Bessel Integrals, Periods and Zeta Numbers
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Mathematical Physics. The day in mathematical physics: rigorous models of physical systems.
Kai: We're looking at the paper: "Bessel Integrals, Periods and Zeta Numbers".
Mira: Here's the gist of the paper. we present a summary of recent and older results on Bessel integrals and their relation with zeta numbers The continuous fractions for ζ(3), ζ(2) and ψ1(1/3)−ψ1(2/3) are presented, some of them related to (1), others sustained by numerical PSLQ evidence<ref:1209.1036#pg5>.
Kai: Review of Quantum Mechanics and Bessel Integrals The paper reviews results from
5: concerning quantum mechanics and Bessel integrals<ref:1209.1036#pg3>.
Mira: The random magnetic impurity model describes a quantum particle in a plane coupled to a random distribution of Aharonov-Bohm fluxes perpendicular to the plane<ref:1209.1036#pg3>.
Kai: Perturbative expansions of the partition function lead to multiple integrals on products of modified Bessel functions, such as Iρ2α4 = Z ∞ 0 du K0(u) 2 (uK1(u))2<ref:1209.1036#pg3>.
Title and authors: Kai: So, we're looking at this paper today, "Bessel Integrals, Periods and Zeta Numbers." It touches on a lot of ground there. It starts by bringing together results from older work on Bessel integrals and their connections to zeta numbers <ref:1209.1036#pg1>.
Mira: Yeah, it sets up this framework where the continuous fractions for zeta numbers like zeta three, zeta two, and that stuff involving psi one over three are presented, some of which are linked to the main integral form (one) <ref:1209.1036#pg1>. It's a very deep connection.
Lev: From what I see on page two, they start by looking at quantum mechanics and the random magnetic impurity model five <ref:1209.1036#pg1,the random magnetic impurity model>. They use perturbative expansions of the partition function to get multiple integrals involving modified Bessel functions, like those for I rho two alpha four <ref:1209.1036#pg2>.
Kai: And that's where things get interesting because they show how these integrals can be related back to simpler ones through direct integration and integration by parts <ref:1209.1036#pg4>. It’s a pathway to connecting these complicated forms to the zeta numbers we study.
Mira: They seem to infer a kind of counting rule from several results, like results three through eight <ref:1209.1036#pg4>, suggesting that an integration with an odd power of u reduces the Bessel weight by one <ref:1209.1036#pg4>.
Lev: That mapping they propose is pretty specific: they suggest that the integral with four as the weight, R zero over zero du K zero of u four, is like zeta five <ref:1209.1036#pg4>.
Kai: And then for the integral with three as a weight, I-Bessel function has no weight at all, so that is like zeta two <ref:1209.1036#pg4>.
Mira: They also suggest that the integral R over zero du u three K zero of u two K one of u two, and x K one of x I one of x K zero of x two, is like zeta three times zeta two <ref:1209.1036#pg4>. This whole idea fits this Bessel integral to zeta number concept <ref:1209.1036#pg5>.
Lev: Then they look at a family of integrals called I over kappa n j, defined by a specific integral form <ref:1209.1036#pg6>. Integration by parts there yields this recurrence relation involving matrices whose determinant depends on the parity of kappa <ref:1209.1036#pg6>.
Title and authors: Kai: That leads to sub-families based on the parity of n minus j, which means that by inverting that relationship, all integrals in a specific sub-family become linear combinations with rational coefficients of initial conditions related to numbers like zeta five <ref:1209.1036#pg6>.
Mira: And they claim this process determines a basis for these integrals, which leads to irrationality claims for certain values of kappa <ref:1209.1036#pg7>.
Lev: They examine the irrationality of those basis elements, specifically the ones with even and odd weights respectively <ref:1209.1036#pg7>. For weight kappa equal to four, they look at a continuous fraction for zeta k minus one <ref:1209.1036#pg10>.
Kai: That iteration is related to Apéry’s proof for the irrationality of zeta three <ref:1209.1036#pg10>. For weight kappa equal to three, they get a continuous fraction related to psi one over three minus psi one over two thirds <ref:1209.1036#pg15>.
