Radiative decays of the h c meson from lattice QCD

arXiv:2607.25707 · hep-lat · Submitted 2026-07-28 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: High Energy Physics: Lattice. The day in lattice field theory: what the simulations put on the table.

Kai: I'm Kai.

Mira: And I'm Mira, here with Lev.

Kai: We're looking at the paper: "Radiative decays of the h c meson from lattice QCD".

Mira: Here's the gist of the paper. Radiative decays of the hc meson from lattice QCD explore transitions to γη and γη′ states to investigate discrepancies with experimental results <ref:2607.25707#pg2>.

Kai: The work explores, for the first time in lattice QCD, the radiative decays to γη and γη′ of the J P C = 1+− charmonium state, hc <ref:2607.25707#pg2>.

Mira: The study builds upon previous calculations of J/ψ decays to these final states by incorporating novel technology to access hc as an excited state and separate independent electric dipole and longitudinal form–factors <ref:2607.25707#pg2>.

Kai: The specific decays to the lightest isoscalar pseudoscalars, hc → γη and hc → γη′, have been measured with good statistical precision at BES III <ref:2607.25707#pg2>.

Title and authors: Kai: So summarizing this paper, they are doing a world first lattice QCD calculation of multipole transition form factors for hc to gamma eta and gamma eta prime evaluated on the same m pi equal three hundred ninety one MeV lattices used in references three and four.

Mira: They are using optimized operators following from diagonalization of a matrix of two point correlation functions to access hc as an excited state and separate the electric dipole and longitudinal form factors.

Lev: And they introduce a procedure for simultaneous fitting of the time dependencies of multiple three point correlation functions to get a single value for J on page two of that work <ref:2607.25707#pg1>.

Kai: This allows them to extract values for the invariant form factors, E one Q two and C one Q two, which represent multipoles of electric dipole and longitudinal nature respectively.

Mira: The total width for the Dalitz decays follows from integrating over Q two from two m L squared to m hc minus m eta prime squared on page two of that work <ref:2607.25707#pg1>.

Lev: So they are setting up the calculation to determine the Q two dependence of these form factors, which is necessary for that integration.

Kai: The procedure leads to a set of values of E one Q two and C one Q two discretely sampled in Q two for both hc to gamma eta and hc to gamma eta prime on page two of that work <ref:2607.25707#pg1>.

Mira: And the radiative decay rates for hc to gamma eta and hc to gamma eta prime are controlled by E one zero, as presented in equation two on page two of that work <ref:2607.25707#pg1>.

Lev: So E one zero is the value they are using to characterize these decay rates, which is a single number derived from the form factor values.

Kai: The best estimates for this zero Q two form factor are found to be ehc to c gamma eta prime zero is equal to zero point zero five five six with a three zero error in giga-electron volts on page two of that work <ref:2607.25707#pg1>.

Mira: That specific number, ehc to c gamma eta prime zero is equal to zero point oh five five six with a three zero error in giga-electron volts on page two of that work, is the central finding for the form factor at zero Q two <ref:2607.25707#pg2>.

Lev: It's a concrete number we can compare against experimental data, but I’m wondering how much systematic uncertainty is baked into those estimates before we even look at the Q two dependence.

Kai: They do provide those values discretely sampled in Q two for both hc to gamma eta and hc to gamma eta prime on page two of that work <ref:2607.25707#pg1>.

Mira: And they also have these values for the decay rates controlled by E one zero, as presented in equation two on page two of that work <ref:2607.25707#pg1>.

Lev: So we need to understand how those discrete points map onto a continuous dependence so we can actually talk about the Q two dependence properly.

Kai: The procedure for simultaneous fitting of time dependencies is what allows them to do that, which is what they are doing on page two of that work <ref:2607.25707#pg1>.

Mira: That technique is essentially how they handle the complexity of the three point correlation functions to get a single value for J on page two of that work <ref:2607.25707#pg1>.

Lev: It’s a way to control the noise or systematic effects from those different time slices by ensuring the fit parameters, like J and delta t, are common across all time steps.

Mira: And this is a bit of a simplification because it means they only have to deal with a single form factor for the psi prime decays, which makes the analysis simpler in that specific case.

The paper's summary: Kai: Now let’s look at the improvements they suggest for this work. They use optimized operator technology to access the first excited vector state, which is psi prime on page three of that work <ref:2607.25707#pg1>.

Mira: That means they simplify things by only having to deal with a single form factor for those psi prime decays, which makes the analysis simpler in that specific case.

Lev: That simplification is useful if you are trying to get a baseline result quickly, but it might hide complexity if the physical process actually involves interference between different states.

Kai: The paper also introduces the procedure for simultaneous fitting of time dependencies of multiple three point correlation functions to get a single value of J on page two of that work <ref:2607.25707#pg1>.

Lev: This seems like a useful technique for controlling the noise or systematic effects from those different time slices by ensuring the fit parameters, like J and delta t, are common across all time steps.

Kai: The paper also uses various parameterizations, including dipole and exponential forms as well as a conformal mapping into a variable z of Q two on page two of that work <ref:2607.25707#pg1>.

Mira: That allows for comprehensive characterization of the Q two range explored in the calculation, so they aren't stuck with just one mathematical shape.

Lev: A conformal mapping is powerful because it can sometimes map a complex region onto a simpler one, but we need to be careful that we don't introduce artifacts in our parameterization.

Kai: And they compare their real photon transition rates by comparing estimates from various parameterizations of the Q two dependence of E one Q two with values extracted from BESIII experimental decay widths on page two of that work <ref:2607.25707#pg1>.

