Electronic Transport in Amorphous Solids: Hopping, VRH and a‑Si
In amorphous solids, the absence of structural order leads to significant differences in electronic structure and transport properties compared to crystalline solids. Key distinctions include the presence of localized states and a continuously distributed density of states rather than a well-defined band gap. These localized states profoundly modify how charge carriers move through the material.
Conduction via Extended States
Conduction in amorphous solids can occur through both extended states (outside the mobility gap) and localized states (within the mobility gap). For extended states, the conductivity formalism is similar to crystalline solids, where conductivity scales with mobility and charge density. However, the mobility itself is significantly different due to two main factors:
- Trapped States: Amorphous solids contain numerous "trap states" within the mobility gap. Electrons excited into the conduction or valence band can get trapped in these localized states, temporarily ceasing to contribute to electrical conduction. They only resume conduction once thermally released back into the extended states. This trapping effect is accounted for by a factor (1 - f_trap), where f_trap is the fraction of time a carrier is trapped, effectively reducing mobility.
- Structural Disorder: Even the "free mobility" (mu_0, ignoring trapping) in amorphous solids is lower than in crystalline counterparts. This is because structural disorder causes much greater scattering of carriers.
Combined, these factors result in apparent mobility for extended states in amorphous materials being orders of magnitude lower than in crystalline solids. For example, electron drift mobility in amorphous silicon is typically a few cm²/Vs, compared to over 1,000 cm²/Vs in crystalline silicon. This low mobility makes amorphous solids unsuitable for high-speed electronic devices, though they are used in applications like displays and solar cells where gigahertz operation is not required.
The apparent mobility in amorphous solids exhibits an Arrhenius dependence, increasing with temperature. At higher temperatures, carriers have more energy and spend less time trapped in localized states. In the high-temperature limit, f_trap approaches zero, and the measured mobility approaches the free mobility, which is still significantly lower than in crystalline materials. Extended-state conduction is thus a thermally activated process due to detrapping.
Conduction via Localized States: Hopping
Unlike extended states where electrons roam freely, in localized states, electrons are confined to the vicinity of defects. For electrical conduction to occur, these electrons must "hop" from one localized state to another.
Fixed-Range Hopping (in Doped Crystalline Semiconductors)
This phenomenon can be observed even in crystalline semiconductors with dopants. At cryogenic temperatures, carriers can be frozen out onto donor atoms, creating spatially separated localized states. When an electric field is applied, electrons can hop between these dopant atoms via quantum mechanical tunneling.
The probability of tunneling depends on the wave function overlap between neighboring localized states. If the spacing between two states is R, the wave function overlap scales exponentially as exp(-2αR), where α gauges the spatial confinement of the wave function. This exponential decay means hopping is practically limited to nearest neighbors, a mechanism known as fixed-range hopping.
Variable Range Hopping (VRH) in Amorphous Solids
In amorphous semiconductors, localized states exist within the mobility gap. Hopping conduction between these states has some key differences:
- Fermi Level Proximity: Hopping conduction primarily involves localized states near the Fermi level. States far above the Fermi level are empty, lacking carriers, while states far below are filled, lacking empty states for hopping.
- Energy Difference and Phonons: Hopping can occur between states with different energies. To conserve energy, this process must involve a phonon:
- Hopping from high to low energy: phonon emission.
- Hopping from low to high energy: phonon absorption. This makes hopping a two-particle process. The probability of finding a phonon with the correct energy is approximated by a Boltzmann distribution, scaling as exp(-ΔE/kT), where ΔE is the energy difference.
- Optimal Hopping Distance: The total hopping probability depends on both wave function overlap (exp(-2αR)) and phonon availability (exp(-ΔE/kT)). Maximizing this probability involves a trade-off:
- Shorter R increases wave function overlap.
