-
Probabilistic Prediction of Material Stability: Integrating Convex Hulls into Active Learning
Authors:
Andrew Novick,
Diana Cai,
Quan Nguyen,
Roman Garnett,
Ryan Adams,
Eric Toberer
Abstract:
Active learning is a valuable tool for efficiently exploring complex spaces, finding a variety of uses in materials science. However, the determination of convex hulls for phase diagrams does not neatly fit into traditional active learning approaches due to their global nature. Specifically, the thermodynamic stability of a material is not simply a function of its own energy, but rather requires e…
▽ More
Active learning is a valuable tool for efficiently exploring complex spaces, finding a variety of uses in materials science. However, the determination of convex hulls for phase diagrams does not neatly fit into traditional active learning approaches due to their global nature. Specifically, the thermodynamic stability of a material is not simply a function of its own energy, but rather requires energetic information from all other competing compositions and phases. Here we present Convex hull-aware Active Learning (CAL), a novel Bayesian algorithm that chooses experiments to minimize the uncertainty in the convex hull. CAL prioritizes compositions that are close to or on the hull, leaving significant uncertainty in other compositions that are quickly determined to be irrelevant to the convex hull. The convex hull can thus be predicted with significantly fewer observations than approaches that focus solely on energy. Intrinsic to this Bayesian approach is uncertainty quantification in both the convex hull and all subsequent predictions (e.g., stability and chemical potential). By providing increased search efficiency and uncertainty quantification, CAL can be readily incorporated into the emerging paradigm of uncertainty-based workflows for thermodynamic prediction.
△ Less
Submitted 23 February, 2024;
originally announced February 2024.
-
All "roads" lead to rocksalt structure
Authors:
Matthew Jankousky,
Helen Chen,
Andrew Novick,
Vladan Stevanović
Abstract:
Rocksalt is the most common crystal structure among binary compounds. Moreover, no long-lived, metastable polymorphs are observed in compounds with the rocksalt ground state. We investigate the absence of polymorphism and the dominance of the rocksalt structure via first-principles random structure sampling and transformation kinetics modeling of three rocksalt compounds: MgO, TaC, and PbTe. Rando…
▽ More
Rocksalt is the most common crystal structure among binary compounds. Moreover, no long-lived, metastable polymorphs are observed in compounds with the rocksalt ground state. We investigate the absence of polymorphism and the dominance of the rocksalt structure via first-principles random structure sampling and transformation kinetics modeling of three rocksalt compounds: MgO, TaC, and PbTe. Random structure sampling reveals that for all three systems the rocksalt local minimum on the potential energy surface is far larger than any other, making it statistically the most significant structure. The kinetics modeling shows that virtually all other relevant structures, including most of the 457 known A$_1$B$_1$ prototypes, transform rapidly to rocksalt, making it the only option at ambient conditions. These results explain the absence of polymorphism in binary rocksalts, answer why rocksalt is the structure of choice for high-entropy ceramics and disordered ternary nitrides, and help understand/predict metastable states in crystalline solids more generally.
△ Less
Submitted 9 February, 2024;
originally announced February 2024.
-
Mechanistically-guided materials chemistry: synthesis of new ternary nitrides, CaZrN$_2$ and CaHfN$_2$
Authors:
Christopher L. Rom,
Andrew Novick,
Matthew J. McDermott,
Andrey A. Yakovenko,
Jessica R. Gallawa,
Gia Thinh Tran,
Dominic C. Asebiah,
Emily N. Storck,
Brennan C. McBride,
Rebecca C. Miller,
Amy L. Prieto,
Kristin A. Persson,
Eric Toberer,
Vladan Stevanović,
Andriy Zakutayev,
James R. Neilson
Abstract:
Recent computational studies have predicted many new ternary nitrides, revealing synthetic opportunities in this underexplored phase space. However, synthesizing new ternary nitrides is difficult, in part because intermediate and product phases often have high cohesive energies that inhibit diffusion. Here, we report the synthesis of two new phases, calcium zirconium nitride (CaZrN$_2$) and calciu…
▽ More
Recent computational studies have predicted many new ternary nitrides, revealing synthetic opportunities in this underexplored phase space. However, synthesizing new ternary nitrides is difficult, in part because intermediate and product phases often have high cohesive energies that inhibit diffusion. Here, we report the synthesis of two new phases, calcium zirconium nitride (CaZrN$_2$) and calcium hafnium nitride (CaHfN$_2$), by solid state metathesis reactions between Ca$_3$N$_2$ and $M$Cl$_4$ ($M$ = Zr, Hf). Although the reaction nominally proceeds to the target phases in a 1:1 ratio of the precursors via Ca$_3$N$_2$ + $M$Cl$_4$ $\rightarrow$ Ca$M$N$_2$ + 2 CaCl$_2$, reactions prepared this way result in Ca-poor materials (Ca$_xM_{2-x}$N$_2$, $x<1$). A small excess of Ca$_3$N$_2$ (ca. 20 mol\%) is needed to yield stoichiometric Ca$M$N$_2$, as confirmed by high-resolution synchrotron powder X-ray diffraction. In situ synchrotron X-ray diffraction studies reveal that nominally stoichiometric reactions produce Zr$^{3+}$ intermediates early in the reaction pathway, and the excess Ca$_3$N$_2$ is needed to reoxidize Zr$^{3+}$ intermediates back to the Zr$^{4+}$ oxidation state of CaZrN$_2$. Analysis of computationally-derived chemical potential diagrams rationalizes this synthetic approach and its contrast from the synthesis of MgZrN$_2$. These findings additionally highlight the utility of in situ diffraction studies and computational thermochemistry to provide mechanistic guidance for synthesis.
