【作者】 段德芳；

【导师】 崔田；

【作者基本信息】 吉林大学 ， 凝聚态物理， 2011， 博士

【摘要】 高压下分子体系的物理行为一直以来都是物理、化学、材料、地球以及行星科学的一个热点研究课题。分子体系中分子间相互作用较弱,很容易受到压力的影响而缩短分子间的距离,导致晶格的重新排列,引起结构相变;分子间距离的缩小同时会增加相邻电子轨道的重合程度,进而导致电子相变。Wigner提出所有的分子体系在压力作用下一定会塌陷,形成了密堆积结构,并且在足够高的压力下会转变为金属。在实际的分子体系当中,针对不同的分子晶体,分子解离形成的新结构及物理机制、金属化的成因、及超导转变温度随压力的变化都会表现出多样性,并被实验证实。第一种类型是典型的双原子分子晶体H₂, N₂, O₂,和I₂等等,它们在高压下表现出非常复杂的相图。第二种类型是含氢分子体系,如H₂O, NH₃,和HF等等,它们在高压下会发生氢键对称化等新奇的现象。分子晶体MI₄（M=Ge,Sn）是另外一类例子,它们经历了压力导致的非晶,并且变成了金属玻璃。影响分子晶体的各种因素中,分子的形状和分子间的相互作用是比较重要的,因为化学键的各向异性、晶体结构和电子特性主要依赖于形成晶体的分子形状和分子间的相互作用,压力效应也主要是通过影响这些因素而显现出来的。近年来,随着理论的完善和计算机的发展,第一性原理计算方法已成为凝聚态物理、量子化学和材料科学中重要的研究手段。金属氢是二十一世纪最为重要的十大物理问题之一,最有可能是室温超导体和优质能源材料。为了获得金属氢,最有效的方法是加压。但是由于氢分子内部键合非常强,至今没有获得金属氢。在这样的背景之下,人们开始研究与氢类似的双原子分子体系,比如卤族元素固体的高压行为,以便获得有助于研究金属氢的规律。在本论文中,我们选择了四种不同的含卤族元素的典型分子晶体,用基于密度泛函理论的第一性原理计算方法研究了它们在高压下的分子解离、金属化及超导电性等一系列物理行为,取得一些创新性的成果。（1）非静水压对原子相碘超导电性的影响。实验结果表明碘的原子相具有超导特性,它的第II相（底心正交结构,bco）和第III相（体心四方结构,bct）的超导温度T_c随压力增大逐渐降低,相变到第IV相（面心正交结构,fcc）时,T_c开始升高。这种随压力变化的超导行为和机制一直没能得到很好的解释,在此背景之下,我们利用第一性原理方法研究了碘单原子相在高压下的超导电性。计算结果表明在静水压作用下,碘的第II和III相的超导温度T_c与实验值符合的很好,但是第IV相的T_c随压力增大而降低,这与其它理论计算符合,而与实验观测到的变化趋势相反。为了探索实验观测和理论计算差异的原因,并且考虑到实验由于未使用传压介质,可能会产生非静水压效应,所以研究了两种非静水压效应对单原子相的超导电性的影响。通过各向异性压缩fcc结构,得到了两个新的非静水压结构:体心四方（nonhy-bct）和面心正交（nonhy-fco）。发现在第一种非静水压下,nonhy-bct的T_c随压力增大而降低,不符合实验变化趋势,而在第二种非静水压下,nonhy-fco的T_c随压力增大而升高,与实验值符合的很好。进一步的研究表明,nonhy-fco的T_c随压力升高主要归因于费米面处态密度的升高和电声矩阵元<I²>的增大。计算结果很好的解释了长期以来一直没有得到很好解释的实验现象,表明非静水压可以产生的一些新奇特性。（2）高压下溴原子相的结构和超导电性。通过高压X光衍射,实验上已经获得碘单原子相的相变序列和结构,但是由于受到实验条件的限制,溴原子相的结构到目前为止还不是很清楚。我们通过第一性原理计算成功的模拟了单原子溴的相变序列:体心正交结构（bco,相II,空间群Immm）在126 GPa转变为体心四方结构（bct,相III,空间群I4/mmm）,之后在157 GPa转变为面心立方结构（fcc,相IV,空间群Fm-3m）,这个面心立方结构在300 GPa时都能稳定存在。我们计算了不同压力下bct结构的能量随晶格常数a的变化情况,发现总能与a的关系曲线呈现双势阱特征,两个能量最低点分别对应bct和fcc结构,用这个双势阱模型成功的解释了bct转变到fcc时晶格常数a突变的原因。另外我们计算了单原子相的超导电性,在100 GPa的超导转变温度为1.46 K与实验上观测到的1.5 K非常吻合。并且发现超导转变温度在第II相、第III相、第IV相每个相区间都随压力的增大而降低,与碘的单原子相的变化相似,进一步的理论计算揭示了超导电性的物理机制,发现第IV相的电声相互作用参数λ随压力增大降低主要是由声子软化的逐渐消退引起的。（3）高压下HBr和HCl的氢键对称化和超导电性。氢键是分子内或分子间的一种弱相互作用,它在物理、生物、化学、材料等领域扮演着一个非常重要的角色,是目前人们研究的热门领域之一。氢键对称化是高压下的一个重要现象,实验上观测到HBr和HCl在高压下会发生氢键对称化。另外,因为HBr和HCl含氢元素,在高压下它们还有比较有趣的特性—超导,最近实验和理论上都发现很多富氢体系如SiH₄,GeH₄,SnH₄,YH₃,ScH₃和LaH₃等等在高压下具有较高的超导转变温度。卤族元素氢化物HBr和HCl是最简单的氢键双原子分子晶体,对它们进行研究可以对其它氢键体系提供理论指导。我们对HBr和HCl的氢键、结构和超导在高压下的变化进行了系统的研究。计算结果表明HBr和HCl分别在25 GPa和40 GPa发生氢键对称化,并同时发生了结构相变（Cmc2₁→Cmcm）,氢键对称化主要是由拉伸对称模式A₁软化导致的。氢键对称化后HBr和HCl的Cmcm相的横声学支分别在160 GPa和250 GPa出现虚频,说明这个结构不再稳定,将会有新的结构出现。我们通过移动原子位置得到一个新相（第V相）,这个新相是属于P21/m空间群的单斜结构,仍保持氢键对称化特性。对HBr,这个结构稳定存在134 GPa到196 GPa之间,之后分解为Br₂和H₂。对HCl,第IV相在233 GPa转变到第V相,但一直到400 GPa都未出现分解。进一步研究了这个新相的超导,发现HBr在160 GPa下的超导转变温度为27³4 K,HCl在280 GPa的超导转变温度为9¹4 K,并且随压力的增大超导转变温度升高。（4）四碘化锡（SnI₄）分子晶体的压致非晶和再结晶。自从Mishima等人发现高密度非晶冰以来,在很多体系中都发现了压力导致的非晶化,如Si,Sn,和P等等。四碘化锡（SnI₄）在高压下也具有非晶态,尽管有大量的实验研究了SnI₄的高压行为,但是压力导致的SnI₄的非晶结构依然存在争议。另外,N. Hamaya等人发现非晶态在更高的压力64 GPa又结晶为新的晶体,但是这个晶体的结构并未确定。我们的研究结果显示,晶格常数在25 GPa突然劈裂,体积明显降低,同时,分子内Sn-I的距离突然增大,而分子内和分子间碘和碘的距离突然降低。所有这些信息都表明四面体SnI₄分子晶体发生解离,与实验上观测在25 GPa转变为非晶态符合。另外采用经典分子动力学方法得到了SnI₄的非晶结构,它的XRD谱和对分布函数与实验结果非常符合。在60 GPa,晶格常数又发生突变,说明非晶态再结晶转变为晶体相III（CP-III）,与实验发现在61 GPa出现的晶体相吻合。并且首次得到了CP-III的结构,具有P2₁/c空间群,这个结构的XRD谱也与实验结果非常符合。（5）三碘化硼（BI₃）的分子解离及其超导电性。实验测量BI₃分子晶体在6.2 GPa分子解离,此时碘原子的排列为面心立方结构,但是由于硼原子质量相对碘太小,所以无法得到硼原子的结构信息。电阻测量表明BI₃在23 GPa金属化,27 GPa的超导转变温度为0.5 K,但是高压下BI3的金属化以及超导电性的机制尚不清楚。我们首次得到了三碘化硼高压新相（第II相）的结构,具有P2₁/c空间群,这个结构的XRD与实验结果符合的很好。通过焓曲线发现第I相（P6₃/m）在5.6 GPa转变为第II相（P2₁/c）,与实验测量值6.2 GPa非常吻合,并且通过声子和弹性计算证明了新结构的动力学和力学稳定性。我们还计算了BI₃的电子结构,发现第II相在30 GPa由绝缘体转变为金属,这主要是由能带交叠引起的。通过电声相互作用计算了BI3的超导电性,在60 GPa的超导温度为0.5 K。更多还原