Mira: That comparison is a way to cross-validation method to assess the accuracy of lattice QCD results against what we see experimentally at BESIII.

Lev: Cross-validation is always a good thing, but we have to remember that the experimental data itself has its own uncertainties, and they need to be comparable on both sides.

Kai: They conclude that relevant signals with a high degree of statistical precision can be obtained by use of contemporary techniques on page one of that work <ref:2607.25707#pg1>.

Mira: So they are saying these new techniques give us a way to get results with good statistical precision using contemporary methods.

Lev: That’s the goal, but I wonder if the practical overhead for achieving that high precision is too much for current real hardware setups.

The paper's improvements: Kai: So to wrap up on "Radiative decays of the h c meson from lattice QCD," they find that their most likely origin of the observed discrepancy with respect to experiment common to both J psi decaying and hc decaying is some property of the eta and eta prime on this particular lattice on page one of that work <ref:2607.25707#pg1>.

Mira: That means the issue isn't in how we model charmonium decays generally, but perhaps in how those specific light mesons behave on this particular lattice setup.

Lev: So they are pointing toward something specific to the gauge field configurations used in this calculation rather than a general failure of the theory itself.

Lev: I think the next step they suggest, considering multi-hadron production like J psi decaying to gamma pi pi and coupled channel J psi decaying to gamma kappa kappa on page one of that work, would be a good way to test if this effect holds up <ref:2607.25707#pg1>.

Lev: And I think that's a solid plan because if we can't get a clean signal here, we need better tools for sampling different parts of the configuration space on other lattices.

Conclusion: Kai: So we’re wrapping up this look at "Radiative decays of the h c meson from lattice QCD." Basically, they used optimized operators to access that hc state as an excited state and they successfully separated the electric dipole and longitudinal form factors for those decays to gamma eta and gamma eta prime.

Mira: Right, so we’re talking about characterizing those decay rates using a covariant vector current matrix element decomposition, which gives them those invariant form factors, E one Q two and C one Q two.

Lev: And they used that simultaneous fitting of multiple three point correlation functions to get a single value for J. That’s the mechanism they used to control the noise from those different time slices.

Kai: They found some concrete numbers for these form factors at zero momentum, E one zero and C one zero, and those estimates are around zero point zero five five six in giga-electron volts.

Mira: Those are important because they’re the inputs you need to calculate the total width for the Dalitz decays by integrating over that Q two range.

Lev: But we have to remember they also found that across the entire Q two range, their signal is broadly compatible with zero, which suggests a certain level of consistency.

Kai: The conclusion is that whatever discrepancy we see with experimental results from BESIII for both J psi decaying and hc decaying might actually be due to some property specific to the eta and eta prime mesons on this particular lattice.

Mira: They also compared their computed ratio, Ehc to c gamma eta prime zero over Ehc to c gamma eta zero, which is about two point nine with a three point nine error, against the J psi ratio they found on those same lattices.

Lev: The comparison between the ratios for J psi and hc shows approximate agreement when you remove the phase space differences between them. That’s a key piece of evidence.

Kai: They suggest that the production of eta prime is larger than eta production, which is what we would usually expect in this system based on standard understanding.

Mira: It points toward something specific about how those light mesons are behaving within the gauge field configurations they used for this calculation.

Lev: So even though they found a discrepancy, the immediate conclusion is that the setup itself might be different from what’s happening in nature.

Kai: It’s interesting because it shows that using contemporary techniques gives us a way to get results with high statistical precision on these kinds of charmonium transitions.

Mira: And while they have this result for the hc state, they acknowledge that the sensitivity of the eta prime system to detailed properties of gauge field configurations might not be shared by other light meson final states.

Lev: So the next move is looking at multi-hadron production like J psi decaying to gamma pi pi and coupled channel J psi decaying to gamma kappa kappa on page one of that work.

Kai: That’s the direction they are heading, looking at more complex final states to see if we can get a clearer picture.

Mira: It’s a logical progression since the current results give us a strong hint about where to look next in this field.

Adelaide University · College of William and Mary · Thomas Jefferson National Accelerator Facility

hep-lat

Submitted: 2026-07-28

Updated: 2026-07-28

DOI: 10.1103/3s33-c5n5

Code: https://github.com/JeffersonLab/mg_proto

License: http://creativecommons.org/licenses/by/4.0/

Importance score: 81/100

The gist: Radiative decays of the hc meson from lattice QCD explore transitions to γη and γη′ states to investigate discrepancies with experimental results <ref:2607.25707#pg2>.

Key concepts

Electric Dipole (E1) Form-Factor
This is a mathematical description used to characterize how the charmonium state decays into a photon and a light meson. It represents one type of transition, similar to an electric dipole moment in quantum mechanics, and its Q² dependence describes how this strength changes as the momentum transfer squared varies.
Longitudinal (C1) Form-Factor
This form factor describes the second type of transition component, often called longitudinal. It is crucial for fully describing the decay process by accounting for all possible polarization states of the decaying particle and providing a complete picture of the transition matrix element.
Three-Point Correlation Functions
These are specific mathematical tools used in lattice QCD to calculate decay processes involving three particles (like an initial state, a final photon, and a light meson). Simultaneous fitting of their time dependencies helps reliably extract the properties of the decaying particle, hc.
Q² Dependence
Q² is a variable representing the momentum transfer squared in the decay process. Determining how form factors like E1(Q²) change with Q² allows researchers to map out the full range of physical information contained within these transitions.

