Showing 1–50 of 53 results for author: Huang, E W

Unleashing the Power of LLMs as Multi-Modal Encoders for Text and Graph-Structured Data

Authors: Jiacheng Lin, Kun Qian, Haoyu Han, Nurendra Choudhary, Tianxin Wei, Zhongruo Wang, Sahika Genc, Edward W Huang, Sheng Wang, Karthik Subbian, Danai Koutra, Jimeng Sun

Abstract: Graph-structured information offers rich contextual information that can enhance language models by providing structured relationships and hierarchies, leading to more expressive embeddings for various applications such as retrieval, question answering, and classification. However, existing methods for integrating graph and text embeddings, often based on Multi-layer Perceptrons (MLPs) or shallow… ▽ More Graph-structured information offers rich contextual information that can enhance language models by providing structured relationships and hierarchies, leading to more expressive embeddings for various applications such as retrieval, question answering, and classification. However, existing methods for integrating graph and text embeddings, often based on Multi-layer Perceptrons (MLPs) or shallow transformers, are limited in their ability to fully exploit the heterogeneous nature of these modalities. To overcome this, we propose Janus, a simple yet effective framework that leverages Large Language Models (LLMs) to jointly encode text and graph data. Specifically, Janus employs an MLP adapter to project graph embeddings into the same space as text embeddings, allowing the LLM to process both modalities jointly. Unlike prior work, we also introduce contrastive learning to align the graph and text spaces more effectively, thereby improving the quality of learned joint embeddings. Empirical results across six datasets spanning three tasks, knowledge graph-contextualized question answering, graph-text pair classification, and retrieval, demonstrate that Janus consistently outperforms existing baselines, achieving significant improvements across multiple datasets, with gains of up to 11.4% in QA tasks. These results highlight Janus's effectiveness in integrating graph and text data. Ablation studies further validate the effectiveness of our method. △ Less

Submitted 14 October, 2024; originally announced October 2024.

Charge Susceptibility and Kubo Response in Hatsugai-Kohmoto-related Models

Authors: Yuhao Ma, Jinchao Zhao, Edwin W. Huang, Dhruv Kush, Barry Bradlyn, Philip W. Phillips

Abstract: We study in depth the charge susceptibility for the band Hatsugai-Kohmoto (HK) and orbital (OHK) models. As either of these models describes a Mott insulator, the charge susceptibility takes on the form of a modified Lindhard function with lower and upper Hubbard bands, thereby giving rise to a multi-pole structure. The particle-hole continuum consists of hot spots along the $ω$ vs $q$ axis arisin… ▽ More We study in depth the charge susceptibility for the band Hatsugai-Kohmoto (HK) and orbital (OHK) models. As either of these models describes a Mott insulator, the charge susceptibility takes on the form of a modified Lindhard function with lower and upper Hubbard bands, thereby giving rise to a multi-pole structure. The particle-hole continuum consists of hot spots along the $ω$ vs $q$ axis arising from inter-band transitions. Such transitions, which are strongly suppressed in non-interacting systems, are obtained here because of the non-rigidity of the Hubbard bands. This modified Lindhard function gives rise to a plasmon dispersion that is inversely dependent on the momentum, resulting in an additional contribution to the conventional f-sum rule. This extra contribution originates from a long-range diamagnetic contribution to the current. This results in a non-commutativity of the long-wavelength ($q\rightarrow 0$) and thermodynamic ($L\rightarrow\infty$) limits. When the correct limits are taken, we find that the Kubo response computed with either open or periodic boundary conditions yields identical results that are consistent with the continuity equation contrary to recent claims. We also show that the long wavelength pathology of the current noted previously also plagues the Anderson impurity model interpretation of dynamical mean-field theory (DMFT). Coupled with our previous work\cite{mai20231} which showed that HK is the correct $d=\infty$ limit of the Hubbard model, we arrive at the conclusion that single-orbital HK=DMFT. △ Less

Submitted 27 September, 2024; v1 submitted 11 September, 2024; originally announced September 2024.

Comments: typos corrected and figures edited

Covering a Graph with Dense Subgraph Families, via Triangle-Rich Sets

Authors: Sabyasachi Basu, Daniel Paul-Pena, Kun Qian, C. Seshadhri, Edward W Huang, Karthik Subbian

Abstract: Graphs are a fundamental data structure used to represent relationships in domains as diverse as the social sciences, bioinformatics, cybersecurity, the Internet, and more. One of the central observations in network science is that real-world graphs are globally sparse, yet contains numerous "pockets" of high edge density. A fundamental task in graph mining is to discover these dense subgraphs. Mo… ▽ More Graphs are a fundamental data structure used to represent relationships in domains as diverse as the social sciences, bioinformatics, cybersecurity, the Internet, and more. One of the central observations in network science is that real-world graphs are globally sparse, yet contains numerous "pockets" of high edge density. A fundamental task in graph mining is to discover these dense subgraphs. Most common formulations of the problem involve finding a single (or a few) "optimally" dense subsets. But in most real applications, one does not care for the optimality. Instead, we want to find a large collection of dense subsets that covers a significant fraction of the input graph. We give a mathematical formulation of this problem, using a new definition of regularly triangle-rich (RTR) families. These families capture the notion of dense subgraphs that contain many triangles and have degrees comparable to the subgraph size. We design a provable algorithm, RTRExtractor, that can discover RTR families that approximately cover any RTR set. The algorithm is efficient and is inspired by recent results that use triangle counts for community testing and clustering. We show that RTRExtractor has excellent behavior on a large variety of real-world datasets. It is able to process graphs with hundreds of millions of edges within minutes. Across many datasets, RTRExtractor achieves high coverage using high edge density datasets. For example, the output covers a quarter of the vertices with subgraphs of edge density more than (say) $0.5$, for datasets with 10M+ edges. We show an example of how the output of RTRExtractor correlates with meaningful sets of similar vertices in a citation network, demonstrating the utility of RTRExtractor for unsupervised graph discovery tasks. △ Less

Submitted 23 July, 2024; originally announced July 2024.

All Against Some: Efficient Integration of Large Language Models for Message Passing in Graph Neural Networks

Authors: Ajay Jaiswal, Nurendra Choudhary, Ravinarayana Adkathimar, Muthu P. Alagappan, Gaurush Hiranandani, Ying Ding, Zhangyang Wang, Edward W Huang, Karthik Subbian

Abstract: Graph Neural Networks (GNNs) have attracted immense attention in the past decade due to their numerous real-world applications built around graph-structured data. On the other hand, Large Language Models (LLMs) with extensive pretrained knowledge and powerful semantic comprehension abilities have recently shown a remarkable ability to benefit applications using vision and text data. In this paper,… ▽ More Graph Neural Networks (GNNs) have attracted immense attention in the past decade due to their numerous real-world applications built around graph-structured data. On the other hand, Large Language Models (LLMs) with extensive pretrained knowledge and powerful semantic comprehension abilities have recently shown a remarkable ability to benefit applications using vision and text data. In this paper, we investigate how LLMs can be leveraged in a computationally efficient fashion to benefit rich graph-structured data, a modality relatively unexplored in LLM literature. Prior works in this area exploit LLMs to augment every node features in an ad-hoc fashion (not scalable for large graphs), use natural language to describe the complex structural information of graphs, or perform computationally expensive finetuning of LLMs in conjunction with GNNs. We propose E-LLaGNN (Efficient LLMs augmented GNNs), a framework with an on-demand LLM service that enriches message passing procedure of graph learning by enhancing a limited fraction of nodes from the graph. More specifically, E-LLaGNN relies on sampling high-quality neighborhoods using LLMs, followed by on-demand neighborhood feature enhancement using diverse prompts from our prompt catalog, and finally information aggregation using message passing from conventional GNN architectures. We explore several heuristics-based active node selection strategies to limit the computational and memory footprint of LLMs when handling millions of nodes. Through extensive experiments & ablation on popular graph benchmarks of varying scales (Cora, PubMed, ArXiv, & Products), we illustrate the effectiveness of our E-LLaGNN framework and reveal many interesting capabilities such as improved gradient flow in deep GNNs, LLM-free inference ability etc. △ Less

Submitted 20 July, 2024; originally announced July 2024.

Enhanced pair-density-wave vertices in a bilayer Hubbard model at half-filling

Authors: Fangze Liu, Xu-Xin Huang, Edwin W. Huang, Brian Moritz, Thomas P. Devereaux

Abstract: Motivated by the pair-density-wave (PDW) state found in the one-dimensional Kondo-Heisenberg chain, we report on a determinant quantum Monte Carlo DQMC study of pair-fields for a two-dimensional half-filled Hubbard layer coupled to an itinerant, non-interacting layer with one electron per site. In a specific range of interlayer hopping, the pairing vertex associated with PDW order becomes more att… ▽ More Motivated by the pair-density-wave (PDW) state found in the one-dimensional Kondo-Heisenberg chain, we report on a determinant quantum Monte Carlo DQMC study of pair-fields for a two-dimensional half-filled Hubbard layer coupled to an itinerant, non-interacting layer with one electron per site. In a specific range of interlayer hopping, the pairing vertex associated with PDW order becomes more attractive than that for uniform d-wave pairing, although both remain subdominant to the leading antiferromagnetic correlations at half-filling. Our result sheds light on where one potentially may find a PDW state in such a model. △ Less

Submitted 9 October, 2024; v1 submitted 1 April, 2024; originally announced April 2024.

An Interpretable Ensemble of Graph and Language Models for Improving Search Relevance in E-Commerce

Authors: Nurendra Choudhary, Edward W Huang, Karthik Subbian, Chandan K. Reddy

Abstract: The problem of search relevance in the E-commerce domain is a challenging one since it involves understanding the intent of a user's short nuanced query and matching it with the appropriate products in the catalog. This problem has traditionally been addressed using language models (LMs) and graph neural networks (GNNs) to capture semantic and inter-product behavior signals, respectively. However,… ▽ More The problem of search relevance in the E-commerce domain is a challenging one since it involves understanding the intent of a user's short nuanced query and matching it with the appropriate products in the catalog. This problem has traditionally been addressed using language models (LMs) and graph neural networks (GNNs) to capture semantic and inter-product behavior signals, respectively. However, the rapid development of new architectures has created a gap between research and the practical adoption of these techniques. Evaluating the generalizability of these models for deployment requires extensive experimentation on complex, real-world datasets, which can be non-trivial and expensive. Furthermore, such models often operate on latent space representations that are incomprehensible to humans, making it difficult to evaluate and compare the effectiveness of different models. This lack of interpretability hinders the development and adoption of new techniques in the field. To bridge this gap, we propose Plug and Play Graph LAnguage Model (PP-GLAM), an explainable ensemble of plug and play models. Our approach uses a modular framework with uniform data processing pipelines. It employs additive explanation metrics to independently decide whether to include (i) language model candidates, (ii) GNN model candidates, and (iii) inter-product behavioral signals. For the task of search relevance, we show that PP-GLAM outperforms several state-of-the-art baselines as well as a proprietary model on real-world multilingual, multi-regional e-commerce datasets. To promote better model comprehensibility and adoption, we also provide an analysis of the explainability and computational complexity of our model. We also provide the public codebase and provide a deployment strategy for practical implementation. △ Less

Submitted 1 March, 2024; originally announced March 2024.

