-
Scalable Exact Inference in Multi-Output Gaussian Processes
Authors:
Wessel P. Bruinsma,
Eric Perim,
Will Tebbutt,
J. Scott Hosking,
Arno Solin,
Richard E. Turner
Abstract:
Multi-output Gaussian processes (MOGPs) leverage the flexibility and interpretability of GPs while capturing structure across outputs, which is desirable, for example, in spatio-temporal modelling. The key problem with MOGPs is their computational scaling $O(n^3 p^3)$, which is cubic in the number of both inputs $n$ (e.g., time points or locations) and outputs $p$. For this reason, a popular class…
▽ More
Multi-output Gaussian processes (MOGPs) leverage the flexibility and interpretability of GPs while capturing structure across outputs, which is desirable, for example, in spatio-temporal modelling. The key problem with MOGPs is their computational scaling $O(n^3 p^3)$, which is cubic in the number of both inputs $n$ (e.g., time points or locations) and outputs $p$. For this reason, a popular class of MOGPs assumes that the data live around a low-dimensional linear subspace, reducing the complexity to $O(n^3 m^3)$. However, this cost is still cubic in the dimensionality of the subspace $m$, which is still prohibitively expensive for many applications. We propose the use of a sufficient statistic of the data to accelerate inference and learning in MOGPs with orthogonal bases. The method achieves linear scaling in $m$ in practice, allowing these models to scale to large $m$ without sacrificing significant expressivity or requiring approximation. This advance opens up a wide range of real-world tasks and can be combined with existing GP approximations in a plug-and-play way. We demonstrate the efficacy of the method on various synthetic and real-world data sets.
△ Less
Submitted 17 July, 2020; v1 submitted 14 November, 2019;
originally announced November 2019.
-
GP-ALPS: Automatic Latent Process Selection for Multi-Output Gaussian Process Models
Authors:
Pavel Berkovich,
Eric Perim,
Wessel Bruinsma
Abstract:
A simple and widely adopted approach to extend Gaussian processes (GPs) to multiple outputs is to model each output as a linear combination of a collection of shared, unobserved latent GPs. An issue with this approach is choosing the number of latent processes and their kernels. These choices are typically done manually, which can be time consuming and prone to human biases. We propose Gaussian Pr…
▽ More
A simple and widely adopted approach to extend Gaussian processes (GPs) to multiple outputs is to model each output as a linear combination of a collection of shared, unobserved latent GPs. An issue with this approach is choosing the number of latent processes and their kernels. These choices are typically done manually, which can be time consuming and prone to human biases. We propose Gaussian Process Automatic Latent Process Selection (GP-ALPS), which automatically chooses the latent processes by turning off those that do not meaningfully contribute to explaining the data. We develop a variational inference scheme, assess the quality of the variational posterior by comparing it against the gold standard MCMC, and demonstrate the suitability of GP-ALPS in a set of preliminary experiments.
△ Less
Submitted 17 December, 2019; v1 submitted 5 November, 2019;
originally announced November 2019.
-
AFLOW-CHULL: Cloud-oriented platform for autonomous phase stability analysis
Authors:
Corey Oses,
Eric Gossett,
David Hicks,
Frisco Rose,
Michael J. Mehl,
Eric Perim,
Ichiro Takeuchi,
Stefano Sanvito,
Matthias Scheffler,
Yoav Lederer,
Ohad Levy,
Cormac Toher,
Stefano Curtarolo
Abstract:
$\textit{A priori}$ prediction of phase stability of materials is a challenging practice, requiring knowledge of all energetically-competing structures at formation conditions. Large materials repositories $\unicode{x2014}$ housing properties of both experimental and hypothetical compounds $\unicode{x2014}$ offer a path to prediction through the construction of informatics-based, $\textit{ab-initi…
▽ More
$\textit{A priori}$ prediction of phase stability of materials is a challenging practice, requiring knowledge of all energetically-competing structures at formation conditions. Large materials repositories $\unicode{x2014}$ housing properties of both experimental and hypothetical compounds $\unicode{x2014}$ offer a path to prediction through the construction of informatics-based, $\textit{ab-initio}$ phase diagrams. However, limited access to relevant data and software infrastructure has rendered thermodynamic characterizations largely peripheral, despite their continued success in dictating synthesizability. Herein, a new module is presented for autonomous thermodynamic stability analysis implemented within the open-source, $\textit{ab-initio}$ framework AFLOW. Powered by the AFLUX Search-API, AFLOW-CHULL leverages data of more than 1.8 million compounds currently characterized in the AFLOW.org repository and can be employed locally from any UNIX-like computer. The module integrates a range of functionality: the identification of stable phases and equivalent structures, phase coexistence, measures for robust stability, and determination of decomposition reactions. As a proof-of-concept, thorough thermodynamic characterizations have been performed for more than 1,300 binary and ternary systems, enabling the identification of several candidate phases for synthesis based on their relative stability criterion $\unicode{x2014}$ including 18 promising $C15_{b}$-type structures and two half-Heuslers. In addition to a full report included herein, an interactive, online web application has been developed showcasing the results of the analysis, and is located at aflow.org/aflow-chull.
