Fund Project:Project supported by the National Natural Science Foundation of China (Grant No. 51775066)
Received Date:02 May 2020
Accepted Date:29 May 2020
Available Online:28 September 2020
Published Online:05 October 2020
Abstract:Graphene has a wide range of applications in the fields of electricity, chemistry, biomedicine, and lubrication. But the strong van der Waals interaction between graphene sheets makes it easy to aggregate in preparation process, difficult to produce and put into practical applcation on a large-scale. There are many methods to prevent the graphene sheets from aggregating, such as reducing the size of sheets, adjusting the interaction between solvent and graphene, and using dispersant. Another possible method is to turn the sheet graphene into a three-dimensional structure like the crumpled paper. Compared with sheet graphene, the crumpled graphene ball has excellent aggregation-resistant. The current research on crumpled graphene ball mainly focuses on the effect of the initial structure of graphene sheet on the structure stability of the crumpled ball, but rarely involves the effect of functional groups. In this paper, ReaxFF molecular dynamics is used to simulate the crumpling process of graphene oxide sheet. The effect of functional groups (hydroxyl, epoxy) on the crumpling behavior and the stability of the crumpled ball of graphene oxide are studied. Graphene sheet oxidized by hydroxyl exhibits a push-up crumpling behavior. Graphene sheet oxidized by epoxy exhibits a layer-to-layer fitted crumpling behavior. Different crumpling behavior will lead to the difference in final crumpled ball structure. By analyzing the relationship between the atomic level potential energy incremental distribution and the distribution of broken and formed C—C bonds, we find that the broken and formed C—C bonds mainly occur in areas with a large degree of deformation, and the epoxy group has a stronger weakening effect on the C—C bond connected to it than the hydroxyl group. The release process of graphene oxide crumpled ball is simulated to study its structural stability. The stability of graphene oxide crumpled ball depends on the number of the broken and formed C—C bonds, that is, the more the number of broken and formed C—C bonds, the more stable the structure is, and under the same oxidation rate, the stability of the crumpled ball structure increases with the proportion of epoxy groups increasing. This study shows that the stability of graphene oxide crumpled ball structure can be controlled by changing the relative proportion of functional groups. Keywords:graphene oxide/ crumpling behavior/ stability/ molecular dynamics simulation
图1是氧化石墨烯在径向压缩作用下褶皱过程示意图. 将片层放置在虚拟球形盒子中, 片层上原子受到指向球心的径向力, 力的大小为 图 1 氧化石墨烯片层在径向压缩作用下褶皱过程示意图 Figure1. Schematic diagram of the crumpling process of graphene oxide sheet under radial compression.
图2给出了不同含氧基团比例修饰石墨烯片层褶皱过程原子结构演变. 其中, 蓝色表示石墨烯片层上的碳原子, 红色表示含氧基团中的氧原子, 白色表示含氧基团以及石墨烯片层边缘上的氢原子. 片层四个角弯曲程度以及折痕深度受含氧基团的影响. 当σ = 0 (全羟基修饰)时, 褶皱开始后片层的四个角向上向下弯曲, 片层中部有轻微起伏, 并伴有折痕出现, 随着褶皱的进行片层由边缘向中部推进, 整个过程呈现出“推进式”的褶皱行为. 当σ = 1 (全环氧基修饰)时, 氧化石墨烯片层四个角弯曲的程度增大, 片层中部凹陷, 随着褶皱的进行被凹陷隔开的片层开始贴合, 整个过程呈现出片层与片层“贴合式”的褶皱行为. 当σ = 0.5时, 片层四个角的弯曲程度以及片层中部起伏程度(出现凹槽)都有所增加. 片层从σ = 0到σ = 1褶皱行为上的转变可能与石墨烯片层表面含氧基团间范德华力、氢键、静电的相互作用以及片层抵抗变形的能力有关[20]. 图 2 不同含氧基团比例修饰石墨烯片层的褶皱过程(? = 50%) Figure2. Atomistic configurations during the crumpling process of graphene oxide sheets with various ratios of oxygen functional groups (? = 50%).
片层在褶皱过程中表现出两种褶皱行为, “推进式”以及“贴合式”的褶皱行为, 不同的褶皱行为将导致褶皱最终结构的差异. 图3是从不同视角观察到的石墨烯片层在两种褶皱行为下的褶皱最终结构(? = 50%和σ = 0, ? = 50%和σ = 1). 当σ = 0 (全羟基修饰)时, 正视图下的褶皱结构主要以折痕与折痕的堆叠为主, 这里把折痕与折痕堆叠的位置称为脊, 侧视图下的褶皱结构无明显折痕与折痕的堆叠现象, 以局部片层的贴合为主. 当σ = 1(全环氧基修饰)时, 两种视角下折痕与片层的分布都较为均匀. 图 3 含氧基团类型对石墨烯片层褶皱过程最终结构的影响(? = 50%) Figure3. Final structures resulted from the crumpling process of graphene oxide sheets with σ = 0 and 1 (? = 50%).
