Zsigmondy's theorem

From HandWiki
Short description: On prime divisors of differences two nth powers

In number theory, Zsigmondy's theorem, named after Karl Zsigmondy, states that if [math]\displaystyle{ a\gt b\gt 0 }[/math] are coprime integers, then for any integer [math]\displaystyle{ n \ge 1 }[/math], there is a prime number p (called a primitive prime divisor) that divides [math]\displaystyle{ a^n-b^n }[/math] and does not divide [math]\displaystyle{ a^k-b^k }[/math] for any positive integer [math]\displaystyle{ k\lt n }[/math], with the following exceptions:

  • [math]\displaystyle{ n=1 }[/math], [math]\displaystyle{ a-b=1 }[/math]; then [math]\displaystyle{ a^n-b^n=1 }[/math] which has no prime divisors
  • [math]\displaystyle{ n=2 }[/math], [math]\displaystyle{ a+b }[/math] a power of two; then any odd prime factors of [math]\displaystyle{ a^2-b^2=(a+b)(a^1-b^1) }[/math] must be contained in [math]\displaystyle{ a^1-b^1 }[/math], which is also even
  • [math]\displaystyle{ n=6 }[/math], [math]\displaystyle{ a=2 }[/math], [math]\displaystyle{ b=1 }[/math]; then [math]\displaystyle{ a^6-b^6=63=3^2\times 7=(a^2-b^2)^2 (a^3-b^3) }[/math]

This generalizes Bang's theorem,[1] which states that if [math]\displaystyle{ n\gt 1 }[/math] and [math]\displaystyle{ n }[/math] is not equal to 6, then [math]\displaystyle{ 2^n-1 }[/math] has a prime divisor not dividing any [math]\displaystyle{ 2^k-1 }[/math] with [math]\displaystyle{ k\lt n }[/math].

Similarly, [math]\displaystyle{ a^n+b^n }[/math] has at least one primitive prime divisor with the exception [math]\displaystyle{ 2^3+1^3=9 }[/math].

Zsigmondy's theorem is often useful, especially in group theory, where it is used to prove that various groups have distinct orders except when they are known to be the same.[2][3]

History

The theorem was discovered by Zsigmondy working in Vienna from 1894 until 1925.

Generalizations

Let [math]\displaystyle{ (a_n)_{n\ge1} }[/math] be a sequence of nonzero integers. The Zsigmondy set associated to the sequence is the set

[math]\displaystyle{ \mathcal{Z}(a_n) = \{n \ge 1 : a_n \text{ has no primitive prime divisors}\}. }[/math]

i.e., the set of indices [math]\displaystyle{ n }[/math] such that every prime dividing [math]\displaystyle{ a_n }[/math] also divides some [math]\displaystyle{ a_m }[/math] for some [math]\displaystyle{ m \lt n }[/math]. Thus Zsigmondy's theorem implies that [math]\displaystyle{ \mathcal{Z}(a^n-b^n)\subset\{1,2,6\} }[/math], and Carmichael's theorem says that the Zsigmondy set of the Fibonacci sequence is [math]\displaystyle{ \{1,2,6,12\} }[/math], and that of the Pell sequence is [math]\displaystyle{ \{1\} }[/math]. In 2001 Bilu, Hanrot, and Voutier[4] proved that in general, if [math]\displaystyle{ (a_n)_{n\ge1} }[/math] is a Lucas sequence or a Lehmer sequence, then [math]\displaystyle{ \mathcal{Z}(a_n) \subseteq \{ 1 \le n \le 30 \} }[/math] (see OEISA285314, there are only 13 such [math]\displaystyle{ n }[/math]s, namely 1, 2, 3, 4, 5, 6, 7, 8, 10, 12, 13, 18, 30). Lucas and Lehmer sequences are examples of divisibility sequences.

It is also known that if [math]\displaystyle{ (W_n)_{n\ge1} }[/math] is an elliptic divisibility sequence, then its Zsigmondy set [math]\displaystyle{ \mathcal{Z}(W_n) }[/math] is finite.[5] However, the result is ineffective in the sense that the proof does not give an explicit upper bound for the largest element in [math]\displaystyle{ \mathcal{Z}(W_n) }[/math], although it is possible to give an effective upper bound for the number of elements in [math]\displaystyle{ \mathcal{Z}(W_n) }[/math].[6]

See also

References

  1. A. S. Bang (1886). "Taltheoretiske Undersøgelser". Tidsskrift for Mathematik. 5 (Mathematica Scandinavica) 4: 70–80.  And Bang, A. S. (1886). "Taltheoretiske Undersøgelser (continued, see p. 80)". Tidsskrift for Mathematik 4: 130–137. 
  2. Montgomery, H. "Divisibility of Mersenne Numbers." 17 Sep 2001.
  3. Artin, Emil (August 1955). "The Orders of the Linear Groups". Comm. Pure Appl. Math. 8 (3): 355-365. doi:10.1002/cpa.3160080302. 
  4. Y. Bilu, G. Hanrot, P.M. Voutier, Existence of primitive divisors of Lucas and Lehmer numbers, J. Reine Angew. Math. 539 (2001), 75-122
  5. J.H. Silverman, Wieferich's criterion and the abc-conjecture, J. Number Theory 30 (1988), 226-237
  6. P. Ingram, J.H. Silverman, Uniform estimates for primitive divisors in elliptic divisibility sequences, Number theory, Analysis and Geometry, Springer-Verlag, 2010, 233-263.

External links