△ Less
Submitted 18 June, 2018;
originally announced June 2018.
-
Mechanical Properties of Schwarzites - A Fully Atomistic Reactive Molecular Dynamics Investigation
Authors:
Cristiano F. Woellner,
Tiago Botari,
Eric Perim,
Douglas S. Galvao
Abstract:
Schwarzites are crystalline, 3D porous structures with stable negative curvature formed of sp2-hybridized carbon atoms. These structures present topologies with tunable porous size and shape and unusual mechanical properties. In this work, we have investigated the mechanical behavior under compressive strains and energy absorption of four different Schwarzites, through reactive molecular dynamics…
▽ More
Schwarzites are crystalline, 3D porous structures with stable negative curvature formed of sp2-hybridized carbon atoms. These structures present topologies with tunable porous size and shape and unusual mechanical properties. In this work, we have investigated the mechanical behavior under compressive strains and energy absorption of four different Schwarzites, through reactive molecular dynamics simulations, using the ReaxFF force field as available in the LAMMPS code. We considered two Schwarzites families, the so-called Gyroid and Primitive and two structures from each family. Our results also show they exhibit remarkable resilience under mechanical compression. They can be reduced to half of their original size before structural failure (fracture) occurs.
△ Less
Submitted 17 January, 2018;
originally announced January 2018.
-
The AFLOW Fleet for Materials Discovery
Authors:
Cormac Toher,
Corey Oses,
David Hicks,
Eric Gossett,
Frisco Rose,
Pinku Nath,
Demet Usanmaz,
Denise C. Ford,
Eric Perim,
Camilo E. Calderon,
Jose J. Plata,
Yoav Lederer,
Michal Jahnátek,
Wahyu Setyawan,
Shidong Wang,
Junkai Xue,
Kevin Rasch,
Roman V. Chepulskii,
Richard H. Taylor,
Geena Gomez,
Harvey Shi,
Andrew R. Supka,
Rabih Al Rahal Al Orabi,
Priya Gopal,
Frank T. Cerasoli
, et al. (26 additional authors not shown)
Abstract:
The traditional paradigm for materials discovery has been recently expanded to incorporate substantial data driven research. With the intent to accelerate the development and the deployment of new technologies, the AFLOW Fleet for computational materials design automates high-throughput first principles calculations, and provides tools for data verification and dissemination for a broad community…
▽ More
The traditional paradigm for materials discovery has been recently expanded to incorporate substantial data driven research. With the intent to accelerate the development and the deployment of new technologies, the AFLOW Fleet for computational materials design automates high-throughput first principles calculations, and provides tools for data verification and dissemination for a broad community of users. AFLOW incorporates different computational modules to robustly determine thermodynamic stability, electronic band structures, vibrational dispersions, thermo-mechanical properties and more. The AFLOW data repository is publicly accessible online at aflow.org, with more than 1.7 million materials entries and a panoply of queryable computed properties. Tools to programmatically search and process the data, as well as to perform online machine learning predictions, are also available.
△ Less
Submitted 1 December, 2017;
originally announced December 2017.
