We are delighted to announce the upcoming Fifth Gogny conference, building on the success of previous editions held in Bruyères-le-Châtel (2015), Madrid (2017), Livermore (2019), and Granada (2022). This new edition is scheduled to take place in downtown Paris, from December 10th to 13th, 2024. The entire conference will take place at Campus des Cordeliers, located at 15 rue de l'école de médecine, in the 6th district.
The aim of these conference series is to regularly bring together researchers working on finite-range nuclear interactions and associated many-body approaches. The traditional themes encompass:
In this 2024 edition, a session will be dedicated to quantum computing and artificial intelligence.
Program and participation
The scientific program is made from the recommandations of our International Advisory Committee, ensuring that the presentations span the developments of all the international community. We are pleased to offer complimentary lunches and coffee breaks for all participants throughout the conference.
A limited number of seats are available for scholar wishing to attend the conference and follow the recent progresses of the field.
Social Events
We are delighted to host the conference dinner at the renowned Musée d'Orsay restaurant on Thursday, December 12th. Participants will enjoy an exclusive, complimentary tour of the museum an hour and a half before dinner, providing a unique opportunity to appreciate its extensive collection of French Impressionist masterpieces.
The Fifth Gogny conference is jointly organized by CEA, Laboratoire Kastler Brossel, IJCLab/CNRS and Sorbonne Université.
|
|
|
|
Describing nuclear phenomena across a wide range of energy scales from hundreds of MeV in binding energies to fractions of an MeV for low-lying excitations—remains a long-standing challenge in nuclear physics. The ab initio method is a systematically improvable approach for quantitatively describing nuclei using the finest resolution scale possible while maximizing its predictive capabilities. In this talk, I will highlight recent advances in ab initio nuclear structure calculations, focusing on developments in chiral nuclear forces and methods for estimating uncertainties in theoretical predictions.
One of the most well-known methods to deal with many-body systems is the Hartree-Fock (HF) approximation. In this talk we show how the application of renormalization group (RG) techniques within the HF approximation leads to the Skyrme and Gogny forces. Unlike the application of the RG to the two-body system, where we end up with two infrared fixed points that describe weakly and strongly interacting systems respectively, here we find a unique fixed point that generate contact-range interactions that are independent of the cutoff. This leads to the reinterpretation of Gogny and Skyrme forces as renormalizable effective forces.
The key ingredient for mean-field calculations in nuclear structure is the effective interaction which models the strong force in the nuclear medium. Such interactions usually depend on a set of parameters fitted to properties nuclei and infinite nuclear matter. These interactions can suffer several limitations and problems. For example, since they are usually adjusted on properties of observed nuclei close to the valley of stability, their predictive power for exotic and super-heavy nuclei may be questionable. Furthermore, unphysical finite-size instabilities can sometimes appear when these interactions are used to calculate properties of nuclei which have not been considered to constrain their parameters. These instabilities can appear in various channels and therefore have scalar, vector, isoscalar or isovector characters. It was shown that the formalism of the linear response in infinite-nuclear matter can be used to avoid such instabilities for the construction of zero-range interaction (of Skyrme type). Although such a formalism was also developed for finite-range interactions (Gogny type), the calculations for the linear response are much more time-consuming and can hardly be incorporated in the procedure used to fit their parameters. I will discuss how the scalar-isovector instabilities are related to the distributions of protons and neutrons in nuclei and how, in turn, information on charge density distributions can be used to prevent these instabilities. Beside the avoidance of instabilities, information about the charge distribution may lead to a better balance between the different contributions to the binding energy of nuclei and their evolution with mass and asymmetry. I will show that the use of constraints on charge distributions from a set of chosen nuclei can be used to avoid the appearance of scalar-isovector instabilities and discuss how this could improve the predictive power of the mean-field calculations.
In this talk, I will discuss a recent extension of the Gogny interaction including both finite range spin-orbit and tensor terms. I will present nuclear matter results and finite nuclei ones obtained at the mean-field and beyond levels, in particular spectroscopy in odd nuclei.
Nuclear structure physics endeavors a robust and accurate description of the way an arbitrary number of nucleons self-organizes and gets disorganized in nuclei. To achieve this goal, nuclear physics has entered an era of reformulation of its standard phenomenological models into bona fide effective field theories (EFTs). However, the empirical microscopic model offering the best compromise between predictive power and numerical complexity - the so-called energy density functional (EDF) - has so far resisted all attempts of reformulation into an EFT. This presentation examines various attempts to derive the EDF approach from first principles.
The nuclear charge radius can provide information about the nuclear interactions and the nuclear structure. While global trend of the charge radii is guided by the nuclear bulk properties, its local variation is affected by various nuclear structure effects. Due to recent advances in experimental techniques, new high-precision data on charge radii on various isotopic chains allows testing nuclear structure models more thoroughly. I will discuss the predictions of the nuclear charge radii on various isotopic chains with DFT-based models, concentrating in particularly on results obtained with the Fayans EDF. Presently used Fayans EDF parametrizations were adjusted with a focus on nuclear charge radii. Recent measurements in K, Ag, and Pd isotopic chains have shown that while DFT-based models can predict overall trends well, there are, nevertheless, some deficiencies. Typically, Fayans EDF seems to reproduce the experimental trend better than Skyrme EDF models, however, the magnitude of the odd-even staggering of the radii is often overestimated. In addition, I will also discuss the similar pattern of differential charge radii in even-even isotopic chains from Ca to Zn with DFT and ab-initio-based nuclear structure models, and octupole deformation properties of Fayans EDF in the Actinide region.
