On Regular Higher Power Rational Diophantine Triples
Alen Andrašek · 2026
Read On Regular Higher Power Rational Diophantine Triples by Alen Andrašek (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-2604-17018.txt.
Introduction
Andrašek studies rational Diophantine triples where each pair product plus one is a fourth power. He produces non-trivial infinite families of such triples with positive elements for fourth powers. The paper also looks briefly at the sixth- and eighth-power cases, explaining why the method resists extension to higher exponents.
Catalog record
| CREATOR | Alen Andrašek |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 36946 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.17018
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
On Regular Higher Power Rational Diophantine Triples Alen Andrašek Abstract . A rational Diophantine m m -tuple is a set { a 1 , … , a m } \{a_{1},\ldots,a_{m}\} of distinct nonzero rational numbers such that a i a j + 1 a_{i}a_{j}+1 is a square for all 1 ≤ i < j ≤ m 1\leq i<j\leq m . Similarly, we may ask when a i a j + 1 a_{i}a_{j}+1 is a k k -th power. Here, we study the case k = 4 k=4 and produce some non-trivial infinite families of such triples. We show that there are infinitely many triples with positive elements for k = 4 k=4 . We also briefly consider the k = 6 k=6 (sextic) and k = 8 k=8 (octic) cases, explaining the difficulties in extending the method to higher exponents. Keywords : Higher Power Rational Diophantine Triples, Regular Diophantine Triples, Quartic Diophantine Triples 1. Introduction A rational Diophantine m -tuple is a set \{a_{1},\ldots,a_{m}\} of distinct nonzero rational numbers such that a_{i}a_{j}+1 is a square for all 1\leq i<j\leq m . Sets with this property were first studied by Diophantus (see, e.g. [ 6 ] ) who found some particular examples such as \displaystyle\left\{\frac{1}{16},\frac{33}{16},\frac{17}{4},\frac{105}{16}\right\}. Fermat found the integer quadruple \{1,3,8,120\} and Euler extended it to a rational quintuple \displaystyle\left\{1,3,8,120,\frac{777480}{8288641}\right\}. Stoll [ Stoll_2019 ] proved that this extension was unique. Dujella [ 9 ] proved that the size m of an integer Diophantine m -tuple is at most 5 and it was proved that an integer Diophantine quintuple does not exist (see [ 13 ] ). On the other hand, there exist one-parameter infinite families for rational Diophantine sextuples (see [ 10 ] , [ 8 ] , [ 7 ] ). No rational Diophantine septuple is known and no upper bounds on the size m of a rational Diophantine m -tuple have been given. For more background on Diophantine m -tuples, see [ 10 ] . A k -th power rational Diophantine m -tuple is a set \{a_{1},\ldots,a_{m}\} of distinct nonzero rationals such that a_{i}a_{j}+1=t_{ij}^{k} for some rational t_{ij} . Dujella and Bugeaud [ 3 ] give the following examples: in the case k=3 , an integer triple is given by \displaystyle\{2,171,25326\} and in the case k=4 by \displaystyle\{1352,9539880,9768370\}. Batta, Szikszai and Tengely showed in [ 1 ] that there exists a rational k -th power Diophantine triple for any positive integer k by giving explicit parametric families. For even k , the set (1) \displaystyle\left\{-\frac{1}{r^{\frac{k}{2}}},\frac{r^{k}-1}{r^{\frac{k}{2}}},r^{\frac{k}{2}}\right\} is a Diophantine triple over \mathbb{Q}(r) . Similarly, they also give parametric families for odd k . In the case k=3 , they show that every cubic Diophantine pair except \{-1,1\} and \{-3,3\} can be extended to a Diophantine triple (this was also shown by Byeon and Fuchs [ 4 ] ). They also exhibited a parametric family of third power Diophantine quadruples. On the other hand, Dujella and Bugeaud in [ 3 ] showed that the size m of a k -th power integer m -tuple is bounded for any k but again no bound is known in the rational case. In this work, we will mainly study the case k=4 (which we will refer to as the quartic case). We introduce the notion of a regular higher power triple (Definition 2 ) in complete analogy with the classical definition of regularity (see [ 10 ] ). We show that regular quartic triples correspond to points on a surface \displaystyle\frac{s^{2}r^{2}-1}{s^{2}-r^{2}}=t^{2}, which we interpret as an elliptic curve with a parameter. By finding parametric points of