23.2.褶皱过程的变形分布、C—C断键与成键 -->
3.2.褶皱过程的变形分布、C—C断键与成键
随着褶皱程度的加深, 氧化石墨烯片层会出现键的拉伸、旋转和断裂等弹性和塑性变形行为. 研究发现, 石墨烯片层加上不同的含氧基团后, 含氧基团对片层上键的影响程度是不一样的[27]. 为了揭示含氧基团对石墨烯片层变形行为的影响, 对褶皱过程原子级势能变化与石墨烯片层CC—C断键、成键进行了研究. 为了表征含氧基团与石墨烯片层C—C断键、成键以及变形之间的关系, 做出了褶皱过程断键、成键原子对以及原子势能增量的位置分布. 石墨烯片层C—C断键、成键判别标准分别为1.8与1.5 ?[31], 势能增量通过计算相应原子当前时刻与初始时刻的势能变化得到. 图4给出了褶皱过程势能增量分布, C—C断键、成键原子对位置分布. 其中, 黑色空心圆表示断键原子, 橘黄色空心圆表示成键原子. 势能增量用彩条表示, 彩条的颜色从蓝色到红色, 表示片层变形程度由低到高. 图4中蓝色和浅蓝色区域对应的变形程度低, 绿色区域对应的变形程度相对较高, 并且绿色区域的彩条呈线条式的分布, 对应褶皱结构上的折痕, 线条与线条相交的地方为脊. 当σ = 0(全羟基修饰)时, 断键、成键原子对分布在图中绿色区域. 当σ = 1 (全环氧基修饰)时, 在褶皱程度低时 (ρg = 0.45), 断键、成键原子对在蓝色、浅蓝色和绿色区域都有分布. 在褶皱程度高时(ρg = 0.34), 绿色区域断键、成键原子有所增多, 蓝色和浅蓝色的区域依然有断键、成键原子对存在. 图 4 褶皱过程中的势能增量分布、C—C断键和成键原子对位置分布(? = 15%) (a) σ = 0; (b) σ = 1 Figure4. Distributions of the potential energy increment and the distribution of broken and formed C—C bonds during the crumpling process of graphene oxide sheets with (a) σ = 0 and (b) σ = 1.
片层的褶皱变形主要集中在折痕与脊处, 其他地方的变形程度相对较小. 当σ = 0(全羟基修饰)时, 断键、成键原子主要分布在变形程度大的地方, 也就是褶皱结构上的折痕与脊处. 当σ = 1(全环氧基修饰)时, 断键、成键原子不再只是分布在变形程度大的地方, 片层上变形程度小的地方也有出现, 这些地方为环氧基所连碳原子间的C—C键. 通过σ = 0和1的对比可以发现, 相比于羟基, 环氧基会削弱与其相连的C–C键, 使与其相连的C—C键容易断开. 图5给出了不同氧化率石墨烯片层褶皱过程断键、成键数量变化, 其中实心表示断键原子的数量变化, 空心表示成键原子的数量变化. 通过对片层褶皱过程结构的观察(图2), 以ρg = 0.45为分界点, 将褶皱过程分为两个阶段, 图5中以黑色竖直虚线隔开. 虚线左边为第一阶段, 这个阶段片层变形程度低, 褶皱结构主要以局部片层和折痕为主. 虚线右边为第二阶段, 这个阶段片层变形程度高, 褶皱结构基本已成球形. 当σ = 0(全羟基修饰)时, 在第一阶段, 片层断键、成键数量较少; 在第二阶段, 也即片层褶皱为球形后, 片层断键、成键数量明显增多. 当σ = 1(全环氧基修饰)时, 片层在第一阶段也有较多的断键、成键, 随着褶皱的进行断键、成键数量不断增多, 在第二阶段尤为明显. 图 5 不同氧化率石墨烯片层褶皱过程的C—C断键、成键数量变化 (a) σ = 0; (b) σ = 1 Figure5. Variations in the number of broken and formed C—C bonds during the crumpling process of graphene oxide sheets with (a) σ = 0 and (b) σ = 1.
