在什么条件下绝缘体能转变成导体?

高温?高压?强磁场?强电场?
关注者
578
被浏览
166,422

40 个回答

金属绝缘体相变是凝聚态物理中一个非常重要的问题。这学期在上一门相关的课,借此复习一下。

1. Wilson转变

Wilson转变通常是由高压所导致的。从紧束缚模型,我们可以知道,当晶体中的原子间距减小时,各原子的同一轨道会发生混合,导致能级展宽产生能带,如下图所示。

晶格常数越小,轨道交叠越多,能带越宽。假设原来费米面落在两个能带中间。随着外界压强增大,晶格常数变小,能带变宽。当能带宽到费米面上下两个能带交叠时,价带顶的部分电子会转移到导带底,此时费米面附近态密度大于零,绝缘体变成金属。如下图所示。

2. Mott绝缘体

上面说的Wilson转变纯粹在能带论范围内就可以解释,然而Mott绝缘体需要考虑电子间相互作用。还是从紧束缚模型出发,如果考虑当两个电子占据同一个原子的同一轨道时存在库伦排斥能,则电子的哈密顿量可以写成

H=-t\sum_{<i,j> \sigma}^{}{c_{i\sigma}^\dagger c_{j\sigma}}+U\sum_{i}^{}{n_{i,\uparrow}n_{i,\downarrow}}

此即Hubbard模型。i,j表示不同的原子的轨道,\sigma表示自旋,n是轨道的电子数。第一项表示不同原子轨道的混合(可以想象成电子在不同轨道间跳跃使动能减小)带来的能量展宽,就是紧束缚模型;第二项表示占据同一轨道电子的库伦排斥。

传统的紧束缚模型没有考虑电子电子相互作用。假设费米面在能带中心(按照能带论应该是金属)。如果能带展宽W小于U,该能带将分裂成两个子带,下面的带电子占据的是空态,上面的带电子占据的态已经有一个电子占据(电子跳跃需要克服库伦排斥),因而能量较高,如下图所示。

Mott绝缘体的一个例子是Y_{1-x}Ca_xTiO_3,是钙钛矿结构的过渡金属氧化物,Ti占据体心。Y和Ca只提供电子和占据位置。

如果x=0,由于Y是3价的,Ti的外层电子为3d1,能带恰好是半满。然而由于较大的U,这是一个Mott绝缘体。如果掺杂Ca,相当于引入空穴,Ti的填充少于半满,因而逐渐变为导体。

Mott绝缘体往往具有有趣的磁性质,很多是反铁磁材料。

3. Anderson局域化

Anderson局域化考虑的是有较多缺陷的金属。当金属中有比较少的缺陷时,缺陷所起的作用就是对电子的布洛赫波产生散射,从而贡献一部分电阻。然而,当缺陷较多时,电子将从巡游电子转变为局域化的电子。考虑下面的哈密顿量

H=\sum_{i}^{}{n_i\epsilon_i}+\sum_{i,j}^{}{t_{ij}c_i^\dagger c_j}

这里每处的E_i是一个分布范围为V的随机变量。这个随机分布的变量代表金属中的无规则分布的缺陷、杂质等。如果每个原子轨道的E_i相同(即V=0),则与紧束缚模型无区别,电子在不同态之间跳跃,从而可以导电。如果E_i的分布范围较大,大于紧束缚模型下的带宽W,那么可以想象这时电子都局域在每一个态上,不会在不同态之间跳跃,即产生局域化。如果介于两者之间,则能带中心附近的态仍为巡游电子态,能带两边的态被局域化。这两者的边界称为迁移率边。如下图所示。

在实验上,我们可以通过调节缺陷的数目(迁移率边位置)或者调节费米面的位置来实现金属绝缘体相变。例如下图,\alpha粒子破坏产生的缺陷越多,迁移率边越靠近能带中心。当费米面落到局域态中时,金属就变成了绝缘体。

4. Peierls转变

Peierls转变通常发生在一维体系中。一维等间距的具有一个价电子的原子链是不稳定的。此时,第一布里渊区恰好被填充了一半。如果晶格发生畸变,使得晶格常数变为原来的2倍(一个原胞有2个原子),则第一布里渊区变为原来的一半,并且电子的能量因带隙的出现而降低,从而使整个晶体能量降低。如下图:

。 。 。 。 。 。 。 。 →

。。 。。 。。 。。

这种相变和电荷密度波往往同时出现。

下面是一种有机一维体系TTF-TCNQ的电导率随温度的变化,可以明显看出低温下由导体变为绝缘体。

参考资料:1. 课程讲义

2. 黄昆、韩汝琦 《固体物理学》,高等教育出版社

前面几位答主做了很好的回答,下面将说一个金属——绝缘体——金属转变的例子,再简单说下mott转变。

1)金属——绝缘体——金属转变文献

这项工作由Ivan I. Naumov and Russell J. Hemley发表在PRL上 Origin of Transitions between Metallic and Insulating States in Simple Metals

论文摘要

Unifying principles that underlie recently discovered transitions between metallic and insulating states in elemental solids under pressure are developed. Using group theory arguments and first-principles calculations, we show that the electronic properties of the phases involved in these transitions are controlled by symmetry principles. The valence bands in these systems are described by simple and composite band representations constructed from localized Wannier functions centered on points unoccupied by atoms, and which are not necessarily all symmetrical. The character of the Wannier functions is closely related to the degree of s−p(−d) hybridization and reflects multicenter chemical bonding in these insulating states. The conditions under which an insulating state is allowed for structures having an integer number of atoms per primitive unit cell as well as reentrant (i.e., metal-insulator-metal) transition sequences are detailed, resulting in predictions of behavior such as phases having band-contact lines. The general principles developed are tested and applied to the alkali and alkaline earth metals, including elements where high-pressure insulating phases have been reported (e.g., Li, Na, and Ca).

大概是在说

最近发现,元素固体在压力下的金属—绝缘态之间转变的统一原理得到发展。利用群论理论和第一原理计算,我们证明了这些跃迁过程中所涉及的相的电子性质是由对称原理所控制的。这些系统中的价带由简单的组合表示法描述,由局部的Wannier函数构造,以原子的点为中心,而这些点并不一定都是对称的。Wannier函数的特征与s-p(-d)杂化的程度密切相关,并反映了这些绝缘状态下的多中心化学键合。一个绝缘态允许的结构,每个晶胞有整数原子数的结构,以及可重复的金属-绝缘金属转变序列,这导致了一些行为的预测,例如带接触的相。所发展的一般原理被测试和应用于碱金属和碱土金属,包括已经报道了高压绝缘相(例如Li,Na和Ca)的元素。