Comments: Accepted to The Web Conference 2024 (Industry)

ACM Class: H.3.3; I.2.7; J.7

The emergence of antiferromagnetic correlations and Kondo-like features in a two-band model for infinite-layer nickelates

Authors: Fangze Liu, Cheng Peng, Edwin W. Huang, Brian Moritz, Chunjing Jia, Thomas P. Devereaux

Abstract: We report a determinant quantum Monte Carlo study of a two-band model, inspired by infinite-layer nickelates, focusing on the influence of interlayer hybridization between $3d_{x^2-y^2}$ orbitals derived from Ni (or Ni and O) in one layer and rare-earth ($R$) 5d orbitals in the other layer, hereafter the NI and $R$ layers, respectively. For a filling with one electron shared between the two layers… ▽ More We report a determinant quantum Monte Carlo study of a two-band model, inspired by infinite-layer nickelates, focusing on the influence of interlayer hybridization between $3d_{x^2-y^2}$ orbitals derived from Ni (or Ni and O) in one layer and rare-earth ($R$) 5d orbitals in the other layer, hereafter the NI and $R$ layers, respectively. For a filling with one electron shared between the two layers on average, interlayer hybridization leads to "self-doped" holes in the Ni layer and the absence of antiferromagnetic ordering, but rather the appearance of spin-density and charge-density stripe-like states. As the interlayer hybridization increases, both the Ni and $R$ layers develop antiferromagnetic correlations, even though either layer individually remains away from half-filling. For hybridization within an intermediate range, roughly comparable to the intralayer nearest-neighbor hopping $t_{\text{Ni}}$, the model develops signatures of Kondo-like physics. △ Less

Submitted 30 April, 2024; v1 submitted 24 January, 2024; originally announced January 2024.

ForeSeer: Product Aspect Forecasting Using Temporal Graph Embedding

Authors: Zixuan Liu, Gaurush Hiranandani, Kun Qian, Eddie W. Huang, Yi Xu, Belinda Zeng, Karthik Subbian, Sheng Wang

Abstract: Developing text mining approaches to mine aspects from customer reviews has been well-studied due to its importance in understanding customer needs and product attributes. In contrast, it remains unclear how to predict the future emerging aspects of a new product that currently has little review information. This task, which we named product aspect forecasting, is critical for recommending new pro… ▽ More Developing text mining approaches to mine aspects from customer reviews has been well-studied due to its importance in understanding customer needs and product attributes. In contrast, it remains unclear how to predict the future emerging aspects of a new product that currently has little review information. This task, which we named product aspect forecasting, is critical for recommending new products, but also challenging because of the missing reviews. Here, we propose ForeSeer, a novel textual mining and product embedding approach progressively trained on temporal product graphs for this novel product aspect forecasting task. ForeSeer transfers reviews from similar products on a large product graph and exploits these reviews to predict aspects that might emerge in future reviews. A key novelty of our method is to jointly provide review, product, and aspect embeddings that are both time-sensitive and less affected by extremely imbalanced aspect frequencies. We evaluated ForeSeer on a real-world product review system containing 11,536,382 reviews and 11,000 products over 3 years. We observe that ForeSeer substantially outperformed existing approaches with at least 49.1\% AUPRC improvement under the real setting where aspect associations are not given. ForeSeer further improves future link prediction on the product graph and the review aspect association prediction. Collectively, Foreseer offers a novel framework for review forecasting by effectively integrating review text, product network, and temporal information, opening up new avenues for online shopping recommendation and e-commerce applications. △ Less

Submitted 7 October, 2023; originally announced October 2023.

Particle-hole asymmetric ferromagnetism and spin textures in the triangular Hubbard-Hofstadter model

Authors: Jixun K. Ding, Luhang Yang, Wen O. Wang, Ziyan Zhu, Cheng Peng, Peizhi Mai, Edwin W. Huang, Brian Moritz, Philip W. Phillips, Benjamin E. Feldman, Thomas P. Devereaux

Abstract: In a lattice model subject to a perpendicular magnetic field, when the lattice constant is comparable to the magnetic length, one enters the "Hofstadter regime," where continuum Landau levels become fractal magnetic Bloch bands. Strong mixing between bands alters the nature of the resulting quantum phases compared to the continuum limit; lattice potential, magnetic field, and Coulomb interaction m… ▽ More In a lattice model subject to a perpendicular magnetic field, when the lattice constant is comparable to the magnetic length, one enters the "Hofstadter regime," where continuum Landau levels become fractal magnetic Bloch bands. Strong mixing between bands alters the nature of the resulting quantum phases compared to the continuum limit; lattice potential, magnetic field, and Coulomb interaction must be treated on equal footing. Using determinant quantum Monte Carlo (DQMC) and density matrix renormalization group (DMRG) techniques, we study this regime numerically in the context of the Hubbard-Hofstadter model on a triangular lattice. In the field-filling phase diagram, we find a broad wedge-shaped region of ferromagnetic ground states for filling factor $ν\leq 1$, bounded below by filling factor $ν= 1$ and bounded above by half-filling the lowest Hofstadter subband. We observe signatures of SU(2) quantum Hall ferromagnetism at filling factors $ν=1$ and $ν=3$. The phases near $ν=1$ are particle-hole asymmetric, and we observe a rapid decrease in ground state spin polarization consistent with the formation of skyrmions only on the electron doped side. At large fields, above the ferromagnetic wedge, we observe a low-spin metallic region with spin correlations peaked at small momenta. We argue that the phenomenology of this region likely results from exchange interaction mixing fractal Hofstadter subbands. The phase diagram derived beyond the continuum limit points to a rich landscape to explore interaction effects in magnetic Bloch bands. △ Less

Submitted 1 March, 2024; v1 submitted 14 September, 2023; originally announced September 2023.

Comments: 10+7 pages, 6+13 figures

Communication-Free Distributed GNN Training with Vertex Cut

Authors: Kaidi Cao, Rui Deng, Shirley Wu, Edward W Huang, Karthik Subbian, Jure Leskovec

Abstract: Training Graph Neural Networks (GNNs) on real-world graphs consisting of billions of nodes and edges is quite challenging, primarily due to the substantial memory needed to store the graph and its intermediate node and edge features, and there is a pressing need to speed up the training process. A common approach to achieve speed up is to divide the graph into many smaller subgraphs, which are the… ▽ More Training Graph Neural Networks (GNNs) on real-world graphs consisting of billions of nodes and edges is quite challenging, primarily due to the substantial memory needed to store the graph and its intermediate node and edge features, and there is a pressing need to speed up the training process. A common approach to achieve speed up is to divide the graph into many smaller subgraphs, which are then distributed across multiple GPUs in one or more machines and processed in parallel. However, existing distributed methods require frequent and substantial cross-GPU communication, leading to significant time overhead and progressively diminishing scalability. Here, we introduce CoFree-GNN, a novel distributed GNN training framework that significantly speeds up the training process by implementing communication-free training. The framework utilizes a Vertex Cut partitioning, i.e., rather than partitioning the graph by cutting the edges between partitions, the Vertex Cut partitions the edges and duplicates the node information to preserve the graph structure. Furthermore, the framework maintains high model accuracy by incorporating a reweighting mechanism to handle a distorted graph distribution that arises from the duplicated nodes. We also propose a modified DropEdge technique to further speed up the training process. Using an extensive set of experiments on real-world networks, we demonstrate that CoFree-GNN speeds up the GNN training process by up to 10 times over the existing state-of-the-art GNN training approaches. △ Less

Submitted 6 August, 2023; originally announced August 2023.

You Only Transfer What You Share: Intersection-Induced Graph Transfer Learning for Link Prediction

Authors: Wenqing Zheng, Edward W Huang, Nikhil Rao, Zhangyang Wang, Karthik Subbian

Abstract: Link prediction is central to many real-world applications, but its performance may be hampered when the graph of interest is sparse. To alleviate issues caused by sparsity, we investigate a previously overlooked phenomenon: in many cases, a densely connected, complementary graph can be found for the original graph. The denser graph may share nodes with the original graph, which offers a natural b… ▽ More Link prediction is central to many real-world applications, but its performance may be hampered when the graph of interest is sparse. To alleviate issues caused by sparsity, we investigate a previously overlooked phenomenon: in many cases, a densely connected, complementary graph can be found for the original graph. The denser graph may share nodes with the original graph, which offers a natural bridge for transferring selective, meaningful knowledge. We identify this setting as Graph Intersection-induced Transfer Learning (GITL), which is motivated by practical applications in e-commerce or academic co-authorship predictions. We develop a framework to effectively leverage the structural prior in this setting. We first create an intersection subgraph using the shared nodes between the two graphs, then transfer knowledge from the source-enriched intersection subgraph to the full target graph. In the second step, we consider two approaches: a modified label propagation, and a multi-layer perceptron (MLP) model in a teacher-student regime. Experimental results on proprietary e-commerce datasets and open-source citation graphs show that the proposed workflow outperforms existing transfer learning baselines that do not explicitly utilize the intersection structure. △ Less

Submitted 18 June, 2023; v1 submitted 27 February, 2023; originally announced February 2023.

Comments: Accepted in TMLR (https://openreview.net/forum?id=Nn71AdKyYH)

Quantitative assessment of the universal thermopower in the Hubbard model

Authors: Wen O. Wang, Jixun K. Ding, Edwin W. Huang, Brian Moritz, Thomas P. Devereaux

Abstract: As primarily an electronic observable, the room-temperature thermopower $S$ in cuprates provides possibilities for a quantitative assessment of the Hubbard model. Using determinant quantum Monte Carlo, we demonstrate agreement between Hubbard model calculations and experimentally measured room-temperature $S$ across multiple cuprate families, both qualitatively in terms of the doping dependence an… ▽ More As primarily an electronic observable, the room-temperature thermopower $S$ in cuprates provides possibilities for a quantitative assessment of the Hubbard model. Using determinant quantum Monte Carlo, we demonstrate agreement between Hubbard model calculations and experimentally measured room-temperature $S$ across multiple cuprate families, both qualitatively in terms of the doping dependence and quantitatively in terms of magnitude. We observe an upturn in $S$ with decreasing temperatures, which possesses a slope comparable to that observed experimentally in cuprates. From our calculations, the doping at which $S$ changes sign occurs in close proximity to a vanishing temperature dependence of the chemical potential at fixed density. Our results emphasize the importance of interaction effects in the systematic assessment of the thermopower $S$ in cuprates. △ Less

Submitted 3 November, 2023; v1 submitted 25 February, 2023; originally announced February 2023.