Terminology

Summary

Radiative decays of the hc meson from lattice QCD explore transitions to γη and γη′ states to investigate discrepancies with experimental results <ref:2607. p. 1

Where the paper says it

The use of optimized operator technology in charmonium also allows us access to the first–excited vector state, the ψ ′ , corresponding to the experimental ψ(3686) meson. We will compute transitions ψ ′ → γη(′) , presenting first lattice QCD results for this process, finding a signal that is statistically compatible with zero on the current lattices.

25707#pg2>.

Introduction and Motivation

The work explores, for the first time in lattice QCD, the radiative decays to γη and γη′ of the J P C = 1+− charmonium state, hc <ref:2607.25707#pg2>. The study builds upon previous calculations of J/ψ decays to these final states by incorporating novel technology to access hc as an excited state and separate independent electric dipole and longitudinal form–factors <ref:2607. p. 0

Where the paper says it

Edwards3, ‡ (for the Hadron Spectrum Collaboration) 1CSSM, Department of Physics, Adelaide University, Adelaide, 5005, South Australia, Australia 2Department of Physics, College of William and Mary, Williamsburg, VA 23187, USA 3Thomas Jefferson National Accelerator Facility, 12000 Jefferson Avenue, Newport News, VA 23606, USA (Dated: July 29, 2026) We explore, for the first time in lattice QCD, the radiative decays to γη and γη′ of the J P C = 1+− charmonium state, hc. This work expands upon a previous calculation of J/ψ decays to the same final states, incorporating novel technology to access the hc where it is an excited state and to separate the independent electric dipole and longitudinal form–factors.

25707#pg2>. The specific decays to the lightest isoscalar pseudoscalars, hc → γη and hc → γη′, have been measured with good statistical precision at BES III <ref:2607. p. 0
Where the paper says it

Radiative decays to light hadron systems in which the cc¯ pair annihilate contribute with branching fractions typically at the parts– per–thousand level [1]. The specific decays to the lightest isoscalar pseudoscalars, hc → γη and hc → γη′ , have been measured with good statistical precision at BES III [2].

25707#pg2>.

Accessing the hc and η′ at Various Momenta

To separate the electric dipole and longitudinal form–factors in the hc → γη(′) process, it proves useful to consider both the helicity 0 and helicity ±1 components of J P C = 1+− hc at many values of momentum <ref:2607.25707#pg2>. Optimized operators following from diagonalization of a matrix of two-point correlation functions allow for the reliable projection out of hc <ref:2607. p. 1

Where the paper says it

Optimized operators following from diagonalization of a matrix of two–point correlation functions computed with a basis of fermion bilinear operators, as described in Ref. [4], allow us to reliably project out the hc.

25707#pg2>. The helicity 0 components subduce into moving–frame irreps which also contain J P C = 0−− and 2−− as well as higher J <ref:2607. p. 1
Where the paper says it

Contributions of higher excited states appearing as curvature at early times in this plot appear to be somewhat larger than for the J/ψ, but still modest. The helicity 0 components of the hc subduce into moving–frame irreps which also contain J P C = 0−− and 2 −− as well as higher J.

25707#pg2>.

Extraction of Multipole Form-Factors from Three-Point Correlation Functions

The decays hc → γη(′) can be characterized by a covariant vector–current matrix–element decomposition <ref:2607.25707#pg2>. The invariant form–factors, E1(Q2) and C1(Q2), represent multipoles of electric dipole and ‘longitudinal’ nature respectively <ref:2607. p. 3

Where the paper says it

4 This decomposition implements conservation of the vector current, 0 = (p − p ′ )µ η (′) (p ′ ) j µ(0) hc(p, λ) , by construction. The invariant form–factors, E1(Q2 ) and C1(Q2 ), represent multipoles of electric dipole and ‘longitudinal’ nature respectively [10].

25707#pg2>. The total width for the Dalitz decay follows from the integration over q2 from (2ml)2 to (mhc − mη′)2, which requires determining the Q2 dependence of these form–factors <ref:2607. p. 3
Where the paper says it

In terms of q 2 ≡ −Q2 , the differential decay rate for this process can be expressed as, dΓ dq2 d cos θ ⋆ ℓ = 1 27 α 2 |q| m2 hc |p ⋆ ℓ | p q 2 1 q 2 1 + 4m2 ℓ q 2 |E1| 2 + q 2 |C1| 2 + 4|p ⋆ ℓ | 2 q 2 |E1| 2 − q 2 |C1| 2 cos2 θ ⋆ ℓ , (3) where |q| represents the magnitude of momentum of the η (′) in the rest–frame of the decaying hc, and (|p ⋆ ℓ |, θ⋆ ℓ ) represent the magnitude and direction of momentum of the ℓ + in the rest–frame of the ℓ +ℓ − system, relative to the axis defined by the momentum of the ℓ +ℓ − pair in the hc rest–frame4 . The total width for the Dalitz decay follows from the (straightforward) integration over θ ⋆ ℓ , and the integration over q 2 from (2mℓ) 2 to (mhc − mη(′) ) 2 , which requires us to have determined the Q2 dependence of the form–factors E1(Q2 ) and C1(Q2 ) over that range.

25707#pg2>.

Form-Factor Determination and Results

The procedure leads to a set of values of E1(Q2) and C1(Q2) discretely sampled in Q2 for both hc → γη and hc → γη′ <ref:2607.25707#pg2>. The radiative decay rates for hc → γη and hc → γη′ are controlled by E1(0), as presented in Eqn. p. 0

Where the paper says it

This work expands upon a previous calculation of J/ψ decays to the same final states, incorporating novel technology to access the hc where it is an excited state and to separate the independent electric dipole and longitudinal form–factors. Results for the radiative decay rates, computed on three-flavor lattices with mπ ∼ 391 MeV show a suppression relative to experiment that is similar to that observed in the prior calculation of J/ψ → γη(′) .