Mira: But here's the caveat: the numerical convergence for these approximations doesn't seem fast enough to actually prove the irrationality of those corresponding Bessel integrals <ref:1209.1036#pg7>.
Lev: They also show that integrals like R over zero du u K zero of u n and R over zero du u three K zero of u n are periods <ref:1209.1036#pg19>. Integration of these further leads to expressions with logarithms, such as the integral over four leading to seven over eight zeta three <ref:1209.1036#pg26>.
Kai: And that integral over I zero of u K zero of u three is like psi one over three minus psi one over two thirds <ref:1209.1036#pg26>, which suggests a deeper meaning for Bessel integrals as periods <ref:1209.1036#pg27>.
Mira: The paper concludes by suggesting that if the irrationality claim on those basis elements, specifically those with even and odd weights, could be proven, it would provide a proof of the irrationality of zeta five <ref:1209.1036#pg31>.
Lev: They mention that the systematic search for Apéry-like numbers associated with these continuous fractions is still ongoing <ref:1209.1036#pg31>.
Kai: And they suggest that because Bessel integrals fall into the category of periods, it might point to a deeper meaning we haven't fully grasped yet <ref:1209.1036#pg22>.
Title and authors: Mira: Before we wrap up, let me add that they suggest that for any two odd positive integers n and m, the sum of those specific modified zeta numbers can be rewritten as a linear combination involving one, the integral over zero du K zero of u six du, the integral over zero du u three K zero of u six du, and zeta three and zeta five <ref:1209.1036#pg31>.
Lev: Finding those coefficients in that linear combination could potentially be done by generalizing integration by parts procedures from earlier work on multiple integrals to double nested integrals <ref:1209.1036#pg31>.
Kai: That sounds like a massive undertaking, trying to find these connections across different weight classes, like those involving weights six and eight.
Mira: And the numerical PSLQ evidence they found for something like I rho two alpha six being equal to one three zero over zero du u K zero of u six du plus one twenty over zero du u three K zero of u six du minus two five minus one sixteen zeroth zeta five <ref:1209.1036#pg3> is quite concrete.
Lev: That numerical identity involving zeta five is a key piece of evidence they use to support the mapping between Bessel integrals and zeta numbers <ref:1209.1036#pg5>.
Kai: So, to sum up, this paper, "Bessel Integrals, Periods and Zeta Numbers," connects these complex integrals to known constants through a counting rule derived from their structure.
Mira: It shows that these integrals behave like periods and suggests that proving the irrationality of some of these specific Bessel integral basis elements could lead to a proof for the irrationality of zeta five <ref:1209.1036#pg31>.
Lev: And they point out that finding the coefficients in those linear combinations between sums of double nested integrals and combinations of zeta numbers like zeta three and zeta five is a path forward <ref:1209.1036#pg31>.
Kai: It seems like the next step is to use these methods to systematically search for more Apéry-like numbers in this space.
Mira: That's the direction they are heading, looking for those specific structures that link different integral weights together.
Lev: We have a lot of work left on how this would translate into something runnable on actual hardware, but the theoretical framework presented here is definitely getting more detailed with respect to these deep mathematical objects.
The paper's summary: Kai: So, we're looking at this paper today about Bessel integrals and their link to zeta numbers. It basically summarizes some recent work connecting these complex integrals to these famous constants like zeta three and zeta five.
Mira: Right, it sets up this idea where they show how you can derive continuous fractions for those numbers using the structure of these Bessel integrals, especially for weights three and four.
Lev: So they're suggesting a kind of counting rule, a way to map the integral's weight onto a specific zeta number. That mapping fits into this bigger concept of linking Bessel integrals directly to periods.
Kai: They are basically saying that if you look at how these integrals behave, you can find patterns—like an odd power reducing the weight by one—and those patterns suggest which zeta number it relates to.
Mira: The real punch here is when they move from just mapping them to actually finding these integral periods. They show that some of these Bessel integrals *are* periods, which is a big deal because it gives the integrals a more fundamental physical meaning.
Lev: From an engineering standpoint, if those integrals are periods, it means we might be able to use them as stable quantities in some physical system, maybe even in quantum mechanics where we usually deal with those partition functions.
