Amorphous Semiconductors: Beyond Traditional Glasses
Amorphous semiconductors, particularly those composed of Group IV elements like amorphous silicon (a-Si), amorphous germanium, and amorphous carbon, are often referred to as "tetrahedral glasses" in some literature. However, this classification is technically inaccurate. True glasses possess a metastable disordered structure and exhibit a glass transition. Amorphous semiconductors, while disordered, lack this crucial glass transition.
The formation of these materials differs significantly from conventional glasses. Instead of cooling a melt, amorphous semiconductors are typically prepared by vapor phase deposition (physical or chemical vapor deposition) directly from a gas phase, forming thin amorphous films. This process bypasses the glass transition step entirely. Furthermore, when heated, these amorphous semiconductors crystallize well before showing any signs of softening or glass transition, unlike traditional glasses which would revert to a supercooled liquid.
Structural Characteristics and Electronic Properties
The amorphous structure of these semiconductors, like silicon, retains a distorted version of its crystalline counterpart (diamond cubic). While the four-fold coordination of silicon atoms is largely maintained, bond angles are distorted, and local coordination can change. A key structural signature is the presence of "dangling bonds," where a silicon atom has an unpaired electron, making it three or four-coordinated with one lone electron.
These dangling bonds profoundly impact the electronic properties. In a conventional semiconductor, the band gap between the valence and conduction bands is empty. However, dangling bonds introduce "mid-gap defects" – states within the band gap. These mid-gap states act as recombination centers, quickly capturing excited electrons from the conduction band and recombining them with holes in the valence band. This phenomenon, known as Shockley-Read-Hall (SRH) recombination, renders pure amorphous silicon largely useless for electronics, as it rapidly annihilates any excited charge carriers.
Hydrogenation: A Game Changer for Amorphous Silicon
The utility of amorphous silicon was revolutionized by the discovery that introducing hydrogen can "passivate" these dangling bonds. Early methods of preparing amorphous silicon, such as evaporation, produced pure a-Si with its inherent electronic limitations.
Chemical vapor deposition (CVD) processes, particularly those using silane (SiH4) and hydrogen as a carrier gas, proved transformative. In a plasma environment, SiH4 breaks down into various radicals (SiH4, SiH2, SiH, and mobile hydrogen ions). These radicals then cluster on a substrate to form the amorphous film. Crucially, the resulting film incorporates a significant amount of hydrogen atoms, leading to "hydrogenated amorphous silicon" (a-Si:H).
Hydrogen, being a small and mobile atom with one excess electron, readily combines with the dangling bonds, forming stable Si-H bonds. This process effectively annihilates the mid-gap defect states, making a-Si:H a useful electronic material.
Applications in Macroelectronics
Hydrogenated amorphous silicon is now widely used in macroelectronic devices, which prioritize large area and cost-effectiveness over miniaturization and high speed. Examples include displays and solar cells.
Advantages of a-Si:H in these applications include:
- Large Area Deposition: Unlike crystalline silicon, which is limited by wafer size, a-Si:H can be deposited on vast glass sheets (meters in width), enabling efficient, high-throughput, and low-cost manufacturing.
- Substrate Versatility: Amorphous films do not require single-crystal substrates for epitaxial growth, allowing deposition on any material, including inexpensive glass, which is ideal for displays and solar cells.
While a-Si:H exhibits electronic properties comparable to crystalline semiconductors (e.g., ability to be doped p-type or n-type, forming p-n junctions), it also possesses unique characteristics that will be discussed in future lectures. The critical role of hydrogen in passivating defects is paramount.
Metallic Glasses: Overcoming Crystallization
Metallic glasses, or amorphous metals, differ significantly from network glasses in their bonding nature. Network glasses (like oxides or chalcogenides) form amorphous structures easily due to the directionality of covalent bonds, which creates open networks that can accommodate structural disorder. Metallic bonds, however, arise from interactions between outer shell electrons and atomic nuclei, favoring close-packed structures to lower energy. This is why most crystalline metals exhibit close-packed or near-close-packed arrangements.
