Download Full Article
This article is available as a downloadable PDF with complete code listings and syntax highlighting.
Introduction
The Riemann Hypothesis remains a central pillar of analytic number theory, asserting that the non-trivial zeros of the Riemann zeta function lie precisely on the critical line where the real part is 1/2. The source paper arXiv:hal-01845847 explores this conjecture through the lens of the Mertens function, denoted as M(x), which is the summatory function of the Mobius function. By analyzing the integral representation of the reciprocal of the zeta function, the paper establishes a framework of inequalities that investigate the supremum of the real parts of the zeros, often denoted as beta*.
This investigation is significant because it shifts the focus from the zeros themselves to the growth properties of the integral tail of the Mertens function. The core contribution of this analysis is the identification of a mathematical tension between the upper bounds derived from assumed zero distributions and the lower bounds dictated by the density of square-free integers. This article provides a technical breakdown of these arguments and proposes computational pathways for further verification.
Mathematical Background
To analyze the findings in arXiv:hal-01845847, we define the primary arithmetic objects. The Mobius function mu(n) is the parity of the number of prime factors for square-free integers, and zero otherwise. The Mertens function is defined as the sum of mu(n) for all n up to x. A fundamental identity connects these to the zeta function:
1/zeta(s) = s * integral from 1 to infinity of M(u) / u^(s+1) du
The paper focuses on the tail of this integral, defined for a truncation point x as the integral from x to infinity of M(u) / u^(s+1) du. The research utilizes the parameter beta*, representing the supremum of the real parts of non-trivial zeros. Under the Riemann Hypothesis, beta* would equal 1/2. The analysis assumes a growth rate for the Mertens function where |M(u)| is bounded by m * u^beta for some constant m and any beta greater than beta*.
Main Technical Analysis
Upper and Lower Bound Divergence
The technical core of the paper involves comparing two distinct bounds for the magnitude of the integral tail. The first is an upper bound derived from the assumed growth of the Mertens function. For a complex variable s = sigma + it, the magnitude of the integral tail is constrained by the term (sigma * m * x^(beta - sigma)) / ((sigma - beta) * sqrt(sigma^2 + t^2)).
Conversely, the lower bound is established by examining the density of square-free integers, which is known to be 6/pi^2. The paper derives a lower estimate of the form ((1 - epsilon) / sqrt(sigma^2 + t^2)) * [6 * x^(1 - sigma) / (pi^2 * (sigma - 1)) + O(x^(1/2 - sigma))] - (m * x^(beta - sigma) / sqrt(sigma^2 + t^2)).
The Ratio Contradiction
The refutation argument hinges on the ratio of these two bounds as x tends toward infinity. If the Riemann Hypothesis holds and beta* is less than 1, the ratio of the lower bound to the upper bound is dominated by a factor of x^(1 - beta). As x increases, this factor grows without limit. This implies that for any beta less than 1, the lower bound will eventually exceed the upper bound, which is a logical impossibility. Therefore, the paper concludes that beta* must equal 1, which would mean the Riemann Hypothesis is false.
Novel Research Pathways
- Characterization of T-sets: Investigating the density and measure of the sets of imaginary values t where the lower bounds are maximized. Understanding if these sets become increasingly sparse as x grows is critical to validating the contradiction.
- Smoothing Kernels: Applying smoothing functions to the Mertens integral to determine if the contradiction persists when the sharp cutoff at x is replaced by a weighted decay. This would test if the result is a fundamental property of Mobius distribution or an artifact of the integral truncation.
- Resonance Methods: Utilizing resonance techniques to find specific frequencies t that correlate with the signs of mu(n), thereby providing a more rigorous basis for the lower bound estimates without relying on triangle inequality saturation.
Computational Implementation
The following Wolfram Language code allows for the numerical exploration of the integral tail and the growth of the ratio between the theoretical lower and upper bounds.
(* Section: Mertens Integral Tail and Ratio Analysis *)
(* Purpose: Compare numerical integral values against paper bounds *)
ClearAll[mertensTail, lowBound, upBound, sigma, beta, xVal];
(* Parameters based on arXiv:hal-01845847 *)
sigma = 1.2;
beta = 0.5; (* RH Assumption *)
m = 1.0;
epsilon = 0.05;
(* Mertens function via Mobius sum *)
mFunc[u_] := Mertens[Floor[u]];
(* Numerical tail integral at t=0 *)
mertensTail[x_?NumericQ, s_?NumericQ] :=
NIntegrate[mFunc[u] / u^(s + 1), {u, x, Infinity},
Method -> "GlobalAdaptive", PrecisionGoal -> 3];
(* Lower bound leading term *)
lowBound[x_, s_] := ((1 - epsilon)/s) * ((6 * x^(1 - s))/(Pi^2 * (s - 1)));
(* Upper bound leading term *)
upBound[x_, s_, b_] := (s * m * x^(b - s))/(s - b);
(* Generate Comparison Data *)
xValues = Table[x, {x, 50, 500, 50}];
ratios = Table[{x, lowBound[x, sigma] / upBound[x, sigma, beta]}, {x, xValues}];
(* Plotting the growth of the ratio R(x) *)
ListLinePlot[ratios,
PlotLabel -> "Growth of Lower/Upper Bound Ratio (beta=0.5)",
AxesLabel -> {"x", "Ratio R(x)"},
PlotStyle -> Orange,
Mesh -> All]
Conclusions
The analysis provided by arXiv:hal-01845847 offers a provocative challenge to the current understanding of the zeta function's zeros. By demonstrating an asymptotic incompatibility between the square-root growth permitted by the Riemann Hypothesis and the density of square-free integers, the work suggests that the Mertens function must exhibit larger oscillations than previously assumed. The most promising next step involves a rigorous measure-theoretic study of the sets of frequencies where these lower bounds are realized, as this remains the most delicate point in the refutation logic.
References
- arXiv:hal-01845847 - Primary source on Mertens integral bounds.
- arXiv:hal-01728058 - Related research on Mobius summatory functions.
- Titchmarsh, E. C. (1986). The Theory of the Riemann Zeta-Function. Oxford University Press.