△ Less
Submitted 30 October, 2023;
originally announced October 2023.
-
3D photonics for ultra-low energy, high bandwidth-density chip data links
Authors:
Stuart Daudlin,
Anthony Rizzo,
Sunwoo Lee,
Devesh Khilwani,
Christine Ou,
Songli Wang,
Asher Novick,
Vignesh Gopal,
Michael Cullen,
Robert Parsons,
Alyosha Molnar,
Keren Bergman
Abstract:
Artificial intelligence (AI) hardware is positioned to unlock revolutionary computational abilities across diverse fields ranging from fundamental science [1] to medicine [2] and environmental science [3] by leveraging advanced semiconductor chips interconnected in vast distributed networks. However, AI chip development has far outpaced that of the networks that connect them, as chip computation s…
▽ More
Artificial intelligence (AI) hardware is positioned to unlock revolutionary computational abilities across diverse fields ranging from fundamental science [1] to medicine [2] and environmental science [3] by leveraging advanced semiconductor chips interconnected in vast distributed networks. However, AI chip development has far outpaced that of the networks that connect them, as chip computation speeds have accelerated a thousandfold faster than communication bandwidth over the last two decades [4, 5]. This gap is the largest barrier for scaling AI performance [6, 7] and results from the disproportionately high energy expended to transmit data [8], which is two orders of magnitude more intensive than computing [9]. Here, we show a leveling of this long-standing discrepancy and achieve the lowest energy optical data link to date through dense 3D integration of photonic and electronic chips. At 120 fJ of consumed energy per communicated bit and 5.3 Tb/s bandwidth per square millimeter of chip area, our platform simultaneously achieves a twofold improvement in both energy consumption and bandwidth density relative to prior demonstrations [10, 11]. These improvements are realized through employing massively parallel 80 channel microresonator-based transmitter and receiver arrays operating at 10 Gb/s per channel, occupying a combined chip footprint of only 0.32 mm2. Furthermore, commercial complementary metal-oxide-semiconductor (CMOS) foundries fabricate both the electronic and photonic chips on 300 mm wafers, providing a clear avenue to volume scaling. Through these demonstrated ultra-energy efficient, high bandwidth data communication links, this work eliminates the bandwidth bottleneck between spatially distanced compute nodes and will enable a fundamentally new scale of future AI computing hardware without constraints on data locality.
△ Less
Submitted 2 October, 2023;
originally announced October 2023.
-
Process Variation-Aware Compact Model of Strip Waveguides for Photonic Circuit Simulation
Authors:
Aneek James,
Anthony Rizzo,
Yuyang Wang,
Asher Novick,
Songli Wang,
Robert Parsons,
Kaylx Jang,
Maarten Hattink,
Keren Bergman
Abstract:
We report a novel process variation-aware compact model of strip waveguides that is suitable for circuit-level simulation of waveguide-based process design kit (PDK) elements. The model is shown to describe both loss and -- using a novel expression for the thermo-optic effect in high index contrast materials -- the thermo-optic behavior of strip waveguides. A novel group extraction method enables…
▽ More
We report a novel process variation-aware compact model of strip waveguides that is suitable for circuit-level simulation of waveguide-based process design kit (PDK) elements. The model is shown to describe both loss and -- using a novel expression for the thermo-optic effect in high index contrast materials -- the thermo-optic behavior of strip waveguides. A novel group extraction method enables modeling the effective index's ($n_{\mathrm{eff}}$) sensitivity to local process variations without the presumption of variation source. Use of Euler-bend Mach-Zehnder interferometers (MZIs) fabricated in a 300~mm wafer run allow model parameter extraction at widths up to 2.5~$μ$m (highly multi-mode) with strong suppression of higher-order mode excitation. Experimental results prove the reported model can self-consistently describe waveguide phase, loss, and thermo-optic behavior across all measured devices over an unprecedented range of optical bandwidth, waveguide widths, and temperatures.
