You are on page 1of 6

MATHEMATICS OF COMPUTATION

Volume 00, Number 0, Pages 000000


S 0025-5718(XX)0000-0

NEW RESULTS FOR SORLIS CONJECTURE


ON ODD PERFECT NUMBERS - PART III
JOSE ARNALDO B. DRIS

Abstract. If N = q k n2 is an odd perfect number given in Eulerian form,


then Sorlis conjecture predicts that k = q (N ) = 1. In this article, we give
bounds for the real number satisfying

(q k )
(n)

<<

qk
,
n

when k > 1.

1. Introduction
If N is a positive integer, then we denote the sum of the divisors of N by (N ).
The positive integer N is said to be perfect if (N ) = 2N . It is currently an open
problem to determine whether there are infinitely many even perfect numbers, or if
there are any odd perfect numbers. Ochem and Rao [6] recently obtained the lower
bound N > 101500 for an odd perfect numbers magnitude, and a lower bound of
1062 for its largest component (i.e., divisor pa ||N with p prime). This improves on
previous results by Brent, Cohen and te Riele [1] in 1991 and Cohen [2] in 1987,
respectively.
An odd perfect number N = q k n2 is said to be given in Eulerian form if q is
prime with q k 1 (mod 4) and gcd(q, n) = 1. (The number q is called the
Euler prime, while the component q k is referred to as the Euler factor. Note that,
since q is prime and q 1 (mod 4), then q 5.)
In his Ph. D. thesis, Sorli [7] conjectured that k = q (N ) = 1.1
The author conjectured in [3] that the components q k and n are related by the
inequality q k < n.
We denote the abundancy index I of the positive integer x as
I(x) =

(x)
.
x

2. Bounds for , where

(q k )
(n)

<<

qk
n

Let N = q k n2 be an odd perfect numbergiven in Eulerian form.


Recall from the article [3] that I(q k ) < 3 2 < I(n). This implies that
(q k )
qk
< .
(n)
n
Received by the editor July 13, 2013.
2010 Mathematics Subject Classification. Primary 11A05; Secondary 11J25, 11J99.
1Key Words and Phrases: odd perfect number, Sorlis conjecture, Euler prime
c
XXXX
American Mathematical Society

JOSE ARNALDO B. DRIS

Without loss of generality, we may assume that there is a positive real number
such that inequality
(q k )
qk
<<
(n)
n
is true. Equivalently, we also have a positive real number such that the inequality
1
(n)
n
< = <
qk

(q k )
holds. Note that 6= 1 and =
6 1, because of the biconditional
q k < n (q k ) < (n).
(See [4]).
Assume that q k < n and k > 1. (Note that this implies that 0 < < 1.) The
inequality
(q k )
qk
<<
(n)
n
then implies that
 k


q
(q k )

< 0,
(n)
n
from which it follows that
" (qk ) #
(q k ) q k
1
n
+ <
+ .
(n)

(n)
n
k
q

Since we have the biconditional q k < n (q k ) < (n), we get the bounds
r
125
(q k )
<
(n)
128
(see [5]), and
qk
<1
n
from which we get the upper bound
(q k ) q k
+
<1+
(n)
n

125
.
128

Now we get a lower bound for the quantity


" (qk ) #
1
n
+
(n)
k
q

in terms of . It suffices to get a lower bound for


(q k )
n
(n)
qk

To this end, recall from the article [5] that, for the case q k < n and k > 1, we have
the bounds
(q k )
(n)
1<
< k < 2.
n
q

NEW RESULTS FOR SORLIS CONJECTURE

ON ODD PERFECT NUMBERS - PART III 3

In particular, we have the lower bound


(q k )
n
(n)
qk

1
<
2

Consequently, we get
1
1
+ <
2

" (qk ) #
n
(n)
qk

(q k ) q k
+ <
+
<1+
(n)
n

125
,
128

from which it follows that


r
2

2 2 1 +

125
128

!
+ 1 < 0.

Solving this last inequality, we get


q
q



1
1 
1
1 
160 10 6 < <
160 10 6.
16 + 5 10
16 + 5 10 +
32
32
32
32
Next, suppose that n < q k and k > 1. This time, our assumed inequality takes
the form
n
1
(n)
< = <
.
qk

(q k )
By the biconditional n < q k (n) < (q k ), we know that
0 < < 1.
It follows that




n
(n)

< 0,
(q k )
qk

which then gives


1

"

(n)
qk
(q k )
n

#
+ <

(n)
n
+ .
(q k ) q k

Again, since we have the biconditional n < q k (n) < (q k ), we obtain the
bounds
(n)
<1
(q k )
and
r
n
4 125
<
qk
128
(see [5]), from which we get the upper bound
r
(n)
n
4 125
+ k <1+
.
k
(q ) q
128
As before, we derive a lower bound for the quantity
" (n) #
1
qk
+
(q k )

JOSE ARNALDO B. DRIS

Similarly, it suffices to get a lower bound for


in terms of .
(n)
qk
(q k )
n

Again, recall from the article [5] that, for the case n < q k and k > 1, we have the
bounds
(n)
(q k )
1< k <
< 2.
q
n
In particular, we have the lower bound
1
<
2

(n)
qk
(q k )
n

Consequently, we now obtain


" (n) #
r
1
1
(n)
n
qk
4 125

+ <
+ <
+ k <1+
,
k
(q k )

(q ) q
128
2

n
from which it follows that
r
22 2 1 +

125
128

!
+ 1 < 0.

