2.4 Carrier drift and mobility

Carrier drift fundamentals

When you apply an electric field across a semiconductor, charge carriers (electrons and holes) start moving in a directed way. This directed motion is called carrier drift, and it's one of the two main transport mechanisms in semiconductors (the other being diffusion). Getting a solid handle on drift and mobility is essential for understanding how transistors, diodes, and sensors actually work.

Drift velocity definition

Drift velocity ($v_d$) is the average net velocity that charge carriers pick up when an electric field pushes them through the crystal. Without a field, carriers bounce around randomly due to thermal energy and their average net displacement is zero. The field adds a small directed component on top of that random motion.

The relationship is straightforward:

$v_d = \mu E$

where $\mu$ is the carrier mobility and $E$ is the applied electric field strength. The actual drift velocity depends on whether you're dealing with electrons or holes, the material itself, and how strong the field is.

Electric field influence

The applied electric field $E$ is what drives drift. Electrons (negative charge) accelerate opposite to the field direction, while holes (positive charge) accelerate in the same direction as the field.

Increasing the field strength increases drift velocity, but only up to a point. At very high fields, carriers reach a saturation velocity because scattering events become so frequent that additional field strength can't speed them up further. In silicon, this saturation velocity is roughly $10^7$ cm/s.

Carrier mobility concept

Carrier mobility ($\mu$) quantifies how easily a charge carrier moves through a semiconductor under an electric field. It's defined as:

$\mu = \frac{v_d}{E}$

The units are $\text{cm}^2/\text{V·s}$. Higher mobility means carriers respond more readily to the field, which translates directly into better electrical conductivity and faster device operation.

Drift current

Drift current is the electric current that results from carrier drift. It's the primary current mechanism in most semiconductor devices when a voltage is applied.

Current density equation

The drift current density $J$ (charge flowing per unit area per unit time) is:

$J = qn\mu E$

where $q$ is the elementary charge ($1.6 \times 10^{-19}$ C), $n$ is the carrier concentration, and $\mu$ is the mobility. This follows directly from $J = qnv_d$ and substituting $v_d = \mu E$.

For electrons, the conventional current direction is opposite to their motion (since they carry negative charge). For holes, the current flows in the same direction as their drift.

Electron vs hole current

Both electrons in the conduction band and holes in the valence band contribute to drift current. The total drift current density is the sum of both contributions:

$J_{total} = J_n + J_p = q(n\mu_n + p\mu_p)E$

where $n$ and $p$ are electron and hole concentrations, and $\mu_n$ and $\mu_p$ are their respective mobilities. Even though electrons and holes drift in opposite directions, their currents add together because they carry opposite charges.

Conductivity and resistivity

Conductivity ($\sigma$) measures how well a material conducts current. From the drift current equation, you can see that:

$\sigma = q(n\mu_n + p\mu_p)$

so that $J = \sigma E$. Conductivity depends on both how many carriers you have and how easily they move.

Resistivity ($\rho$) is simply the inverse:

$\rho = \frac{1}{\sigma}$

This is the quantity you'd use to calculate resistance for a given device geometry.

Carrier mobility

Mobility is one of the most important material parameters in semiconductor physics. It determines conductivity, device speed, and ultimately which material you'd choose for a given application.

Mobility dependencies

Three main factors control mobility:

• Effective mass: Determined by the band structure. Lighter effective mass means higher mobility, since carriers accelerate more easily.
• Scattering mechanisms: Collisions with the lattice (phonons) and with ionized impurities slow carriers down and reduce mobility.
• Temperature: Affects both the intensity of lattice vibrations and the thermal velocity of carriers, changing how often scattering occurs.

Lattice scattering effects

Lattice scattering (also called phonon scattering) happens when carriers interact with the thermal vibrations of atoms in the crystal. As temperature rises, atoms vibrate more vigorously, so carriers scatter more frequently and mobility drops.

This is the dominant scattering mechanism at high temperatures and in lightly doped (intrinsic) semiconductors where there aren't many impurities to scatter off of.

Impurity scattering impact

Ionized dopant atoms and crystal defects create local disturbances in the electric potential that deflect passing carriers. The more ionized impurities present, the more scattering occurs and the lower the mobility.

