MoirÉ superlattice simulation
A neural network-based deep learning method simulates moiré superlattices, overcoming traditional computational limitations to uncover new quantum phases and patterns in moiré materials, enhancing the understanding of electron correlations.
Patent Information
- Application Number
- PCT/CN2024/087836
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2024-04-15
- Publication Date
- 2025-10-23
AI Technical Summary
Current computational methods for simulating moiré superlattices struggle to accurately capture strong electron correlation effects, limiting the discovery of new quantum phases and phases in moiré materials due to their reliance on traditional approaches like Monte Carlo and Hartree-Fock approximation.
Employing a neural network to determine ground state wave functions and generate moiré patterns based on property information, utilizing a deep learning architecture to simulate moiré superlattices, allowing for the representation of many-body quantum states without traditional ansatz assumptions.
The deep learning approach provides a more accurate and efficient simulation of moiré superlattices, uncovering new quantum phases and patterns, including Mott insulators, generalized Wigner crystals, and Wigner molecular crystals, beyond the limitations of traditional methods.
Smart Images

Figure CN2024087836_23102025_PF_FP_ABST
Abstract
Description
MOIR* SUPERLATTICE SIMULATIONFIELD
[0001] The present disclosure generally relates to the field of computer, and more specifically, to methods, devices, and computer program products for moiré superlattice simulation.BACKGROUND
[0002] Moiré superlattices, spanning from twisted graphene, transition metal dichalcogenides (TMDs) and other systems, have attracted the most significant research interest in condensed matter physics over the past decade. They provide a flexible platform to tune electronic, magnetic and optical properties, and explore strongly correlated and topological phenomena such as superconductivity and correlated insulating behavior.SUMMARY
[0003] In a first aspect of the present disclosure, there is provided a method of moiré superlattice simulation. The method includes: obtaining property information of one or more particles in a moiré superlattice, the property information at least comprising initial position information of the one or more particles; determining, by using a neural network, respective ground state wave functions of the one or more particles based on the property information; and generating a moiré pattern of the moiré superlattice based on the respective ground state wave functions, the moiré pattern representing a particle density distribution across the moiré superlattice.
[0004] In a second aspect of the present disclosure, there is provided an electronic device. The electronic device comprises: a computer processor coupled to a computer-readable memory unit, the memory unit comprising instructions that when executed by the computer processor implements a method according to the first aspect of the present disclosure.
[0005] In a third aspect of the present disclosure, there is provided a computer program product, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by an electronic device to cause the electronic device to perform a method according to the first aspect of the present disclosure.
[0006] This Summary is provided to introduce a selection of concepts in a simplified form that are further described below in the Detailed Description. This Summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used to limit the scope of the claimed subject matter.BRIEF DESCRIPTION OF THE DRAWINGS
[0007] Through the more detailed description of some embodiments of the present disclosure in the accompanying drawings, the above and other objects, features, and advantages of the present disclosure will become more apparent, wherein the same reference generally refers to the same components in the embodiments of the present disclosure.
[0008] FIG. 1 illustrates an example environment in which example embodiments of the present disclosure can be implemented;
[0009] FIG. 2 illustrates an example workflow of moiré superlattice simulation based on deep learning according to some embodiments of the present disclosure;
[0010] FIGS. 3A-3D illustrate example diagrams of simulated density pattern of WSe2 / WS2 hetero-bilayer according to some embodiments of the present disclosure;
[0011] FIGS. 4A-4F illustrate example diagrams of simulated density pattern of WS2 homo-bilayer according to some embodiments of the present disclosure;
[0012] FIG. 5 illustrates an example flowchart of a method of moiré superlattice simulation according to some embodiments of the present disclosure;
[0013] FIG. 6 illustrates a block diagram of an electronic device in which various embodiments of the present disclosure can be implemented.DETAILED DESCRIPTION
[0014] Principle of the present disclosure will now be described with reference to some embodiments. It is to be understood that these embodiments are described only for the purpose of illustration and help those skilled in the art to understand and implement the present disclosure, without suggesting any limitation as to the scope of the disclosure. The disclosure described herein can be implemented in various manners other than the ones described below.
[0015] In the following description and claims, unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skills in the art to which this disclosure belongs.
[0016] References in the present disclosure to “one embodiment, ” “an embodiment, ” “an example embodiment, ” and the like indicate that the embodiment described may include a particular feature, structure, or characteristic, but it is not necessary that every embodiment includes the particular feature, structure, or characteristic. Moreover, such phrases are not necessarily referring to the same embodiment. Further, when a particular feature, structure, or characteristic is described in connection with an example embodiment, it is submitted that it is within the knowledge of one skilled in the art to affect such feature, structure, or characteristic in connection with other embodiments whether or not explicitly described.
[0017] It shall be understood that although the terms “first” and “second” etc. may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another. For example, a first element could be termed a second element, and similarly, a second element could be termed a first element, without departing from the scope of example embodiments. As used herein, the term “and / or” includes any and all combinations of one or more of the listed terms.
[0018] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of example embodiments. As used herein, the singular forms “a” , “an” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms “comprises” , “comprising” , “has” , “having” , “includes” and / or “including” , when used herein, specify the presence of stated features, elements, and / or components etc., but do not preclude the presence or addition of one or more other features, elements, components and / or combinations thereof.
[0019] Principle of the present disclosure will now be described with reference to some embodiments. It is to be understood that these embodiments are described only for the purpose of illustration and help those skilled in the art to understand and implement the present disclosure, without suggesting any limitation as to the scope of the disclosure. The disclosure described herein can be implemented in various manners other than the ones described below. In the following description and claims, unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skills in the art to which this disclosure belongs.
[0020] It may be understood that data involved in the present technical solution (including but not limited to the data itself, the acquisition or use of the data) should comply with requirements of corresponding laws and regulations and relevant rules.
[0021] It may be understood that, before using the technical solutions disclosed in various embodiment of the present disclosure, the user should be informed of the type, scope of use, and use scenario of the information involved in the present disclosure in an appropriate manner in accordance with relevant laws and regulations, and the user’s authorization should be obtained.
