k) is not integer for p>2.
Therefore, the only possible value is k=2.
Answer:
The only positive integer k such that
n+11i=0∑n(in)k
is an integer for every positive integer n is k=2.