Explicit density computations for certain sets of primes in Lucas sequences
Joaquim Cera Da Conceição · 2026
Read Explicit density computations for certain sets of primes in Lucas sequences by Joaquim Cera Da Conceição (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-20014.txt.
Introduction
Cera Da Conceição derives formulas in closed form for the Dirichlet density of primes p where a fixed integer d divides the rank of appearance of p in a Lucas sequence. The rank of appearance is the least positive integer n with p dividing the nth term, when such an n exists. Covering all Lucas sequences and all positive d, the results complete Sanna's 2022 work. The paper thus settles prime densities for rank divisibility in full generality.
Catalog record
| CREATOR | Joaquim Cera Da Conceição |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 56510 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.20014
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Explicit density computations for certain sets of primes in Lucas sequences Joaquim Cera Da Conceição Abstract . Let U U be a Lucas sequence, p p be prime, and ρ U ( p ) \rho_{U}(p) be the rank of appearance of p p in U U . We derive closed-form formulas for the Dirichlet density of primes p p for which d | ρ U ( p ) d\mid\rho_{U}(p) , where d ≥ 1 d\geq 1 is a fixed integer. Our results complete the work of Sanna ( 2022 2022 ) by covering all U U and all d ≥ 1 d\geq 1 . 1991 Mathematics Subject Classification 11B39, 11R45, 11R18, 11R32 1. Introduction A Lucas sequence U=U(a_{1},a_{2})=(U_{n})_{n\geq 0} with non-zero parameters a_{1},a_{2}\in\mathbb{Z} is defined by U_{0}=0 , U_{1}=1 , and U_{n+2}=a_{1}U_{n+1}-a_{2}U_{n} for all n\geq 0 . Given a prime number p , we denote by \rho_{U}(p) the smallest positive n at which p\mid U_{n} . This integer is called the rank of appearance of p in U and exists for all p\nmid a_{2} . In this paper, we study the Dirichlet density of the set \mathcal{R}_{U}(d) of primes p\nmid a_{2} such that \rho_{U}(p) is divisible by a fixed integer d\geq 1 . This was first studied by Hasse [ 6 , 7 ] for sequences U(a+1,a) with a\in\mathbb{Z} square-free, |a|\geq 2 , and d a prime number. He settled the existence and an explicit formula of the Dirichlet density. For instance, Hasse found a density of 17/24 in the case |a|=d=2 , and d^{2}/(d^{2}-1) otherwise. Following his work, many authors considered this problem. Let f_{U}(X)=X^{2}-a_{1}X+a_{2} be the characteristic polynomila of U . The problem was completely solved by Wiertelak [ 19 , 20 , 21 ] when f_{U}(X) is reducible. In 1985, Lagarias [ 10 ] computed the Dirichlet density of \mathcal{R}_{U}(2) for three Lucas sequences with irreducible polynomial. In particular, he proved that \mathcal{R}_{F}(2) has density 2/3 , where F=U(1,-1) is the Fibonacci sequence. Later on, the Fibonacci case was completed for all d\geq 1 by Cubre and Rouse [ 3 ] through the study of algebraic groups. More recently, Sanna [ 14 ] made an important contribution to this problem by computing the Dirichlet density for all odd d\geq 1 with 3\nmid d if the splitting field of f_{U} is \mathbb{Q}(\sqrt{-3}) . This was done using a method of Moree [ 12 ] . Prior to Sanna’s result, Ballot [ 1 ] considered the case of the splitting field being \mathbb{Q}(i) or \mathbb{Q}(\sqrt{-3}) , and d\in\{2,4\} or d\in\{3,6\} respectively. We want to stress that a cyclotomic splitting field causes many problems when d is divisible by one of 2 or 3 . Let a,b\in\bar{\mathbb{Q}} be the roots of f_{U} and K=\mathbb{Q}(a) . We denote by \Delta=a_{1}^{2}-4a_{2} the discriminant of f_{U} and \gamma=a/b its root quotient. Then, the set \mathcal{R}_{U}(d) is equal, up to finitely many exceptions, to \mathcal{R}_{\gamma}(d)=\{p:p\nmid 2a_{2}\Delta\text{ and }d\mid\mathrm{ord}_{\pi}(\gamma)\text{ for all $\pi\in\mathcal{O}_{K}$ and $p=\pi\cap\mathbb{Z}$}\}, where \mathcal{O}_{K} is the ring of integers of K and \ord_{\pi}(\gamma) is the order of \gamma\bmod\pi in the multiplicative group (\mathcal{O}_{K}/\pi)^{\times} . Indeed, by [ 14 , Lemma 4.1] , we have \rho_{U}(p)=\mathrm{ord}_{\pi}(\gamma) . The goal of our paper is to find the Dirichlet density, if it exists, of \mathcal{R}_{U}(d) in all the cases left by Sanna’s theorem. Moreover, we want to write it in a closed-form