Comments: 7 pages, 4 figures. Supplementary Information: 9 pages, 7 figures

Journal ref: Nat. Commun. 14, 7064 (2023)

Search Behavior Prediction: A Hypergraph Perspective

Authors: Yan Han, Edward W Huang, Wenqing Zheng, Nikhil Rao, Zhangyang Wang, Karthik Subbian

Abstract: Although the bipartite shopping graphs are straightforward to model search behavior, they suffer from two challenges: 1) The majority of items are sporadically searched and hence have noisy/sparse query associations, leading to a \textit{long-tail} distribution. 2) Infrequent queries are more likely to link to popular items, leading to another hurdle known as \textit{disassortative mixing}. To add… ▽ More Although the bipartite shopping graphs are straightforward to model search behavior, they suffer from two challenges: 1) The majority of items are sporadically searched and hence have noisy/sparse query associations, leading to a \textit{long-tail} distribution. 2) Infrequent queries are more likely to link to popular items, leading to another hurdle known as \textit{disassortative mixing}. To address these two challenges, we go beyond the bipartite graph to take a hypergraph perspective, introducing a new paradigm that leverages \underline{auxiliary} information from anonymized customer engagement sessions to assist the \underline{main task} of query-item link prediction. This auxiliary information is available at web scale in the form of search logs. We treat all items appearing in the same customer session as a single hyperedge. The hypothesis is that items in a customer session are unified by a common shopping interest. With these hyperedges, we augment the original bipartite graph into a new \textit{hypergraph}. We develop a \textit{\textbf{D}ual-\textbf{C}hannel \textbf{A}ttention-Based \textbf{H}ypergraph Neural Network} (\textbf{DCAH}), which synergizes information from two potentially noisy sources (original query-item edges and item-item hyperedges). In this way, items on the tail are better connected due to the extra hyperedges, thereby enhancing their link prediction performance. We further integrate DCAH with self-supervised graph pre-training and/or DropEdge training, both of which effectively alleviate disassortative mixing. Extensive experiments on three proprietary E-Commerce datasets show that DCAH yields significant improvements of up to \textbf{24.6\% in mean reciprocal rank (MRR)} and \textbf{48.3\% in recall} compared to GNN-based baselines. Our source code is available at \url{https://github.com/amazon-science/dual-channel-hypergraph-neural-network}. △ Less

Submitted 28 November, 2022; v1 submitted 23 November, 2022; originally announced November 2022.

Comments: WSDM 2023

TuneUp: A Simple Improved Training Strategy for Graph Neural Networks

Authors: Weihua Hu, Kaidi Cao, Kexin Huang, Edward W Huang, Karthik Subbian, Kenji Kawaguchi, Jure Leskovec

Abstract: Despite recent advances in Graph Neural Networks (GNNs), their training strategies remain largely under-explored. The conventional training strategy learns over all nodes in the original graph(s) equally, which can be sub-optimal as certain nodes are often more difficult to learn than others. Here we present TuneUp, a simple curriculum-based training strategy for improving the predictive performan… ▽ More Despite recent advances in Graph Neural Networks (GNNs), their training strategies remain largely under-explored. The conventional training strategy learns over all nodes in the original graph(s) equally, which can be sub-optimal as certain nodes are often more difficult to learn than others. Here we present TuneUp, a simple curriculum-based training strategy for improving the predictive performance of GNNs. TuneUp trains a GNN in two stages. In the first stage, TuneUp applies conventional training to obtain a strong base GNN. The base GNN tends to perform well on head nodes (nodes with large degrees) but less so on tail nodes (nodes with small degrees). Therefore, the second stage of TuneUp focuses on improving prediction on the difficult tail nodes by further training the base GNN on synthetically generated tail node data. We theoretically analyze TuneUp and show it provably improves generalization performance on tail nodes. TuneUp is simple to implement and applicable to a broad range of GNN architectures and prediction tasks. Extensive evaluation of TuneUp on five diverse GNN architectures, three types of prediction tasks, and both transductive and inductive settings shows that TuneUp significantly improves the performance of the base GNN on tail nodes, while often even improving the performance on head nodes. Altogether, TuneUp produces up to 57.6% and 92.2% relative predictive performance improvement in the transductive and the challenging inductive settings, respectively. △ Less

Submitted 26 August, 2023; v1 submitted 26 October, 2022; originally announced October 2022.

Exact solution for finite center-of-mass momentum Cooper pairing

Authors: Chandan Setty, Jinchao Zhao, Laura Fanfarillo, Edwin W. Huang, Peter J. Hirschfeld, Philip W. Phillips, Kun Yang

Abstract: Pair density waves (PDWs) are superconducting states formed by ``Cooper pairs" of electrons containing a non-zero center-of-mass momentum. They are characterized by a spatially modulated order parameter and may occur in a variety of emerging quantum materials such as cuprates, transition metal dichalcogenides (TMDs) and Kagome metals. Despite extensive theoretical and numerical studies seeking PDW… ▽ More Pair density waves (PDWs) are superconducting states formed by ``Cooper pairs" of electrons containing a non-zero center-of-mass momentum. They are characterized by a spatially modulated order parameter and may occur in a variety of emerging quantum materials such as cuprates, transition metal dichalcogenides (TMDs) and Kagome metals. Despite extensive theoretical and numerical studies seeking PDWs in a variety of lattices and interacting settings, there is currently no generic and robust mechanism that favors a modulated solution of the superconducting order parameter in the presence of time reversal symmetry. Here, we study the problem of two electrons subject to an anisotropic ($d$-wave) attractive potential. We solve the two-body Schrodinger wave equation exactly to determine the pair binding energy as a function of the center-of-mass momentum. We find that a modulated (finite momentum) pair is favored over a homogeneous (zero momentum) solution above a critical interaction. Using this insight from the exact two-body solution, we construct a BCS-like variational many-body wave function and calculate the free energy and superconducting gap as a function of the center-of-mass momentum. A zero temperature analysis of the energy shows that the conclusions of the two-body problem are robust in the many-body limit. Our results lay the theoretical and microscopic foundation for the existence of PDWs. △ Less

Submitted 21 September, 2022; originally announced September 2022.

Comments: 24 pages, 6 figures

Journal ref: Phys. Rev. B 108, 174506 (2023)

The Wiedemann-Franz law in doped Mott insulators without quasiparticles

Authors: Wen O. Wang, Jixun K. Ding, Yoni Schattner, Edwin W. Huang, Brian Moritz, Thomas P. Devereaux

Abstract: Many metallic quantum materials display anomalous transport phenomena that defy a Fermi liquid description. Here, we use numerical methods to calculate thermal and charge transport in the doped Hubbard model and observe a cross-over separating high- and low-temperature behaviors. Distinct from the behavior at high temperatures, the Lorenz number $L$ becomes weakly doping dependent and less sensiti… ▽ More Many metallic quantum materials display anomalous transport phenomena that defy a Fermi liquid description. Here, we use numerical methods to calculate thermal and charge transport in the doped Hubbard model and observe a cross-over separating high- and low-temperature behaviors. Distinct from the behavior at high temperatures, the Lorenz number $L$ becomes weakly doping dependent and less sensitive to parameters at low temperatures. At the lowest numerically accessible temperatures, $L$ roughly approaches the Wiedemann-Franz constant $L_0$, even in a doped Mott insulator that lacks well-defined quasiparticles. Decomposing the energy current operator indicates a compensation between kinetic and potential contributions, which may help to clarify the interpretation of transport experiments beyond Boltzmann theory in strongly correlated metals. △ Less

Submitted 1 December, 2023; v1 submitted 19 August, 2022; originally announced August 2022.

Comments: 15 pages, 4 figures. Supplementary Materials: 23 pages, 15 figures

Journal ref: Science 382, 1070-1073 (2023)

Interaction-driven Spontaneous Ferromagnetic Insulating States with Odd Chern Numbers

Authors: Peizhi Mai, Edwin W. Huang, Jiachen Yu, Benjamin E. Feldman, Philip W. Phillips

Abstract: Motivated by recent experimental work on moiré systems in a strong magnetic field, we compute the compressibility as well as the spin correlations and Hofstadter spectrum of spinful electrons on a honeycomb lattice with Hubbard interactions using the determinantal quantum Monte Carlo method. While the interactions in general preserve quantum and anomalous Hall states, emergent features arise corre… ▽ More Motivated by recent experimental work on moiré systems in a strong magnetic field, we compute the compressibility as well as the spin correlations and Hofstadter spectrum of spinful electrons on a honeycomb lattice with Hubbard interactions using the determinantal quantum Monte Carlo method. While the interactions in general preserve quantum and anomalous Hall states, emergent features arise corresponding to an antiferromagnetic insulator at half-filling and other incompressible states following the Chern sequence $\pm (2N+1)$. These odd integer Chern states exhibit strong ferromagnetic correlations and arise spontaneously without any external mechanism for breaking the spin-rotation symmetry. Analogs of these magnetic states should be observable in general interacting quantum Hall systems. In addition, the interacting Hofstadter spectrum is qualitatively similar to the experimental data at intermediate values of the on-site interaction. △ Less

Submitted 20 February, 2023; v1 submitted 17 May, 2022; originally announced May 2022.

Comments: 8 pages, 5 figures and a supplement

Fluctuating intertwined stripes in the strange metal regime of the Hubbard model

Authors: Edwin W. Huang, Tianyi Liu, Wen O. Wang, Hong-Chen Jiang, Peizhi Mai, Thomas A. Maier, Steven Johnston, Brian Moritz, Thomas P. Devereaux

Abstract: Strongly correlated electron systems host a variety of poorly understood correlations in their high temperature normal state. Unlike ordered phases defined by order parameters, these normal state phases are often defined through unconventional properties such as strange metallic transport or spectroscopic pseudogaps. Characterizing the microscopic correlations in the normal state is necessary to e… ▽ More Strongly correlated electron systems host a variety of poorly understood correlations in their high temperature normal state. Unlike ordered phases defined by order parameters, these normal state phases are often defined through unconventional properties such as strange metallic transport or spectroscopic pseudogaps. Characterizing the microscopic correlations in the normal state is necessary to elucidate mechanisms that lead to these properties and their connection to ground state orders. Here we establish the presence of intertwined charge and spin stripes in the strange metal normal state of the Hubbard model using determinant quantum Monte Carlo calculations. The charge and spin density waves constituting the stripes are fluctuating and short-ranged, yet they obey a mutual commensurability relation and remain microscopically interlocked, as evidenced through measurements of three-point spin-spin-hole correlation functions. Our findings demonstrate the ability of many-body numerical simulations to unravel the microscopic correlations that define quantum states of matter. △ Less

Submitted 17 February, 2022; originally announced February 2022.

