自然（20220210出版）一周论文导读

　　编译 | 未玖
　　Nature,  10 February 2022, VOL 602, ISSUE 7896
　　《自然》 2022年2月10日，第602卷，7896期
　　天文学 Astronomy
　　A white dwarf accreting planetary material determined from X-ray observations
　　X射线观测到白矮星吸积行星物质
　　作者：Tim Cunningham, Peter J. Wheatley, Pier-Emmanuel Tremblay, Boris T. Gänsicke, George W. King, Odette Toloza, et al.
　　链接：
　　https://www.nature.com/articles/s41586-021-04300-w
　　摘要
　　大多数白矮星的大气受到重元素的污染，这些重元素预计会在短时间内从可见层下沉。这被解释为白矮星不断吸积小行星、彗星和巨行星碎片的标志。探测一些白矮星周围的碎片盘和行星碎片凌星现象支持了这一设想。
　　然而，光球金属只是持续吸积的间接证据，推断的吸积率和母体成分在很大程度上取决于白矮星大气中扩散和混合过程的模型。
　　研究组报道了被污染白矮星G29–38的X射线4.4σ探测。根据测得的X射线光度，他们推导出瞬时吸积率MX=1.63 109 g s 1，且独立于恒星大气模型。该比率高于过去对G29–38光球丰度研究所作的估计，表明可能需要对流过冲来模拟白矮星碎片聚集的光谱。
　　研究组测量出kBT= 0.5   0.2 keV的低等离子体温度，证实了此前预测的白矮星以低吸积率吸积的爆炸方式。
　　Abstract
　　The atmospheres of a large proportion of white dwarf stars are polluted by heavy elements that are expected to sink out of visible layers on short timescales. This has been interpreted as a signature of ongoing accretion of debris from asteroids, comets and giant planets. This scenario is supported by the detection of debris discs and transits of planetary fragments around some white dwarfs. However, photospheric metals are only indirect evidence for ongoing accretion, and the inferred accretion rates and parent body compositions heavily depend on models of diffusion and mixing processes within the white dwarf atmosphere. Here we report a 4.4σ detection of X-rays from a polluted white dwarf, G29–38. From the measured X-ray luminosity, we derive an instantaneous accretion rate of MX=1.63 109 g s 1, which is independent of stellar atmosphere models. This rate is higher than estimates from past studies of the photospheric abundances of G29–38, suggesting that convective overshoot may be needed to model the spectra of debris-accreting white dwarfs. We measure a low plasma temperature of kBT = 0.5   0.2 keV, corroborating the predicted bombardment solution for white dwarfs accreting at low accretion rates.
　　物理学 Physics
　　Real-space visualization of intrinsic magnetic fields of an antiferromagnet
　　反铁磁体内禀磁场的实空间可视化
　　作者：Yuji Kohno, Takehito Seki, Scott D. Findlay, Yuichi Ikuhara & Naoya Shibata
　　链接：
　　https://www.nature.com/articles/s41586-021-04254-z
　　摘要
　　原子尺度的磁性结构表征，是材料和器件中纳米磁性设计与调控的核心。然而，在该维度上，磁场的实空间可视化一直颇具挑战性。近年来，原子分辨率差分相衬扫描透射电子显微镜（DPC STEM）已能够直接成像单个原子内部的电场分布。
　　研究组展示了在无磁场环境中使用原子分辨率DPC STEM实现反铁磁赤铁矿（α-Fe2O3）内部磁场分布的实空间可视化。
　　在去除原子电场引起的相移分量并通过单元平均法提高信噪比后，研究组实现了α-Fe2O3内禀磁场的实空间可视化。这些研究结果为许多磁性结构的实空间表征提供了新途径。
　　Abstract
　　Characterizing magnetic structures down to atomic dimensions is central to the design and control of nanoscale magnetism in materials and devices. However, real-space visualization of magnetic fields at such dimensions has been extremely challenging. In recent years, atomic-resolution differential phase contrast scanning transmission electron microscopy (DPC STEM) has enabled direct imaging of electric field distribution even inside single atoms. Here we show real-space visualization of magnetic field distribution inside antiferromagnetic haematite (α-Fe2O3) using atomic-resolution DPC STEM in a magnetic-field-free environment. After removing the phase-shift component due to atomic electric fields and improving the signal-to-noise ratio by unit-cell averaging, real-space visualization of the intrinsic magnetic fields in α-Fe2O3 is realized. These results open a new possibility for real-space characterization of many magnetic structures.
　　Ferroelectric incommensurate spin crystals
　　铁电不相称自旋晶体
　　作者：Dorin Rusu, Jonathan J. P. Peters, Thomas P. A. Hase, James A. Gott, Gareth A. A. Nisbet, Jörg Strempfer, et al.