[0022] For example, in response to receiving an active request from the user, prompt information is sent to the user to explicitly inform the user that the requested operation will need to acquire and use the user’s information. Therefore, the user may independently choose, according to the prompt information, whether to provide the information to software or hardware such as electronic devices, applications, servers, or storage media that perform operations of the technical solutions of the present disclosure.
[0023] As an optional but non-limiting implementation, in response to receiving an active request from the user, the way of sending prompt information to the user, for example, may include a pop-up window, and the prompt information may be presented in the form of text in the pop-up window. In addition, the pop-up window may also carry a selection control for the user to choose “agree” or “disagree” to provide the information to the electronic device.
[0024] It may be understood that the above process of notifying and obtaining the user authorization is only illustrative and does not limit the implementation of the present disclosure. Other methods that satisfy relevant laws and regulations are also applicable to the implementation of the present disclosure.
[0025] As used herein, the term “model” is referred to as an association between an input and an output learned from training data, and thus a corresponding output may be generated for a given input after the training. The generation of the model may be based on a machine learning technique. In general, a machine learning model may be built, which receives input information and makes predictions based on the input information. For example, a classification model may predict a class of the input information among a predetermined set of classes. As used herein, “model” may also be referred to as “machine learning model” , “learning model” , “machine learning network” , or “learning network, ” which are used interchangeably herein.
[0026] Example environment
[0027] FIG. 1 illustrates a block diagram of an example environment 100 in which various embodiments of the present disclosure may be implemented. In the environment 100 of FIG. 1, an electronic device 120 includes or is deployed with a simulation system 110. The simulation system 110 is configured to predict a moiré pattern 102 of a moiré superlattice 101. The moiré pattern 102 may represent a particle density distribution across the moiré superlattice 101. As used herein, the moiré pattern may be also referred to as a density pattern. In embodiments of the present disclosure, a particle may be an electron or a hole.
[0028] The moiré superlattice may include two or more layers of two-dimensional material. In some embodiments, the moiré superlattice 101 may include two layers of the same material, and thus may be referred to as a hetero-bilayer structure, for example, WSe2 / WS2 hetero-bilayer. In some embodiment, the moiré superlattice 101 may include two layers of the different materials, and thus may be referred to as a homo-bilayer structure, for example, WS2 homo-bilayer. In the following, the moiré superlattice may be also referred to as a moiré material.
[0029] In FIG. 1, the electronic device 110 may be any system with computing power, such as various computing devices / systems, terminal devices, servers, etc. The terminal device can be any type of mobile terminal, fixed terminal, or portable terminal, including mobile phones, desktop computers, laptops, netbooks, tablets, media computers, multimedia tablets, or any combination of the aforementioned, including accessories and peripherals of these devices or any combination thereof. The server may include but are not limited to mainframe, edge computing nodes, computing devices in cloud environment, etc.
[0030] As briefly mentioned above, the moiré superlattices provide a flexible platform to tune electronic, magnetic, and optical properties, and explore strongly correlated and topological phenomena such as superconductivity and correlated insulating behavior. For example, recent discoveries of new quantum phases, including the generalized Wigner crystal, known for their unique symmetry and particle arrangements, have provided deep insights into the behaviors of strongly correlated systems. These insights have been further enriched by Wigner molecular crystals emerging from multi-electron artificial atoms in twisted WS2 homo-bilayer, underscoring the tunable feature of moiré superlattice.
[0031] Theoretically, a grand challenge remains that a general theoretical approach to handle strong electron correlation effects of moiré system is lacking, and the current investigations largely rely on traditional computational methods, such as classic Monte Carlo and Hartree-Fock approximation. While these calculations may provide qualitative explanations to experimental results and make other predictions, it is questionable whether all essential correlated physics has been captured. In recent years, powerful approaches based on deep learning architecture have been developed to treat many-body quantum problems. Utilizing the universality and expressiveness of neural networks, many-body quantum states may be well-represented without assuming a limited traditional ansatz, and new insights to the correlation effects are revealed. However, these deep learning approaches may only be applied to lattice models, chemical molecules, uniform electron gases, small solids, and isolated Wigner molecules.
[0032] Embodiments of the present disclosure propose solutions for moiré superlattice simulation. According to embodiments of the present disclosure, property information of one or more particles in a moiré superlattice is obtained. The property information at least comprises initial position information of the one or more particles. Respective ground state wave functions of the one or more particles based on the property information is determined by using a neural network. A moiré pattern of the moiré superlattice is generated based on the respective ground state wave functions. The moiré pattern represents a particle density distribution across the moiré superlattice.
[0033] In this way, the moiré pattern can be simulated by means of deep learning. In this way, not only experimental research of moiré superlattices can be explained, but also unexplored moiré patterns for further investigation can be discovered.
[0034] Theoretical foundation of the present disclosure and example embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings.
[0035] Deep learning architecture
[0036] Reference is now made to FIG. 2, which illustrates an example workflow 200 of moiré superlattice simulation based on deep learning according to some embodiments of the present disclosure. The workflow 200 may be implemented by the simulation system 110 as shown in FIG. 1. In the example, a moiré superlattice 201 is simulated. Specifically, property information of one or more particles in the moiré superlattice 201 may be obtained. The one or more particles may be one or more holes, or one or more electrons. The property information at least comprises initial position information of the one or more particles. For example, the initial position information may include initial position coordinates of these particles. In some embodiments, the property information may further include spin state information of the one or more particles. These embodiments will be described below in detail.
[0037] A neural network 210 may be used to determine respective ground state wave functions of the one or more particles based on the property information. The neural network 210 may have any suitable structure, and protection scope of the present disclosure is not limited in this regard. In the following, a wave function determined by using a neural network may be also referred to as a “neural network wave function” .