Kai: They also touch on irrationality claims for specific versions of these integrals, like weight four and weight three. It's a bit tricky because the numerical tests they run don't quite prove the irrationality yet.
Mira: True, the convergence isn't fast enough to definitively prove it right now. But they do give us some continuous fractions that look really close to what we expect for numbers like zeta three and psi one over three minus psi one over two thirds.
Lev: That’s where it gets interesting because those approximations are already getting pretty good, which is exactly the kind of evidence you need for a theoretical claim.
Kai: So, the implication is that if they can prove those irrationality claims on the basis elements—those fundamental building blocks—it could actually give us a proof for the irrationality of zeta five.
Mira: That would be huge because it connects this whole world of special functions and integrals back to one of the most famous numbers in mathematics in a very direct way.
Lev: It's also interesting because they hint at finding coefficients for linear combinations involving sums of double nested integrals that link things like zeta three and zeta five. That's a promising direction for finding new PSLQ identities.
Kai: So, to wrap up this summary, they’re suggesting that these Bessel integrals aren't just random mathematical curiosities; they are periods with deep connections to the world of zeta numbers.
Mira: And if we can nail down those proofs about irrationality and find those linking coefficients, it opens up a new avenue for connecting different types of integrals together.
Lev: That's a lot to unpack, but it’s definitely pointing toward some very specific structures in the math that could have physical relevance.
The paper's improvements: Kai: So, we're looking at what the authors suggest as improvements to their original paper on Bessel integrals and zeta numbers <ref:1209.1036#pg1>. They are basically saying they need a better way to prove those irrationality claims for the integral basis elements.
Mira: Right, they point out that relying only on numerical convergence isn't enough for those irrationality proofs involving weight four and weight three. They need something stronger than just seeing the numbers get close.
Lev: From my side, I think what they’re really suggesting is a systematic way to find these rational approximations using continuous fractions, which is a solid theoretical tool. It moves it from just "looks like" to actually being tested against known irrationality proofs.
Kai: So they want a method where you can take those specific integral forms and use the continuous fractions they derived to perform actual checks for irrationality on things like the integral with weight zero over zero du u four.
Mira: They are also looking at how to generalize these constructions. They’re thinking about taking the success they had with weights four and three and seeing if you can extend that logic to higher orders, like weight six or even eight.
Lev: That extension is key for me because those higher orders are exactly what we need to model real quantum error-correction protocols. If you want a stable algorithm, you need the math to scale up properly, and this generalization seems to be the path there.
Kai: So they’re trying to see if these patterns hold when the weight increases beyond what they've tested so far, which would give us more of those structural relationships with zeta numbers.
Mira: Exactly. The goal is to find a way to use the mapping between Bessel integrals and periods we found earlier to build these larger, more powerful algebraic constructions.
Lev: If they can succeed in generalizing the integration by parts procedures across these different weight classes, it means we could systematically hunt for new PSLQ identities that link sums of these nested integrals directly to combinations of zeta three and zeta five.
Kai: That would be incredibly useful because finding those exact coefficients is what we need to confirm the mapping they are proposing between the integrals and the constants.
Mira: It's a path forward, but it still relies on being able to handle these double nested integrals without getting lost in complexity, which is where you need that systematic approach.
Lev: So, the paper isn't just about one result; it’s laying out a whole strategy for how we can use this integral structure to systematically probe the landscape of irrationality and constant relationships.
Conclusion: Kai: So we're wrapping up on "Bessel Integrals, Periods and Zeta Numbers." It boils down to these complex integrals being linked to fundamental constants through a counting rule derived from their structure.
Mira: Right, it shows that these integrals aren't just random functions; they act like periods, which gives them a deeper physical interpretation in condensed matter or quantum systems.
Lev: I think the most important part for me is that they're showing how to systematically search for new links between different integral weights and zeta numbers through double nested integration procedures.
Kai: So if we can find those coefficients in the linear combinations involving zeta three and zeta five, it gives us a concrete way to understand how these different mathematical objects interact.
Mira: It really suggests that the structure of Bessel integrals might be the key to finding new PSLQ identities, which are essentially ways to prove irrationality for certain numbers.
