If you search for the id "A216264" on oeis.org, you would find a table of a(n), n=1..60. One interesting thing is that the site said that it was Mikhail Rubinchik who calculated a(26) to a(60), which happened to be out of the brute-force range. What is really disappointing is that in this problem, n may be 61, I think it's that guys's trick to play with us. Another interesting fact is that this guy also invented and introduced the Palindromic Tree. So, I deduce that the solution to this problem is somehow related to this data structure.