

A081858


Numbers n such that 2n+1 divides 2^n1.


5



0, 3, 8, 11, 15, 20, 23, 35, 36, 39, 44, 48, 51, 56, 63, 68, 75, 83, 95, 96, 99, 111, 116, 119, 120, 128, 131, 135, 140, 155, 156, 168, 170, 176, 179, 183, 191, 200, 204, 215, 216, 219, 224, 228, 231, 239, 243, 251, 260, 280, 284, 288, 296, 299, 300, 303, 308
(list;
graph;
refs;
listen;
history;
text;
internal format)



OFFSET

1,2


COMMENTS

From Chris Boyd, Mar 16 2014: (Start)
n is a term if and only if n=0, 2n+1 is a prime of the form 8k+1, or 2n+1 is an Euler pseudoprime satisfying 2^n == 1 mod 2n+1.
Case 1: 0 is a term. Case 2, 2n+1 prime: by Euler's criterion and the quadratic character of 2, 2^n == 1 mod 2n+1 only if 2n+1 is of the form 8k+1. Case 3, 2n+1 composite: 2^n == 1 mod 2n+1 only if 2n+1 is one of the subset of Euler pseudoprimes satisfying 2^n == 1 mod 2n+1.
The first term for which 2n+1 is a qualifying Euler pseudoprime is n=170.
The first Euler pseudoprime that does not correspond to a term is 3277, because 2^((32771)/2) == 1 mod 3277. (End)


LINKS

Table of n, a(n) for n=1..57.


FORMULA

n such that A002326(n)n: since 2^n == 1 mod 2n+1, n must be a multiple of the order of 2 mod 2n+1.


PROG

(PARI) isok(n) = !((2^n1) % (2*n+1)); \\ Michel Marcus, Dec 04 2013
(PARI) for(n=0, 400, if(n%znorder(Mod(2, 2*n+1))==0, print1(n", "))) \\ Chris Boyd, Mar 16 2014, after Michael Somos in A002326


CROSSREFS

Cf. A014664.
Sequence in context: A257336 A234431 A310282 * A145837 A111132 A188473
Adjacent sequences: A081855 A081856 A081857 * A081859 A081860 A081861


KEYWORD

nonn


AUTHOR

Benoit Cloitre, Apr 11 2003


EXTENSIONS

Formula corrected by Chris Boyd, Mar 16 2014


STATUS

approved