Nuclear shell model (NSM) and variational mean-field approximations and their beyond-mean-field extensions are the two main workhorses in the calculation of nuclear structure observables. In the last few years, thanks to the availability of methods like Monte Carlo Shell Model (MCSM) or the projected generator coordinate method (PGCM) and its multiple variants, NSM individual states can be interpreted in terms of collective coordinates defined in the intrinsic frame (e.g., multipole deformations). On the other hand, the evaluation of occupation numbers of self-consistent spherical orbits within the PGCM framework also allows for a NSM-like interpretation of the nuclear states provided by this method. These theoretical developments are very useful to, e.g., fully understand collective phenomena like shape-evolution, shape-mixing and shape-coexistence, as well as the appearance/degradation of magic numbers. In this contribution I will show some examples of how these two (in principle) different approaches are actually two sides of the same coin.
In the spirit of the present conference, one should recall that Daniel Gogny, in his seminal paper where the force that bears his name was introduced, said that one should leave room for the evaluation of the “correction … due to the quasiparticle-vibrational coupling”. The accuracy that one can reach through theoretical calculations of single-particle (or quasi-particle) energies has been the subject of long debates. In this talk, we will discuss recent calculations that (a) use state-of-the-art Energy Density Functionals (EDFs), (b) include pairing and tensor forces, and (c) include the coupling with collective vibrations. Results will be compared with existing data as well as with recent data from ISOLDE. We will also compare the results of similar calculations for collective states, like low-lying multipole states or giant resonances, with experimental findings. Implications for the nuclear equation of state will be also addressed. As in the case of the single-particle spectra, we aim to assess the merits and the limitations of state-of-the-art calculations based on Energy Density Functionals (EDFs), including quasiparticle-vibrational coupling. We conclude by discussing the predictive power of currently used EDFs.
How collective modes emerge in nuclei, especially in cases where the effects of proximity to the particle continuum are large, remains one of the core problems in nuclear theory. Furthermore, such collective modes have a profound effect on reactions that occur in stellar interiors and nuclear technology applications. Thus, a good description of astrophysical reaction rates needs to properly take into account the clustering (and deformation) aspects of nuclear structure. We will discuss progress in the first-principle description of nuclear reactions, as well as recent applications of the no-core shell model with continuum (NCSMC) to various reactions of astrophysical interest, the description of alpha clustering.
Gamow-Teller excitations and beta-decay are studied within the subtracted second random-phase approximation (SSRPA), based on Skyrme functionals.. The comparison with the conventional random-phase approximation (RPA) results and experimental data is discussed, showing the improvement obtained within the SSRPA. In particular, it is found that the amount of Gamow-Teller strength obtained in SSRPA is much lower than the RPA one, and it agrees better with experimental data [1,2]. The inclusion of two-particle-two-hole configurations is responsible for this quenching, avoiding thus the use of any “ad-hoc” quenching factors normally adopted in this kind of studies. The beta-decay half lives are also calculated and discussed, showing that also in this case the inclusion of the two-particle-two-hole configurations allows for a better description of the experimental values [1,3], without any use of the “quenching factor” usually needed in RPA. This result may have implications for the computation of nuclear matrix elements in the same framework for neutrinoless double-beta decay.
References:
[1] D. Gambacurta, M. Grasso, and J. Engel Phys. Rev. Lett. 125, 212501, (2020)
[2] D. Gambacurta and M. Grasso, Phys. Rev. C 105, 014321, (2022)
[3] D. Gambacurta and M. Grasso, in preparation
Neutrinoless double-beta decay, a sought-after process in which two nucleons in a nucleus beta-decay simultaneously without emitting neutrinos, is a promising probe for physics beyond the Standard Model. It would not only demonstrate lepton-number violation but also shed light on the unknown nature of neutrinos. However, extracting the interesting physics from experiments requires knowledge on nuclear matrix elements(NMEs), which currently form a major obstacle. In my talk, I will cover recents improvements on the NMEs in two widely used frameworks: the proton-neutron quasiparticle random-phase approximation (pnQRPA) and nuclear shell model (NSM). I will discuss how correlations with other observables---that can be or have been measured---can help constrain the values of the NMEs. I will also show first results for the NMEs with complete next-to-next-to-leading-order terms in medium-mass to heavy nuclei.
The Projected Generator Coordinate Method (PGCM) is a powerful and versatile many-body method that has been used for decades to study low-energy nuclear structure. In particular, the PGCM is particularly well suited to describe collective nuclear phenomena such as deformation or pairing. Also, a great strength of the method is the proper treatment of quantum numbers associated with the symmetries of the Hamiltonian that permits the unambiguous evaluation of the principal observables of interest in nuclear spectroscopy (spin and parity, electric quadrupole and magnetic dipole moments, reduced transition probabilities, ...). In the past, PGCM calculations have been mostly performed using phenomenological representations of the nuclear Hamiltonian. In recent years, however, we also developed numerical tools to perform advanced PGCM calculations based on microscopic Hamiltonians derived from chiral effective field theory. In this presentation, I will discuss the progress to anchor the PGCM in an ab initio philosophy and show recent results concerning nuclear structure and related areas.
