betaFeedback

WorksMathematics

On the exponential Diophantine equation (Pn(k))x+(Pr(k))x=Pm(k)\left(P_{n}^{(k)}\right)^{x}+\left(P_{r}^{(k)}\right)^{x}=P_{m}^{(k)}

Abstract

The Pell numbers satisfy Pn2+Pn+12=P2n+1P_n^2+P_{n+1}^2=P_{2n+1}, and Bravo, Luca and Pomeo recently found all the cases in which the sum of the xx-th powers of two consecutive kk-generalized Pell numbers is again a kk-generalized Pell number. We ask what happens when the two indices are arbitrary, that is, we solve (Pn(k))x+(Pr(k))x=Pm(k)\big(P_n^{(k)}\big)^x+\big(P_r^{(k)}\big)^x=P_m^{(k)} in positive integers k≥2k\ge2, n≥rn\ge r, mm and xx. The answer is that nothing new appears: apart from 1x+1x=21^x+1^x=2, every solution has consecutive indices. The proof combines linear forms in logarithms with an auxiliary Fibonacci equation which does not depend on kk, and, for large kk, with the fact that the dominant root of the sequence is very close to φ2φ^2. For the reduction step, the method of Dujella and Pethő fails for large indices and has to be replaced by a criterion of Legendre type.

Introduction

Among the many identities satisfied by the Pell numbers 0,1,2,5,12,29,70,169,…0,1,2,5,12,29,70,169,\ldots, one of the most pleasant is

Pn2+Pn+12=P2n+1(1)P_n^2+P_{n+1}^2=P_{2n+1} \tag*{(1)}

meaning that the sum of the squares of two consecutive Pell numbers is again a Pell number, as in 22+52=292^2+5^2=29 or 52+122=1695^2+12^2=169. The Fibonacci numbers have the same property,

Fn2+Fn+12=F2n+1,F_n^2+F_{n+1}^2=F_{2n+1},

and it is natural to ask what happens with higher powers. Marques and Togbé [13] showed that if Fnx+Fn+1xF_n^x+F_{n+1}^x is a Fibonacci number for all sufficiently large nn, then x∈{1,2}x\in\{1,2\}; Luca and Oyono [12] then proved that Fnx+Fn+1x=FmF_n^x+F_{n+1}^x=F_m has no solution at all with n≥2n\ge2 and x≥3x\ge3, and Gómez and Luca [9] extended this to the kk-generalized Fibonacci numbers. For the Pell numbers the question was settled in [17], by Rihane, B. Faye, Luca and Togbé: Pnx+Pn+1x=PmP_n^x+P_{n+1}^x=P_m has only the solutions given by (1). For the kk-generalized Pell numbers, Hasanalizade [10] treated the squares, and Bravo, Luca and Pomeo [6] recently solved the equation

(Pn(k))x+(Pn+1(k))x=Pm(k)\left(P_n^{(k)}\right)^x+\left(P_{n+1}^{(k)}\right)^x=P_m^{(k)}

completely.

In a companion paper [8] we study the equation

(Pn+1(k))x−(Pn−1(k))x=2Pm(k)\left(P_{n+1}^{(k)}\right)^x-\left(P_{n-1}^{(k)}\right)^x=2P_m^{(k)}

which comes from two other Pell identities,

Pn+1−Pn−1=2PnandPn+12−Pn−12=2P2n.P_{n+1}-P_{n-1}=2P_n \qquad\text{and}\qquad P_{n+1}^2-P_{n-1}^2=2P_{2n}.

While working on it, we asked ourselves what is really special about the consecutive indices in (1). Could a sum such as P7x+P3xP_7^x+P_3^x, with indices far from each other, be a kk-Pell number? For the Fibonacci numbers, the analogous question with an arbitrary gap was considered recently by Alvarenga, Chaves, Ramos, Silva and Sosa [1]. In this paper we answer it for the kk-generalized Pell numbers. Let k≥2k\ge2 and let P(k)=(Pn(k))nP^{(k)}=\left(P_n^{(k)}\right)_n be the kk-generalized Pell sequence. We study

(Pn(k))x+(Pr(k))x=Pm(k)(2)\left(P_n^{(k)}\right)^x+\left(P_r^{(k)}\right)^x=P_m^{(k)} \tag*{(2)}

in positive integers k≥2k\ge2, n≥rn\ge r, mm and xx, with no condition on the gap n−rn-r.

