附件中的文件是我写的文章,主要讨论应变对RuO2磁有序的影响,你帮我看看文章总结一下
Tạo vào: 2 tháng 1, 2026
Trả lời bằng GPT-5.2 Thinking bởi Chat01
Tạo vào: 2 tháng 1, 2026
Trả lời bằng GPT-5.2 Thinking bởi Chat01
附件中的文件是我写的文章,主要讨论应变对RuO2磁有序的影响,你帮我看看文章总结一下
下面是我对你这篇《基于第一性原理计算的 RuO₂ 应变诱发磁有序》的结构化总结(按“研究问题—方法—结果—机制—意义/结论”来梳理):
你把 RuO₂ 作为“潜在交错磁体”的候选材料,但针对“是否存在磁有序”这一点,近期实验上出现了较大争议。文章聚焦于:外延整体应变是否能触发 AFM?在临界附近的局部畸变是否能进一步诱导 SDW 成为基态? (基于第一性原理计算的RuO2应变诱发磁有序.docx)
Ru 原子磁矩随 ε 增大从 0 开始出现并增长,得到单胞的临界应变 εc=0.7404(无 SOC);考虑 SOC 后 εc≈0.7340,临界点基本不被 SOC 颠覆。(基于第一性原理计算的RuO2应变诱发磁有序.docx)
在简谐畸变下,Δh 从 8‰/14‰/20‰ 增大时,εc 只“缓慢下降”(从 0.7404 到约 0.72/0.69/0.66),说明磁有序出现与否主要由整体 ε 决定,局部畸变更多在细节层面调制磁矩与能量竞争。(基于第一性原理计算的RuO2应变诱发磁有序.docx)
你定义 ΔE = E_SDW − E_AFM 来判定谁更稳定。在 ε≈εc 时,ΔE 随 Δh 呈非单调,并在一定幅度(例如 Δh≥12‰)时可变为负值,意味着 SDW 成为更低能态;且你指出外延薄膜里 10‰ 量级局部畸变是常见的,因此这种机制在实验上是“有机会发生”的。(基于第一性原理计算的RuO2应变诱发磁有序.docx) (基于第一性原理计算的RuO2应变诱发磁有序.docx)
你提出假设:SDW 在波节处本征地趋向“低磁矩/近无磁”,当局部压缩进一步压低该区域磁矩时,SDW 反而能在波节处获得相对 AFM 的能量优势,从而推动 ΔE 变负;并给出可检验判断:** ΔE 的主要贡献来自波节附近**。(基于第一性原理计算的RuO2应变诱发磁有序.docx)
你计算得到 ΔE per cell 随周期 n 先降后升,并观察到“Δh 越小,最稳定 SDW 所需 n 越大”,最终提炼出更易稳定 SDW 的特征波长范围 约 7–10 nm(对应 16a–22a)。(基于第一性原理计算的RuO2应变诱发磁有序.docx)
你进一步提出:临界附近且存在局部畸变时,AFM 与 SDW 的能量差仅 10–100 μeV(~10 GHz),可解释实验上观察到的“低能激发”与传统 AFM THz 自旋波尺度的差异。(基于第一性原理计算的RuO2应变诱发磁有序.docx)
总体上,你给出一个统一图像:整体应变决定“是否进入 AFM”;临界附近叠加局部畸变决定“是否可能出现 SDW 作为潜在基态”,从而为 RuO₂ 薄膜磁性实验结论分歧提供解释,并提出通过应力/畸变维度调控自旋输运的新思路。(基于第一性原理计算的RuO2应变诱发磁有序.docx)
如果你愿意,我也可以在这个总结基础上,帮你再做一版更“投稿友好”的一页式摘要(背景→方法→三条结果→一句话意义),或者把你文中“关键图(Fig.1–6)各自想表达的结论”用一句话逐张对齐,方便你检查叙事是否紧凑。
你注意到本文想是通过整体应变和局部畸变来影响RuO2磁性的,但本文目前的题目是Strain induced RuO2 magnetic order
from first-principles calculations,我想在题目中体现局部畸变的作用,但导师认为这可能太繁琐,我该怎么办(我打算投的期刊是prb)?
你注意到本文想是通过整体应变和局部畸变来影响RuO2磁性的,但本文目前的题目是Strain induced RuO2 magnetic order
from first-principles calculations,我想在题目中体现局部畸变的作用,但导师认为这可能太繁琐,我该怎么办(我打算投的期刊是prb)?
你导师担心“太繁琐”其实是对的:APS/PRB 的风格里,标题要自洽、简洁、信息量高,避免塞太多限定语/术语。(APS Journals)
但你这篇的关键新意之一,确实是“临界整体应变附近,局部畸变可以把 SDW 拉成更稳定的态”,而不仅是“应变诱导 AFM”。(基于第一性原理计算的RuO2应变诱发磁有序.docx)
你可以把“local distortion”作为一个并列驱动因素或一个限定机制写进标题,而不是把模型细节(harmonic/δ、n、Δ)写进去。PRB 读者看到 local distortions 就够了,细节放摘要/正文。
我建议你优先考虑这两种写法(都符合“短而信息密度高”):
这一类标题等价于告诉读者:你讲的是“整体应变 + 局部畸变”共同决定磁性相图/基态选择(与你文章主线一致)。(基于第一性原理计算的RuO2应变诱发磁有序.docx)
如果你更希望读者一眼看到“SDW”这条新意,就用这类。
在 PRB 里这句通常可省略,因为方法在摘要/计算细节里会写清楚;去掉它能显著减短标题、也更像 PRB 常见风格。(APS Journals)
如果你想保留“第一性原理”标签,建议改成更短的前缀式:
你可以把上面写法 A 当作折中:只多加 local distortions 两个词,但把你论文的核心叙事(整体应变决定是否磁有序、局部畸变决定临界附近的 SDW 竞争)准确放进标题。(基于第一性原理计算的RuO2应变诱发磁有序.docx)
同时你也可以让导师放心:模型细节不进标题,而是在摘要第一段第二句就点明“near critical strain, local distortions stabilize SDW (10–100 μeV; 7–10 nm)”——这样标题不繁琐,但卖点不会丢。
如果你愿意,我可以按你现在 PRB 的写法习惯(大小写、连字符、是否用 epitaxial/global)帮你把候选标题再精修到 3 个“最像 PRB 文章标题”的最终版本。
能不能把局部畸变换成局部应变
能不能把局部畸变换成局部应变
可以,而且从你这篇的建模方式看,“局部应变”其实比“局部畸变”更贴切——你是在固定整体应变 ε 的前提下,让每个单胞的晶格常数 a 在空间上起伏(Δh / Δδ 控制幅值、n 控制周期),这本质上就是“非均匀/局域应变场(local strain inhomogeneity / fluctuations)”的一个模型化表达。(基于第一性原理计算的RuO2应变诱发磁有序.docx)
但我建议你在英文标题里不要直接写成 bare 的 local strain(太短时容易让人误解成“又施加了一种新的宏观应变分量”),更 PRB 友好的写法通常是:
这样既简洁,又能准确对应你“整体应变 + 局部起伏”这条主线。(基于第一性原理计算的RuO2应变诱发磁有序.docx)
A. 最像你原题、改动最小(导师一般最容易接受)
B. 更推荐:把“局部应变”的物理含义说清(仍然不长)
2) Epitaxial strain and local strain inhomogeneity induce magnetic order in RuO₂
3) Magnetic order in RuO₂ driven by epitaxial strain and local strain fluctuations
C. 如果你希望标题直接“卖”你最亮的结果(临界附近 + 局部效应 → SDW)
4) Local strain fluctuations near critical epitaxial strain stabilize a spin-density wave in RuO₂
(这句和你文中“临界附近局部畸变可诱导 SDW 成为更低能态”的结论是对齐的。(基于第一性原理计算的RuO2应变诱发磁有序.docx))
如果你把标题写“local strain …”,我建议你在模型首次出现的位置用一句话把术语钉死,比如:
“We model local strain fluctuations (i.e., local lattice-constant modulations) at fixed global strain ε …”
这和你现在对“整体应变 ε + 局部起伏 Δ”的定义完全一致,也能避免“你到底是畸变还是应变?”这种语义争论。(基于第一性原理计算的RuO2应变诱发磁有序.docx)
如果你告诉我你更想突出哪条主结果(“应变诱导 AFM”还是“临界区 SDW”),我可以把上面 4 个再精修成最像 PRB 标题的 2–3 个最终版本(大小写、连字符、是否用 epitaxial/global 等都一起统一)。
在本文中能不能把局部畸变换成局部应变?