△ Less
Submitted 4 January, 2023;
originally announced January 2023.
-
The Mixing Thermodynamics and Local Structure of High-entropy Alloys from Randomly Sampled Ordered Configurations
Authors:
Andrew Novick,
Quan Nguyen,
Roman Garnett,
Eric Toberer,
Vladan Stevanović
Abstract:
A general method is presented for modeling high entropy alloys as ensembles of randomly sampled, ordered configurations on a given lattice. Statistical mechanics is applied post hoc to derive the ensemble properties as a function of composition and temperature, including the free energy of mixing and local structure. Random sampling is employed to address the high computational costs needed to mod…
▽ More
A general method is presented for modeling high entropy alloys as ensembles of randomly sampled, ordered configurations on a given lattice. Statistical mechanics is applied post hoc to derive the ensemble properties as a function of composition and temperature, including the free energy of mixing and local structure. Random sampling is employed to address the high computational costs needed to model alloys with a large number of components. Doing so also provides rigorous convergence criteria, including the quantification of noise due to random sampling, and an estimation of the number of additional samples required to lower this noise to the needed/desired levels. This method is well-suited for a variety of cases: i) high entropy alloys, where standard lattice models are costly; ii) "medium" entropy alloys, where both the entropy and enthalpy play significant roles; and iii) alloys with residual short-range order. Binary to 5-component alloys of the group-IV chalcogenides are used as case examples, for which the predicted miscibility shows excellent agreement with experiment.
△ Less
Submitted 23 November, 2022;
originally announced November 2022.
-
Fabrication-Robust Silicon Photonic Devices in Standard Sub-Micron Silicon-on-Insulator Processes
Authors:
Anthony Rizzo,
Utsav Dave,
Asher Novick,
Alexandre Freitas,
Samantha P. Roberts,
Aneek James,
Michal Lipson,
Keren Bergman
Abstract:
Perturbations to the effective refractive index from nanometer-scale fabrication variations in waveguide geometry plague high index-contrast photonic platforms including the ubiquitous sub-micron silicon-on-insulator (SOI) process. Such variations are particularly troublesome for phase-sensitive devices such as interferometers and resonators, which exhibit drastic changes in performance as a resul…
▽ More
Perturbations to the effective refractive index from nanometer-scale fabrication variations in waveguide geometry plague high index-contrast photonic platforms including the ubiquitous sub-micron silicon-on-insulator (SOI) process. Such variations are particularly troublesome for phase-sensitive devices such as interferometers and resonators, which exhibit drastic changes in performance as a result of these fabrication-induced phase errors. In this Letter, we propose and experimentally demonstrate a design methodology for dramatically reducing device sensitivity to silicon width variations. We apply this methodology to a highly phase-sensitive device, the ring-assisted Mach Zehnder interferometer (RAMZI), and show comparable performance and footprint to state-of-the-art devices while substantially reducing stochastic phase errors from etch variations. This decrease in sensitivity is directly realized as energy savings by significantly lowering the required corrective thermal tuning power, providing a promising path towards ultra-energy-efficient large-scale silicon photonic circuits.
△ Less
Submitted 23 May, 2022;
originally announced May 2022.
-
Integrated Kerr frequency comb-driven silicon photonic transmitter
Authors:
Anthony Rizzo,
Asher Novick,
Vignesh Gopal,
Bok Young Kim,
Xingchen Ji,
Stuart Daudlin,
Yoshitomo Okawachi,
Qixiang Cheng,
Michal Lipson,
Alexander L. Gaeta,
Keren Bergman
Abstract:
The exponential growth of computing needs for artificial intelligence and machine learning has had a dramatic impact on data centre energy consumption, which has risen to environmentally significant levels. Using light to send information between compute nodes can dramatically decrease this energy consumption while simultaneously increasing bandwidth. Through wavelength-division multiplexing with…
▽ More
The exponential growth of computing needs for artificial intelligence and machine learning has had a dramatic impact on data centre energy consumption, which has risen to environmentally significant levels. Using light to send information between compute nodes can dramatically decrease this energy consumption while simultaneously increasing bandwidth. Through wavelength-division multiplexing with chip-based microresonator Kerr frequency combs, independent information channels can be encoded onto many distinct colours of light in the same optical fibre for massively parallel data transmission with low energy. While previous demonstrations have relied on benchtop equipment for filtering and modulating Kerr comb wavelength channels, data centre interconnects require a compact on-chip form factor for these operations. Here, we demonstrate the first integrated silicon photonic transmitter using a Kerr comb source. The demonstrated architecture is scalable to hundreds of wavelength channels, enabling a fundamentally new class of massively parallel terabit-scale optical interconnects for future green hyperscale data centres.
△ Less
Submitted 8 September, 2021;
originally announced September 2021.