In recent times, it has become commonplace to mention the unification of structure and reaction nuclear theory as one of the hot topics in low-energy nuclear physics. This interest is, of course, not new, but some present circumstances might have made it more acute. First, the experimental access to very weakly bound or unbound nuclei has blurred the limits between structure and reaction theory. Second, the fast development of computational tools and resources has rendered scattering problems tractable with bound states techniques. We will also address some ideas in the path to another important unification: the theory of direct and compound nucleus reactions. This line of research is important in order to address important processes, such as capture reactions, involving nuclei away from the stability valley, where an unusually low level density calls for the description of a transition between the statistical and direct reaction regimes.
The compound nucleus theory plays the central role in calculating nuclear reaction processes at relatively low energies, and the statistical treatment for the compound resonances leads to the so-called Hauser-Feshbach theory with width fluctuation correction. The Hauser-Feshbach theory is particularly important for providing reliable nuclear reaction data not only in nuclear technology development but also fundamental science development such as astrophysics. The theory of Hauser and Feshbach is to replace the average compound decay width by the optical model transmission coefficient, and the imposed unitarity gives rise to reduction in the cross sections for all the $a \ne b$ channels due to the elastic enhancement. The statistical properties of the compound nucleus reaction can be studied by a stochastic S-matrix that includes Gaussian Orthogonal Ensemble (GOE) in the propagator. This also allows us to study the average S-matrix behavior when strongly coupled channels exist. In addition to the main framework of the Hauser-Feshbach model, practical calculations often require detailed knowledge of nuclear structure too, which might be informed by measurements or theoretical predictions. The theoretical modeling for nuclear structure includes a wide spectrum of fully microscopic approaches to phenomenological models. In this talk, we present recent progress in the statistical compound nucleus theory, and applications to nuclear data in various nuclear technology fields.
Nuclear reactions are one of the most diverse probes to study nuclear properties. They are also used to extract astrophysically relevant information. However, the reaction theory needed for the interpretation of the measurements contains significant uncertainties. I will review the recent efforts on performing Bayesian analyses of reaction models and the efforts to quantify the uncertainties coming from the optical potential.
The past couple of decades have seen tremendous advances in nuclear structure and reaction theory. Innovative theory frameworks for describing the nuclear many-body system, increasingly powerful computers, and opportunities for confronting theory predictions with data on unstable nuclei, have been driving the field. An important goal is to move from phenomenological ingredients in reaction calculations to predictive theories based on microscopic frameworks. I will discuss ongoing efforts aimed at integrating microscopic descriptions of nuclear structure into reaction predictions for medium-mass and heavy nuclei. I will highlight areas where such efforts can improve nuclear data evaluations and also enable indirect measurements of important reaction cross sections.
Within the last 20 years, nuclear-structure calculations have made very significant progresses. More and more nuclei can now be computed ab initio. This prowess is not solely limited to the bottom of the valley of stability, but reaches the nucleon driplines [1]. Testing the predictions of ab initio calculations away from stability requires a way to bridge these many-body results to usual few-body models of reactions. Halo-EFT (or Cluster-EFT) provides a simple, yet effective, description of halo nuclei. It is built upon the clear separation of scales in these exotic nuclei: a compact and tightly-bound core to which a diffuse halo is loosely bound [2,3]. This description can be easily included within a model of reactions, such as transfer [4], breakup [5], or knockout [6]. By fitting the low-energy constants of this description to the ab initio predictions, we can test the reliability of these models. Conversely, starting from reaction observables, we can also infer key structure informations such as the one-nucleon separation energy or the asymptotic normalisation constant of the bound state [7].
References: [1] A. Calci, P. Navrátil, R. Roth, J. Dohet-Eraly, S. Quaglioni, G. Hupin, Phys. Rev. Lett. 117, 242501 (2016).
[2] C.A. Bertulani, H.-W. Hammer, U. Van Kolck, Nucl. Phys. A 712, 37–58 (2002).
[3] H.-W. Hammer, C. Ji, D.R. Phillips, J. Phys. G 44, 103002 (2017).
[4] J. Yang, P. Capel, Phys. Rev. C 98, 054602 (2018).
[5] P. Capel, D.R. Phillips, H.-W. Hammer, Phys. Rev. C 98, 034610 (2018).
[6] C. Hebborn, P. Capel, Phys. Rev. C 104, 024616 (2021). [7] P. Capel, D.R. Phillips et al. Eur. Phys. J. A 59, 273 (2023).
Based on a momentum-space in-medium folding model, we disclose the universal separability of the optical potential, revealing its radial and non- locality features at beam energies in the range 40-400 MeV and target mass numbers in the range 40≤A≤208. From this microscopic study we find that the nonlocality form factor is inherently complex and of hydrogenic nature, affecting both central and spin-orbit components of the potential. A striking outcome from this study is the consistent appearance of a nodal point in the imaginary radial form factor, notably suppressing surface absorption peaks, in evident contrast with Woods-Saxon’s assumption of an absorptive peak at the nuclear surface. Our analysis reveals that the complex radial form factor can effectively be represented as convolutions of uniform spherical distribution with a Gaussian form factor and a Yukawa term. These ro- bust microscopically driven findings offer new ways for investigating nuclear reactions beyond the restricting Woods-Saxon and Perey-Buck assumptions.
