1.School of Computational Science and Electronics, Hunan Institute of Engineering, Xiangtan 411104, China 2.College of Physics, Mechanical and Electrical Engineering, Jishou University, Jishou 416000, China 3.Hunan Provincial Key Laboratory of Vehicle Power and Transmission Systems, Hunan Institute of Engineering, Xiangtan 411104, China
Fund Project:Projects supported by the National Natural Science Foundation of China (Grant Nos. 51701071, 51871096) and the Natural Science Foundation of Hunan Province, China (Grant Nos. 2016JJ5028, 2018JJ3100)
Received Date:20 February 2021
Accepted Date:14 April 2021
Available Online:07 June 2021
Published Online:05 September 2021
Abstract:Bimetallic core-shell nanoparticles such as NiCu are of great interest not only due to their excellent stability, selectivity, and magnetic and catalytic properties, but also because they are tunable by changing the morphology, surface element distribution, and particle size of the nanoparticles. The surface segregation and structural features of NiCu bimetallic nanoparticles, the deposition growth and the surface diffusion of Cu adsorbed atoms on the Ni substrate surface are studied by using molecular dynamics and the Montero method combined with embedded atomic potential. The results show that the Cu atom has a strong tendency of surface segregation. With the increase of concentration of Cu atoms, Cu atoms preferentially occupy the vertex, edge, (100), and (111) facet of nanoparticles due to the difference in configuration energy between Cu atoms and surface Ni atoms with different coordination numbers after the exchange, and finally form perfect Ni-core/Cu-shell nanoparticles. When growth temperature T = 400 K, the Ni-core/Cu-shell structure formed is the most stable. By observing the NiCu core-shell structure’s growth sequence, it is found that a few Ni atoms are replaced by Cu atoms on the step edge of the Ni substrate. The diffusion energy barrier of Cu atoms adsorbed on a Ni substrate surface is calculated by using the nudged elastic band method. The results show that Cu atoms adsorbed need to overcome a large ES barrier for both exchange and diffusion, making it difficult to diffuse between the facets of Ni substrate surface in a temperature range of 200–800 K. The lowest energy barrier for the diffusion of Cu atoms between facets of Ni substrate surface is 0.43 eV, and the diffusion path is from (111) facet to (100) facet. In contrast to Ni substrate, Ni atoms deposited on Cu substrate can easily migrate from the (111) facet to the (100) facet with a diffusion energy barrier of only about 0.12 eV, and at the present simulated temperature, Ni adsorbed atoms are unable to migrate on the (100) facet, resulting in a growth configuration toward an octahedral shape with its eight apex angles almost occupied by Ni atoms. In this paper, a new idea and method are provided for the preliminary design of NiCu nano-catalysts from atoms. Keywords:NiCu bimetallic nanoparticles/ surface segregation/ diffusion/ growth/ core-shell structure
长期以来, 纳米材料的熔化都是理论和实验研究的主要课题[45,46]. 为研究原子数N = 586—4033具有TOC形貌的Ni, Cu纳米粒子的熔点, 在0—2000 K的加热过程中, 纳米粒子的温度增量每步为50 K. 考虑到熔点附近的温度波动, 温度增量减小为10 K[47]. 图2所示为Ni, Cu纳米粒子每原子的平均势能和林德曼指数随温度的变化关系. 其中, 基于原子间距波动的Lindeman 指数能够定量地描述一个系统的熔化机制, 又由于在Lindeman 指数中包含了均方根的概念, 因此它能够描述每原子的一个流动状态, 指数越高, 说明该原子运动越无序, 流动性越强. 对于一个具有N个原子的系统, 系统中每原子的Lindeman 指数定义为均方根键长波动, 其表达式为[4] 图 2 具有TOC结构的Ni和Cu纳米粒子的每原子势能和林德曼指数随温度的变化 Figure2. Relations of the potential energy per atom and Lindemann index as a function of temperature for the Ni and Cu nanoparticles with TOC structures.