当σ = 0(全羟基修饰)时, 片层的断键、成键来源于褶皱变形, 褶皱过程第一阶段变形程度低, 断键、成键数量相对较少. 当片层褶皱为球形后, 变形的程度增大, 断键、成键数量增多. 当σ = 1(全环氧基修饰)时, 由于环氧基对C—C键的削弱作用, 使得环氧基所连的C—C键容易断开, 所以在褶皱过程的第一阶段也有较多的断键、成键, 随后由于变形程度增大断键、成键数量进一步增加. 图6给出了不同含氧基团比例石墨烯片层断键、成键数量变化. 同一含氧基团比例下, 成键数量低于断键数量. 褶皱过程第一阶段, 这时片层变形程度低, 但仍然有较多的断键、成键, 并且断键、成键的数量随着σ的增大而增多. 片层的断键、成键数量随着σ的增大而增多, 除了环氧基对C—C键的削弱作用外, 还有一个原因, 当氧原子连续线性分布在石墨烯六元环一侧时, 六元环会发生自发的解环行为, 这也会导致石墨烯片层断键、成键数量增加[32]. 图 6 不同含氧基团比例石墨烯片层褶皱过程的C—C断键、成键数量变化(? = 50%) Figure6. Variations in the number of broken and formed C—C bonds during the crumpling process of graphene oxide sheets with ? = 50%.
23.3.褶皱球形结构的稳定性 -->
3.3.褶皱球形结构的稳定性
将片层褶皱之后总原子数密度(~200 nm–3)远大于块体石墨的原子数密度(113.75 nm–3)时的褶皱结构无约束释放. 待释放稳定后, 取稳定后的ρg值进行分析, ρg值越小, 说明结构越容易稳定. 图7(a)给出了褶皱球结构释放时刻碳原子数密度和释放稳定后ρg值随氧化率的变化. 图7(a)中, 右边的纵坐标表示褶皱球结构释放稳定后的ρg值, 左边的纵坐标表示褶皱球结构释放时刻的碳原子数密度, 碳原子数密度越大, 褶皱程度越大. 当褶皱球结构总原子数密度为~200 nm–3时, 其对应的碳原子数密度随着?的增大而减小, 释放稳定的ρg值随褶皱球结构释放时刻碳原子数密度的减小而增大. 由图7(a)可以看到, 片层褶皱程度随着?的变化而变化, ?越小, 片层压缩到同一总原子数密度时的碳原子数密度越大, 对应的褶皱程度越大, 释放后结构越容易稳定. 图 7 (a) 释放开始时刻的碳原子数密度与释放稳定后的ρg值; (b) 释放开始时刻的C—C断键、成键数量 (σ = 0) Figure7. (a) Number density of C atoms at the beginning of release process and the ρg value at the end of release process; (b) Number of broken and formed C—C bonds at the beginning of release process for the graphene oxide sheets with σ = 0.
片层褶皱程度越大, 褶皱球结构越容易稳定. 而褶皱程度大的结构变形大, 变形主要表现在断键、成键上. 为了进一步解释片层褶皱程度与褶皱球结构稳定性的关系, 计算了褶皱球结构释放时刻的C—C断键、成键数量. 图7(b)给出了褶皱球结构释放时刻C—C断键、成键数量随氧化率的变化, 褶皱球结构释放时刻的C—C断键、成键数量随着?的升高而降低, 这与片层褶皱程度随着?的变化关系一致. 这说明片层上C—C断键、成键数量会随着褶皱程度的增大而增多, C—C断键、成键数量越多, 褶皱球结构释放后越容易稳定. 图8(a)给出了褶皱球结构释放时刻碳原子数密度和释放稳定后ρg值随含氧基团比例的变化. 当褶皱球结构总原子数密度为~200 nm–3时, 其对应的碳原子数密度随着σ的增大而增大, 释放稳定的ρg值随着褶皱球结构释放时刻碳原子数密度的减小而增大. 片层的褶皱程度随着σ的变化而变化, σ越大, 片层压缩到同一总原子数密度时的碳原子数密度越大, 对应的褶皱程度越大, 褶皱球结构越容易稳定. 图8(b)给出了褶皱球结构释放时刻C—C断键、成键数量随着含氧基团比例的变化, 褶皱球结构释放时刻C—C断键、成键数量随σ的变化与褶皱程度随σ的变化规律一致. 这同样说明褶皱球结构的稳定性与片层上C—C断键、成键数量有关, C—C断键、成键数量越多, 褶皱球结构释放后越容易稳定. 图 8 (a) 释放开始时刻的碳原子数密度与释放稳定后的ρg值; (b) 释放开始时刻的C—C断键、成键数量 (? = 50%) Figure8. (a) Number density of C atoms at the beginning of release process and the ρg value at the end of release process; (b) Number of broken and formed C—C bonds at the beginning of release process for the graphene oxide sheets with ? = 50%.