Unit 1 - Practice Quiz

Semiconductors are a class of materials with electrical conductivity values falling between those of highly conductive metals and non-conductive insulators.

Silicon is the cornerstone of the modern electronics industry due to its abundance, stability, and well-understood properties.

Increasing the temperature of an intrinsic semiconductor provides more thermal energy, which excites more electrons from the valence band to the conduction band, increasing the number of free charge carriers and thus increasing conductivity.

Resistivity () is a measure of a material's intrinsic ability to oppose the flow of electric current. It is the reciprocal of conductivity, related by the formula .

In direct bandgap semiconductors, an electron can drop from the conduction band to the valence band and release its energy directly as a photon (light), which is a very efficient process for light emission.

In indirect bandgap semiconductors, a change in momentum is required for an electron to transition. This change in momentum is accommodated by the emission or absorption of a phonon, which represents a quantum of lattice vibration.

Silicon is the second most abundant element in the Earth's crust and is primarily extracted from sand or quartz, which are both forms of silicon dioxide ().

Metallurgical Grade Silicon (MGS) is the first product in the silicon purification process. It is about 98-99% pure and serves as the feedstock for producing higher purity silicon.

A single-crystalline material has a periodic and repeating crystal lattice structure throughout its entire volume, without any interruptions or changes in orientation.

Polycrystalline materials consist of numerous small individual crystals (grains) that are separated by grain boundaries. The orientation of the crystal lattice is different in each grain.

Electronic Grade Silicon (EGS) must be extraordinarily pure, often referred to as "nine-nines" (9N) pure or better, to ensure the precise electrical properties needed for semiconductor devices.

The Siemens process refines silicon-containing gases (like trichlorosilane) to deposit ultra-pure polycrystalline silicon onto slim silicon rods, which then serve as the raw material for single-crystal growth.

The Czochralski method involves dipping a small seed crystal into a crucible of molten silicon and then slowly withdrawing it while rotating to form a large, cylindrical single-crystal ingot.

At the high temperatures required to melt silicon, the molten silicon slowly dissolves the inner walls of the quartz () crucible, introducing oxygen atoms into the crystal.

The Float Zone method does not use a crucible to contain the melt, which eliminates the primary source of oxygen contamination seen in the CZ method, resulting in higher purity silicon crystals.

In the FZ process, a polycrystalline rod is held vertically, and a radio frequency (RF) coil creates a narrow molten zone. This zone is moved along the rod, melting and re-solidifying the silicon to form a single crystal.

The Bridgman method is particularly well-suited for growing single crystals of compound semiconductors, such as GaAs or InP, which can be challenging to grow with other methods.

The large, continuous single-crystal cylinder grown by methods like Czochralski or Float Zone is called an ingot or a boule before it is sliced into wafers.

The first step in wafer preparation is to slice the cylindrical ingot into many thin, circular discs using a high-precision saw, typically a diamond-edged blade or wire saw.

Flats or notches are ground onto the side of the ingot to provide a reference for the crystal's specific atomic orientation (e.g., <100> or <111>) and conductivity type (p-type or n-type), which is critical for subsequent fabrication steps.

During the CZ process, the hot molten silicon reacts with the quartz () crucible, causing some of the crucible to dissolve. This incorporates oxygen atoms into the silicon melt, which are then integrated into the growing crystal as interstitial impurities.

In a direct bandgap material like GaAs, an electron at the conduction band minimum can recombine with a hole at the valence band maximum while conserving momentum, releasing energy efficiently as a photon (light). In an indirect bandgap material like Si, this recombination requires the involvement of a phonon to conserve momentum, making the radiative process much less probable.

Grain boundaries are interfaces between different crystal orientations in a polycrystalline material. These interfaces contain defects and dangling bonds that trap charge carriers and act as scattering centers, which impedes the flow of electrons and holes, thereby reducing carrier mobility and overall device performance.

The Siemens process primarily uses the thermal decomposition and reduction of high-purity trichlorosilane () gas with hydrogen. The reaction, , deposits very pure polycrystalline silicon onto thin, heated silicon filaments.

CMP is the final and most crucial polishing step. It uses a chemical slurry and mechanical abrasion to produce a highly planar (flat) surface across the entire wafer with an extremely smooth, damage-free finish, which is essential for subsequent photolithography and device fabrication steps.

In the Float Zone method, a molten zone is passed through a vertical polysilicon rod without the need for a crucible to contain the melt. This eliminates the primary source of contamination found in the CZ method, which is the dissolution of the quartz () crucible into the melt, leading to significantly lower oxygen and carbon concentrations.

Phosphorus is a Group V element, acting as a donor impurity in silicon (Group IV). Each phosphorus atom donates one free electron to the conduction band, significantly increasing the electron concentration (). Since conductivity is given by , the increase in leads to a large increase in conductivity.