【Abstract】 The effect of high pressure on molecular systems has been a central issue of fundamental physics and chemistry as well as planetary sciences. Cohesion of simple molecular solids occurs through forces of very different strengths: covalent, ionic, van der Waals, and hydrogen bonds. Pressure drives materials to states of higher density and gives rise to competition among those chemical bonds, structural instabilities, and changes in electronic properties. A simple picture suggests that all molecular systems must collapse on compression to form closed-packed structures and go over into metallic states at sufficient high pressures. However, the diversity of a process toward their destruction in real substances has manifested itself in numerous experimental observations. For instance, diatomic molecules crystals H₂, N₂, O₂, and I₂ are widely known to exhibit unexpected phases and complex phase diagram. Another class is hydrogen-containing molecules, such as H₂O, NH₃, and HCl, they turn out to be hydrogen bond symmetrization under high pressure. Metal tetraiodides MI₄ （M=Ge, Sn） are examples of another class. They undergo pressure-induced amorphization and become metallic glasses, which are quite common in materials having tetrahedral coordination. Among various factors affecting on the response of molecular crystals to compression, the shape of a molecule and intermolecular interactions may be of particular importance, because the anisotropy of chemical bonds, crystal structure, and electronic properties strongly depend on the shape of molecules composing a crystal and intermolecular interactions between the individual building blocks. Most recently, with the help of improved theory and computational capability, ab initio calculation based on the density functional theory has been used widely in the condensed matter physics, quantum chemistry, and material science. And it has been a common research tools except for theoretical and experimental method. Metal hydrogen is the most important one of top ten physical problems in the 21st century. Most likely it is a room temperature superconductors and good energy materials. Pressure is the most effective method to obtain metal hydrogen. Because the hydrogen molecule internally bonding is very strong, so far, there is no metal hydrogen observed in laboratory. As Ashcroft pointed out, hydrogen is regarded naturally as the first element of the halogen group. Therefor, we detailedly study the pressure induced molecular dissociation, metallization and superconductivity of solid halogens and related halide, which are also valuable for providing insight into the metallic hydrogen.（1） Effect of nonhydrostatic pressure on superconductivity of monatomic iodine. The superconductivity of iodine had been successfully discovered with T_c = 1.2 K at 28 GPa. It was reported that the T_c of monatomic iodine decreased with pressure at first but started to increase with pressure for the highest-symmetry phase IV, the face-centered cubic phase. The mechanism of such a superconductivity with pressure is still unclear. So, we have presented an ab initio investigation of the hydrostatic and nonhydrostatic pressure effects on the superconductivity of monatomic iodine.It is shown that the T_c of both phase II and phase III under hydrostatic pressures are in agreement with the