-
Mechanical Properties and Fracture Patterns of Graphene (Graphitic) Nanowiggles
Authors:
Rafael A. Bizao,
Tiago Botari,
Eric Perim,
Nicola M. Pugno,
Douglas S. Galvao
Abstract:
Graphene nanowiggles (GNW) are graphene-based nanostructures obtained by making alternated regular cuts in pristine graphene nanoribbons. GNW were recently synthesized and it was demonstrated that they exhibit tunable electronic and magnetic properties by just varying their shape. Here, we have investigated the mechanical properties and fracture patterns of a large number of GNW of different shape…
▽ More
Graphene nanowiggles (GNW) are graphene-based nanostructures obtained by making alternated regular cuts in pristine graphene nanoribbons. GNW were recently synthesized and it was demonstrated that they exhibit tunable electronic and magnetic properties by just varying their shape. Here, we have investigated the mechanical properties and fracture patterns of a large number of GNW of different shapes and sizes using fully atomistic reactive molecular dynamics simulations. Our results show that the GNW mechanical properties are strongly dependent on its shape and size and, as a general trend narrow sheets have larger ultimate strength and Young's modulus than wide ones. The estimated Young's modulus values were found to be in a range of ~ 100-1000 GPa and the ultimate strength in a range of ~ 20-110 GPa, depending on GNW shape. Also, super-ductile behaviour under strain was observed for some structures.
△ Less
Submitted 3 February, 2017;
originally announced February 2017.
-
Spectral descriptors for bulk metallic glasses based on the thermodynamics of competing crystalline phases
Authors:
Eric Perim,
Dongwoo Lee,
Yanhui Liu,
Cormac Toher,
Pan Gong,
Yanglin Li,
W. Neal Simmons,
Ohad Levy,
Joost J. Vlassak,
Jan Schroers,
Stefano Curtarolo
Abstract:
Metallic glasses have attracted considerable interest in recent years due to their unique combination of superb properties and processability. Predicting bulk metallic glass formers from known parameters remains a challenge and the search for new systems is still performed by trial and error. It has been speculated that some sort of "confusion" during crystallization of the crystalline phases comp…
▽ More
Metallic glasses have attracted considerable interest in recent years due to their unique combination of superb properties and processability. Predicting bulk metallic glass formers from known parameters remains a challenge and the search for new systems is still performed by trial and error. It has been speculated that some sort of "confusion" during crystallization of the crystalline phases competing with glass formation could play a key role. Here, we propose a heuristic descriptor quantifying confusion and demonstrate its validity by detailed experiments on two well-known glass forming alloy systems. With the insight provided by these results, we develop a robust model for predicting glass formation ability based on the spectral decomposition of geometrical and energetic features of crystalline phases calculated ab-initio in the AFLOW high throughput framework. Our findings indicate that the formation of metallic glass phases could be a much more common phenomenon than currently estimated, with more than 17% of binary alloy systems being potential glass formers. Our approach is capable of pinpointing favorable compositions, overcoming a major bottleneck hindering the discovery of new materials. Hence, it is demonstrated that smart descriptors, based solely on the energetics and structure of competing crystalline phases calculated from first-principles and available in online databases, others the sought-after key for accelerated discovery of novel metallic glasses.
△ Less
Submitted 16 July, 2016; v1 submitted 3 June, 2016;
originally announced June 2016.
-
Mechanical and Structural Properties of Graphene-like Carbon Nitride Sheets
Authors:
J. M. de Sousa,
T. Botari,
E. Perim,
R. A. Bizao,
Douglas S. Galvao
Abstract:
Carbon nitride-based nanostructures have attracted special attention (from theory and experiments) due to their remarkable electromechanical properties. In this work we have investigated the mechanical properties of some graphene-like carbon nitride membranes through fully atomistic reactive molecular dynamics simulations. We have analyzed three different structures of these CN families, the so-ca…
▽ More
Carbon nitride-based nanostructures have attracted special attention (from theory and experiments) due to their remarkable electromechanical properties. In this work we have investigated the mechanical properties of some graphene-like carbon nitride membranes through fully atomistic reactive molecular dynamics simulations. We have analyzed three different structures of these CN families, the so-called graphene-based g-CN, triazine-based g-C3N4 and heptazine-based g-C3N4. The stretching dynamics of these membranes was studied for deformations along their two main axes and at three different temperatures: 10K, 300K and 600K. We show that g-CN membranes have the lowest ultimate fracture strain value, followed by heptazine-based and triazine-based ones, respectively. This behavior can be explained in terms of their differences in terms of density values, topologies and types of chemical bonds. The dependency of the fracture patterns on the stretching directions is also discussed.