好的,这个问题我等元旦后再跟导师讨论,现在帮我翻译文章吧,先翻译如下部分,如果有一定的修改记得跟我说:
RuO2作为一种潜在的交错磁体,其有望成为可精确调控自旋输运的新一代自旋电子学材料,但其磁有序有无近期受到广泛质疑。本文基于第一性原理计算,以TiO2基底外延RuO2薄膜为例,系统研究了在外延应变作用下RuO₂薄膜的磁有序行为。当整体应变足够强时,RuO₂薄膜可由顺磁态转变为反铁磁(AFM)态。在磁应变临界状态附近引入足够强度的局部畸变(非均匀应变)时,自旋密度波态(SDW)可成为比AFM更稳定的基态,其能量差在10–100 μeV量级,特征波长约为7–10 nm。外延应变诱导的磁有序机制可为理解RuO₂薄膜交错磁性的实验争议提供了统一的微观图像,揭示了整体应变和局部畸变在调控RuO₂磁性及自旋输运中的关键作用。
可以,但我建议不要直接一刀切把所有“局部畸变”都替换成“局部应变”,而是用更精确、也更不容易被审稿人挑语义的表述:
原因是:你模型里“局部畸变”本质上是在固定整体应变 ε 的前提下,让局部晶格常数 发生空间起伏(由 Δ 和 n 控制),这在物理上确实等价于“非均匀局部应变场”。但如果只写“局部应变(local strain)”,有些读者会误以为你又施加了一个额外的宏观应变分量。你文中其实也已经把它解释成“不改变整体应变 ε 的局部畸变”这一点了。(基于第一性原理计算的RuO2应变诱发磁有序.docx)
最稳妥的写法:首次出现用
局部畸变(非均匀/局部应变涨落)
后文为了简洁,可以统一简称“局部应变不均匀”或“局部应变涨落”。
English (translated and lightly polished):
RuO, proposed as a potential altermagnet, is expected to serve as a new-generation spintronic material enabling precise control of spin transport; however, the very existence of magnetic order in RuO has recently been widely questioned. Here, based on first-principles calculations and taking epitaxial RuO films grown on TiO substrates as an example, we systematically investigate the magnetic ordering of RuO thin films under epitaxial strain. We find that a sufficiently large global strain can drive RuO films from a nonmagnetic state to an antiferromagnetic (AFM) state. Near the strain-driven magnetic critical point, sufficiently strong local strain inhomogeneity (i.e., inhomogeneous strain modeled as local lattice-constant modulations) can stabilize a spin-density-wave (SDW) state over the AFM state, with an energy difference on the order of 10–100 eV and a characteristic wavelength of about 7–10 nm. The strain-induced magnetic-ordering mechanism provides a unified microscopic picture for the experimental controversy regarding altermagnetism in RuO thin films, and highlights the key roles of global strain and local strain inhomogeneity in tuning RuO magnetism and spin transport. (基于第一性原理计算的RuO2应变诱发磁有序.docx)
如果你接下来把文章其它部分也发我(或直接按章节),我可以继续按同样标准翻译,并顺手把“局部畸变/局部应变”相关术语在全文做一致性处理。
翻译如下部分:
近年来,交错磁性(altermagnetism)的研究迅速升温。交错磁体在实空间不产生净磁矩,但由于晶体对称性较低,可在动量空间诱导明显的自旋劈裂,从而兼具铁磁与反铁磁体系的部分特征,因此受到了自旋电子学方向这类关注微观调控材料磁性方向学科的重视。与此同时,交错磁体的高频响应、无杂散场优势以及由晶格各向异性驱动的输运特性,也为磁性材料的拓展了新的应用领域。相关研究已报道了交错磁体存在反常霍尔效应[41–44]、THz 发射以及自旋–电荷互相转换[45–47]等现象,这为新一代自旋电子学器件的开发与 MRAM、GMR、TMR 等技术的改良提供了新思路[48–50]。在此背景下,相关工作对RuO2较大的自旋劈裂能与较高 Néel 温度的预测[1],以及围绕 RuO2 反铁磁性的理论铺垫[2–4]和关于RuO2的 交错磁体自旋劈裂现象(ASSE) 研究实验的快速推进[5],共同推动 RuO2 成为该方向的研究热点。
下面是你给的这段 INTRODUCTION 中文的英文翻译(按 PRB 常见写法做了轻微润色,同时尽量保持你原意与信息结构不变)。(基于第一性原理计算的RuO2应变诱发磁有序.docx)
In recent years, research on altermagnetism has rapidly gained momentum. Altermagnets exhibit no net magnetization in real space; however, owing to their reduced crystal symmetry, they can host a pronounced spin splitting in momentum space, thereby combining key features of both ferromagnets and antiferromagnets. As a result, altermagnets have attracted considerable attention in spintronics, where microscopic control of magnetic order is central. Meanwhile, the high-frequency response, stray-field-free advantage, and lattice-anisotropy–driven transport properties of altermagnets have opened new opportunities for expanding the application landscape of magnetic materials. A number of phenomena have been reported in altermagnets, including the anomalous Hall effect [41–44], terahertz (THz) emission, and spin–charge interconversion [45–47], which provide new ideas for developing next-generation spintronic devices and for improving technologies such as MRAM, GMR, and TMR [48–50]. In this context, the prediction of a large spin-splitting energy and a high Néel temperature in RuO [1], together with theoretical groundwork on antiferromagnetism in RuO [2–4] and rapid experimental progress on the altermagnetic spin-splitting effect (ASSE) in RuO [5], have jointly propelled RuO to become a research hotspot in this field.
如果你希望更“硬核/更物理论文腔”(比如强调 spin-symmetry breaking、momentum-dependent spin splitting without SOC 等),我也可以给你一个更偏理论物理表达的版本。
翻译下一段:
尽管 RuO2 被寄予推动自旋电子学体系发展的厚望,其磁有序性质却长期存在显著争议。早期研究认为 RuO2 具有较弱的反铁磁序,并利用共振 X 射线散射讨论了块体与薄膜中的反铁磁信号,给出薄膜磁相关长度约为 5 nm[2,6]。但另一方面,也有相当多实验支持 RuO2 更接近 Pauli 顺磁金属:μSR 结果指出体系内部缺乏可观测的动态磁场与磁有序[7,8];部分 ARPES 结果亦显示能带结构更符合非磁有序特征[9,10];量子振荡(QO)实验对反铁磁序的存在提出了更强的挑战[11]。与此同时,亦有文献强调 RuO2 的磁结构本身极为脆弱,例如费米面附近费米速度偏小,使其磁性对外界扰动高度敏感[13,14]。考虑到 RuO2具有一定库仑排斥且带宽较宽,属于强共价的中等耦合体系,其复杂磁效应并不意外。因此,有观点提出氧空位、外延应变以及表面/界面对称性破缺等非本征因素可能是薄膜中磁性的重要成因[13,15,16];而这些因素在薄膜体系中往往较为显著。事实上,多数报告反铁磁磁有序的实验主要基于 RuO2 薄膜,论证RuO2交错磁性的相关自旋输运实验亦多发生在薄膜体系[17,18]。更有研究指出通过多电子/少电子杂质调控电子数以改变 Néel 矢量方向并增强反铁磁性表征[19,20],这进一步凸显非本征效应对 RuO2 的关键影响。
Despite the great promise of RuO for advancing spintronic platforms, its magnetic ordering has long remained highly controversial. Early studies suggested that RuO hosts relatively weak antiferromagnetic order, and resonant x-ray scattering was used to identify antiferromagnetic signatures in both bulk and thin-film samples, yielding a magnetic correlation length of about 5 nm in films [2,6]. On the other hand, a substantial body of experiments supports RuO being closer to a Pauli-paramagnetic metal: SR results indicate the absence of observable internal dynamic magnetic fields and long-range magnetic order [7,8]; some ARPES measurements also show a band structure more consistent with a nonmagnetic state [9,10]; and quantum oscillation (QO) experiments pose an even stronger challenge to the existence of antiferromagnetic order [11]. Meanwhile, several works have emphasized that the magnetic state of RuO is intrinsically fragile—for instance, the relatively small Fermi velocity near the Fermi surface makes its magnetism highly susceptible to external perturbations [13,14]. Given that RuO features a finite Coulomb repulsion together with a relatively wide bandwidth, placing it in a strongly covalent, intermediate-coupling regime, such complex magnetic behavior is not unexpected. Accordingly, it has been proposed that extrinsic factors—such as oxygen vacancies, epitaxial strain, and symmetry breaking at surfaces or interfaces—may play an important role in generating magnetism in thin films [13,15,16], where these effects are often pronounced. Indeed, most experiments reporting antiferromagnetic order are based on RuO thin films, and spin-transport experiments supporting altermagnetism in RuO have also predominantly been carried out in film geometries [17,18]. Moreover, it has been reported that tuning the electron count via electron- or hole-type impurities can modify the Néel-vector orientation and enhance the experimental signatures of antiferromagnetism [19,20], further highlighting the critical impact of extrinsic effects on RuO. (基于第一性原理计算的RuO2应变诱发磁有序.docx)
翻译下一段:
基于上述认识,越来越多工作开始聚焦应变对 RuO2 的影响,以及应变诱导的晶格畸变如何改变电输运与磁有序来实现器件层面的调制[21–23]。这一研究路线的必要性尤为突出:RuO2 脆弱的费米面结构使其磁结构对晶格常数异常敏感。已有研究表明,掺杂可通过改变晶格而显著影响其磁性[53];也有文章指出在特定应变条件下 RuO2 薄膜可出现超导现象[54–56],甚至在化学催化研究中,RuO2 的晶格参数对活化能的巨大影响也被反复强调[62]。受此启发,实验上亦开始探索通过薄膜厚度实现基底–薄膜耦合强度的精细调控[51,58];而应用导向的研究则进一步提出利用外延应变对 RuO2 的自旋输运进行细节控制,从而扩展自旋电子学器件的调控维度[52,59]。总体而言,应变对 RuO2 磁性的影响已在理论、实验与应用三个层面受到系统关注。
下面是这段的英文翻译(按 PRB 常见的 introduction 语气做了轻微润色,但不改你的逻辑与信息点): (基于第一性原理计算的RuO2应变诱发磁有序.docx)
Building on these considerations, a growing number of studies have begun to focus on how strain influences RuO, and how strain-induced lattice distortions modify charge transport and magnetic ordering to enable device-level tuning [21–23]. This research direction is particularly well motivated because the fragile Fermi-surface structure of RuO renders its magnetic state unusually sensitive to the lattice constants. Previous work has shown that doping can markedly affect its magnetism by altering the lattice [53]. It has also been reported that RuO thin films may exhibit superconductivity under specific strain conditions [54–56]. Even in studies of chemical catalysis, the strong impact of RuO lattice parameters on the activation energy has been repeatedly emphasized [62]. Motivated by these insights, experiments have started to explore fine control of substrate–film coupling via the film thickness [51,58]. Application-oriented studies have further proposed using epitaxial strain to control the details of spin transport in RuO, thereby expanding the tuning “knobs” for spintronic devices [52,59]. Overall, the influence of strain on RuO magnetism has attracted systematic attention at the theoretical, experimental, and device/application levels.
