-
Generative deep-learning reveals collective variables of Fermionic systems
Authors:
Raphaël-David Lasseri,
David Regnier,
Mikaël Frosini,
Marc Verriere,
Nicolas Schunck
Abstract:
Complex processes ranging from protein folding to nuclear fission often follow a low-dimension reaction path parameterized in terms of a few collective variables. In nuclear theory, variables related to the shape of the nuclear density in a mean-field picture are key to describing the large amplitude collective motion of the neutrons and protons. Exploring the adiabatic energy landscape spanned by…
▽ More
Complex processes ranging from protein folding to nuclear fission often follow a low-dimension reaction path parameterized in terms of a few collective variables. In nuclear theory, variables related to the shape of the nuclear density in a mean-field picture are key to describing the large amplitude collective motion of the neutrons and protons. Exploring the adiabatic energy landscape spanned by these degrees of freedom reveals the possible reaction channels while simulating the dynamics in this reduced space yields their respective probabilities. Unfortunately, this theoretical framework breaks down whenever the systems encounters a quantum phase transition with respect to the collective variables. Here we propose a generative-deep-learning algorithm capable of building new collective variables highly representative of a nuclear process while ensuring a differentiable map** to its Fermionic wave function. Within this collective space, the nucleus can evolve continuously from one of its adiabatic quantum phase to the other at the price of crossing a potential energy barrier. This approach applies to any Fermionic system described by a single Slater determinant, which encompasses electronic systems described within the density functional theory.
△ Less
Submitted 14 June, 2023;
originally announced June 2023.
-
Building Surrogate Models of Nuclear Density Functional Theory with Gaussian Processesand Autoencoders
Authors:
Marc Verriere,
Nicolas Schunck,
Irene Kim,
Petar Marević,
Kevin Quinlan,
Michelle N. NGo,
David Regnier,
Raphael David Lasseri
Abstract:
From the lightest Hydrogen isotopes up to the recently synthesized Oganesson (Z=118), it is estimated that as many as about 3000 atomic nuclei could exist in nature. Most of these nuclei are too short-lived to be occurring on Earth, but they play an essential role in astrophysical events such as supernova explosions or neutron star mergers that are presumed to be at the origin of most heavy elemen…
▽ More
From the lightest Hydrogen isotopes up to the recently synthesized Oganesson (Z=118), it is estimated that as many as about 3000 atomic nuclei could exist in nature. Most of these nuclei are too short-lived to be occurring on Earth, but they play an essential role in astrophysical events such as supernova explosions or neutron star mergers that are presumed to be at the origin of most heavy elements in the Universe. Understanding the structure, reactions, and decays of nuclei across the entire chart of nuclides is an enormous challenge because of the experimental difficulties in measuring properties of interest in such fleeting objects and the theoretical and computational issues of simulating strongly-interacting quantum many-body systems. Nuclear density functional theory (DFT) is a fully microscopic theoretical framework which has the potential of providing such a quantitatively accurate description of nuclear properties for every nucleus in the chart of nuclides. Thanks to high-performance computing facilities, it has already been successfully applied to predict nuclear masses, global patterns of radioactive decay like $β$ or $γ$ decay, and several aspects of the nuclear fission process such as, e.g., spontaneous fission half-lives. Yet, predictive simulations of nuclear spectroscopy or of nuclear fission, or the quantification of theoretical uncertainties and their propagation to applications, would require several orders of magnitude more calculations than currently possible. However, most of this computational effort would be spent into generating a suitable basis of DFT wavefunctions. Such a task could potentially be considerably accelerated by borrowing tools from the field of machine learning and artificial intelligence. In this paper, we review different approaches to applying supervised and unsupervised learning techniques to nuclear DFT.
△ Less
Submitted 6 September, 2022;
originally announced September 2022.