2 <ref:2607. p. 0
Where the paper says it

In practice this means that the hc often appears together with the J/ψ as an excited state such that we cannot simply rely upon ground–state dominance at large Euclidean times – we will use a further application of optimized operator technology to access it reliably. The arXiv:2607.

25707#pg2>. The best estimates for the form–factor at Q2 = 0 are found to be Ehc→c γη1 (0) = 0. p. 5
Where the paper says it

We presented fits using a dipole form and two exponential–based forms, as well as a rather flexible approach based upon a conformal mapping of Q2 into a variable, z, in terms of which a low–order polynomial description could be found. Evaluating these various fitted parameterizations at Q2 = 0 yielded a set of estimates of the real photon transition.

0195(19) GeV and Ehc→c γη′1 (0) = 0. p. 3
Where the paper says it

In order to obtain E1 and C1 within lattice QCD, we nonperturbatively evaluate the three–point correlation functions, C(t, ∆t) = 0 Ωη(′) (∆t) j(t) Ω† hc (0) 0 , for all current insertion times, t, between the fixed source– sink separation, ∆t. The use of optimized operators to interpolate the hc and η (′) leads us to expect dominance of those states in the time–dependence such that, C˜(t, ∆t) ≡ q 2Eη(′) 2Ehc e Eη(′) (∆t−t) e Ehc t C(t, ∆t) = ⟨η (′) |j(0)|hc⟩ + .

0556(30) GeV <ref:2607. p. 0
Where the paper says it

In practice this means that the hc often appears together with the J/ψ as an excited state such that we cannot simply rely upon ground–state dominance at large Euclidean times – we will use a further application of optimized operator technology to access it reliably. The arXiv:2607.

25707#pg2>.

Conclusions

The most likely origin of the observed discrepancy with respect to experiment common to both J/ψ → γη(′) and hc → γη(′) is some property of the η and η′ on this particular lattice <ref:2607.25707#pg2>. The computed η′/η ratio, Ehc→c γη'1 (0)/Ehc→c γη1 (0) = 2.9(3), can be compared to the corresponding ratio for J/ψ decays computed on the same lattices, F J/ψ→γη' (0)/F J/ψ→γη(0) = 3.3(3), which is compatible within errors <ref:2607. p. 1

Where the paper says it

We will compute transitions ψ ′ → γη(′) , presenting first lattice QCD results for this process, finding a signal that is statistically compatible with zero on the current lattices. We refer the reader to Ref.

25707#pg2>. The dimensionless ratio comparing the strength of the electric dipole process in hc decays to the magnetic dipole process in J/ψ decays, removing differing phase–space, shows approximate agreement for both η and η′ <ref:2607. p. 1
Where the paper says it

ACCESSING THE hc AND η (′) AT VARIOUS MOMENTA In order to separate the electric dipole and longitudinal form–factors in the hc → γη(′) process, it proves useful to consider both the helicity 0 and helicity ±1 components of the J P C = 1+− hc at many values of momentum. In–flight, the helicity ±1 components subduce into irreps which also contain J P C = 1−− (see Table I of Ref.

25707#pg2>. The analysis suggests that the production of η′ is larger than η production is of course expected in the usual understanding of this system <ref:2607. p. 0
Where the paper says it

The timelike Q 2 dependence of both form–factors together influence the Dalitz decays, hc → ℓ +ℓ −η (′) , which are as yet unmeasured experimentally. INTRODUCTION Radiative decays of the J/ψ and other heavy-quark vector mesons copiously produced in e +e − collisions serve as a factory for the production of light meson resonances in a process that is theoretically relatively clean.

25707#pg2>.

Pruning redundant reasoning tokens

The work utilizes optimized operator technology to access the first excited vector state, the ψ′ <ref:2607.25707#pg2>. The presence of only a single form-factor simplifies the analysis for ψ′ → γη and ψ′ → γη′ <ref:2607. p. 5

Where the paper says it

FORM–FACTORS The procedure described above leads to a set of values of E1(Q2 ) and C1(Q2 ) discretely sampled in Q2 for both hc → γη and hc → γη′ . [4] we described the use of explicit parameterizations to capture the Q2 dependence of the single form–factor for each of J/ψ → γη and J/ψ → γη′ .

25707#pg2>. The results extracted using a single source–sink separation, ∆t/at = 20, with ψ′ momenta in the range nψ'2 ≤ 4, show that across the entire Q2 range, the signal is broadly compatible with zero <ref:2607.25707#pg2>. The quality of fit appears to suggest that a polynomial in z of at least order 2 is required to describe the data over the wide Q2 range considered <ref:2607. p. 7
Where the paper says it

[4] and in Appendix D. The quality of fit appears to suggest that a polynomial in z of at least order 2 is required to describe the data over the wide Q2 range considered.

25707#pg2>. The presence of more points in the region around Q2 = 0 (due to the larger mass of η′ relative to η) leads to a smaller parameterization dependence on E1(0) <ref:2607. p. 8
Where the paper says it

f for fits which successfully capture the trend of the data reflect the larger degree of scatter in the data points. The presence of more points in the region around Q2 = 0 (due to the larger mass of the η ′ relative to η) leads to a smaller parameterization dependence on E1(0).