[0038] As an example, to explore the quantum phases of moiré materials, a neural network may be employed to represent the many-body wave function. The neural network wave function ψnet may have the following form:
[0039] where rj represents a position or coordinate of the jth particle, and ki represents a phase factor, “det” represents the Slater determinant. Particle features ri are converted to be periodic and fed into permutation equivalent neural networks. Quasi orbitals are then constructed, which combined with Slater determinant, forming a periodic, complex-valued and anti-symmetry wave function for solids.
[0040] Compared with traditional wave functions, in the Eq. (1) , ψnet explicitly includes the features of each particle (rj, r≠j) in the formulation, permitting it to capture all the correlations. Millions of parameters are embedded in the neural network, rendering it a powerful and convenient tool to express quantum states of distinct symmetries.
[0041] After the ground state wave functions are determined, the simulation system 110 may generate a moiré pattern 102 of the moiré superlattice 101 based on the respective ground state wave functions of the one or more particles. The moiré pattern 202 represents a particle density distribution across the moiré superlattice 201. The moiré pattern 102 may be computed based on the respective ground state wave functions in any suitable manner, and the protection scope of the present disclosure is not limited in this regard.
[0042] Now some example embodiments regarding the determination of the ground state wave functions are described in detail.
[0043] In some embodiments, an optimization process 220 may be applied to update the neural network 210 to achieve ground states. For example, the neural network 210 may be used to generate respective initial wave functions of the one or more particles based on the property information. Then, the neural network 210 (specifically, parameters of the neural network 210) may be updated iteratively based on the respective initial wavefunctions and an energy function for the moiré superlattice. For example, the energy function may be a Hamiltonian of the moiré superlattice, and the optimization process 210 may be based on a Variational Monte Carlo (VMC) method. The respective ground state wave functions may be generated by using the updated neural network 210, or in other words, by using the optimized neural network 210.
[0044] Given the neural network wave function and the Hamiltonian of the moiré superlattice, the VMC method may be employed to efficiently optimize parameters of the neural network. VMC is based on the variational principle of quantum mechanics, i.e. the ground state is the lower bound of the solution to the stationary Schrodinger equation. Hence the VMC training of the neural network is an unsupervised process, further guaranteeing the reliability of our deep learning architecture. As illustrated in FIG. 2, a neural network wave function ψ is built for moiré systems. Variational Monte Carlo is then applied to optimize ψ and reach the ground state.
[0045] In some embodiments, the energy function used in the optimization process 220 may comprise a plurality of energy items. For example, the energy function may include an item associated with a kinetic energy of each of the one or more particles, which is also referred to as a kinetic energy item. For another example, the energy function may include an item associated with a moiré potential of each of the one or more particles, which is also referred to as a moiré potential item. In some embodiments, a moiré potential of a given particles of the one or more particles is based on a position of the given particle in a moiré cell of the moiré superlattice, as will be described below with reference to FIG. 4A and FIG. 4D. For a further example, the energy function may include an item associated with electrostatic interaction (such as Coulomb interaction) between different particles of the one or more particles, which is also referred to as an electrostatic interaction item.
[0046] Moiré superlattice appears when two layers of van der Waals material overlap each other, with mismatched lattice constants and twisted angles. The superlattice formed features a significantly large lattice constant and contains thousands of atoms in the unit cell, which makes it unrealistic to employ the full ab initio Hamiltonian of a moiré material in the current workflow. Therefore, in the present disclosure, the low-energy effective Hamiltonian may be as following:
[0047] where the first item is an example of the kinetic energy item, the second item VM (ri) is an example of the moiré potential item, and the third item is an example of the electrostatic interaction item. This model describes doping holes of these materials near fermi surface and m* refers to the effective mass of valence band edge. VM (ri) is moiré potential felt by holes across the whole moiré materials. bi denotes the reciprocal vector of moiré cell, and V, φ are parameters depending on the material. When moiré materials get twisted to critical angles, flat band appears and Coulomb interaction becomes dominant, which accounts for the last term in Eq. (2) with ∈ referring to effective dielectric constant. A uniform positive charge background is also necessary to make the whole system neutral and remove divergence in Coulomb interactions.
[0048] The effective Hamiltonian may greatly simplify the problem but has included the key feature of Coulomb interactions, so it has been successfully used in studies of TMD hetero-bilayers such as WSe2 / WS2 and Γ-valley homo-bilayer such as WS2. It is worth noting that electron correlations in such an effective Hamiltonian is as non-trivial as the ab initio Hamiltonian, and an exact solution remains extremely difficult. The most prevalent Hartree-Fock method treats particles independent from each other, which would lose the correlation effects. More accurate methods such as the full configuration interaction theory exist, but they are temporally only applicable to small molecules. The deep learning wave function approach proposed by embodiments of the present disclosure achieves an optimal balance of accuracy and efficiency, making it a very promising tool for studying moiré systems.
[0049] In some embodiments, the energy function may be converted or rescaled so as to handle the wide range of length scale in the moiré superlattice. For example, a scaling factor may be employed. The energy function (for example, the Hamiltonian) may be converted from a first unit of length to a second unit of length larger than the first unit of length. The neural network 210 may be updated iteratively based on the respective initial wave functions and the converted energy function.
[0050] In an example, the coordinates may be rescaled as for computational convenience. The Hamiltonian in Eq. (2) may be converted as:
[0051] where aM represents the scaling factor. By rescaling the Hamiltonian, the computational efficiency for the moiré superlattice can be improved.
[0052] Some example embodiments regarding the energy function for optimizing the neural network are described above. Now some example embodiments regarding the updating of the neural network are now described.
[0053] In some embodiments, a system energy of the moiré superlattice 201 may be predicted based on the respective initial wave functions and the energy function for the moiré superlattice. Then, the neural network may be updated iteratively by minimizing the predicted system energy until a predetermined condition is met, for example until an energy change is below a threshold.
[0054] For example, the neural network 210 may be optimized to minimize the rescaled Hamiltonian whose gradient may be as below:
[0055] where <…> denotes the expectations of operators with |ψ|2 distribution. Moreover, Kronecker factored curvature estimator (KFAC) optimizer may be employed which significantly outperforms traditional optimizer in energy minimization.
