《自然》（20231207出版）一周论文导读

 

编译 | 冯维维

Nature, Volume 624 Issue 7990, 7 December 2023

《自然》 第624卷，7990期，2023年12月7日

  ?

物理学Physics

Self-assembled photonic cavities with atomic-scale confinement

具有原子尺度约束的自组装光子腔

▲ 作者：Ali Nawaz Babar, Thor August Schimmell Weis, Konstantinos Tsoukalas, Shima Kadkhodazadeh, Guillermo Arregui, Babak Vosoughi Lahijani & S?ren Stobbe

▲ 链接：

https://www.nature.com/articles/s41586-023-06736-8

▲ 摘要：

尽管自组装纳米技术的研究取得了巨大的进展，如大分子、纳米线和二维材料，但从纳米尺度到宏观尺度的合成自组装方法仍然不可扩展，不如生物自组装。

相比之下，平面半导体技术由于其固有的可扩展性而产生了巨大的技术影响，但它似乎无法达到自组装的原子尺寸。研究者使用表面力，包括卡西米尔-范德华相互作用，来确定自组装和自对准悬浮硅纳米结构，尽管只使用传统光刻和蚀刻，其空洞特征远低于传统光刻和蚀刻的长度尺度。

该方法具有显著的鲁棒性，自组装阈值单调依赖于数千个被测器件的所有控制参数。研究者通过制造任何其他已知方法都无法制造的纳米结构来说明这些概念的潜力：波导耦合高Q硅光子腔，将电信光子限制在2纳米的气隙中，宽高比为100，对应于比衍射极限低100倍以上的模式体积。

扫描透射电子显微镜测量证实了制造亚纳米尺寸设备的能力。研究者表示该技术将自组装的原子尺寸与平面半导体的可扩展性相结合，是迈向新一代制造技术的第一步。

▲ Abstract：

Despite tremendous progress in research on self-assembled nanotechnological building blocks, such as macromolecules, nanowires and two-dimensional materials, synthetic self-assembly methods that bridge the nanoscopic to macroscopic dimensions remain unscalable and inferior to biological self-assembly. By contrast, planar semiconductor technology has had an immense technological impact, owing to its inherent scalability, yet it seems unable to reach the atomic dimensions enabled by self-assembly. Here, we use surface forces, including Casimir–van der Waals interactions, to deterministically self-assemble and self-align suspended silicon nanostructures with void features well below the length scales possible with conventional lithography and etching, despite using only conventional lithography and etching. The method is remarkably robust and the threshold for self-assembly depends monotonically on all the governing parameters across thousands of measured devices. We illustrate the potential of these concepts by fabricating nanostructures that are impossible to make with any other known method: waveguide-coupled high-Q silicon photonic cavities that confine telecom photons to 2?nm air gaps with an aspect ratio of 100, corresponding to mode volumes more than 100 times below the diffraction limit. Scanning transmission electron microscopy measurements confirm the ability to build devices with sub-nanometre dimensions. Our work constitutes the first steps towards a new generation of fabrication technology that combines the atomic dimensions enabled by self-assembly with the scalability of planar semiconductors.

Single-molecule electron spin resonance by means of atomic force microscopy

原子力显微镜下的单分子电子自旋共振

▲ 作者：Lisanne Sellies, Raffael Spachtholz, Sonja Bleher, Jakob Eckrich, Philipp Scheuerer & Jascha Repp

▲ 链接：

https://www.nature.com/articles/s41586-023-06754-6

▲ 摘要：

理解和控制开放量子系统中的退相干是科学研究的基础，而实现长相干时间对于量子信息处理至关重要。

尽管单个系统已经取得了很大的进展，并且单自旋的电子自旋共振（ESR）已经被证明具有纳米级分辨率，但在许多复杂的固态量子系统中，对退相干的理解最终需要将环境控制到原子尺度，这可能通过扫描探针显微镜及其原子和分子表征和操作能力来实现。

因此，最近在扫描隧道显微镜中实现的ESR是实现这一目标的一个里程碑，并很快被相干振荡的演示和真实空间原子分辨率的核自旋所遵循。原子操纵甚至激发了实现第一个人工原子尺度量子器件的雄心。然而，这种方法固有的基于电流的传感限制了相干时间。