-
Microscopic description of the self-conjugate $^{108}$Xe and $^{104}$Te $α$-decay chain
Authors:
F. Mercier,
J. Zhao,
R. -D Lasseri,
J. -P. Ebran,
E. Khan,
T. Niksic,
D. Vretenar
Abstract:
A microscopic calculation of half-lives for the recently observed $^{108}$Xe $\to$ $^{104}$Te $\to$ $^{100}$Sn $α$-decay chain is performed using a self-consistent framework based on energy density functionals. The relativistic density functional DD-PC1 and a separable pairing interaction of finite range are used to compute axially-symmetric deformation energy surfaces of $^{104}$Te and $^{108}$Xe…
▽ More
A microscopic calculation of half-lives for the recently observed $^{108}$Xe $\to$ $^{104}$Te $\to$ $^{100}$Sn $α$-decay chain is performed using a self-consistent framework based on energy density functionals. The relativistic density functional DD-PC1 and a separable pairing interaction of finite range are used to compute axially-symmetric deformation energy surfaces of $^{104}$Te and $^{108}$Xe as functions of quadrupole, octupole and hexadecupole collective coordinates. Dynamic least-action paths are determined that trace the $α$-particle emission from the equilibrium deformation to the point of scission. The calculated half-lives: 197 ns for $^{104}$Te and 50 $μ$s for $^{108}$Xe, are compared to recent experimental values of the half-lives of superallowed $α$-decay of $^{104}$Te: $< 18$ ns, and $^{108}$Xe: 58$^{+106}_{-23}$ $μ$s.
△ Less
Submitted 2 March, 2020;
originally announced March 2020.
-
Localisation and alpha radioactivity
Authors:
J. -P. Ebran,
E. Khan,
R. -D. Lasseri
Abstract:
Relativistic energy density functional approaches are known to well describe nuclear states which involve alpha clusters. Here, alpha emitting nuclei are analysed through the behavior of the spatial localisation of nucleonic states, calculated with an axially deformed RHB approach over the nuclear chart. The systematic occurrence of more localised valence states, having a n = 1 radial quantum numb…
▽ More
Relativistic energy density functional approaches are known to well describe nuclear states which involve alpha clusters. Here, alpha emitting nuclei are analysed through the behavior of the spatial localisation of nucleonic states, calculated with an axially deformed RHB approach over the nuclear chart. The systematic occurrence of more localised valence states, having a n = 1 radial quantum number, allows to pinpoint nuclei in agreement with experimentally known alpha-emitters. The cases of 212Po and 104Te are investigated, showing the concomitant contributions of the pseudospin symmetry and the presence of n = 1 states, on the alpha preformation probability. The impact of the localisation of valence states, on alpha preformation probability, is then analysed. It allows to study shell effects on this probability, over isotopic and isotonic chains. Finally, a phenomenological law is also provided, relating this probability to the radial quantum number of the valence states.
△ Less
Submitted 27 November, 2020; v1 submitted 21 January, 2020;
originally announced January 2020.
-
Alpha-particle condensation: a nuclear quantum phase transition
Authors:
J. -P. Ebran,
M. Girod,
E. Khan,
R. D. Lasseri,
P. Schuck
Abstract:
When the density of a nuclear system is decreased, homogeneous states undergo the so-called Mott transition towards clusterised states, e.g. alpha clustering, both in nuclei and in nuclear matter. Here we investigate such a quantum phase transition (QPT) by using microscopic energy density functional (EDF) calculations both with the relativistic and the Gogny approaches on the diluted $^{16}$O nuc…
▽ More
When the density of a nuclear system is decreased, homogeneous states undergo the so-called Mott transition towards clusterised states, e.g. alpha clustering, both in nuclei and in nuclear matter. Here we investigate such a quantum phase transition (QPT) by using microscopic energy density functional (EDF) calculations both with the relativistic and the Gogny approaches on the diluted $^{16}$O nucleus. The evolution of the corresponding single-particle spectrum under dilution is studied, and a Mott-like transition is predicted at about 1/3 of the saturation density. Complementary approaches are used in order to understand this QPT. A study of spatial localisation properties as a function of the density allows to derive a value of the Mott density in agreement with the one obtained by fully microscopic calculations in $^{16}$O and in nuclear matter. Moreover a study of the spontaneous symmetry breaking of the rotational group in $^{16}$O, down to the discrete tetrahedral one, provides further insight on the features displayed by the single-particle spectrum obtained within the EDF approach.The content of the tetrahedrally deformed A-nucleon product state in terms of spherical particle-hole configurations is investigated. Finally a study of quartet condensation and the corresponding macroscopic QPT is undertaken in infinite matter.
△ Less
Submitted 16 December, 2019;
originally announced December 2019.
-
Taming nuclear complexity with a committee of multilayer neural networks
Authors:
R. -D. Lasseri,
D. Regnier,
J. -P. Ebran,
A. Penon
Abstract:
We demonstrate that a committee of deep neural networks is capable of predicting the ground-state and excited energies of more than 1800 atomic nuclei with an accuracy akin to the one achieved by state-of-the-art nuclear energy density functionals (EDFs) and a major speed-up. An active learning strategy is proposed to train this algorithm with a minimal set of 210 nuclei. This approach enables fut…
▽ More
We demonstrate that a committee of deep neural networks is capable of predicting the ground-state and excited energies of more than 1800 atomic nuclei with an accuracy akin to the one achieved by state-of-the-art nuclear energy density functionals (EDFs) and a major speed-up. An active learning strategy is proposed to train this algorithm with a minimal set of 210 nuclei. This approach enables future fast studies of the influence of EDFs parametrizations on structure properties over the whole nuclear chart and suggests that for the first time an artificial intelligence successfully encoded the laws of nuclear deformation.