[0056] Example Wigner crystals
[0057] The proposed solution for moiré superlattice simulation may be used to a moiré superlattice of a particular filling rate of particles in a moiré cell. In some embodiments, to obtain the property information, a number of moiré cells to be simulated and a filling rate of particles in a moiré cell may be determined first. Then, the initial position information may be determined by distributing the one or more particles uniformly over the number of moiré cells according to the filing rate. Examples are now described with reference to Wigner crystals.
[0058] A Wigner crystal may refer to a special crystal which is composed purely of electrons. As the electron gas becomes sparse, the kinetic energy of electrons is dominated by the Coulomb potential, then electrons will spontaneously organize themselves to present periodicity. The seek for Wigner crystal has continued ever since, leading to many important discoveries in condensed matter physics in the past century. In recent years, the seek has been extended to generalized Wigner crystal in moiré materials.
[0059] In some embodiments, the filling rate may be below a predetermined value (for example 1) to indicate one or less than one particle is distributed within a moiré cell.
[0060] FIGS. 3A-3D illustrates example diagrams of calculated electron density patterns of WSe2 / WS2 hetero-bilayer according to some embodiments of the present disclosure. Specifically, in FIG. 3A, a structure 301 of the WSe2 / WS2 superlattice is shown. In FIG. 3B, the moiré potential 302 of WSe2 / WS2 is shown, where aM denotes corresponding length of a moiré cell and is set to 8.3 nm. In FIG. 3C, simulated moiré patterns 311, 312 and 313 with a triangular phase at fillings v = 1, 2 / 3, 1 / 3 are shown. In FIG. 3D, simulated moiré patterns 321, 322 and 323 with a stripe phase at fillings v = 1 / 2, 2 / 5, 1 / 6. Numbers above the figure denote the particle fillings v.
[0061] As an example, the proposed neural network may be employed to simulate WSe2 / WS2 2 hetero-bilayer (as shown in FIG. 3A) at low particle fillings v ≤ 1, and the results are plotted in FIGS. 3C-3D. The results manifest that particles indeed crystallize at low fillings. The first interesting observation is that even at v = 1 it is found to be a Mott insulator phase instead of metal, highlighting the importance of electron correlation in this system. Going towards lower fractional fillings, we can see new phases of generalized Wigner crystal emerging. With v = 1 / 3, 2 / 3, the C3 symmetry of the moiré lattice is preserved, leading to commensurate Wigner crystal phases. The symmetry is found to be broken at fillings of v = 1 / 2, 2 / 5, 1 / 6, where a range of stripe phases appears. The appearance of stripe phases highlights a delicate competition between attractive moiré potential and long-ranged Coulomb repulsion.
[0062] Beyond the aforementioned fractional filling, recent research has been extended to multiple integer fillings of moiré superlattice, wherein multiple particles are assigned to a single moiré cell. These particles aggregate and form a “molecule" , naturally residing at the minimum of the moiré potential. Upon the activation of Coulomb repulsion, these molecules slightly disperse and form the so-called Wigner molecular crystal.
[0063] In some embodiments, if the filling rate is greater than the predetermined value (for example, 1) to indicate that two or more particles are distributed within a same moiré cell, spin states of the two or more particles need to be considered. For example, for each moiré cell of the number of moiré cells, respective spin directions of the two or more particles may be determined based on a spin state of the moiré cell. Then, spin state information may be determined as a portion of the property information based on the respective spin directions of the two or more particles.
[0064] These molecular crystals are recently observed in WSe2 / WS2 hetero-bilayer of 2H configuration, prompting to employ the proposed neural network to explain this observation. FIGS. 4A-4F illustrates example diagrams of calculated density patterns of WS2 homo-bilayer according to some embodiments of the present disclosure. Specifically, tuples above the figure denote the particle filling rates and magnetic number of spin (v, Sz) .
[0065] Quantum states with different fillings v and spin quantum numbers Sz are investigated and the results are plotted in FIGS. 4A-4F.
[0066] In FIG. 4A, a moiré potential 401 of 2H WS2 homo-bilayer is shown, where aM denotes length of moiré cell and is set to 9.8 nm. Circles denote regions AB, BS / S, BW / W with different stack patterns. As described above, moiré potential of a given particles of the one or more particles is based on a position of the given particle in a moiré cell of the moiré superlattice. For example, particles in the regions AB, BS / S, BW / W have corresponding moiré potentials. In FIGS. 4B-4C, calculated ground state patterns 411, 412, 413 and excited state patterns 421, 422, 423 for WS2 at different particle fillings v and magnetic number Sz.
[0067] In FIG. 4D, a moiré potential 402 of 3R WS2 homo-bilayer is shown. Particles in the regions AA, BW / S, BS / W have corresponding moiré potentials. In FIGS. 4E-4F, calculated moiré patterns 431, 432, 433, 441 and 442 in 3R WS2 are shown with different quantum numbers and twisted angles θ.
[0068] At v = 2 fillings, Sz = 0 state is found has the lowest energy in which two spin opposite holes are all placed in AB region. Remarkably, if their spins are aligned parallel, one hole will depart due to Pauli exclusion and transfers to the BW / W region, forming a charge-transfer insulator. A more intricate situation happens at v = 3, in which it is found that the Sz = 1 / 2 state presents a rather similar trimer pattern with Sz = 3 / 2 and a slightly lower energy differs 0.3 meV / hole. Concerning v = 4, Sz = 1 state is identified as the ground state, exhibiting a solid triangle density pattern. It’s also intriguing to note that this triangle will get flipped if a strong magnetic field is activated and make the system fully polarized Sz = 2.