△ Less
Submitted 3 June, 2016;
originally announced June 2016.
-
Predicting Bulk Metallic Glass Forming Ability with the Thermodynamic Density of Competing Crystalline States
Authors:
Eric Perim,
Dongwoo Lee,
Yanhui Liu,
Cormac Toher,
Pan Gong,
Yanglin Li,
W. Neal Simmons,
Ohad Levy,
Joost Vlassak,
Jan Schroers,
Stefano Curtarolo
Abstract:
Despite two decades of studies, the formation of metallic glasses, very promising systems for industrial applications, still remains mostly unexplained. This lack of knowledge hinders the search for new systems, still performed with combinatorial trial and error. In the past, it was speculated that some sort of "confusion" during crystallization could play a key role during their formation. In thi…
▽ More
Despite two decades of studies, the formation of metallic glasses, very promising systems for industrial applications, still remains mostly unexplained. This lack of knowledge hinders the search for new systems, still performed with combinatorial trial and error. In the past, it was speculated that some sort of "confusion" during crystallization could play a key role during their formation. In this article, we propose a heuristic descriptor quantifying such confusion. It is based on the "thermodynamic density of competing crystalline states", parameterized from high-throughput ab-initio calculations. The existence of highly enthalpy-degenerate but geometrically-different phases frustrates the crystallization process and promotes glass formation. Two test beds are considered. A good and a bad glass-former, CuZr and NiZr, are experimentally characterized with high-throughput synthesis. The experimental results corroborate the capability of the heuristic descriptor in predicting glass forming ability through the compositional space. Our analysis is expected to deepen the understanding of the underlying mechanisms and to accelerate the discovery of novel metallic glasses.
△ Less
Submitted 1 February, 2016; v1 submitted 29 January, 2016;
originally announced January 2016.
-
Novel Nanoscroll Structures from Carbon Nitride Layers
Authors:
Eric Perim,
Douglas S. Galvao
Abstract:
Nanoscrolls consist of sheets rolled up into a papyrus-like form. Their open ends produce great radial flexibility, which can be exploited for a large variety of applications, from actuators to hydrogen storage. They have been successfully synthesized from different materials, including carbon and boron nitride. In this work we have investigated, through fully atomistic molecular dynamics simulati…
▽ More
Nanoscrolls consist of sheets rolled up into a papyrus-like form. Their open ends produce great radial flexibility, which can be exploited for a large variety of applications, from actuators to hydrogen storage. They have been successfully synthesized from different materials, including carbon and boron nitride. In this work we have investigated, through fully atomistic molecular dynamics simulations, the dynamics of scroll formation for a series of graphene-like carbon nitride (CN) two-dimensional systems: g-CN, triazine-based (g-C3N4), and heptazine-based (g-C3N4). Carbon nitride (CN) structures have been attracting great attention since their prediction as super hard materials. Recently, graphene-like carbon nitride (g-CN) structures have been synthesized with distinct stoichiometry and morphologies. By combining these unique CN characteristics with the structural properties inherent to nanoscrolls new nanostructures with very attractive mechanical and electronic properties could be formed. Our results show that stable nanoscrolls can be formed for all of CN structures we have investigated here. As the CN sheets have been already synthesized, these new scrolled structures are perfectly feasible and within our present-day technology.
△ Less
Submitted 1 February, 2015;
originally announced February 2015.
