Global Journal of Mathematical Sciences: Theory and Practical. ISSN 0974-3200 Volume 5, Number 2 (2013), pp. 117-122 © International Research Publication House http://www.irphouse.com

On Arithmetic Functions

Bhabesh Das

Department of Mathematics, Gauhati University, Guwahati-781014, Assam, India E-mail: [email protected]

Abstract

In this paper, we intend to establish some inequalities on Jordan totient function. Using arithmetic functions like the Dedekind’s function , the function , Euler’s totient function  and their compositions we gave some results on perfect numbers and generalized perfect numbers.

Mathematics Subject Classification: 11A25, 11A41, 11Y70

Keywords: , 2-hyperperfect number, super-perfect number, super-hyperperfect number.

1. Introduction Let  ()n denote the sum of divisors of n i.e. ()n  ∑| , where (1) 1. It is well- known that n is said to be a perfect number if(n ) 2 n . Euclid and Euler showed that even perfect numbers are of the form n 2p1 (2 p  1) , where p and 2p  1 are primes. The prime of the form 2p  1 is called and at this moment there are exactly 48 such primes, which mean that there are 48 even perfect numbers [11]. On the other hand, no odd perfect number is known, but no one has yet been able to prove conclusively that none exist. In [5] the author provided detailed information on perfect numbers. Positive n with the property(n ) 2 n  1 is called , while that of(n ) 2 n  1, quasi-perfect number. For many result and conjectures on this topic, see [4]. D.Suryanarayana [10] introduced the notation of super-perfect numbers. A positive integer n is said to be super- perfect number if (  (n )) 2 n and even super-perfect numbers are of the form 2 p1 such that 2p  1 are Mersenne primes. It is asked in this paper and still unsolved whether there are odd super-perfect numbers. Miloni and Bear [8] introduced the concept of

118 Bhabesh Das k  hyperperfect number and they conjecture that there are k  hyperperfect numbers for everyk  . A. Beg and K. Fogarasi gave a table of k  hyperperfect numbers in [1] and defined positive integer n as k  hyperperfect number if ()n  + .They proved that if n 3k1 (3 k  2) , where 3k  2 is prime, then n is a 2  hyperperfect number. Perfect numbers are 1 hyper- perfect numbers. If (  (n ))  + , then n is called super-hyperperfect number. In [1] authors conjectured that all super-hyperperfect numbers are of the formn  3p1 , where p and are primes. The Jordan’s totient function and Dedekind’s function are respectively defined as 1 1 J( n ) nk (1  ) and (n ) n (1  ) , where p runs through the distinct k  pk  p prime divisors ofn . Let by convention Jk (1) 1,  (1) 1. Euler’s totient function is 1 defined as (n ) n (1  ) i.e. J()() n  n . All these functions are multiplicative,  p 1 i.e. satisfy the functional equation f()()() mn f m f n for g.c.d(m , n ) 1. In [9] the author defines   -perfect,   -perfect,   -perfect numbers, where "" denotes composition.

Basic symbols and notations:  ()n  Sum of divisors of n ,  ()n  Dedekind’s ,

Jk () n  Jordan totient function, ()n  Euler’s totient function, a b  a divides b, a Œbb  a does not divide b, pr{} n  set of distinct prime divisors of n .

Main Results k Proposition2.1. Jk()() ab a J k b , for any a, b  2 , k  , with equality only if, pr{}{} a pr b , where pr{} a denotes the set of distinct prime factors of a Proof: Let 1 1 1    J() ab ak b k (1  )(1  )(1  ) ab  p..  q  r , then k k  k  k p a,, pŒŒ bq a, q b r a r b p q r 1 1 1 Since (1 )  1, so J( ab ) ak b k (1  ) (1  )  a k J ( b ) Thus,  pk kqk  r k k k   Jk()() ab a J k b , for any a, b  2 . If pr{}{} a pr b , then ab  p.  q and p a, p b qŒ a, q b k hence Jk()() ab a J k b .

On Arithmetic Functions 119

k Proposition2.2. If pr{}{} a pr b , then Jk( ab ) ( a  1) J k ( b ) for any a, b  2 , k  .

Proof: Let a  p  q  and b  p  r  , where p are the common prime factors, and q pr{} a are such that q pr{} b , so  1 and ,  ,   0 and 1 1 1 1 pr{}{} a pr b . J() ab ak b k (1  )(1  )(1  )  a k (1  )() J b kpk  q k  r k  q k k 1 1 1 1 1 But, 1  1   1   1   (1  ) so, k  k k  k k a() p  q  q  q q 1 1 Jaba()k (1  )() Jba  k (1  )()( Jba  k  1)() Jb This implies, k qk k a k k k k Jk( ab ) ( a  1) J k ( b ) , for any a, b  2 .