[0069] Up to now, only Wigner molecular crystal of C3 symmetry has been observed in the experiment and more symmetry classes remain to be discovered. In the present disclosure, the proposed network may be employed to explore D6 symmetric Wigner molecular crystal in 3R WS2 homo-bilayer, which differs from 2H configuration by opposite orientations of layers. In 3R WS2, energies of different Bernal regions (BW / Sand BS / W) coincide with each other, leading to its underlying D6 symmetry. The calculated density patterns of different quantum number (v, Sz) are plotted in FIG. 4E. Different Sz states are all calculated which shows same pattern as FIG. 4E, proving the robustness of the findings. It may be seen that at small twisted angle θ = 1.1°, honeycomb crystal forms at v = 2, manifesting the D6 symmetry of the material. As doping more holes to v = 6, nearby holes will exclude each other and form two triangles with opposite orientations. At even larger large filling v = 9, additional doped holes build “glue” between each trimers and form “covalent bond” . The present disclosure tries to find Kagome lattice via reducing the moiré length aM. As the moiré length decreases, the size of Wigner molecule becomes comparable with aM and Kagome lattice finally emerges which is plotted in FIG. 4F. However, it is also found that Kagome lattice appears as an excited state and Sz = 1 / 2 state turns out to be the ground state, which means Kagome lattice will only emerge under strong interactions.
[0070] Moiré materials hosts various novel phenomena, tightly related to strong correlation. Despite the fruitful phase has been discovered, more phases remain untouched due to computational and experimental limitations. In the present disclosure, a deep learning approach is described to discover moiré pattern of different TMD materials. Numerous phases with different symmetry appear in our simulation, proving it to be a general and accurate framework to study moiré materials. With the proposed solution, more moiré patterns can be provided for discover. Moreover, the proposed solution may be extended to other problems, such as multilayer materials, quantum transparent and anomalous Hall effect.
[0071] In the present disclosure, a neural network wave function-based deep learning methodology is developed for moiré systems. The neural network wave function is trained via variational quantum Monte Carlo in an unsupervised manner. Focusing on the family of TMD materials over a wide range of particle fillings and effective moiré potentials, a sequence of intriguing phases may be unveiled, including the Mott insulator, the generalized Wigner crystal, and the Wigner molecular crystal. New stripe phases of generalized Wigner crystal are identified with fractional fillings. At integer fillings, electron correlations lead to a rich diagram of Wigner superlattices, including the recently observed Wigner molecular crystals and other unobserved exotic phases such as the generalized Kagome crystal and the Wigner covalent crystal. The present disclosure underscores the pivotal role of electron correlations captured by deep learning in unlocking and leveraging the quantum phenomena inherent in moiré superlattices.
[0072] Example process and device
[0073] FIG. 5 illustrates a flowchart of a method 500 for moiré superlattice simulation in accordance with some example implementations of the present disclosure. The method 500 may be implemented at the electronic device 120 (for example, the simulation system 110) as illustrated in FIG. 1. At a block 510, property information of one or more particles in a moiré superlattice is obtained. The property information at least comprises initial position information of the one or more particles. At a block 520, respective ground state wave functions of the one or more particles based on the property information is determined by using a neural network. At a block 530, a moiré pattern of the moiré superlattice is generated based on the respective ground state wave functions. The moiré pattern represents a particle density distribution across the moiré superlattice.
[0074] In some embodiments, determining the respective ground state wave functions of the one or more particles comprises: generating, by using the neural network, respective initial wave functions of the one or more particles based on the property information; updating the neural network iteratively based on the respective initial wave functions and an energy function for the moiré superlattice; and generating the respective ground state wave functions by using the updated neural network.
[0075] In some embodiments, updating the neural network iteratively based on the respective initial wave functions and an energy function for the moiré superlattice comprises: predicting a system energy of the moiré superlattice based on the respective initial wave functions and the energy function for the moiré superlattice; and updating the neural network iteratively by minimizing the predicted system energy until a predetermined condition is met.
[0076] In some embodiments, the energy function comprises: an item associated with a kinetic energy of each of the one or more particles, an item associated with a moiré potential of each of the one or more particles, and an item associated with electrostatic interaction between different particles of the one or more particles.
[0077] In some embodiments, a moiré potential of a given particles of the one or more particles is based on a position of the given particle in a moiré cell of the moiré superlattice.
[0078] In some embodiments, the method 500 further comprises: converting, based on a scaling factor, the energy function from a first unit of length to a second unit of length larger than the first unit of length, and wherein the neural network is updated iteratively based on the respective initial wave functions and the converted energy function.
[0079] In some embodiments, obtaining the property information of the one or more particles in the moiré superlattice comprises: determining a number of moiré cells to be simulated and a filling rate of particles in a moiré cell; and determining the initial position information by distributing the one or more particles uniformly over the number of moiré cells according to the filing rate.
[0080] In some embodiments, the filling rate is greater than a predetermined value to indicate that two or more particles are distributed within a same moiré cell, and wherein obtaining the property information of the one or more particles in the moiré superlattice further comprises: determining, for each moiré cell of the number of moiré cells, respective spin directions of the two or more particles based on a spin state of the moiré cell; and determining spin state information as a portion of the property information based on the respective spin directions of the two or more particles.
[0081] In some implementations of the present disclosure, there is provided an electronic device, comprising a computer processor coupled to a computer-readable memory unit, the memory unit comprising instructions that when executed by the computer processor implements a method of multi-agent debate. The method comprises: obtaining property information of one or more particles in a moiré superlattice, the property information at least comprising initial position information of the one or more particles; determining, by using a neural network, respective ground state wave functions of the one or more particles based on the property information; and generating a moiré pattern of the moiré superlattice based on the respective ground state wave functions, the moiré pattern representing a particle density distribution across the moiré superlattice.
[0082] In some embodiments, determining the respective ground state wave functions of the one or more particles comprises: generating, by using the neural network, respective initial wave functions of the one or more particles based on the property information; updating the neural network iteratively based on the respective initial wave functions and an energy function for the moiré superlattice; and generating the respective ground state wave functions by using the updated neural network.
[0083] In some embodiments, updating the neural network iteratively based on the respective initial wave functions and an energy function for the moiré superlattice comprises: predicting the system energy of the moiré superlattice based on the respective initial wave functions and an energy function for the moiré superlattice; and updating the neural network iteratively by minimizing the predicted system energy until a predetermined condition is met.