如果你希望全文统一用 electrical transport(而不是 charge transport),我也可以把后续段落都按同一术语风格统一。
翻译下一段:
本文以常见的 TiO2基底外延RuO2 薄膜体系为例,将基底对 RuO2 的影响在近似为对 RuO2 晶格的整体应变,并据此系统讨论晶格局部畸变对 RuO2 磁性的作用。我们指出,即便暂不触及目前仍具争议的 DFT+U 处理,当晶格畸变达到一定程度时,RuO2 亦可从顺磁态转变为反铁磁态。进一步地,在局部畸变(应变非均匀性)较强时,自旋密度波态(SDW)可能相较于 AFM 更为稳定,其与 AFM 的能量差可达到10~100μeV量级。上述结果为理解 RuO2 薄膜体系中丰富的非本征磁有序提供了一个一致的微观图景,并有助于阐明应变对 RuO2 薄膜磁性及自旋输运行为的影响机制。
Using the widely studied system of RuO thin films epitaxially grown on TiO substrates as an example, we approximate the substrate effect as a uniform (global) strain applied to the RuO lattice, and on this basis systematically investigate how local lattice distortions affect the magnetism of RuO. We show that, even without invoking the still-debated DFT+ treatment, RuO can undergo a transition from a paramagnetic state to an antiferromagnetic (AFM) state once the lattice distortion becomes sufficiently strong. Furthermore, when the local distortion (i.e., strain inhomogeneity) is sufficiently pronounced, a spin-density-wave (SDW) state may become energetically more favorable than the AFM state, with an energy difference reaching the 10–100 eV scale. These results provide a consistent microscopic picture for the rich extrinsic magnetic ordering observed in RuO thin films, and help clarify the mechanisms by which strain influences the magnetism and spin-transport behavior of RuO films. (基于第一性原理计算的RuO2应变诱发磁有序.docx)
翻译下一部分:
本文第一性原理计算基于 VASP 程序包,在密度泛函理论(DFT)框架下采用 PBE 交换-相关泛函,并使用相应的赝势描述价电子与离子实之间的相互作用[28–33]。所用赝势为 Ru_pv(4p64d65s1),O(2s22p4),Ti(3p63d34s1)。平面波截断能取 520 eV;占据数展宽参数设为0.15 eV,并采用ISMEAR=1的展宽方案。本文涉及的 RuO2与 TiO2晶格常数列于Table Ⅰ。
对于金红石结构单胞计算,布里渊区采用Γ-centered, 9913的KPOINTS;态密度(DOS)计算使用更密的181826的KPOINTS。对于沿a轴扩胞得到的超胞计算,k 点网格相应选为Γ-centered,1913的KPOINTS,其 DOS 计算使用21826网格。
在结构收敛方面,我们将单胞计算中离子弛豫收敛标准设为 Hellmann–Feynman 力小于0.001eV/Å。对于包含大量超胞的扫描计算,为在保证精度的同时控制计算量,我们对考虑在超胞计算中将 O 原子受力控制在0.05eV/Å 以内。该阈值的合理性讨论与更严格弛豫设置下的对比结果以及使用的方法放入补充材料S3[63]。
数据后处理与可视化主要使用 VESTA 与 VASPKIT[34,60]。正文结果以不含 SOC、且不引入 DFT+U的计算为主;SOC 与 DFT+U的对比与补充分析统一置于补充材料S1,S2[63]。
下面是你这段 COMPUTATIONAL METHOD 的英文翻译(按 PRB 常见写法做了轻微润色,但不改信息点)。(基于第一性原理计算的RuO2应变诱发磁有序.docx) (基于第一性原理计算的RuO2应变诱发磁有序.docx)
Our first-principles calculations were performed using the VASP package within the framework of density functional theory (DFT). The Perdew–Burke–Ernzerhof (PBE) exchange–correlation functional was adopted, and the interaction between valence electrons and ionic cores was described using the corresponding pseudopotentials (PAW datasets) [28–33]. The valence configurations were Ru_pv (4p4d5s), O (2s2p), and Ti (3p3d4s). The plane-wave energy cutoff was set to 520 eV. A smearing width of 0.15 eV was used with the ISMEAR = 1 scheme. The lattice constants of RuO and TiO employed in this work are listed in Table I. (基于第一性原理计算的RuO2应变诱发磁有序.docx)
For calculations of the rutile primitive cell, the Brillouin zone was sampled using a Γ-centered 9×9×13 -point mesh; a denser 18×18×26 mesh was used for density-of-states (DOS) calculations. For supercells expanded along the axis, a Γ-centered 1×9×13 -point mesh was adopted, and the DOS was computed using a 2×18×26 mesh. (基于第一性原理计算的RuO2应变诱发磁有序.docx)
For structural convergence, the ionic relaxation criterion for primitive-cell calculations was set to a Hellmann–Feynman force < 0.001 eV/Å. For extensive scans involving a large number of supercells, to control the computational cost while maintaining accuracy, we required the forces on O atoms to be within 0.05 eV/Å in supercell calculations. The rationale for this threshold, together with comparisons to more stringent relaxation settings and the methodology used, is provided in Supplementary Material S3 [63]. (基于第一性原理计算的RuO2应变诱发磁有序.docx)
Data post-processing and visualization were mainly carried out using VESTA and VASPKIT [34,60]. The main-text results are primarily based on calculations without spin–orbit coupling (SOC) and without DFT+; comparisons and additional analyses involving SOC and DFT+ are presented in Supplementary Materials S1 and S2 [63]. (基于第一性原理计算的RuO2应变诱发磁有序.docx)
如果你继续发下一段(比如整体应变 ε 的定义 + Fig.1 那段),我可以把公式前后的衔接也按 PRB 习惯一起顺一下。
翻译下一部分:
RuO2 是一种对应变敏感的材料。其费米面结构的脆弱性使得磁性状态容易受到晶格应变的显著调控。鉴于 TiO2 与 RuO2 晶格匹配良好,其相似的金红石结构易于外延RuO2单晶,制备得到较好的自旋输运平台。以 TiO2 基底诱导的晶格应变为切入点讨论 RuO2 的应变诱发磁有序是合理的。为定量表征 TiO2 基底对 RuO2 外延薄膜的整体应变效应,我们引入整体应变参量 ε,以描述为 RuO2 晶格由初始RuO2晶格应变到 TiO2 晶格的程度[Fig. 1(a)],当 ε → 0 时晶格回到 RuO2 本征值;当 ε → 1 时晶格趋近 TiO2。其晶格常数公式如下:
aε=εaTiO2+1−εaRuO2
cε=εcTiO2+1−εcRuO2
(Fig. 1b)展示了Ru原子磁矩随着ε的变化关系。随着 ε 增大,体系可由顺磁态转变为反铁磁态,表现为 Ru 原子磁矩由无到有并逐渐增大。通过更精细的扫描可得到RuO2单胞的相变临界点 εc=0.7404(No SOC)。在考虑 SOC 后,临界点为 εc=0.7340,因此上述结果不会因为考虑SOC而发生颠覆性的变化。与无 SOC 情况非常接近;差异主要体现在磁矩幅值上。当 ε>εc 时,考虑 SOC 的磁矩相较无 SOC 情况约降低 10%。因此正文以下主要讨论无 SOC 的结果,SOC 与无 SOC 的异同将在补充材料 S2中给出[63]。
RuO is a strain-sensitive material. The fragility of its Fermi-surface structure makes its magnetic state susceptible to strong modulation by lattice strain. Given the good lattice match between TiO and RuO, their similar rutile structures facilitate the epitaxial growth of single-crystalline RuO, providing a favorable platform for spin-transport studies. Therefore, it is natural to discuss strain-induced magnetic order in RuO by taking the substrate-induced lattice strain from TiO as a starting point. (基于第一性原理计算的RuO2应变诱发磁有序.docx)
To quantify the global strain effect of the TiO substrate on epitaxial RuO films, we introduce a global strain parameter , which describes the degree to which the RuO lattice is continuously deformed from its intrinsic lattice toward that of TiO [Fig. 1(a)]. When , the lattice returns to the intrinsic RuO values; when , it approaches the TiO lattice. (基于第一性原理计算的RuO2应变诱发磁有序.docx)
The lattice constants are defined as
Figure 1(b) shows the evolution of the Ru magnetic moment as a function of . With increasing , the system can undergo a transition from a paramagnetic (nonmagnetic) state to an antiferromagnetic (AFM) state, as evidenced by the Ru moment emerging from zero and gradually increasing. A finer scan yields a critical strain (without SOC) for the RuO primitive cell. Upon including SOC, the critical point shifts slightly to , indicating that the above conclusion is not qualitatively altered by SOC; the main difference lies in the moment magnitude. For , SOC reduces the moment by about 10% compared with the no-SOC results. Therefore, in the following we primarily discuss the results without SOC, while the similarities and differences between SOC and no-SOC cases are provided in Supplementary Material S2 [63]. (基于第一性原理计算的RuO2应变诱发磁有序.docx)
帮我翻译下一段:
基底诱导的晶格应变不仅发生在TiO2基底外延RuO2薄膜体系中,也常见于MgF2,Al2O3等基底[64]。考虑到 RuO2薄膜本身的磁不稳定性,常见的局部畸变如基底-薄膜异质结导致的非理想生长,薄膜外延缺陷、界面或局域松弛都有可能对RuO2磁有序产生一定影响(对于非外延情况可能会更加显著)。因此,仅通过整体应变ε来讨论体系的现实情况是不够的,因为整体应变只能突出其整体性质而非局部细节,而薄膜局部磁性细节可能会对体系产生较大影响。我们在考虑整体应变ε的同时需要引入局部畸变(非均匀应变)来讨论薄膜体系来研究其对磁有序的作用。具体做法如下:沿 (100) 方向对 RuO2 扩胞,构建 n×1×1 的超胞作为局部畸变(非均匀应变)模型,其中 n 为周期(以单胞数计)。在超胞中保持 b=aε、c=cε 不变,仅对每个单胞的晶格常数 a 施加按单胞序号 i(i=1,2,…, n_i=1,2,…,n)变化的规律性的畸变(非均匀应变)(Fig. 2a)。我们考虑两类局部畸变(非均匀应变)模型:简谐畸变(应变)和δ畸变(应变)。简谐畸变晶格畸变(非均匀应变)周期为 n,在超胞内产生呈周期性分布的畸变(非均匀应变),其中Δh 表示局部畸变(应变非均匀性)幅值。简谐函数的周期性保证了整体应变 ε 不被改变,更偏向理想化建模,便于理解连续起伏的一般效应。δ畸变(应变)是晶格畸变周期为 n,晶格常数a在超胞内呈δ函数式变化的畸变(非均匀应变):其中Δδ 表征超胞中最大拉伸单胞的相对幅值;其余单胞作相应补偿,保证整体应变 ε 不被改变。其可用于模拟薄膜中局域强畸变(非均匀应变)的情形。上述两类局部畸变(非均匀应变)晶格常数计算公式如下:
Substrate-induced lattice strain is not unique to RuO films epitaxially grown on TiO; it is also commonly encountered for other substrates such as MgF and AlO [64]. Considering the intrinsic magnetic instability of RuO thin films, typical local structural imperfections—including nonideal growth associated with the substrate–film heterojunction, epitaxial defects, interfacial effects, or local relaxations—may all affect the magnetic ordering of RuO (and the impact could be even more pronounced in nonepitaxial cases). Therefore, describing the realistic film system solely by a global strain parameter is insufficient: global strain captures the overall lattice deformation but misses local details, while local magnetic features in thin films may have a substantial influence on the system. Accordingly, in addition to the global strain , we introduce local distortions (strain inhomogeneity) to examine their roles in magnetic ordering.