infinite order, we find some parametric families of triples. This is done in a couple of ways. In all cases, the points are constructed by requiring an additional condition on r,s . From these we find rational parametrizations for r,s . To the author’s knowledge, no parametric solutions were known, except for (\ref{BST_parametrization}) . Using the new expressions, we show that there are infinitely many triples with all elements positive (the above parametric family ( 1 ) always has one negative element). Finally, we comment on ideas of how to generalize these ideas to 2k -th powers, in particular sextic (i.e. 6 -th power) triples. This leads to the system \displaystyle X^{k}+Y^{k} \displaystyle=Z^{k}+W^{k}, \displaystyle XYZW \displaystyle=\square, where X,Y,Z,W are rational numbers. We use the notation \square to denote an unspecified rational square. 2. Quartic Regular Triples First, let us fix some notation with the following definition. The notation is in complete analogy with the usual notation for Diophantine triples. Definition 1 . The triple \{a,b,c\} of distinct nonzero rationals is a quartic Diophantine triple if (2) \displaystyle\begin{split}ab+1&=r^{4},\\ ac+1&=s^{4},\\ bc+1&=t^{4},\end{split} for some rational r,s,t . We want to find rational a,b,c that satisfy such equations. Immediately, we can satisfy the first two equations by defining b,c so that (3) \displaystyle\begin{split}b&=\frac{r^{4}-1}{a},\\ c&=\frac{s^{4}-1}{a}.\end{split} This leaves us with the problem of determining when the third equation in ( 2 ) is satisfied. To simplify things, we introduce an analogue to the concept of a regular Diophantine triple, in which case the third equation reduces to that of an elliptic curve over \mathbb{Q}(r) . Definition 2 . With notation as in Definition 1 , a quartic Diophantine triple is called regular if a=s^{2}-r^{2} (or equivalently if c=a+b+2r^{2} ). To find a triple \{a,b,c\} such that it is a quartic Diophantine triple, express b,c as in ( 3 ) and plug into the third equation to find that, \displaystyle t^{4}=bc+1 \displaystyle=\frac{r^{4}-1}{a}\frac{s^{4}-1}{a}+1 \displaystyle=\frac{(r^{4}-1)(s^{4}-1)+a^{2}}{a^{2}}. Now we can use regularity, \displaystyle t^{4}=bc+1 \displaystyle=\frac{(r^{4}-1)(s^{4}-1)+(s^{2}-r^{2})^{2}}{(s^{2}-r^{2})^{2}} \displaystyle=\frac{(s^{2}r^{2}-1)^{2}}{(s^{2}-r^{2})^{2}}. Taking square roots, we are left to determine when \displaystyle\frac{s^{2}r^{2}-1}{s^{2}-r^{2}}=\pm t^{2}. Without loss of generality, we may assume the plus sign, for a minus sign would just amount to switching r and s . Proposition 1 . Suppose r,s are rationals such that s,r\neq\pm 1 , s\neq\pm r and (4) \displaystyle\frac{s^{2}r^{2}-1}{s^{2}-r^{2}} \displaystyle=t^{2}. Then (a,b,c) is a rational quartic Diophantine triple with a=s^{2}-r^{2} , b=\frac{r^{4}-1}{a}=t^{2}-r^{2} , c=\frac{s^{4}-1}{a}=s^{2}+t^{2} . Proof. We have already proved that this is a sufficient condition for a quartic Diophantine triple to exist. Notice that ( 4 ) can be rewritten as \displaystyle r^{2}s^{2}+r^{2}t^{2}=1+s^{2}t^{2}. We may rewrite this as \displaystyle r^{4}-1 \displaystyle=(r^{2}-t^{2})(r^{2}-s^{2}). Now r^{4}-1=ab and s^{2}-r^{2}=a , so upon dividing by a we get that \displaystyle b=t^{2}-r^{2}. Finally, c=a+b+2r^{2}=s^{2}+t^{2} . This proves the expressions for b,c . ∎ We shall see that there are infinitely many rational solutions to equation ( 4 ) and, in fact, there is even an integral solution as indicated by a computer search. Some smaller integer r,s are given by \displaystyle(r,s)=(337,339),(337,3107),(507,1242). These are the only ones with r<s<10000 . Unfortunately, the third pair does not give an integer Diophantine triple (a,b,c) (the corresponding t is not an integer in that case). The first two pairs give the same regular triple \displaystyle(a,b,c)=(1352,9539880,9768370), as already found by Dujella and Bugeaud [ 3 ] . Thus, the only known integer quartic triple is in fact regular. We also note that the rational parametric triple ( 1 ) is regular. 2.1. A Special Case Notice that the pair (r,s)=(337,339) satisfies s=r+2 . More generally, let s=r+\alpha . Then ( 4 ) becomes (5) \displaystyle\frac{r^{2}(r+\alpha)^{2}-1}{\alpha(2r+\alpha)}=t^{2}. In the special case that \alpha=2 one has cancellation of polynomial factors in the