[0084] In some embodiments, the energy function comprises: an item associated with a kinetic energy of each of the one or more particles, an item associated with a moiré potential of each of the one or more particles, and an item associated with electrostatic interaction between different particles of the one or more particles.
[0085] In some embodiments, a moiré potential of a given particles of the one or more particles is based on a position of the given particle in a moiré cell of the moiré superlattice.
[0086] In some embodiments, the method 500 further comprises: converting, based on a scaling factor, the energy function from a first unit of length to a second unit of length larger than the first unit of length, and wherein the neural network is updated iteratively based on the respective initial wave functions and the converted energy function.
[0087] In some embodiments, obtaining the property information of the one or more particles in the moiré superlattice comprises: determining a number of moiré cells to be simulated and a filling rate of particles in a moiré cell; and determining the initial position information by distributing the one or more particles uniformly over the number of moiré cells according to the filing rate.
[0088] In some embodiments, the filling rate is greater than a predetermined value to indicate that two or more particles are distributed within a same moiré cell, and wherein obtaining the property information of the one or more particles in the moiré superlattice further comprises: determining, for each moiré cell of the number of moiré cells, respective spin directions of the two or more particles based on a spin state of the moiré cell; and determining spin state information as a portion of the property information based on the respective spin directions of the two or more particles.
[0089] In some implementations of the present disclosure, there is provided an electronic device, comprising a computer processor coupled to a computer-readable memory unit, the memory unit comprising instructions that when executed by the computer processor implements a method of multi-agent debate. The method comprises: obtaining property information of one or more particles in a moiré superlattice, the property information at least comprising initial position information of the one or more particles; determining, by using a neural network, respective ground state wave functions of the one or more particles based on the property information; and generating a moiré pattern of the moiré superlattice based on the respective ground state wave functions, the moiré pattern representing a particle density distribution across the moiré superlattice.
[0090] FIG. 6 illustrates a block diagram of an electronic device 600 in which various embodiments of the present disclosure can be implemented. It would be appreciated that the electronic device 600 shown in FIG. 6 is merely for purpose of illustration, without suggesting any limitation to the functions and scopes of the present disclosure in any manner. The electronic device 600 may be used to implement the above method 600. As shown in FIG. 6, the electronic device 600 may be a general-purpose electronic device. The electronic device 600 may at least comprise one or more processors or processing units 610, a memory 620, a storage unit 630, one or more communication units 640, one or more input devices 650, and one or more output devices 660.
[0091] The processing unit 610 may be a physical or virtual processor and can implement various processes based on programs 625 stored in the memory 620. In a multi-processor system, multiple processing units execute computer executable instructions in parallel so as to improve the parallel processing capability of the electronic device 600. The processing unit 610 may also be referred to as a central processing unit (CPU) , a microprocessor, a controller, or a microcontroller.
[0092] The electronic device 600 typically includes various computer storage medium. Such medium can be any medium accessible by the electronic device 600, including, but not limited to, volatile and non-volatile medium, or detachable and non-detachable medium. The memory 620 can be a volatile memory (for example, a register, cache, Random Access Memory (RAM) ) , a non-volatile memory (such as a Read-Only Memory (ROM) , Electrically Erasable Programmable Read-Only Memory (EEPROM) , or a flash memory) , or any combination thereof. The storage unit 630 may be any detachable or non-detachable medium and may include a machine-readable medium such as a memory, flash memory drive, magnetic disk, or another other media, which can be used for storing information and / or data and can be accessed in the electronic device 600.
[0093] The electronic device 600 may further include additional detachable / non-detachable, volatile / non-volatile memory medium. Although not shown in FIG. 6, it is possible to provide a magnetic disk drive for reading from and / or writing into a detachable and non-volatile magnetic disk and an optical disk drive for reading from and / or writing into a detachable non-volatile optical disk. In such cases, each drive may be connected to a bus (not shown) via one or more data medium interfaces.
[0094] The communication unit 640 communicates with a further electronic device via the communication medium. In addition, the functions of the components in the electronic device 600 can be implemented by a single computing cluster or multiple computing machines that can communicate via communication connections. Therefore, the electronic device 600 can operate in a networked environment using a logical connection with one or more other servers, networked personal computers (PCs) or further general network nodes.
[0095] The input device 650 may be one or more of a variety of input devices, such as a mouse, keyboard, tracking ball, voice-input device, and the like. The output device 660 may be one or more of a variety of output devices, such as a display, loudspeaker, printer, and the like. By means of the communication unit 640, the electronic device 600 can further communicate with one or more external devices (not shown) such as the storage devices and display device, with one or more devices enabling the user to interact with the electronic device 600, or any devices (such as a network card, a modem, and the like) enabling the electronic device 600 to communicate with one or more other electronic devices, if required. Such communication can be performed via input / output (I / O) interfaces (not shown) .
[0096] In some embodiments, instead of being integrated in a single device, some, or all components of the electronic device 600 may also be arranged in cloud computing architecture. In the cloud computing architecture, the components may be provided remotely and work together to implement the functionalities described in the present disclosure. In some embodiments, cloud computing provides computing, software, data access and storage service, which will not require end users to be aware of the physical locations or configurations of the systems or hardware providing these services. In various embodiments, the cloud computing provides the services via a wide area network (such as Internet) using suitable protocols. For example, a cloud computing provider provides applications over the wide area network, which can be accessed through a web browser or any other computing components. The software or components of the cloud computing architecture and corresponding data may be stored on a server at a remote position. The computing resources in the cloud computing environment may be merged or distributed at locations in a remote data center. Cloud computing infrastructures may provide the services through a shared data center, though they behave as a single access point for the users. Therefore, the cloud computing architectures may be used to provide the components and functionalities described herein from a service provider at a remote location. Alternatively, they may be provided from a conventional server or installed directly or otherwise on a client device.
[0097] The functionalities described herein can be performed, at least in part, by one or more hardware logic components. For example, and without limitation, illustrative types of hardware logic components that can be used include Field-Programmable Gate Arrays (FPGAs) , Application-specific Integrated Circuits (ASICs) , Application-specific Standard Products (ASSPs) , System-on-a-chip systems (SOCs) , Complex Programmable Logic Devices (CPLDs) , and the like.