Thermodynamics of correlated electrons in a magnetic field

Authors: Jixun K. Ding, Wen O. Wang, Brian Moritz, Yoni Schattner, Edwin W. Huang, Thomas P. Devereaux

Abstract: The Hofstadter-Hubbard model captures the physics of strongly correlated electrons in an applied magnetic field, which is relevant to many recent experiments on Moiré materials. Few large-scale, numerically exact simulations exists for this model. In this work, we simulate the Hubbard-Hofstadter model using the determinant quantum Monte Carlo (DQMC) algorithm. We report the field and Hubbard inter… ▽ More The Hofstadter-Hubbard model captures the physics of strongly correlated electrons in an applied magnetic field, which is relevant to many recent experiments on Moiré materials. Few large-scale, numerically exact simulations exists for this model. In this work, we simulate the Hubbard-Hofstadter model using the determinant quantum Monte Carlo (DQMC) algorithm. We report the field and Hubbard interaction strength dependence of charge compressibility, fermion sign, local moment, magnetic structure factor, and specific heat. The gross structure of magnetic Bloch bands and band gaps determined by the non-interacting Hofstadter spectrum is preserved in the presence of $U$. Incompressible regions of the phase diagram have improved fermion sign. At half filling and intermediate and larger couplings, a strong orbital magnetic field delocalizes electrons and reduces the effect of Hubbard $U$ on thermodynamic properties of the system. △ Less

Submitted 13 August, 2022; v1 submitted 17 November, 2021; originally announced November 2021.

Comments: v2: 11 pages, 7 figures; accepted version

Journal ref: Communications Physics 5, 204 (2022)

Cold Brew: Distilling Graph Node Representations with Incomplete or Missing Neighborhoods

Authors: Wenqing Zheng, Edward W Huang, Nikhil Rao, Sumeet Katariya, Zhangyang Wang, Karthik Subbian

Abstract: Graph Neural Networks (GNNs) have achieved state-of-the-art performance in node classification, regression, and recommendation tasks. GNNs work well when rich and high-quality connections are available. However, their effectiveness is often jeopardized in many real-world graphs in which node degrees have power-law distributions. The extreme case of this situation, where a node may have no neighbor… ▽ More Graph Neural Networks (GNNs) have achieved state-of-the-art performance in node classification, regression, and recommendation tasks. GNNs work well when rich and high-quality connections are available. However, their effectiveness is often jeopardized in many real-world graphs in which node degrees have power-law distributions. The extreme case of this situation, where a node may have no neighbors, is called Strict Cold Start (SCS). SCS forces the prediction to rely completely on the node's own features. We propose Cold Brew, a teacher-student distillation approach to address the SCS and noisy-neighbor challenges for GNNs. We also introduce feature contribution ratio (FCR), a metric to quantify the behavior of inductive GNNs to solve SCS. We experimentally show that FCR disentangles the contributions of different graph data components and helps select the best architecture for SCS generalization. We further demonstrate the superior performance of Cold Brew on several public benchmark and proprietary e-commerce datasets, where many nodes have either very few or noisy connections. Our source code is available at https://github.com/amazon-research/gnn-tail-generalization. △ Less

Submitted 13 March, 2022; v1 submitted 8 November, 2021; originally announced November 2021.

Comments: Published as a conference paper in ICLR 2022

Generic character of charge and spin density waves in superconducting cuprates

Authors: Sangjun Lee, Edwin W. Huang, Thomas A. Johnson, Xuefei Guo, Ali A. Husain, Matteo Mitrano, Kennan Lu, Alexander V. Zakrzewski, Gilberto de la &ntilde, Yingying Peng, Sang-Jun Lee, Hoyoung Jang, Jun-Sik Lee, Young Il Joe, William B. Dorisese, Paul Szypryt, Daniel S. Swetz, Adam A. Aczel, Gregory J. Macdougall, Steven A. Kivelson, Eduardo Fradkin, Peter Abbamonte

Abstract: Understanding the nature of charge density waves (CDW) in cuprate superconductors has been complicated by material specific differences. A striking example is the opposite doping dependence of the CDW ordering wavevector in La-based and Y-based compounds, the two families where charge ordering is strongest and best characterized. Here we report a combined resonant soft X-ray scattering (RSXS) and… ▽ More Understanding the nature of charge density waves (CDW) in cuprate superconductors has been complicated by material specific differences. A striking example is the opposite doping dependence of the CDW ordering wavevector in La-based and Y-based compounds, the two families where charge ordering is strongest and best characterized. Here we report a combined resonant soft X-ray scattering (RSXS) and neutron scattering study of charge and spin density waves in isotopically enriched La$_{1.8-x}$ Eu$_{0.2}$ Sr$_{x}$ CuO$_{4}$ over a range of doping $0.07 \leq x \leq 0.20$. For all dopings studied by RSXS, we find that the CDW amplitude is approximately temperature-independent and develops well above experimentally accessible temperatures. Surprisingly, the CDW ordering wavevector shows a non-monotonic temperature dependence, with a sudden change occurring at temperatures near the SDW onset temperature. We describe this behavior with a Landau-Ginzburg theory for an incommensurate CDW in a metallic system with a finite charge compressibility and CDW-SDW coupling. Our Landau-Ginzburg analysis suggests that the ordering wavevector at high temperatures decreases with increased doping. This behavior is opposite to the trend at low temperatures and highly reminiscent of the doping dependence seen in YBa$_2$ Cu$_3$ O$_{6+δ}$ , suggesting a common origin of the CDW in hole-doped cuprate superconductors. △ Less

Submitted 26 October, 2021; originally announced October 2021.

Comments: Manuscript submitted to PNAS; ArXiv version formatted in Revtex4-2; Supplementary Material included as appendices; 20 pages, 14 figures (6 in the main text), and 41 references

Magnon heat transport in a two-dimensional Mott insulator

Authors: Wen O. Wang, Jixun K. Ding, Brian Moritz, Edwin W. Huang, Thomas P. Devereaux

Abstract: Whether or not anomalies in the thermal conductivity in insulating cuprates can be attributed to antiferromagnetic order and magnons in a 2D Mott insulator remains an intriguing open question. To shed light on this issue, we investigate the thermal conductivity $κ$ and its relationship with the specific heat $c_v$ in the half-filled 2D single-band Hubbard model, using the numerically exact determi… ▽ More Whether or not anomalies in the thermal conductivity in insulating cuprates can be attributed to antiferromagnetic order and magnons in a 2D Mott insulator remains an intriguing open question. To shed light on this issue, we investigate the thermal conductivity $κ$ and its relationship with the specific heat $c_v$ in the half-filled 2D single-band Hubbard model, using the numerically exact determinant quantum Monte Carlo algorithm and maximum entropy analytic continuation. At low temperatures where the charge degrees of freedom are gapped-out and $c_v$ exhibits a clear magnon peak, we observe that thermal conductivity $κ$ also tends to form a peak at similar temperatures. Reducing temperature further produces a sharp upturn in $κ$, associated with an increasing mean-free path. We identify this as the high-temperature side of the anomalous peak in insulating cuprates, where the mean-free path eventually is cut-off by other scattering effects, including phonons, disorder, and physical size. Different scattering effects in our model are identified and analyzed in the thermal diffusivity. △ Less

Submitted 18 April, 2022; v1 submitted 2 September, 2021; originally announced September 2021.

Comments: 6 pages, 4 figures. Supplementary Material: 11 pages, 11 figures

Journal ref: Phys. Rev. B 105, L161103 (2022)

Determinantal Quantum Monte Carlo solver for Cluster Perturbation Theory

Authors: Edwin W. Huang, Yao Wang

Abstract: Cluster Perturbation Theory (CPT) is a technique for computing the spectral function of fermionic models with local interactions. By combining the solution of the model on a finite cluster with perturbation theory on intra-cluster hoppings, CPT provides access to single-particle properties with arbitrary momentum resolution while incurring low computational cost. Here, we introduce Determinantal Q… ▽ More Cluster Perturbation Theory (CPT) is a technique for computing the spectral function of fermionic models with local interactions. By combining the solution of the model on a finite cluster with perturbation theory on intra-cluster hoppings, CPT provides access to single-particle properties with arbitrary momentum resolution while incurring low computational cost. Here, we introduce Determinantal Quantum Monte Carlo (DQMC) as a solver for CPT. Compared to the standard solver, exact diagonalization (ED), the DQMC solver reduces finite size effects through utilizing larger clusters, allows study of temperature dependence, and enables large-scale simulations of a greater set of models. We discuss the implementation of the DQMC solver for CPT and benchmark the CPT+DQMC method for the attractive and repulsive Hubbard models, showcasing its advantages over standard DQMC and CPT+ED simulations. △ Less

Submitted 23 July, 2021; originally announced July 2021.

Intertwined states at finite temperatures in the Hubbard model

Authors: E. W. Huang, W. O. Wang, J. K. Ding, T. Liu, F. Liu, X. -X. Huang, B. Moritz, T. P. Devereaux

Abstract: Significant advances in numerical techniques have enabled recent breakthroughs in the study of various properties of the Hubbard model - a seemingly simple, yet complex model of correlated electrons that has been a focus of study for more than half a century. In particular, it captures the essence of strong correlations, and is believed to possess various emergent, low energy states and collective… ▽ More Significant advances in numerical techniques have enabled recent breakthroughs in the study of various properties of the Hubbard model - a seemingly simple, yet complex model of correlated electrons that has been a focus of study for more than half a century. In particular, it captures the essence of strong correlations, and is believed to possess various emergent, low energy states and collective excitations characteristic of cuprate high-temperature superconducting materials. While a thorough review of all activity is not possible here, we have focused the discussion on our recent work using unbiased, numerically exact, ``brute force", finite temperature quantum Monte Carlo methods. Our various studies reveal a rich variety of quantum liquid crystal phases, and complementary transport properties, which answer some questions, but certainly raise others concerning ``strange metal" behavior and the ultimate fate of quasiparticles in the Hubbard model. △ Less

Submitted 17 June, 2021; originally announced June 2021.

Comments: 10 pages, 12 figures

Journal ref: J. Phys. Soc. Jpn. 90, 111010 (2021)

Superconductivity, Charge-Density-Waves, and Bipolarons in the Holstein model

Authors: B. Nosarzewski, E. W. Huang, Philip M. Dee, I. Esterlis, B. Moritz, S. A. Kivelson, S. Johnston, T. P. Devereaux

Abstract: The electron-phonon (e-ph) interaction remains of great interest in condensed matter physics and plays a vital role in realizing superconductors, charge-density-waves (CDW), and polarons. We study the two-dimensional Holstein model for e-ph coupling using determinant quantum Monte Carlo across a wide range of its phase diagram as a function of temperature, electron density, dimensionless e-ph coup… ▽ More The electron-phonon (e-ph) interaction remains of great interest in condensed matter physics and plays a vital role in realizing superconductors, charge-density-waves (CDW), and polarons. We study the two-dimensional Holstein model for e-ph coupling using determinant quantum Monte Carlo across a wide range of its phase diagram as a function of temperature, electron density, dimensionless e-ph coupling strength, and the adiabatic ratio of the phonon frequency to the Fermi energy. We describe the behavior of the CDW correlations, the competition between superconducting and CDW orders and polaron formation, the optimal conditions for superconductivity, and the transition from the weak-coupling regime to the strong-coupling regime. Superconductivity is optimized at intermediate e-ph coupling strength and intermediate electron density, and the superconducting correlations increase monotonically with phonon frequency. The global maximum for superconductivity in the Holstein model occurs at large phonon frequency, the limit where an attractive Hubbard model effectively describes the physics. △ Less

Submitted 8 July, 2021; v1 submitted 16 March, 2021; originally announced March 2021.