Elemental metals, with their uniformly sized atoms, readily crystallize, making it challenging to form them in an amorphous state. For a long time, it was believed impossible. It wasn't until the 1960s that rapid quenching techniques for metal melts allowed the preparation of amorphous metals, initially as thin foils.
The breakthrough for "bulk metallic glasses" (BMGs) came from a Japanese scientist who developed compositions that could form millimeter or even centimeter-thick amorphous components. This required significantly enhancing the "glass-forming ability" – the material's resistance to crystallization.
Empirical Rules for Bulk Metallic Glass Formation
To maximize glass-forming ability, three empirical rules were proposed:
- Multi-component System: The system must consist of three or more elements.
- Significant Size Difference: There must be a substantial size difference between the constituent atoms.
- Negative Enthalpy of Mixing: The elements must have a negative enthalpy of mixing, indicating a thermodynamic preference to stay mixed rather than segregate.
The first two rules are topological in nature, focusing on atomic size rather than chemical reactivity. In metals, maximizing packing density is key to lowering energy. To avoid periodic close-packed structures, a hierarchical arrangement of atoms with different sizes is favored. Smaller atoms can fill the interstitial spaces between larger atoms, creating a densely packed yet disordered structure. This geometric problem, involving radius ratios and atomic fractions, is central to the stability of metallic glasses.
Properties of Amorphous Metals
The absence of lattice defects like dislocations and grain boundaries in amorphous metals leads to distinct properties compared to their crystalline counterparts:
- Chemical Resistance: Grain boundaries in polycrystalline metals are regions of higher structural disorder and energy, making them susceptible to chemical attack. Amorphous metals lack these weak links, exhibiting significantly higher chemical resistance.
- Mechanical Properties: In conventional metals, plastic deformation occurs through the movement of dislocations. Amorphous metals, lacking dislocations, deform differently. They possess a much higher elastic limit and yield strength (often an order of magnitude higher) than crystalline metals. This means they can withstand greater stress before undergoing plastic deformation. While they do yield under sufficient stress, they do so catastrophically through "strain localization" rather than gradual plastic flow.
The processing benefits and unique properties of amorphous metals, particularly BMGs, have led to interest in applications requiring high mechanical strength, such as components for electronic devices.
Other Amorphous Materials and Compositional Conventions
Beyond the discussed categories, a vast array of amorphous materials exists:
- Phosphate Glasses: Phosphorus is the glass-forming element. These are useful in bioactive glasses due to their biocompatibility and dissolution in body fluids.
- Heavy Metal Oxides: Compounds of oxygen with elements like tellurium or transition metals. Some form glasses, while others (e.g., amorphous titanium dioxide) are amorphous but not considered true glasses because they crystallize before glass transition.
- Halide Glasses and Alloys: Similar to chalcogenide and oxide glasses, but with halides (e.g., fluorine, chlorine) as the other elements. Often used for infrared-transmitting fibers (e.g., ZBLAN).
- Amorphous Minerals and Biominerals: Examples include opal and components of biological structures. Lobster shells, for instance, utilize amorphous calcium carbonate in a hierarchical structure to achieve both flexibility and strength.
Representing Glass Compositions
Conventions exist for labeling glass compositions:
- Oxide Glasses: Network modifiers are listed in increasing valence order, followed by the glass network former. For example, K2O-CaO-SiO2 (potassium, calcium, then silicon dioxide). Compositions can be expressed as molar fraction or by weight.
- Metallic Glasses: Elements are listed in decreasing order of amount. Vitreloy (Zr-Ti-Cu-Ni-Be) is a common example.
These are customary conventions, and exceptions can occur.
Glass Formation Theories: Kinetics and Topology
The formation of glass involves cooling a melt rapidly enough to bypass crystallization, leading to a supercooled liquid that then undergoes glass transition to form a solid glass. Three key questions arise:
- How fast is "fast enough" to avoid crystallization? This is addressed by the kinetic theory of glass formation.