[0098] Program code for carrying out the methods of the subject matter described herein may be written in any combination of one or more programming languages. The program code may be provided to a processor or controller of a general-purpose computer, special purpose computer, or other programmable data processing apparatus such that the program code, when executed by the processor or controller, causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely or partly on a machine, executed as a stand-alone software package partly on the machine, partly on a remote machine, or entirely on the remote machine or server.
[0099] In the context of this disclosure, a machine-readable medium may be any tangible medium that may contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include but not limited to an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of the machine-readable storage medium would include an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random-access memory (RAM) , a read-only memory (ROM) , an erasable programmable read-only memory (EPROM or Flash memory) , an optical fiber, a portable compact disc read-only memory (CD-ROM) , an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
[0100] Further, while operations are illustrated in a particular order, this should not be understood as requiring that such operations are performed in the particular order shown or in sequential order, or that all illustrated operations are performed to achieve the desired results. In certain circumstances, multitasking and parallel processing may be advantageous. Likewise, while several specific implementation details are contained in the above discussions, these should not be construed as limitations on the scope of the subject matter described herein, but rather as descriptions of features that may be specific to particular embodiments. Certain features that are described in the context of separate embodiments may also be implemented in combination in a single implementation. Rather, various features described in a single implementation may also be implemented in multiple embodiments separately or in any suitable sub-combination.
[0101] Although the subject matter has been described in language specific to structural features and / or methodological acts, it is to be understood that the subject matter specified in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are disclosed as example forms of implementing the claims.
[0102] From the foregoing, it will be appreciated that specific embodiments of the presently disclosed technology have been described herein for purposes of illustration, but that various modifications may be made without deviating from the scope of the disclosure. Accordingly, the presently disclosed technology is not limited except as by the appended claims.
[0103] Embodiments of the subject matter and the functional operations described in the present disclosure can be implemented in various systems, digital electronic circuitry, or in computer software, firmware, or hardware, including the structures disclosed in this specification and their structural equivalents, or in combinations of one or more of them. Embodiments of the subject matter described in this specification can be implemented as one or more computer program products, i.e., one or more modules of computer program instructions encoded on a tangible and non-transitory computer readable medium for execution by, or to control the operation of, data processing apparatus. The computer readable medium can be a machine-readable storage device, a machine-readable storage substrate, a memory device, a composition of matter effecting a machine-readable propagated signal, or a combination of one or more of them. The term “data processing unit” or “data processing apparatus” encompasses all apparatus, devices, and machines for processing data, including by way of example a programmable processor, a computer, or multiple processors or computers. The apparatus can include, in addition to hardware, code that creates an execution environment for the computer program in question, e.g., code that constitutes processor firmware, a protocol stack, a database management system, an operating system, or a combination of one or more of them.
[0104] A computer program (also known as a program, software, software application, script, or code) can be written in any form of programming language, including compiled or interpreted languages, and it can be deployed in any form, including as a stand-alone program or as a module, component, subroutine, or other unit suitable for use in a computing environment. A computer program does not necessarily correspond to a file in a file system. A program can be stored in a portion of a file that holds other programs or data (e.g., one or more scripts stored in a markup language document) , in a single file dedicated to the program in question, or in multiple coordinated files (e.g., files that store one or more modules, sub programs, or portions of code) . A computer program can be deployed to be executed on one computer or on multiple computers that are located at one site or distributed across multiple sites and interconnected by a communication network.
[0105] Processors suitable for the execution of a computer program include, by way of example, both general and special purpose microprocessors, and any one or more processors of any kind of digital computer. Generally, a processor will receive instructions and data from a read only memory or a random access memory or both. The essential elements of a computer are a processor for performing instructions and one or more memory devices for storing instructions and data. Generally, a computer will also include, or be operatively coupled to receive data from or transfer data to, or both, one or more mass storage devices for storing data, e.g., magnetic, magneto optical disks, or optical disks. However, a computer need not have such devices. Computer readable media suitable for storing computer program instructions and data include all forms of nonvolatile memory, media, and memory devices, including by way of example semiconductor memory devices, e.g., EPROM, EEPROM, and flash memory devices. The processor and the memory can be supplemented by, or incorporated in, special purpose logic circuitry.
[0106] It is intended that the specification, together with the drawings, be considered exemplary only, where exemplary means an example. As used herein, the use of “or” is intended to include “and / or” , unless the context clearly indicates otherwise.
[0107] While the present disclosure contains many specifics, these should not be construed as limitations on the scope of any disclosure or of what may be claimed, but rather as descriptions of features that may be specific to particular embodiments of particular disclosures. Certain features that are described in the present disclosure in the context of separate embodiments can also be implemented in combination in a single implementation. Conversely, various features that are described in the context of a single implementation can also be implemented in multiple embodiments separately or in any suitable sub-combination. Moreover, although features may be described above as acting in certain combinations and even initially claimed as such, one or more features from a claimed combination can in some cases be excised from the combination, and the claimed combination may be directed to a sub-combination or variation of a sub-combination.
[0108] Similarly, while operations are illustrated in the drawings in a particular order, this should not be understood as requiring that such operations be performed in the particular order shown or in sequential order, or that all illustrated operations be performed, to achieve desirable results. Moreover, the separation of various system components in the embodiments described in the present disclosure should not be understood as requiring such separation in all embodiments. Only a few embodiments and examples are described and other embodiments, enhancements and variations can be made based on what is described and illustrated in the present disclosure.