experimental data, while the T_c of phase IV under hydrostatic pressures decreases with increasing pressure, contrary to the experimental results. In order to explore the origin of difference between experimental and theoretical results, we have studied the effect of non-hydrostatic pressure on the superconductivity of monatomic iodine, and found that the symmetry of phase IV changes from face-centered cubic to face-centered orthorhombic （fco） under anisotropic stresses. Further calculations show that the T_c of this fco structure increases with increasing pressure, in good agreement with the experimental results, which is mainly attributed to the non-hydrostatic pressure-induced enhancement of the electronic density of states at the Fermi level and electron-phonon coupling matrix element <I²>.（2） Crystal structure and superconducting properties of monatomic bromine under high pressure. The monatomic phase transition sequence of iodine had been observed by X-ray diffraction experiment. A similar scheme of phase transformations can be expected for bromine, but experimental results are much scarcer than those in the iodine case which restricts our understanding of the nature of bromine under high pressure. So, the crystal structure and superconducting properties of monatomic bromine under high pressure have been studied by ab initio calculations.We have found the following phase transition sequence with increasing pressure: from body-centered orthorhombic （bco, phase II） to body-centered tetragonal structure （bct, phase III） at 126 GPa, then to face-centered cubic structure （fcc, phase IV） at 157 GPa, which is stable at least up to 300 GPa. The calculated superconducting critical temperature T_c = 1.46 K at 100 GPa is consistent with the experimental value of 1.5 K. In addition, our results of T_c decreases with increasing pressure in all the monatomic phases of bromine, similar to monatomic iodine. Further calculations show that the decrease ofλwith pressure in the phase IV is mainly attributed to the weakening of the“soft”vibrational mode caused by pressure.（3） Hydrogen bond symmetrization and superconducting phase of HBr and HCl under high pressure. Hydrogen bonds are quite pervasive in a broad range of fields including physics, chemistry, biology, and materials sciences. Besides, hydrogen bond symmetrization is an important high pressure phenomenon. The pressure-induced hydrogen bond symmetrization in hydrogen halides （HBr, HCl, and DCl） have also been observed by Raman and infrared measurements. In addition, hydrogen in HBr and HCl can lead to other interesting properties under high pressure. Recently, theoretical or experimental studies have reported that these hydrogen compounds such as SiH₄, GeH₄, SnH₄, YH₃, ScH₃ and LaH₃ present a high superconducting critical temperature. HBr and HCl are simple diatomic molecules forming hydrogen bond in condensed state. Therefore, the studies on HBr and HCl can provide theoretical guidance to other hydrogen bonding system.Ab initio calculations are performed to probe the hydrogen bonding, structural and superconducting behaviors of HBr and HCl under high pressure. The calculated results show that the hydrogen bond symmetrization （Cmc2₁→Cmcm transition） of HBr and HCl occurs at 25 and 40 GPa, respectively, which can be attributed to the symmetry stretching A₁ mode softening. After hydrogen bond symmetrization, a pressure-induced soft transverse