You are on page 1of 8

A Sucient Condition for Disproving Descartess Conjecture on Odd Perfect Numbers

Jose Arnaldo B. Dris

Department of Mathematics, Far Eastern University Nicanor Reyes Street, Sampaloc, Manila Philippines

Abstract Let (x) be the sum of the divisors of x. If N is odd and (N ) = 2N , then the odd perfect number N is said to be given in Eulerian form if N = q k n2 where q is prime with q k 1 (mod 4) and gcd(q, n) = 1. In this note, we show that q < n implies that Descartess conjecture (previously Sorlis conjecture), k = q (N ) = 1, is not true. This then implies an unconditional proof for the biconditional k = q (N ) = 1 n < q. Lastly, following a recent result of Cohen and Sorli, we show that if q < n, then either q > 5 or k > 5 is true. Keywords: Sorlis conjecture, odd perfect number, abundancy index 2010 MSC: 11A05, 11J25, 11J99

This research work was completed under a Research Assistance Grant from Far Eastern University. Email address: jdris@feu.edu.ph, jabdris@yahoo.com.ph (Jose Arnaldo B. Dris)

A Sucient Condition for Disproving Descartess Conjecture on Odd Perfect Numbers

Jose Arnaldo B. Dris
Department of Mathematics, Far Eastern University Nicanor Reyes Street, Sampaloc, Manila Philippines

1. Introduction If J is a positive integer, then we write (J ) for the sum of the divisors of J . A number L is perfect if (L) = 2L. We denote the abundancy index I of the positive integer x as I (x) = (x)/x. An even perfect number M is said to be given in Euclidean form if M = (2p 1) 2p1 where p and 2p 1 are primes. We call Mp = 2p 1 the Mersenne prime factor of M . Currently, there are only 48 known Mersenne primes [9], which correspond to 48 even perfect numbers. An odd perfect number N is said to be given in Eulerian form if N = q k n2 where q is prime with q k 1 (mod 4) and gcd(q, n) = 1. We call q k the Euler part of N while n2 is called the non-Euler part of N . (We will call q the Euler prime factor of N .) It is currently unknown whether there are innitely many even perfect numbers, or whether any odd perfect numbers exist. It is widely believed that there is an innite number of even perfect numbers. On the other hand, no examples for an odd perfect number have been found (despite extensive computer searches), nor has a proof for their nonexistence been established. Ochem and Rao [14] 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 ra ||N with r prime). This improves on previous results by Brent, Cohen and te Riele [2] in 1991 and Cohen [3] in 1987, respectively.
This research work was completed under a Research Assistance Grant from Far Eastern University. Email address: jdris@feu.edu.ph, jabdris@yahoo.com.ph (Jose Arnaldo B. Dris)

November 10, 2013

In a recent preprint, Nielsen [12] claims to have obtained the lower bound (N ) 10 for the number of distinct prime factors of N , improving on his last result (N ) 9 (see [13]). For the largest prime factor p(N ) of an odd perfect number N , Goto and Ohno proved in 2008 [8] that p(N ) > 108 , improving on Jenkins result that p(N ) > 107 in 2003 [11]. Sorli conjectured in [16] that k = q (N ) = 1 always holds. Just recently, Beasley [1] points out that Descartes was the rst to conjecture k = q (N ) = 1 in a letter to Marin Mersenne in 1638, with Frenicles subsequent observation occurring in 1657. Beasleys observations are corroborated by similar lines from Jaromas paper [10]. Dris referred to Descartess conjecture in [5] and [6] (as well as in MathOverow http://mathoverflow.net) as Sorlis conjecture. Dris conjectured in [6] and [7] that the components q k and n are related by the inequality q k < n. This conjecture was made on the basis of the result 3 k I (q ) < 2 < I (n). 2. Main Results The proof of the following lemma is trivial. Lemma 2.1. Let N = q k n2 be an odd perfect number given in Eulerian form. If n < q k then the biconditional k = 1 n < q is true. Proof. Suppose that N = q k n2 is an odd perfect number given in Eulerian form. If n < q k , then k = 1 = n < q follows trivially by assuming k = 1 and substituting through to n < q k = q. Unconditionally, the other implication {n < q = k = 1} follows from [6]. This nishes the proof. We are now ready to prove our main result. Theorem 2.2. Let N = q k n2 be an odd perfect number given in Eulerian form. If q < n, then k 5. Proof. Suppose that N = q k n2 is an odd perfect number given in Eulerian form. Assume further that q < n. Now, either the implication k = 1 = n < q is true, or not. The said implication is false if and only if both k=1 and q<n 3

