Charles E. Diesendruck

Krill Prize 2018
Technion

Charles E. Diesendruck

New polymers for Modern Chemistry

Our research is on the field of polymer mechanochemistry and the synthesis of new functional polymers. We look at how mechanical force changes the chemistry of a polymer, and how changing the structure of molecules changes the mechanical properties of plastic materials. In recent years, we have been studying the use of intramolecular covalent cross-links to develop polymers that mimic the structure of proteins. We’ve made polymers that are “invincible” to force in solution and we are on the pathway to produce materials with high stiffness and elasticity. Finally, we have also been looking at the chemical decomposition of membranes for alkaline fuel cells. Using fundamental organic chemistry, significant improvements have been developed in testing new membranes, and we proposed new solutions to a technological problem that could change the energy storage and production landscape

Yakov Babichenko

Krill Prize 2018
Technion

Yakov Babichenko

Complexity and Learning of Equilibria.

Equilibria are the central solution concepts in game theory. However, these solutions impose strong assumptions on the rationality of the players. Players know the utilities
of all players they are facing, players are able to compute an equilibrium, and finally after computing equilibria players are able to agree on the same equilibrium and
play accordingly. This criticism raises the following natural question:

1. Can players learn to play an equilibrium using some natural learning rule that does not require the above strong assumptions?
2. If so, how fast can players learn it?
3. How hard is equilibrium computation?

In my research I am trying to address these question by considering the equilibrium computation task in various computational models; E.g., communication complexity
and query complexity.

Ayelet Erez

Krill Prize 2018
Weismann Institute

Ayelet Erez

Deciphering unique metabolic changes in complex diseases for diagnosis and treatment

As a physician- scientist- geneticist, my research challenge is to untangle the genetic and metabolic alterations that lead to disease development, for diagnostic and therapeutic applications.

To characterize the variables that differentiate healthy cells from the pathological ones, we use advanced genetic and metabolic systems that involve cell models, mice, and patient samples. By stably labeling metabolites, we are able to track the path of a specific metabolite in the healthy cell and compare it to the corresponding path in the patient’s cell. Because multiple changes occur along disease course, we correlate our findings with samples from patients with inborn errors of metabolism (IEM), to dissect the direct consequences of a specific metabolic aberration.

Revealing a unique metabolic anomaly in patients, enables us to suggest diagnostic and therapeutic options, which identify and target specifically the metabolic pathway that is essential for the pathological cell, thereby diagnosing, slowing down or even stopping the disease progression.

Adi Salomon

Krill Prize 2018
Bar-Ilan University

Adi Salomon

light matter interaction at the nanoscale

Energy transfer processes between light and molecules and among molecules themselves play important roles in nature, with photosynthesis probably being the best-known example. In my laboratory at the Bar-Ilan Nanocenter, we study such light-matter interactions at the nanoscale, focusing on organic molecules. To do so, we use metallic nanostructures to concentrate the light energy. Surface-plasmons, light-driven, collective oscillations of the metal’s free electrons, allow tuning, enhancing and confining the electromagnetic field to a tiny, sub-wavelength volume. In this context, the overall goal of my wosrk is to modify molecules using plasmonic modes as a photonic environment, or even as a ‘photonic catalyst’.

Having this goal in mind, we have developed during the last four years metallic systems with unique properties. These systems are composed either of well-defined metallic nanostructures, or, more recently, of a large-scale nanoporous metallic network. Both of these systems are different from conventional plasmonic devices, and they are complementary to each other in many aspects. The nanofabricated surfaces we produce are like an artificial leaf, on which the light energy can be funneled to a desired ‘reaction center’. We use this confined light as a ‘photonic reagent’, which opens new photochemical reaction channels, or which modifies the potential energy barrier along a given reaction coordinate and thus enhances or inhibits a specific reaction channel. The idea here is to open new routes for photochemical reactions on surfaces by controlling the electromagnetic-field properties at the metal surface.