研究者展示了泵探针ESR原子力显微镜（AFM）检测电子自旋跃迁之间的非平衡态的单个并五苯分子。这些跃迁的光谱表现出亚纳米电子伏特的光谱分辨率，允许局部区分分子，只是在它们的同位素配置不同。

此外，电子自旋可以在数十微秒内进行相干操纵。我们预计单分子ESR-AFM可以与原子操作和表征相结合，从而为了解原子定义良好的量子元素中退相干的原子起源和基础量子传感实验铺平道路。

▲ Abstract：

Understanding and controlling decoherence in open quantum systems is of fundamental interest in science, whereas achieving long coherence times is critical for quantum information processing. Although great progress was made for individual systems, and electron spin resonance (ESR) of single spins with nanoscale resolution has been demonstrated, the understanding of decoherence in many complex solid-state quantum systems requires ultimately controlling the environment down to atomic scales, as potentially enabled by scanning probe microscopy with its atomic and molecular characterization and manipulation capabilities. Consequently, the recent implementation of ESR in scanning tunnelling microscopy represents a milestone towards this goal and was quickly followed by the demonstration of coherent oscillations and access to nuclear spins with real-space atomic resolution. Atomic manipulation even fuelled the ambition to realize the first artificial atomic-scale quantum devices. However, the current-based sensing inherent to this method limits coherence times. Here we demonstrate pump–probe ESR atomic force microscopy (AFM) detection of electron spin transitions between non-equilibrium triplet states of individual pentacene molecules. Spectra of these transitions exhibit sub-nanoelectronvolt spectral resolution, allowing local discrimination of molecules that only differ in their isotopic configuration. Furthermore, the electron spins can be coherently manipulated over tens of microseconds. We anticipate that single-molecule ESR-AFM can be combined with atomic manipulation and characterization and thereby paves the way to learn about the atomistic origins of decoherence in atomically well-defined quantum elements and for fundamental quantum-sensing experiments.

化学Chemistry

Scaling deep learning for materials discovery

扩展深度学习用于材料发现

▲ 作者：Amil Merchant, Simon Batzner, Samuel S. Schoenholz, Muratahan Aykol, Gowoon Cheon & Ekin Dogus Cubuk

▲ 链接：

https://www.nature.com/articles/s41586-023-06735-9

▲ 摘要：

新型功能材料使从清洁能源到信息处理等技术应用取得根本性突破。从微芯片到电池和光伏，无机晶体的发现一直受到昂贵的试错方法的阻碍。同时，随着数据和计算量的增加，语言、视觉和生物学的深度学习模型也显示出了新兴的预测能力。

研究者展示了大规模训练的图网络可以达到前所未有的泛化水平，将材料发现的效率提高了一个数量级。在持续研究中发现的4.8万个稳定晶体的基础上，效率的提高使人们能够在目前的凸壳下发现220万个结构，其中许多结构超出了人类以前的化学直觉。

这项研究代表了人类已知的稳定物质的一个数量级的扩展。最终凸包上的稳定发现将用于筛选技术应用，正如作者对分层材料和固体电解质候选物的演示一样。

在稳定结构中，736个已经独立实验实现。数以亿计的第一性原理计算的规模和多样性也为下游应用解锁了建模能力，特别是导致高度精确和强大的学习原子间势，可用于凝聚态分子动力学模拟和高保真离子电导率零射预测。

▲ Abstract：

Novel functional materials enable fundamental breakthroughs across technological applications from clean energy to information processing. From microchips to batteries and photovoltaics, discovery of inorganic crystals has been bottlenecked by expensive trial-and-error approaches. Concurrently, deep-learning models for language, vision and biology have showcased emergent predictive capabilities with increasing data and computation. Here we show that graph networks trained at scale can reach unprecedented levels of generalization, improving the efficiency of materials discovery by an order of magnitude. Building on 48,000 stable crystals identified in continuing studies, improved efficiency enables the discovery of 2.2 million structures below the current convex hull, many of which escaped previous human chemical intuition. Our work represents an order-of-magnitude expansion in stable materials known to humanity. Stable discoveries that are on the final convex hull will be made available to screen for technological applications, as we demonstrate for layered materials and solid-electrolyte candidates. Of the stable structures, 736 have already been independently experimentally realized. The scale and diversity of hundreds of millions of first-principles calculations also unlock modelling capabilities for downstream applications, leading in particular to highly accurate and robust learned interatomic potentials that can be used in condensed-phase molecular-dynamics simulations and high-fidelity zero-shot prediction of ionic conductivity.