Specifically, we construct supercells by expanding RuO along the (100) direction as a model for local distortions (inhomogeneous strain), where denotes the modulation period measured in the number of primitive cells. Within the supercell, we keep and fixed, and apply a prescribed modulation only to the lattice constant of each primitive cell as a function of the cell index [Fig. 2(a)]. We consider two types of local-distortion (inhomogeneous-strain) models: harmonic and -type distortions. For the harmonic case, the distortion has period and forms a spatially periodic modulation within the supercell, with denoting the amplitude of the local distortion (strain inhomogeneity). The periodicity of the harmonic function ensures that the global strain remains unchanged, making it an idealized model that is convenient for understanding generic effects of smooth fluctuations. For the -type case, the distortion also has period , but the lattice constant varies in a -function-like manner within the supercell: characterizes the relative amplitude of the most stretched cell, while the remaining cells provide compensating distortions so that the global strain is conserved. This model is useful for mimicking situations with strongly localized distortions (strain inhomogeneity) in thin films, closer to defect-like scenarios. The lattice-constant constructions for these two models are given by: (基于第一性原理计算的RuO2应变诱发磁有序.docx)
如果你把后面的“公式如下”那两条公式贴出来,我可以把这一段最后一句补全成带公式的最终版本,并把符号()在全文统一到最标准的 PRB 记号。
翻译下一部分:
考虑到局部畸变对各处的晶格常数影响不同,不同位置的Ru原子磁矩有所不同(Fig. 2a),根据初步计算,我们注意到在晶格拉伸中心处Ru原子的磁矩最大,晶格压缩中心处Ru原子的磁矩最小,这意味着拉伸有利于磁有序,压缩抑制磁有序,与单胞的计算结果相同。之后我们将用最大磁矩和最小磁矩来衡量这种调幅反铁磁的磁性。我们暂取 n=9,来系统考察 ε 与 Δ 对磁有序的影响。结果显示(Fig. 2b):在简谐畸变下,随着 Δh 从 8‰,14‰ 增至 20‰,而最大拉伸处出现磁有序(μmax大于0.01μB)所需应变由 0.7404 缓慢降低至约 0.72、0.69 与 0.66。而最大压缩处的磁矩在ε<εc时基本不产生较强的磁有序,但在ε> εc时也开始迅速增长。这意味着ε对体系整体磁性调制是主要的,不同局部畸变体系的整体应变仍需达到εc附近才能产生整体性的磁有序。只是这个过程由于局部畸变(应变非均匀性)的存在导致不同区域磁有序程度有所差别,具体表现为拉伸区域产生磁有序(μmax大于0.01μB)比压缩区域产生磁有序所需的应变要比压缩产生磁有序所需的应变要小,两者产生磁有序并不同步。这意味着存在局部畸变(非均匀应变)体系没有临界应变的说法,而是类似非晶熔化存在一个熔化温度区间一样,局部畸变(非均匀应变)体系存在一个应变磁有序区间。且这个应变磁有序区间受到εc的绝对调制。(部分原子磁矩分布数据放在补充材料S4[63]中)
当采用 δ-type畸变(应变)时,在 Δδ=8‰, 14‰, 20‰ 下最大拉伸处产生磁性的临界应变分别约为 0.73、0.70、0.67,最大压缩处的磁矩在ε<εc时几近为0,在ε> εc时开始迅速增长。与简谐畸变给出的结果非常接近(Fig. 2c)。尽管 δ 畸变(应变)在空间分布形式上已显著不同于简谐畸变(应变),但在相同幅值参数都表现出相近的最大拉伸处临界应变数值。且其整体磁性都表现出较强的整体应变依赖性,这表明:磁有序产生与否主要由整体畸变 ε 决定,而局部畸变(应变非均匀性)更多改变的是磁矩分布而非相变边界本身。
Because local distortions affect the lattice constants differently at different positions, the Ru magnetic moments become position dependent [Fig. 2(a)]. From preliminary calculations, we find that the Ru moment is maximized near the center of the tensile region and minimized near the center of the compressive region. This indicates that tensile strain favors magnetic ordering whereas compressive strain suppresses it, consistent with the primitive-cell results. In the following, we characterize this amplitude-modulated antiferromagnetism using the maximum and minimum local moments. We set and systematically examine the effects of and on magnetic ordering.
As shown in Fig. 2(b), for the harmonic distortion, increasing from 8‰ and 14‰ to 20‰ leads to a slow reduction of the global strain required for the tensile maximum to develop magnetic order (defined by ): the corresponding critical strain decreases from 0.7404 to approximately 0.72, 0.69, and 0.66. By contrast, the moment at the maximum compression remains nearly zero for , but starts to increase rapidly once . This implies that the overall magnetic response is governed primarily by the global strain : even in the presence of local distortions, the system generally needs to approach the vicinity of to establish global magnetic order. The role of local distortion (strain inhomogeneity) is mainly to produce spatially nonuniform onset—magnetism emerges in tensile regions at smaller than in compressive regions, i.e., the two do not develop magnetic order simultaneously. Consequently, for systems with local distortions it is more appropriate to speak of a strain window for the onset of magnetic order rather than a single sharply defined critical strain, and this window is ultimately constrained by (and anchored to) . Representative spatial distributions of the atomic moments are provided in Supplementary Material S4 [63]. (基于第一性原理计算的RuO2应变诱发磁有序.docx)
When the -type distortion is adopted, the critical strains for magnetism at the maximum tensile site are approximately 0.73, 0.70, and 0.67 for and , respectively, while the moment at the maximum compression site remains nearly zero for and rises rapidly for . These results are very close to those obtained for harmonic distortions [Fig. 2(c)]. Although the spatial profile of the -type distortion differs substantially from a harmonic fluctuation, it yields a similar critical strain at the tensile maximum for the same amplitude , and the overall magnetism remains strongly dependent on the global strain. This indicates that whether magnetic order emerges is determined mainly by the global strain , whereas local distortions (strain inhomogeneity) primarily modify the moment distribution rather than the phase boundary itself. (基于第一性原理计算的RuO2应变诱发磁有序.docx)
如果你希望全文把“局部畸变”统一替换为“局部应变不均匀/涨落”,我也可以在后续翻译里帮你把术语统一成 local strain inhomogeneity / fluctuations,并在第一次出现处加一句定义来彻底消除歧义。
你没有必要把我的括号里"非均匀应变"以及局部畸变这些忽略,因为这个稿还在改i,我目前还没想好用哪个,所以我用括号括起来方便之后选择。
你没有必要把我的括号里"非均匀应变"以及局部畸变这些忽略,因为这个稿还在改i,我目前还没想好用哪个,所以我用括号括起来方便之后选择。
明白了——接下来我会严格保留你括号里的备选/解释用语(例如“局部畸变(非均匀应变)”我就翻成 local distortion (inhomogeneous strain),不再替你提前“二选一”或省略括号内容),方便你之后统一替换风格。
下面把你介意的那两段我按“括号保留版”重新给一遍(含必要的英文润色,但不删括号信息):
Substrate-induced lattice strain occurs not only in epitaxial RuO films grown on TiO substrates, but is also common for substrates such as MgF and AlO [64]. Considering the magnetic instability of RuO thin films, common local distortions (inhomogeneous strain)—such as non-ideal growth caused by the substrate–film heterojunction, epitaxial defects in the film, interfacial effects, or local relaxations—may all influence the magnetic ordering of RuO (and the effect may be even more pronounced in non-epitaxial cases). Therefore, discussing realistic films solely in terms of the global strain is insufficient, because global strain highlights overall properties rather than local details, while local magnetic details in thin films may have a significant impact on the system. Hence, in addition to the global strain , we introduce local distortions (inhomogeneous strain) to investigate their roles in magnetic ordering.