In this talk I will describe our formalism to treat odd nuclei and apply it in particular to study different properties of isotope chains, as binding energies, deformation, neutron separation energies, charge radii, etc. Our approximation is based in a HF+BCS model, where the finite range Gogny force is used as effective nuclear interaction; even-even and even-odd nuclei can be described within the framework of this approach. We analize isotopes chains from oxygen to calcium.
Neutron-rich nuclei provide important insights to nuclear forces and to the nuclear equation of state. Advances in ab initio methods combined with new opportunities with rare isotope beams enable unique explorations of their properties based on nuclear forces applicable over the entire nuclear chart. In this talk, I will present novel chiral low-resolution interactions that accurately describe bulk properties from 16O to 208Pb, including density distributions and neutron skins of neutron-rich nuclei. I will show how neutron skins are narrowly predicted over all nuclei with interesting sensitivities for the most extreme, experimentally unexplored cases.
Despite major and numerous recent progresses within ab initio methods [1], these are not capable yet to describe the whole variety of phenomena over the nuclear chart, and Energy Density Functionals thus remain the tool of choice to such end at present. To build a functional suited for astrophysical applica- tions, one must describe as good as possible nuclear properties, such as masses, of crucial importance for r-process [2], as well as infinite nuclear matter properties of importance for neutron stars [3]. Only this way one can hope to extrapolate reliable nuclear properties up to the neutron driplines. Several competitive nuclear mass models within mean-field framework have appeared over the years to that purpose. Recent ones based on Skyrme interaction give excellent results [4, 5] and exhibit faithful advantages such as low computational cost, however, they are also intrinsicly penalized by required energy cut-offs to avoid non physical divergences. Gogny forces, designed mostly to overcome these problems, come then into play to complete the picture. Fitting the functional free parameters is a tedious task and there exist much less Gogny than Skyrme interactions on the market, even less usable for astrophysical models, most of them having a root mean square over masses as high as 2-6 MeV. Motivated by an extension of the Gogny force [6] including a third gaussian in the central term, recents results [7] have lead us to think we could build a parametrization fulfilling previously mentioned require- ments. I will present the freshly obtained D3G3M Gogny-Hartree-Fock-Bogoliubov nuclear mass model [8] by means of a systematic fitting protocole over infinite nuclear matter and nuclear masses similar to the one adopted for D1M [10], exhibting both a rms as low as 800 keV and excellent nuclear matter properties, and I will show the improvements made compared to other Gogny forces.
References [1] H. Hergert, Front. Phys. 8 (2020) [2] M. Arnould, S. Goriely, and K. Takahashi , Phys. Rep. 450, 97 (2007). [3] M. Oertel, M. Hempel, T. Klähn, and S. Typel Rev Mod. Phys. 89, 015007 (2017) [4] G. Grams, W. Ryssens et al. Eur. Phys. J. A 59, 270 (2023) [5] G. Grams, N. Shchechilin et al. arXiv preprint 2411.08007 (2024) [6] D. Davesne, P. Becker et al., Acta Physica Polonica B 48, 3 (2017) [7] L. Batail, D. Davesne et al., The European Physical Journal A, (2023) [8] L. Batail, S. Goriely, S. Peru, S. Hilaire, D. Davesne, A. Pastore submitted 2024 [9] M. Wang et al., Chinese Phys. C 45 030003 (2021) [10] S. Goriely, S. Hilaire, M. Girod, S. Peru, Phys. Rev. Letters 102, 242501 (2009).
The 3He(alpha,gamma)7Be radiative capture reaction plays a key role in the creation of elements in stars as well as in the production of solar neutrinos, the observation of which is one of the main tools to study the properties of our sun. Since accurate experimental measurements of this fusion cross section at solar energies are difficult due to the strong Coulomb repulsion between the reactants, the onus falls on theory to provide a robust means for extrapolating from the region where experimental data is available down to the desired astrophysical regime. I will present the first microscopic calculations of 3He(alpha,gamma)7Be with explicit inclusion of three-nucleon forces. Our prediction of the astrophysical S factor qualitatively agrees with experimental data. We further incorporate experimental bound-state and scattering data in our calculation to better understand the origin of the remaining difference. This process reveals that there is insufficient repulsion in the 1/2+ channel of our model space to reproduce elastic-scattering data. This deficit suggests that 3He(alpha,gamma)7Be probes aspects of the nuclear force that are not currently well-constrained.
In recent years, significant progress has been made in ab initio calculations for nuclear structure. Most of them are restricted to relative light nuclei. Studies of medium-heavy and heavy nuclei are based on nuclear density functional theory. Very successful relativistic and non-relativistic functionals are available nowadays. However, most of them are phenomenological functionals. Therefore, studying the connection of such functionals to ab initio nucleon-nucleon forces is essential. Non-relativistic Brueckner-Hartree-Fock theory was a starting point of ab initio investigations in nuclear structure in the fifties and sixties. It failed because three-body forces were not included at that time. Later, it was found that the relativistic Brueckner-Hartree-Fock (RBHF) theory can reproduce the saturation properties of nuclear matter. In this contribution, we discuss recent developments in RBHF theory for infinite nuclear matter, finite nuclei, and applications for neutron stars.