△ Less
Submitted 25 February, 2020; v1 submitted 9 October, 2019;
originally announced October 2019.
-
ADG: Automated generation and evaluation of many-body diagrams I. Bogoliubov many-body perturbation theory
Authors:
Pierre Arthuis,
Thomas Duguet,
Alexander Tichai,
Raphaël-David Lasseri,
Jean-Paul Ebran
Abstract:
We describe the first version (v1.0.0) of the code ADG that automatically (1) generates all valid Bogoliubov many-body perturbation theory (BMBPT) diagrams and (2) evaluates their algebraic expression to be implemented for numerical applications. This is achieved at any perturbative order $p$ for a Hamiltonian containing both two-body (four-legs) and three-body (six-legs) interactions (vertices).…
▽ More
We describe the first version (v1.0.0) of the code ADG that automatically (1) generates all valid Bogoliubov many-body perturbation theory (BMBPT) diagrams and (2) evaluates their algebraic expression to be implemented for numerical applications. This is achieved at any perturbative order $p$ for a Hamiltonian containing both two-body (four-legs) and three-body (six-legs) interactions (vertices). The automated generation of BMBPT diagrams of order $p$ relies on elements of graph theory, i.e., it is achieved by producing all oriented adjacency matrices of size $(p+1) \times (p+1)$ satisfying topological Feynman's rules. The automated evaluation of BMBPT diagrams of order $p$ relies both on the application of algebraic Feynman's rules and on the identification of a powerful diagrammatic rule providing the result of the remaining $p$-tuple time integral. The diagrammatic rule in question constitutes a novel finding allowing for the straight summation of large classes of time-ordered diagrams at play in the time-independent formulation of BMBPT. Correspondingly, the traditional resolvent rule employed to compute time-ordered diagrams happens to be a particular case of the general rule presently identified. The code ADG is written in Python2.7 and uses the graph manipulation package NetworkX. The code is also able to generate and evaluate Hartree-Fock-MBPT (HF-MBPT) diagrams and is made flexible enough to be expanded throughout the years to tackle the diagrammatics at play in various many-body formalisms that already exist or are yet to be formulated.
△ Less
Submitted 4 September, 2018;
originally announced September 2018.
-
Single-particle spatial dispersion and clusters in nuclei
Authors:
J. -P. Ebran,
E. Khan,
R. -D Lasseri,
D. Vretenar
Abstract:
The spatial dispersion of the single-nucleon wave functions is analyzed using the self-consistent mean-field framework based on nuclear energy density functionals, and with the harmonic oscillator approximation for the nuclear potential. It is shown that the dispersion depends on the radial quantum number n, but displays only a very weak dependence on the orbital angular momentum. An analytic expr…
▽ More
The spatial dispersion of the single-nucleon wave functions is analyzed using the self-consistent mean-field framework based on nuclear energy density functionals, and with the harmonic oscillator approximation for the nuclear potential. It is shown that the dispersion depends on the radial quantum number n, but displays only a very weak dependence on the orbital angular momentum. An analytic expression is derived for the localization parameter that explicitly takes into account the radial quantum number of occupied single-nucleon states. The conditions for single-nucleon localization and formation of cluster structures are fulfilled in relatively light nuclei with $A \leq 30$ and $n=1$ states occupied. Heavier nuclei exhibit the quantum liquid phase of nucleonic matter because occupied levels that originate from $n > 1$ spherical states are largely delocalized. Nevertheless, individual $α$-like clusters can be formed from valence nucleons filling single-particle levels originating from $n=1$ spherical mean-field states.
△ Less
Submitted 14 May, 2018;
originally announced May 2018.