Elad Gross

Krill Prize 2018
Hebrew University

Elad Gross

Mechanism-driven regulation of reactivity and selectivity in organocatalysis

The preparation of many materials that are essential for our daily life in a modern society, such as plastics, fuels and fertilizers, heavily relies on the use of catalysts. Although catalysts have been used in the chemical industry for more than a century, there are many details about their structure-reactivity correlations which are not yet clear. One of the questions that I find most intriguing in catalysis research is identifying how the physical and chemical properties of catalysts influence their reactivity. In order to address this goal I use state of the art spectroscopic tools to detect the locations in which chemical reactions occur on the surface of single particles. Using this approach, we have recently identified that chemical reactivity primarily occurs on the periphery of metallic particles while lower reactivity was recorded at the center of the particle. These conclusions will enable the development of optimized catalysts based on rational design.

Emmanuel Levy

Krill Prize 2018
Weizmann Institute

Emmanuel Levy

“Soon after the start of my Ph.D, I met with my advisor and a post-doctoral researcher who needed to carry out an analysis of protein structures. They presented me with the problem of compiling a non-redundant dataset of protein complexes to analyze their structure and evolution. They asked me if I was interested in starting my Ph.D with this project. While it was supposed to take 6 months at most, it evolved and became the focus of my career”.

Protein-protein interactions from atomic to cellular scales

Dr. Levy aims to discover novel functional and evolutionary properties of proteins, which are the main actors of cellular functions. Proteins in a cell are much like people in a company: they move and interact with each other and often form teams or “protein complexes.” During his PhD studies, Dr. Levy studied the structure of protein complexes and discovered common routes for their evolution. During his post-doctoral work, he studied the broader picture of all protein interactions within a cell. He proposed the concept that many “promiscuous” interactions exist and showed that proteins are under evolutionary pressure to avoid such counter-productive interactions. Today, Dr. Levy’s group is investigating protein interactions from the atomic to the cellular scales, understanding how even single point mutations can drive a dramatic re-organization of proteins within the cell, and how proteins can be programmed to assemble into biomaterials.

Anat Milo

Krill Prize 2018
Ben-Gurion University

Anat Milo

Mechanism-driven regulation of reactivity and selectivity in organocatalysis

From energy conversion, through the production of materials, to the synthesis of natural products, pharmaceutics, and fine-chemicals – catalysts have been key to advancing our modern lifestyle. However, most industrial catalytic processes rely on toxic, expensive metals that have to be rigorously removed from the products. Organo¬catalysis offers a greener alternative, but often requires high catalyst loading, suffers from repro-ducibility issues, and is tailored for a limited range of substrates. To overcome these limitations and unlock organocatalysts’ full potential, it is necessary to uncover the structural origin of their reactivity and selectivity, thereby providing a knowledge-driven approach to their design and optimization. The Milo research group synergistically integrates experimental, computational, and statistical methods to design and construct modular organocatalyst libraries and provide a powerful strategy for discovering and optimizing selective catalytic reactions.

Avi Shroeder

Krill Prize 2017
Technion

Dr. Avi Shroeder

Personalized Cancer Nano-Medicines
Our ultimate goal is to help patients recover from metastatic cancer, by
tailoring a personalized treatment to each patient. We believe this far-reaching
goal will be achieved through fundamental research that bridges basic
sciences with engineering.

Specifically, we develop nanotechnologies that target the diseased tissue,
where they perform a programmed therapeutic task, including repairing
genetic errors and retrieving patient-specific information from the disease site. For this, we engineer nanotechnologies that target miniature molecularmachines, proteins and genes (DNA, RNA) to primary tumors and metastasis, for diagnosing and treating cancer in a personalized manner. Our motivation stems from the unfortunate fact that more than one-third of cancer patients are administered an ineffective medication – a medicine the
cancerous tissue does not respond to, while inducing harsh side-effects. We
sought to change this reality by developing an innovative technology that
personalizes cancer care. Our research approach is to employ nanotechnology devices, which are millions of times smaller than a cancer cell, that circulate in surveillance inside the body, detecting diseased tissue, and performing a regenerative medical task at the target. For example, we programmed these nanotechnology devices to diagnose, before treatment begins, which cancer medicine is best for each patient. Specifically, we engineered nanoparticles that contain miniscule amounts of various medicines, each sufficient to kill a single cancer cell. Each drug was labeled with a short DNA barcode, that enables detecting the potency of each medicine inside the patient’s tumor. The nanoparticles are injected intravenously and target the tumor site. In the tumor, the nanoparticles are taken up by the tumor cells, and patient-specific therapeutic information is retrieved from the tumor and analyzed to predict the best medicinal treatment approach.

