OFFSET
0,1
COMMENTS
As D. W. Wilson observes, this is similar to the Riesel/Sierpinski problem and there is e.g. no prime of the form 2^k - 777149, which is divisible by 3,5,7,13,19,37 or 73 if k is in 1+2Z, 2+4Z, 4+12Z, 8+12Z, 12+36Z, 0+36Z resp. 24+36Z. Already for n=935 it is difficult to find a solution. Is this linked to the fact that 2n+1=1871 is member of a prime quadruple (A007530) and quintuple (A022007)? - M. F. Hasler, Apr 07 2008
LINKS
T. D. Noe, Table of n, a(n) for n = 0..934
F. Firoozbakht, M. F. Hasler, Variations on Euclid's formula for Perfect Numbers, JIS 13 (2010) #10.3.1.
EXAMPLE
MATHEMATICA
Table[k = 1; While[2^k < n || ! PrimeQ[2^k - n], k++]; k, {n, 1, 1869, 2}] (* T. D. Noe, Mar 18 2013 *)
PROG
(PARI) A096502(n, k)={ k || k=log(n)\log(2)+1; n=2*n+1; while( !ispseudoprime(2^k++-n), ); k } /* will take a long time for n=935... */ - M. F. Hasler, Apr 07 2008
CROSSREFS
KEYWORD
nonn,hard
AUTHOR
Labos Elemer, Jul 09 2004
STATUS
approved