【作者】 李丽；

【导师】 许斌； 李木森；

【作者基本信息】 山东大学 ， 材料学， 2008， 博士

【摘要】 人造金刚石单晶不仅具有硬度高、抗腐蚀、高耐磨等优异性能，还具有优良的光学、声学、热学和电学性质，不断表现出其在现代科学技术和发展中的重要作用。目前，最具有工业生产规模与广泛应用价值的金刚石单晶合成方法仍然是高温高压触媒法。高温高压金刚石的生长机理对于指导工业生产金刚石无疑具有重要意义，但由于高温高压下在线检测的困难性，理论研究的难度仍然过大，造成目前学术观点尚未统一，尤其是对金刚石生长的碳源这一关键问题仍存在较大争议，近年来金刚石生长机理方面的研究投入也较少，对机理的研究依然是一个重大的探索性课题。课题组前期对合成金刚石后的触媒及包覆着金刚石单晶的金属包膜的组织、结构、成分等进行了系统的实验表征，根据前期的实验结果，本文利用余氏理论和程氏理论计算分析了高温高压触媒法金刚石生长中各物相的价电子结构及界面的电子密度，从电子结构角度研究了金刚石生长的碳源问题及触媒的催化作用，探讨了高温高压金刚石生长机理；同时结合热力学理论，解释触媒在其中的变化过程，从热力学角度进一步分析了金刚石生长的碳源问题。从而为金刚石的机理研究开辟了一条新途径，并提出了触媒成分设计的新思路。本文首先根据材料的热膨胀本质和广义虎克定律，利用晶体的线膨胀系数和弹性常数，建立了晶格常数与温度和压力之间的关系。运用该方法计算六方石墨在不同温度和压力下的晶格常数，所得结果与前人的实验结果非常接近，验证了本文计算方法的可行性。进而计算了金刚石合成过程中各物相（金刚石、石墨、Fe₃C、γ-（Fe，Ni）等）的晶格常数随温度和压力的变化，为高温高压条件下晶体的的价电子结构分析提供了计算基础。根据价电子理论，异相界面的电子密度应连续，则在金刚石晶体生长中，碳源相与金刚石界面的电子密度应保持连续，这是金刚石生长要满足的边界条件。本文对金刚石和石墨的价电子结构分析表明：常温常压下，金刚石和两种石墨各主要晶面之间的最小电子密度差在80％左右，而1600 K、5．5 GPa时最小电子密度差在60％左右，虽然由于温度和压力的作用，金刚石和石墨之间的电子密度有所接近，但仍要要远远大于10％，即在一级近似下是不连续的，不能满足金刚石生长的边界条件。因此从电子结构角度看，触媒法金刚石晶体生长的直接碳源不是来自石墨。另外，石墨结构中共价键的键能随温度和压力变化并不明显，最强键上的键能约为240 kJ／mol，平面网层之间的共价键能非常小，靠范德华力结合。在金刚石合成过程中，部分石墨需以C原子形式溶入似熔态触媒，与触媒合金形成碳化物或间隙固溶体。已有研究证实选择铁基触媒合成金刚石具有较好的应用前景及较高的学术研究价值，课题组前期对合成后的Fe-Ni触媒及包膜进行了系统的研究，发现在包膜／金刚石界面存在着大量的Fe₃C和γ-（Fe，Ni），并推测Fe₃C为金刚石晶体生长的碳源相，γ-（Fe，Ni）为催化相。因此，本文以Fe-Ni-C系金刚石晶体生长为例探讨金刚石的生长机理，对合成后包膜中主要物相的价电子结构及界面的电子密度进行了计算分析。对Fe₃C的价电子结构及Fe₃C／金刚石界面电子密度的分析表明：高温高压状态下，Fe₃C／金刚石界面的电子密度在一级近似下是连续的，能够满足金刚石生长的边界条件。因而，高温高压触媒法合成金刚石，并不是由石墨结构直接转变为金刚石结构，而是C原子集团不断从Fe₃C中脱落，转移到与之电子密度相近的金刚石界面上，进而完成金刚石晶体的生长。另外，Fe₃C的两个主要晶面同金刚石（111）晶面的电子密度连续，可以解释金刚石包裹体中薄片状Fe₃C同金刚石（111）面存在着平行的位向关系这一现象。对γ-（Fe，Ni）的价电子结构及γ-（Fe，Ni）／Fe₃C界面电子密度的分析则发现：γ-（Fe,Ni）／Fe₃C界面的电子密度在一级近似下是连续的，这表明在金刚石生长过程中γ-（Fe，Ni）起着促使Fe₃C分解的作用即催化作用。可见，价电子理论分析结果与前期实验表征结果是相吻合的。为了分析不同触媒的催化作用，进而尝试从电子理论上指导触媒的成分设计，本文对采用过渡族金属（Fe、Ni、Mn、Co）及其合金为触媒合成金刚石过程中可能形成的各种Me₃C型碳化物与金刚石界面以及不同成分配比的γ-Me固溶体与相应Me₃C界面的电子密度分别进行了分析，结果表明：各Me₃C型碳化物与金刚石界面的电子密度以及各γ-Me固溶体与相应Me₃C界面的电子密度在一级近似下均连续，从而可以认为金刚石生长的碳源相和催化相分别为Me₃C和γ-Me固溶体。不同碳化物与金刚石界面的电子密度连续性不同，与金刚石界面保持电子密度连续性越好，结构转化所需要越过的化学势垒越低，也就越容易转变为金刚石结构，在相同的合成条件下，合成的金刚石品质更好。Fe、Ni、Mn、Co合金碳化物与金刚石界面的电子密度连续性基本上都分别好于其单金属碳化物；所有碳化物中，Mn和Co基碳化物与金刚石的电子密度连续性最好；Fe基碳化物与金刚石界面的电子密度连续性好于Ni基碳化物，其中（Fe,Ni）₃C／金刚石界面的电子密度连续性最好。不同元素组成及不同成分配比的γ-Me固溶体与相应Me₃C界面的电子密度连续性也不同，连续性越好，越易促使Me₃C分解，则金刚石的生长速度越快。其中，γ-（Fe，Ni）随着Ni含量的增加，与Fe₃C界面的电子密度连续性基本上呈逐渐变差的趋势。体现实际合成工艺中为，随着Fe-Ni触媒中Ni含量的增加，金刚石的生长速度逐渐变慢，这与金刚石合成实验相吻合。从电子结构的角度提出了良好的触媒剂所应具备的三个条件：高温高压下能与石墨形成Me₃C型碳化物；Me₃C型碳化物与金刚石生长界面有较高的电子密度连续性；γ-Me固溶体与Me₃C型碳化物界面的电子密度连续性适中。根据价电子理论分析的结果，本文对高温高压触媒法金刚石的生长进行了热力学分析，在计算中考虑了体积随温度和压力的变化，结果表明：在金刚石形成之前就有大量Fe₃C形成，而在触媒法合成金刚石的温度和压力范围内，Fe₃C（?）C（金刚石）+3γ-Fe反应的自由能变化和石墨（?）金刚石相变的自由能均为负值，但前者比后者更负，即前者更容易发生。因此，从热力学角度来看，Fe₃C的形成降低了石墨转变为金刚石所要越过的势垒，使用铁基触媒合成金刚石晶体的生长来源于Fe₃C的分解而不是石墨的直接转变。同时，得出了在1200K以上石墨一金刚石的平衡曲线P-T关系：P_eq（GPa）=1．036+0．00236T（K），这一结果与Bundy计算的平衡线比较接近，从而验证了本文热力学计算方法的可行性。本文基于价电子理论和热力学理论的计算分析，均支持了“高温高压触媒法合成金刚石单晶的生长来自于Me₃C型碳化物的分解，而非石墨结构的直接转变”这一论述。对高温高压触媒法金刚石的生长过程可以总结以下：更多还原