-
A Brief Review on Syntheses, Structures and Applications of Nanoscrolls
Authors:
E. Perim,
L. D. Machado,
D. S. Galvao
Abstract:
Nanoscrolls are papyrus-like nanostructures which present unique properties due to their open ended morphology. These properties can be exploited in a plethora of technological applications, leading to the design of novel and interesting devices. During the past decade, significant advances in the synthesis and characterization of these structures have been made, but many challenges still remain.…
▽ More
Nanoscrolls are papyrus-like nanostructures which present unique properties due to their open ended morphology. These properties can be exploited in a plethora of technological applications, leading to the design of novel and interesting devices. During the past decade, significant advances in the synthesis and characterization of these structures have been made, but many challenges still remain. In this mini review we provide an overview on their history, experimental synthesis methods, basic properties and application perspectives.
△ Less
Submitted 22 January, 2015;
originally announced January 2015.
-
Inorganic Graphenylene: A Porous Two-Dimensional Material With Tunable Band Gap
Authors:
Eric Perim,
Ricardo Paupitz,
Pedro Alves da Silva Atreto,
Douglas Galvão
Abstract:
By means of ab initio calculations we investigate the possibility of existence of a boron nitride (BN) porous two-dimensional nanosheet which is geometrically similar to the carbon allotrope known as biphenylene carbon. The proposed structure, which we called Inorganic Graphenylene (IGP), is formed spontaneously after selective dehydrogenation of the porous Boron Nitride (BN) structure proposed by…
▽ More
By means of ab initio calculations we investigate the possibility of existence of a boron nitride (BN) porous two-dimensional nanosheet which is geometrically similar to the carbon allotrope known as biphenylene carbon. The proposed structure, which we called Inorganic Graphenylene (IGP), is formed spontaneously after selective dehydrogenation of the porous Boron Nitride (BN) structure proposed by Ding et al. We study the structural and electronic properties of both porous BN and IGP and it is shown that, by selective substitution of B and N atoms with carbon atoms in these structures, the band gap can be significantly reduced, changing their behavior from insulators to semiconductors, thus opening the possibility of band gap engineering for this class of two-dimensional materials.
△ Less
Submitted 30 September, 2014;
originally announced October 2014.
-
Mechanical Properties and Fracture Dynamics of Silicene Membranes
Authors:
T. Botari,
E. Perim,
P. A. S. Autreto,
A. C. T. van Duin,
R. Paupitz,
D. S. Galvao
Abstract:
As graphene became one of the most important materials today, there is a renewed interest on others similar structures. One example is silicene, the silicon analogue of graphene. It share some the remarkable graphene properties, such as the Dirac cone, but presents some distinct ones, such as a pronounced structural buckling. We have investigated, through density functional based tight-binding (DF…
▽ More
As graphene became one of the most important materials today, there is a renewed interest on others similar structures. One example is silicene, the silicon analogue of graphene. It share some the remarkable graphene properties, such as the Dirac cone, but presents some distinct ones, such as a pronounced structural buckling. We have investigated, through density functional based tight-binding (DFTB), as well as reactive molecular dynamics (using ReaxFF), the mechanical properties of suspended single-layer silicene. We calculated the elastic constants, analyzed the fracture patterns and edge reconstructions. We also addressed the stress distributions, unbuckling mechanisms and the fracture dependence on the temperature. We analysed the differences due to distinct edge morphologies, namely zigzag and armchair.
△ Less
Submitted 7 August, 2014;
originally announced August 2014.
-
Novel Nanoscroll Structures from Carbon Nitride Layers
Authors:
Eric Perim,
Douglas S. Galvao
Abstract:
Nanoscrolls (papyrus-like nanostructures) are very attractive structures for a variety of applications, due to their tunable diameter values and large accessible surface area. They have been successfully synthesized from different materials. In this work we have investigated, through fully atomistic molecular dynamics simulations, the dynamics of scroll formation for a series of graphenelike carbo…
▽ More
Nanoscrolls (papyrus-like nanostructures) are very attractive structures for a variety of applications, due to their tunable diameter values and large accessible surface area. They have been successfully synthesized from different materials. In this work we have investigated, through fully atomistic molecular dynamics simulations, the dynamics of scroll formation for a series of graphenelike carbon nitride (CN) twodimensional systems: gCN, triazinebased gC3N4, and heptazinebased gC3N4. Our results show that stable nanoscrolls can be formed for all of these structures. Possible synthetic routes to produce these nanostructures are also addressed.