formula, meaning that only finitely many operations are required to compute it. By the above, the sets \mathcal{R}_{U}(d) and \mathcal{R}_{\gamma}(d) have the same Dirichlet density, when it exists. Therefore, we study the set \mathcal{R}_{\gamma}(d) , which is more suitable for some of our calculations. In the statement of [ 14 , Theorem 1.1] , Sanna assumed that d is odd and not divisible by 3 when \Delta_{K}=-3 . However, these restrictions are not used in the proofs of the existence of the density, nor of the upper bound. Indeed, [ 14 , Lemma 5.1] is stated without them and, while they appear in the statement of [ 14 , Lemma 5.3] , the proof does not invoke them. The proof of the main theorem only relies on these lemmas and [ 14 , Lemmas 5.2 and 5.4] . It is in the latter that the assumption on d is required to compute a closed-form formula of the density. Let x>1 and denote by \mathcal{R}_{\gamma}(d;x) the number of primes p\in\mathcal{R}_{\gamma}(d) with p\leq x . Let K_{n,d}=K(\zeta_{n},\gamma^{1/d}) , where d\mid n are integers and \zeta_{n} is a primitive n -th root of unity. We write \mathrm{Li} , \omega , and \varphi for the logarithmic integral, the number of distinct prime divisors, and the Euler totient functions respectively. With a slight change of the proof of [ 14 , Lemma 5.2] , the theorem can be restated, with our notation, as follows: Theorem 1 . Let d be an integer. There exists an absolute constant B>0 , such that for every x>\exp(Bd^{40}) , we have \mathcal{R}_{\gamma}(d;x)=\delta_{\gamma}(d)\mathrm{Li}(x)+\mathcal{O}_{\gamma}\left(\frac{d}{\varphi(d)}\cdot\frac{x(\log\log x)^{\omega(d)}}{(\log x)^{9/8}}\right), where (1) \delta_{\gamma}(d)=\sum_{v\mid d^{\infty}}\sum_{u\mid d}\frac{\mu(u)(1+[\sigma_{u,v}\text{ exists}])}{[K_{dv,uv}:\mathbb{Q}]}, and [\sigma_{u,v}\text{ exists}] is a boolean that checks the existence of a \sigma_{u,v} in Gal(K_{dv,uv}/\mathbb{Q}) satisfying \sigma_{u,v}(\sqrt{\Delta})=-\sqrt{\Delta} , \sigma_{u,v}(\zeta_{dv})=\zeta_{dv}^{-1} , and \sigma_{u,v}(\gamma^{1/uv})=\gamma^{-1/uv} . In this paper, we compute closed-form formulas for the series defined in ( 1 ) in all the cases left by Sanna [ 14 ] . This includes the more difficult cases of a cyclotomic splitting field K discussed above. We note that a particular case of our results was obtained independently by Luo, Hong, and Liu [ 11 ] . They extended Sanna’s result to even integers d\geq 2 under the assumption that K is not cyclotomic and \gamma is not an n -th power in K for all n\geq 2 . They used a method similar to Cubre and Rouse [ 3 ] . To simplify our calculations, we separate \delta_{\gamma}(d) into the smaller values (2) \delta_{\gamma}^{+}(d):=\sum_{v\mid d^{\infty}}\sum_{u\mid d}\frac{\mu(u)}{[K_{dv,uv}:\mathbb{Q}]}\quad\text{and}\quad\delta_{\gamma}^{-}(d):=\sum_{v\mid d^{\infty}}\sum_{u\mid d}\frac{\mu(u)[\sigma_{u,v}\text{ exists}]}{[K_{dv,uv}:\mathbb{Q}]}, according to the sum 1+[\sigma_{u,v}\text{ exists}] . The number \delta_{\gamma}^{+}(d) corresponds to the Dirichlet density of \mathcal{R}_{\gamma}^{+}(d) , the set of primes p\nmid a_{2} whose rank is divisible by d and with Legendre symbol (\Delta/p)=1 . For \delta_{\gamma}^{-}(d) , we have (\Delta/p)=-1 and we denote by \mathcal{R}_{\gamma}^{-}(d) the corresponding set. The second section is divided in two parts. First, we give results on powers in cyclotomic extensions of a quadratic field K . For instance, we explicit a 4 -th root of a norm 1 element \gamma\in\mathbb{Q}(i) in Lemma 3 . Next, we prove several results on cyclotomic-Kummer extensions of K . They are extensions of the form K(\zeta_{n},\gamma^{1/d}) , where n and d\mid n are positive integers, and \zeta_{n} is a primitive n -th root of unity. We first compute the degree over \mathbb{Q} of such extensions, and then give necessary and sufficient conditions for special automorphisms to exist in their Galois groups. In Section 3 , we compute closed-form formulas for the density \delta_{\gamma}(d) defined in ( 1 ). Let \zeta\in K be a root of unity and h(\zeta) be the largest n such that \gamma\in(K^{\times})^{n} . We prove our formulas under the assumption that h(1)\geq h(\zeta) for all roots of unity \zeta\in K . In the fourth section, we show that if h(1) is not maximal, then computing \delta_{\gamma}(d) can be