Specifically, we build an supercell by expanding RuO along the (100) direction as a local distortion (inhomogeneous strain) model, where is the period (counted in the number of unit cells). In the supercell, we keep and fixed, and apply a prescribed, unit-cell-index-dependent distortion (inhomogeneous strain) only to the lattice constant for each unit cell indexed by [Fig. 2(a)]. We consider two types of local distortion (inhomogeneous strain) models: harmonic distortion (strain) and -type distortion (strain). For the harmonic distortion (strain), the lattice distortion (inhomogeneous strain) has a period and produces a spatially periodic modulation within the supercell, where denotes the amplitude of the local distortion (strain inhomogeneity). The periodicity of the harmonic function ensures that the global strain is not changed, making it a more idealized model that facilitates understanding the generic effects of continuous fluctuations. The -type distortion (strain) also has a period , but the lattice constant varies in a -function-like manner within the supercell: characterizes the relative amplitude of the most stretched unit cell in the supercell; the remaining unit cells provide compensating distortions so that the global strain remains unchanged. This model can be used to mimic situations with local strong distortion (inhomogeneous strain) in thin films. The lattice-constant construction formulas for these two types of local distortion (inhomogeneous strain) are given by: (基于第一性原理计算的RuO2应变诱发磁有序.docx)
注:你原文里 “i(i=1,2,…, n_i=1,2,…,n)” 这处我按语境在英文里写成了 ,否则英文会显得像符号写错(你若想保留原貌也可以,我可以照排版原样放进去)。(基于第一性原理计算的RuO2应变诱发磁有序.docx)
Considering that local distortions (inhomogeneous strain) affect the lattice constants differently at different positions, the Ru magnetic moments vary with position [Fig. 2(a)]. Based on preliminary calculations, we note that the Ru moment is largest at the center of the tensile region and smallest at the center of the compressive region, implying that tension favors magnetic ordering whereas compression suppresses magnetic ordering, consistent with the primitive-cell results. In the following, we use the maximum and minimum moments to characterize the magnetism of this amplitude-modulated antiferromagnet. We temporarily set to systematically examine the effects of and on magnetic ordering. (基于第一性原理计算的RuO2应变诱发磁有序.docx)
As shown in Fig. 2(b), under harmonic distortion (strain), as increases from 8‰ and 14‰ to 20‰, the global strain required for magnetic ordering at the maximum tensile position (defined by ) decreases slowly from 0.7404 to approximately 0.72, 0.69, and 0.66. Meanwhile, the moment at the maximum compressive position is nearly zero for , but begins to increase rapidly for . This indicates that plays the dominant role in controlling the overall magnetism: for systems with different local distortions (strain inhomogeneity), the global strain still needs to approach the vicinity of to establish overall magnetic ordering. The presence of local distortions (inhomogeneous strain) mainly causes different regions to develop magnetic order to different extents; specifically, the tensile region develops magnetic ordering (with ) at a smaller strain than the compressive region, i.e., the onset is not simultaneous. This suggests that for systems with local distortions (inhomogeneous strain), it is less appropriate to define a single critical strain; instead, similar to how amorphous melting occurs over a temperature interval rather than at a fixed melting point, there exists a strain window for magnetic ordering in systems with local distortions (inhomogeneous strain), and this window is ultimately modulated by . (Part of the atomic-moment distributions is provided in Supplementary Material S4 [63].) (基于第一性原理计算的RuO2应变诱发磁有序.docx)
When the -type distortion (strain) is adopted, for and , the critical strains for magnetism at the maximum tensile position are approximately 0.73, 0.70, and 0.67, respectively. The moment at the maximum compressive position remains nearly zero for and starts to increase rapidly for , in close agreement with the harmonic-distortion results [Fig. 2(c)]. Although the spatial distribution of the -type distortion (strain) differs substantially from that of the harmonic distortion (strain), it yields similar critical strains at the maximum tensile position for the same amplitude parameter. Moreover, the overall magnetism still exhibits a strong dependence on the global strain, indicating that the emergence of magnetic order is mainly determined by the global strain , while local distortions (strain inhomogeneity) primarily modify the moment distribution rather than the phase boundary itself. (基于第一性原理计算的RuO2应变诱发磁有序.docx)
如果你同意这个处理方式,我后面翻译会按一个固定规则走:中文里括号怎么放,我英文里就怎么保留(必要时只做语法顺句、不做术语替你取舍)。你把下一段继续贴我就行。
翻译下一部分:
基于上述整体畸变与局部畸变的系统计算,我们可以将主要结果概括为以下两点:
(1) RuO2 的磁结构对晶格畸变高度敏感;当整体应变达到一定程度 ,体系由顺磁态转变为反铁磁态。该结论在引入 SOC 后基本不变。
(2)在整体应变 ε 给定时, 局部畸变Δ 确实可以实现对磁性一定程度的调控;但体系的磁有序边界仍主要受 ε 控制。
尽管实验中难以实现与模型完全一致的整体应变,但外延薄膜体系中可实现的整体应变水平通常可达到本文参数的量级51,64。同时 ε 的具体取值取决于基底与膜厚等非本征因素,这也为当前 RuO2 薄膜磁性表征的多样性提供了一种自然解释路径。更重要的是,在顺磁-铁磁转变中,RuO2的能带并未发生突变(Fig. 3a)。从费米面来说,只有红色部分发生了较大的变化,蓝色部分只有轻微的变化(Fig. 3b)。这与既有的ARPES和QO的实验观测是相容的。因此,整体应变+局部畸变驱动的磁有序机制在一定程度上缓解了相关能带实验对RuO2磁有序的质疑。
下面是你这段的英文翻译(保持你原句结构与术语,且不删你文中的括号/符号表达): (基于第一性原理计算的RuO2应变诱发磁有序.docx)
Based on the systematic calculations of the above global distortion/strain and local distortion, our main results can be summarized as follows:
(1) The magnetic structure of RuO is highly sensitive to lattice distortion; when the global strain reaches a certain level, the system undergoes a transition from a paramagnetic state to an antiferromagnetic (AFM) state. This conclusion remains essentially unchanged after including SOC.
(2) For a given global strain , the local distortion can indeed modulate the magnetism to some extent; however, the phase boundary of magnetic ordering is still mainly controlled by .
Although it is difficult to realize a global strain identical to our model in experiments, the global strain level achievable in epitaxial thin films is typically on the order of the parameters considered here [51,64] (in this work, at the critical strain, changes by 21‰ and changes by 35‰ compared with the unstrained case; in the above references, the measured strain within a few nanometers from the interface is about 10‰–40‰). Meanwhile, the specific value of depends on extrinsic factors such as the substrate and film thickness, which provides a natural route to rationalize the diversity of current experimental characterizations of RuO thin-film magnetism.
More importantly, across the paramagnetic–ferromagnetic transition, the band structure of RuO does not exhibit an abrupt change [Fig. 3(a)]. In terms of the Fermi surface, only the red portion shows a pronounced modification, while the blue portion changes only slightly [Fig. 3(b)]. This is consistent with existing ARPES and QO measurements. Therefore, the magnetic-ordering mechanism driven by global strain + local distortions can, to some extent, alleviate concerns raised by band-structure experiments regarding magnetic ordering in RuO.
你这里写的是“顺磁-铁磁转变 / paramagnetic–ferromagnetic transition”,但在前文你的主线是“顺磁 → 反铁磁(AFM)”。从上下文逻辑看,这里很可能应为 paramagnetic–AFM transition(或更宽泛写 paramagnetic–magnetically ordered transition)。你可以回头对照 Fig. 3 对应的计算态到底是 FM 还是 AFM,再决定这里用哪个更准确。
翻译下一段:
在对RuO2的反铁磁序研究逐渐深入时,相关研究报道了RuO2-Py薄膜体系的反铁磁IC-SDW现象,并且深入讨论了RuO2自旋密度波与表层Py自旋波耦合导致的低能激发[17]。该低能激发与之前计算文章认为的RuO2自旋波谱不同[24],其GHz频段的低能激发与反铁磁体的THz激发有一定矛盾,揭示了RuO2丰富的反铁磁性质。相关实验也认为RuO2体系存在一定的应变-自旋涨落[57]。考虑到RuO2费米面不稳定,在靠近(pi,0,pi/2)处的能带较为平坦,费米速度较低,有产生费米面嵌套,形成自旋密度波的潜质[3],因此晶格畸变、温度效应、层间交换耦合(IEC)、以及强关联等非本征因素,都可能改变其磁不稳定性格局并影响潜在的 SDW 倾向[25–27,61]。这意味着,在特定外界条件下,RuO2 可能出现 SDW 态,并展现出区别于 AFM 的磁动力学响应。
事实上,晶格-SDW 耦合在部分材料的 SDW 形成中起决定作用[38],但这与DFT计算软件的底层逻辑相斥;在不考虑晶格-SDW 耦合时,AFM 通常在能量上更占优,从(Fig. 4b)计算可以看出,在不施加局部应变的情况下,SDW态比AFM态多出的能量(ΔE)随着磁有序的建立后迅速增加到meV量级。其物理原因在于:SDW 的波节处 Néel 矢量发生翻转并伴随磁矩局域削弱,会带来交换能惩罚和由于波节处Ru原子无磁性导致的额外能量代价,从而使 SDW 更难在自洽计算中成为低能解。不过,RuO2薄膜本身就存在整体应变和局部畸变,因此通过引入先验的周期性局部畸变来计算SDW态是合理的,这也与Cr 的 SDW 能量计算思路相似[37]。
下面是这段的英文翻译(我保留了你原文中的括号内容,如 (IC-SDW)、(IEC)、 等;行文做了少量 PRB 风格的顺句处理,但不替你“二选一”改术语): (基于第一性原理计算的RuO2应变诱发磁有序.docx)
As studies on antiferromagnetic order in RuO have progressed, related work has reported an antiferromagnetic incommensurate spin-density-wave (IC-SDW) state in RuO–Py thin-film systems, and discussed in depth the low-energy excitations arising from the coupling between the RuO spin-density wave and the spin waves in the surface Py layer [17]. This low-energy excitation differs from the RuO spin-wave spectrum predicted in earlier theoretical calculations [24]. In particular, the GHz-range low-energy mode appears to be at odds with the THz excitations typically expected for antiferromagnets, revealing the rich antiferromagnetic behavior of RuO. Related experiments have also suggested the presence of certain strain–spin fluctuations in RuO [57]. Considering the instability of the RuO Fermi surface—where the bands are relatively flat near , implying a low Fermi velocity and a tendency toward Fermi-surface nesting, hence the potential for forming a spin-density wave [3]—extrinsic factors such as lattice distortion, temperature effects, interlayer exchange coupling (IEC), and strong correlations may all reshape its magnetic-instability landscape and affect the tendency toward an SDW state [25–27,61]. This implies that, under specific external conditions, RuO may host an SDW state and exhibit magnetic dynamical responses distinct from those of the AFM state.
In fact, lattice–SDW coupling can play a decisive role in SDW formation in some materials [38], but this is in tension with the underlying logic of standard DFT total-energy calculations. Without explicitly accounting for lattice–SDW coupling, the AFM state is typically energetically favored. As can be seen from Fig. 4(b), in the absence of local strain, the excess energy of the SDW state relative to the AFM state, , increases rapidly to the meV scale once magnetic order is established. Physically, at the SDW nodes the Néel vector reverses and the local moment is strongly suppressed, which leads to an exchange-energy penalty and an additional energetic cost associated with the essentially nonmagnetic Ru atoms at the nodes, making it difficult for the SDW to emerge as a low-energy self-consistent solution. However, RuO thin films inherently involve global strain and local distortions, so it is reasonable to evaluate the SDW state by introducing an a priori periodic local distortion, which is also similar to the strategy used in total-energy calculations of the SDW in Cr [37].