- What happens during glass transition? This involves understanding the "laboratory definition" of glass transition.
- What is the structural origin of glass network stability? This is explained by structural theories, particularly the topological constraint theory.
Kinetic Theory of Glass Formation
Crystallization from a supercooled liquid requires both nucleation and growth.
- Nucleation: The formation of a stable crystalline nucleus. The thermodynamic driving force for nucleation (ΔG) increases with the degree of supercooling (melting point minus crystallization temperature). However, an energy barrier exists due to the interface energy between the solid nucleus and the liquid. This leads to an optimal temperature for nucleation, where there's a balance between sufficient thermodynamic driving force and atomic mobility.
- Growth: The expansion of the crystalline nucleus. Similar to nucleation, growth also exhibits an optimal temperature due to the interplay of thermodynamic and kinetic factors.
Since both nucleation and growth rates peak at optimal temperatures, the overall crystallization rate also peaks at a specific temperature. Glass scientists aim to avoid this temperature to prevent vitrification.
Time-Temperature-Transformation (TTT) Diagrams
TTT diagrams characterize the time required for crystallization under isothermal conditions. They typically show a "nose" shape, representing the fastest crystallization rate at a specific temperature. To form a glass, the cooling process must completely bypass this crystallization region.
The critical cooling line is tangential to the nose of the TTT diagram. Its slope defines the critical cooling rate (Rc), the minimum cooling rate needed to avoid crystallization. A smaller Rc indicates a more stable glass. For example, silica has an Rc of 10^-5 K/s (very stable), while elemental metals can have Rc values of billions of K/s (very unstable). Bulk metallic glasses like Vitreloy have an Rc of about 1 K/s, enabling their bulk formation.
The critical cooling rate dictates the maximum thickness (dmax) of a sample that can be prepared in an amorphous state:
dmax = α * ΔT / Rc
where α is thermal diffusivity and ΔT is the temperature difference between the liquidus and lowest temperature. Thicker samples cool slower internally, making it harder to bypass crystallization. This explains why early metallic glasses were thin foils, while BMGs can be millimeters or centimeters thick.
Different glass preparation techniques offer varying cooling rates: - Air quench: Slow - Liquid quench: Faster - Droplet spray/Melt spinning: Much faster (used for metallic glasses) - Thin film vapor deposition: Extremely fast (enables amorphous forms of materials like a-Si or a-TiO2 that cannot be made in bulk)
Laboratory Definition of Glass Transition
Glass transition is a complex phenomenon where a supercooled liquid transforms into a glass. Its properties are thermal history dependent; different cooling rates lead to different glass structures, volumes, and enthalpies.
The concept of a "potential energy landscape" helps visualize this. The landscape represents the system's potential energy as a function of its atomic configuration (3N coordinates for N atoms). It features a global minimum (crystalline state) and numerous local minima (laboratory glass states).
The "laboratory definition" of glass transition is based on statistical mechanics: - Liquid is ergodic: At high temperatures, atoms have enough energy to explore all possible microscopic states (local minima). - Glass is non-ergodic: At low temperatures, the system gets trapped in one or a few local minima, lacking the energy to transition between them.
Glass transition is an "ergodicity breakdown." Slower cooling allows the system more time to settle into deeper local minima, resulting in lower potential energy and enthalpy. Faster cooling traps the system in higher-energy states.
The transition between microstates is characterized by the "intervalley transition time" (τ), which is thermally activated. The criterion for a glass is when the observation time (t_observation) is less than τ. This implies that whether a system is considered a glass depends on the duration of the experiment. This also explains the inherent stochastic nature and slight property variations observed even in macroscopically similar glass samples.
Topological Constraint Theory
This theory explains glass stability by analyzing the network topology. It classifies networks into three types based on the number of constraints versus degrees of freedom:
- Flexible (Floppy/Under-constrained): Number of constraints < Number of degrees of freedom.