A view of the localized electrons in the unusual insulating state of Li under pressure, courtesy of Russell Hemley and Ivan Naumov. Credit: Russell Hemley and Ivan Naumov
图1.Electronic properties of hP4−Li for lattice parameters a=2.784 Å , c=3.873 Å (90 GPa). The parameters are intentionally kept the same as for hP4−Na at 320 GPa (see Ref. 4). (a) Band structure; the size of red circles is proportional to the s character of the Bloch wave functions. (b),(c) Valence charge density and crystal potential in the (110) plane, respectively. The ions are centered in the blue regions.
图2. Bottom and top valence band wave functions at Γ in hP4−Na at 320 GPa (corresponding to Γ+1 and Γ−4 energy levels) in the primitive unit cell. (a),(b) Isocontours of Γ+1 at ±2.5 and ±1.1 , respectively (yellow for + and light blue for − ). (c),(d) Same as (a),(b) but for Γ−4 .
图3. Bottom and top valence band wave functions at Γ in Aba2−Li at 80 GPa in the primitive unit cell. (a),(b) Isocontours of the bottom wave function at ±1.2 and ±1.1 , respectively. (c),(d) Same as (a),(b) but for the top wave function at ±1.8 and ±1.2 .
图4. Electronic properties of fcc-Ca at a=4.706 Å and 19.0 GPa. (a) Band structure. (b) Valence charge density in the (100) planes passing through the Ca atoms ( electrons/Bohr^3 ).
New work from Carnegie's Russell Hemley and Ivan Naumov hones in on the physics underlying the recently discovered fact that some metals stop being metallic under pressure. Their work is published in Physical Review Letters.
最新研究表明,一些金属在一定压力条件下,不再处于金属状态,Russell Hemley和Ivan Naumov.在物理结构上的发现为这一最新研究奠定了基础。他们的成果也发表在了Physical Review Letters上。
Metals are compounds that are capable of conducting the flow of electrons that make up an electric current. Other materials, called insulators, are not capable of conducting an electric current. At low temperatures, all materials can be classified as either insulators or metals.
金属是能够促进电子流动并形成电流的化合物。其他材料则被称为绝缘体,且无法形成电流。低温时,所有材料都可分为绝缘体或金属。
Insulators can be pushed across the divide from insulator to metal by tuning their surrounding conditions, particularly by placing them under pressure. It was long believed that once such a material was converted into a metal under pressure, it would stay that way forever as the pressure was increased. This idea goes back to the birth of quantum mechanics in the early decades of the last century.
通过改变周围条件,尤其是在施压条件下,绝缘体可以变为金属。传统观念认为,一旦某种材料在受压后变成了金属,随着压力的增加,它会永久保持这种金属的状态。这种观念的形成可以追溯到上个世纪量子力学最初诞生的几十年。
But it was recently discovered that certain groups of metals become insulating under pressure-a remarkable finding that was not previously thought possible.
但最新研究表明,某些特定的金属在受到压力后会变成绝缘体——这个惊人的发现在之前被认为是不可能的。
For example, lithium goes from being a metallic conductor to a somewhat resistant semiconductor under around 790,000 times normal atmospheric pressure (80 gigapascals) and then becomes fully metallic again under around 1.2 million times normal atmospheric pressure (120 gigapascals). Sodium enters an insulating state at pressures of around 1.8 million times normal atmospheric pressure (180 gigapascals). Calcium and nickel are predicted to have similar insulating states before reverting to being metallic.
例如,锂在79万倍的标准大气压(80吉帕斯卡)下,可从金属变为半导体。在120万倍的标准大气压(120吉帕斯卡)下可以重新变为金属。钠在180万倍的标准大气压(120吉帕斯卡)下可以变为绝缘体。钙和镍在转换为金属前,也被认为具有相似的绝缘状态。
Hemley and Naumov wanted to determine the unifying physics framework underlying these unexpected metal-to-insulator-to-metal transitions.
对于这种不寻常的金属到绝缘体再到金属之间的反复转换,Hemley和Naumov想找出一种统一的物理学原理对其作出解释。
"The principles we developed will allow for predictions of when metals will become insulators under pressure, as well as the reverse, the when-insulators-can-become-metals transition," Naumov said.Naumov
声明:“我们的研究理论将会对在一定压力环境下,金属到绝缘体、以及绝缘体到金属之间的转换作出预测。”
The onsets of these transitions can be determined by the positions of electrons within the basic structure of the material. Insulators typically become metallic by a reduction in the spacing between atoms in the material. Hemley and Naumov demonstrated that for a metal to become an insulator, these reduced-spacing overlaps must be organized in a specific kind of asymmetry that was not previously recognized. Under these conditions, electrons localize between the atoms and do not freely flow as they do in the metallic form.
这种转换可从金属基本结构中电子的位置着手。绝缘体变为导体,是由于物质中原子间距的减小造成的。Hemley和Naumov表明,金属变为绝缘体时,这些减小的间距重叠部分必须以某种特定的、不对称的方式进行组合。这种不对称的结构,与之前人们所理解的并不一样。在这种条件下,电子集中在原子之间。此外,与处于金属状态下不同的是,他们不能自由移动。
"This is yet another example of how extreme pressure is an important tool for advancing our understanding principles of the nature of materials at a fundamental level. The work will have implications for the search for new energy materials." Hemley said.
Hemley声称:“这个例子表明,超高压能够极大地提升我们基本层面上对于物质本质的认知水平。这个研究对于寻找新能源物质有着十分重要的意义。”

>>>>>>>>>后面再补一点mott相变的内容

2)mott转变

 mott transition至今仍是一个未能充份理解的现象,只知道它发生在转移金属硫族化合物和金属氧化物的过程中。受热驱动下可以发生金属—绝缘体转变。莫特转变是凝聚态中的金属 - 非金属转变。 由于电场的作用,势能在原子的平衡位置附近变得更尖锐(指数地),电子变得局域化,并且不再能够传导电流。

Mott绝缘体

* 在一定的条件下,一个基态是绝缘体的晶体,如果忽略关联能,则可能错误地把它当作金属

* 这样的绝缘体称为Mott绝缘体,这种金属——绝缘体的转变称为Mott转变

* MnO、CoO、NiO等就是这样的绝缘体

* 高温超导大多是这种掺杂Mott绝缘体

考察单电子孤立原子,互相靠拢形成一维晶体

* 原子互相靠近,孤立原子的分裂能级展宽成能带

* 假定能带宽度在-B/2到B/2的能量范围,总带宽为B

• 电子填充能带至半满时,即从-B/2填充到0

* 电子平均能为-B/4,即比孤立原子能级下降了-B/4

* 也即由于原子互相靠拢,平均每个电子获能B/4

相对于孤立原子,能级下降就是金属的内聚力

* 原因是波函数交迭成扩展态,平滑,因而电子动能减少(与波矢k有关),总能量减少

• 显然,带宽B与原子交叠因而与晶格常数有关,a趋于无穷大,B趋于0。

轨道全空、或单电子占据、或双电子占据的几率分别是1/4、1/2、1/4。 因此,要使单电子占据变成双电子占据或轨道全空,每个电子平均势能消耗为U/4。退局域得到能量B/4,局域化需能量U/4,因此如果U>B,将发生关联引起的局域