如果你下一段里有“局部畸变(非均匀应变)”这种括号,我也会继续逐字保留括号来翻译。
翻译下一部分:
具体而言,我们先构建"整体应变 + 局部畸变"的模型来研究局部畸变对 RuO2 薄膜磁性的影响。考虑到局部畸变周期是SDW周期的一半,因此我们需要构建包含2个局部畸变周期的局部畸变超胞取计算SDW。以 n=8的正弦畸变模型为例(Fig. 4a),我们构建 16 × 1 × 1 的局部畸变模型,其局部畸变晶格常数计算公式如下:
Harmonic Distortion: ai = aε1−Δhcos(2i−1)θ2n, b = aε , c = cε
其中 i 为单胞序号,Δh 为局部畸变幅值参数。
我们定义ΔE=E_SDW-E_AFM,并选取不同 Δh 计算 ΔE,以评估 SDW 与 AFM 的相对稳定性(Fig. 4c)。当 ε>εc 时,ΔE 会随 Δh 的减小而迅速增大;要实现 ΔE<0,所需的局部畸变强度会随 ε 的增大而变得更苛刻,这一趋势与既有文献的结论一致[36]。当 ε<εc 时,体系整体磁有序更难建立,会使 SDW 的出现变得更困难。值得注意的是,当 ε≈εc 时,ΔE 随 Δh 呈现先增大后减小的非单调行为,并在 Δh≥12‰ 时 SDW 成为能量更低的态。考虑到RuO2外延薄膜中10‰量级的局部畸变较为常见,这表明 SDW 更可能在 ε 临界附近且伴随一定局部畸变时成为潜在的基态;当 ε 远离临界区间时,体系更倾向以顺磁态或 AFM 形式存在。总体上,局部畸变是 SDW 形成的重要诱因。
Specifically, we first construct a “global strain + local distortion” model to investigate how local distortions affect the magnetism of RuO thin films. Since the period of the local distortion is half of the SDW period, we need to build a locally distorted supercell containing two local-distortion periods in order to compute the SDW state. Taking the sinusoidal distortion model with as an example [Fig. 4(a)], we construct a locally distorted supercell. The lattice constants are defined as follows:
Here, is the unit-cell index, and is the amplitude parameter of the local distortion.
We define , and evaluate the relative stability of the SDW and AFM states by computing for different values [Fig. 4(c)]. When , increases rapidly as decreases; to achieve , the required local-distortion strength becomes more stringent with increasing , consistent with previous literature [36]. When , global magnetic order is more difficult to establish, making the emergence of the SDW state more challenging. Notably, when , exhibits a nonmonotonic dependence on , first increasing and then decreasing, and the SDW becomes the lower-energy state for . Considering that local distortions on the order of are fairly common in epitaxial RuO thin films, this suggests that the SDW state is more likely to become a potential ground state near the critical when accompanied by a certain degree of local distortion. When is far away from the critical regime, the system tends to remain in the paramagnetic state or adopt the AFM state. Overall, local distortion is an important driving factor for SDW formation.
翻译下一部分:
在此基础上,我们提出一个用于解释 SDW 可能成为基态的假设:随着局部畸变幅值 Δh 增强,SDW 态在波节本征地满足磁矩趋近于零,当压缩区域推动局域磁矩衰减时,SDW 反而能在波节处获得相对 AFM 的能量优势,从而使 ΔE更可能变为负值。由此可推导出一个可检验的结论:ΔE 的主要贡献应来自波节附近。考虑到 AFM 的非局域性,波节处的磁矩在 AFM 中不应立刻降为零,而是随 Δh 增强而缓慢下降;相反,在磁矩最大的区域,AFM 与 SDW 的局域磁矩应趋于一致,因此对 ΔE 的净贡献应较小。我们在 ε=εc条件下,对不同n情况进行了计算。通过Fig. 5a可以看出,当Δh ≥5‰时,ΔE随着Δh的增大而逐渐减小。同时我们可以比较 AFM 与 SDW 的 Ru 局域磁矩分布(Fig. 5b):两种磁结构的最大磁矩均随 Δh 增大而增大,并逐步趋同,这与拉伸区域磁矩趋同、对 ΔE 贡献较小的预言一致;而 AFM 的最小磁矩则随 Δh 增强而降低,反映出波节区域磁矩被持续压低但不为0。因此波节处为 AFM 与 SDW之间磁矩差异的主要来源,从而支撑"ΔE 主要由波节贡献"的判断。
我们也可得到第二个可检验推论:随着周期 n 增大,波节附近更易形成"几近无磁"的区域,SDW 相对 AFM 的能量优势应逐渐显现。为此我们在 ε=εc下计算不同 Δh 时 ΔE per cell (不同周期ΔE per cell易于讨论)及 AFM/SDW 磁矩随 n 的变化。结果显示:ΔE per cell随 n 先减小后增大(Fig. 5c),而AFM的波节处磁矩则随 n 增大迅速减小(Fig. 5d)。其物理图像可以理解为两种能量的竞争:当 n 从较小值增大时,波节附近的"几近无磁"区域扩大且波长变长,交换能也随也随之降低,使 SDW更容易作为基态产生;但当 n 进一步增大时,近无磁区域逐步超过AFM非局域性的限制,AFM 的波节磁矩也逼近零并趋近 SDW,从而导致 ΔE per cell 回升并可趋近于零,同时更长的周期也削弱了相邻磁有序区域之间的耦合,使两种态在磁矩幅值上更趋同。我们还观察到:Δh 越小,使 ΔE per cell 达到最小值所需的 n 越大,即畸变幅度越小,最稳定的 SDW 所需的波长越长。当Δh = 8‰, 14‰, 20‰时, ΔE per cell的最小值对应的n为8,9,11,考虑到上述Δh比较符合现实情况51。这意味着SDW的特征波长为16a22a,即710 nm。
On this basis, we propose a hypothesis to explain why the SDW state may become the ground state: as the local-distortion amplitude increases, the SDW state intrinsically satisfies that the magnetic moment approaches zero at the node region (波节), and when the compressive region further suppresses the local moment, the SDW can instead gain an energetic advantage over the AFM state around the nodes, making more likely to become negative. This leads to a testable conclusion: the dominant contribution to should come from the vicinity of the nodes (波节). Considering the nonlocal nature of AFM, the moment at the nodes in the AFM state should not immediately drop to zero, but should decrease gradually with increasing ; in contrast, in regions where the moment is maximal, the local moments of AFM and SDW should tend to become identical, and thus the net contribution to should be small.
We perform calculations for different at . As shown in Fig. 5(a), when , decreases gradually with increasing . We further compare the spatial distributions of Ru local moments in AFM and SDW [Fig. 5(b)]: the maximum moments in both magnetic structures increase with and gradually converge, consistent with the prediction that the tensile region moments converge and contribute little to . Meanwhile, the minimum moment of the AFM state decreases as increases, indicating that the moment in the node region is continuously suppressed but does not become zero. Therefore, the node region is the primary source of the moment difference between AFM and SDW, which supports the statement that “ is mainly contributed by the nodes (波节).”
We can also obtain a second testable implication: as the period increases, a “nearly nonmagnetic” region is more likely to form around the nodes (波节), and the energetic advantage of SDW relative to AFM should gradually emerge. To this end, at we calculate per cell (不同周期 per cell易于讨论) for different , as well as the evolution of AFM/SDW moments with . The results show that per cell first decreases and then increases with [Fig. 5(c)], while the AFM moment at the nodes decreases rapidly as increases [Fig. 5(d)]. This physical picture can be understood as a competition between two energy contributions: as increases from small values, the “nearly nonmagnetic” region near the nodes expands and the wavelength becomes longer, reducing the exchange-energy cost and making it easier for the SDW to become the ground state; however, when increases further, the nearly nonmagnetic region gradually exceeds the constraint imposed by the nonlocality of AFM, so that the AFM node moment also approaches zero and becomes closer to the SDW, causing per cell to increase again and potentially approach zero. Meanwhile, a longer period also weakens the coupling between neighboring magnetically ordered regions, making the moment magnitudes of the two states more similar.
We further observe that the smaller is, the larger is required to reach the minimum of per cell; i.e., the smaller the distortion amplitude, the longer the wavelength required for the most stable SDW. When , the minima of per cell occur at , respectively, and these values are reasonably consistent with experimental situations [51] (晶格常数涨落为10‰–30‰). This implies that the characteristic SDW wavelength is –, i.e., about 7–10 nm.
继续翻译下一部分,尽量把文章翻译的更偏prb:
我们分析了局部畸变模型下的态密度。以 ε=εc、n=8、Δh=24‰ 的简谐畸变为例,我们对比了不同位置Ru的态密度(Fig. 6a)。非波节区域的Ru 呈现出典型的自旋劈裂特征,进一步的计算指出态密度变化是连续的。
上述结果表明:在理想的周期性畸变模型中,SDW 是RuO2潜在的基态。但在现实情况中,真实薄膜中的局部畸变是无序起伏的。SDW 的空间结构也会随之改变。考虑到诱发SDW的局部畸变没有固定波长,因此我们可以构建一套可人为调节压缩区域位置的局部畸变模型,从而实现对 SDW 波节钉扎位置的调控(具体实现方式见补充材料S5[63])。计算结果显示:SDW 的波节会被最大压缩区域中心钉扎,并跟着最大压缩区域移动(Fig. 6b)。这一方面说明波节位置对局部畸变具有一定的响应;另一方面也表明 RuO2 中"SDW 波长"并非严格唯一常数,而是与局部畸变的特征长度(以及其空间分布)密切相关。
在现实中,由缺陷、界面台阶或局域松弛导致数个晶胞的局部畸变相当常见,其影响范围往往局限在 nm 的局域区域(插入引用文献)。为刻画这类"有限范围的强畸变",我们构建更具现实指向的 δ 局部畸变模型来模拟真是情况:其局部晶格畸变定义为:
δ type Distortion: ai=aε(1−Δδ),1≤i≤n0 or n+1≤i≤n+n0,aε1+n0 n−n0 Δδ,otherwise, b=aε, c=cε.