- Smaller ΔE increases phonon probability. In amorphous solids, a larger search distance (R) might lead to finding a state with a smaller energy difference (ΔE), even if it means a slightly reduced wave function overlap. This leads to an optimal hopping distance that is a function of temperature. This is why it's called Variable Range Hopping (VRH).
The optimal hopping distance (R) is found by minimizing the exponent of the hopping probability, which involves relating ΔE to R using the density of states (g) at the Fermi level. The relationship derived is R ∝ T^(-1/4). This means that as temperature decreases, the optimal hopping distance increases, allowing the system to find states with smaller energy differences.
Plugging this optimal R back into the hopping probability leads to the well-known Mott's T^(-1/4) law for conductivity: σ ∝ exp[-(T_0/T)^(1/4)]. This contrasts with the Arrhenius law (σ ∝ exp(-E_a/kT) or T^(-1)) for extended states.
Temperature Dependence of Conductivity
On an Arrhenius plot (log conductivity vs. inverse temperature), extended-state conduction appears as a straight line. VRH, however, shows a curved dependence, with a less steep slope at lower temperatures.
- High Temperature: Extended-state conduction dominates because there are plenty of free carriers and it's easier to move them in extended states than to force hopping.
- Low Temperature (Cryogenic): The number of free carriers in extended states decreases exponentially, making extended-state conduction negligible. In this regime, hopping becomes the dominant conduction mechanism.
This transition is typical for amorphous semiconductors: linear behavior at high temperatures (Arrhenius) and Mott's T^(-1/4) law at low temperatures. Experimental data, such as conductivity measurements in amorphous glass systems (e.g., arsenic, selenium, tellurium, copper), show excellent agreement with Mott's law at low temperatures, confirming the strong contribution of hopping.
Interestingly, at very low temperatures, amorphous silicon can sometimes exhibit higher electrical conductivity than crystalline silicon because crystalline semiconductors have very little contribution from hopping.
Electronic Transport Summary
- Ionic Conduction: Conductivity scales with 1/T times a Boltzmann factor. In amorphous systems, barrier heights and depths for ion hopping vary statistically, leading to slight deviations from ideal Arrhenius behavior.
- Electronic Transport:
- Electronic Structure: Amorphous materials lack crystal momentum (k) as a good quantum number, so band structure is represented by density of states, not dispersion diagrams.
- Localized States: Presence of localized states creates band tails and gap states, contributing to VRH conduction, which is absent in crystalline semiconductors.
Applications of Amorphous Silicon: Macroelectronics
Amorphous silicon (a-Si) is a crucial material for macroelectronic devices, which require large-area fabrication, unlike microelectronics that prioritize miniaturization (Moore's Law). Examples include displays and solar cells, where "bigger is better" and cost per area is a key metric.
Crystalline silicon is unsuitable for large-area applications due to limitations in wafer size (max 300mm) and the need for single-crystalline substrates. Amorphous silicon offers several advantages:
Large-Area, Low-Temperature, Monolithic Deposition:
- Large Area: Can be deposited over several square meters, enabling high throughput and low cost per area using Plasma-Enhanced Chemical Vapor Deposition (PECVD).
- Low Temperature: Deposition temperatures (e.g., 250-300°C) are low enough not to damage existing devices or substrates.
- Monolithic: Can be directly deposited onto target substrates without lattice matching constraints, simplifying processing and improving yield.
- Passivation: PECVD allows for simultaneous deposition of passivation layers (SiO2, SiN) in the same chamber, a crucial advantage for semiconductor performance.
Doping Capacity: Historically, amorphous silicon was considered useless because it couldn't be effectively doped due to "Fermi level pinning." This occurs because numerous dangling bonds (unsatisfied chemical bonds with a single electron) create mid-gap states. Dopant electrons would fill these mid-gap states instead of shifting the Fermi level, requiring impractically high doping concentrations (e.g., 10^20 cm^-
to saturate them.