Proposition2.3. Let n  3, and suppose that n is  -, then k k Jk ( ( n )) (2  1) n .

Proof: If n is a  -deficient number, then  (n ) 2 n . Again for any n  3,  ()n is an even number. Using Proposition 2.1, we get 1 J(2) a ak J (2)  a k 2(1 k  )  a k (2 k  1) Let u 2 a , an even number, then k k 2k uk { (n )}k (2n )k J( u ) (2k  1) .Therefore, J( ( n )) (2k  1) (2k  1)  (2 k  1)n k k 2k k 2k 2k .

k Proposition 2.4. For any n  3, Jk ( ( n ))  n

Proof: First remark that for any n  3, ()n is even number and ()n n . Let (n ) 2 u , an even integer. Proceeding as Proposition 2.3, we obtain (2k 1) (2 k  1) J(2 u ) uk (2 k  1) , i.e. J(()){()}. n  nk  n k  n k k k 2k 2 k

k k Proposition2.5. If n is an even perfect number, then Jk ( ( n )) (2  1) n .

Proof: If n is an even perfect number, then n 2p1 (2 p  1), where 2p  1 is a k k k Mersenne prime and (n ) 2 n , so Jk(()) n j k (2) n  n j k (2)  (2  1) n .

Proposition2.6. If n be the product of distinct even perfect numbers n1, n 2 ,..... nr , then 2 a. (a) ()n ()r1  ( n )(  n ).....(  n ) 3 1 2 r

120 Bhabesh Das b. (n  ( n )) 2r n2 . r1 c. (  (n )) 2  (  ( n1 ))  (  ( n 2 )).....  (  ( nr ))

pi1 p i Proof: Let n n1 n 2... nr , where ni 2 (2  1) are distinct perfect numbers with

pi Mersenne primes 2 1( pi primes, i1,2,...... , r )

2 pi 2 (2p11) (2 p 2  1).....(2  pr  2) a.  (ni ) 3.2 for each i , so  (n ) 3.2 2  ()r 1 (n )(  n ).....(  n ) . 3 1 2 r

pi1 p i 1 r1 b. (ni ) 2 (2  1) for each i , and clearly (n ) 2  ( n1 ).  ( n 2 )......  ( nr ) , so (n ) 2r1 n .Therefore (n  ( n ))  ( n )  ( n )  2r n2 .

2(pi  1) (2p11) (2 p 2  1)  .....(2 pr  2) (  (ni )) 2 , for eachi , so(  (n ))  (3.2 ) = 2(2p11) (2 p 2  1)  .....(2 pr  2) r1 = 2 (  (n1 ))(   ( n 2 )).....(   ( nr ))

Proposition 2.7. If n be the product of distinct 2-hyperperfect numbers n1, n 2 ,..... nr , then 3 a. ()n ()r1  ( n )(  n ).....(  n ) 2 1 2 r 3 b. ()n ()r1  ( n )(  n ).....(  n ) 4 1 2 r

ki1 k i Proof: Let n n1 n 2... nr , where ni 3 (3  2) distinct 2-hyperperfect numbers with primes are 3ki  2 ( i1,2,3,..... r ).

ki1 k i 1 a. (ni ) 2.3 (3  1) , for eachi .

2 Then (n ) .3(1)(1).....(1)k1  k 2    kr  (3 k 1  1  1)(3 k 2  1  1)...... (3 k r  1  1) 3 3  ( )r1 (n )(  n )...... (  n ) (b)  (n ) 22 3ki2 (3 k i  1) , for each i . 2 1 2 r i 4 3  (n ) .3(k1 1)  ( k 2  1)  ....  ( kr  1) (3 k 1  1)(3 k 2  1)...... (3 k r  1)  ()r1 (n )(  n ).....(  n ) . 3 4 1 2 r

(  (n )) Proposition2.8. If n is a super-hyperperfect number, then(  (n ))  . 3

Proof: From definition of super-hyperperfect number we know that n  3p1 , where p 3p  1 4n and are primes.  (n ) 4.3 p2 , so (  (n )) 4.3p3  and (n ) 2.3p2 , so 2 9

On Arithmetic Functions 121

4n (  (n )) (  (n )) 4.3p2  .Therefore (  (n ))  . 3 3

Proposition2.9. If n be the product of distinct super-hyperperfect numbers n1, n 2 ,..... nr , then 3r2 a. (()) n .(  n )(  n )...... (  n ) 2r 1 1 2 r 3r 1 b. (()) n .()( n  n ).....(  n ) 22r 3 1 2 r

pi 1 Proof: Let n n1 n 2... nr , where ni  3 are distinct super-hyperperfect numbers with 3pi  1 p and are primes (i 1,2,3,..... r .). Again (n ) 2.3pi 2 and (n ) 4.3pi 2 . i 2 i i

