Answer: The Derivative That Divides 2016
Key Idea / Intuition
The key is to think about what j-th derivatives of integer-coefficient polynomials look like at integers. The natural basis for integer-valued polynomials is not {1,x,x2,โฆ} but the binomial coefficients (nxโ). Taking j derivatives of xn produces n(nโ1)โฏ(nโj+1)xnโj, and at integers this must be divisible by 2016 for all valid nโฅj. So we need j! to be divisible by 2016, and we want the smallest such j.
More precisely: the j-th derivative of xn at x=k is n(nโ1)โฏ(nโj+1)knโj, a product of j consecutive integers times knโj. The "worst case" is n=j, giving j!. So we need 2016โฃj!.
Formal Proof / Solution
Step 1: Factor 2016.
2016=25โ
32โ
7
Step 2: Reduce to divisibility of j!.
Every polynomial with integer coefficients can be written uniquely in the basis of falling factorials / binomial coefficients. It suffices to check the monomials p(x)=xn for nโฅj (since lower-degree polynomials have zero j-th derivative). The j-th derivative of xn is
dxjdjโxn=n(nโ1)โฏ(nโj+1)xnโj.
At an integer k, this equals (nโj)!n!โโ
knโj.
The hardest case to make divisible by 2016 is n=j, k=1 (or any k coprime to 2016), giving:
p(j)(1)=j!โ
10=j!.
So we need 2016โฃj!.
Step 3: Conversely, if 2016โฃj!, does it always work?
For general nโฅj and integer k:
(nโj)!n!โโ
knโj=j!โ
(jnโ)โ
knโj.
If 2016โฃj!, then 2016 divides j!โ
(jnโ)โ
knโj for all integers k and all nโฅj. So divisibility of j! by 2016 is both necessary and sufficient.
Step 4: Find the smallest j with 2016โฃj!.
We need 25โ
32โ
7โฃj!.
- The factor 7 requires jโฅ7 (since 7!=5040 contains one factor of 7).
- Check j=7: 7!=5040=24โ
32โ
5โ
7. This gives only 24, but we need 25. So 7โค our requirement โ wait, actually the issue is the power of 2: 7! contributes โ7/2โ+โ7/4โ=3+1=4 factors of 2. We need 5.
- Check j=8: 8!=40320. Powers of 2 in 8!: โ8/2โ+โ8/4โ+โ8/8โ=4+2+1=7โฅ5. โ
Powers of 3: โ8/3โ+โ8/9โ=2+0=2โฅ2. โ
Powers of 7: โ8/7โ=1โฅ1. โ
So 2016โฃ8! but 2016โค7! (since 7! only has 24, missing one factor of 2).
Conclusion:
j=8โ