【Abstract】 Man-made diamond single crystals not only have the excellent performances of high rigidity, corrosion resistance and high wearing resistance but also the fine properties of optics, acoustics, thermotics and electricity, which make them play important roles in the development of morden science and technology. Undoubtedly, diamond growth mechanism is significant to instructing the commercial production of diamond. However, due to the difficulty of on-line observation at high temperature and high pressure （HPHT） the difficulty of theoretical study is so hard that the academic viewpoints about the growth mechanism especially the key problem about carbon source are still not consistent, and few efforts have been put into the mechanism research in these years. The study on diamond growth mechanism is still a crucial groping subject.Our team has investigated the pattern, composition and structure of catalyst and metal film surrounding as-grown diamond systematically with many kinds of experimental methods in earlier stage. In this paper, according to the former experimental results gained by our team the valence electron structure（VES） of phases existent during the course of diamond growth and the electron density of interfaces were calculated and analyzed with EET and TFDC theory, and the carbon source problem and effect of catalyst were analyzed. Thereby, the diamond crystal growth was investigated in the viewpoint of electron structure. At the meantime, the carbon source problem was analyzed further combined with thermodynamic theory. Accordingly, a new theoretical path was explored to fulfill the study of diamond synthesis mechanism, and a new thought to design the composition of catalyst was put forward.According to the essence of thermal expansion and generalized Hooke’s law, the relation between lattice constant of crystal and temperature and pressure were established based on the linear thermal expansion coefficient and elastic constant of crystal. The errors between the lattice constants of graphite at different temperature and pressure calculated with this method and the experimental data are small, which validate the feasibility of the calculational method in this paper. Then the changes of the lattice constants of phases existent during the course of diamond growth with temperature and pressure were calculated to supply a basis for the calculation of VES of crystal at HPHT.According to the valence electron theory, the electron density of carbon source phase/diamond must be continuous, which is the boundary condition for diamond crystal growth. The VES analysis of diamond and graphite shows that the minimum electron density differences between the common planes of diamond and graphite are about 80% at normal temperature and pressure, while they are about 60% at 1600 K and 5.5Gpa. Although their electron densities are approaching with the increace of temperature and pressure, the difference is extraordinarily bigger than 10%, that is, they are not continuous at the first approximation. This discontinuity can’t satisfy the boundary condition of diamond crystal growth. Therefore from the viewpoint of electron structure, the cabon