An autonomous laboratory for the accelerated synthesis of novel materials

一个加速合成新材料的自主实验室

▲ 作者：Nathan J. Szymanski, Bernardus Rendy, Yuxing Fei, Rishi E. Kumar, Tanjin He, David Milsted, Matthew J. McDermott, Max Gallant, Ekin Dogus Cubuk, Amil Merchant, Haegyeom Kim, Anubhav Jain, Christopher J. Bartel, Kristin Persson, Yan Zeng & Gerbrand Ceder

▲ 链接：

https://www.nature.com/articles/s41586-023-06734-w

▲ 摘要：

为了缩小新材料的计算筛选和实验实现之间的差距，研究者引入了自主实验室A-Lab，用于无机粉末的固态合成。该平台使用计算、文献中的历史数据、机器学习（ML）和主动学习来计划和解释使用机器人进行的实验结果。

在17天的连续运行中，A- Lab从58个目标中实现了41种新化合物，包括各种氧化物和磷酸盐，这些目标是使用材料项目和谷歌深度思维的大规模从头算相稳定性数据确定的。

合成配方由基于文献的自然语言模型提出，并使用基于热力学的主动学习方法进行优化。分析失败的合成为改进现有的材料筛选和合成设计技术提供了直接和可行的建议。

高成功率证明了人工智能驱动平台在自主材料发现方面的有效性，并激励了计算、历史知识和机器人技术的进一步整合。

▲ Abstract：

To close the gap between the rates of computational screening and experimental realization of novel materials, we introduce the A-Lab, an autonomous laboratory for the solid-state synthesis of inorganic powders. This platform uses computations, historical data from the literature, machine learning (ML) and active learning to plan and interpret the outcomes of experiments performed using robotics. Over 17?days of continuous operation, the A-Lab realized 41 novel compounds from a set of 58 targets including a variety of oxides and phosphates that were identified using large-scale ab initio phase-stability data from the Materials Project and Google DeepMind. Synthesis recipes were proposed by natural-language models trained on the literature and optimized using an active-learning approach grounded in thermodynamics. Analysis of the failed syntheses provides direct and actionable suggestions to improve current techniques for materials screening and synthesis design. The high success rate demonstrates the effectiveness of artificial-intelligence-driven platforms for autonomous materials discovery and motivates further integration of computations, historical knowledge and robotics.

气候和生态Climate & Ecology

Aligning climate scenarios to emissions inventories shifts global benchmarks

将气候情景与排放清单相结合会改变全球基准

▲ 作者：Matthew J. Gidden, Thomas Gasser, Giacomo Grassi, Nicklas Forsell, Iris Janssens, William F. Lamb, Jan Minx, Zebedee Nicholls, Jan Steinhauser & Keywan Riahi

▲ 链接：

ttps://www.nature.com/articles/s41586-023-06724-y

▲ 摘要：

评估在实现《巴黎协定》方面取得的全球进展，需要始终如一地衡量各国针对模拟缓解途径采取的总体行动和承诺。

然而，国家温室气体清单（NGHGIs）和人为排放的科学评估遵循不同的陆地碳通量核算惯例，导致目前的排放估计值存在很大差异，这一差距将随着时间的推移而扩大。研究者使用最先进的方法和土地碳循环模拟器，将政府间气候变化专门委员会评估的缓解途径与国家温室气体地理信息系统进行比较。

研究结果发现，当使用NGHGI公约计算时，关键的全球缓解基准变得更难实现，这既需要更早的二氧化碳净零排放时间，也需要更低的累积排放量。

此外，减弱自然碳清除过程，如碳施肥，可以掩盖人为的陆地清除努力，其结果是，到2100年，全球温室气体地理区域的陆地碳通量最终可能成为排放源。研究结果对全球盘点很重要，表明各国需要提高各自气候目标的集体雄心，以保持与全球温度目标的一致。

