Martin A. Fisher School of Physics

August 31, 2021

Arnab Datta, Brandeis University
Host: John Wardle
"Energy cost of protein gradient formation in cells"
Abstract: Cells make protein gradients for various purposes, such as establishing position information in development or defining cell polarity in the process of cell division. Two classes of mechanisms for maintaining protein gradients in cells have been reported in the literature: i. those that combine protein diffusion and degradation, and ii. mechanisms that involve active transport of proteins by molecular motors. An example of the first mechanism is the Bicoid protein gradient in the Drosophila embryo, which provides positional information to the nuclei during development [1]. A Smy1 gradient along actin cables in budding yeast cells regulates cable length and is formed by active transport of the proteins by myosin motors to the bud neck [2]. Establishing and maintaining these protein gradients require cells to expend energy. In this talk I examine different mechanisms of gradient formation in cells and estimate the energy costs associated with them. I also consider the scaling of the energy expenditure with cell size for the two different models of gradient formation and discuss when one mechanism is energetically less costly than the other.
References:
[1] O. Grimm, M. Coppey, and E. Wieschaus, Development. (2010) 137:2253-64.
[2] J. A. Eskin, A. Rankova, A. B. Johnston, S. L. Alioto, and B. L. Goode, Mol Biol Cell. (2016) 27:828-37.

Daichi Hayakawa, Brandeis University
Host: John Wardle
"Programming the self-assembly of tubules using DNA origami triangles as building blocks"
Abstract: DNA origami is a method by which a single-stranded DNA scaffold is folded into some prescribed shape by hundreds of user-designed DNA ‘staple’ strands. Historically, this process has been used to make intricate 3D nanostructures with sub-nanometer precision. In this talk, I will discuss a new route for using DNA origami to make colloidal particles that then self-assemble into well-defined geometrical structures. More specifically, I use DNA origami to make triangular subunits and control their binding angles to program the assembly of nanotubes. The nanotubes are assembled from one type of triangle whose three edges bind to themselves at prescribed dihedral angles and are programmed by the DNA sequence design. I show that DNA origami triangles, each roughly 50 nanometers in size, can assemble into rigid tubules reaching tens of micrometers in length. Interestingly, I find that there is a distribution in width and the chirality of the assembled tubes, suggesting that our DNA origami colloids could be flexible or that the kinetic pathway toward tube closure plays a vital role in determining the final structure. Further, I discuss a method for limiting such tube type distributions by increasing the number of subunit types involved in the assembly.

January 14, 2022

James Cho, Flatiron Center for Computational Astrophysics and Brandeis University
Host: Albion Lawrence

Abstract: 'Exoplanets' is an exciting, new field of astrophysics. The field has grown rapidly since the first discovery of an exoplanet around a Sun-like star, only ~25 years ago – for which the Nobel prize in physics was recently awarded. With thousands of exoplanets now detected, accurate characterization of their atmospheres – in particular, their composition, weather, and climate – has become the next critical step in understanding them. The characterization is not only crucial for understanding current observations, it is crucial for ultimately assessing whether planets can harbor life. In this talk, how physics and mathematics are used to address these complex problems as well as what we currently understand about the problems are presented. The presentation will focus on the structure and evolution of ‘exo-storms’ and the variability they induce that could be observed with current missions and future ones that are soon to come online.

September 27, 2021

Brian Swingle, Brandeis University

The physics of quantum chaos is relevant for a broad range of problems, from the origin of hydrodynamics in quantum systems to the nature of black holes in quantum gravity, and it is increasingly experimentally accessible thanks to rapid developments in highly-controlled quantum systems. However, this breadth has led to a myriad of different characterizations of chaos, with the interrelations between them often still mysterious. I will describe progress towards a more unified framework by highlighting an emerging set of connections between three key manifestations of chaos: information scrambling, fluctuating hydrodynamics, and random matrix theory.