Lev: That systematic search for those identities is what I’m most interested in, because those kinds of relations are exactly what you need when you're trying to build robust error correction codes that run on actual hardware.
Kai: It points toward a future where we can use these integral structures as a way to test the limits of mathematical proofs for things like zeta five.
Mira: And it opens up the door for applying this same idea to other areas where we have these types of integrals appearing, like in field theory or statistical mechanics.
Lev: It’s definitely a solid piece of theoretical groundwork, but there's still a lot of work needed to translate these abstract integral relationships into something that runs reliably on real quantum hardware.
LPTMS · CNRS UMR 8626 · University Paris-Sud
math-ph, math.MP
Submitted: 2012-09-05
Updated: 2026-11-16
Comments: 30 pages, no figure
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Importance score: 75/100
The gist: we present a summary of recent and older results on Bessel integrals and their relation with zeta numbers The continuous fractions for ζ(3), ζ(2) and ψ1(1/3)−ψ1(2/3) are presented…
Key concepts
- Bessel Integrals
- These are specific types of mathematical integrals involving modified Bessel functions. The paper investigates how different forms of these integrals, defined by parameters like weight ($\kappa$), can be related to known mathematical constants through integration and recurrence relations.
- Zeta Numbers ($\zeta(s)$)
- Zeta numbers are values of the Riemann zeta function at positive integers (like $\zeta(2)$ or $\zeta(3)$). The paper explores how certain Bessel integrals can be shown to be related to these fundamental constants, suggesting a deep connection between these special functions and number theory.
- Continuous Fractions
- A continuous fraction is an expression representing a number as an infinite sequence of integers separated by horizontal bars. The paper uses these fractions derived from the analysis of Bessel integrals to investigate whether certain related numbers are irrational, similar to how Apéry's proof works for $\zeta(3).
- Periods
- In mathematics, a period often refers to a specific type of integral or function that exhibits periodic behavior. The authors suggest that the Bessel integrals under study fall into this category, implying they possess a deeper meaning beyond just being complex integrals.
Terminology
Summary
we present a summary of recent and older results on Bessel integrals and their relation with zeta numbers The continuous fractions for ζ(3), ζ(2) and ψ1(1/3)−ψ1(2/3) are presented, some of them related to (1), others sustained by numerical PSLQ evidence<ref:1209. p. 01036#pg5>.
Review of Quantum Mechanics and Bessel Integrals
The paper reviews results from [5] concerning quantum mechanics and Bessel integrals<ref:1209.1036#pg3>. The random magnetic impurity model describes a quantum particle in a plane coupled to a random distribution of Aharonov-Bohm fluxes perpendicular to the plane<ref:1209. p. 01036#pg3>. Perturbative expansions of the partition function lead to multiple integrals on products of modified Bessel functions, such as Iρ2α4 = Z ∞ 0 du K0(u) 2 (uK1(u))2<ref:1209. p. 11036#pg3>. These integrals are related to simple integrals and zeta numbers through direct integration and integration by parts<ref:1209. p. 21036#pg4>.
Bessel Integral to Zeta Number Mapping
The authors infer a counting rule
from the results (3, 4, 5, 6, 7, 8) that an integration R ∞ 0 du un with n odd diminishes the Bessel weight by one<ref:1209.1036#pg4>. This suggests a mapping where R ∞ 0 du K0(u)4 is like ζ(5), R ∞ 0 du uI0(u)K0(u)3 is like ζ(2), and R u 0 dx xK1(x)I1(x)K0(x)2 is like ζ(3)ζ(2)<ref:1209. p. 21036#pg4>. This mapping fits the Bessel integral → zeta number
concept<ref:1209.1036#pg5>.
Recurrence and Basis Construction
The paper examines a family of Bessel integrals I(κ)n,j defined by a specific integral form<ref:1209.1036#pg6>. Integration by parts yields a recurrence relation (13) involving matrices whose determinant depends on the parity of κ<ref:1209.1036#pg6>. This leads to sub-families based on the parity of n - j<ref:1209. p. 41036#pg6>. By inverting (16), all integrals in sub-family (17) are linear combinations with rational coefficients of initial conditions, which are related to numbers like ζ(5)<ref:1209. p. 41036#pg6>. The basis for these integrals is determined by the initial conditions, leading to irrationality claims for certain values of κ<ref:1209. p. 91036#pg7>.