are true. Since the implication n < q k = {k = 1 n < q } is true (by Lemma 2.1), it follows that if {k = 1} {n < q } is false, then q k < n is true. Thus, if the implication k = 1 = n < q is not true, then it must be the case that q = q k < n. On the other hand, if the implication k = 1 = n < q is true, then the biconditional k = 1 n < q would then be true. This gives the two cases: q < qk < n or q < n < qk . Thus, if the implication k = 1 = n < q is true, then either q < qk < n or q < n < qk . Note that this last disjunction is actually an exclusive-or (). Let us summarize what we have obtained so far. (Again, note that we are assuming that q < n.) If {k = 1 = n < q }, then q = q k < n holds. If the implication k = 1 = n < q is true, then { q < q k < n} { q < n < q k } is true.

By the contrapositive of the rst implication, we obtain {q = q k < n} = {k = 1 = n < q }. Consequently, since we know that n < q = k = 1 [6], we have {q = q k < n} {n < q k 5}. But by using the second implication above, we get {q = q k < n} = {k = 1 = n < q } = {{q < q k < n} {q < n < q k }}. Consequently, we have: {n < q k 5} = {{q < q k < n} {q < n < q k }}. Hence, we either have {n < q k 5} = {q < q k < n}, or {n < q k 5} = {q < n < q k }, but not both. Now, under the assumption q < n, n < q is false. We therefore conclude that either k 5 = {q < q k < n}, or k 5 = {q < n < q k }, is true, but not both. (Observe that if one of these two implications is true, the other one must be false.) Either way, k 5 must be true. This completes the proof. Remark 2.3. Note that the contrapositive of Theorem 2.2 is equivalent to the truth of the implication k = 1 = n < q. Thus, even if q < n is not true, we now know that the biconditional k = 1 n < q is true. In particular, if the Descartes-Frenicle-Sorli conjecture (i.e., k = q (N ) = 1) is true, then n < q follows, which would then imply that the Euler prime q is the largest prime factor of the odd perfect number N = q k n2 = qn2 . We can then derive the lower bound q > 10500 from [14]. This signicantly improves the currently known results on a lower bound for the largest prime factor of N [8], and also a lower bound for the largest (prime-power) component of N [14]. 5

Further to the statements in Remark 2.3, and building on a recent result from [4], we have the following corollary to Theorem 2.2. Corollary 2.4. Let N = q k n2 be an odd perfect number given in Eulerian form. If q < n, then either 5 < q or 5 < k . Proof. Suppose that N = q k n2 is an odd perfect number given in Eulerian form with q < n. From Theorem 2.2, k 5. Since q is the Euler prime of N , q 1 (mod 4) implies that q 5. Thus, the smallest possible value of the Eulerian component q k is 55 . Assume that q = 5 and k = 5. This contradicts the rst number q k = 55 in Cohen and Sorlis list of impossible Eulerian components for an odd perfect number in [4]. Consequently, if q < n, then it follows that 5 < q or 5 < k , as desired. 3. Remaining Open Problems The author (together with Keneth Adrian P. Dagal) has uploaded a preprint to arXiv (albeit currently on hold), that contains a proof for the inequality q < n, if N = q k n2 is an odd perfect number given in Eulerian form. (A copy is available online via http://www.scribd.com/doc/182737454.). The preprint is currently titled The Abundancy Index of Divisors of Odd Perfect Numbers - Part III and has been submitted to Princeton Universitys Annals of Mathematics. The remaining open problems should be obvious to the reader. The interested student of elementary number theory and recreational mathematics is invited to peruse OEIS sequence A228059 [15] for a (possibly) feasible computational approach to proving that the conjunction {k = q (N ) = 1} {q < n} always holds for an odd perfect number N = q k n2 . More information is available in the survey article [5]. If successful, such a computational project would then eectively prove that there are no odd perfect numbers. If the project is unsuccessful, then Descartes may have been right from the very beginning - which could also be the reason why Mersenne had an uncanny way of determining which primes p 257 would make 2p 1 a prime number as well. Quoting verbatim from Jaroma [10]: Primes of the form 2p 1 are called Mersenne primes. They are named in honor of the 17th century priest, Fr. M. Mersenne (1588-1648), who claimed that such numbers are prime provided that p {2, 3, 5, 7, 13, 17, 19, 31, 67, 127, 257} and are composite for all other values of p 257. Although Mersennes conjecture contained ve mistakes, it had taken more than 300 years for mathematicians to discover them all.