Mott转变的条件就是看U>B,还是B>U

* U对原子间距不敏感。当原子间距增大,能带变窄,即B减小;因此U>B时,将出现Mott转变,金属绝缘体。反之,当原子靠拢,能带变宽, 即B增大,B>U时,将从绝缘体变为金属

未完....


3.Anderson 绝缘

晶态到无序态

晶体周期性,由Bloch定理,才导致电子共有化,如果无序呢?

* 杂质,缺陷等都会导致周期性破缺,会在禁带中导致局域态

• Anderson提出,无序可导致局域态

* 如果局域态正好处于费米能级Anderson绝缘体!

• 无序分类

* 成分无序,拓扑无序,结构无序、关联无序,等等

安德森转变

1958年Anderson在“物理评论”上发表了一篇题为“扩散在某些无规格子中的消失”的文章,这篇文章开创了物理学的一个新领域,也正是这篇文章使人们对电子在无序体系中的行为特征有了初步了解。Anderson指出,处在局域态上的电子对传导没有贡献,随无序度增加,体系电子态会局域化。后人称这种局域化为Anderson局域化(Andersonlocalization)。这样,改变系统的无序度,可使系统由金属型变为绝缘体,人们称这类转变为Anderson转变(Anderson transition)。他考虑电子在一维(1D)点阵上运动,点阵位具有随机势,电子只被允许在最邻近位间跃迁。结果表明,如果无序很强,波函数可以变成按指数局域化。1979年 Anderson 又和Abrahams等发表文章提出了一种著名的局域标度理论,此理论证实,对于无相互作用、存在非关联无序 的所有1D、2D系统,只要存在任意大小的无序,波函数都是指数衰减的。仅在3D中,单电子态可以在弱局域或态密度足够小的情形保持扩展态。

基本思想

具有严格周期性的有序晶格其波函数具有平移不变性,满足Bloch定理,电子处于扩展态。

所有电子在有序晶格中作公有化运动->扩展态

在晶体中引入缺陷—>周期性结构破坏—>局域在杂质附近

此时,波函数是指数衰减形式

大体来说,就是杂质或缺陷导致周期性结构破坏,电子的 Bloch 波函数需要修正。随着杂质浓度升高,电子波函数将发生衰减,逐渐从扩展态转变为局域态。结果就是金属变为绝缘体。

Anderson 局域

无平移对称性,波矢k不再是描述电子态的好量子数

TBA(紧束缚近似)—>无序系统

假定有一周期势如图(a)所示,每个原子由一方势阱表示并只有一个价电子,在孤子原子极限下占据在图中原子势阱处水平短线表示的束缚能级上。在晶体中,这一原子能级因波函数的交叠关联展宽成宽度为B的能带。无序可以两种形式引人,一种是每一格点相对于平衡位置有一无规偏移,另一种是原子位置保持在格点上,势阱的深度、因为束缚能级从一个格点到另一个格点无规变化(图(b))。安德森的讨论采用后一种无序情形。
在体系的长程有序消失后,波矢k不再是描述电子态的好量子数,因此对无序体系电子态的研究,广泛采用紧束缚近似,从院子轨道波函数,或Wannier函数出发来讨论。

4.结构畸变,比如Peierls相变导致的金属绝缘体转变

Peierl's distortion of a 1-d periodic lattice.

单价原子,排列成如图的一维晶体。如果每隔一个原子发生了如图的原子移动。用空晶格模型+微扰方法,分析原子结构变化前后的能带变化,并得到能带示意图。

The lowest Bloch band of an undistorted 1D lattice.


The lowest Bloch bands of a distorted 1D lattice. Energy gaps appear in {\displaystyle ka=\pm \pi /2} {\displaystyle ka=\pm \pi /2} as a result of the Peierls' instability.

原胞变大,布里渊区变小,原来一条能带折叠成两条能带,根据微扰方法,边界上简并的能

带将发生分裂。原来原胞内只一个原子,一个电子;现在原胞内有两个原子,两个电子原

半满的能带,现在全满,导带变为了新的价带+空带,金属将变为绝缘体。