To date, we demonstrated this approach improves the management of triplenegative breast cancer.

We plan to :

1) use this system to study how the primary tumor, versus the metastasis, respond to medication,

2) to expand our knowledge base to additional cancer types, including pancreatic and ovarian cancer, and

3) to develop a system that predicts, before beginning treatment, harsh side-effects a patient may undergo in response to medication.

Nanotechnology grants unprecedented accuracy, enabling us to diagnose and
treat malignancies, while reducing side effects and improving patient’ qualityof-
life.

Carmel Rotchild

Krill Prize 2017
Technion

Dr. Carmel Rotchild

  • New thermodynamic ideas for solar energy: In principle, converting inefficient parts of the solar spectrum to wavelengths most suitable for photovoltaics should double their efficiency. Confronting this problem using conventional optical frequency conversion methods is known to be very challenging because they require huge intensities and high coherence far above solar radiation.

Recently, we discovered what we believe is a breakthrough in both fundamental and applied science. The radiance of thermal emission, as described by Planck’s law, depends only on the emissivity and temperature of a body, and increases monotonically with temperature rise at any emitted wavelength. Nonthermal radiation, such as photoluminescence (PL), is a fundamental light–matter interaction that conventionally involves the absorption of an energetic photon, thermalization, and the emission of a red-shifted photon. In this quantum process, radiation is governed by the photon rate conservation and thermodynamically described by the chemical potential. Until recently, the role of rate conservation when thermal excitation is significant had not been studied in any nonthermal radiation, leaving open many questions; for example, what is the overall emission rate if a high quantum efficiency PL material is heated to a temperature

where it thermally emits a rate of 50 photons/sec at its bend edge, while in parallel, the PL is excited at a rate of 100 photons/sec? We discovered that the answer is an overall rate of 100 blue-shifted photons/sec. In contrast to thermal emission, the PL rate is conserved if the temperature increases, while each photon is blue-shifted. A further rise in temperature leads to an abrupt transition to thermal emission where the photon rate increases sharply. We also demonstrated how endothermic-PL generates orders of magnitude more energetic photons than thermal emission at similar temperatures. These findings show that PL is an ideal optical heat pump, and can harvest thermal losses in photovoltaics with theoretical maximal efficiency of 70%, and a practical device that aims to reach 48% efficiency. These discoveries have already  led to two publications in leading journals (Optica, Nature Communication and its Optics and Photonics News). A paper was also published in the proceedings as an invited talk at SPIE, the largest conference on optics. Based on our proposed device, we were invited to join in publishing with leading researchers in the field (Eli Yablonovich, Gang Chen, Martin Green, and more) on the “Roadmap on the Optical Energy Conversion”. These publications of my group put us in the vanguard of research on the thermodynamics of radiation. As a result of these achievements, I have been invited to give numerous invited talks and my student, Dr. Assaf Manor, was awarded an ‘Adams Fellowship’. We are currently aiming to demonstrate a fully operational energy conversion device with record efficiency as part of our ERC grant on “new thermodynamic ideas for frequency conversion and photovoltaics”. Due to its ‘all optical’ nature, and low temperature operation, such a device has a significant chance of becoming a disruptive technology in photovoltaics.