October 5, 2021

Lee Roberts, Boston University
Host: Aram Apyan

Abstract: The Standard Model provides a very precise prediction of the muon’s magnetic anomaly aμ = (gμ - 2)/2, the deviation from 2 of the gyromagnetic ratio gμ. In his seminal 1926 paper, P.A.M. Dirac predicted that for electrons ge = 2, but experiments then revealed that ge was slightly larger than 2. The reason was to be found in Quantum Mechanics, and the first radiative correction to ge, calculated by Julian Schwinger, explained a deviation of order 0.1 %. Today, the Standard Model predicts the value of aμ to a precision of ± 0.36 parts per million (ppm). Dedicated experiments have measured aμ to ± 0.35 ppm precision. Therefore, precision measurements of the anomaly provide a stringent test of the Standard Model’s completeness, since Nature knows about all forces that could contribute to the muon’s magnetism, including those from New Physics that has not yet been discovered.

I will briefly review the intellectual history that began with the discovery of spin and the g-factor of the electron and its role on the development of Modern Physics. I will then focus on the new measurement of the muon magnetic anomaly that was recently reported by Fermilab experiment E989. The result determined from the first data set collected in 2018 has a precision of 0.46 ppm, and agrees well with the previous result obtained at Brookhaven National Laboratory at the beginning of this century. The combined experimental value exhibits tension with the Standard Model value.  

October 12, 2021

Aram Harrow, Massachusetts Institute of Technology
Host: Aram Apyan

Abstract: The appeal of quantum computing is based on the fact that simulating N quantum systems on a classical computer takes time exponential in N. This exponential hardness is known to hold even for shallow quantum circuits, meaning unitary dynamics that run for a constant amount of time. We show that when the quantum circuits are made of random gates on a 2D geometry, they are not always exponentially hard to simulate. Instead, we give evidence for a phase transition in computational difficulty as the depth and local dimension are varied. Our evidence consists of (1) fast classical simulations of random circuits on a 400x400 grid of qubits, (2) a mapping to the order/disorder transition in an associated stat mech model, and (3) a proof that some circuit families are easy to simulate approximately but hard to simulate exactly. Our algorithms are based on tensor network contraction and mapping the 2D random unitary circuit to a 1D process consisting of alternating rounds of random local unitaries and weak measurements.

January 19, 2022

Special Division of Science Seminar

*Please note that this seminar will take place in Gerstenzang 123.

Chanda Prescod-Weinstein, University of New Hampshire

October 26, 2021

Piyush Grover, University of Nebraska, Lincoln
Host: Seth Fraden

Abstract: The geometrical framework of dynamical systems theory was originally developed by Poincare to study the chaotic dynamics of the gravitational three-body problem. In this talk, we will first discuss the application of this theory to explain the (sometimes puzzling) motion of celestial bodies, as well as to design non-intuitive fuel-efficient space missions to moon and beyond. A typical particle meanders through the phase space of an N-body problem by travelling on invariant manifolds that connect different equilibria and periodic orbits. These invariant manifolds, created by the competing gravitational forces, act as `interplanetary superhighways'. Next, we will discuss the far reaching generalizations of this framework in the context of hydrodynamics, including our recent work in 2D active nematics. Hydrodynamics equations give rise to a deterministic nonlinear dynamical system evolving in an infinite dimensional phase space. The dominant flow structures are understood in terms of Exact Coherent Structures (ECS) and the invariant manifolds connecting them. An ECS is (generically unstable) stationary, periodic, quasiperiodic, or traveling wave solution of the hydrodynamic equations. A finite set of ECS, together with their invariant manifolds, constitutes a reduced-order but exact characterization of the global phase space. Though each ECS is non-turbulent, this representation has been shown to be adequate for describing high Reynolds number turbulent flows of passive fluids, which appear as chaotic trajectories meandering through the phase space and visiting the neighborhoods of different ECS in a recurring fashion. We provide evidence that the ECS and their invariant manifolds also act as an organizing template for the complicated spatiotemporal motion of active fluid turbulence.