- Stressed Rigid (Over-constrained): Number of constraints > Number of degrees of freedom.
- Isostatic (Critically Constrained): Number of constraints = Number of degrees of freedom.
Constraints include bond-stretching (fixed bond length) and bond-bending (fixed bond angle). Degrees of freedom are 3N for N atoms.
- Flexible Networks: Have shallow energy barriers between local minima in the potential energy landscape. This allows the system to easily relax towards the global minimum (crystallization), making them kinetically unstable.
- Stressed Rigid Networks: Characterized by deep global minima, indicating a large thermodynamic driving force towards crystallization. Any deviation from the ideal crystal structure incurs a high energy penalty. Even small crystalline-like clusters can easily stabilize and propagate, leading to poor amorphous stability.
The optimal glass stability is achieved with an isostatic network, which strikes a balance between kinetic and thermodynamic factors. It provides a sufficiently high energy barrier to prevent crystallization while avoiding an excessively large thermodynamic driving force.
Zachariasen's rules, though empirical and applicable only to oxide glasses, foreshadowed this theory by focusing on topological features like coordination numbers and polyhedral sharing. These rules emphasize small coordination numbers (typical of covalent bonds), corner-sharing polyhedra (maximizing flexibility), and the formation of a 3D network (at least three shared corners per polyhedron). These rules highlight that the valent state and connectivity, rather than chemical reactivity, are crucial for glass formation.
The number of constraints primarily depends on the valent state of the constituent atoms. While cooling rate can subtly influence some constraint relaxation, it generally does not alter the fundamental topological architecture to the first order. Doping, by changing the valent state, can directly impact network topology and thus glass stability.
The next lecture will delve into the quantitative derivation of constraints and the conditions for achieving an isostatic network to maximize glass stability.
Takeaways
- Amorphous semiconductors like a‑Si are disordered but lack a true glass transition because they are made by vapor deposition and crystallize before softening.
- Dangling bonds in a‑Si create mid‑gap defect states that cause rapid Shockley‑Read‑Hall recombination, making pure a‑Si unsuitable for electronic devices.
- Incorporating hydrogen during CVD passivates dangling bonds by forming Si‑H bonds, eliminating mid‑gap states and enabling the widely used hydrogenated amorphous silicon (a‑Si:H) for displays and solar cells.
- Bulk metallic glasses require multi‑component alloys with large atomic size mismatches and negative mixing enthalpy, which together raise the glass‑forming ability and lower the critical cooling rate.
- Glass stability is maximized in isostatic networks where the number of constraints equals degrees of freedom, balancing kinetic and thermodynamic factors and preventing crystallization.
Frequently Asked Questions
Why does hydrogen passivation improve the electronic performance of amorphous silicon?
Hydrogen atoms bond to dangling silicon bonds, converting them into stable Si‑H bonds and removing mid‑gap defect states. This eliminates Shockley‑Read‑Hall recombination centers, dramatically increases carrier lifetime, and allows a‑Si:H to be doped and used in large‑area devices such as solar cells and displays.
What are the three empirical rules for bulk metallic glass formation?
The three empirical rules for bulk metallic glass formation are a multi‑component composition, a significant atomic size difference, and a negative enthalpy of mixing. These criteria promote dense, disordered packing that hinders crystallization, thereby increasing glass‑forming ability and lowering the critical cooling rate needed to produce bulk amorphous parts.
Who is MIT OpenCourseWare on YouTube?
MIT OpenCourseWare is a YouTube channel that publishes videos on a range of topics. Browse more summaries from this channel below.
Does this page include the full transcript of the video?
Yes, the full transcript for this video is available on this page. Click 'Show transcript' in the sidebar to read it.
Helpful resources related to this video
If you want to practice or explore the concepts discussed in the video, these commonly used tools may help.
Links may be affiliate links. We only include resources that are genuinely relevant to the topic.