(p 1)  ( p  1)  ......  ( p  1) 2 (p 1)  ( p  1)  ......  ( p  1) (p 1)  ( p  1)  ......  ( p  2) a. (n )  (31 2 r )  .3 1 2 r  2.3 1 2 r , so, 3 r2 2 (p 1)  ( p  1)  ......  ( p  1) 3 (  (n )) .3 1 2 r  . (n )  ( n )......  ( n ) . 9 2r1 1 2 r

(p 1)  ( p  1)  ...... ( p  1) 4 (p 1)  ( p  1)  ...... ( p  1) (p 1)  ( p  1)  ......  ( p  2) (n )  (31 2 r )  .3 1 2 r  4.3 1 2 r , so b. 3 8 3r1 (  (n ) .3(p1 1)  ( p 2  1)  ...... ( pr  1)  . (n )  ( n ).....  ( n ) . 3 22r 3 1 2 r

Proposition2.10. (a) There are infinitely many n such that (J2 ( n )) J 2 (  ( n )) . (b) J(()) n There are infinitely many n such that  (J ( n ))  2 . (c) There are infinitely 2 2 2 many n such that (J2 ( n )) J 2 (  ( n ))  2 n . (d) There are infinitely many n such that 2 J2( ( n )) n   ( J 2 ( n )) .

a a2 a 2 Proof: (a) is valid for n  2 for any a  1, since J 2 (2 ) 3.2 , 3 3n2  (J (2a )) 3.22 a 1  n 2 and  (2a ) 3.2 a1 , J ( (2a )) 3.22 a 1  . This implies 2 2 2 2 a b a b2 a 1 2 b  1 (J2 ( n )) J 2 (  ( n )) . For (b) put n  2 3 , (a 1, b  1) . Then J2 (2 3 ) 2 3 , 4 ((23))Ja b  (22 a 1 3 2 b  1 )2  2 a  2 3 2 b  1  n 2 . Again  (2a 3 b ) 2 a1 3 b , 2 3 8 J(()) n J( (2a 3 b )) 22 a 3 3 2 b  1  n 2 , so  (J ( n ))  2 . To prove (c) put n  2a 7 b , 2 3 2 2 576 (a , b  1) J (2a 7 b ) 32 .2 2 a 2 7 2 b  2 ,  (J (2a 7 b ))  n2 , and (2a 7 b ) 3.2 a2 7 b  1 , 2 2 343

122 Bhabesh Das

4608 J( (2a 7 b ))  n2 , So (J ( n )) J (  ( n ))  2 n2 . 2 2401 2 2 288 To prove (d), put n5a ,( a  1) J(5a ) 24.52 a 2 , ( J (5 a ))  n 2 , 2 2 125 576 (5a ) 6.5 a1 ,J (  (5 a ))  n 2 , 2 625 2 So, J2( ( n )) n   ( J 2 ( n )) .

Proposition2.11. (a) There are infinitely many n such that (J2 ( n )) J 2 (  ( n )) (())J n (b) There are infinitely many n such that J( ( n ))  2 . 2 3

Proof: (a) is valid for n  3a for any a 1 . (b) is valid for n  2a 3 b for any a1, b  1 . Then an easy computation shows that 2 2 (())J n J( ( n ))  n2 and (())J n n2 , so J( ( n ))  2 . 2 27 2 9 2 3

References:

[1] A. Bege, K. Fogarasi, Generalized perfect numbers, Acta Univ. Sapientiae, Mathematica, 1, 1, (2009) 73-82. [2] L.E.Dickson, History of the Theory of Numbers, Vol.1, Stechert New York 1934. [3] R.Gur, N. Bircan, On Perfect Numbers and their Relations, ijcm, vol.5, 2010.no.27, 1337-1346. [4] R.K. Guy: Unsolved problems in number theory, Third ed., Springer Verlag, 2004. [5] O. Knill, The oldest open problem in mathematics, NEU Math Circle, December 2, 2007. [6] J. Luquette, Perfect numbers, math.bu.edu/people/kost/teaching/Ma341/Joe.pdf. [7] J. S. McCranie, A study of hyperperfect numbers, Journal of Integer Sequences, Vol.3.(2000). Article 00.1.3. [8] D. Minoli, R. Bear, Hyperperfect numbers, Pi, Mu Epsilon J., 6 (1975), 153- 157 [9] J. Sandor: On the composition of some arithmetic functions, Studia Babes Bolyai University, Math., 34 (1989), 7-14 [10] D.Suryanarayana, Super perfect numbers, Elem.Matg.24 (1969), 16-17. [11] Great Internet Mersenne Prime Search (GIMPS)