Chapter 1
STRUCTURE OF MINERAL SURFACES
Andrew S. Gibson and John P. LaFemina
CONTENTS
I. Introduction
A. Purpose and Scope of Chapter
B. Basic Nomenclature
C. Principles Governing Surface Structures
1. Principle 1: Stable Surfaces Are
2. Principle 2: Rehybridize the Dangling Bond Charge Density
3. Principle 3: Form an Insulating Surface
4. Principle 4: Conserve Near-Neighbor
5. Principle 5: Kinetics Are Important
D. Chapter Organization
II. Methods of Surface Structure Determination
A. Experimental Methods
1. Scanning Probe Microscopies
2. Electron Diffraction Methods
B. Theoretical Methods
1. Quantum Mechanical Methods
2. Classic Potential Methods
3. Comparison of the Methods
III. Structure of Mineral Surfaces
A. Zincblende
B. Wurtzite
1. Wurtzite (100) Surface
2. Wurtzite (110) Surface
3. Wurtzite (0001) and (000) Surfaces
C. Rocksalt
D. Rutile Surfaces
1. Rutile (110) Surfaces
2. Rutile (100) Surfaces
3. Rutile (111) Surfaces
E. Perovskite Surfaces
F. Corundurn Surfaces
G. Silica Surfaces
H. Carbonate Surfaces
IV. Discussion
A. Autocompensation and the Rehybridization of Dangling Bond Charge Density
B. Surface Stress and the Importance of Surface Topology
C. Ionic vs. Covalent Bonding: Implications for Surface Structure
D. Areas for Future Research
Acknowledgments
References
A. Purpose and Scope of Chapter
There is no more fundamental property of a surface or an interface than its atomic geometry.1,2 The physical arrangement of the atoms controls every aspect of the physics and chemistry of the interface. The goal of this chapter is to review the current state of the art in experimental and theoretical determinations of surface and interface geometry. To a large extent we limit the scope of the discussion to mineral surfaces for which experimental determinations or theoretical predictions have been made, rather than make conjectures concerning the properties of surfaces which may be environmentally or geologically interesting, but about which nothing is really known. We will demonstrate, however, that the atomic structure of mineral surfaces can be qualitatively understood, and predicted, using a set of five simple principles. These principles, based upon fundamental chemistry and physics, have been derived from over 20 years of research on semiconductor surfaces and interfaces,1, 2, 3, 4 and have recently begun to be applied to mineral oxides.5 These principles offer the opportunity to make predictions for some mineral surfaces that have not, as yet, been studied experimentally.
The range of minerals for which detailed experimental or theoretical information is available, unfortunately, is limited.6 Experimental limitations derive mainly from the paucity of well-characterized single-crystal materials for study. In addition, the insulating nature of these materials limits the use of charged particle spectroscopies, such as low-energy electron diffraction (LEED),7 which can be used to provide quantitative information on surface structures. This lack of detailed experimental information has also limited the development of semiempirical and empirical quantum mechanical models; models which, because of the complexity associated with ab initio methods, have traditionally been the first to characterize surface relaxations and reconstructions.2 These difficulties will be explored in more detail in Section II.
In this chapter we will focus on the subset of mineral systems for which the most detailed understanding of surface atomic structures exists. The results of computational studies, at all levels of approximation, and experimental surface structure determinations by LEED,7 X-ray photoelectron diffraction (XPD),8 Auger photoelectron diffraction (APD),8 ion scattering,9 and scanning probe microscopies10 will be presented and discussed in the context of the surface structure principles described in detail later in the chapter.
The level of presentation in this chapter is such that readers familiar with fundamental concepts in solid state atomic structure and chemical bonding should have no difficulty. Good undergraduate-level texts include: Introduction to Solid State Physics by C. Kittel11 and Chemistry in Two Dimensions: Surfaces by Gabor A. Somorjai.12 More in-depth treatments of these topics can be found in several excellent graduate-level texts: Physics at Surfaces by Andrew Zangwill;13 Solids and Surfaces: A Chemistās View of Bonding in Extended Structures by Roald Hoffmann;14 Atomic and Electronic Structure of Surfaces: Theoretical Foundations by M. Lanoo and P. Friedel;15 and Electronic Structure and the Properties of Solids: The Physics of the Chemical Bond by W.A. Harrison.16
It is useful, at this time, to review some basic nomenclature that will be used throughout this chapter. A surface is said to be relaxed if it displays the same symmetry as the bulk material. Reconstructed surfaces, on the other hand, display a surface symmetry different from the bulk. Relaxed and reconstructed surfaces are designated with the label (n Ć m), where n and m represent the ratio of the reconstructed to unreconstructed surface translation vectors for the two directions parallel to the surface. Relaxed surfaces, therefore, carry the designation (1 Ć 1). In some cases an additional designation of p (for primitive) or c (for centered) is added to more fully reflect the symmetry of the surface unit cell. Finally, if the surface translation vectors for the reconstructed surface are rotated from the surface translation vectors of the unreconstructed (or truncated bulk) surface the designation RĪĀ° is added, where Ī is the angle of rotation.
When discussing the mechanisms and driving forces for surface relaxations and reconstructions, we will repeatedly refer to the surface dangling bond charge density. This is the excess charge density remaining at the surface in the ādanglingā bonds which were used to bind the surface atoms to their, now missing, neighbors in the bulk material. This dangling bond charge density is localized at the surface in surface states which can be primarily derived from the surface cations (cation-derived surface states) or surface anions (anion-derived surface states). As will be seen, this is an enormously useful concept, not only on semiconductors, but also in ionic materials.
C. Principles Governing Surface Structures
Although the set of mineral surfaces that have had their surface atomic and electronic structure examined in detail is small, there is much that can be learned from these studies that is generally applicable.5,6 As stated earlier, there exists a set of simple physical and chemical principles, derived from the study of semiconductor surfaces,1, 2, 3, 4 that can be used to qualitatively understand and predict the surface relaxations and reconstructions which occur at mineral surfaces. In this section each of these principles will be described.
1. Principle 1: Stable Surfaces Are Autocompensated
Typically, bulk materials are stable when the bonding orbitals (or bands) are fully occupied and the antibonding orbitals (or bands) are fully...