Irrationality Claims and Continuous Fractions
The paper investigates the irrationality of the basis elements (28) and (29) for even and odd weights respectively<ref:1209.1036#pg7>. For weight κ = 4, the analysis leads to a continuous fraction z(k − 1) = −16k/6 squared + 9k + 15k 2/10k cubed + z(k)<ref:1209. p. 81036#pg10>. This iteration is related to Apéry’s proof of ζ(3) irrationality<ref:1209. p. 81036#pg10>. For weight κ = 3, the analysis leads to a continuous fraction related to ψ1(1/3) − ψ1(2/3)<ref:1209. p. 01036#pg15>. The numerical convergence for these approximations does not seem sufficiently fast to prove the irrationality of the corresponding Bessel integrals. p. 9
Integration and Periods
The authors show that R ∞ 0 du uK0(u)n and R ∞ 0 du u3K0(u)n are periods<ref:1209.1036#pg19>. The integration of these integrals further leads to expressions involving logarithms, such as Z ∞ 0 du uK0(u)4 = 7/8 ζ(3)<ref:1209. p. 11036#pg26> and Z ∞ 0 du uI0(u)K0(u)3 = ψ1(1/3) − ψ1(2/3)<ref:1209. p. 01036#pg26>. This suggests a deeper meaning for Bessel integrals as periods<ref:1209. p. 01036#pg27>.
Conclusion
The paper concludes by suggesting that if the irrationality claim on (28) and (29) could be proven, it would provide a proof of the irrationality of ζ(5)<ref:1209.1036#pg31>. The systematic search for Apéry-like numbers associated with these continuous fractions is ongoing. p. 0 The fact that Bessel integrals fall in the category of periods might also be an indication of a deeper meaning yet to be understood<ref:1209. p. 0 p. 01036#pg22>.
References
[1] J. Desbois, C. p. 0 Furtlehner and S. Ouvry, Random Magnetic Impurities and the Landau Problem, Nuclear Physics B[FS] 453 (1995) 759. p. 0
[2] C. Furtlehner and S. Ouvry, Integrals involving four Macdonald functions and their relation to 7ζ(3)/2, arXiv:math-ph/0306004. p. 0
[3] C. Furtlehner and S. Ouvry, private communication. p. 0<ref:1209.1036#pg3>
[4] S. Ouvry, Random Aharonov-Bohm vortices and some exact families of integrals, JSTAT: Theory and Experiment (2005) P09004. p. 0
[5] S. Mashkevich and S. Ouvry, Random Aharonov-Bohm vortices and some exact families of integrals: Part II, JSTAT: Theory and Experiment (2008) P03018. p. 0<ref:1209.1036#pg3>
[6] http://mathworld.wolfram.com/PSLQAlgorithm.html, and references therein.
[7] as far as Apéry’s work is concerned see R. p. 6 Apéry, Irrationalité de ζ(2) et ζ(3), Astérisque, 61:11-13, 1979 and the review A. p. 0 van der Poorten, A proof that Euler missed…Apéry’s proof of the irrationality of ζ(3), An informal Report
Math. Intelligencer, 1:195-203, 1979.
[8] see for example M. p. 3 Konsevitch and D. Zagier, Periods (2001) IHES/M/01/22. p. 0
[9] D. Zagier, Integrals solutions of Apéry-like recurrence equations, in: Groups and Symmetries: From Neolithic Scots to John McKay, CRM Proc. Lecture Notes 47 (Amer. p. 9 Math. Soc., Providence, RI, 2009), 349-366.
[10] (70) and (72) are listed in G. Almkvist and W. Zudilin, Differential equations, mirror maps and zeta values. p. 0 In Mirror symmetry. V, AMS/IP Stud. Adv. Math. 38, Amer. Math. Soc., Providence, RI (22006), 481-515.