4. Acknowledgements The author profusely thanks Pascal Ochem for his comments on an earlier version of this preprint that paved the way for most of the results in this article. The author also wishes to express his gratitude to Peter Acquaah for helpful e-mail exchanges on odd perfect numbers. He would like to thank Severino Gervacio (his M. Sc. thesis adviser) for suggesting that he try to investigate the Descartes-Frenicle-Sorli conjecture rst, and for serving as his / their research mentor at Far Eastern University - Manila (http://www.feu.edu.ph/manila/. Lastly, the author thanks the anonymous referee(s) who have made invaluable comments and suggestions for improving the quality of this paper. References [1] B. D. Beasley, Euler and the ongoing search for odd perfect numbers, ACMS 19th Biennial Conference Proceedings, Bethel University, May 29 to Jun. 1, 2013, http://godandmath.files.wordpress.com/2013/07/acms-2013-proceedings.pdf, pages 21-31. [2] 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. [3] 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. [4] G. L. Cohen, R. M. Sorli, On odd perfect numbers and even 3-perfect numbers, Integers, 12 (2012), Article #A6, http://www.emis.de/journals/INTEGERS/papers/a6self/a6self.pdf, ISSN 1867-0652, doi:http://dx.doi.org/10.1515/integers-2012-0036. [5] J. A. B. Dris, Euclid-Euler heuristics for (odd) perfect numbers, (Oct. 2013) http://arxiv.org/abs/1310.5616. [6] 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. [7] J. A. B. Dris, Solving the Odd Perfect Number Problem: Some Old and New Approaches, M. Sc. thesis, De La Salle University, Manila, Philippines, 2008, http://arxiv.org/abs/1204.1450. [8] T. Goto, Y. Ohno, Odd perfect numbers have a prime factor exceeding 108 , Math. Comp., 77 (2008), 1859-1868, doi:http://dx.doi.org/10.1090/S0025-5718-08-02050-4. 7

[9] Various contributors, Great Internet Mersenne Prime http://www.mersenne.org/, last accessed: 11/10/2013.

Search,

[10] J. H. Jaroma, The imperfect Fibonacci and Lucas numbers, Irish Math. Soc. Bull., 63 (2009), 33-43, http://www.maths.tcd.ie/pub/ims/bull63/S6302.pdf, ISSN 07915578. [11] P. .M. Jenkins, Odd perfect numbers have a prime factor exceeding 107 , Math. Comp., 72 (2003), 1549-1554, doi:http://dx.doi.org/10.1090/S0025-5718-03-01496-0. [12] P. P. Nielsen, Odd perfect numbers, Diophantine equations, and upper bounds, preprint, http://math.byu.edu/~pace/BestBound_web.pdf. [13] P. P. Nielsen, Odd perfect numbers have at least nine distinct prime factors, Math. Comp., 76 (2007), 2109-2126, doi:http://dx.doi.org/10.1090/S0025-5718-07-01990-4. [14] 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. [15] T. D. Noe, Online Encyclopedia of Integer Sequences, sequence A228059, http://oeis.org/A228059, last accessed: 11/03/2013. [16] 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.