acoustic （TA） phonon mode of Cmcm phase is identified, and a unique metallic phase with monoclinic structure of P2₁/m （4 molecules/cell） for both compounds is revealed by ab initio phonon calculations. This phase preserves the symmetric hydrogen bond and is stable in the pressure range from 134 GPa to 196 GPa for HBr and above 233 GPa for HCl, while HBr is predicted to decompose into Br₂+H₂ above 196 GPa. Perturbative linear-response calculations predict that the phase P2₁/m is a superconductor with T_c of 27³4 K for HBr at 160 GPa and 9¹4 K for HCl at 280 GPa.（4） Pressure-induced amorphization and recrystal of tin tetraiodide molecular crystal. In recent years, pressure-induced amorphization （PIA） has attracted extensive experimental and theoretical interest such as H₂O, SiO₂, P, etc. Tin tetraiodide SnI₄ molecular crystal is also observed to show amorphous under pressure. Although there have been a variety of experimental studies on SnI₄ under pressure, the structure of PIA forms is still controversial. In addition, the amorphous recrystallizes to a nonmolecular crystalline phase III （CP-III） at 61 GPa, but the crystal structure is not clear. Here we report an ab intio study that reveals the mechanisms controlling PIA in SnI₄, provides important insights pertaining PIA phenomena at large, and gives the structure of CP-III.Full geometry optimization show that, at 25 GPa, the lattice constants abrubtly split, volume significantly decreases, and distance of intramolecular Sn-I increase abruptly, while intermolecular and intramolecular I-I decreases suddenly. These indicate the tetrahedral molecular dissociate at 25 GPa, which is in good agreement with the experiment results. In addition, we obtained an amorphous structure through the classical molecular dynamics. The XRD and radial distribution function of this structure is consistent with the experimental resuts. At 60 GPa, the lattice constants changed abrubtly, indicating that the crystal phase （CP-III） occured. We firstly got a structure of CP-III with space group P2₁/c which has 40 atoms in unit cell. The XRD of our calculated structure is consistent with the experiment measure, indicating that our predicted structure is correct.（5） Pressure-induced molecular dissociation and superconductivity of boron triiodide. It is reported that BI₃ molecular structure transform to a monatomic phase at 6.2 GPa with the face-centered-cubic lattice of iodine atoms by X-ray diffraction experiment. Since the atomic X-ray scattering power of boron is only 5/53 of that of iodine or less, the boron atoms were not detected. So, the crystal structure of new phase is not clear. The monatomic phase becomes metallic at 23 GPa and exhibits superconductivity above 27 GPa by resistivity measurements.We got a new structure of BI₃ with space group P2₁/c which has 4 moleculars in unit cell. The XRD of P2₁/c is in good agreement with the experiment measure, indicating that our predicted structure is correct. The pahse I （P6₃/m） transforms to pahse II （P2₁/c） at 5.6 GPa, which is in well agreement with the experimental results. Moreover, the P2₁/c structure is dynamical and mechenical stability by phonon and elastic calculation. Another phase transition from insulator to metal phase occurs at 30 GPa which is primarily attributed to the band overlap. Perturbative linear-response calculations predict that the phase P2₁/c is a superconductor with T_c of 0.5 K.更多还原