Ramanujan Primes and Negative Pell’s Equation

  • V. SangeethaDepartment of Mathematics National College(Autonomous) (Affiliated to Bharathidasan University), Trichy - 620001,Tamil Nadu, India.
  • T. AnupreethiPG and Research Department of Mathematics.
  • S. Manju SomanathDepartment of Mathematics, National College(Autonomous)( Affiliated to Bharathidasan University), Trichy - 620 001, Tamil Nadu, India.

Abstract

The Diophantine equation under consideration to find the non-zero integral solution are of the type x = (Rp)y − (R1) which represents a very general type of negative Pell’s equation formed using Ramanujan primes as coefficients. The equation are all formed using the first 10 Ramanujan primes

st

2,11,17,29,41,47,59,67,71 & 97. Fixing the 1 Ramanujan prime R1 = 2, we seek for solutions of Pell’s

th

equation with coefficients Rp, where Rp denotes the p Ramanujan prime.

MSC Classification Number: 11D09,11D99.

Keywords:Diophantine equationRamanujan primeNegative Pell’s equationIntegral solutionsBrah-

magupta lemma.

1. Introduction

The Diophantine equation x − Dy = 1, also known as Pell’s equation, is known to have an infinite number of integer solutions for any square - free positive integer D > 1 [2]. The solutions stem from a single fundamental solution found in Algebraic Number Theory. Brahmagupta and Bhaskara were the first to study the classical Pell’s equation, and Lagrange developed the entire theory.There are an infinite number of integer solutions (xn, yn) to Pell’s equation x −Dy = ±1 for n ≥ 1. Since all other solutions to this equation can be (easily) derived from the first non-trivial positive integer solution (x1, y1) (in this case, x1 or x1 + y1 D is minimum), it is known as the fundamental solution. Let x − Dy = 1 have a fundamental solution (x1, y1). Then, for any integer n ≥ 2, the n-th positive solution of this equation, (xn, yn), is defined by the equality xn + yn D = (x1 + y1 D) .The corresponding negative Pell’s equation, x − Dy = −1, (1) is considerably more mysterious. For (1) to be solvable in integers, it is required that D ≡ 1 or 2 (mod 4), and that all odd prime divisors of D have forms that are congruent to 1 modulo 4. These conditions, however, do not guarantee the existence of a solution. One may visit A031396 in the OEIS [16] for some known numbers D = 2, 5, 10, 13, 17, 26, 29, 37, 41, 50, ... such that (1) is solvable in integers x and y.For several decades, certain number theorists have been committed to developing standards for the solvability of (1). A few requirements are Qr listed in [2, 10, 11, 17]. According to Newman’s [11] demonstration, (1) can be solved in integers if D = pi, pi where r = 2 or r is odd, and pi’s are primes congruent to 1 modulo 4 and satisfy = −1 for all 1 ≤ i, j ≤ r pj with i = ̸ j. The standard Jacobi symbol is in this case. Conversely, Mollin [10] found a relationship between x − Dy = 1 and the negative Pell’s equation (1).

He demonstrated that (1) can be solved in integers x and y if and only if x0 ≡ −1 (mod 2D) is satisfied by the fundamental solution (x0, y0) of x − Dy = 1. Despite the significance of these findings, their approaches are constrained if D ≫ 1. For instance, there are numerous situations where the fundamental solutions are fairly large. Thus, extensive computations are required to verify these conditions. Positive Pell’s equations have recently been found to have well-known solvability criteria. A primitive Pythagorean triple for D can be found by using the length of its period as a guide. Another method is to compute the simple continued fraction of √ D.[18-20] A prime number that satisfies a result about the prime-counting function proved by Srinivasa Ramanujan is known as a Ramanujan prime in mathematics. As he points out, Bertrand’s postulate was initially proven by Chebyshev. Ramanujan presented a revised demonstration of the concept in 1919. Ramanujan arrived to a generalized result at the end of the two-page published work. Consequently, 2, 11, 17, 29, and 41 are the initial five Ramanujan primes.

Take note that Rn must be a prime number in order for it to exist:π(x) − π and, hence,π(x) must increase x Rn by obtaining another prime at x = Rn. Since π(x) − π can increase by atmost 1,π(Rn) − π = n. In

2 2 t th

this paper the negative Pell’s equation x = (Rp)y − (R1) formed using the p Ramanujan prime Rp, are analysed for obtaining non-zero integer solution for various choice of t and are illustrated for primes less then 100.