25707#pg2>. The fit parameter values for all presented fits are provided in Appendix D <ref:2607. p. 7
Where the paper says it

We observe from the fit curves that the paucity of discrete points around Q2 = 0 does lead to a degree of parameterization dependence in estimates of E1(0), but one that is still modest. The fit parameter values for all presented fits are provided in Appendix D.

25707#pg2>. The computed hc → γη(′) form–factor values for real photons lie significantly below the experimental values <ref:2607. p. 0
Where the paper says it

Computing for a large number of different meson momenta, we mapped out the timelike Q2 dependence, which can be accessed through Dalitz decays, J/ψ → ℓ +ℓ − η (′) , for leptons ℓ = e, µ [6, 7]. This prior calculation, performed with heavier than physical light–quark masses, such that mπ ∼ 391 MeV, showed that the techniques applied are effective in determining statistically clean results, but the real photon transition rates extracted for J/ψ → γη and J/ψ → γη′ both lie significantly below the experimental values.

25707#pg2>.

How the training signal is built

The authors employ a procedure for simultaneous fitting of time–dependencies of multiple three-point correlation functions to obtain a single value of J <ref:2607.25707#pg2>. This procedure involves systematically removing the same number of points from the source and/or the sink on every ∆t simultaneously, while ensuring that at least two timeslices are retained on the smallest ∆t <ref:2607. p. 4

Where the paper says it

In the current calculation we again compute on multiple time–separations, in this case ∆t/at = {16, 20, 24, 28}, but we now introduce a procedure of simultaneous fitting. Allowing for all possible time–windows in each ∆t would lead to an unmanageable number of fits, and as such we adopt a simplified scheme where, beginning with the complete set of timeslices, we systematically remove the same number of points from the source and/or the sink on every ∆t simultaneously, while ensuring that at least two timeslices are retained on the smallest ∆t.

25707#pg2>. The fit parameters, a subset of J, Asrc, δEsrc, Asnk, δEsnk are chosen to be common for all ∆t <ref:2607. p. 4
Where the paper says it

6 are considered with just a constant, a constant plus a source exponential, a constant plus a sink exponential, or a constant with both a source and sink exponential. The fit parameters, a subset of {J, Asrc, δEsrc, Asnk, δEsnk}, are chosen to be common for all ∆t.

25707#pg2>.

Where the gains come from

The study demonstrates that relevant signals with a high degree of statistical precision can be obtained by use of contemporary techniques <ref:2607.25707#pg2>. The particular sensitivity of the η, η′ system to detailed properties of the gauge–field configurations may not be shared by other light–meson final states <ref:2607. p. 0

Where the paper says it

Could there be unusually large discretization errors that are not visible as departures from continuum–like behavior at a single lattice spacing? Might the properties of η and η ′ relevant to this process be particularly sensitive to any imperfect topological sampling of gauge fields on these particular lattices?

p. 0
Where the paper says it

Could there be unusually large discretization errors that are not visible as departures from continuum–like behavior at a single lattice spacing? Might the properties of η and η ′ relevant to this process be particularly sensitive to any imperfect topological sampling of gauge fields on these particular lattices?

25707#pg2>. The results show that relevant signals with a high degree of statistical precision can be obtained by use of contemporary techniques <ref:2607.25707#pg2>. The authors conclude that the most likely origin of the observed discrepancy is some property of the η and η′ on this particular lattice <ref:2607. p. 0
Where the paper says it

This prior calculation, performed with heavier than physical light–quark masses, such that mπ ∼ 391 MeV, showed that the techniques applied are effective in determining statistically clean results, but the real photon transition rates extracted for J/ψ → γη and J/ψ → γη′ both lie significantly below the experimental values. The origin of this discrepancy remains unknown.

25707#pg2>. The next step will be to consider multi–hadron production, such as J/ψ → γ ππ and the coupled–channel J/ψ → γ KK <ref:2607.25707#pg2>. The results of this paper and Ref. p. 1
Where the paper says it

We will compute transitions ψ ′ → γη(′) , presenting first lattice QCD results for this process, finding a signal that is statistically compatible with zero on the current lattices. We refer the reader to Ref.

p. 1
Where the paper says it

We will compute transitions ψ ′ → γη(′) , presenting first lattice QCD results for this process, finding a signal that is statistically compatible with zero on the current lattices. We refer the reader to Ref.

p. 1
Where the paper says it

We will compute transitions ψ ′ → γη(′) , presenting first lattice QCD results for this process, finding a signal that is statistically compatible with zero on the current lattices. We refer the reader to Ref.

[4] show that relevant signals with a high degree of statistical precision can be obtained by use of contemporary techniques <ref:2607.25707#pg2>. The authors acknowledge support from the U.S. Department of Energy contract DE-SC0018416 at William & Mary <ref:2607. p. 0
Where the paper says it

Dudek,2, † and Robert G. Edwards3, ‡ (for the Hadron Spectrum Collaboration) 1CSSM, Department of Physics, Adelaide University, Adelaide, 5005, South Australia, Australia 2Department of Physics, College of William and Mary, Williamsburg, VA 23187, USA 3Thomas Jefferson National Accelerator Facility, 12000 Jefferson Avenue, Newport News, VA 23606, USA (Dated: July 29, 2026) We explore, for the first time in lattice QCD, the radiative decays to γη and γη′ of the J P C = 1+− charmonium state, hc.

25707#pg2>. The material is based upon work supported by the U. p. 5
Where the paper says it

[4] we described the use of explicit parameterizations to capture the Q2 dependence of the single form–factor for each of J/ψ → γη and J/ψ → γη′ . We presented fits using a dipole form and two exponential–based forms, as well as a rather flexible approach based upon a conformal mapping of Q2 into a variable, z, in terms of which a low–order polynomial description could be found.