We began with a computer search, for 2≤k≤802 \le k \le80, n≤45n \le45 and x≤25x \le25. It found nothing new. Apart from the trivial 1x+1x=21^x + 1^x = 2, every solution it produced had r=n−1r = n - 1 and was already in the list of [6]; the curious solution 25+15=33=P5(3)2^5 + 1^5 = 33 = P^{(3)}_5, for example, appears for k=3k = 3 only, and it is the only non-trivial solution with an exponent larger than 2. The main result of the paper is that the search did not miss anything.

Theorem 1.1. All the solutions of (2) in positive integers k≥2k \ge2, n≥r≥1n \ge r \ge1, mm and xx are

(a) (n,r,m,x)=(1,1,2,x)(n,r,m,x) = (1,1,2,x) for every k≥2k \ge2 and x≥1x \ge1, namely 1x+1x=21^x + 1^x = 2;

(b) (n,r,m,x)=(2,1,3,2)(n,r,m,x) = (2,1,3,2) for every k≥2k \ge2, namely 22+12=52^2 + 1^2 = 5;

(c) (n,r,m,x)=(n,n−1,2n−1,2)(n,r,m,x) = (n,n-1,2n-1,2) for k=2k = 2 and every n≥2n \ge2, namely (1);

(d) (n,r,m,x)=(2,1,5,5)(n,r,m,x) = (2,1,5,5) for k=3k = 3, namely 25+15=33=P5(3)2^5 + 1^5 = 33 = P^{(3)}_5.

So the answer to our question is negative. Part of the explanation is a fact observed by Kılıç in [11]. He show that the first k+1k + 1 non-zero terms of P(k)P^{(k)} are the Fibonacci numbers of odd index, Pj(k)=F2j−1P^{(k)}_j = F_{2j-1} for 1≤j≤k+11 \le j \le k + 1. As long as all the indices in (2) are at most k+1k + 1, the equation does not see kk at all, and we are led to the following statement, which we prove first.

Theorem 1.2. The equation

F2n−1x+F2r−1x=F2m−1(3)F^{x}_{2n-1} + F^{x}_{2r-1} = F_{2m-1} \tag*{(3)}

has no solution in integers n≥3n \ge3, 1≤r≤n1 \le r \le n, m≥1m \ge1 and x≥2x \ge2.

The general strategy is the one of [6] and of our companion paper, but the gap d=n−rd = n - r is a new unknown, and this changes the argument in several places. When dd is large, PrxP^x_r is negligible compared with PnxP^x_n, and a linear form in the logarithms of gg, α\alpha and PnP_n is very small. When dd is small, PrxP^x_r is not negligible, and we have to keep the factor 1+α−dx1 + \alpha^{-dx} in a second form; this form can vanish, and for k=2k = 2 it does vanish, precisely because of (1). The case r=nr = n gives the equation 2Pnx=Pm2P^x_n = P_m; to see that the corresponding form does not vanish we needed a lower bound for the discriminant of the characteristic polynomial of P(k)P^{(k)} (see our Lemma 2.6).

There was also an obstacle we did not expect. For large nn, log⁡Pn\log P_n is extremely close to log⁡g+nlog⁡α\log g + n \log\alpha, so that the inhomogeneous part of the linear form is almost a combination of 1 and of the irrational number used in the reduction. The method of Dujella and Pethő then fails. We then replaced it, for these nn, by a criterion of Legendre type in which the small perturbation is controlled.

For large kk the dominant root α\alpha of P(k)P^{(k)} is very close to φ2\varphi^2, and all the linear forms can be replaced by forms in Q(5)\mathbb{Q}(\sqrt{5}) whose coefficients do not depend on kk. This gives k<1.57×1045k < 1.57 \times10^{45}, and a few rounds of reduction bring this down to k≤604k \le604. Table 1 summarizes the structure of the proof.

RangeToolResult
x=1x = 1, n≤2n \le2, (k,x,d)=(2,2,1)(k,x,d) = (2,2,1)elementary, [2, 16]§ 3
n≤k+1n \le k + 1, m≤k+1m \le k + 1Theorem 1.2 (forms in Q(5)\mathbb{Q}(\sqrt{5}))no solution
all kkforms in Q(α)\mathbb{Q}(\alpha), Matveevx<X(k)x < X(k), Prop. 5.3
k≥Kabsk \ge K_{\mathrm{abs}}forms in Q(5)\mathbb{Q}(\sqrt{5})none, Prop. 6.3
K0<k<KabsK_0 < k < K_{\mathrm{abs}}reduction in Q(5)\mathbb{Q}(\sqrt{5})none, Prop. 6.4
2≤k≤K02 \le k \le K_0reduction in Q(α)\mathbb{Q}(\alpha), searchProp. 7.1

Table 1. Structure of the proof.