Comments: 9 pages, 6 figures

Journal ref: Phys. Rev. B 103, 235156 (2021)

Numerical approaches for calculating the low-field dc Hall coefficient of the doped Hubbard model

Authors: Wen O. Wang, Jixun K. Ding, Brian Moritz, Yoni Schattner, Edwin W. Huang, Thomas P. Devereaux

Abstract: Using determinant Quantum Monte Carlo, we compare three methods of evaluating the dc Hall coefficient $R_H$ of the Hubbard model: the direct measurement of the off-diagonal current-current correlator $χ_{xy}$ in a system coupled to a finite magnetic field (FF), $χ_{xy}^{\text{FF}}$; the three-current linear response to an infinitesimal field as measured in the zero-field (ZF) Hubbard Hamiltonian,… ▽ More Using determinant Quantum Monte Carlo, we compare three methods of evaluating the dc Hall coefficient $R_H$ of the Hubbard model: the direct measurement of the off-diagonal current-current correlator $χ_{xy}$ in a system coupled to a finite magnetic field (FF), $χ_{xy}^{\text{FF}}$; the three-current linear response to an infinitesimal field as measured in the zero-field (ZF) Hubbard Hamiltonian, $χ_{xy}^{\text{ZF}}$; and the leading order of the recurrent expansion $R_H^{(0)}$ in terms of thermodynamic susceptibilities. The two quantities $χ_{xy}^{\text{FF}}$ and $χ_{xy}^{\text{ZF}}$ can be compared directly in imaginary time. Proxies for $R_H$ constructed from the three-current correlator $χ_{xy}^{\text{ZF}}$ can be determined under different simplifying assumptions and compared with $R_H^{(0)}$. We find these different quantities to be consistent with one another, validating previous conclusions about the close correspondence between Fermi surface topology and the sign of $R_H$, even for strongly correlated systems. These various quantities also provide a useful set of numerical tools for testing theoretical predictions about the full behavior of the Hall conductivity for strong correlations. △ Less

Submitted 9 July, 2021; v1 submitted 8 March, 2021; originally announced March 2021.

Comments: 13 pages, 7 figures

Journal ref: Phys. Rev. Research 3, 033033 (2021)

Doped Mott Insulators Break $\mathbb Z_2$ Symmetry of a Fermi Liquid: Stability of Strongly Coupled Fixed Points

Authors: Edwin W. Huang, Gabriele La Nave, Philip W. Phillips

Abstract: Because Fermi liquids are inherently non-interacting states of matter, all electronic levels below the chemical potential are doubly occupied. Consequently, the simplest way of breaking Fermi liquid theory is to engineer a model in which some of those states are singly occupied keeping time-reversal invariance intact. We show that breaking an overlooked local-in-momentum space $\mathbb Z_2$ symmet… ▽ More Because Fermi liquids are inherently non-interacting states of matter, all electronic levels below the chemical potential are doubly occupied. Consequently, the simplest way of breaking Fermi liquid theory is to engineer a model in which some of those states are singly occupied keeping time-reversal invariance intact. We show that breaking an overlooked local-in-momentum space $\mathbb Z_2$ symmetry of a Fermi liquid does precisely this. As a result, while the Mott transition from a Fermi liquid is correctly believed to obtain without the breaking of any continuous symmetry, a discrete symmetry is broken. This symmetry breaking serves as an organizing principle for Mott physics whether it arises from the tractable Hatsugai-Kohmoto (HK) model or the intractable Hubbard model. That both are controlled by the same fixed point we establish through a renormalization group analysis. An experimental manifestation of this fixed point is the onset of particle-hole asymmetry, a widely observed phenomenon in strongly correlated systems. Theoretically, the singly-occupied region of the spectrum gives rise to a surface of zeros of the single-particle Green function, denoted as the Luttinger surface. Using K-homology, we show that the Bott topological invariant guarantees the stability of this surface to local perturbations. Our proof demonstrates that the strongly coupled fixed point only corresponds to those Luttinger surfaces with co-dimension $p+1$ with $p$ odd. We conclude that the Hubbard and HK models both lie in the same high temperature universality class and are controlled by the broken $\mathbb Z_2$ symmetry quartic fixed point. △ Less

Submitted 21 March, 2022; v1 submitted 4 March, 2021; originally announced March 2021.

Comments: Published version in Nature Physics

Journal ref: Nature Physics, xxx (2022)

Tendencies of enhanced electronic nematicity in the Hubbard model and a comparison with Raman scattering on high-temperature superconductors

Authors: Tianyi Liu, Daniel Jost, Brian Moritz, Edwin W. Huang, Rudi Hackl, Thomas P. Devereaux

Abstract: The pseudogap regime of the cuprate high-temperature superconductors is characterized by a variety of competing orders, the nature of which are still widely debated. Recent experiments have provided evidence for electron nematic order, in which the electron fluid breaks rotational symmetry while preserving translational invariance. Raman spectroscopy, with its ability to symmetry resolve low energ… ▽ More The pseudogap regime of the cuprate high-temperature superconductors is characterized by a variety of competing orders, the nature of which are still widely debated. Recent experiments have provided evidence for electron nematic order, in which the electron fluid breaks rotational symmetry while preserving translational invariance. Raman spectroscopy, with its ability to symmetry resolve low energy excitations, is a unique tool that can be used to assess nematic fluctuations and nematic ordering tendencies. Here, we compare results from determinant quantum Monte Carlo simulations of the Hubbard model to experimental results from Raman spectroscopy in $\text{La}_{2-x}\text{Sr}_{x}\text{CuO}_{4}$, which show a prominent increase in the $B_{1g}$ response around 10% hole doping as the temperature decreases, indicative of a rise in nematic fluctuations at low energy. Our results support a picture of nematic fluctuations with $B_{1g}$ symmetry occurring in underdoped cuprates, which may arise from melted stripes at elevated temperatures. △ Less

Submitted 13 April, 2021; v1 submitted 19 January, 2021; originally announced January 2021.

Comments: 8 pages, 7 figures

Journal ref: Phys. Rev. B 103, 134502 (2021)

Strong-coupling mechanism of the pseudogap in small Hubbard clusters

Authors: Edwin W. Huang

Abstract: In the hole-doped cuprates, the pseudogap refers to a suppression of the density of states at low energies, in the absence of superconducting long-range order. Numerous calculations of the Hubbard model show a pseudogap in the single-particle spectra, with striking similarities to photoemission and tunneling experiments on cuprates. However, no clear mechanism has been established. Here, we solve… ▽ More In the hole-doped cuprates, the pseudogap refers to a suppression of the density of states at low energies, in the absence of superconducting long-range order. Numerous calculations of the Hubbard model show a pseudogap in the single-particle spectra, with striking similarities to photoemission and tunneling experiments on cuprates. However, no clear mechanism has been established. Here, we solve the Hubbard model on $2\times2$ clusters by exact diagonalization, with integration over twisted boundary conditions. A pseudogap is found in the single-particle density of states with the following characteristics: a decreasing energy scale and onset temperature for increased hole-doping, closure at a critical hole doping near 15\%, absence upon electron-doping, particle-hole asymmetry indicated by the location of the gap center, and persistence in the strong-coupling limit of $U/t \to \infty$. Studying the many-body excitation spectrum reveals that the pseudogap in single-particle spectra is due to orthogonality between bare electrons and the lowest energy excitations for $U/t \gtrsim 8$. △ Less

Submitted 23 October, 2020; originally announced October 2020.

Extracting correlation effects from Momentum-Resolved Electron Energy Loss Spectroscopy (M-EELS): Synergistic origin of the dispersion kink in Bi$_{2.1}$Sr$_{1.9}$CaCu$_2$O$_{8+x}$

Authors: Edwin W. Huang, Kridsanaphong Limtragool, Chandan Setty, Ali A. Husain, Matteo Mitrano, Peter Abbamonte, Philip W. Phillips

Abstract: We employ Momentum-Resolved Electron Energy Loss Spectroscopy (M-EELS) on Bi2.1Sr1.9CaCu2O8+x to resolve the issue of the kink feature in the electron dispersion widely observed in the cuprates. To this end, we utilize the GW approximation to relate the density response function measured in in M-EELS to the self-energy, isolating contributions from phonons, electrons, and the momentum dependence o… ▽ More We employ Momentum-Resolved Electron Energy Loss Spectroscopy (M-EELS) on Bi2.1Sr1.9CaCu2O8+x to resolve the issue of the kink feature in the electron dispersion widely observed in the cuprates. To this end, we utilize the GW approximation to relate the density response function measured in in M-EELS to the self-energy, isolating contributions from phonons, electrons, and the momentum dependence of the effective interaction to the decay rates. The phononic contributions, present in the M-EELS spectra due to electron-phonon coupling, lead to kink features in the corresponding single-particle spectra at energies between 40 meV and 80 meV, independent of the doping level. We find that a repulsive interaction constant in momentum space is able to yield the kink attributed to phonons in ARPES. Hence, our analysis of the M-EELS spectra points to local repulsive interactions as a factor that enhances the spectroscopic signatures of electron-phonon coupling in cuprates. We conclude that the strength of the kink feature in cuprates is determined by the combined action of electron-phonon coupling and electron-electron interactions. △ Less

Submitted 6 October, 2020; originally announced October 2020.

Journal ref: Phys. Rev. B 103, 035121 (2021)

The Future of the Correlated Electron Problem

Authors: A. Alexandradinata, N. P. Armitage, Andrey Baydin, Wenli Bi, Yue Cao, Hitesh J. Changlani, Eli Chertkov, Eduardo H. da Silva Neto, Luca Delacretaz, Ismail El Baggari, G. M. Ferguson, William J. Gannon, Sayed Ali Akbar Ghorashi, Berit H. Goodge, Olga Goulko, G. Grissonnanche, Alannah Hallas, Ian M. Hayes, Yu He, Edwin W. Huang, Anshul Kogar, Divine Kumah, Jong Yeon Lee, A. Legros, Fahad Mahmood , et al. (22 additional authors not shown)

Abstract: A central problem in modern condensed matter physics is the understanding of materials with strong electron correlations. Despite extensive work, the essential physics of many of these systems is not understood and there is very little ability to make predictions in this class of materials. In this manuscript we share our personal views on the major open problems in the field of correlated electro… ▽ More A central problem in modern condensed matter physics is the understanding of materials with strong electron correlations. Despite extensive work, the essential physics of many of these systems is not understood and there is very little ability to make predictions in this class of materials. In this manuscript we share our personal views on the major open problems in the field of correlated electron systems. We discuss some possible routes to make progress in this rich and fascinating field. This manuscript is the result of the vigorous discussions and deliberations that took place at Johns Hopkins University during a three-day workshop January 27, 28, and 29, 2020 that brought together six senior scientists and 46 more junior scientists. Our hope, is that the topics we have presented will provide inspiration for others working in this field and motivation for the idea that significant progress can be made on very hard problems if we focus our collective energies. △ Less

Submitted 13 July, 2022; v1 submitted 1 October, 2020; originally announced October 2020.