The diameter of the growing ingot is a sensitive function of the temperature at the solid-liquid interface. This temperature is controlled by carefully adjusting the heater power and the pull rate. A faster pull rate tends to decrease the diameter, while a slower pull rate increases it. Rotation is used to create a uniform temperature and dopant distribution.

In an electric arc furnace at temperatures around 2000°C, quartzite (a form of ) is reacted with carbon sources like coal or wood chips. The carbon reduces the silicon dioxide according to the overall reaction: . The resulting silicon is about 98-99% pure and is called Metallurgical Grade Silicon.

Flats were used on smaller wafers (<200 mm) to indicate crystal orientation and dopant type. For larger wafers (200 mm and 300 mm), flats were replaced by a small notch. This change was made to reduce wasted wafer area and to improve the mechanical integrity and balance of the wafer during high-speed rotation in processing equipment.

The Bridgman method involves melting the material in a sealed ampoule and slowly passing it through a temperature gradient. This sealed environment is excellent for growing compound semiconductors like GaAs because it helps maintain the stoichiometry by preventing the more volatile element (like Arsenic) from evaporating, which is a major challenge in open systems like the CZ method.

In an indirect bandgap semiconductor, the conduction band minimum and the valence band maximum are at different values of crystal momentum (). For an electron to transition between these points, it must change both energy and momentum. The energy is released (often as heat, or sometimes a photon), and the momentum change is facilitated by the emission or absorption of a phonon.

Resistivity is the inverse of conductivity (). For an intrinsic semiconductor, . As temperature rises, thermal energy creates more electron-hole pairs, causing the intrinsic carrier concentration () to increase exponentially. This dramatic increase in far outweighs the slight decrease in mobility due to increased lattice scattering, causing the overall conductivity to rise and resistivity to fall.

The transistor is built on a single-crystal silicon substrate, where the active channel is formed. A thin layer of silicon dioxide acts as a high-quality insulator (gate dielectric). A layer of heavily doped polycrystalline silicon is then deposited on top of the oxide to serve as the gate electrode, which controls the flow of current in the channel.

MGS (98-99% pure) contains many impurities like iron, aluminum, and boron. By converting it to a liquid at room temperature, such as (boiling point ~31.8°C), it can be purified using fractional distillation. This process effectively separates the from impurities that have different boiling points, which is a key step in achieving the high purity required for EGS.

A segregation coefficient of 0.8 means that for every 10 dopant atoms in the melt at the interface, only 8 are incorporated into the solid crystal. The remaining 2 are rejected back into the melt. As the crystal grows, this continuous rejection of the dopant causes its concentration in the remaining liquid to gradually increase.

The FZ method relies on the surface tension of the molten silicon to hold the zone together. This becomes mechanically unstable for large diameters. Therefore, it is very difficult and costly to grow uniform, defect-free FZ crystals with the large diameters (300mm) required for modern, high-volume manufacturing, for which the CZ method is better suited.

Heavily doped polycrystalline silicon (poly-Si) is widely used as the gate electrode in MOS transistors because it is stable at high processing temperatures and has a suitable work function. It is also used as the active channel material in thin-film transistors (TFTs), which are commonly found in LCD displays.

For an n-type semiconductor, the conductivity , where is the electron concentration, which equals the donor concentration. First, calculate conductivity: . Resistivity is the inverse of conductivity: .

Slicing and lapping are abrasive mechanical processes that create a layer of microscopic cracks and defects on the wafer's surface and subsurface. Chemical etching, typically using a mixture of nitric and hydrofluoric acids, is performed to isotropically remove this damaged layer, relieving stress and improving the crystalline quality of the surface before polishing.

As the crystal grows, the dopant (with ) is rejected, increasing its concentration in the remaining melt. The Burton-Prim-Slichter model shows that the effective segregation coefficient, , increases with the pull rate. By carefully programming an increase in the pull rate during growth, more dopant can be incorporated into the solid, compensating for its rising concentration in the melt and helping to maintain a more uniform resistivity along the ingot's length.

The purification of SiHCl₃ before it enters the Siemens reactor relies heavily on fractional distillation. The effectiveness of distillation depends on the difference in boiling points of the components. Because the boiling point of BCl₃ is very close to that of SiHCl₃, it is extremely difficult to separate them, allowing boron to be carried over into the reactor and contaminate the final EGS product.

Excess noise in an APD is dominated by the statistical nature of the impact ionization process. McIntyre's theory shows that noise is minimized when only one type of carrier (either electrons or holes) can cause impact ionization. This corresponds to an ionization coefficient ratio () that is either very large or very close to zero. Many direct bandgap III-V materials used for APDs, like InP/InGaAs systems, can be engineered to have a large difference between electron and hole ionization rates (a very small ), leading to lower excess noise compared to silicon, where the rates are more similar.