source for diamond growth with the method of catalyst at HPHT does not come from graphite directly. The bonds’ energy of graphite changes unconspicuously with the change of temperature and pressure. The energy of the strongest bond is about 240 kJ/mol, while the bond energy among the parallel layer planes is very small and the planes are mainly bonded by Van der Waals force. During the diamond growth, parts of the graphite dissolve into the melting catalyst with the manner of C atoms, then form the carbides or solid solutions with metal or alloy catalyst.The former studies about metal catalyst have verified the excellent application prospect and academic value of diamond synthesis with Fe-based catalyst. Our team investigated systematically the Fe-Ni catalyst and metal film after diamond synthesis, and found that there were a lot of Fe₃C andγ-（Fe,Ni） phases on the interface of film/diamond, and considered Fe₃C andγ-（Fe,Ni） as carbon source phase and catalysis phase respectively. Accordingly, in this paper the diamond growth mechanism was investigated with the example of diamond growth from Fe-Ni-C by calculating and analyzing the VES of main phases in metal film and the electron density of interface. The analyses on the VES of Fe₃C and electron density of Fe₃C/diamond interface show that the electron density of the interface is continuous at the first approximation, which can satisfy the boundary condition of diamond growth. Therefore, the carbon source of diamond growth with catalyst at HPHT comes from the carbon atom groups separated from Fe₃C instead of the direct transformation of graphite structure. Moreover, the electron densities of two main crystal planes of Fe₃C are continuous with that of （111） plane of diamond, which can explain perfectly the parallel direction relationship between Fe₃C in the inclusion of diamond crystal and the （111） plane of diamond. The analyses on the VES ofγ-（Fe,Ni） and electron density of Fe₃C/γ-（Fe,Ni） interface show that the electron density of the interface is continuous at the first approximation, which illuminates thatγ-（Fe,Ni） plays a role of catalysis phase, that is,γ-（Fe,Ni） improves the decomposition of Fe₃C. Thus it can be seen that the results based on valence electron theory analysis meet with the former experimental results gained by our team.In order to analyze the effect of different kinds of catalyst and instruct designing catalyst composition with electron theory, the electron density of the diamond/carbide interfaces which are probably formed during the diamond growth with transition group metals （Fe, Ni, Mn, Co） and their alloys as catalyst and the interfaces formed ofγ-Me solid solutions with different mixture ratio and Me₃C corresponding. The results show that the electron densities across Me₃C/diamond and y-Me/Me₃C interfaces are all continuous at the first approximation, which indicates Me₃C andγ-Me can be considered as carbon source phase and catalysis phase respectively. The electron density continuities of carbide/diamond