Comments: 55 pages, 19 figures

Observation of Pines' Demon in Sr$_2$RuO$_4$

Authors: A. A. Husain, E. W. Huang, M. Mitrano, M. S. Rak, S. I. Rubeck, X. Guo, H. Yang, C. Sow, Y. Maeno, B. Uchoa, T. C. Chiang, P. E. Batson, P. W. Phillips, P. Abbamonte

Abstract: The characteristic excitation of a metal is its plasmon, which is a quantized collective oscillation of its electron density. In 1956, David Pines predicted that a distinct type of plasmon, dubbed a "demon," could exist in three-dimensional metals containing more than one species of charge carrier. Consisting of out-of-phase movement of electrons in different bands, demons are acoustic, electrical… ▽ More The characteristic excitation of a metal is its plasmon, which is a quantized collective oscillation of its electron density. In 1956, David Pines predicted that a distinct type of plasmon, dubbed a "demon," could exist in three-dimensional metals containing more than one species of charge carrier. Consisting of out-of-phase movement of electrons in different bands, demons are acoustic, electrically neutral, and do not couple to light, so have never been detected in an equilibrium, three-dimensional metal. Nevertheless, demons are believed to be critical for diverse phenomena including phase transitions in mixed-valence semimetals, optical properties of metal nanoparticles, "soundarons" in Weyl semimetals, and high temperature superconductivity in, for example, metal hydrides. Here, we present evidence for a demon in Sr$_2$RuO$_4$ from momentum-resolved electron energy-loss spectroscopy (M-EELS). Formed of electrons in the $β$ and $γ$ bands, the demon is gapless with a room temperature velocity $v=1.065 \pm 0.12 \times 10^5$ m/s and critical momentum $q_c=0.08$ reciprocal lattice units. Its spectral weight violates low-energy partial sum rules, affirming its neutral character. Our study confirms a 66-year old prediction and suggests that demons may be a pervasive feature of multiband metals. △ Less

Submitted 27 May, 2022; v1 submitted 13 July, 2020; originally announced July 2020.

Comments: 14 pages, 4 figures ; Revised to include a complete analysis of the gapless collective mode

DC Hall coefficient of the strongly correlated Hubbard model

Authors: Wen O. Wang, Jixun K. Ding, Brian Moritz, Edwin W. Huang, Thomas P. Devereaux

Abstract: The Hall coefficient is related to the effective carrier density and Fermi surface topology in noninteracting and weakly interacting systems. In strongly correlated systems, the relation between the Hall coefficient and single-particle properties is less clear. Clarifying this relation would give insight into the nature of transport in strongly correlated materials that lack well-formed quasiparti… ▽ More The Hall coefficient is related to the effective carrier density and Fermi surface topology in noninteracting and weakly interacting systems. In strongly correlated systems, the relation between the Hall coefficient and single-particle properties is less clear. Clarifying this relation would give insight into the nature of transport in strongly correlated materials that lack well-formed quasiparticles. In this work, we investigate the DC Hall coefficient of the Hubbard model using determinant quantum Monte Carlo in conjunction with a recently developed expansion of magneto-transport coefficients in terms of thermodynamic susceptibilities. At leading order in the expansion, we observe a change of sign in the Hall coefficient as a function of temperature and interaction strength, which we relate to a change in the topology of the apparent Fermi surface. We also combine our Hall coefficient results with optical conductivity values to evaluate the Hall angle, as well as effective mobility and effective mass based on Drude theory of metals. △ Less

Submitted 30 July, 2020; v1 submitted 27 February, 2020; originally announced February 2020.

Journal ref: npj Quantum Mater. 5, 51 (2020)

Exact Superconducting Instability in a Doped Mott Insulator

Authors: Philip W. Phillips, Luke Yeo, Edwin W. Huang

Abstract: Because the cuprate superconductors are doped Mott insulators, it would be advantageous to solve even a toy model that exhibits both Mottness and superconductivity. We consider the Hatsugai-Kohmoto model, an exactly solvable system that is a prototypical Mott insulator above a critical interaction strength at half filling. Upon doping or reducing the interaction strength, our exact calculations sh… ▽ More Because the cuprate superconductors are doped Mott insulators, it would be advantageous to solve even a toy model that exhibits both Mottness and superconductivity. We consider the Hatsugai-Kohmoto model, an exactly solvable system that is a prototypical Mott insulator above a critical interaction strength at half filling. Upon doping or reducing the interaction strength, our exact calculations show that the system becomes a non-Fermi liquid metal with a superconducting instability. In the presence of a weak pairing interaction, the instability produces a thermal transition to a superconducting phase, which is distinct from the BCS state, as evidenced by a gap-to-transition temperature ratio exceeding the universal BCS limit. The elementary excitations of this superconductor are not Bogoliubov quasiparticles but rather superpositions of doublons and holons, composite excitations signaling that the superconducting ground state of the doped Mott insulator inherits the non-Fermi liquid character of the normal state. An unexpected feature of this model is that it exhibits a superconductivity-induced transfer of spectral weight from high to low energies as seen in the cuprates as well as a suppression of the superfluid density relative to that in BCS theory. △ Less

Submitted 1 April, 2020; v1 submitted 2 December, 2019; originally announced December 2019.

Comments: 12 pages, 6 figures

Journal ref: Nature Physics 16, 1175-1180 (2020)

Emergence of Quasiparticles in a Doped Mott Insulator

Authors: Yao Wang, Yu He, Krzysztof Wohlfeld, Makoto Hashimoto, Edwin W. Huang, Donghui Lu, Sung-Kwan Mo, Seiki Komiya, Chunjing Jia, Brian Moritz, Zhi-Xun Shen, Thomas P. Devereaux

Abstract: How a Mott insulator develops into a weakly coupled metal upon doping is a central question to understanding various emergent correlated phenomena. To analyze this evolution and its connection to the high-$T_c$ cuprates, we study the single-particle spectrum for the doped Hubbard model using cluster perturbation theory on superclusters. Starting from extremely low doping, we identify a heavily ren… ▽ More How a Mott insulator develops into a weakly coupled metal upon doping is a central question to understanding various emergent correlated phenomena. To analyze this evolution and its connection to the high-$T_c$ cuprates, we study the single-particle spectrum for the doped Hubbard model using cluster perturbation theory on superclusters. Starting from extremely low doping, we identify a heavily renormalized quasiparticle dispersion that immediately develops across the Fermi level, and a weakening polaronic side band at higher binding energy. The quasiparticle spectral weight roughly grows at twice the rate of doping in the low doping regime, but this rate is halved at optimal doping. In the heavily doped regime, we find both strong electron-hole asymmetry and a persistent presence of Mott spectral features. Finally, we discuss the applicability of the single-band Hubbard model to describe the evolution of nodal spectra measured by angle-resolved photoemission spectroscopy (ARPES) on the single-layer cuprate La$_{2-x}$Sr$_x$CuO$_4$ ($0 \le x \le 0.15$). This work benchmarks the predictive power of the Hubbard model for electronic properties of high-$T_c$ cuprates. △ Less

Submitted 15 November, 2020; v1 submitted 19 September, 2019; originally announced September 2019.

Comments: 7 pages, 5 figures

Journal ref: Commun. Phys. 3, 210 (2020)

Predicting Onset of Dementia in Parkinson's Disease Patients

Authors: Dhruv Agarwal, Abhishek Srivastava, Edward W Huang

Abstract: Alzheimer's disease (AD) and Parkinson's disease (PD) are the two most common neurodegenerative disorders in humans. Because a significant percentage of patients have clinical and pathological features of both diseases, it has been hypothesized that the patho-cascades of the two diseases overlap. Despite this evidence, these two diseases are rarely studied in a joint manner. In this paper, we util… ▽ More Alzheimer's disease (AD) and Parkinson's disease (PD) are the two most common neurodegenerative disorders in humans. Because a significant percentage of patients have clinical and pathological features of both diseases, it has been hypothesized that the patho-cascades of the two diseases overlap. Despite this evidence, these two diseases are rarely studied in a joint manner. In this paper, we utilize clinical, imaging, genetic, and biospecimen features to cluster AD and PD patients into the same feature space. By training a machine learning classifier on the combined feature space, we predict the disease stage of patients two years after their baseline visits. We observed a considerable improvement in the prediction accuracy of Parkinson's dementia patients due to combined training on Alzheimer's and Parkinson's patients, thereby affirming the claim that these two diseases can be jointly studied. △ Less

Submitted 2 June, 2019; originally announced June 2019.

Comments: 7 pages

Pairing and non-Fermi liquid behavior in partially flat-band systems

Authors: Sharareh Sayyad, Edwin W. Huang, Motoharu Kitatani, Mohammad-Sadegh Vaezi, Zohar Nussinov, Abolhassan Vaezi, Hideo Aoki

Abstract: While multiband systems are usually considered for flat-band physics, here we study one-band models that have flat portions in the dispersion to explore correlation effects in the 2D repulsive Hubbard model in an intermediate coupling regime. The FLEX+DMFT~(the dynamical mean-field theory combined with the fluctuation exchange approximation) is used to show that we have a crossover from ferromagne… ▽ More While multiband systems are usually considered for flat-band physics, here we study one-band models that have flat portions in the dispersion to explore correlation effects in the 2D repulsive Hubbard model in an intermediate coupling regime. The FLEX+DMFT~(the dynamical mean-field theory combined with the fluctuation exchange approximation) is used to show that we have a crossover from ferromagnetic to antiferromagnetic spin fluctuations as the band filling is varied, which triggers a crossover from triplet to singlet pairings with a peculiar filling dependence that is dominated by the size of the flat region in the dispersion. A curious manifestation of the flat part appears as larger numbers of nodal lines associated with pairs extended in real space. We further detect non-Fermi liquid behavior in the momentum distribution function, frequency dependence of the self-energy and spectral function. These indicate correlation physics peculiar to flat-band systems. △ Less

Submitted 11 December, 2019; v1 submitted 23 March, 2019; originally announced March 2019.