We recently carried out a systematic calculation for beta decay half-lives within proton-neutron QRPA. Our formalism adopted a Gogny type forces for both isospin T=0 and T=1 pairing channels.
This ensures us a reliable for beta-decay calculations within the QRPA. Delayed-neutron branching ratios are also estimated by a Hauser-Feshbach statistical model with beta-strength functions calculated by the QRPA. We will discuss the results comparing with
those of different models.
The Gogny interaction has never been fully exploited for astrophysical calculations due to its typical poor isovector properties. Over the years several attempts have been made to improve these features, but only recently some success has been achieved. In my talk, I will discuss several properties of infinite nuclear matter with strong isospin asymmetry toillustrate how different Gogny interactions do perform in such a system. I will show that the most recent parametrisations are compatible with a fair reproduction of both systems: nuclei and neutron stars thus solving a long standing question related to the limitations of a finite range interaction.
Nuclear fission occurs when a nucleus splits into smaller nuclei, re-
leasing a significant amount of energy. Although nuclear fission was dis-
covered more than seventy years ago, accurately predicting its behavior
based on the basic constituents of nuclei remains challenging due to the
extremely high dimensionality of the quantum space involving numer-
ous particles. Hence, an approximation scheme is necessary to describe
fissioning systems and simplify their complex nature into a more man-
ageable form. The nuclear density functional theory is such a framework
that can predict nuclear properties for most elements on the nuclear
chart. However, it is still limited by computational constraints. As a
result, most fully microscopic implementations have only considered two
collective degrees of freedom, such as the elongation and the asymme-
try of the fissioning system. Recently, we enhanced this approach by
incorporating a third degree of freedom. This presentation explores our
improvements to the theory and discusses the results we obtained with
it.
Atomic nuclei are complex many-body systems with a number of constituents ranging from very few to several hundreds. Among the difficult aspects, nuclei are self-bound systems that require the treatment of a continuum of wave functions in the Hilbert space. The nuclear strong interaction is scarcely known and highly non-perturbative, with the onset of multi-body interaction beyond the usual interaction of particles two-by-two. Nuclear physics also belongs to problems that face the exponential growth of the Hilbert space when the number of constituents increases. For these reasons, the exact treatment of these systems on classical computers, starting from the interaction, is still restricted to a few percent of the nuclear chart. Quantum technologies and associated quantum algorithms appear in this context as disruptive technologies that might surpass the current limitations in the coming years. I have recently initiated a long-term project to explore using quantum computers and quantum information for nuclear physics and related many-body problems. Inspired by strategies used in classical computing, several novel approaches have been proposed to obtain the ground or low-lying states in many-body systems. A significant effort has been made to use the symmetry breaking/symmetry restoration method. Based on the use of projectors through phase estimation, quantum oracles, or classical shadow, the Quantum Variation After Variation was formulated. Several methods were proposed to access excited states, including the Quantum Krylov, Quantum equation of motion, or Quantum Generator Coordinate Method. After reviewing and illustrating these methods, the current status and future challenges in using quantum computers for atomic nuclei will be discussed.
We present quantum algorithms for two appoaches to nuclear structure: Firstly, the quantum imaginary time approach to state preparation which we use to produce solutions to nuclear density functional theory in the case of simplified Skyrme interactions: Secondly, a variational quantum eigensolver approach to the nucelar shell model, where we make use of particular physics-inspired ansatz choices to target ground and low-lying excited states in Ni-58. Simulated quantum calculations are presented, along with discussion of possible implementation on real hardware, and future perspectives on large-scale calculations.
Digital and analog Quantum Computing provide in perspective an exceptional natural tool to solve time-dependent quantum many-body problems. A straightforward application might be that of predicting the outcome of nuclear reactions in a completely non-perturbative way once the nucleon-nucleon interaction Hamiltonian is given. In this talk some first attempt to implement, at least this scheme in a partial way and on test cases will be illustrated. In particular the potential achievements will be presented together with the current limitations and some strategies that can be used to go towards the solution of problems of interest.
Correlations present in the ground and excited states of nuclear systems are analyzed in terms of quantum information tools like quantum discord or various types of entropies. To establish the link several exactly soluble models are explored. Some preliminary results using more sophisticated setups will be presented
The last decade has put artificial intelligence in the spotlight both in science and in our daily life. This boom takes its origin from the wealth of recently designed machine learning algorithms such as generative adversarial networks (2014) or transformers (2017). An other important key to its success is its large accessibility through open source libraries such as TensorFlow and PyTorch steamed by the increasing power and availability of GPUs. A natural question in such a context is: Can we import techniques or ideas from the field of artificial intelligence to nuclear physics or the other way around ?
In this talk I will survey the main applications of machine learning to nuclear experiments and theory. I will try to emphasize which of these techniques are mature for applications, proof of principle, or ideas for future explorations.