　　链接：
　　https://www.nature.com/articles/s41586-021-04260-1
　　摘要
　　铁性材料，尤其是铁磁体，在特定的电和力学边界条件下，可形成复杂的拓扑自旋结构，如漩涡和斯格明子。
　　在专用铁电系统中，尤其是在PbTiO3/SrTiO3等铁电-绝缘体超晶格中，人们已观察到简单的涡状电偶极子拓扑结构，但后来由于其高去极化场，被证明是一个模型系统。迄今为止，还没有实验观察到由Dzyaloshinskii–Moriya相互作用（DMi）驱动的有序磁自旋晶格的等效电偶极。
　　研究组探讨了夹在SrRuO3电极间的单一PbTiO3外延层的畴结构。他们观察到周期性的顺时针和逆时针铁电旋涡，沿其环形核心受到二阶调制。计算结果支持的拓扑结构是具有两个正交周期调制的迷宫状图案，形成了一个不相称的极性晶体，其铁电性类似于最近在铁磁材料中发现的不相称自旋晶体。
　　这些发现进一步模糊了突现铁磁和铁电拓扑之间的边界，为未来磁DMi驱动相的电对应物的实验实现扫清了道路。
　　Abstract
　　Ferroics, especially ferromagnets, can form complex topological spin structures such as vortices and skyrmions when subjected to particular electrical and mechanical boundary conditions. Simple vortex-like, electric-dipole-based topological structures have been observed in dedicated ferroelectric systems, especially ferroelectric–insulator superlattices such as PbTiO3/SrTiO3, which was later shown to be a model system owing to its high depolarizing field. To date, the electric dipole equivalent of ordered magnetic spin lattices driven by the Dzyaloshinskii–Moriya interaction (DMi) has not been experimentally observed. Here we examine a domain structure in a single PbTiO3 epitaxial layer sandwiched between SrRuO3 electrodes. We observe periodic clockwise and anticlockwise ferroelectric vortices that are modulated by a second ordering along their toroidal core. The resulting topology, supported by calculations, is a labyrinth-like pattern with two orthogonal periodic modulations that form an incommensurate polar crystal that provides a ferroelectric analogue to the recently discovered incommensurate spin crystals in ferromagnetic materials. These findings further blur the border between emergent ferromagnetic and ferroelectric topologies, clearing the way for experimental realization of further electric counterparts of magnetic DMi-driven phases.
　　材料科学 Materials Science
　　A highly distorted ultraelastic chemically complex Elinvar alloy
　　一种高度变形超弹性化学复杂的艾林瓦合金
　　作者：Q. F. He, J. G. Wang, H. A. Chen, Z. Y. Ding, Z. Q. Zhou, L. H. Xiong, et al.
　　链接：
　　https://www.nature.com/articles/s41586-021-04309-1
　　摘要
　　高性能超弹性金属的研发具有超高强度、大弹性应变极限和温度不敏感的弹性模量（艾林瓦效应），对于从致动器、医疗设备到高精度仪器的各种工业应用都至关重要。由于位错易滑动，体晶金属的弹性应变极限通常小于1%。
　　形状记忆合金（包括胶质金属和应变玻璃合金）的弹性应变极限可高达几个百分点，虽然这是伪弹性的结果，且伴随着巨大的能量耗散。近年来，化学性质复杂的合金，如＂高熵＂合金，因其良好性能引发了人们广泛的研究兴趣。
　　在这项工作中，研究组报道了一种化学复杂的合金，其原子尺寸错配较大，常规合金通常无法承受。该合金在室温下具有较高的弹性应变极限（约2%）和极低的内耗（小于2    10 4 ）。
　　更有趣的是，这种合金表现出非凡的艾林瓦效应，在室温和627 （900K）之间的弹性模量近乎恒定，迄今为止报道的现有合金均无法与之比拟。
　　Abstract
　　The development of high-performance ultraelastic metals with superb strength, a large elastic strain limit and temperature-insensitive elastic modulus (Elinvar effect) are important for various industrial applications, from actuators and medical devices to high-precision instruments. The elastic strain limit of bulk crystalline metals is usually less than 1 per cent, owing to dislocation easy gliding. Shape memory alloys—including gum metals and strain glass alloys—may attain an elastic strain limit up to several per cent, although this is the result of pseudo-elasticity and is accompanied by large energy dissipation. Recently, chemically complex alloys, such as ‘high-entropy’ alloys, have attracted tremendous research interest owing to their promising properties. In this work we report on a chemically complex alloy with a large atomic size misfit usually unaffordable in conventional alloys. The alloy exhibits a high elastic strain limit (approximately 2 per cent) and a very low internal friction (less than 2    10 4 ) at room temperature. More interestingly, this alloy exhibits an extraordinary Elinvar effect, maintaining near-constant elastic modulus between room temperature and 627 degrees Celsius (900 kelvin), which is, to our knowledge, unmatched by the existing alloys hitherto reported.
