What is Amorphous silicon?

Silicon without long-range periodicity. a-Si:H, with its dangling bonds terminated by hydrogen, is used in thin-film transistors and thin-film solar cells.

Japanese name: アモルファスシリコン

Amorphous silicon appears in 2 practice questions across 1 topic, including Glasses and amorphous materials.

Field: Glasses and amorphous materials

Structural features

In amorphous silicon, almost every Si atom is covalently bonded to four Si atoms, just as in the crystal. However, the bond angles deviate from the crystalline 109.5° by about 10° on average and the bond lengths also vary slightly, so long-range periodicity is lost. Such a structure is represented by a continuous random network of tetrahedra linked with disorder. Because of strain in the network, some atoms cannot complete four bonds and dangling bonds remain. This disorder and these dangling bonds, absent in the crystal, are major factors that determine the electrical properties.

Density of states and electrical properties

In a crystal, the edges of the valence and conduction bands are sharp and there are no electronic states between them. In a-Si:H, network disorder creates tail states that extend exponentially from the band edges into the gap. Electrons in tail states are localized and hard to move, and the energy span between the boundaries with mobile states is called the mobility gap. The optical band gap is about 1.7–1.8 eV. Dangling bonds create levels near the middle of the gap. In 1975 Spear and LeComber showed that the conduction type of films in which defects had been reduced by hydrogen could be controlled by adding phosphorus or boron, opening the way to electronic devices.

Light-induced degradation and solar cell structure

a-Si:H absorbs visible light strongly, but its carriers can diffuse only a short distance, so solar cells use a pin structure in which an undoped i-layer a few hundred nm thick lies between the p- and n-layers, and the internal field extracts the carriers. Under continued illumination there is the Staebler–Wronski effect, in which dangling bonds increase and efficiency falls; efficiency drops early in use and then becomes nearly constant. The degradation recovers on heating to about 150 °C. The stabilized efficiency of single-junction cells is only about 10% even for the best laboratory cells, and improvements were sought with tandem cells stacked with microcrystalline silicon and other designs.

Roles in displays and crystalline silicon solar cells

Because a-Si:H can be deposited uniformly on large glass substrates by plasma CVD at about 300 °C or below, it has been used in the thin-film transistors (TFTs) that switch the pixels of liquid crystal displays. However, the field-effect mobility of its TFTs is small, about 0.5–1 cm²/(V·s), so for high resolution and for driving organic light-emitting diodes, oxide semiconductors (IGZO) at about 10 cm²/(V·s) and low-temperature polysilicon at around 100 cm²/(V·s) have come into use. In crystalline silicon solar cells, on the other hand, heterojunction cells in which thin a-Si:H layers passivate surface defects support high efficiencies.

Try it: questions about Amorphous silicon

Amorphous silicon for solar cells and thin-film transistors is made by plasma CVD from silane as a-Si:H containing about 10% hydrogen. What is the main reason?

  1. AHydrogen acts as an interstitial donor and makes the film n-type even without phosphorus doping
  2. BHydrogen promotes crystallization of Si and turns the film into high-mobility polycrystalline silicon, improving its properties
  3. CHydrogen terminates Si dangling bonds and so reduces defect levels in the band gap
Show answer and explanation

Answer: C. Hydrogen terminates Si dangling bonds and so reduces defect levels in the band gap

Amorphous silicon without hydrogen contains as many as about 10¹⁹–10²⁰ cm⁻³ dangling bonds without a partner because of network disorder, and the levels they create near the middle of the band gap trap electrons and holes. As a result, adding dopants barely moves the Fermi level, and carriers generated by light recombine quickly. When hydrogen terminates these with Si-H bonds, the defect density falls to about 10¹⁵–10¹⁶ cm⁻³, making n-type and p-type doping with phosphorus or boron possible. Hydrogen does not act as a donor, and under ordinary deposition conditions the film stays amorphous (silane strongly diluted with hydrogen can also give microcrystalline films, but that serves a different purpose). The Staebler–Wronski effect, in which prolonged illumination increases defects, is also related to this hydrogen.

Which correctly compares the properties of hydrogenated amorphous silicon (a-Si:H) and crystalline silicon?

  1. AIts disordered atomic arrangement means it absorbs almost no visible light, so it is used for transparent electrodes in place of ITO
  2. BIt absorbs visible light more strongly than crystalline silicon, so films under 1 µm suffice, but its electron mobility is 2 to 3 orders lower
  3. CIts electron mobility is higher than that of crystalline silicon, so it is widely used in place of it for transistors in high-speed microprocessors
Show answer and explanation

Answer: B. It absorbs visible light more strongly than crystalline silicon, so films under 1 µm suffice, but its electron mobility is 2 to 3 orders lower

Crystalline silicon is an indirect-gap semiconductor that absorbs the longer wavelengths of visible light weakly, so solar cells usually use substrates 100 µm thick or more. In a-Si:H there is no long-range periodicity, so the constraint of conserving momentum in a transition is relaxed and its visible absorption coefficient is larger than that of the crystal, letting a film 1 µm thick or less absorb the light. On the other hand, carriers are trapped in band-tail states created by the disordered network, so the electron mobility is about 1–10 cm²/(V·s), 2 to 3 orders of magnitude lower than the roughly 1400 cm²/(V·s) of crystalline silicon. It is therefore unsuited to high-speed logic circuits, but taking advantage of large-area deposition at about 300 °C, it has been used for pixel TFTs in liquid crystal displays and for thin-film solar cells. With a band gap of about 1.7 eV it absorbs visible light strongly and cannot be a transparent electrode.

FAQ

What is the difference between amorphous silicon and polycrystalline silicon?

Polycrystalline silicon is an aggregate of crystal grains, and within each grain the atoms are arranged periodically. Amorphous silicon has no crystal grains and lacks long-range periodicity throughout. Polycrystalline silicon has far higher mobility, while amorphous silicon has the advantage of being easy to deposit uniformly over large areas at low temperature.

What happens when amorphous silicon is heated?

Heating above a few hundred °C drives out hydrogen and increases defects, and holding it for long above about 600 °C, or melting it momentarily with a laser, crystallizes it into polycrystalline silicon. Low-temperature polysilicon TFTs are made by crystallizing an a-Si film on a glass substrate with an excimer laser.

Where have amorphous silicon solar cells been used?

Because they generate power well even in weak light such as indoor fluorescent lighting, and can be made thin, cheaply and over large areas, they have been widely used to power small devices such as calculators and watches. For outdoor power generation, higher-efficiency crystalline silicon solar cells dominate.

Amorphous silicon in the explanations

Questions that mention Amorphous silicon (2)

  1. Amorphous silicon for solar cells and thin-film transistors is made by plasma CVD from silane as a-Si:H containing about 10% hydrogen. What is the main reason?
  2. Which correctly compares the properties of hydrogenated amorphous silicon (a-Si:H) and crystalline silicon?

Related terms

Indirect-bandgap semiconductor Chemically strengthened glass Metallic glass Network former Network modifier Non-bridging oxygen Reduced glass transition temperature Tempered glass

Back to the glossary