S. Department of Energy, Office of Science, Office of Nuclear Physics under Contract No. p. 0
Where the paper says it

Dudek,2, † and Robert G. Edwards3, ‡ (for the Hadron Spectrum Collaboration) 1CSSM, Department of Physics, Adelaide University, Adelaide, 5005, South Australia, Australia 2Department of Physics, College of William and Mary, Williamsburg, VA 23187, USA 3Thomas Jefferson National Accelerator Facility, 12000 Jefferson Avenue, Newport News, VA 23606, USA (Dated: July 29, 2026) We explore, for the first time in lattice QCD, the radiative decays to γη and γη′ of the J P C = 1+− charmonium state, hc.

89243126CSC000213 <ref:2607. p. 0
Where the paper says it

In practice this means that the hc often appears together with the J/ψ as an excited state such that we cannot simply rely upon ground–state dominance at large Euclidean times – we will use a further application of optimized operator technology to access it reliably. The arXiv:2607.

25707#pg2>. This work used clusters at Jefferson Laboratory under the USQCD Initiative and the LQCD ARRA project <ref:2607. p. 0
Where the paper says it

Edwards3, ‡ (for the Hadron Spectrum Collaboration) 1CSSM, Department of Physics, Adelaide University, Adelaide, 5005, South Australia, Australia 2Department of Physics, College of William and Mary, Williamsburg, VA 23187, USA 3Thomas Jefferson National Accelerator Facility, 12000 Jefferson Avenue, Newport News, VA 23606, USA (Dated: July 29, 2026) We explore, for the first time in lattice QCD, the radiative decays to γη and γη′ of the J P C = 1+− charmonium state, hc. This work expands upon a previous calculation of J/ψ decays to the same final states, incorporating novel technology to access the hc where it is an excited state and to separate the independent electric dipole and longitudinal form–factors.

p. 0
Where the paper says it

Edwards3, ‡ (for the Hadron Spectrum Collaboration) 1CSSM, Department of Physics, Adelaide University, Adelaide, 5005, South Australia, Australia 2Department of Physics, College of William and Mary, Williamsburg, VA 23187, USA 3Thomas Jefferson National Accelerator Facility, 12000 Jefferson Avenue, Newport News, VA 23606, USA (Dated: July 29, 2026) We explore, for the first time in lattice QCD, the radiative decays to γη and γη′ of the J P C = 1+− charmonium state, hc. This work expands upon a previous calculation of J/ψ decays to the same final states, incorporating novel technology to access the hc where it is an excited state and to separate the independent electric dipole and longitudinal form–factors.

25707#pg2>. The software codes Chroma [30], QUDA [31, 32], QUDA-MG [33], QPhiX [34], MG PROTO [35], and QOPQDP [36, 37] were used <ref:2607. p. 0
Where the paper says it

In practice this means that the hc often appears together with the J/ψ as an excited state such that we cannot simply rely upon ground–state dominance at large Euclidean times – we will use a further application of optimized operator technology to access it reliably. The arXiv:2607.

25707#pg2>. The authors thank their colleagues within the Hadron Spectrum Collaboration for their continued assistance <ref:2607. p. 0
Where the paper says it

Dudek,2, † and Robert G. Edwards3, ‡ (for the Hadron Spectrum Collaboration) 1CSSM, Department of Physics, Adelaide University, Adelaide, 5005, South Australia, Australia 2Department of Physics, College of William and Mary, Williamsburg, VA 23187, USA 3Thomas Jefferson National Accelerator Facility, 12000 Jefferson Avenue, Newport News, VA 23606, USA (Dated: July 29, 2026) We explore, for the first time in lattice QCD, the radiative decays to γη and γη′ of the J P C = 1+− charmonium state, hc.

25707#pg2>. This work contributes to the goals of the U. p. 0
Where the paper says it

Edwards3, ‡ (for the Hadron Spectrum Collaboration) 1CSSM, Department of Physics, Adelaide University, Adelaide, 5005, South Australia, Australia 2Department of Physics, College of William and Mary, Williamsburg, VA 23187, USA 3Thomas Jefferson National Accelerator Facility, 12000 Jefferson Avenue, Newport News, VA 23606, USA (Dated: July 29, 2026) We explore, for the first time in lattice QCD, the radiative decays to γη and γη′ of the J P C = 1+− charmonium state, hc. This work expands upon a previous calculation of J/ψ decays to the same final states, incorporating novel technology to access the hc where it is an excited state and to separate the independent electric dipole and longitudinal form–factors.

S. Department of Energy ExoHad Topical Collaboration, Contract No. p. 0
Where the paper says it

Dudek,2, † and Robert G. Edwards3, ‡ (for the Hadron Spectrum Collaboration) 1CSSM, Department of Physics, Adelaide University, Adelaide, 5005, South Australia, Australia 2Department of Physics, College of William and Mary, Williamsburg, VA 23187, USA 3Thomas Jefferson National Accelerator Facility, 12000 Jefferson Avenue, Newport News, VA 23606, USA (Dated: July 29, 2026) We explore, for the first time in lattice QCD, the radiative decays to γη and γη′ of the J P C = 1+− charmonium state, hc.

DE-SC0023598 <ref:2607. p. 0
Where the paper says it

In practice this means that the hc often appears together with the J/ψ as an excited state such that we cannot simply rely upon ground–state dominance at large Euclidean times – we will use a further application of optimized operator technology to access it reliably. The arXiv:2607.