　　人工智能 Artificial Intelligence
　　Outracing champion Gran Turismo drivers with deep reinforcement learning
　　通过深度强化学习超越Gran Turismo冠军级赛车手
　　作者：Peter R. Wurman, Samuel Barrett, Kenta Kawamoto, James MacGlashan, Kaushik Subramanian, Thomas J. Walsh, et al.
　　链接：
　　https://www.nature.com/articles/s41586-021-04357-7
　　摘要
　　人工智能的许多潜在应用涉及与人类交互时在物理系统中做出实时决策。赛车就是这种情况的一个典型代表；赛车手必须执行复杂的战术性操作以超越或阻挡对手，同时在牵引力极限下驾驶赛车。
　　赛车模拟，比如PlayStation游戏Gran Turismo，忠实地再现了真实赛车的非线性控制挑战，同时也封装了复杂的多代理交互。
　　研究组介绍了他们如何训练Gran Turismo的AI代理，使其能够与世界上最优秀的电子竞技赛车手相匹敌。他们将最先进的、无模型的深度强化学习算法与混合场景训练相结合，来学习一种综合控制策略，将卓越的速度与令人印象深刻的战术相结合。
　　此外，研究组还构建了一个奖励函数，使AI代理能够在遵守重要但灵活的赛车规则的同时保持竞争力。最终研究组的AI代理Gran Turismo Sophy崭露头角，在正面竞争中战胜了四名世界顶级Gran Turismo赛车手。
　　通过描述如何训练冠军级别的赛车手，研究组展示了使用这些技术来控制复杂动态系统的机遇和挑战，在这些领域中，AI代理必须尊重灵活定义的人类规则。
　　Abstract
　　Many potential applications of artificial intelligence involve making real-time decisions in physical systems while interacting with humans. Automobile racing represents an extreme example of these conditions; drivers must execute complex tactical manoeuvres to pass or block opponents while operating their vehicles at their traction limits. Racing simulations, such as the PlayStation game Gran Turismo, faithfully reproduce the non-linear control challenges of real race cars while also encapsulating the complex multi-agent interactions. Here we describe how we trained agents for Gran Turismo that can compete with the world’s best e-sports drivers. We combine state-of-the-art, model-free, deep reinforcement learning algorithms with mixed-scenario training to learn an integrated control policy that combines exceptional speed with impressive tactics. In addition, we construct a reward function that enables the agent to be competitive while adhering to racing’s important, but under-specified, sportsmanship rules. We demonstrate the capabilities of our agent, Gran Turismo Sophy, by winning a head-to-head competition against four of the world’s best Gran Turismo drivers. By describing how we trained championship-level racers, we demonstrate the possibilities and challenges of using these techniques to control complex dynamical systems in domains where agents must respect imprecisely defined human norms.
　　地球科学 Earth Science
　　Superionic iron alloys and their seismic velocities in Earth’s inner core
　　超离子铁合金及其在地球内核中的地震速度
　　作者：Yu He, Shichuan Sun, Duck Young Kim, Bo Gyu Jang, Heping Li & Ho-kwang Mao
　　链接：
　　https://www.nature.com/articles/s41586-021-04361-x
　　摘要
　　地球内核（IC）的密度低于纯铁，这表明其内部存在轻元素。硅、硫、碳、氧和氢被认为是候选元素，人们研究了铁-轻元素合金的性能以约束IC成分。轻元素对铁合金的地震速度、熔化温度和热导率有很大影响。然而，人们很少考虑IC中轻元素的状态。
　　研究组利用第一性原理分子动力学模拟，发现在IC条件下，六方密排铁中的氢、氧和碳转变为超离子态，表现出像流体一样的高扩散系数。这表明IC可以处于超离子态，而非正常固态。
　　流体轻元素导致地震速度大幅降低，接近IC的地震学观测值。横波波速的大幅降低为软IC提供了一种解释。此外，轻元素对流对IC地震学结构和磁场也有潜在影响。
　　Abstract
　　Earth’s inner core (IC) is less dense than pure iron, indicating the existence of light elements within it. Silicon, sulfur, carbon, oxygen and hydrogen have been suggested to be the candidates, and the properties of iron–light-element alloys have been studied to constrain the IC composition. Light elements have a substantial influence on the seismic velocities, the melting temperatures and the thermal conductivities of iron alloys. However, the state of the light elements in the IC is rarely considered. Here, using ab initio molecular dynamics simulations, we find that hydrogen, oxygen and carbon in hexagonal close-packed iron transform to a superionic state under the IC conditions, showing high diffusion coefficients like a liquid. This suggests that the IC can be in a superionic state rather than a normal solid state. The liquid-like light elements lead to a substantial reduction in the seismic velocities, which approach the seismological observations of the IC. The substantial decrease in shear-wave velocity provides an explanation for the soft IC. In addition, the light-element convection has a potential influence on the IC seismological structure and magnetic field.