Claims
1.A method of moiré superlattice simulation, comprising:obtaining property information of one or more particles in a moiré superlattice, the property information at least comprising initial position information of the one or more particles;determining, by using a neural network, respective ground state wave functions of the one or more particles based on the property information; andgenerating a moiré pattern of the moiré superlattice based on the respective ground state wave functions, the moiré pattern representing a particle density distribution across the moirésuperlattice.2.The method of claim 1, wherein determining the respective ground state wave functions of the one or more particles comprises:generating, by using the neural network, respective initial wave functions of the one or more particles based on the property information;updating the neural network iteratively based on the respective initial wave functions and an energy function for the moiré superlattice; andgenerating the respective ground state wave functions by using the updated neural network.3.The method of claim 2, wherein updating the neural network iteratively based on the respective initial wave functions and an energy function for the moiré superlattice comprises:predicting a system energy of the moiré superlattice based on the respective initial wave functions and the energy function for the moiré superlattice; andupdating the neural network iteratively by minimizing the predicted system energy until a predetermined condition is met.4.The method of any of claims 2-3, wherein the energy function comprises:an item associated with a kinetic energy of each of the one or more particles,an item associated with a moiré potential of each of the one or more particles, andan item associated with electrostatic interaction between different particles of the one or more particles.5.The method of claim 4, wherein a moiré potential of a given particles of the one or more particles is based on a position of the given particle in a moiré cell of the moiré superlattice.6.The method of claim 2, further comprising:converting, based on a scaling factor, the energy function from a first unit of length to a second unit of length larger than the first unit of length, andwherein the neural network is updated iteratively based on the respective initial wave functions and the converted energy function.7.The method of any of claims 1-6, wherein obtaining the property information of the one or more particles in the moiré superlattice comprises:determining a number of moiré cells to be simulated and a filling rate of particles in a moiré cell; anddetermining the initial position information by distributing the one or more particles uniformly over the number of moiré cells according to the filing rate.8.The method of claim 7, wherein the filling rate is greater than a predetermined value to indicate that two or more particles are distributed within a same moiré cell, andwherein obtaining the property information of the one or more particles in the moirésuperlattice further comprises:determining, for each moiré cell of the number of moiré cells, respective spin directions of the two or more particles based on a spin state of the moiré cell; anddetermining spin state information as a portion of the property information based on the respective spin directions of the two or more particles.9.An electronic device, comprising a computer processor coupled to a computer-readable memory unit, the memory unit comprising instructions that when executed by the computer processor implements a method of moiré superlattice simulation, the method comprising:obtaining property information of one or more particles in a moiré superlattice, the property information at least comprising initial position information of the one or more particles;determining, by using a neural network, respective ground state wave functions of the one or more particles based on the property information; andgenerating a moiré pattern of the moiré superlattice based on the respective ground state wave functions, the moiré pattern representing a particle density distribution across the moirésuperlattice.10.The device of claim 9, wherein determining the respective ground state wave functions of the one or more particles comprises:generating, by using the neural network, respective initial wave functions of the one or more particles based on the property information;updating the neural network iteratively based on the respective initial wave functions and an energy function for the moiré superlattice; andgenerating the respective ground state wave functions by using the updated neural network.11.The device of claim 10, wherein updating the neural network iteratively based on the respective initial wave functions and an energy function for the moiré superlattice comprises:predicting a system energy of the moiré superlattice based on the respective initial wave functions and the energy function for the moiré superlattice; andupdating the neural network iteratively by minimizing the predicted system energy until a predetermined condition is met.12.The device of any of claims 10-11, wherein the energy function comprises:an item associated with a kinetic energy of each of the one or more particles,an item associated with a moiré potential of each of the one or more particles, andan item associated with electrostatic interaction between different particles of the one or more particles.13.The device of claim 12, wherein a moiré potential of a given particles of the one or more particles is based on a position of the given particle in a moiré cell of the moiré superlattice.14.The device of claim 10, the method further comprising:converting, based on a scaling factor, the energy function from a first unit of length to a second unit of length larger than the first unit of length, andwherein the neural network is updated iteratively based on the respective initial wave functions and the converted energy function.15.The device of any of claims 9-14, wherein obtaining the property information of the one or more particles in the moiré superlattice comprises:determining a number of moiré cells to be simulated and a filling rate of particles in a moiré cell; anddetermining the initial position information by distributing the one or more particles uniformly over the number of moiré cells according to the filing rate.16.The device of claim 15, wherein the filling rate is greater than a predetermined value to indicate that two or more particles are distributed within a same moiré cell, andwherein obtaining the property information of the one or more particles in the moirésuperlattice further comprises:determining, for each moiré cell of the number of moiré cells, respective spin directions of the two or more particles based on a spin state of the moiré cell; anddetermining spin state information as a portion of the property information based on the respective spin directions of the two or more particles.17.A computer program product, the computer program product comprising a non-transitory computer readable storage medium having program instructions embodied therewith, the program instructions executable by an electronic device to cause the electronic device to perform a method of moiré superlattice simulation, the method comprises:obtaining property information of one or more particles in a moiré superlattice, the property information at least comprising initial position information of the one or more particles;determining, by using a neural network, respective ground state wave functions of the one or more particles based on the property information; andgenerating a moiré pattern of the moiré superlattice based on the respective ground state wave functions, the moiré pattern representing a particle density distribution across the moirésuperlattice.18.The product of claim 17, wherein determining the respective ground state wave functions of the one or more particles comprises:generating, by using the neural network, respective initial wave functions of the one or more particles based on the property information;updating the neural network iteratively based on the respective initial wave functions and an energy function for the moiré superlattice; andgenerating the respective ground state wave functions by using the updated neural network.19.The product of claim 18, wherein updating the neural network iteratively based on the respective initial wave functions and an energy function for the moiré superlattice comprises:predicting a system energy of the moiré superlattice based on the respective initial wave functions and the energy function for the moiré superlattice; andupdating the neural network iteratively by minimizing the predicted system energy until a predetermined condition is met.20.The product of any of claims 18-19, wherein the energy function comprises:an item associated with a kinetic energy of each of the one or more particles,an item associated with a moiré potential of each of the one or more particles, andan item associated with electrostatic interaction between different particles of the one or more particles.
Citation Information
Patent Citations
Data processing method and device for quantum chemistry system
CN114758733A
Crystal property prediction model training method and device and computer equipment
CN115346616A
Micro-groove particle sorting experiment particle detection method based on deep learning
CN117011758A
Rearrangable cellular neutral net
JP2000348009A