Solving this last inequality, we obtain


 1q
 1q

1
1
4
4
4
4
4 + 250
16 + 8 250 + 5 10 < <
4 + 250 +
16 + 8 250 + 5 10.
8
8
8
8
But we know that
n
1
(n)
< = <
.
qk

(q k )
Consequently,
(q k )
1
qk
<= <
(n)
n

and we therefore have


1
1
1

<= < 1
.
 1p
 1p
4
4
4
1

4 + 250 +
16 + 8 250 + 5 10
4 + 250
16 + 8 4 250 + 5 10
8

We now have the following theorem.


Theorem 2.1. Let N = q k n2 be an odd perfect number given in Eulerian form.
Suppose that k > 1.
k

)
q
If q k < n and the real number satisfies (q
(n) < < n , then
q
q



1 
1
1 
1
16 + 5 10
160 10 6 < <
16 + 5 10 +
160 10 6.
32
32
32
32

If n < q k and the real number satisfies


1
8

4+

(q k )
(n)

1
1
p

<= <
4

250 +
16 + 8 250 + 5 10


1
8

<<
1
8

4+

qk
n ,

then
1

250

1
8

.
16 + 8 4 250 + 5 10

NEW RESULTS FOR SORLIS CONJECTURE

ON ODD PERFECT NUMBERS - PART III 5

Remark 2.2. Note the rational approximations


q


1 
1
16 + 5 10
160 10 6 0.29536,
32
32
q



1
1
16 + 5 10 +
160 10 6 1.69285,
32
32
1
p

0.5882475271,


4
1
250 + 18 16 + 8 4 250 + 5 10
8 4+
and
1

3.3999292948.

 1p
4
4
1
250

16
+
8
250
+
5
10
4
+
8
8
Remark 2.3. Let N = q k n2 be an odd perfect number given in Eulerian form, and
suppose that k > 1.
k
)
qk
If q k < n and the real number satisfies (q
(n) < < n , then the bounds for
given by the inequality
q
q


1 
1
1 
1
16 + 5 10
160 10 6 < <
16 + 5 10 +
160 10 6
32
32
32
32
k

)
q
are trivial, as compared to the bounds 21 < (q
(n) < < n < 1, as given in page 5
of the article [5].
k
)
qk
Likewise, if n < q k and the real number satisfies (q
(n) < < n , then the
bounds for given by the inequality
1
1

<< 1

 1p
 1p
4
4
4
1
250 + 8 16 + 8 250 + 5 10
250 8 16 + 8 4 250 + 5 10
8 4+
8 4+

are also trivial, as compared to the bounds 1 <


page 5 of the article [5].

(q k )
(n)

< <

qk
n

< 2, as given in

3. Conclusion
In this paper, we have attempted to obtain bounds for the real number satisfying
(q k )
qk
<<
(n)
n
by considering two (mutually exclusive) cases, under the assumption that Sorlis
conjecture for odd perfect numbers is false.
Notice that, if the upper bound I(n) < 2 can be improved (which can be done
by using an idea from [8]; see [5]), then we can potentially get sharper bounds for
the quantities
(q k )
(n)
and
qk
.
n
The consequential bounds thus obtained can then be iteratively inputted into the
same algorithm (considered in this paper) as before, and thereby obtain either an
optimal bound, or maybe a contradiction, in the particular cases q k < n or n < q k .
(This paper is currently a work in progress.)

JOSE ARNALDO B. DRIS

4. Acknowledgments
The author sincerely thanks the anonymous referees who have made several
suggestions, which helped in improving the style of the paper.
References
1. R. P. Brent, G. L. Cohen, H. J. J. te Riele, Improved techniques for lower bounds for
odd perfect numbers, Math. Comp. 57 (1991), 857-868, doi:http://dx.doi.org/10.1090/
S0025-5718-1991-1094940-3.
2. G. L. Cohen, On the largest component of an odd perfect number, J. Austral. Math. Soc. Ser.
A, 42 (1987), 280-286, doi:http://dx.doi.org/10.1017/S1446788700028251.
3. J. A. B. Dris, The abundancy index of divisors of odd perfect numbers, J. Integer Seq., 15
(Sep. 2012), Article 12.4.4, https://cs.uwaterloo.ca/journals/JIS/VOL15/Dris/dris8.html,
ISSN 1530-7638.
4. J. A. B. Dris, New results for Sorlis conjecture on odd perfect numbers, to appear in Int. J.
Pure Appl. Math., preprint:http://arxiv.org/abs/1302.5991.
5. J. A. B. Dris, New results for Sorlis conjecture on odd perfect numbers - Part II, preprint:http:
//arxiv.org/abs/1303.2329.
6. P. Ochem, M. Rao, Odd perfect numbers are greater than 101500 , Math. Comp., 81 (2012),
1869-1877, doi:http://dx.doi.org/10.1090/S0025-5718-2012-02563-4
7. R. M. Sorli, Algorithms in the Study of Multiperfect and Odd Perfect Numbers, Ph. D. Thesis,
University of Technology, Sydney, 2003, http://epress.lib.uts.edu.au/research/handle/
10453/20034.
8. J. Ward, Does Ten Have a Friend?, http://arxiv.org/pdf/0806.1001v2.pdf.
Far Eastern University, Nicanor Reyes Street, Sampaloc, Manila, Philippines
E-mail address: jadris@feu.edu.ph, jabdris@yahoo.com.ph