  • New thermodynamic idea for extreme optical frequency up-conversion: Frequency up- conversion of a few low-energy photons into a single high-energy photon contributes to imaging, light sources, and detection. The up-converting of many photons, however, exhibits negligible efficiency. Up-conversion through laser heating is an efficient means to generate energetic photons, yet the spectrally broad thermal emission and the challenge of operating at high temperatures limit its practicality. We recently developed and demonstrated experimentally up- conversion by excitation of a steady-state nonthermal-equilibrium population, which induces steady, narrow emission at a practical bulk temperature. Specifically, we used a 10.6mm laser to resonantly excite vibronic states in silica that are coupled to emitters operating at 980nm. The result is a narrow emission with 4% total efficiency and up-converted radiance that far exceeds the device’s possible black-body radiation. Such efficiency is better by many orders of magnitude than ever reported and opens the way for the development of new light sources with record efficiencies. We recently published these results in ASC Photonics. In another paper, currently under review, we use the same method to demonstrate experimentally that in contrast to thermal excitation, different co-doped emitters compete with each other for energy transferred from the ‘hot’ vibronic modes in a similar way as in photoluminescence (down conversion), even though the photon energy is increased 10-fold. In addition, the same method is used to generate 13-, 16- and 20-fold narrow up-conversion at record efficiencies. Further development of this method using tailored materials will put us in a position to present new and more efficient UV and X-ray
  • New thermodynamic idea for solar powered laser and on chip high-Q lasers: The optical conversion of incoherent solar radiation into a bright, coherent laser beam enables the application of nonlinear optics to solar energy conversion and energy storage. Prior to our work, solar powered lasers only operated under highly concentrated sunlight (above 3000 suns), which makes them inapplicable. As Yablonovich presented in his paper ‘Thermodynamics of Daylight- pumped Lasers’, for lasers to operate with non-concentrated solar radiation, it requires a super lasing material that has a ratio between pump absorption and absorption at lasing wavelength that exceeds 105. This ratio is nearly three orders of magnitude higher than known lasing materials. In my postdoc, I showed that such a material can be tailored by cascaded energy transfer between available materials. Now, using this method, we were able to have demonstrate low threshold solar lasers operating at 200 suns. This work was published in Scientific Reports, and co- authored with my MIT supervisor, Marc Balo. Although this concept is my idea, and my student, Sergey Nechayev, did most of the experiments, it was done in Marc’s lab at MIT as a result of the delay in my lab In a second paper recently published in Scientific

Reports, with the same authors, I developed a new analytic tool for an incoherent pump for photonic devices. The simulation is based on a modified net radiation method where we incorporated photoluminescence and is much simpler and more reliable than existing Monte- Carlo or coherent simulations. We used this method for a solar powered laser and revealed, for the first time, that pump absorption cannot be maximal as in conventional lasers, but must be optimized due to the tradeoff between pump absorption and absorption at lasing wavelength. A third paper on the thermodynamic efficiency limit of solar powered lasers, and its analog to solar cells is currently under review. These publications put us in the forefront of incoherent pumped laser research.

Shiri Chechik

Krill Prize 2017
Tel Aviv University

Dr. Shiri Chechik

Much of the research done on the foundations of algorithms, assumes that the algorithm  has full knowledge  of the input data and on future changes to the data. In spite of the theoretical appeal  and simplicity  of  this setting, the assumption that the algorithm has “full knowledge” does not always hold. Uncertainty and partial knowledge arise in many settings, for example:

  • where the data is very large (e.g. so called “large data” derived from internet transactions activity, where merely reading the entire data once is impossible, and only sampling is possible),
  • where the data is not static, i.e., changes occur over time (e.g., social networks where information is fluid, and often time critical),
  • where processing of the data is distributed over computation nodes, and each node has only local information. The large body of work in the literature on streaming, dynamic, online, fault-tolerant, and distributed algorithms, in fact, deals with variations of the cases described

Graph data is by and large the most ubiquitous form of data used in practice. Graphs are used for representing networks (e.g. social networks, semantic networks, communication networks and more general flow networks, epidemiology networks, biologic interaction networks), as well as complicated linear algebraic operators (matrices), pairwise similarity relations between abstract objects and so on. In spite of the importance of graph data in the algorithm design community, our current understanding of these problems in the streaming, dynamic, fault tolerant and distributed algorithm setting is very incomplete. In particular:

  • There are major discrepancies between the efficiency of deterministic and randomized
  • The situation concerning amortized individual worst case guarantees is not well understood.
  • It is unclear how performance changes as a function of the E.g., what type of changes are considered in dynamic settings? How much information may be transmitted in distributed settings, etc.

Dealing with such “partial knowledge” and uncertainty settings, while maintaining the quality of the results, raises complex issues that often add new and non-trivial mathematical challenges. I plan to study algorithmic questions at the interface with graph theory, data structures and distributed computing. In particular, I am interested in exploring the limits of uncertainty environments and determine what can and cannot be achieved in this setting. Designing solutions that handle uncertainty and partial knowledge is a huge field that covers many different dis_ciplines, such as distributed networks, biology (where often the network is too big), online algorithms, streaming, etc. I plan to focus on three main fundamental areas, where I have identified gaps in our understanding and where I believe that progress will have a significant impact.  The first area relates to dynan1ic algorithms,  that is, designing algorithms that maintain some key functionality or property of the graph, while the graph itself changes over time. The second area concerns distributed graph algorithms and in particular symmetry breaking problems (e.g., coloring, maximal independent sot, etc.). The third involves social networks, and in particular designing efficient algorithms for extracting useful information from social networks. I will next describe in more detail some of the problems I plan to study.