The structure of low-nuclear density nuclear matter is of great importance to the physics of neutron star crusts. One of the most significant aspects of this structure is the transition from roughly spherical neutron rich nuclei to uniform matter. Models for both of these extremes exist but the transition is less easily understood. In this presentation I will discuss my results from variational Monte Carlo calculations using neural-network quantum states which show that this method can model nuclear matter in this density region efficiently and more accurately than other Monte Carlo methods. The results to be shown come from calculations at several densities and proton fractions using a pionless effective field theory Hamiltonian. From these results I will show predictions for clustering, symmetry energy, and proton fraction for the beta-equilibrated ground state.
Strongly correlated quantum many-body systems pose great computational challenges throughout a plethora of research fields, including quantum chemistry, condensed matter, atomic, and nuclear physics. The difficulty arises from the exponential scaling of the Hilbert space with the total number of single-particle degrees of freedom, e.g. spin, orbital, lattice site, depending on the specific system. Numerical accurate solutions can be obtained by configuration interaction (CI) methods, which express the fully interacting wave function of fermionic systems as a linear combination of Slater determinant basis states and compute the coefficients of this expansion by solving the Hamiltonian eigenvalue problem. Unfortunately, CI becomes computationally impractical already for small systems. Here, we develop an active learning approach to iteratively select the most relevant Slater determinants and construct an approximate wave function for a multiconfigurational Dirac-Hartree-Fock atomic structure Hamiltonian [1] or for effective second quantization model Hamiltonians describing quantum many-body systems and quantum clusters in the realm of solid-state physics [2]. The large CI computation is thereby replaced by a series of smaller ones performed on an iteratively expanding basis subset managed by a neural network. This deep learning approach to selected CI allows to reduce computational complexity and successfully address larger many-body systems.
References:
[1] P. Bilous, A. Palffy, and F. Marquardt, Phys. Rev. Lett.131,133002 (2023)
[2] P. Bilous, L. Thirion. H. Mecke, M. Haverkort, A. Palffy and P. Hansmann, arXiv:2406.00151 (2024)
Nuclear astrophysics aims at describing the nuclear properties occurring in and powering astrophysical objects, as well as the cosmic origin of chemical elements found in the Universe. During this talk, I will present some recent progress and future challenges in the field of nuclear data for astrophysics, with a particular focus on the nucleosynthesis of heavy elements.
The phenomenology of atomic nuclei is bewildering: many thousands of different isotopes exist, each of which can differ dramatically from systems with just one less nucleon. This diversity makes nuclear structure interesting by itself, but the field is also crucial to progress in other domains: from precision searches for beyond-standard-model physics to the chemistry of superheavy nuclei. The need for nuclear data is particularly large in astrophysics: simulations of phenomena such as nucleosynthesis or neutron star mergers rely on the properties of dense matter at the extremes of isospin, density and temperature. Models based on nuclear energy density functionals (EDFs) represent our current best hope to provide this data; such approach describes a nucleus in terms of its constituent nucleons while the equations remain sufficiently tractable for global application at the cost of introducing a modest amount of parameters. We are building a new class of EDF-based models aimed at providing all necessary data to astrophysical applications: the Brussels-Skyrme-on-a-Grid or BSkG-series [1, 2, 3]. These models all achieve a description of thousands of masses and several hundred charge radii with a root-mean-square error of less than 800 keV and 0.03 fm, respectively. These models accord the nucleus an extreme amount of freedom: nuclear shapes range from spheres and axially symmetric ellipsoids but also exhibit triaxial deformation, reflection asymmetry, non-zero angular momentum or all of these combined! The BSkG-series thus leverages spontaneous symmetry breaking as much as possible, which makes bench- marks with many types of new experimental data possible across the entire nuclear chart [5, 6, 7]. These models do not just describe bulk properties such as masses and radii, but their reach also extends to pseudo-observables that serve as input to reaction models, predictions for dense matter in neutron stars and fission properties; for the latter in particular, the BSkG models offer an accuracy of less than 0.5 MeV [4, 3] w.r.t. the available empirical fission barriers of actinide nuclei. In this contribution, I will start by discussing the BSkG series of models and discuss some key points on how spontaneous symmetry breaking helps us improve our global description of the properties of nuclei. After this more general introduction, I will dig deeper into the properties of the latest model: BSkG5 is as performant as the earlier entries in the series, but is based on a so-called N2LO form of the Skyrme EDF [8]. While involved and complicated to use numerically, this analytical form has several advantages. First, it allows a zero-range Skyrme to incorporate (part of) the physics of finite range interactions. Second, it gives us the freedom to combine an excellent description of nuclear properties with a stiff equation of state that can support the existence of heavy neutron stars without having to rely on additional density dependencies. Finally, all of these attractive properties can be achieved with less parameters than the preceding models.
Nuclear reaction models, implemented in nuclear reactions codes, require the knowledge of several inputs such as masses, nuclear levels and level densities, optical models, photon strength functions, and sometimes fission paths. For decades, analytical expressions have been used in nuclear reaction codes, due to the freedom they offer to the user to modify their associated parameters in order to fit cross sections. The development of computational resources has opened a new era, roughly 20 years ago, by allowing the systematic calculation of these ingredients from microscopic approaches and their use through tables stored in databases. During this period, several approaches have been developed to improve step by step the physical content of these models. We will review these efforts, focussing in particular on those performed using the Gogny force as basic input, and will show where we are now and what we foresee as future improvements.
