Compiled by: Unjiu
Nature, February 29, 2024, Volume 626, Issue 8001
Title: Astronomy: Dwarf Galaxies as Primary Photon Sources for Cosmic Reionization
Abstract: Identifying the sources responsible for cosmic reionization, a crucial phase transition occurring approximately 600–800 million years post-Big Bang, has sparked ongoing debate. Some models advocate for quasars, citing their high ionizing emissivity and escape fractions (fesc). Conversely, others argue that bright galaxies, with their elevated fesc values, emit sufficient ionizing radiation to drive this process. Additionally, a subset of studies suggests that the abundance of faint galaxies, when integrated with models of stellar-mass-dependent ionizing efficiency and fesc, may dominate cosmic reionization. However, due to their extreme faintness, comprehensive spectroscopic studies of low-mass galaxies have been lacking. Here, we present an analysis of eight ultra-faint galaxies (in a restricted field) during the reionization epoch, exhibiting absolute magnitudes ranging from MUV?≈??17?mag to ?15?mag (down to 0.005L?). Our findings reveal that faint galaxies in the initial billion years of the Universe emit ionizing photons with log[ξion?(Hz?erg?1)]= 25.80±0.14, surpassing commonly assumed values by a factor of four. If this field represents the widespread distribution of faint galaxies, the photon emission rate exceeds that required for reionization, even with escape fractions as low as 5%. Materials Science
High fatigue resistance in a titanium alloy via near-void-free 3D printing
Authors: Zhan Qu, Zhenjun Zhang, Rui Liu, Ling Xu, Yining Zhang, Xiaotao Li, et al.
Abstract:
The promising potential of 3D printing, also known as additive manufacturing (AM), for structural materials has been hindered by their underwhelming fatigue performance. This deficiency often stems from the presence of microvoids generated during the current printing processes. Thus, the question arises: can eliminating these microvoids offer a viable solution for significantly enhancing the fatigue resistance of void-free AM (Net-AM) alloys?
By understanding the asynchronism of phase transformation and grain growth, our research team has successfully developed a Net-AM processing technique, which effectively reconstructs an almost void-free AM microstructure within the Ti-6Al-4V titanium alloy. We have identified the fatigue resistance of such AM microstructures and demonstrated that they exhibit a high fatigue limit of approximately 1 GPa, surpassing the fatigue resistance of all AM and forged titanium alloys, as well as other metallic materials.
These findings confirm the high fatigue resistance of Net-AM microstructures and highlight the potential advantages of AM processing in producing structural components with maximum fatigue strength. This bodes well for the further application of AM technologies in engineering fields.
Site-specific reactivity of stepped Pt surfaces driven by stress release
Authors: Guangdong Liu, Arthur J. Shih, Huiqiu Deng, Kasinath Ojha, Xiaoting Chen, Mingchuan Luo, et al. Link:
https://www.nature.com/articles/s41586-024-07090-z
Abstract:
Heterogeneous catalysts are extensively utilized in facilitating chemical reactions. While it's understood that these reactions typically occur on catalyst surfaces, only specific surface sites exhibit high catalytic activity. Thus, the core of catalysis research lies in identifying these active sites and optimizing their functionality. A conventional approach involves categorizing active sites based on different surface configurations, such as terraces and steps.
However, this simplistic classification often leads to significant errors in predicting catalyst activity and qualitative uncertainties regarding active sites, thereby constraining opportunities for catalyst design. Using stepped Pt(111) surfaces and the electrochemical oxygen reduction reaction (ORR) as a case study, we illustrate that the primary cause of these discrepancies is a simplified categorization overlooking atomic site-specific reactivity driven by surface stress release.
Specifically, surface stress release at steps induces non-uniform strain fields, with compression rates reaching up to 5.5%, resulting in distinct electronic structures and reactivity for terrace atoms with identical local coordination. This leads to an atomic site-specific enhancement of ORR activity. For terrace atoms adjacent to both sides of the step edge, the enhancement is up to 50 times higher than that of atoms in the middle of the terrace, offering avenues to control ORR reactivity by adjusting terrace widths or managing external stress.