2. Preliminaries

Theorem 0.1. If x1, y1 is the fundamental solution of x − Dy = 1,then every positive solution of the equation is given by xn, yn where xn and yn are the integers determined from xn + yn D = (x1 + y1 D) , n = 1, 2, 3 · · ·

For illustration,consider the fundamental solution x1 = 24 & y1 = 5 of x − 23y = 1. A second positive solution (x2, y2) can be obtained from the formula √ √ √ x2 + y2 23 = (24 + 5 23) = 1151 + 240 23

This gives x2 = 1151, y2 = 240.These integers also satisfy the equation

2 2 2 2

x − 23y = 1,since 1151 − 23y = 1.

∴ Any positive solution can be calculated using the Theorem 2.1.

Theorem 0.2. Suppose that the equation x − Dy = −1 (2) has solution in positive integers and (x0, y0) be its minimal solution.The general solution to (2) is given by (xn, yn)n≥0 where xn = x0un +Dy0vn,yn = y0un +x0vn and (un, vn)n≥0 is the general solution to Pell’s equation u − Dv = 1.

Theorem 0.3. Let p be a prime. The negative Pell’s equation x − py = −1 is solvable if and only if p = 2 or p ≡ 1 (mod 4).

2.1 Testing the solubility of the negative Pell’s equation

Suppose D is a positive integer and not a perfect square. Then the negative Pell’s equation x − Dy = −1 is soluble if and only if D is expressible as

2 2 2 2

D = a + b , gcd(a, b) = 1, a and b positive, b odd and the Diophantine equation −bv + 2avw + bw = 1 has a solution(the case of solubility occurs for exactly one such (a,b)) .

The Algorithm

( 2.1.1 ). Find all expressions of D as a sum of two relatively prime squares using cornacchia’s method.If none exist, the negative Pell’s equation is not soluble.

(2.1.2). For each representation D = a + b ,gcd(a, b) = 1, a and b positive, b odd,test the solubility of −bv + 2avw + bw = 1 using Lagrange-Matthews algorithm.If solutions exist,the negative Pell’s equation is soluble.

(2.1.3). If each representation yields no solution,then the negative Pell’s equation is insoluble.

3. Main Results

2 2 t th

The equation under consideration in the negative Pell’s equation x = (Rp)y − (R1) , where Rp denotes the p Ramanujan prime and R1 = 2,the first Ramanujan Prime.

System-I: Diophantine equation x = 11y − 2 ,with R2 = 11 and R1 = 2.

This system concerns with negative Pell’s equation x = 11y − 2 , t ∈ N.Here the prime R2 = 11 cannot be expressed as the sum of two relatively prime squares.Thus by the condition (2.1.1) of the algorithm,the negative Pell’s equation x − 11y = −1 is not soluble. Thus we can conclude that the negative Pell’s equation x = 11y − 2 is insoluble.

System-II: Diophantine Equation x = 17y − 2 with R3 = 17, R1 = 2.

The prime R3 = 17 satisfies all the conditions of Theorem 2.3,and hence we conclude that the negative Pell’s equation x = 17y − 2 ,t ∈ N is solvable and infinitely many positive integer solutions are obtained for the various choices of t.The more general Pell’s equation u = 17v + 1 (I) can be viewed and its general solution (un, vn) is given by

un = fn; vn = √ gn, (II)

Choice-I: t is an even number Case 1:t=2, The Pell’s equation under consideration is

with the initial solution x0 = 8; y0 = 2.Applying Theorem 2.2, a series of non-zero integer solutions to (3) are discovered as

yn = √ [2 17fn + 8gn] ,where fn and gn are defined as in (II)

The recurrence relations satisfied by the solutions of (3) are given by

yn+2 − 66yn+1 + yn = 0 Case 2:t = 2k + 2, where 1 ≤ k ≤ 7, The Pell’s equation thus obtained is with initial solution x0 = 2 , y0 = 2 .Applying Brahmagupta lemma between the solutions(x0, y0) and (un, vn) from (II),the sequence of non-zero distinct integer solutions to (4) are obtained as

xn = [fn + 17gn]

yn = √ [ 17fn + gn].