25707#pg2>. The authors acknowledge support from the U.S. Department of Energy, Office of Science, Office of Advanced Scientific Computing Research and Office of Nuclear Physics, Scientific Discovery through Advanced Computing (SciDAC) program <ref:2607.25707#pg2>. Also acknowledged is support from the Exascale Computing Project (17-SC20-SC), a collaborative effort of the U. p. 0
Where the paper says it

required us to consider large values of meson momentum in order to sample the decay kinematics corresponding to nearly real photons, Q2 ≈ 0. Computing for a large number of different meson momenta, we mapped out the timelike Q2 dependence, which can be accessed through Dalitz decays, J/ψ → ℓ +ℓ − η (′) , for leptons ℓ = e, µ [6, 7].

S. Department of Energy Office of Science and the National Nuclear Security Administration <ref:2607. p. 0
Where the paper says it

Dudek,2, † and Robert G. Edwards3, ‡ (for the Hadron Spectrum Collaboration) 1CSSM, Department of Physics, Adelaide University, Adelaide, 5005, South Australia, Australia 2Department of Physics, College of William and Mary, Williamsburg, VA 23187, USA 3Thomas Jefferson National Accelerator Facility, 12000 Jefferson Avenue, Newport News, VA 23606, USA (Dated: July 29, 2026) We explore, for the first time in lattice QCD, the radiative decays to γη and γη′ of the J P C = 1+− charmonium state, hc.

25707#pg2>. MB is supported by the Australian Research Council grants DP240102839 and DP260104445 <ref:2607. p. 0
Where the paper says it

In practice this means that the hc often appears together with the J/ψ as an excited state such that we cannot simply rely upon ground–state dominance at large Euclidean times – we will use a further application of optimized operator technology to access it reliably. The arXiv:2607.

25707#pg2>. This work used clusters at Jefferson Laboratory under the USQCD Initiative and the LQCD ARRA project <ref:2607.25707#pg2>. The DiRAC component of CSD3 was funded by BEIS capital funding via STFC capital grants ST/P002307/1 and ST/R002452/1 and STFC operations grant ST/R00689X/1 <ref:2607.25707#pg2>. Other components were provided by Dell EMC and Intel using Tier-2 funding from the Engineering and Physical Sciences Research Council (capital grant EP/P020259/1) <ref:2607.25707#pg2>. The DiRAC is part of the National E-Infrastructure <ref:2607. p. 0
Where the paper says it

Dudek,2, † and Robert G. Edwards3, ‡ (for the Hadron Spectrum Collaboration) 1CSSM, Department of Physics, Adelaide University, Adelaide, 5005, South Australia, Australia 2Department of Physics, College of William and Mary, Williamsburg, VA 23187, USA 3Thomas Jefferson National Accelerator Facility, 12000 Jefferson Avenue, Newport News, VA 23606, USA (Dated: July 29, 2026) We explore, for the first time in lattice QCD, the radiative decays to γη and γη′ of the J P C = 1+− charmonium state, hc.

25707#pg2>. Also used was an award of computer time provided by the U. p. 4
Where the paper says it

The fit parameters, a subset of {J, Asrc, δEsrc, Asnk, δEsnk}, are chosen to be common for all ∆t. Each time–window and parameterization variation is considered and the fit characterized by the value of the Akaike Information Criterion (AIC), with the 30 largest values retained and used in a model average [12] to yield a single value of J with, we argue, a conservatively estimated uncertainty.

S. Department of Energy INCITE program and supported in part under an ALCC award, and resources at: the Oak Ridge Leadership Computing Facility, which is a DOE Office of Science User Facility supported under Contract DE-AC05-00OR22725 <ref:2607. p. 0
Where the paper says it

Dudek,2, † and Robert G. Edwards3, ‡ (for the Hadron Spectrum Collaboration) 1CSSM, Department of Physics, Adelaide University, Adelaide, 5005, South Australia, Australia 2Department of Physics, College of William and Mary, Williamsburg, VA 23187, USA 3Thomas Jefferson National Accelerator Facility, 12000 Jefferson Avenue, Newport News, VA 23606, USA (Dated: July 29, 2026) We explore, for the first time in lattice QCD, the radiative decays to γη and γη′ of the J P C = 1+− charmonium state, hc.

25707#pg2>. The National Energy Research Scientific Computing Center (NERSC), a U. p. 0
Where the paper says it

Dudek,2, † and Robert G. Edwards3, ‡ (for the Hadron Spectrum Collaboration) 1CSSM, Department of Physics, Adelaide University, Adelaide, 5005, South Australia, Australia 2Department of Physics, College of William and Mary, Williamsburg, VA 23187, USA 3Thomas Jefferson National Accelerator Facility, 12000 Jefferson Avenue, Newport News, VA 23606, USA (Dated: July 29, 2026) We explore, for the first time in lattice QCD, the radiative decays to γη and γη′ of the J P C = 1+− charmonium state, hc.

S. Department of Energy Office of Science User Facility located at Lawrence Berkeley National Laboratory, operated under Contract No. p. 0
Where the paper says it

Dudek,2, † and Robert G. Edwards3, ‡ (for the Hadron Spectrum Collaboration) 1CSSM, Department of Physics, Adelaide University, Adelaide, 5005, South Australia, Australia 2Department of Physics, College of William and Mary, Williamsburg, VA 23187, USA 3Thomas Jefferson National Accelerator Facility, 12000 Jefferson Avenue, Newport News, VA 23606, USA (Dated: July 29, 2026) We explore, for the first time in lattice QCD, the radiative decays to γη and γη′ of the J P C = 1+− charmonium state, hc.