▲ Abstract：

Taking stock of global progress towards achieving the Paris Agreement requires consistently measuring aggregate national actions and pledges against modelled mitigation pathways. However, national greenhouse gas inventories (NGHGIs) and scientific assessments of anthropogenic emissions follow different accounting conventions for land-based carbon fluxes resulting in a large difference in the present emission estimates, a gap that will evolve over time. Using state-of-the-art methodologies and a land carbon-cycle emulator, we align the Intergovernmental Panel on Climate Change (IPCC)-assessed mitigation pathways with the NGHGIs to make a comparison. We find that the key global mitigation benchmarks become harder to achieve when calculated using the NGHGI conventions, requiring both earlier net-zero CO2 timing and lower cumulative emissions. Furthermore, weakening natural carbon removal processes such as carbon fertilization can mask anthropogenic land-based removal efforts, with the result that land-based carbon fluxes in NGHGIs may ultimately become sources of emissions by 2100. Our results are important for the Global Stocktake6, suggesting that nations will need to increase the collective ambition of their climate targets to remain consistent with the global temperature goals.

Integrated global assessment of the natural forest carbon potential

天然林碳潜力的全球综合评估

▲ 作者：Lidong Mo, Constantin M. Zohner, Peter B. Reich, Jingjing Liang, Sergio de Miguel, Gert-Jan Nabuurs, Susanne S. Renner, Johan van den Hoogen, Arnan Araza, Martin Herold, Leila Mirzagholi, Haozhi Ma, Colin Averill, Oliver L. Phillips, Javier G. P. Gamarra, Iris Hordijk, Devin Routh, Meinrad Abegg, Yves C. Adou Yao, Giorgio Alberti, Angelica M. Almeyda Zambrano, Braulio Vilchez Alvarado, Esteban Alvarez-Dávila, Patricia Alvarez-Loayza, …Thomas W. Crowther Show authors

▲ 链接：

https://www.nature.com/articles/s41586-023-06723-z

▲ 摘要：

森林是一个重要的陆地碳汇，但土地利用和气候的人为变化大大缩小了这一系统的规模。用于量化全球森林碳损失的遥感估算具有相当大的不确定性，缺乏全面的地面评估来对这些估算进行基准测试。

研究者结合了几种地面来源和卫星来源的方法来评估农业和城市土地以外的全球森林碳潜力的规模。尽管存在区域差异，但这些预测在全球范围内显示出显著的一致性，地面来源和卫星估算值之间的差异仅为12%。

目前，全球森林碳储量明显低于自然潜力，低人类足迹地区总亏缺226 Gt（模型范围为151 ~ 363 Gt）。这一潜力的大部分（61%，139亿吨碳当量）位于有森林的地区，在这些地区，生态系统保护可以使森林恢复到成熟。其余39%（87亿吨碳当量）的潜力存在于森林被砍伐或破碎的地区。

虽然森林不能替代减排，但研究结果支持这样一种观点，即保护、恢复和可持续管理多样化的森林为实现全球气候和生物多样性目标做出了宝贵的贡献。

▲ Abstract：

Forests are a substantial terrestrial carbon sink, but anthropogenic changes in land use and climate have considerably reduced the scale of this system. Remote-sensing estimates to quantify carbon losses from global forests are characterized by considerable uncertainty and we lack a comprehensive ground-sourced evaluation to benchmark these estimates. Here we combine several ground-sourced6 and satellite-derived approaches to evaluate the scale of the global forest carbon potential outside agricultural and urban lands. Despite regional variation, the predictions demonstrated remarkable consistency at a global scale, with only a 12% difference between the ground-sourced and satellite-derived estimates. At present, global forest carbon storage is markedly under the natural potential, with a total deficit of 226 Gt (model range = 151–363 Gt) in areas with low human footprint. Most (61%, 139 Gt C) of this potential is in areas with existing forests, in which ecosystem protection can allow forests to recover to maturity. The remaining 39% (87 Gt C) of potential lies in regions in which forests have been removed or fragmented. Although forests cannot be a substitute for emissions reductions, our results support the idea that the conservation, restoration and sustainable management of diverse forests offer valuable contributions to meeting global climate and biodiversity targets.