interfaces are different with carbide, moreover, the better the electron density continuity of interface, the lower the chemical potential getting over to fulfill structure transformation, and the easier the transformation to diamond structure, so the better the diamond quality in the same synthesis condition. The electron density continuity of Fe, Ni, Mn, Co alloying carbide/diamond interfaces are mostly better than that of monometallic carbide/ diamond interfaces; the electron density continuity of the Mn based and Co based carbide/diamond interfaces are better than other carbides; the electron density continuity of the （Fe,Ni）₃C/diamond is the best among Fe based and Ni based carbides. The electron density continuitiy ofγ-Me/Me₃C interfaces are different with the different compositon ofγ-Me. With the increasing of continuity, the growth rate of diamond crystal quickens. The electron density continuity ofγ-（Fe,Ni）/Fe3C interface ebbs with the adding of the Ni content. That is, the growth rate of diamond crystal becomes lower with the increase of the Ni content of Fe-Ni catalyst in the same synthesis condition. The VES analyses of the effect of catalyst mostly meet with the experimental results. Accordingly, a new thought about the design of catalyst composition can be put forward that the Me₃C type carbides must be formed by the effect of catalyst and graphite; the electron density continuity across Me₃C/diamond interface is high; and the electron density continuity acrossγ-Me/Me₃C is appropriate.According to the results of valence electron theory analysis, the diamond growth was also analyzed with thermodynamic theory, and the changes of volume with temperature and pressure were involved in the calculation. The results show that the Fe₃C phases have been formed before diamond nucleation; at the temperature and ressure range of the diamond synthesis method with catalyst, the Gibbs free energies of Fe₃C（?）C（diamond）+3γ-Fe and graphite（?）diamond are all negative, but the former is more little than the latter, which means the former will take place more easily.Therefore, form the viewpoint of thermodynamics the formation of Fe₃C reduces the potential energy of transformation from graphite to diamond, and the diamond crystal growth with Fe based catalyst comes from the decomposition of Fe₃C instead of the direct transformation from graphite structure to diamond structure. Moreover, the P-T （Pressure-Temperature） equilibrium of P_eq（GPa）=1.036+0.002367（K） is gained, which is closer to the equilibrium gained by Bundy. Thereby the feasibility of the thermodynamic calculational method in this paper is verified.The analyses based on the valence electron theory and thermodynamics both hold up the viewpoint that the carbon source of diamond crystal growth with the method of catalyst at HPHT comes from the decomposition of carbide instead of the direct transformation from graphite to diamond. Accordingly, the diamond growth with the method of catalyst at HPHT can be described as following.更多还原