[11] J. M. Borwein and B. Salvy, A proof of recursion for Bessel moments (2007). p. 8
[12] W. Zudilin, A third order Apéry-like recursion for ζ(5), arXiv:math/0206178. p. 0
Appendix A
-
Trying to integrate R ∞ 0 du uI0(u)K0(u)n using again the integral representation (53) and (u = 2√t), one can obtain an expression again in terms of un−2, vn−2 and wn−2 but somehow more involved than (67).
-
More on integrating R ∞ 0 du uK0(u)n, R ∞ 0 du u3K0(u)n and R ∞ 0 du uI0(u)K0(u)n n = 3 Z ∞ 0 du uK0(u)3 =?<ref:1209.1036#pg23>.
Appendix B
From the iteration (16) it is easy to get for κ ≥ 5 higher order recursions generalizing the recursions (42, 51) for κ = 4, 3.<ref:1209.1036#pg27>.
One can for example start from the inverse iteration x(k + 1)y(k + 1)z(k + 1) = α(k)β(k)γ(k)δ(k)ǫ(k)ζ(k)<ref:1209. p. 81036#pg28>.
Appendix C
Coming back to (2,9), use Z ∞ 0 g(u)du Z ∞ u f(x)dx = Z ∞ 0 f(u)du Z u 0 g(x)dx to rewrite Iρ2α6 as.
The authors infer that for any two odd positive integers n, m one should have that ˜ζ(fn, gm)+˜ζ(fm, gn) can be rewritten as a linear combination of 1, R ∞ 0 du K0(u)6 du, R ∞ 0 du u3K0(u)6 du and ζ(3) and ζ(5)<ref:1209.1036#pg31>. Finding the coefficients of the linear combination can be in principle obtained by generalizing integration by part procedures in [1, 2, 4, 5] to double nested integrals<ref:1209. p. 11036#pg31>. The fact that Bessel integrals fall in the category of periods might also be an indication of a deeper meaning yet to be understood<ref:1209.1036#pg22>.
From the iteration (16) it is easy to get for κ ≥ 5 higher order recursions generalizing the recursions (42, 51) for κ = 4, 3.<ref:1209.1036#pg27>.
One can for example start from the inverse iteration x(k + 1)y(k + 1)z(k + 1) = α(k)β(k)γ(k)δ(k)ǫ(k)<ref:1209. p. 81036#pg28>.
From the iteration (16) it is easy to get for κ ≥ 5 higher order recursions generalizing the recursions (42, 51) for κ = 4, 3.<ref:1209.1036#pg27>.
One can for example start from the inverse iteration x(k + 1)y(k + 1)z(k + 1) = α(k)β(k)γ(k)δ(k)<ref:1209. p. 81036#pg28>.
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
System can perform irrationality checks for specific Bessel integrals by finding rational approximations derived from continuous fractions, as shown in Section 3.3 and Section 41-45, allowing it to numerically test claims about numbers like I(4)0,0 or I(3)0,0.
-
System can derive linear combinations of double nested integrals that are related to zeta numbers by generalizing integration by part procedures across multiple weight classes (like those involving weights 6 and 8), potentially leading to proofs for the irrationality of zeta numbers like ζ(5) via PSLQ identities (e.g., identity 69).
-
The system can solve and analyze high-order linear recurrence relations associated with Bessel integrals, such as recursion (95), to determine asymptotic behaviors and derive continuous fractions, which is useful for identifying the irrationality of related quantities like ψ1(1/3)−ψ1(2/3) in the weight 3 case.
-
The system can utilize the mapping between Bessel integrals and periods to generalize algebraic constructions from lower weights (like κ=4 or κ=3) to higher orders (κ >= 5), potentially leading to a systematic search for new PSLQ identities that link sums of double nested integrals to combinations of zeta numbers like ζ(3) and ζ(5).
Sources
- Integrals involving four Macdonald functions and their relation to 7zeta(3)/2
- A third-order Apery-like recursion for $\zeta(5)$
Related papers
- Existence of a fluid phase for the two-dimensional one-component plasma
- The Huang-Yang conjecture for the low-density Fermi gas
- Anderson localisation in infinite and finite resonator crystals with long-range interactions
- Free energy and spectral edge of the SYK model
- Anderson localization for 1-d quasi-periodic Schr"odinger operators with degenerate weights
- Irregular SLE(4) martingales and isomonodromic deformations