Elementary Bounds and the Riemann Hypothesis: Analyzing the Robin Criterion

Download Full Article

This article is available as a downloadable PDF with complete code listings and syntax highlighting.

Introduction

The Riemann Hypothesis (RH) stands as the most significant unsolved problem in pure mathematics, asserting that all non-trivial zeros of the Riemann zeta function zeta(s) lie on the critical line where the real part of s is 1/2. While the hypothesis is typically framed in the context of complex analysis and the distribution of prime numbers, a powerful alternative approach exists within the realm of elementary number theory. This approach translates the analytic properties of zeta(s) into inequalities involving arithmetic functions, most notably the sum-of-divisors function sigma(n).

The research paper arXiv:hal-04494457v2, authored by Frank Vega, investigates this elementary frontier. The core of the analysis rests upon the Robin criterion, which established in 1984 that the Riemann Hypothesis is equivalent to the assertion that sigma(n) is less than e^gamma * n * log(log n) for all n greater than 5040, where gamma is the Euler-Mascheroni constant.

Mathematical Background

The sum-of-divisors function, denoted as sigma(n), is a multiplicative function defined as the sum of all positive divisors of n. Its relationship to the Riemann zeta function is established through a specific identity where the sum of sigma(n) weighted by n to the power of negative s relates to the product of zeta(s) and zeta(s-1). This identity shows that the behavior of sigma(n) is governed by the poles and zeros of the zeta function.

In 1913, Gronwall proved the maximum growth of sigma(n), but it was Guy Robin who refined this by showing its equivalence to the Riemann Hypothesis. This shifted the focus to superabundant numbers and colossally abundant numbers. A number n is colossally abundant if there exists an exponent epsilon greater than zero such that the ratio of sigma(n) to n raised to the power of 1 plus epsilon is maximized. These numbers are the most likely candidates to violate the Robin inequality.

Main Technical Analysis

The Robin Ratio and Colossally Abundant Numbers

The core technical contribution of arXiv:hal-04494457v2 involves the analysis of the Robin ratio G(n) = sigma(n) / (n log log n). Vega examines the behavior of this ratio specifically for colossally abundant numbers. The paper argues that for any counterexample to the Robin inequality to exist, it must occur at one of these extremely abundant integers.

The analysis begins by expressing sigma(n)/n as a product over the prime factors of n. As n becomes colossally abundant, the exponents of its prime factors are determined by a parameter epsilon. Vega focuses on the logarithmic form of the Robin inequality, using prime product representation to approximate the ratio using sums of reciprocals of primes.

Asymptotic Growth and the Chebyshev Function

A significant portion of the analysis in arXiv:hal-04494457v2 is dedicated to the relationship between the Robin ratio and the distribution of primes via the Chebyshev function, which sums the logarithms of primes. The paper demonstrates that the difference between the discrete sum over primes and the continuous integral representation maintains a specific bound that prevents the Robin ratio from exceeding the threshold e^gamma.

Vega employs a method of log-density analysis, considering the sequence of colossally abundant numbers and the associated values of the Robin ratio. By analyzing the increments between consecutive colossally abundant numbers, the paper attempts to show that the increase in the divisor ratio is always compensated by the increase in the logarithmic bound.

Novel Research Pathways

Generalized Robin Criteria: Extending the methodology of arXiv:hal-04494457v2 to investigate the Generalized Riemann Hypothesis (GRH) for Dirichlet L-functions.
Arithmetic Progressions: Analyzing the behavior of the Robin ratio for integers whose prime factors are restricted to specific arithmetic progressions, linking local distributions to the growth of arithmetic functions.
Computational Sieve Methods: Developing parallelized algorithms to explore the Robin ratio in the dark regions of extremely large integers (e.g., 10^10000), providing an empirical check on the analytic bounds.

Computational Implementation

(* Section: Robin Inequality and Zeta Verification *)
(* Purpose: This code calculates the Robin Ratio for primorials and *)
(* compares the growth to the values of the Zeta function. *)

Module[{
  maxPrimes = 100, 
  gamma = EulerGamma, 
  primorials, 
  robinRatios, 
  threshold, 
  zetaCheck
},
  (* Generate the first n primorials *)
  primorials = Table[Product[Prime[i], {i, 1, k}], {k, 3, maxPrimes}];
  
  (* Define the Robin Ratio G(n) = sigma(n) / (n * log(log(n))) *)
  RobinRatio[n_] := DivisorSigma[1, n] / (n * Log[Log[n]]);
  
  (* Calculate ratios for the primorial sequence *)
  robinRatios = Table[
    {Log[Log[primorials[[k]]]], RobinRatio[primorials[[k]]]}, 
    {k, 1, Length[primorials]}
  ];
  
  (* Include Zeta and ZetaZero for comparative growth analysis *)
  zetaCheck = Table[N[Abs[Zeta[1/2 + I*Im[ZetaZero[n]]]], 20], {n, 1, 5}];
  
  (* The Robin bound constant e^gamma *)
  threshold = Exp[gamma];
  
  (* Output the maximum ratio found and the check result *)
  Print["Maximum Robin Ratio found: ", Max[robinRatios[[All, 2]]]];
  Print["Robin Bound (e^gamma): ", N[threshold, 15]];
  Print["Zeta Magnitudes on Critical Line: ", zetaCheck];
  
  ListLinePlot[
    robinRatios,
    AxesLabel -> {"log log n", "G(n)"},
    PlotLabel -> "Robin Ratio vs. Mertens Constant",
    Epilog -> {Red, Dashed, InfiniteLine[{0, threshold}, {1, 0}]}
  ]
]

Conclusions

The analysis of arXiv:hal-04494457v2 provides a compelling look at the Riemann Hypothesis through the lens of elementary number theory. By focusing on the Robin inequality and the growth patterns of colossally abundant numbers, the paper bridges the gap between the analytic properties of the zeta function and the discrete behavior of the divisor function. The most promising avenue for immediate progress appears to be the refinement of error terms in prime-sum approximations and the computational exploration of near-colossally abundant numbers.

References

Vega, F. (2024). A proof of the Riemann hypothesis. arXiv:hal-04494457v2

Robin, G. (1984). Grandes valeurs de la fonction somme des diviseurs et hypothèse de Riemann. Journal de Mathématiques Pures et Appliquées, 63, 187-213.