图3所示为200 K下原子数为586的NiCu双金属纳米粒子中Cu的分布随Cu浓度的变化. 从图3可以看出, 随着Cu原子浓度的增加, 不仅Cu原子整体的表面呈现偏析行为, 而且Cu原子在纳米粒子的表面也展现出多位点的偏析竞争. 其中Cu原子的占位依次是顶点、边、(100) 面和 (111) 面, 然后是体内. 利用密度泛函理论 (density functional theory, DFT) 方法, Wang等[52]最近对13原子和55原子的Cu和Cu-Ni核壳纳米粒子的结构、稳定性、电荷转移、电子和磁性进行研究, 发现Ni核/Cu壳比Cu核/Ni壳纳米粒子拥有更低的能量优势. Quaino等[34]采用晶格蒙特卡洛和分子动力学模拟方法对CUB形貌的NiCu纳米粒子进行研究, 发现Cu原子展现出很强的表面偏析. 一般而言, 表面能小的金属原子倾向于占据纳米粒子的表面, 有利于系统能量的降低确保稳定结构形成, 偏析的NiCu纳米合金的原子构型是由表面能、晶格畸变和结合能之间的平衡决定的[20]. 图 3 在200 K, 586原子的Ni1–xCux双金属纳米粒子中Cu的分布随Cu浓度的变化 Figure3. Variation of the fractional composition of Cu at different surface sites as a function of overall Cu composition for 586 atom nanoparticles at T = 200 K.
其中EA是系统为自由表面时的总能量, S是体系自由表面的表面积, Ebulk表示在周期性条件下系统的总能量. TOC形貌的纯Ni, Cu纳米粒子的表面能随尺寸的关系如图4所示, 随着纳米粒子尺寸的增加, 其表面能逐渐减少并趋向一个极限值, 主要是由于小尺寸的纳米粒子具有更高比例的表面原子. 从图4可以看出, 在相同尺寸下, Cu纳米粒子比Ni纳米粒子拥有更低的表面能. 此外, 相比金属Cu, 金属Ni具有更低的内聚能, 二者的内聚能分别为–3.54和–4.45 eV/atom[42]. 因此, Cu纳米粒子相对Ni纳米粒子更趋向于纳米粒子表面. 图 4 Ni, Cu纳米粒子的表面能随原子尺寸的关系 Figure4. Variation of the surface energy of Ni and Cu nanoparticles as atomic number.
偏析 (通过目前的MC + MD模拟获得) 和随机 (无序固溶体) 的Ni1–xCux 纳米粒子形成热结果如图5所示, 为了减少误差, 通过1000个构型对形成热结果进行统计平均, 纳米粒子每个原子的平均形成热 $ \left(?H\right) $[55]计算如下式: 图 5 随机和偏析结构的Ni1–xCux双金属纳米粒子的形成热与Cu原子浓度的变化关系. 实线为偏析结构, 虚线为随机结构 Figure5. The formation energy of Ni1–xCux bimetallic nanoparticles with random and segregation structure varies with the concentration of Cu atom. The solid line is a segregation structure, and the dotted line is a random structure.
表1对于586原子的TOC纳米粒子, 内部Cu原子与表面Ni原子交换后构型能差值 (?E, 单位: eV/atom), 数字6, 7, 8, 9和12分别表示顶点、边缘、(100) 面、(111) 面和体内的位置 Table1.Configuration energy difference values (?E, unit: eV/ atom) for a bulk Cu atom exchanging with surface Ni atoms for the TOC nanoparticles of 586 atoms. The numbers 6, 7, 8, 9, and 12 are represent the sites at the vertex, edge, (100) facet, (111) facet, and bulk, respectively
图 6 在200 K, 586个原子TOC形貌的Ni1–xCux双金属纳米粒子结构. 橘红色和海蓝色的球分别表示Cu原子和Ni原子 Figure6. Configurations and cross-section snapshots of 586 atom TOC Ni1–xCux nanoparticles at T = 200 K. The orange-red and sea-blue balls represent Cu atoms and Ni atoms, respectively.