Comments: 7+2 pages

Journal ref: Phys. Rev. B 101, 014501 (2020)

Angle-resolved photoemission spectroscopy of a Fermi-Hubbard system

Authors: Peter T. Brown, Elmer Guardado-Sanchez, Benjamin M. Spar, Edwin W. Huang, Thomas P. Devereaux, Waseem S. Bakr

Abstract: Angle-resolved photoemission spectroscopy (ARPES) measures the single-particle excitations of a many-body quantum system with both energy and momentum resolution, providing detailed information about strongly interacting materials. ARPES is a direct probe of fermion pairing, and hence a natural technique to study the development of superconductivity in a variety of experimental systems ranging fro… ▽ More Angle-resolved photoemission spectroscopy (ARPES) measures the single-particle excitations of a many-body quantum system with both energy and momentum resolution, providing detailed information about strongly interacting materials. ARPES is a direct probe of fermion pairing, and hence a natural technique to study the development of superconductivity in a variety of experimental systems ranging from high temperature superconductors to unitary Fermi gases. In these systems a remnant gap-like feature persists in the normal state, which is referred to as a pseudogap. A quantitative understanding of pseudogap regimes may elucidate details about the pairing mechanisms that lead to superconductivity, but developing this is difficult in real materials partly because the microscopic Hamiltonian is not known. Here we report on the development of ARPES to study strongly interacting fermions in an optical lattice using a quantum gas microscope. We benchmark the technique by measuring the occupied single-particle spectral function of an attractive Fermi-Hubbard system across the BCS-BEC crossover and comparing to quantum Monte Carlo calculations. We find evidence for a pseudogap in our system that opens well above the expected critical temperature for superfluidity. This technique may also be applied to the doped repulsive Hubbard model which is expected to exhibit a pseudogap at temperatures close to those achieved in recent experiments. △ Less

Submitted 13 March, 2019; originally announced March 2019.

Journal ref: Nature Physics (2019)

Enhanced correlations and superconductivity in weakly interacting partially flat band systems: a determinantal quantum Monte Carlo study

Authors: Edwin W. Huang, Mohammad-Sadegh Vaezi, Zohar Nussinov, Abolhassan Vaezi

Abstract: Motivated by recent experiments realizing correlated phenomena and superconductivity in 2D van der Waals devices, we consider the general problem of whether correlation effects may be enhanced by modifying band structure while keeping a fixed weak interaction strength. Using determinantal quantum Monte Carlo, we study the 2D Hubbard model for two different band structures: a regular nearest-neighb… ▽ More Motivated by recent experiments realizing correlated phenomena and superconductivity in 2D van der Waals devices, we consider the general problem of whether correlation effects may be enhanced by modifying band structure while keeping a fixed weak interaction strength. Using determinantal quantum Monte Carlo, we study the 2D Hubbard model for two different band structures: a regular nearest-neighbor tight-binding model, and a partially flat band structure containing a non-dispersing region, with identical total non-interacting bandwidth $W$. For both repulsive and attractive weak interactions ($|U| \ll W$), correlated phenomena are significantly stronger in the partially flat model. In the repulsive case, even with $U$ an order of magnitude smaller than $W$, we find the presence of a Mott insulating state near half-filling of the flat region in momentum space. In the attractive case, where generically the ground state is superconducting, the partially flat model exhibits significantly enhanced superconducting transition temperatures. These results suggest the possibility of engineering correlation effects in materials by tuning the non-interacting electronic dispersion. △ Less

Submitted 21 November, 2018; originally announced November 2018.

Journal ref: Phys. Rev. B 99, 235128 (2019)

Fermi surface reconstruction in electron-doped cuprates without antiferromagnetic long-range order

Authors: J. -F. He, C. R. Rotundu, M. S. Scheurer, Y. He, M. Hashimoto, K. Xu, Y. Wang, E. W. Huang, T. Jia, S. -D. Chen, B. Moritz, D. -H. Lu, Y. S. Lee, T. P. Devereaux, Z. -X. Shen

Abstract: Fermi surface (FS) topology is a fundamental property of metals and superconductors. In electron-doped cuprate Nd2-xCexCuO4 (NCCO), an unexpected FS reconstruction has been observed in optimal- and over-doped regime (x=0.15-0.17) by quantum oscillation measurements (QOM). This is all the more puzzling because neutron scattering suggests that the antiferromagnetic (AFM) long-range order, which is b… ▽ More Fermi surface (FS) topology is a fundamental property of metals and superconductors. In electron-doped cuprate Nd2-xCexCuO4 (NCCO), an unexpected FS reconstruction has been observed in optimal- and over-doped regime (x=0.15-0.17) by quantum oscillation measurements (QOM). This is all the more puzzling because neutron scattering suggests that the antiferromagnetic (AFM) long-range order, which is believed to reconstruct the FS, vanishes before x=0.14. To reconcile the conflict, a widely discussed external magnetic field-induced AFM long-range order in QOM explains the FS reconstruction as an extrinsic property. Here, we report angle-resolved photoemission (ARPES) evidence of FS reconstruction in optimal- and over-doped NCCO. The observed FSs are in quantitative agreement with QOM, suggesting an intrinsic FS reconstruction without field. This reconstructed FS, despite its importance as a basis to understand electron-doped cuprates, cannot be explained under the traditional scheme. Furthermore, the energy gap of the reconstruction decreases rapidly near x=0.17 like an order parameter, echoing the quantum critical doping in transport. The totality of the data points to a mysterious order between x=0.14 and 0.17, whose appearance favors the FS reconstruction and disappearance defines the quantum critical doping. A recent topological proposal provides an ansatz for its origin. △ Less

Submitted 12 November, 2018; originally announced November 2018.

Journal ref: PNAS 116, 3449-3453 (2019)

Three dimensional collective charge excitations in electron-doped cuprate superconductors

Authors: M. Hepting, L. Chaix, E. W. Huang, R. Fumagalli, Y. Y. Peng, B. Moritz, K. Kummer, N. B. Brookes, W. C. Lee, M. Hashimoto, T. Sarkar, J. F. He, C. R. Rotundu, Y. S. Lee, R. L. Greene, L. Braicovich, G. Ghiringhelli, Z. X. Shen, T. P. Devereaux, W. S. Lee

Abstract: High temperature cuprate superconductors consist of stacked CuO2 planes, with primarily two dimensional electronic band structures and magnetic excitations, while superconducting coherence is three dimensional. This dichotomy highlights the importance of out-of-plane charge dynamics, believed to be incoherent in the normal state, yet lacking a comprehensive characterization in energy-momentum spac… ▽ More High temperature cuprate superconductors consist of stacked CuO2 planes, with primarily two dimensional electronic band structures and magnetic excitations, while superconducting coherence is three dimensional. This dichotomy highlights the importance of out-of-plane charge dynamics, believed to be incoherent in the normal state, yet lacking a comprehensive characterization in energy-momentum space. Here, we use resonant inelastic x-ray scattering (RIXS) with polarization analysis to uncover the pure charge character of a recently discovered collective mode in electron-doped cuprates. This mode disperses along both the in- and, importantly, out-of-plane directions, revealing its three dimensional nature. The periodicity of the out-of-plane dispersion corresponds to the CuO2 plane distance rather than the crystallographic c-axis lattice constant, suggesting that the interplane Coulomb interaction drives the coherent out-of-plane charge dynamics. The observed properties are hallmarks of the long-sought acoustic plasmon, predicted for layered systems and argued to play a substantial role in mediating high temperature superconductivity. △ Less

Submitted 29 October, 2018; originally announced October 2018.

Comments: This is the version of first submission. The revised manuscript according to peer reviews is now accepted by Nature and will be published online on 31st Oct., 2018

Biexciton Condensation in Electron-hole Doped Hubbard Bilayers -- A Sign-Problem-Free Quantum Monte Carlo Study

Authors: Xuxin Huang, Martin Claassen, Edwin W. Huang, Brian Moritz, Thomas P. Devereaux

Abstract: The bilayer Hubbard model with electron-hole doping is an ideal platform to study excitonic orders due to suppressed recombination via spatial separation of electrons and holes. However, suffering from the sign problem, previous quantum Monte Carlo studies could not arrive at an unequivocal conclusion regarding the presence of phases with clear signatures of excitonic condensation in bilayer Hubba… ▽ More The bilayer Hubbard model with electron-hole doping is an ideal platform to study excitonic orders due to suppressed recombination via spatial separation of electrons and holes. However, suffering from the sign problem, previous quantum Monte Carlo studies could not arrive at an unequivocal conclusion regarding the presence of phases with clear signatures of excitonic condensation in bilayer Hubbard models. Here, we develop a determinant quantum Monte Carlo (DQMC) algorithm for the bilayer Hubbard model that is sign-problem-free for equal and opposite doping in the two layers, and study excitonic order and charge and spin density modulations as a function of chemical potential difference between the two layers, on-site Coulomb repulsion, and inter-layer interaction. In the intermediate coupling regime and in proximity to the SU(4)-symmetric point, we find a biexcitonic condensate phase at finite electron-hole doping, as well as a competing $(π,π)$ charge density wave (CDW) state. We extract the Berezinskii-Kosterlitz-Thouless (BKT) transition temperature from superfluid density and a finite size scaling analysis of the correlation functions, and explain our results in terms of an effective biexcitonic hardcore boson model. △ Less

Submitted 17 July, 2019; v1 submitted 17 September, 2018; originally announced September 2018.

Comments: 6 pages, 4 figures

Journal ref: Phys. Rev. Lett. 124, 077601 (2020)

Dispersion, damping, and intensity of spin excitations in the single-layer (Bi,Pb)$_{2}$(Sr,La)$_{2}$CuO$_{6+δ}$ cuprate superconductor family

Authors: Y. Y. Peng, E. W. Huang, R. Fumagalli, M. Minola, Y. Wang, X. Sun, Y. Ding, K. Kummer, X. J. Zhou, N. B. Brookes, B. Moritz, L. Braicovich, T. P. Devereaux, G. Ghiringhelli

Abstract: Using Cu-$L_3$ edge resonant inelastic x-ray scattering (RIXS) we measured the dispersion and damping of spin excitations (magnons and paramagnons) in the high-$T_\mathrm{c}$ superconductor (Bi,Pb)$_{2}$(Sr,La)$_{2}$CuO$_{6+δ}$ (Bi2201), for a large doping range across the phase diagram ($0.03\lesssim p\lesssim0.21$). Selected measurements with full polarization analysis unambiguously demonstrate… ▽ More Using Cu-$L_3$ edge resonant inelastic x-ray scattering (RIXS) we measured the dispersion and damping of spin excitations (magnons and paramagnons) in the high-$T_\mathrm{c}$ superconductor (Bi,Pb)$_{2}$(Sr,La)$_{2}$CuO$_{6+δ}$ (Bi2201), for a large doping range across the phase diagram ($0.03\lesssim p\lesssim0.21$). Selected measurements with full polarization analysis unambiguously demonstrate the spin-flip character of these excitations, even in the overdoped sample. We find that the undamped frequencies increase slightly with doping for all accessible momenta, while the damping grows rapidly, faster in the (0,0)$\rightarrow$(0.5,0.5) nodal direction than in the (0,0)$\rightarrow$(0.5,0) antinodal direction. We compare the experimental results to numerically exact determinant quantum Monte Carlo (DQMC) calculations that provide the spin dynamical structure factor $S(\textbf{Q},ω)$ of the three-band Hubbard model. The theory reproduces well the momentum and doping dependence of the dispersions and spectral weights of magnetic excitations. These results provide compelling evidence that paramagnons, although increasingly damped, persist across the superconducting dome of the cuprate phase diagram; this implies that long range antiferromagnetic correlations are quickly washed away, while short range magnetic interactions are little affected by doping. △ Less

Submitted 5 August, 2018; originally announced August 2018.