The effective Gogny interactions of the D1 family were established by D. Gogny more than forty years ago to describe simultaneously the mean field and the pairing field corresponding to the nuclear interaction. The most popular Gogny parametrizations, namely D1S, D1N, and D1M, describe accurately the ground-state properties of spherical and deformed finite nuclei all across the mass table obtained with Hartree-Fock-Bogoliubov (HFB) calculations. However, these forces produce a rather soft equation of state (EoS) in neutron matter, which leads to predicting maximum masses of neutron stars well below the observed value of two solar masses. To get rid of this limitation, we have built new Gogny parametrizations by modifying the density dependence of the symmetry energy predicted by the force in such a way that they can be applied to the neutron star domain and can also reproduce the properties of finite nuclei as good as their predecessors. These new parametrizations allow us to obtain stiffer EoSs based on the Gogny interactions, which predict maximum masses of neutron stars around two solar masses. Moreover, other global properties of the star, such as the moment of inertia and the tidal deformability, are in harmony with those obtained with other well tested EoSs based on the SLy4 Skyrme force or the Barcelona-Catania-Paris-Madrid (BCPM) energy density functional. Properties of the core-crust transition predicted by these Gogny EoSs are also analyzed. Using these new Gogny forces, the EoS in the inner crust is obtained with the Wigner-Seitz approximation in the Variational Wigner-Kirkwood approach along with the Strutinsky integral method, which allows one to estimate in a perturbative way the proton shell and pairing corrections. For the outer crust, the EoS is determined basically by the nuclear masses, which are taken from the experiments, wherever they are available, or by HFB calculations performed with these new forces if the experimental masses are not known.
beta-decay rates are fundamental to understanding r-process nucleosynthesis, which is responsible for producing roughly half of the heavy elements. Existing theoretical global calculations of the rates use either Skyrme or relativistic quasiparticle random phase approximation (QRPA). These models yield very different predictions and are limited due to their treatment of nuclear many-body correlations. Many-body correlations are known to determine the low-lying beta-decay strength and consequently the decay half-lives due to their strong sensitivity to phase space. In this talk, I address the inclusion of deformation and the coupling of quasiparticles to like-particle phonons within relativistic QRPA linear response theory. The impact on the beta strength and half-lives will be discussed together with the competition of Gamow-Teller and forbidden transitions.
The Musée d'Orsay is within walking distance from the conference venue. However, if you prefer, you can take the RER C train and get off at the next station, Invalides. Entrance to the museum will proceed through the group entrance, where the organizers will meet everyone at the designated time.
The dinner location is accessible directly from within the Musée d'Orsay.
For your information, you can find details about the restaurant here:
Musiam - Le Restaurant
Recently, new measurements of the fission fragments’ spin showed no correlations between the fragments’ spin. These results have stimulated extensive theoretical discussion about the generation, orientation, and correlation of the fission fragments spin. In this contribution, I will discuss several approaches microscopical and collective to describe the mechanisms responsible for the angular momentum at scission. Although we currently have experimental data about the correlation between the magnitude of the angular momenta of the fragments, the presence of correlation between their direction described by the opening angle distribution is subject to different predictions from various theories. I will show how quantal effects and geometry of the scission configuration can change the opening angle distribution.
First of all, we will present new methods using constraints on overlaps and leading to continuous and regular, adiabatic and excited 1D PES, within the framework of constrained HFB theory. Next, we'll present some of the results that these new PES allow us to study, first at the static level, then by performing dynamics including intrinsic excitations using the SCIM formalism. Results include odd-even proton staggering, in-depth analysis of chemical potentials in deformation, and dynamical evaluation of TKE and TXE.
The scission configuration is intuitively defined in the moment of the fission process when the mother nucleus breaks into two nascent fragments. In the continuous fission process, this configuration may be defined in various ways depending on the applied model. Nevertheless, as many observables are fixed at this point, the definition is important for properly reproducing the experimental data. In the constrained self-consistent calculations, the discontinuity on the potential energy surface exists at scission. It is the so-called scission cliff: a rapid change of energy that accompanies the neck rupture. With the help of the neck thickness parameter, the scission may be described as a continuous process of disappearing the neck. We will describe changes in the distribution of the nuclear matter during the neck rupture and its impact on the scission point.
In 2016, the first simulation of a compound nucleus’ evolution from the outer saddle point to scission, which is the most rapid and highly non-equilibrium stage of fission, to two fully separated fission fragments (FFs) was achieved for realistic initial conditions. This was done in the framework of the time-dependent superfluid local density approximation (TDSLDA) or equivalently time-dependent density functional theory extended to superfluid systems. Since the first study of the fission of 240Pu, TDSLDA has been applied extensively to compound systems 236U, 240Pu, and 252Cf (spontaneous fission), with extensive investigations into energy sharing, the FF intrinsic spins and their correlations, scission neutrons and neck rupture dynamics, entanglement, and the evolution of entropy. The most significant results were demonstrating the essential nature of pairing in fission, and theoretically proving that the dynamics from saddle to scission are overdamped. Very recently this effort has been extended to odd-even nuclei 241Pu, 243Pu, and 239U, and an odd-odd nucleus 238Np. This is the first time the fission dynamics of odd systems have been investigated, beyond the adiabatic approximation, which is inconsistent with the dissipative evolution from saddle to scission. Nuclei with odd number of either protons or neutrons comprise of the majority of nuclear systems and are expected to behave qualitatively different from their even-even counter parts, such as having longer fission times in the case of spontaneous fission. Here I will present the results of our investigations, including the dependence on FF masses, charges, excitation energies, and total kinetic energy (TKE) on spins and excitations of quasiparticles flipped. I will also cover the applicability of the Pauli blocking approximation, taking into account the time-reversal symmetry breaking terms in the evolution Hamiltonian, the redistribution of the occupation probabilities during fission, and the importance of using the full set of the quasi-particle states.