The invention of PECVD changed this. PECVD uses silane (SiH4) and hydrogen gas, incorporating significant amounts of hydrogen (10-15%) into the a-Si film. These hydrogen atoms passivate dangling bonds, forming stable two-electron structures and eliminating electron defects. With reduced mid-gap states, amorphous silicon becomes amenable to doping, enabling the creation of PN junctions and modulation of carrier concentration.
CMOS Compatibility: As a silicon-based material, amorphous silicon is compatible with existing CMOS processes, allowing leverage of established infrastructure and technical expertise from the microelectronics industry.
The "essence of being human is that one does not seek perfection" applies to macroelectronics. Amorphous silicon, despite its defects compared to single-crystalline silicon, is "adequate" for these applications.
Structure and Defects of Amorphous Silicon
Crystalline silicon has a diamond cubic structure with six-membered rings. Amorphous silicon, however, has a random distribution of ring sizes (five-, six-, seven-membered rings), leading to structural disorder and dangling bonds. These dangling bonds are the dominant electron defects.
Dangling bonds create spatially localized defect states that are typically located near the center of the mobility gap. This high density of mid-gap states causes Fermi level pinning, preventing effective doping in unpassivated a-Si.
PECVD Deposition Process
PECVD uses silane (SiH4) and hydrogen plasma to create radicals (SiH3, SiH2, SiH, H). These radicals condense onto a substrate. The surface of the growing film is typically passivated by hydrogen. For new silicon atoms to attach, radicals must extract hydrogen from the surface, creating a dangling bond (active site). Other SiH3 radicals then diffuse to and attach to this site, continuing film growth.
The defect density in a-Si films is temperature-dependent, exhibiting an optimal intermediate temperature (250-300°C) for minimal defects. * Low Temperature: Ineffective surface diffusion of SiH3 leads to inadequate passivation and high defect density. * High Temperature: The endothermic extraction of hydrogen from the surface is favored, also leading to inadequate passivation due to chemical equilibrium shifts.
Electronic and Optical Properties of Amorphous Silicon
| Property | Amorphous Silicon (a-Si:H) | Crystalline Silicon (c-Si) | Implications
Takeaways
- In amorphous solids, lack of long‑range order creates localized trap states within the mobility gap, causing charge carriers to spend time immobilized and dramatically reducing apparent mobility compared with crystalline counterparts.
- Extended‑state conduction in amorphous materials follows an Arrhenius temperature dependence because thermal energy releases carriers from traps, but even the free mobility remains orders of magnitude lower than in crystals.
- Hopping conduction occurs via electrons moving between localized states, and the probability depends on both wave‑function overlap (exp(‑2αR)) and phonon assistance (exp(‑ΔE/kT)), leading to a trade‑off between hopping distance and energy difference.
- Variable‑range hopping (VRH) optimizes this trade‑off, giving an optimal hopping distance R ∝ T^(‑1/4) and resulting in Mott’s conductivity law σ ∝ exp[‑(T₀/T)^(1/4)], which dominates at low temperatures.
- Amorphous silicon’s large‑area, low‑temperature PECVD deposition and hydrogen passivation of dangling bonds enable effective doping and make it suitable for macroelectronic applications such as displays and solar cells despite its lower carrier mobility.
Frequently Asked Questions
Why does amorphous silicon exhibit lower electron mobility than crystalline silicon?
Amorphous silicon’s lower electron mobility stems from structural disorder and abundant trap states that capture carriers, reducing the time they contribute to conduction. The disorder increases scattering, and trapped carriers must be thermally released, so the apparent mobility is orders of magnitude smaller than the free mobility of crystalline silicon.
What is the mechanism behind Mott’s T‑1/4 law in variable‑range hopping?
Mott’s T‑1/4 law arises because in variable‑range hopping the optimal hopping distance grows as temperature falls, allowing electrons to find states with smaller energy differences even if they are farther apart. Minimizing the combined exponential penalties from wave‑function overlap and phonon availability yields a conductivity σ ∝ exp[‑(T₀/T)^(1/4)].
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.