-
Localization of pairing correlations in nuclei within relativistic mean field models
Authors:
R. -D. Lasseri,
J. -P. Ebran,
E. Khan,
N. Sandulescu
Abstract:
We analyze the localization properties of two-body correlations induced by pairing in the framework of relativistic mean field (RMF) models. The spatial properties of two-body correlations are studied for the pairing tensor in coordinate space and for the Cooper pair wave function. The calculations are performed both with Relativistic-Hatree-Bogoliubov (RHB) and RMF+Projected-BCS (PBCS) models and…
▽ More
We analyze the localization properties of two-body correlations induced by pairing in the framework of relativistic mean field (RMF) models. The spatial properties of two-body correlations are studied for the pairing tensor in coordinate space and for the Cooper pair wave function. The calculations are performed both with Relativistic-Hatree-Bogoliubov (RHB) and RMF+Projected-BCS (PBCS) models and taking as examples the nuclei $^{66}$Ni, $^{124}$Sn and $^{200}$Pb. It is shown that the coherence length have the same pattern as in previous non-relativistic HFB calculations, i.e., it is maximum in the interior of the nucleus and drops to a minimum in the surface region. In the framework of RMF+PBCS we have also analysed, for the particular case of $^{120}$Sn, the dependence of the coherence length on the intensity of the pairing force. This analysis indicates that pairing is reducing the coherence length by about 25-30 $\%$ compared to the RMF limit.
△ Less
Submitted 3 May, 2018;
originally announced May 2018.
-
Axially deformed solution of the Skyrme-Hartree-Fock-Bogolyubov equations using the transformed harmonic oscillator basis (III) hfbtho (v3.00): a new version of the program
Authors:
R. Navarro Perez,
N. Schunck,
R. -D. Lasseri,
C. Zhang,
J. Sarich
Abstract:
We describe the new version 3.00 of the code HFBTHO that solves the nuclear Hartree-Fock (HF) or Hartree-Fock-Bogolyubov (HFB) problem by using the cylindrical transformed deformed harmonic oscillator basis. In the new version, we have implemented the following features: (i) the full Gogny force in both particle-hole and particle-particle channels, (ii) the calculation of the nuclear collective in…
▽ More
We describe the new version 3.00 of the code HFBTHO that solves the nuclear Hartree-Fock (HF) or Hartree-Fock-Bogolyubov (HFB) problem by using the cylindrical transformed deformed harmonic oscillator basis. In the new version, we have implemented the following features: (i) the full Gogny force in both particle-hole and particle-particle channels, (ii) the calculation of the nuclear collective inertia at the perturbative cranking approximation, (iii) the calculation of fission fragment charge, mass and deformations based on the determination of the neck (iv) the regularization of zero-range pairing forces (v) the calculation of localization functions (vi)MPI interface for large-scale mass table calculations.
△ Less
Submitted 12 April, 2017;
originally announced April 2017.
-
Lorentz-symmetry test at Planck-scale suppression with nucleons in a spin-polarized $^{133}$Cs cold atom clock
Authors:
H. Pihan-Le Bars,
C. Guerlin,
R. -D. Lasseri,
J. -P. Ebran,
Q. G. Bailey,
S. Bize,
E. Khan,
P. Wolf
Abstract:
We introduce an improved model that links the frequency shift of the $^{133}\text{Cs}$ hyperfine Zeeman transitions $\vert F = 3, m_F> \longleftrightarrow \vert F = 4, m_F >$ to the Lorentz-violating Standard-Model Extension (SME) coefficients of the proton and neutron. The new model uses Lorentz transformations developed to second order in boost and additionally takes the nuclear structure into a…
▽ More
We introduce an improved model that links the frequency shift of the $^{133}\text{Cs}$ hyperfine Zeeman transitions $\vert F = 3, m_F> \longleftrightarrow \vert F = 4, m_F >$ to the Lorentz-violating Standard-Model Extension (SME) coefficients of the proton and neutron. The new model uses Lorentz transformations developed to second order in boost and additionally takes the nuclear structure into account, beyond the simple Schmidt model used previously in SME analyses, thereby providing access to both proton and neutron SME coefficients including the isotropic coefficient $\tilde{c}_{TT}$. Using this new model in a second analysis of the data delivered by the FO2 dual Cs/Rb fountain at Paris Observatory and previously analysed in arXiv:hep-ph/0601024v1, we improve by up to 12 orders of magnitude the present maximum sensitivities (see arXiv:0801.0287v9) on the $\tilde{c}_{Q}$, $\tilde{c}_{TJ}$ and $\tilde{c}_{TT}$ coefficients for the neutron and on the $\tilde{c}_{TT}$ coefficient for the proton, reaching respectively $10^{-20}$, $10^{-17}$, $10^{-13}$ and $10^{-15}$ GeV.
△ Less
Submitted 31 May, 2017; v1 submitted 21 December, 2016;
originally announced December 2016.