DE-AC02-05CH11231 <ref:2607. p. 0
Where the paper says it

In practice this means that the hc often appears together with the J/ψ as an excited state such that we cannot simply rely upon ground–state dominance at large Euclidean times – we will use a further application of optimized operator technology to access it reliably. The arXiv:2607.

25707#pg2>. The Texas Advanced Computing Center (TACC) at The University of Texas at Austin <ref:2607.25707#pg2>. The Extreme Science and Engineering Discovery Environment (XSEDE), which is supported by National Science Foundation Grant No. p. 0
Where the paper says it

Dudek,2, † and Robert G. Edwards3, ‡ (for the Hadron Spectrum Collaboration) 1CSSM, Department of Physics, Adelaide University, Adelaide, 5005, South Australia, Australia 2Department of Physics, College of William and Mary, Williamsburg, VA 23187, USA 3Thomas Jefferson National Accelerator Facility, 12000 Jefferson Avenue, Newport News, VA 23606, USA (Dated: July 29, 2026) We explore, for the first time in lattice QCD, the radiative decays to γη and γη′ of the J P C = 1+− charmonium state, hc.

ACI-1548562 <ref:2607. p. 0
Where the paper says it

In practice this means that the hc often appears together with the J/ψ as an excited state such that we cannot simply rely upon ground–state dominance at large Euclidean times – we will use a further application of optimized operator technology to access it reliably. The arXiv:2607.

25707#pg2>. Blue Waters is a joint effort of the University of Illinois at Urbana-Champaign and its National Center for Supercomputing Applications <ref:2607. p. 0
Where the paper says it

Dudek,2, † and Robert G. Edwards3, ‡ (for the Hadron Spectrum Collaboration) 1CSSM, Department of Physics, Adelaide University, Adelaide, 5005, South Australia, Australia 2Department of Physics, College of William and Mary, Williamsburg, VA 23187, USA 3Thomas Jefferson National Accelerator Facility, 12000 Jefferson Avenue, Newport News, VA 23606, USA (Dated: July 29, 2026) We explore, for the first time in lattice QCD, the radiative decays to γη and γη′ of the J P C = 1+− charmonium state, hc.

25707#pg2>. The authors conclude by restating that the most likely origin of the observed discrepancy with respect to experiment common to both J/ψ → γη(′) and hc → γη(′) is some property of the η and η′ on this particular lattice <ref:2607. p. 0
Where the paper says it

This prior calculation, performed with heavier than physical light–quark masses, such that mπ ∼ 391 MeV, showed that the techniques applied are effective in determining statistically clean results, but the real photon transition rates extracted for J/ψ → γη and J/ψ → γη′ both lie significantly below the experimental values. The origin of this discrepancy remains unknown.

25707#pg2>. The importance of the SU(3) flavor singlet content of these mesons in this process, and the sensitivity of that component to the U(1) axial anomaly, motivates consideration on other lattices where the gauge–field topology sampling is better studied <ref:2607. p. 0
Where the paper says it

Could there be unusually large discretization errors that are not visible as departures from continuum–like behavior at a single lattice spacing? Might the properties of η and η ′ relevant to this process be particularly sensitive to any imperfect topological sampling of gauge fields on these particular lattices?

25707#pg2>. The results of this paper and Ref. [4] show that relevant signals with a high degree of statistical precision can be obtained by use of contemporary techniques <ref:2607.25707#pg2>. The particular sensitivity of the η, η′ system to detailed properties of the gauge–field configurations may not be shared by other light–meson final states <ref:2607.25707#pg2>. The results of this paper and Ref. [4] show that relevant signals with a high degree of statistical precision can be obtained by use of contemporary techniques <ref:2607.25707#pg2>.

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

1

Improvement in lattice QCD calculation of radiative decay rates for hc to gamma eta and gamma eta prime by incorporating novel technology to access hc as an excited state and separating independent electric dipole and longitudinal form factors. This allows for a more rigorous study of the transition matrix element decomposition, which is crucial for understanding the physical process.

2

Improvement in accessing the J P C = 1+− charmonium state hc at various momenta by using optimized operators derived from diagonalization of a matrix of two-point correlation functions. This enables reliable projection out of hc even when it appears as an excited state in certain irreps, such as B1, B2, and E2.

3

Improvement in extracting multipole transition form factors E1(Q2) and C1(Q2) by introducing a procedure for simultaneous fitting of time-dependencies of multiple three-point correlation functions with different source-sink separations. This approach provides a more robust determination of the matrix element J by systematically removing excited state pollution.

4

Improvement in analyzing the Q2 dependence of form factors by using various parameterizations, including dipole and exponential forms, as well as a conformal mapping into a variable z(Q2). This allows for comprehensive characterization of the Q2 range explored in the calculation.

5

Improvement in predicting real photon transition rates by comparing estimates from various parameterizations of the Q2 dependence of E1(Q2) with values extracted from BESIII experimental decay widths. This provides a cross-validation method to assess the accuracy of lattice QCD results.

Abstract

We explore, for the first time in lattice QCD, the radiative decays to γη and γη' of the J PC=1+- charmonium state, h c. This work expands upon a previous calculation of J/ψ decays to the same final states, incorporating novel technology to access the h c where it is an excited state and to separate the independent electric dipole and longitudinal form--factors. Results for the radiative decay rates, computed on three-flavor lattices with m π about 391 MeV show a suppression relative to experiment that is similar to that observed in the prior calculation of J/ψ to γη. The timelike Q squared dependence of both form--factors together influence the Dalitz decays, h c to+- η, which are as yet unmeasured experimentally.

Sources

Related papers