Hence, the discovery of this synergy provides a fresh perspective for understanding the fundamentals of catalytically active atomic sites and the design principles of heterogeneous catalysts. Supramolecular Polymers Form Tactoids through Liquid-Liquid Phase Separation
Abstract:
Liquid-liquid phase separation (LLPS) of biopolymers has emerged as a key process in the formation of membraneless organelles, showcasing diverse biological functionalities. The ongoing exploration into the interplay between LLPS and macromolecular condensation has led to intriguing findings. While synthetic supramolecular polymers represent the non-covalent counterparts of macromolecules, their ability to undergo LLPS has remained largely unreported.
Our study unveils a fascinating phenomenon: continuously elongating fibrils resulting from supramolecular polymerizations of synthetic components drive phase separation, forming highly anisotropic aqueous liquid droplets known as tactoids, through an entropy-driven mechanism. The crowded environment, modulated by dextran concentration, not only influences the kinetics of supramolecular polymerizations but also governs LLPS properties, including phase-separation kinetics, morphology, internal order, fluidity, and mechanical characteristics of the resultant tactoids. Moreover, interfaces such as substrate-liquid and liquid-liquid interfaces accelerate LLPS of supramolecular polymers, facilitating the creation of myriad three-dimensional ordered structures, including meticulously arranged arrays of micrometer-long tactoids on surfaces.
The versatility and potential of supramolecular polymerizations in shaping emerging morphologies are underscored by our findings across various supramolecular polymers, thus opening up a novel domain of materials ranging from intricately structured aqueous solutions stabilized by LLPS to nanoscopic soft matter.
Identifying General Reaction Conditions by Bandit Optimization
Authors: Jason Y. Wang, Jason M. Stevens, Stavros K. Kariofillis, Mai-Jan Tom, Dung L. Golden, Jun Li, et al.
Link: https://www.nature.com/articles/s41586-024-07021-y
Abstract:
There is a pressing need for reaction conditions that can be universally applied across a wide range of substrates, particularly in pharmaceutical and chemical industries. Despite numerous methods available to assess the general applicability of developed conditions, there is a scarcity of universal approaches to efficiently discover these conditions during optimization processes.
Our research group presents the design, implementation, and application of reinforcement learning bandit optimization models to identify generally applicable conditions through efficient condition sampling and evaluation of experimental feedback. Performance benchmarking on existing datasets demonstrates statistically high accuracies in identifying general conditions, with a remarkable up to 31% improvement over baselines that emulate state-of-the-art optimization approaches. Experimental investigations were conducted on palladium-catalyzed imidazole C–H arylation reaction, aniline amide coupling reaction, and phenol alkylation reaction to assess the practical applications and functionalities of the bandit optimization model. In less than 15% of the expert-designed reaction space, our study identified the most generally applicable yet insufficiently studied reaction conditions for the aforementioned reactions.
Online Images Amplify Gender Bias
Authors: Douglas Guilbeault, Solène Delecourt, Tasker Hull, Bhargav Srinivasa Desikan, Mark Chu, & Ethan Nadler Link:
https://www.nature.com/articles/s41586-024-07068-x
Abstract:
Every year, there's a trend where people spend less time reading and more time engaging with images online. Platforms like Google and Wikipedia witness millions of image downloads daily, while social media platforms like Instagram and TikTok thrive on visual content exchange. Simultaneously, news agencies and advertisers are increasingly leveraging images to capture online attention, as they are processed faster, implicitly, and memorably compared to text.
This study reveals that the surge in online images significantly amplifies gender bias, both in terms of its statistical prevalence and psychological impact. Analyzing over a million images from Google, Wikipedia, and the Internet Movie Database (IMDb), along with billions of words from these platforms, we investigated gender associations across 3,495 social categories (e.g., "nurse" or "banker"). The findings indicate that gender bias is consistently more pronounced in images compared to text, regardless of whether the categories are traditionally associated with females or males.
Furthermore, we observe a pronounced underrepresentation of women in online images, a disparity far more severe compared to text, public opinion, and US census data. A nationally representative, preregistered experiment demonstrates that searching for occupation-related images on Google, as opposed to textual descriptions, exacerbates gender bias in participants' perceptions.
Addressing the societal implications of this widespread shift towards visual communication is crucial for fostering a fair and inclusive future for the internet.