Usually fission is described in terms of the evolution of the nuclear shape from the ground state of the nucleus up to the pre-scission configuration. In the microscopic picture, the energy dependence of shape evolution can be studied within the Hartree- Fock-Bogoliubov (HFB) mean field theory. To do so, one needs to arbitrary choose constraints on deformation parameters, associated (most often) with the "geometrical" degrees of freedom, like various multipolar moments or necking. In order to determine fission observables like spontaneous fission lifetimes, one needs to perform dynamic calculations that involve the evaluation of the action integral. This quantity is governed by the interplay between the potential energy landscape and the behaviour of collective inertias. Similar approach, although in a slightly different framework can be applied to determine fission fragments mass yield. Since the HFB equations are numerically expensive, one needs to find a reasonable balance between a proper choice of the most relevant degrees of freedom and computational time. The choice of the right set of degrees of freedom in the theoretical description of fission still remains one of the major challenges for contemporary nuclear structure physics [1,2,3]. Obviously quadrupole and octupole degrees of freedom are essential to describe nuclear evolution towards fission. However one should remember also about the non-negligible role of pairing correlations in fission dynamics which is a well known fact since many decades. It has been shown that an increase in the pairing gap parameter (simple, realistic model), leads to an increase in the penetrability of the fission barrier [4]. The main message from those studies is that there is a strong interplay between the potential energy which increases as a square of the pairing gap and the collective inertia - decreasing as an inverse of the pairing gap squared. As a consequence the least action path leads through regions with a larger pairing gap than the least energy path. There are several ways of including constraints associated with pairing correlations into the fission picture. We will present three options based on considering a pairing gap, the particle number fluctuations and pairing strength. Their impact on the fission description will be discussed.
References:. [1] A. Zdeb, M. Warda, L.M. Robledo, Phys. Rev. C104 014610 (2021). [2] N. Schunck, L.M. Robledo, Rep. Prog. Phys. 79 116301 (2016). [3] H.J. Krappe and K. Pomorski, in Theory of Nuclear Fission, Lectures Notes in Physics Vol. 838 (Springer, Berlin, 2012), p. 207. [4] L.G. Moretto and R.P. Babinet, Phys. Lett. B49 147 (1974).
One of the most challenging problems in nuclear physics is to describe nuclear fission microscopically starting from nucleonic degrees of freedom. Such microscopic description is important particularly for low-energy induced fission, in which the excitation energy of the compound nucleus is relatively low so that an application of the statistical model may be questionable. This includes fission in r-process nucleosynthesis as well as in barrier-top fission. Here we shall discuss our recent attempts with a configuration-interaction approach, for which many-body configurations are constructed based on a constrained density functional theory for shapes along a fission path. We apply this approach to a barrier-top fission of 236U, restricting the model space to seniority zero configurations of neutrons and protons. We find that fission dynamics closely follows that described by the transition state theory, that is the insensitivity of a fission width to post-barrier dynamics as well as a small number of degree of freedom for the distribution of the fission probability. We shall also discuss the sensitivity of the results to a choice of energy functional, such as Skyrme or Gogny functionals, as well as a possible strategy towards more consistent calculations with non-zero seniority configurations.
The task of modelling the dynamics of a fissioning nucleus in order to extract observable quantities with sufficient accuracy to verify and predict experimental outcomes is an ongoing challenge. One category of solutions to this problem is based on a "beyond mean-field" approach utilising the time-dependent generator coordinate method (TDGCM), which simulates the dynamics of a fissioning nucleus as it propagates across a potential energy surface determined by self-consistent solutions of the Hartree-Fock (or similar) equations. The majority of past models in this category have used the Gaussian overlap approximation (GOA) to simplify computations and produce fission observables. It is argued that the inherent assumptions of the TDGCM+GOA approach may limit its suitability for future descriptions of fission aiming to achieve greater precision or to account for previously neglected physics. This motivates a new implementation of the TDGCM which is not only able to simulate fission dynamics without relying on the GOA, but also obtain the desired fission observables up to an arbitrary order of precision. Some preliminary results from applying this model to the well-studied induced fission of 236U are presented and interpreted, and directions for future development are highlighted. Finally, potential extensions to improve the broader TDGCM approach to nuclear fission are discussed.
In recent years, the study of angular momentum in fission fragments has undergone a renaissance. Both theoretical and experimental advancements have reignited the long-standing debate on several unresolved questions, including the generation mechanism of angular momentum in fragments, its mass dependence, and correlations between fragments. One open question is how to use microscopic theory to predict angular momentum distributions across the full range of fragment charges and masses. In this talk, we will present preliminary results obtained by combining symmetry restoration techniques with the time-dependent generator coordinate method.