Comments: 11 pages, 9 figures

Journal ref: Phys. Rev. B 98, 144507 (2018)

Strange metallicity in the doped Hubbard model

Authors: Edwin W. Huang, Ryan Sheppard, Brian Moritz, Thomas P. Devereaux

Abstract: Strange or bad metallic transport, defined by its incompatibility with conventional quasiparticle pictures, is a theme common to strongly correlated materials and ubiquitous in many high temperature superconductors. The Hubbard model represents a minimal starting point for modeling strongly correlated systems. Here we demonstrate strange metallic transport in the doped two-dimensional Hubbard mode… ▽ More Strange or bad metallic transport, defined by its incompatibility with conventional quasiparticle pictures, is a theme common to strongly correlated materials and ubiquitous in many high temperature superconductors. The Hubbard model represents a minimal starting point for modeling strongly correlated systems. Here we demonstrate strange metallic transport in the doped two-dimensional Hubbard model using determinantal quantum Monte Carlo calculations. Over a wide range of doping, we observe resistivities exceeding the Mott-Ioffe-Regel limit with linear temperature dependence. The temperatures of our calculations extend to as low as 1/40 the non-interacting bandwidth, placing our findings in the degenerate regime relevant to experimental observations of strange metallicity. Our results provide a foundation for connecting theories of strange metals to models of strongly correlated materials. △ Less

Submitted 16 January, 2019; v1 submitted 21 June, 2018; originally announced June 2018.

Journal ref: Science 366, 987-990 (2019)

RIXS studies of magnons and bimagnons in the lightly doped cuprate La2-xSrxCuO4

Authors: L. Chaix, E. W. Huang, S. Gerber, X. Lu, C. Jia, Y. Huang, D. E. McNally, Y. Wang, F. H. Vernay, A. Keren, M. Shi, B. Moritz, Z. -X. Shen, T. Schmitt, T. P. Devereaux, W. -S. Lee

Abstract: We investigated the doping dependence of magnetic excitations in the lightly doped cuprate La2-xSrxCuO4 via combined studies of resonant inelastic x-ray scattering (RIXS) at the Cu L3-edge and theoretical calculations. With increasing doping, the magnon dispersion is found to be essentially unchanged, but the spectral width broadens and the spectral weight varies differently at different momenta.… ▽ More We investigated the doping dependence of magnetic excitations in the lightly doped cuprate La2-xSrxCuO4 via combined studies of resonant inelastic x-ray scattering (RIXS) at the Cu L3-edge and theoretical calculations. With increasing doping, the magnon dispersion is found to be essentially unchanged, but the spectral width broadens and the spectral weight varies differently at different momenta. Near the Brillouin zone center, we directly observe bimagnon excitations which possess the same energy scale and doping dependence as previously observed by Raman spectroscopy. They disperse weakly in energy-momentum space, and are consistent with a bimagnon dispersion that is renormalized by the magnon-magnon interaction at the zone center. △ Less

Submitted 2 May, 2018; originally announced May 2018.

Comments: 8 pages, 6 figures

Journal ref: Phys. Rev. B 97, 155144 (2018)

Magnon Splitting Induced by Charge Transfer in the Three-Orbital Hubbard Model

Authors: Yao Wang, Edwin W. Huang, Brian Moritz, Thomas P. Devereaux

Abstract: Understanding spin excitations and their connection to unconventional superconductivity have remained a central issue since the discovery of the cuprates. Direct measurement of the dynamical spin structure factor in the parent compounds can provide key information on important interactions relevant in the doped regime, and variations in the magnon dispersion have been linked closely to differences… ▽ More Understanding spin excitations and their connection to unconventional superconductivity have remained a central issue since the discovery of the cuprates. Direct measurement of the dynamical spin structure factor in the parent compounds can provide key information on important interactions relevant in the doped regime, and variations in the magnon dispersion have been linked closely to differences in crystal structure between families of cuprate compounds. Here, we elucidate the relationship between spin excitations and various controlling factors thought to be significant in high-$T_c$ materials by systematically evaluating the dynamical spin structure factor for the three-orbital Hubbard model, revealing differences in the spin dispersion along the Brillouin zone axis and the diagonal. Generally, we find that the absolute energy scale and momentum dependence of the excitations primarily are sensitive to the effective charge-transfer energy, while changes in the on-site Coulomb interactions have little effect on the details of the dispersion. In particular, our result highlights the splitting between spin excitations along the axial and diagonal directions in the Brillouin zone. This splitting decreases with increasing charge-transfer energy and correlates with changes in the apical oxygen position, and general structural variations, for different cuprate families. △ Less

Submitted 11 June, 2018; v1 submitted 13 November, 2017; originally announced November 2017.

Comments: 5 pages, 4 figures

Journal ref: Phys. Rev. Lett. 120, 246401 (2018)

Breakdown of Migdal-Eliashberg theory; a determinant quantum Monte Carlo study

Authors: Ilya Esterlis, Benjamin Nosarzewski, Edwin W. Huang, Brian Moritz, Thomas P. Devereaux, Douglas J. Scalapino, Steven A. Kivelson

Abstract: The superconducting (SC) and charge-density-wave (CDW) susceptibilities of the two dimensional Holstein model are computed using determinant quantum Monte Carlo (DQMC), and compared with results computed using the Migdal-Eliashberg (ME) approach. We access temperatures as low as 25 times less than the Fermi energy, $E_F$, which are still above the SC transition. We find that the SC susceptibility… ▽ More The superconducting (SC) and charge-density-wave (CDW) susceptibilities of the two dimensional Holstein model are computed using determinant quantum Monte Carlo (DQMC), and compared with results computed using the Migdal-Eliashberg (ME) approach. We access temperatures as low as 25 times less than the Fermi energy, $E_F$, which are still above the SC transition. We find that the SC susceptibility at low $T$ agrees quantitatively with the ME theory up to a dimensionless electron-phonon coupling $λ_0 \approx 0.4$ but deviates dramatically for larger $λ_0$. We find that for large $λ_0$ and small phonon frequency $ω_0 \ll E_F$ CDW ordering is favored and the preferred CDW ordering vector is uncorrelated with any obvious feature of the Fermi surface. △ Less

Submitted 19 February, 2018; v1 submitted 4 November, 2017; originally announced November 2017.

Comments: 7 pages, 7 figures. v2: discussion improved, added supplemental material

Journal ref: Phys. Rev. B 97, 140501 (2018)

Stripe order from the perspective of the Hubbard model

Authors: Edwin W. Huang, Christian B. Mendl, Hong-Chen Jiang, Brian Moritz, Thomas P. Devereaux

Abstract: A microscopic understanding of the strongly correlated physics of the cuprates must account for the translational and rotational symmetry breaking that is present across all cuprate families, commonly in the form of stripes. Here we investigate emergence of stripes in the Hubbard model, a minimal model believed to be relevant to the cuprate superconductors, using determinant quantum Monte Carlo (D… ▽ More A microscopic understanding of the strongly correlated physics of the cuprates must account for the translational and rotational symmetry breaking that is present across all cuprate families, commonly in the form of stripes. Here we investigate emergence of stripes in the Hubbard model, a minimal model believed to be relevant to the cuprate superconductors, using determinant quantum Monte Carlo (DQMC) simulations at finite temperatures and density matrix renormalization group (DMRG) ground state calculations. By varying temperature, doping, and model parameters, we characterize the extent of stripes throughout the phase diagram of the Hubbard model. Our results show that including the often neglected next-nearest-neighbor hopping leads to the absence of spin incommensurability upon electron-doping and nearly half-filled stripes upon hole-doping. The similarities of these findings to experimental results on both electron and hole-doped cuprate families support a unified description across a large portion of the cuprate phase diagram. △ Less

Submitted 20 April, 2018; v1 submitted 7 September, 2017; originally announced September 2017.

Journal ref: npj Quantum Materials 3, 22 (2018)

Doping dependence of ordered phases and emergent quasiparticles in the doped Hubbard-Holstein model

Authors: Christian B. Mendl, Elizabeth A. Nowadnick, Edwin W. Huang, Steven Johnston, Brian Moritz, Thomas P. Devereaux

Abstract: We present determinant quantum Monte Carlo simulations of the hole-doped single-band Hubbard-Holstein model on a square lattice, to investigate how quasiparticles emerge when doping a Mott insulator (MI) or a Peierls insulator (PI). The MI regime at large Hubbard interaction $U$ and small relative electron-phonon coupling strength $λ$ is quickly suppressed upon doping, by drawing spectral weight f… ▽ More We present determinant quantum Monte Carlo simulations of the hole-doped single-band Hubbard-Holstein model on a square lattice, to investigate how quasiparticles emerge when doping a Mott insulator (MI) or a Peierls insulator (PI). The MI regime at large Hubbard interaction $U$ and small relative electron-phonon coupling strength $λ$ is quickly suppressed upon doping, by drawing spectral weight from the upper Hubbard band and shifting the lower Hubbard band towards the Fermi level, leading to a metallic state with emergent quasiparticles at the Fermi level. On the other hand, the PI regime at large $λ$ and small $U$ persists out to relatively high doping levels. We study the evolution of the $d$-wave superconducting susceptibility with doping, and find that it increases with lowering temperature in a regime of intermediate values of $U$ and $λ$. △ Less

Submitted 1 September, 2017; originally announced September 2017.

Comments: 7 pages, 5 figures

Journal ref: Phys. Rev. B 96, 205141 (2017)

Decrease of d-wave pairing strength in spite of the persistence of magnetic excitations in the overdoped Hubbard model

Authors: Edwin W. Huang, Douglas J. Scalapino, Thomas A. Maier, Brian Moritz, Thomas P. Devereaux

Abstract: Evidence for the presence of high energy magnetic excitations in overdoped La$_{2-x}$Sr$_x$CuO$_4$ (LSCO) has raised questions regarding the role of spin-fluctuations in the pairing mechanism. If they remain present in overdoped LSCO, why does $T_c$ decrease in this doping regime? Here, using results for the dynamic spin susceptibility ${\rm Im}χ(q,ω)$ obtained from a determinantal quantum Monte C… ▽ More Evidence for the presence of high energy magnetic excitations in overdoped La$_{2-x}$Sr$_x$CuO$_4$ (LSCO) has raised questions regarding the role of spin-fluctuations in the pairing mechanism. If they remain present in overdoped LSCO, why does $T_c$ decrease in this doping regime? Here, using results for the dynamic spin susceptibility ${\rm Im}χ(q,ω)$ obtained from a determinantal quantum Monte Carlo (DQMC) calculation for the Hubbard model we address this question. We find that while high energy magnetic excitations persist in the overdoped regime, they lack the momentum to scatter pairs between the anti-nodal regions. It is the decrease in the spectral weight at large momentum transfer, not observed by resonant inelastic X-ray scattering (RIXS), which leads to a reduction in the $d$-wave spin-fluctuation pairing strength. △ Less

Submitted 10 May, 2017; originally announced May 2017.

Journal ref: Phys. Rev. B 96, 020503 (2017)