Impurity scattering is most significant at low temperatures (where phonon scattering is weak) and in heavily doped semiconductors. This is why you can't just keep adding dopants to improve conductivity indefinitely; at some point the mobility drop offsets the gain in carrier concentration.

Temperature effects on mobility

Temperature is one of the biggest knobs controlling mobility. The dominant scattering mechanism shifts depending on the temperature range, creating distinct regimes of behavior.

Phonon scattering mechanisms

At elevated temperatures, phonon scattering dominates. Two types matter:

• Acoustic phonon scattering: Carriers interact with low-energy lattice vibrations. This gives a mobility dependence of roughly $\mu \propto T^{-3/2}$.
• Optical phonon scattering: Carriers interact with higher-energy phonons. This becomes important at higher temperatures and is especially significant in polar semiconductors like GaAs, where the lattice has ionic character.

High vs low temperature behavior

• High temperature: Phonon scattering dominates. Mobility decreases as temperature increases.
• Low temperature: Impurity scattering dominates (in doped materials). Mobility actually increases with temperature in this regime, because faster-moving carriers are harder for ionized impurities to deflect.
• Intermediate temperature: Both mechanisms compete. The overall mobility curve typically shows a peak somewhere in this range.

This is why mobility vs. temperature plots for doped semiconductors have a characteristic inverted-V shape.

Intrinsic vs extrinsic semiconductors

In intrinsic semiconductors, there are very few ionized impurities, so phonon scattering controls mobility across most of the temperature range. The $T^{-3/2}$ dependence holds well at moderate to high temperatures.

In extrinsic (doped) semiconductors, impurity scattering adds a second mechanism. At low temperatures, impurity scattering limits mobility. At high temperatures, phonon scattering takes over. The crossover temperature depends on the doping concentration: heavier doping pushes the crossover higher.

Mobility in different semiconductors

Different semiconductor materials have very different mobilities because of differences in band structure, effective mass, and dominant scattering mechanisms. The table below summarizes room-temperature values for common materials:

Silicon and germanium mobility

Silicon is the workhorse of the electronics industry, with electron mobility around 1400 cm²/V·s and hole mobility around 450 cm²/V·s at room temperature. Germanium has notably higher mobilities (3900 and 1900 cm²/V·s for electrons and holes), which is why Ge channels are being explored for high-performance transistors.

III-V semiconductor mobility

III-V compounds like GaAs and InP have very high electron mobilities due to their low electron effective masses. GaAs at ~8500 cm²/V·s is well suited for high-frequency and optoelectronic applications. InP pushes even higher at ~12000 cm²/V·s, making it the material of choice for high-speed electronic and photonic devices. Note that their hole mobilities are generally much lower than their electron mobilities.

Wide bandgap material mobility

Wide bandgap semiconductors like GaN (~1000 cm²/V·s) and SiC (~700 cm²/V·s) have lower electron mobilities than Si or III-V materials. However, their real advantages lie elsewhere: high breakdown electric fields, excellent thermal conductivity, and the ability to operate at high temperatures. These properties make them ideal for power electronics and harsh-environment applications, where raw mobility isn't the primary concern.

Mobility engineering techniques

In practice, engineers don't just accept the intrinsic mobility of a material. Several techniques can boost mobility to improve device performance.

Doping concentration optimization

There's a trade-off with doping. Too little doping means few carriers (low conductivity despite high mobility). Too much doping means heavy impurity scattering crushes mobility. The optimal doping concentration balances these effects to maximize conductivity, $\sigma = qn\mu$, for a given application.

Strain-induced mobility enhancement

Applying mechanical strain to a semiconductor crystal changes the band structure and can reduce the effective mass of carriers. In modern MOSFETs:

• Tensile strain in Si enhances electron mobility
• Compressive strain enhances hole mobility

This is one of the most widely used techniques in modern CMOS manufacturing. Strained silicon channels have been standard in high-performance logic transistors since the 90 nm technology node.

Heterostructure design for mobility

Heterostructures stack layers of different semiconductor materials to confine carriers in a high-mobility region while keeping dopants in a separate layer. This technique, called modulation doping, spatially separates carriers from the ionized impurities that would scatter them.

The result is dramatically higher mobility. AlGaAs/GaAs and InAlAs/InGaAs heterostructures are the basis for high-electron-mobility transistors (HEMTs), which achieve some of the highest operating frequencies of any transistor type.