Here Kabs=1.5694×1045K_{\mathrm{abs}} = 1.5694 \times10^{45} and K0=604K_0 = 604. All the computations were done in Python, the details are in Appendix § 9.

Preliminaries

The generalized Pell sequence

Let k≥2k \ge2 be an integer. The kk-generalized Pell sequence P(k):=(Pn(k))n≥2−kP^{(k)} := (P_n^{(k)})_{n\ge2-k} is defined by

Pn(k)=2Pn−1(k)+Pn−2(k)+⋯+Pn−k(k)for all n≥2,P_n^{(k)} = 2P_{n-1}^{(k)} + P_{n-2}^{(k)} + \cdots+ P_{n-k}^{(k)} \quad\text{for all } n \ge2,

References

  1. [1]R. Alvarenga, A. P. Chaves, M. E. Ramos, M. Silva, M. Sosa, Diophantine equations over the generalized Fibonacci sequences: exploring sums of powers, preprint (2025), arXiv:2501.08899.arxiv.org/abs/2501.08899
  2. [2]M. Bacchi, A. Togbé, On k-Pell numbers close to power of 2, Rev. Colombiana Mat. 58 (2024), no. 1, 67–80. https://doi.org/10.15446/recolma.v58n1.117434
  3. [3]J. J. Bravo, J. L. Herrera, Repdigits in generalized Pell sequences, Arch. Math. (Brno) 56 (2020), 249–262.DOI
  4. [4]J. J. Bravo, J. L. Herrera, F. Luca, On a generalization of the Pell sequence, Math. Bohem. 146 (2021), no. 2, 199–213.DOI
  5. [5]J. J. Bravo, J. L. Herrera, F. Luca, Common values of generalized Fibonacci and Pell sequences, J. Number Theory 226 (2021), 51–71.DOI
  6. [6]J. J. Bravo, F. Luca, F. Pomeo, An exponential Diophantine equation with generalized Pell numbers, Bull. Braz. Math. Soc. (N.S.) 56 (2025), Paper No. 56. https://doi.org/10.1007/s00574-025-00478-z
  7. [7]A. Dujella, A. Pethő, A generalization of a theorem of Baker and Davenport, Quart. J. Math. Oxford Ser. (2) 49 (1998), no. 195, 291–306.DOI
  8. [8]F. Faye, B. Faye, On the exponential Diophantine equation (P⁽ᵏ⁾ₙ₊₁)ˣ − (P⁽ᵏ⁾ₙ₋₁)ˣ = 2P⁽ᵏ⁾ₘ, preprint (2026).
  9. [9]C. A. Gómez, F. Luca, An exponential Diophantine equation related to the sum of powers of two consecutive k-generalized Fibonacci numbers, Colloq. Math. 137 (2014), no. 2, 171–188.
  10. [10]E. Hasanizada, On the Diophantine equation (P⁽ᵏ⁾ₙ)² + (P⁽ᵏ⁾ₙ₊₁)² = P⁽ᵏ⁾ₘ, Bull. Aust. Math. Soc. (2024).
  11. [11]E. Kılıç, On the usual Fibonacci and generalized k-Pell numbers, Ars Combin. 88 (2008), 33–45.
  12. [12]F. Luca, R. Oyono, An exponential Diophantine equation related to power of two consecutive Fibonacci numbers, Proc. Japan Acad. Ser. A Math. Sci. 87 (2011), 45–50.
  13. [13]D. Marques, A. Togbé, On the sum of power of two consecutive Fibonacci numbers, Proc. Japan Acad. Ser. A Math. Sci. 86 (2010), 174–176.DOI
  14. [14]E. M. Matveev, An explicit lower bound for a homogeneous rational linear form in logarithms of algebraic numbers II, Izv. Ross. Akad. Nauk Ser. Mat. 64 (2000), no. 6, 125–180; English translation in Izv. Math. 64 (2000), no. 6, 1217–1269.
  15. [15]M. R. Murty, J. Esmonde, Problems in Algebraic Number Theory, 2nd ed., Graduate Texts in Mathematics 190, Springer, New York, 2005.
  16. [16]M. N. Normenyo, S. E. Rihane, A. Togbé, Fermat and Mersenne numbers in k-Pell sequence, Mat. Stud. 56 (2021), no. 2, 115–123.
  17. [17]S. E. Rihane, B. Faye, F. Luca, A. Togbé, On the exponential Diophantine equation P⁽ᵏ⁾ₙ + P⁽ᵏ⁾ₙ₊₁ = Pₘ, Turkish J. Math. 43 (2019), no. 3, 1640–1649.

Paper details

vlnc:W_k4aj6juw
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC0-1.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Mathematics

Claude (Anthropic)

View statement
Looking for related papers…
Contents