其中 n0 表示每个磁有序区域内被压缩的晶格数,而 n−n0 对应波节(node)附近的拉伸晶格数;上述构造保证在一个周期内 ⟨ai⟩=aε。更接近"局域强畸变+有限作用范围"的薄膜现实情况。
在计算中,我们取 n=10,ε = εc,构建 20×1×1的局部畸变超胞,并通过改变 n0 来调节压缩区域的"宽度",从而考察其对 SDW 与 AFM 相对稳定性的影响(Fig. 6c)。结果显示:当n0 = 1 时,ΔE随 Δδ增大而增大。当n0 = 2,3,4 时,ΔE随 Δδ增大而增大。且ΔE随Δδ变化的斜率也会随着n0的增大而减小。这说明:SDW需要一个相对较大的局部压缩区域,从而实现 SDW 态较于 AFM 态的稳定。这也反映出体系存在SDW与否和薄膜本身的局部压缩情况息息相关,与之前简谐畸变的结果对应。
下面是你这段的 PRB 风格英文翻译(并按你的要求保留你文中括号/占位提示,比如“(插入引用文献)”、以及“波节(node)”这种中英并列): (基于第一性原理计算的RuO2应变诱发磁有序.docx)
We further analyze the density of states (DOS) within the local-distortion model. Taking the harmonic distortion with , , and as an example, we compare the Ru-projected DOS at different positions [Fig. 6(a)]. Ru sites away from the node region (波节) exhibit a clear spin-splitting feature, and additional calculations indicate that the DOS evolution is continuous (smooth).
These results suggest that, within an idealized model with periodic distortions, the SDW state can emerge as a potential ground state of RuO. In realistic thin films, however, local distortions are disordered fluctuations, and the real-space profile of the SDW is expected to adjust accordingly. Since the local distortions that induce an SDW do not possess a unique wavelength, we construct a class of local-distortion models in which the position of the compressed region can be tuned, thereby enabling control over the pinning position of the SDW node (波节) (see Supplementary Material S5 [63] for details). Our calculations show that the SDW node is pinned to the center of the maximally compressed region and follows its motion [Fig. 6(b)]. This, on the one hand, demonstrates that the node position responds to local distortions; on the other hand, it implies that the “SDW wavelength” in RuO is not a strictly unique constant, but is closely tied to the characteristic length scale of the local distortion (and its spatial distribution).
In practice, local distortions extending over several unit cells—caused by defects, interfacial steps, or local relaxations—are rather common, and their influence is often confined to nm-scale regions (插入引用文献). To capture this type of “strong distortion with a finite spatial extent,” we introduce a more realistic -type local-distortion model to mimic the realistic situation. The local lattice distortion is defined as
Here denotes the number of compressed unit cells within each magnetically ordered region, while corresponds to the stretched unit cells near the node (波节, node). This construction ensures over one period, and is closer to the thin-film reality of “local strong distortion + finite interaction range.”
In the calculations, we set and , construct a locally distorted supercell, and tune the “width” of the compressed region by varying , in order to examine its impact on the relative stability of the SDW and AFM states [Fig. 6(c)]. The results show that for , increases with increasing ; for , also increases with , while the slope of decreases as becomes larger. This indicates that stabilizing the SDW relative to the AFM requires a sufficiently extended locally compressed region. It also implies that the existence (or absence) of an SDW is intimately linked to the local-compression conditions in the film, consistent with the trends obtained from the harmonic-distortion analysis.
如果你愿意,我也可以把这一段再进一步“PRB 化”到更硬的语气(比如把 potential ground state 换成更谨慎的 energetically competitive / metastable,以及把“连续”具体化为“no discontinuity across the SDW stabilization regime”等),但我先按你原意最大程度直译+顺句了。
继续下一部分:
综合以上结果,我们可以归纳出两点结论与一条推论:
(1)RuO2薄膜在临界应变附近的局部畸变可诱导 SDW 成为基态。此时 AFM 与 SDW 的能量竞争主要由波节附近的局域结构决定:波节处压缩增强会放大两态在该区域的差异,从而增强 SDW 相对 AFM 的能量优势。当波节附近的"压缩/无磁倾向"区域空间尺度逐步扩大时,ΔE可呈现先减小后回升的非单调行为。
(2) RuO2薄膜的SDW 的特征波长由畸变强度决定。当畸变特征长度约为 7–10 nm 时更有利于 SDW 稳定。若畸变特征长度更短,交换作用会显著提高 Néel 矢量翻转处(波节处)的能量代价,使 SDW 更难占优;若畸变特征长度更长,则 AFM 与 SDW 的能量差趋于减小并接近简并。
(3) 推论:RuO2薄膜体系的低能激发可能源于 AFM 与 SDW 的近简并竞争。 在接近临界应变且存在局部畸变时,AFM 与 SDW 的能量差仅为 10–100 μeV(~10 GHz),显著低于常规反铁磁体常见的 THz 量级自旋波激发能量。
CONCLUSIONS
本文讨论了外延RuO2薄膜体系的磁有序情况,研究了整体应变对RuO2薄膜体系的影响,并讨论了RuO2不同局部畸变产生的非本征磁有序。当基底对RuO2的整体应变足够强时,RuO2可由顺磁态转变为AFM态。在顺磁-AFM应变临界状态,当引入足够强度的局部畸变时,SDW会作为潜在的能量基态出现,SDW和AFM的能量差取决于局部畸变的特征长度和局部畸变的幅值。当局部畸变区域达到数个晶格常数,畸变幅度达到10‰量级时SDW即可作为基态存在。
考虑到外延RuO2 薄膜的磁性对整体应变与局域畸变均表现出显著的敏感性,本文揭示的应力诱发磁有序具有两方面意义:一方面,它有助于解释在不同样品形貌、外延条件与界面环境下对 外延RuO2 薄膜磁性所得结论的分歧,有助于讨论外延RuO2薄膜的反常磁现象;另一方面,本文也为 外延RuO2 薄膜的自旋输运调控提出了新的思路,为实现对外延RuO2薄膜自旋输运性质的调控,提供了一定的计算结果支撑,也为扩展自旋电子学器件的应力调控维度提供了新的途径与想法。
Combining the above results, we arrive at two conclusions and one inference:
(1) Local distortions near the critical global strain can stabilize an SDW ground state in RuO thin films. In this regime, the energetic competition between AFM and SDW is governed primarily by the local environment around the nodes (波节). Enhanced compression at the nodes amplifies the difference between the two states in this region, thereby increasing the energetic advantage of the SDW over the AFM. As the spatial extent of the “compressed / nonmagnetic tendency” region around the nodes increases, can exhibit a nonmonotonic behavior, i.e., it first decreases and then rises again.
(2) The characteristic SDW wavelength in RuO thin films is set by the distortion strength (characteristic length scale). A distortion length of approximately 7–10 nm is most favorable for stabilizing the SDW. If the distortion length is shorter, exchange interactions substantially increase the energy cost associated with the Néel-vector reversal (at the node), making the SDW less competitive. If the distortion length is longer, the energy difference between AFM and SDW decreases and the two states approach near degeneracy.
(3) Inference: the low-energy excitations observed in RuO thin films may originate from the near-degenerate competition between the AFM and SDW states. Close to the critical global strain and in the presence of local distortions, the AFM–SDW energy difference is only 10–100 eV (10 GHz), which is far below the THz-scale magnon energies typical of conventional antiferromagnets.
In this work, we investigate magnetic ordering in epitaxial RuO thin films by analyzing the effects of global strain and local distortions on non-intrinsic magnetic order. When the substrate-induced global strain on RuO is sufficiently strong, RuO undergoes a transition from a paramagnetic state to an AFM state. Near the paramagnetic–AFM critical strain, introducing sufficiently strong local distortions can render an SDW state energetically competitive and potentially stabilize it as the ground state. The energy difference between the SDW and AFM states depends on both the characteristic length scale and the amplitude of the local distortion. In particular, when the distorted region extends over several lattice constants and the distortion amplitude reaches the level, the SDW can become the ground state.
