Congratulations to Curt from Reigate College for cracking this tough nut problem. Here is Curt's solution:

If n= n 1 n , I first assert that for an n>1, then n 1 n is larger than 1

If n>1 then, raising both sides to the power of 1/n yields n 1 n > 1 1 n . So n 1 n is larger than 1 no matter what the value of 1/n. Therefore, one could make n 1 n =1+R. As n 1 n is always larger than 1, R is always a positive real number. Now, raising both sides to the power n one obtains:
n=1+nR+ 1 2 n(n-1) R2 +.

(Incidentally, n>nR, therefore R<1 and so n 1 n <2)

As n is an integer larger than one, at least the third term of expansion must exist. It is clear that:
n > 1 2 n(n-1) R2 2n n(n-1) > R2 2 n-1 >R .

So R0 as n and
n 1 n =1+R<1+ 2 n-1 .

So n 1 n 1 as n.

As a corollary, I will show that 2/(n-1) grows smaller for successive values of n>1. Assume that for some positive B, 2/(n-1)<2/(n+B-1). Now if this is to be true 1/(n-1)<1/(n+b-1), therefore n-1>n+B-1, therefore B<0, contrary to our conditions. Therefore the assertion that for some positive B the value of the function increases is fallacious. This will prove useful later.

For n=1,2, etc. we have n=1, 21/2 , 31/3 , 41/4 = 21/2 ... and we see that the values increase to a maximum of 31/3 and then start to decrease.

Now to prove that 3 is the value of n pertaining to the maximum of this discrete function, I will find an n such that 1+2/(n-1) is less that the cube root of 3. From that point onwards, it is known that for successive values 2/(n-1) decreases, therefore beyond this point the value of n 1 n decreases; more importantly, is less than 3 1 3 . For n=19 this holds true, as it will for subsequent values. The ïn-between" values have been checked and are less than 3 1 3 . If n19 then
n 1 n <1+ 2 18 = 4 3 < 3 1 3