△ Less
Submitted 23 May, 2014;
originally announced May 2014.
-
One-dimensional Silicon and Germanium Nanostructures With No Carbon Analogues
Authors:
Eric Perim,
Ricardo Paupitz,
Tiago Botari,
Douglas S. Galvao
Abstract:
In this work we report new silicon and germanium tubular nanostructures with no corresponding stable carbon analogues. The electronic and mechanical properties of these new tubes were investigated through ab initio methods. Our results show that the structures have lower energy than their corresponding nanoribbon structures and are stable up to high temperatures (500 and 1000 K, for silicon and ge…
▽ More
In this work we report new silicon and germanium tubular nanostructures with no corresponding stable carbon analogues. The electronic and mechanical properties of these new tubes were investigated through ab initio methods. Our results show that the structures have lower energy than their corresponding nanoribbon structures and are stable up to high temperatures (500 and 1000 K, for silicon and germanium tubes, respectively). Both tubes are semiconducting with small indirect band gaps, which can be significantly altered by both compressive and tensile strains. Large bandgap variations of almost 50% were observed for strain rates as small as 3%, suggesting possible applications in sensor devices. They also present high Young's modulus values (0.25 and 0.15 TPa, respectively). TEM images were simulated to help the identification of these new structures.
△ Less
Submitted 2 October, 2014; v1 submitted 9 March, 2014;
originally announced March 2014.
-
Violation of the Universal Behavior of Membranes Inside Cylindrical Tubes at Nanoscale
Authors:
E. Perim,
A. F. Fonseca,
N. M. Pugno,
D. S. Galvao
Abstract:
Recently, it was proposed based on classical elasticity theory and experiments at macroscale, that the conformations of sheets inside cylindrical tubes present a universal behavior. A natural question is whether this behavior still holds at nanoscale. Based on molecular dynamics simulations and analytical modeling for graphene and boron nitride membranes confined inside carbon nanotubes, we show t…
▽ More
Recently, it was proposed based on classical elasticity theory and experiments at macroscale, that the conformations of sheets inside cylindrical tubes present a universal behavior. A natural question is whether this behavior still holds at nanoscale. Based on molecular dynamics simulations and analytical modeling for graphene and boron nitride membranes confined inside carbon nanotubes, we show that the class of universality observed at macroscale is violated at nanoscale. The precise origins of these discrepancies is addressed and proven to be related to both surface and atomistic effects.
△ Less
Submitted 27 February, 2014;
originally announced February 2014.
-
Structure and Dynamics of Boron Nitride Nanoscrolls
Authors:
Eric Perim,
Douglas S. Galvao
Abstract:
Carbon nanoscrolls (CNSs) are structures formed by rolling up graphene layers into a papyruslike shape. CNNs have been experimentally produced by different groups. Boron nitride nanoscrolls (BNNSs) are similar structures using boron nitride instead of graphene layers. In this work we report molecular mechanics and molecular dynamics results for the structural and dynamical aspects of BNNS format…
▽ More
Carbon nanoscrolls (CNSs) are structures formed by rolling up graphene layers into a papyruslike shape. CNNs have been experimentally produced by different groups. Boron nitride nanoscrolls (BNNSs) are similar structures using boron nitride instead of graphene layers. In this work we report molecular mechanics and molecular dynamics results for the structural and dynamical aspects of BNNS formation. Similarly to CNS, BNNS formation is dominated by two major energy contributions, the increase in the elastic energy and the energetic gain due to van der Waals interactions of the overlap** surface of the rolled layers. The armchair scrolls are the most stable configuration while zigzag scrolls are metastable structures which can be thermally converted to armchair. Chiral scrolls are unstable and tend to evolve to zigzag or armchair configurations depending on their initial geometries. The possible experimental routes to produce BNNSs are also addressed.
△ Less
Submitted 31 July, 2009;
originally announced July 2009.