Given the pronounced sensitivity of epitaxial RuO thin-film magnetism to both global strain and local distortions, the strain-induced magnetic ordering mechanism revealed here has two implications. On the one hand, it provides a natural framework to rationalize the disparate experimental conclusions reported for epitaxial RuO films under different morphologies, epitaxial conditions, and interfacial environments, and may help interpret anomalous magnetic responses in these systems. On the other hand, it suggests additional routes for controlling spin transport in epitaxial RuO films, offering computational support for strain-based tuning and expanding the stress/strain control “knobs” available for spintronic devices. (基于第一性原理计算的RuO2应变诱发磁有序.docx)
继续翻译如下部分:
DFT+U在相当多的文章中已经研究过,但只有少量文章讨论了在应变情况下DFT+U的情况[51],且得到了与我们结果相近的结果。我们也讨论了DFT+U在不同ε下的情况。Fig. S1 (a)是通过计算不同U和ε的单胞计算结果,无论ε取值如何,当U<1eV,ε<εc时并不会导致无磁到有磁的转变,可以看出这两个物理量是相互独立影响磁性的。Fig. S2 (b)反映的是ε=εc,n=16,Δh=20‰时随着U变大ΔE,AFM和SDW的磁矩变化情况。可以看出,U在一定程度上抑制了SDW的产生。同时,AFM最小磁矩的变化也体现了由于U而提高的体系非局域性。
考虑SOC后,RuO2系统的磁矩出现了一定的缩小,但εc没有较大变化。这说明在体系磁性有无这点上SOC没有发挥较大作用。但SOC会降低体系的磁矩大小,例如在拉伸和压缩晶格常数a时,单胞晶格常数按如下公式计算:
a = aε1+∆ , b = aε , c = cε
其中Δ为a的拉伸比例,我们计算了ε = εc时不同Δ下的磁矩结果(Fig. 8a),可以看出考虑SOC时的磁矩较于没有SOC下降了近20%,这意味着在在局部畸变模型中的计算中其磁矩具有一定的区别,且意味着无磁性能量惩罚更低。在SDW的计算中我们知道,RuO2的SDW源于波节处的能量优势,但考虑SOC的RuO2局部畸变模型计算结果指出(Fig. 8b),SOC增强了体系的非局域性,这使得需要更大的n(波长)才能实现潜在的SDW基态,这表现在AFM计算中RuO2的最大磁矩比无SOC的情况下小了20%,但RuO2的波节处磁矩反而比无SOC的情况下大了50%。这会导致在考虑SOC情况下RuO2系统的SDW出现条件变得更加苛刻且难以计算。但经计算它并不会颠覆我们之前对RuO2中AFM和SDW的讨论。我们采用δ畸变进行计算,其中n0=n/2 (n0为半整数时只改变多出的一半晶格)。Fig. 8c计算了不同Δδ下ΔE随n的变化,能明显看出尽管Δδ较大,但其最小值尚未出现,这进一步验证了考虑SOC后体系非局域性增强的推测。Fig. 8d计算了不同n下ΔE随Δδ的变化,其结果与之前不考虑SOC类似,这意味着上述讨论放在SOC计算中仍然正确。虽然SOC让SDW成为基态所需临界局部畸变增强,这本质上是因为体系磁非局域性增加,使得采用的局部畸变模型的波长不够长,难以产生SDW。这也意味着不考虑SOC情况下计算得到的SDW波长偏小。
DFT+. DFT+ has been examined in a large number of studies, but only a few works have discussed DFT+ under strain [51], reaching conclusions close to ours. We also investigated the DFT+ behavior as a function of . Figure S1(a) summarizes primitive-cell calculations for different and : regardless of the value of , when eV and , the system does not undergo a transition from a nonmagnetic state to a magnetic state, indicating that these two parameters influence the magnetism in an essentially independent manner. (基于第一性原理计算的RuO2应变诱发磁有序.docx)
Figure S2(b) shows, for , , and , the evolution of as well as the magnetic moments of the AFM and SDW states as increases. One finds that suppresses the emergence of the SDW to some extent. Meanwhile, the variation of the minimum AFM moment also suggests an enhanced nonlocal character of the magnetism induced by . (基于第一性原理计算的RuO2应变诱发磁有序.docx)
SOC. Upon including spin–orbit coupling (SOC), the magnetic moment in RuO is reduced, while changes only weakly. This indicates that SOC does not play a major role in determining whether magnetism appears, but it does reduce the magnitude of the moment. For example, when the lattice constant is tensile/compressive distorted according to
where is the tensile ratio of , our calculations at show that the moment obtained with SOC is reduced by nearly 20% compared with the no-SOC case [Fig. 8(a)]. This implies that, in the local-distortion model, the moments differ quantitatively once SOC is included, and the energetic penalty associated with nonmagnetic regions becomes lower. (基于第一性原理计算的RuO2应变诱发磁有序.docx)
In SDW calculations, the SDW in RuO originates from an energetic advantage at the node region (波节). However, SOC calculations for the locally distorted RuO model indicate that SOC enhances the nonlocality of the magnetic response [Fig. 8(b)], such that a larger (wavelength) is required to realize a potential SDW ground state. Concretely, in AFM calculations the maximum moment is about 20% smaller than in the no-SOC case, whereas the node-region moment becomes about 50% larger. As a result, the SDW stabilization conditions become more stringent and harder to access computationally when SOC is included, although SOC does not overturn our previous discussion of AFM versus SDW in RuO. (基于第一性原理计算的RuO2应变诱发磁有序.docx)
We further perform calculations using the -distortion, taking (when is a half-integer, only the extra half-lattice is modified). Figure 8(c) shows versus for different : even for relatively large , the minimum has not yet appeared, further supporting the conjecture that SOC enhances nonlocality. Figure 8(d) presents as a function of for different , showing trends similar to the no-SOC results, indicating that the above arguments remain valid with SOC included. (基于第一性原理计算的RuO2应变诱发磁有序.docx)
Although SOC increases the critical local distortion required for the SDW to become the ground state, this is essentially because the increased nonlocality makes the wavelength in the adopted local-distortion model not long enough to stabilize the SDW. This also suggests that the SDW wavelength obtained in the no-SOC calculations is underestimated (偏小). (基于第一性原理计算的RuO2应变诱发磁有序.docx)
小提醒(不改你原意,只是排版一致性): 你这段里写的是 Fig. S1(a)、Fig. S2(b),但你当前稿件里补充材料对应位置似乎标成 Fig. 7(a)、Fig. 5(b)(见你文档片段)。后续定稿时建议把补充材料图号统一一下,避免 PRB 审稿人读图时困惑。
继续翻译如下部分:
Hellman-Feynman force反映的是体系中各个原子所受的力,对于超胞体系来说,取EDIFFG=-0.05已经是个相对较小的值。考虑到文章中大量的超胞计算,为了减少离子步迭代,我们可以先将超胞中的每个单胞进行离子弛豫计算(EDIFFG=-0.001),然后将其放入超胞中。这会让大多数体系中O原子所受力从最大0.30eV/A降低到0.04eV/A以下,对于大多数超胞计算即可不需要进行离子弛豫。但考虑到RuO2磁性比较脆弱,因此有必要对更小的EDIFFG进行测试。我们选择计算EDIFFG=-0.005来讨论计算中离子弛豫对体系的影响。相关参数参数n=16,ε=εc,结果放在Graph. 2中。可以看出磁矩和ΔE基本相同。因此,即便对于磁结构脆弱的RuO2体系来说,采用EDIFFG=-0.05仍然是可行的。但如果在超胞计算计算ΔE时不采用上述先单胞弛豫的方法且对超胞计算不设EDIFFG,会导致产生1~5 μeV per cell左右的误差。这会让计算结果不可信。
我们列出了正弦畸变和δ畸变在n=9,ε=0.74,0.70,Δ=20‰,14‰,8‰的情况下AFM态的磁矩分布(Fig. S4a,b)。以及n=16,ε=0.74下不同Δh下的磁矩分布(Fig. S4c,d)。
我们考虑ε=εc,n=10,Δh=1.0%,考虑2011的局部畸变模型。其中n1为参数:
Harmonic distortion: ai=aϵ(1−Δhcos(2i−1)i2n1) i<n1aϵ(1−Δhcos(2i−1)θ2(n−n1))n1≤i<2n, b=aϵ , c=cϵ
可以看出,这个畸变可以通过调整n1实现压缩晶格区域的移动,n1即为压缩晶格区域的中心序数。
下面是你给出的这段内容的 PRB 风格英文翻译(我顺手把 Hellman 改为更规范的 Hellmann–Feynman 拼写;另外把一些“we can see”改成更学术的表述——这些算轻微润色,我会在末尾提醒你可能需要统一的地方)。
DFT+U has been investigated in a substantial number of previous studies; however, only a limited number of works have addressed the DFT+U behavior under strain conditions [51], and they reported results broadly consistent with ours. Here we further examine the role of DFT+U at different values of the global strain parameter ε. Figure S1(a) summarizes unit-cell calculations as functions of U and ε. For all ε considered, when and , no transition from a nonmagnetic to a magnetic state is obtained, indicating that U and ε affect the magnetism in largely independent ways. Figure S2(b) shows the U dependence of as well as the magnetic moments of the AFM and SDW states at , , and . These results suggest that increasing U tends to suppress the stabilization of the SDW state. Meanwhile, the evolution of the minimum AFM moment also indicates an enhancement of the nonlocal character of magnetism induced by U.
Upon including spin–orbit coupling (SOC), the magnetic moments of RuO are reduced, whereas the critical strain remains nearly unchanged. This implies that SOC does not play a decisive role in whether magnetism emerges, but it does reduce the magnitude of the ordered moments. For example, when the lattice constant is tensilely/compressively distorted according to
the moments at decrease by nearly 20% once SOC is included [Fig. 8(a)]. This difference implies that, within the local-distortion model, the magnetic-moment distribution is modified and the energetic penalty associated with locally nonmagnetic configurations is reduced.
In our SDW analysis, the SDW stabilization originates from an energetic advantage near the nodes. However, calculations with SOC for the locally distorted models indicate that SOC enhances the magnetic nonlocality [Fig. 8(b)], such that a larger period (i.e., a longer wavelength) is required to realize a putative SDW ground state. Concretely, in the AFM calculations the maximum Ru moment is reduced by ~20% relative to the no-SOC case, while the moment at the SDW node increases by ~50%. As a result, the conditions for stabilizing the SDW become more stringent and the computations become more demanding. Nevertheless, SOC does not overturn our qualitative conclusions regarding the competition between AFM and SDW in RuO.
We further performed SOC calculations using the δ-type distortion model with (for half-integer , only the extra half cell is modified accordingly). Figure 8(c) shows as a function of for different . Even for relatively large , the minimum of is not reached within our accessible range of , which provides additional support for the SOC-enhanced nonlocality picture. Figure 8(d) shows as a function of for different , exhibiting trends similar to the no-SOC results; therefore, the above discussion remains valid when SOC is included. Overall, SOC increases the local-distortion threshold required for the SDW to become the ground state, essentially because the enhanced nonlocality demands longer distortion wavelengths than those adopted in our practical supercell models. This also suggests that the SDW wavelengths obtained without SOC are likely underestimated.
The Hellmann–Feynman forces quantify the residual forces acting on atoms. For large supercells, using already corresponds to a relatively tight force criterion. Given the large number of supercell calculations in this work, to reduce the number of ionic steps we adopt the following strategy: each primitive cell is first fully relaxed (EDIFFG = −0.001) and then assembled into the supercell. This procedure reduces the maximum force on O atoms from ~0.30 eV/Å to below ~0.04 eV/Å in most cases, so that additional ionic relaxation is unnecessary for the majority of supercell calculations. Because the magnetism of RuO is relatively fragile, we further benchmarked a tighter setting with EDIFFG = −0.005 to assess the influence of ionic relaxation. The corresponding results at and are summarized in Graph 2, showing that both the magnetic moments and remain essentially unchanged. Therefore, even for magnetically fragile RuO, employing the looser criterion EDIFFG = −0.05 is still reliable within our numerical resolution.
By contrast, if one computes in the supercell without applying the above “pre-relaxed unit cell” procedure and without imposing a force threshold (EDIFFG) for the supercell relaxation, an error of about per cell may arise, rendering the results unreliable.
We provide representative distributions of the AFM local moments for both the sinusoidal distortion and the δ-type distortion at , , and [Fig. S4(a,b)]. We also present the moment distributions at and for different values [Fig. S4(c,d)].
We consider , , and , using a locally distorted supercell. We introduce a tunable parameter and define the harmonic distortion as
This construction enables controlled translation of the compressed-lattice region by tuning ; specifically, corresponds to the index of the center of the compressed region.
如果你把下一段也贴出来,我可以继续按同样风格翻译,并顺手帮你把补充材料段落的叙述逻辑调成 PRB 更常见的“现象—结论—解释”的结构。