The Arithmetic Bridge: Mertens Function Bounds and the Riemann Hypothesis

Download Full Article

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

Introduction

The Riemann Hypothesis remains the most significant unsolved problem in pure mathematics, representing a fundamental bridge between the distribution of prime numbers and the analytic properties of the complex zeta function. Specifically, the hypothesis asserts that all non-trivial zeros of the Riemann zeta function, ζ(s), lie on the critical line where the real part of s is exactly 1/2. While the hypothesis is usually framed in terms of the zeros of ζ(s), an equivalent and deeply insightful formulation exists in the language of arithmetic functions, specifically the Möbius function μ(n) and its partial sums, known as the Mertens function M(n).

The research presented in the paper arXiv:hal-01667383v4 by M. Sghiar focuses on this exact intersection. The paper explores the analytic continuation of the reciprocal of the zeta function, 1/ζ(z), from the half-plane Re(z) > 1 to the critical strip. By leveraging the relationship between the Mertens function and the Dirichlet series of 1/ζ(z), Sghiar proposes a methodology to establish that the non-trivial zeros must be confined to the line Re(z) = 1/2. This analysis is crucial because it transforms a question about the location of zeros into a question about the convergence of an integral involving the Mertens function.

Mathematical Background

The foundation of this analysis rests on several key mathematical objects. The Riemann zeta function is defined for Re(s) > 1 by the Dirichlet series ζ(s) = Σ 1/n^s. Its reciprocal can be expressed using the Möbius function μ(n), which is defined as μ(n) = (-1)^k if n is a product of k distinct primes, and 0 otherwise. This gives us 1/ζ(s) = Σ μ(n)/n^s for Re(s) > 1.

The Mertens function M(x) = Σ μ(n) for n ≤ x represents the cumulative sum of these values. A central identity used in arXiv:hal-01667383v4 connects these via a Mellin-type integral transform:

1/ζ(z) = z * integral from 1 to infinity of M(x)/x^z+1 dx

This integral converges for Re(z) > 1 and provides the mechanism for analytic continuation. The key insight is that the growth properties of M(x) directly control the convergence of integrals that define the domain of 1/ζ(s).

Main Technical Analysis

Spectral Properties and Zero Distribution

The source paper explores the spectral properties of the zeta function, particularly in relation to the distribution of its zeros. The author shows that the zeros of the zeta function are closely related to the prime numbers, and that the distribution of these zeros has significant implications for the Riemann Hypothesis. If M(x) exhibits controlled growth, specifically M(x) = O(x^1/2+ε), then the integral representation of the reciprocal zeta function converges for Re(z) > 1/2.

The Impact of Mertens Bounds

A crucial theorem from arXiv:hal-01667383v4 establishes that if |M(x)| < sqrt(x) for all x ≥ 1, then 1/ζ(s) extends analytically to Re(s) > 1/2. While the original Mertens conjecture (|M(x)| < sqrt(x)) was disproven by Odlyzko and te Riele in 1985, the underlying principle remains sound: any bound of the form x^1/2+ε is sufficient to prove the Riemann Hypothesis. The paper argues that by using properties of the Möbius function and prime distributions, one can demonstrate that 1/ζ(z) is defined for Re(z) > 1/2, thereby precluding zeros in that region.

Symmetry and the Functional Equation

The final stage involves the functional equation of the zeta function, which relates ζ(s) to ζ(1-s). This symmetry implies that if ζ(s) has a zero at s = σ + it, it must also have a zero at 1-s. If 1/ζ(z) is analytic for Re(z) > 1/2, then ζ(z) cannot have zeros in that region. By symmetry, it cannot have zeros for Re(z) < 1/2 either, forcing all non-trivial zeros onto the critical line Re(z) = 1/2.

Novel Research Pathways

Refined Oscillation Analysis: Investigate bounds of the form |M(x)| < sqrt(x) * (log log x)^a. This recognizes that the original conjecture may be too strong while maintaining the connection to the Riemann Hypothesis through modified analytic continuation.
Sieve Theory Connections: Connect Mertens function bounds to modern developments in sieve theory, such as the large sieve inequality, to estimate the distribution of square-free integers with specific sign patterns.
Spectral Analysis of the Mertens Operator: Construct a Hilbert space operator where the Mertens function acts as a kernel, linking the analytic continuation of 1/ζ(z) to the resolvent of this operator.

Computational Implementation

(* Section: Mertens and Zeta Convergence Analysis *)
(* Purpose: Visualizing the growth of M(x) and its relation to the critical strip *)

Module[{nMax = 2000, muList, mList, ratioList, zeros, p1, p2, s, approx, actual},
  (* Calculate Mobius and Mertens values *)
  muList = Table[MoebiusMu[n], {n, 1, nMax}];
  mList = Accumulate[muList];
  
  (* Normalized ratio M(n)/sqrt(n) *)
  ratioList = Table[{n, mList[[n]]/Sqrt[n]}, {n, 1, nMax}];
  
  (* Plot 1: Mertens Function vs Sqrt Bounds *)
  p1 = Show[
    ListLinePlot[mList, PlotStyle -> Blue, PlotLabel -> "Mertens Function M(n)"],
    Plot[{Sqrt[x], -Sqrt[x]}, {x, 1, nMax}, PlotStyle -> {Red, Dashed}]
  ];
  
  (* Plot 2: Nontrivial Zeros on the Critical Line *)
  zeros = Table[ZetaZero[k], {k, 1, 15}];
  p2 = ComplexListPlot[zeros, 
    GridLines -> {{1/2}, {}}, 
    PlotStyle -> Red, 
    PlotLabel -> "First 15 Zeta Zeros (Re = 1/2)"
  ];
  
  (* Numerical Verification of 1/Zeta(s) via Integral *)
  s = 0.8 + 10*I;
  approx = s * NSum[mList[[n]] * (n^(-s-1) - (n+1)^(-s-1))/s, {n, 1, nMax - 1}];
  actual = 1/Zeta[s];
  
  Print["Test at s = ", s];
  Print["Integral Approx: ", approx];
  Print["Actual 1/Zeta: ", actual];
  
  GraphicsColumn[{p1, p2}]
]

Conclusions

The analysis of arXiv:hal-01667383v4 underscores the profound relationship between the summatory behavior of the Möbius function and the location of the zeros of the Riemann zeta function. By focusing on the integral representation of 1/ζ(z), the paper provides a framework where the Riemann Hypothesis is treated as a problem of analytic continuation. The most promising avenue for further research lies in the refinement of Mertens bounds through probabilistic methods or the application of this integral technique to wider classes of L-functions.

References

arXiv:hal-01667383v4: M. Sghiar, "The Mertens function and the proof of the Riemann's hypothesis"
Odlyzko, A. M. and te Riele, H. J. J. (1985). "Disproof of the Mertens conjecture." Journal für die reine und angewandte Mathematik.
Titchmarsh, E. C. (1986). "The Theory of the Riemann Zeta-Function." Oxford University Press.