November 2, 2021

November 9, 2021

Martin Bazant, Massachusetts Institute of Technology
Host: Aram Apyan

Abstract

The importance of indoor airborne transmission of COVID-19 is now widely recognized, but public health guidance has yet to provide quantitative measure sto protect against it, beyond the Six-Foot Rule of social distancing.  This talk introduces a simple safety guideline that bounds the ``cumulative exposure time” in an enclosed space, depending on the rates of ventilation and air filtration, dimensions of the room, breathing rate, respiratory activity, face mask use, and infectiousness of exhaled air.  Case studies are presented for classrooms and nursing homes, and an online app is provided to facilitate widespread use of the guideline.  Real-time monitoring of transmission risk can be accomplished by expressing the guideline as a bound on safe carbon dioxide concentrations, and the model can be used to optimize HVAC systems to maximize clean and safe air while minimizing energy costs. [Ref: http://www.mit.edu/~bazant/COVID-19.]

November 16, 2021

November 30, 2021

Pankaj Mehta, Boston University
Host: Kanaya Malakar

ABSTRACT: The towering successes of twentieth century theoretical physics were marked by two guiding principles: (i) the importance of symmetry and (ii) the centrality of minimization principles and energy functionals reflecting equilibrium dynamics. Yet, how we can exploit these principles to develop a theory of living systems is unclear since the biological world is composed of heterogeneous, interacting components operating out of equilibrium. For these reasons, theoretical biological physics requires new ideas that move beyond these two theoretical pillars. Through examples from ecology, neural-inspired machine learning, and gene networks, I will argue that one possible strategy for taming biological complexity is to embrace ideas from random matrix theory and the physics of disordered systems. I will show how, at their core, many of these problems can be thought of as generalized constraint-satisfaction problems, hinting at a new theoretical paradigm for tackling problems not only in biology, but also in other branches of physics such as quantum control.

December 7, 2021

January 18, 2022

Aram Harrow, MIT

Host: Aram Apyan

Abstract: The appeal of quantum computing is based on the fact that simulating N quantum systems on a classical computer takes time exponential in N. This exponential hardness is known to hold even for shallow quantum circuits, meaning unitary dynamics that run for a constant amount of time. We show that when the quantum circuits are made of random gates on a 2D geometry, they are not always exponentially hard to simulate. Instead, we give evidence for a phase transition in computational difficulty as the depth and local dimension are varied. Our evidence consists of (1) fast classical simulations of random circuits on a 400x400 grid of qubits, (2) a mapping to the order/disorder transition in an associated stat mech model, and (3) a proof that some circuit families are easy to simulate approximately but hard to simulate exactly. Our algorithms are based on tensor network contraction and mapping the 2D random unitary circuit to a 1D process consisting of alternating rounds of random local unitaries and weak measurements.

January 25, 2022

Tyler Maunu, Brandeis University

Host: Aram Apyan

Abstract: This talk will discuss some recent work on methods for sampling that utilize ideas from optimal transport. Sampling refers to the problem of generating samples from a probability distribution given access to a potentially unnormalized version of its density. The sampling problem has wide applications ranging from Bayesian inference, generative models, online algorithms, statistical physics, computational biology, and much more. I will begin by outlining the problem of optimal transport and some basic theory behind it. Then, I will discuss its connections to the sampling problem through the idea of gradient flows, and show how this perspective leads to new algorithms for sampling, like the Newton Langevin Algorithm. After this, I will finish by demonstrating how one can combine optimal transport, regularization, and neural networks to form efficient sampling methods for deep generative modeling.

February 1, 2022

JiJi Fan, Brown University

Host: Aram Apyan

Abstract: Axions, periodic pseudo-scalars, enjoy a wide range of phenomenological applications, from solving the strong CP problem to providing dark matter and inflaton candidates. Justifiably, they have been the subject of prolonged theoretical interest, which has intensified over the past few years. In this talk, I will discuss several fun developments in axion probes and models, including: novel astrophysical probes, such as axion echos from supernovae remnants and cosmological distance ladders; and a recently identified source of axion potential from virtual magnetically charged particles.