23.3.生长与扩散 -->
3.3.生长与扩散
以1289原子TOC形貌的Ni核为基底, 通过随机沉积Cu原子至Ni核表面来实现动态的生长模拟, 当沉积的Cu原子数等于800时, 停止沉积. 利用纳米颗粒表面原子数 ($N_{\rm{A}}^{{\rm{surf}}}$)和内部原子数 ($N_{\rm{B}}^{{\rm{bulk}}} $) 定量描述了双金属纳米颗粒结构的生长, 其中A和B分别表示基底原子和注入原子, $N_{\rm{A}}^{{\rm{surf}}} $和$N_{\rm{B}}^{{\rm{bulk}}} $的交叉点定义为核壳团簇的缺陷数 (Ndef), 根据Baletto的描述, Ndef越小, 形成的核壳结构越好[19]. 在温度T = 200—800 K时, $ N_{\rm{Ni}}^{{\rm{surf}}}$和$N_{\rm{Cu}}^{{\rm{bulk}}} $沉积Cu原子数 (Ndep)如图7所示. 在沉积的开始, 随着Ndep的增加, $N_{\rm{Ni}}^{{\rm{surf}}} $几乎是线性减少, $N_{\rm{Cu}}^{{\rm{bulk}}} $则几乎是一个接近于0的小值. 结果表明, 入射的Cu原子占据表面位点, 没有进入Ni基衬底的内部. 随着Ndep持续增加, $N_{\rm{Cu}}^{{\rm{bulk}}} $开始缓慢增加, 最终与$N_{\rm{Ni}}^{{\rm{surf}}} $交叉. 主要原因是后沉积的Cu原子覆盖了已有的Cu壳原子, 形成第二层; 或一部分Cu原子与Ni基底台阶边缘的原子交换, 进入Ni核内部成为体原子[11,12,57]. 从图7可以看出, Ndef处于较低的水平, 说明在整个温度范围内都能形成较好的Ni核/Cu壳结构. 这与之前MC + MD的结果一致, Ni核/Cu壳的纳米粒子更稳定. 当生长温度为400 K时, 对应的交叉点Ndef处于最低水平, 说明温度400 K最利于Ni核/Cu壳结构的形成. 图 7 在温度T = 200, 400, 600和800 K下, $N_{\rm{Ni}}^{{\rm{surf}}} $和$N_{\rm{Cu}}^{{\rm{bulk}}} $与沉积的Cu原子数的函数关系 Figure7. The growth of Cu atoms on the TOC Ni substrate with 1289 atoms at T = 200, 400, 600, and 800 K, the $N_{\rm{Ni}}^{{\rm{surf}}} $ and $N_{\rm{Cu}}^{{\rm{bulk}}} $ as a function of the deposited Cu atoms (Ndep).
为了更加清晰地阐明Ni核/Cu壳的动态生长过程, 给出了T = 400 K时Cu原子在TOC结构的Ni基底表面的生长序列(图8). 从图8(a)—(f)可以看出, 沉积的Cu原子数分别为100, 200, 300, 400, 500和600. 随着Ndep的增加, Cu原子逐渐覆盖Ni核基底, 当Ndep = 600时, 形成相对完整的核壳结构. 类似的还有Al-Mg, Cu-Co和Ni-Ag等[11,12,19]核壳结构. 在生长过程中, 基底台阶边缘上少量Ni原子被沉积的Cu原子所取代. 采用NEB方法系统地研究了Cu吸附原子在镍基表面的 (111) 面到 (100) 面和 (111) 面到 (111) 面之间的扩散势垒, 结果如图9(a)和图9(b)所示. 图 8T = 400 K, Al原子在Ni TOC1289基底上的生长序列 (a) Ndep = 100; (b) Ndep = 200; (c) Ndep = 300; (d) Ndep = 400; (e) Ndep = 500; (f) Ndep = 600. 橘红色和海蓝色的球分别表示Cu原子和Ni原子 Figure8. Growth sequence of NiCu clusters at T = 400 K. The snapshots of the growth simulation for various Ndep of 100, 200, 300, 400, 500, and 600. The orange-red and sea-blue ball show the Cu and the Ni atom, respectively.
图 9 Cu (Ni) 吸附原子在Ni TOC1289 (Cu TOC1289) 基底表面的扩散势垒 (a) Ni-基底 (111)→(100); (b) Ni基底 (111)→(111); (c) Cu基底 (111)→(100); (d) Cu基底 (111)→(111). 橘红色和海蓝色的球分别表示Cu原子和Ni原子 Figure9. The diffusion energy barrier of Cu adatom on the surface of the Ni TOC1289: (a) Ni-base (111)→(100) facet; (b) Ni-base (111)→(111) facet; (c) Cu-base (111)→(100) facet; (d) Cu-base (111)→(111) facet. The orange-red and sea-blue balls represent Cu atoms and Ni atoms, respectively.