Dynamic algorithms In many real-world applications, including communication networks, VLSI design, graphics, and assembly planning, graphs may change over time. This raises the essential need to design solutions that can efficiently adapt to these changes, e.g., GPS navigation systems are required to provide alternative routes in case of congestion, the internet needs to adapt to new users and new pages and to be resilient to potential failures.

In the last three decades there has been a growing interest in dynamic graph algorithms. The trivial approach to solving problems in changing circumstances is to solve the problem from scratch subsequent to every change. E.g., when road  congestion  levels change,  one can recompute  the  new shortest  path,  given the  current  state of affairs. If the new input is somehow close to the previous input, one suspects that this should involve less computation than the trivial approach of forgetting everything and starting from scratch. In fact, this has been shown to be true in a wide (and growing) variety of problems, some of which are described hereinafter.

I plan to study the fundamental dynamic graph algorithms with large discrepancy between the state of the art upper and lower bounds. A partial list of these problems includes: distance oracles, dynamic single source shortest path and all pairs shortest path, decremental single-source reachability (SSR), decremental strongly connected components (SCC), and dynamic connectivity for undirected graphs , etc.

Distributed computing and Symmetry breaking In the last three decades there was a huge shift from central-based computing to distributed-based computing. We are surrounded by distributed networks and they play a central part of our everyday life, that includes, the internet, cellular networks, World Wide Web, wireless sensor networks, aircraft control systems and many more. Even biological systems, ranging from the molecular to  the  organism  level, are  distributer.d.

Distributed Computing is a massive and growing field of study. It deals with settings in which many processors

work in parallel, communicating and coordinating their actions by passing messages. These processors work together in order to achieve a common goal. Symmetry brealdng sits at  the heart  of distributed  computing  theory and is arguably one of the most important problem in distributed network computing. The input iu this setting is a distributed graph, where each node in the graph represents a processor and each processor (node) can transfer messages to only its neighbors in the graph.  Symmetry breaking concerns with studying problems such as Maximal Independent Set (MIS), Maximal Matching (MM), and proper graph Coloring. In the MIS problem, for example, after the algorithm ends, each node !mows if it is in the MIS or not and the set of nodes that are in the MIS forms a valid MIS in the input graph.  The main question in this framework  is whether  these tasks can be solved locally, that is, by exchanging messages between nodes at short distance in the graph.

Besides their theoretical appeal, symmetry breaking algorithms are also used as fundamental  building  block in many applications. (e.g. coloring has been used for scheduling in wireless networks, MIS has been used for resource scheduling for parallel threads in a multi-core environment). This area has been a subject of intensive research since the beginning of the 80s. However, despite extensive study, some major problems remain open and the gaps between upper and lower bounds are in many cases exponential.

Social  Networks   Recently, social networks have gained significant  popularity, attracting millions of users, on a daily-basis. The significant popularity of these sites and the enormous information they capture, make the  task of studying and analyzing these networks extremely important. In  a preliminary study, we address  the  question of how well can one estimate the similarities of two nodes by observing only the connections of a social network.

I plan to continue pursuing a better understanding of social networks, and analyzing both its theoretical and practical aspects. I think that this area is still in its infancy and there are many directions I wish to explore. In particular, I am interested in studying efficient algorithms for extracting useful information from social networks, e.g., finding similarities between users, detecting communities, identifying influential people or suspicious activity within the network. Such information can be used to optimize a variety of applications including search engines and  advertising  delivery systems.  I am also interested  in studying the  behavior  and  propagation  mechanisms of information spreading processes in social networks. More specifically, my research aims to identify basic parameters (of the society, the network, and the type of information) that govern the speed, efficiency, and shape of  the  propagation,  as well as to  develop mechanisms  that control or manipulate  these processes,  e.g.,  change their speed or shape.