February 8, 2022

Jesse Thaler, MIT

Host: Aram Apyan

Abstract: Since the 1960s, particle physicists have developed a variety of data analysis strategies for the goal of comparing experimental measurements to theoretical predictions.  Despite their numerous successes, these techniques can seem esoteric and ad hoc, even to practitioners in the field.  In this talk, I explain how many particle physics concepts and analysis tools have a natural geometric interpretation in an emergent “space” of collider events.  This approach reveals fascinating connections between quantum field theory and machine learning, which I illustrate using public data from the CMS experiment at the Large Hadron Collider.

February 15, 2022

Baylor Fox-Kemper, Brown University

Host: Aram Apyan

Abstract: As our planet warms, changes in the oceans and cryosphere connect to many aspects of societal concern: sea level, acidification, ecosystem impacts, water availability, fisheries, etc.   I will review some of the key assessments of observed and projected changes from the UN Intergovernmental Panel on Climate Change Sixth Assessment Report (AR6) based on my experience as a Coordinating Lead Author, focusing primarily on my chapter which covers changes to the oceans and changes to the cryosphere that affect sea level. Many of the headline assessments are unchanged or consistent with previous IPCC reports, but the advances in the CMIP6 models over the CMIP5 generation, new use of other classes of models such as emulators to extend the results over a broader range of scenarios and constrain temperature and sea level versus observations, new observation types and quantity, and the continuing emergence of climate trends from natural variability provide greater confidence and precision than in previous reports.  In a recent manuscript (Hall & Fox-Kemper, 2021) we show that regional mixed layer depth is significantly correlated with many of these model parameters, although the active layer and the mixed layer are distinct.  As the mixed layer is observable using Argo data, it can be used as an emergent constraint.  Using these correlations and observations from the Argo float network, we revise the ensemble mean and narrow the 66% range of equilibrium climate sensitivity (ECS) for the particular CMIP6 model collection from 4.51 (3.13–5.71) C, to 4.66 (3.88–5.43) C, amounting to a 40% reduction in the span of the uncertainty range.

March 1, 2022

Salvatore Vitale, MIT

Host: Aram Apyan

Abstract: Seven years after the first direct detection of gravitational waves, nearly 100 compact binary mergers have been discovered in the data of the LIGO and Virgo detectors. The masses and spins of these objects - black holes and neutron stars - can be used to tackle multiple problems in physics, astrophysics and cosmology. In this talk I will discuss what can be learned about the properties of the individual gravitational-wave sources and from the overall set of detections. Finally, I will discuss the scientific potential of the next generation of ground-based gravitational-wave detectors.

March 8, 2022

Alex Sushkov, Boston University

Abstract: The dark matter puzzle is one of the most important open problems in modern physics. I will describe some of the approaches that are used to search for non-gravitational interactions of ultralight dark matter, and the fundamental limitations on their sensitivity. The axion is a compelling dark matter candidate, since it resolves the strong-CP problem of quantum chromodynamics. I will focus on precision laboratory-scale experiments that search for axion-like dark matter, aiming to achieve, or circumvent, the quantum limits on their sensitivity.

March 22, 2022

Matthew Lister, Purdue University

Host: John Wardle

Abstract: Supermassive black holes at the centers of distant galaxies represent some of the most powerful particle accelerators in the universe, and are capable of launching highly relativistic plasma jets at near light speeds out to distances of several million light years. Recent observations suggest that these jets are massive neutrino factories, with measured energies of up to several PeV. Despite their ~billion light year distances, it is possible to make high-resolution time lapse movies of extragalactic jets using continent-scale radio interferometer arrays. I will discuss recent results from a long term radio imaging program (MOJAVE), which is regularly observing the brightest jets in the northern sky with the Very Long Baseline Array. Our study seeks to learn why only a handful of jets so far have been seen as neutrino emitters, and to better understand how jet flows are organized and accelerated on scales within a few hundred light years of the black hole.  MOJAVE has also made interesting discoveries regarding high energy gamma-ray emission from very recently launched jets, and found evidence of jet nozzles that oscillate over time. I will discuss the latter and its relevance to the MOJAVE jet PKS 2131-021, recently discovered to be a close-orbit supermassive black hole binary system that is predicted to undergo a spectacular merger roughly 10,000 years from now.