Case 3: t=2k+16, k ∈ N

with initial solution x0 = 59(2 ), y0 = 125(2 ) Applying Theorem 2.2, and using (II) we get infinite number of solutions to (5) as

Choice-II: t is an odd number Case 1: t = 2k + 1, 1 ≤ k ≤ 4 The Pell’s equation to be solved is

The following table represents various values of t with the initial & general solutions to (6).

t = 2k + 1 (x0, y0) (xn, yn) 1 √ 1 √ k = 1 (3,1) xn = [3fn + 17gn] ; yn = √ [ 17fn + 3gn] 1 √ 1 √ k = 2 (6,2) xn = [6fn + 2 17gn] ; yn = √ [2 17fn + 6gn] 1 √ 1 √ k = 3 (5,3) xn = [5fn + 3 17gn] ; yn = √ [3 17fn + 5gn] 1 √ 1 √ k = 4 (10,6) xn = [10fn + 6 17gn] ; yn = √ [6 17fn + 10gn] Case 2: t=2k+9, k ∈ N The Pell’s equation thus obtained is

Let (x0, y0) be the initial solution of (7)given by x0 = 3(2 ), y0 = 11(2 ).

Applying Theorem 2.2, a series of non-zero integer solution to (7) are discovered as

yn = √ [11 17fn + 3gn], where fn and gn are defined as in (II).

The recurrence relations satisfied by the solutions of (7) are given by

yn+2 − 66yn+1 + yn = 0 System -III: Diophantine Equation x = 29y − 2 with R4 = 29, R1 = 2.

Given that R4 = 29 satisfies all the settings of Theorem 2.3,we can conclude that the negative Pell’s equation x = 29y − 2 is solvable in integers.

Consider the case when t is an even number.

In this case,the Pell’s equation is

2 2 2k

Let (x0, y0) be the primary solution of (8) specified by x0 = 5(2 ), y0 = 2 . Analysing the pell’s equation u = 29v + 1 to discover the additional solutions to (8), whose generic solution (un, vn) is provided by un = fn; vn = √ gn

Applying Brahmagupta lemma between the solutions(x0, y0) and (un, vn),the sequence of non-zero distinct integer solutions to (8) are obtained as

The recurrence relations fulfilled by the solutions of (8) are specified by

yn+2 − 1960yn+1 + yn = 0 Observation:This system fails to generate integer solutions for all the odd choices of t.

System-IV:Diophantine Equation x = 41y − 2 with R5 = 41, R1 = 2.

For this particular equation,we consider the prime 41.Using the Algorithm and testing with (a, b) = (4, 5),−bv + 2avw + bw = 1 has a solution (v, w) = (2, 1).Therefore,we can develop this discussion by confirming that the negative pell equation x = 41y − 2 .t ∈ N is solvable in integers.

Consider the pellian equation u = 41v + 1 (III) whose general solution (un, vn) is given by

un = fn; vn = √ gn (IV)

Choice-1: t is an even number

The Pell’s equation to be checked for its solutions is

Assume (x0, y0) be the fundamental solution of (9) given by x0 = 64, y0 = 10 Applying Theorem 2.2, a collection of nonzero different integer solutions to (9) is achieved as The recurrence relation derived by employing the solutions are

yn+2 − 4098yn+1 + yn = 0.

Case2: t = 2k + 2, 1 ≤ k ≤ 3

Let (x0, y0) represent the first solution to (10) proposed by x0 = 5(2 ), y0 = 2 Applying Brahmagupta Lemma between (x0, y0) and (un, vn) from (IV),the series of different non-zero integer solutions are discovered as

Case 3: t = 2k + 8, k ∈ N The Pell’s equation is

Let (x0, y0) represent the initial solutions to (11) offered by x0 = 2 , y0 = 5(2 ) Applying Brahmagupta Lemma between (x0, y0) and (un, vn) from (IV),the sequence of non-zero distinct integer solutions to (11) are obtained as

xn = [fn + 5 41gn]

yn = √ [5 41fn + gn].