Carrier mobility characterization

Measuring mobility accurately is critical for material development and device optimization. Three main techniques are used.

Hall effect measurements

The Hall effect is the most common method for measuring mobility. Here's how it works:

1. Pass a known current through the semiconductor sample.
2. Apply a magnetic field perpendicular to the current flow.
3. The magnetic force deflects carriers to one side, building up a transverse voltage called the Hall voltage.
4. From the Hall voltage, sample dimensions, current, and magnetic field, extract the carrier concentration and mobility.

Hall measurements also reveal the conductivity type (n or p) based on the sign of the Hall voltage.

Field-effect mobility extraction

For transistors (especially MOSFETs), mobility can be extracted directly from device electrical characteristics. You measure the transconductance, which is the change in drain current per unit change in gate voltage ($g_m = dI_D/dV_G$). Combined with device geometry and capacitance, this yields the field-effect mobility of carriers in the channel.

This method gives mobility values specific to the device operating conditions, which can differ from bulk Hall mobility due to surface effects and the vertical electric field in the channel.

Mobility spectrum analysis techniques

Mobility spectrum analysis (MSA) is used when multiple carrier species are present (for example, electrons in different valleys, or both electrons and holes contributing simultaneously). The technique involves:

1. Measuring resistivity and Hall coefficient over a range of magnetic field strengths.
2. Analyzing the magnetic-field dependence to decompose the signal into contributions from each carrier species.
3. Extracting the mobility and concentration of each species individually.

MSA is particularly valuable for complex material systems and heterostructures where simple Hall measurements can't separate overlapping carrier populations.

Mobility models and simulations

Predicting mobility from theory is essential for device simulation and design. Models range from simple analytical formulas to full numerical simulations.

Drude model of carrier transport

The Drude model treats carriers as classical particles that accelerate under the electric field and periodically scatter, losing their directed momentum. The key parameter is the mean free time between collisions ($\tau$), which gives:

$\mu = \frac{q\tau}{m^*}$

where $m^*$ is the effective mass. This simple model captures the basic physics: mobility increases with longer scattering times and lighter effective masses. It's a useful starting point, though it doesn't account for the quantum mechanical details of real band structures.

Monte Carlo simulation of mobility

Monte Carlo methods simulate carrier transport by tracking individual carriers as they drift and scatter through the crystal. The simulation:

1. Initializes a carrier with some energy and momentum.
2. Calculates the time until the next scattering event using probability distributions for each scattering mechanism.
3. Determines the type of scattering and the carrier's new energy/momentum after the event.
4. Repeats for many carriers and averages the results.

This approach can handle complex band structures, multiple scattering mechanisms, and high-field effects that analytical models struggle with.

Hydrodynamic transport models

Hydrodynamic models treat the carrier population as a fluid characterized by local velocity, temperature, and density. They solve conservation equations for carrier number, momentum, and energy simultaneously.

These models are especially useful for high-field phenomena like velocity overshoot (where carriers temporarily exceed the steady-state drift velocity) and hot carrier effects (where carriers gain enough energy to cause reliability problems). They're computationally cheaper than Monte Carlo but more accurate than simple drift-diffusion models for short-channel devices.

Applications of drift and mobility

High-frequency transistor design

High mobility translates directly to faster carrier transit through a device, enabling higher operating frequencies. GaAs HEMTs are used in wireless communication systems, while InP HEMTs push into millimeter-wave and terahertz frequency ranges. The key figure of merit is the transit frequency $f_T$, which scales with carrier velocity (and therefore mobility) in the channel.

Power semiconductor devices

In power devices like power MOSFETs and IGBTs, carrier mobility determines the on-state resistance. Lower on-resistance means lower conduction losses and less heat generation. Wide bandgap materials (GaN, SiC) are increasingly used here: even though their mobilities are lower than silicon, their much higher breakdown fields allow thinner drift regions, which more than compensates for the mobility difference.

Optoelectronic device performance

Drift and mobility matter for optoelectronics in several ways:

• Photodetectors: High mobility enables fast collection of photogenerated carriers, giving shorter response times and higher bandwidth.
• Solar cells: Good mobility reduces the chance that carriers recombine before reaching the contacts, improving energy conversion efficiency.
• LEDs: Higher mobility improves current spreading across the device area, leading to more uniform light emission and better extraction efficiency.