March 29, 2022

Areg Danagoulian, MIT

Host: Aram Apyan

Abstract: Arms control treaties are not sufficient in and of themselves to neutralize the existential threat of nuclear weapons. Technologies are necessary for verifying the authenticity of the nuclear warheads undergoing dismantlement before counting them towards a treaty partner’sobligation. We have developed a neutron-based concept which uses observations of isotope specific nuclear transitions to authenticate a warhead's fissile components.  Most actinides such as uranium and plutonium exhibit unique sets of nuclear resonances when interacting with eV neutrons. When measured, these resonances produce isotope-specific features in the spectral data, thus creating an isotopic-geometric  "fingerprint" of an object. All information in these measurements is encrypted in the physical domain through a process called physical cryptography. Using proof-of-concept experiments, these techniques are shown to reveal no isotopic or geometric information about the weapon, while readily detecting hoaxing attempts. The talk will discuss the policy context, the concept of the experimental techniques, along with results from simulation and experimental measurements.

April 5, 2022

Crystal Noel, Duke University

Host: Brian Swingle

Abstract: Trapped ions have been used for decades as model quantum systems. With excellent isolation from the environment and atomic transitions in the visible spectrum where lasers are available, unparalleled quantum control is possible. These properties make them one of the best candidates as qubits, the fundamental unit of a quantum computer.  At the Duke Quantum Center (DQC), we are building and operating the leading academic quantum computers available using trapped ions. Our systems have modularity, high fidelity, stability, and potential for growth. In this talk, I will introduce you to these systems and discuss recent algorithms and simulations we have implemented. I will discuss not only the future of the DQC as a user facility with many quantum computing systems available for academic use, but also the future of trapped ion quantum computing and what technological advancements are required to scale these systems up to many more qubits.

April 12, 2022

Katia Bertoldi, Harvard

Host: Kanaya Malakar

Abstract: Inflating a rubber balloon leads to a dramatic shape change: a property that is exploited in the
design of soft robots and deployable structures. On the one hand, fluid-driven actuators capable
of complex motion can power highly adaptive and inherently safe soft robots. On the other hand,
inflation can be used to transform seemingly flat shapes into shelters, field hospitals, and space
modules. In both cases, just like the simple balloon, only one input is required to achieve the
desired deformation. This simplicity, however, brings strict limitations: soft actuators are often
restricted to unimodal and slow deformation and deployable structures need a continuous supply
of pressure to remain upright. Here, we embrace multistability as a paradigm to improve the
functionality of inflatable systems. In the first part of this seminar, I exploit snapping instabilities
in spherical shells to decouple the input signal from the output deformation in soft actuators–a
functionality that can be utilized to design a soft machine capable of jumping. In the second part
of the seminar, I draw inspiration from origami to design multistable and inflatable structures at
the meter scale. Because these deployable systems are multistable, pressure can be disconnected
when they are fully expanded, making them ideal candidates for applications such as emergency
sheltering and deep space exploration. Together, these two projects highlight the potential of
multistability in enabling the design and fabrication across various scales of multi-form, multi-
functional, and multi-purpose materials and structures.

April 26, 2022

Anatoli Polkovnikov, Boston University

Host: Kanaya Malakar

Abstract: After briefly reviewing classical chaos and its relation to ergodicity, i.e. to emergence of statistical mechanics, I will talk about the notion of quantum chaos and ergodicity, its present understanding through random matrices. Then I will discuss how we can understand quantum chaos through sensitivity of eigenstates to small perturbations. This sensitivity is expressed through so called fidelity susceptibility or quantum geometric tensor. If time permits I will also mention emergent connections of this probe to recently discovered universal operator spreading and Krylov complexity

May 4, 2022

Paul Francois, McGill University

Note: special time 12:00 pm