Choice-2: t is an odd number

Applying Theorem 2.2 ,a collection of nonzero different integer solutions to (12) are achieved as

Case 2: t = 2k + 3, k ∈ N The Pell’s equation to be examined is

Applying Brahmagupta Lemma between (x0, y0) and (un, vn) from (IV),the sequence of non-zero distinct integer solutions to (13) are obtained as

System-V

In this system, we consider the Ramanujan primes R6 = 47, R7 = 59, R8 = 67 and R9 = 71.We form Pell’s

2 2 t 2 2 t 2 2 t 2 2 t

equations x = 47y − 2 , x = 59y − 2 , x = 67y − 2 and x = 71y − 2 .All these primes fail to satisfy the conditions of Theorem 2.3 and thus we conclude that, the above equations so formed does not possess integer solutions.

System-VI Diophantine Equation x = 97y − 2 with R10 = 97, R1 = 2.

We consider the prime 97 in support of this precise equation. Given that 97 meets every requirement of Theorem 2.3,we can come to the conclusion that negative the Pell’s equation x = 97y − 2 is solvable.The more general Pell’s equation u = 97v + 1 (V)

can be viewed and its general solution (un, vn) is given by

un = fn; vn = √ gn, (VI)

Choice-1:t is an even number.

The Pell’s equation so generated is

Let (x0, y0) represent the initial solution to (14) offered by x0 = 11208, y0 = 1138.Applying Brahmagupta Lemma between (x0, y0) and (un, vn) from (VI),the sequence of non-zero distinct integer solutions to (14) are obtained as

Case 2: t = 2k + 2, k = 1, 2, 3, 4, 5

The sequence of non-zero distinct integer solutions to (15) are obtained as

Using Theorem 2.2,the general solution to x = 97y − 2 is given by

Case 4: t = 2k + 14, k ∈ N The Pell’s equation is

has solutions in positive integer and let (6(2 ), 26(2 )) be its minimal solution.The general solution to (16)is given by (xn, yn) where

Choice-2:t is an odd number

Case 1: t=3,5,7 The Pell’s equation is

The following table represents various values of t with the initial & general solutions to (17).

Case 2: t = 2k + 7, k ∈ N The Pell’s equation is t (x0, y0) (xn, yn) 1 √ 1 √ 3 (325,33) xn = [325fn + 33 97gn] ; yn = √ [33 97fn + 325gn] 1 √ 1 √ 5 (29,3) xn = [29fn + 3 97gn] ; yn = √ [3 97fn + 29gn] 1 √ 1 √ 7 (58,6) xn = [58fn + 6 97gn] ; yn = √ [6 97fn + 58gn]

with initial solution x0 = 19(2 ),y0 = 3(2 ).Applying Brahmagupta lemma between (x0, y0) and (un, vn) from (VI), the progression of non-zero different integer solutions to (18) is found to be

4. Conclusion

In this paper, the equations are all formed restricting the primes less than 100. It can be extended to any bound and also to any prime number and thus delve into the world of Pell’s equations.

5. Author Contribution

Credit authorship contribution statement:

Manju Somanath: Visualization, Validation,Conceptualisation.

V.Sangeetha: Writing – original draft, Validation, Data curation.

T.Anupreethi: Writing – original draft, Visualization,Validation.

Declaration of competing interest:

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References

  1. [1] Anupreethi,T.,Vinmol.K.Jesudas.,Manju Somanath.,Sangeetha,V.,Kannan,J., Vijaya Shanthi,P. (2025). Cryptographic Application of Elliptic Curve generated through the formation of Diophantine triples using Hex Numbers and Pronic Num-bers.Communications on Applied Nonlinear Analysis, 32(9s) (2025), 836-840. https://doi.org/10.52783/cana.v32.4017
  2. [2] Grytczuk, A., Luka, F., & Wojtwicz, M. (2000). The negative Pell equation and Pythagorean triples. Proceedings-japanacademy series a mathematical sciences, 76(6), 91-94. https://www.researchgate.net
  3. [3] Kannan, J., & Somanath, M. (2023). Fundamental perceptions in contemporary number theory. Nova Science Publishers. https://doi.org/10.52305/RRCF4106
  4. [4] Gopalan, M. A., Sangeetha, V., & Somanath, M. (2014). Construction of strong and almost strong rational Diophantine quadruples. JP Journal of Algebra, Number Theory and Applications, 35(1), 35-48. https://www.researchgate
  5. [5] Manju Somanath.,Anupreethi,T., Sangeetha, V. (2022). Lattice points for the quadratic diophantine equation 21(x2+y2)− 19xy = 84z2. International journal of innovative research science, engineering and technology (IJIRSET), 11(1), 5542-5549. https://www.ijirset.com/upload/2022/may/165_LATTICE_HARDCERT1.pdf
  6. [6] Anupreethi, T., Somanath, Manju., & Venugopal, Sangeetha. (2022). Integral solutions of quadratic diophantine equation with two unknowns 11(2 + Ω2 ) = 2(12 Ω−1). Research and Reflections on Education, 20(3a), 108-111. https://www.researchgate.net
  7. [7] Manju Somanath., Sangeetha,V., Anupreethi,T. (2023). Special Dio-triples involving Primes. CRC Press Taylor& Francis Group. https://www.taylorfrancis.com/
  8. [8] Manju Somanath., Anupreethi,T., Sangeetha,V. (2023). Integral solutions of Quadratic Diophantine EquationM2+ N2− 14M + 18N = 130(R2 −1). International Journal of innovative science and Research Technology, 8(11) (2023), 2486-2488. https://www.ijisrt.com/assets/upload/files/IJISRT23NOV2280.pdf
  9. [9] Manju Somanath., Sangeetha,V., Anupreethi,T. (2023). Solutions in integers for the exponential Diophantine equation (Pp)ȷ + 3` = λ 2. International Journal of Science and Applied Research, 10(12),10-14. https://www.ijsar.in/Admin/pdf/SOLUTIONS-IN-INTEGERS-FOR-THE-EXPONENTIAL-DIOPHANTINE-EQUATION.pdf
  10. [10] Mollin, R. A., & Srinivasan, A. (2010). A note on the negative Pell equation. International Journal of Algebra, 4(19), 919-922. https://www.m-hikari.com/ija/ija-2010/ija-17-20-2010/mollinIJA17-20-2010.pdf
  11. [11] Newman, M. (1977). A note on an equation related to the Pell equation. The American Mathematical Monthly, 84(5), 365-
  12. [366] https://doi.org/10.1080/00029890.1977.11994359
  13. [12] Ramanujan, S.,A proof of Bertrand’s postulate,Journal of the Indian Mathematical Society.,11 (1919), 181-182.
  14. [13] Sangeetha, V., Anupreethi, T., & Somanath, M. (2023). Construction of Special dio—triples. Indian Journal Of Science And Technology, 16(39), 3440-3442. https://sciresol.s3.us-east-2.amazonaws.com/IJST/Articles/2023/Issue-39/IJST-2023-1735.pdf
  15. [14] Sangeetha, V., Anupreethi, T., & Somanath, M. (2024). Cryptographic application of elliptic curve generated through centered hexadecagonal numbers. Indian Journal of Science and Technology, 17(20), 2074-2078. https://doi.org/10.17485/IJST/v17i20.1183
  16. [15] Sangeetha,V.,Anupreethi,T., Manju Somanath. (2024). Exponential diophantine equation in two variables. Bulletin of calcutta mathemaical society, 116(4), 425-430.
  17. [16] Sloane, N. J. (2010). The on-line encyclopedia of integer sequences. https://www.academia.edu/down-load/97072750/187495321.pdf#page=221
  18. [17] Sierpi´nski, W. (1964). Elementary Theory of Numbers, Polish Scientific Publishers, Warszawa,
  19. [18] Das, R., Somanath, M., & VA, B. (2024). Solution of negative pell's equation using self primes. Palestine journal of math-ematics, 13(4), P1005. https://openurl.ebsco.com
  20. [19] Kannan, J., Somanath, M., & Raja, K. (2019). On a class of solutions for the hyperbolic diophantine equation. Interna-tional Journal of Applied Mathematics, 32(3), 443-449. https://diogenes.bg/ijam/contents/2019-32-3/6/6.pdf
  21. [20] Titu Andreescu, Dorin Andrica, Ion Cucurezeanu. (2010). An introduction to diophantine equations. Springer New York Dordrecht Heidelberg London. https://doi.org/10.1007/978-0-8176-4549-6