Modern high-yield IC manufacturing relies on intrinsic gettering. In CZ silicon, the high concentration of oxygen is used to intentionally create controlled oxygen precipitates in the wafer's bulk during heat treatment. These precipitates and the surrounding defects act as traps (gettering sites) for fast-diffusing metallic impurities (like Fe, Cu, Ni), pulling them away from the device-active region near the surface. FZ silicon's purity is a disadvantage here, as it lacks the oxygen needed for this powerful self-cleaning mechanism.

CMP is a synergistic process. The slurry chemically modifies (e.g., oxidizes) the surface, and the pad mechanically abrades this softer, modified layer. The Preston equation assumes the mechanical abrasion is the bottleneck. However, at very low velocities, the pad abrades the oxidized layer slower than the chemical slurry can reform it. The process is no longer limited by how fast you can abrade, but by how fast the surface can chemically react. Thus, the MRR plateaus and becomes independent of velocity, a clear deviation from the linear prediction of the Preston equation.

At 300 K, lattice scattering is significant. For Sample A (higher doping) to have the same total mobility as Sample B (lower doping), its impurity scattering contribution must be larger, and its lattice scattering contribution must be identical (as it only depends on T). This means at 300 K, is larger for A. At low temperatures like 77 K, lattice scattering becomes much weaker ( increases significantly), and ionized impurity scattering becomes the dominant mechanism. Since Sample A has a higher dopant concentration (), its ionized impurity scattering will be much stronger than Sample B's. Therefore, its total mobility, now dominated by , will be significantly lower than that of Sample B.

Initially, as doping () increases, more electrons are available to be trapped at the grain boundary, increasing the trapped charge and thus the barrier height. However, as the doping becomes very heavy, the width of the depletion region required to shield the trapped charge shrinks significantly. At extremely high doping levels, the Fermi level moves into the conduction band (degenerate semiconductor), and the depletion regions become so narrow that carriers can easily tunnel through the barrier, effectively reducing its impact and causing the measured barrier height to decrease.

In both the hydrochlorination of MGS and the deposition from SiHCl₃ in the Siemens reactor, a significant amount of silicon tetrachloride (SiCl₄) is produced as a byproduct (). Discarding SiCl₄ would mean wasting up to 75% of the silicon that entered the deposition reactor. By hydrogenating the SiCl₄ byproduct (), it is converted back into the desired trichlorosilane precursor. This recycling loop is crucial for making the process economically viable and resource-efficient.

Constitutional supercooling occurs when the actual temperature of the liquid ahead of the solid-liquid interface is lower than its local equilibrium freezing point. When a dopant with is rejected into the melt, it forms a solute-rich boundary layer. This layer has a lower freezing point than the bulk melt. If the thermal gradient () in the liquid is not steep enough to overcome this freezing point depression, a region of supercooled liquid forms, leading to interface instability and polycrystalline growth. The condition for stability is , showing the problem is worst for high concentration (), low , and low thermal gradient .

For lasing, a high population of electrons must be sustained at the conduction band minimum to recombine radiatively with holes. In Material B, even if electrons are excited to the higher-energy direct conduction valley, they will very quickly lose energy via phonon scattering and relax to the lower-energy indirect valley. Once in the indirect valley, they are 'stuck' because radiative recombination from there is a slow, inefficient, phonon-assisted process. Consequently, a population inversion cannot be achieved, and the material will not lase effectively.

The core advantage of FBR is its energy efficiency. The Siemens process requires heating large silicon rods to over 1000 °C, radiating a huge amount of heat. An FBR operates on a 'bed' of millions of tiny silicon particles, which collectively have an enormous surface area. This allows for a much more efficient transfer of heat and reactants, leading to higher deposition rates and significantly lower electricity consumption (a major cost in polysilicon production) compared to the Siemens process.

The ID saw is a very thin annular blade held under high tension to ensure rigidity. However, any fluctuation in this tension, mechanical vibration in the machine, or non-uniformity in the ingot can cause the blade to oscillate or wobble slightly during the cut. These microscopic deviations from a perfect plane translate directly into the parallel grooves known as saw marks on the wafer surface, which must be removed by subsequent lapping and polishing steps.

A simple transverse or vertical magnetic field effectively dampens melt convection, which is good for reducing oxygen incorporation. However, this can also lead to poor mixing of dopants near the growth interface, resulting in radial non-uniformity. A cusp field is a tailored magnetic field that is strong in the bulk of the melt (away from the crystal) to suppress the large-scale convection currents responsible for oxygen transport, but weak or zero at the center, directly under the crystal. This allows for controlled, rotation-driven convection in the critical region where dopant incorporation occurs, providing the 'best of both worlds': low oxygen and good dopant uniformity.

The Hall effect preferentially measures higher-mobility carriers, while drift mobility is an average over all carriers. The Hall scattering factor accounts for the fact that carrier scattering times (and thus mobilities) are dependent on carrier energy, and carriers have a thermal distribution of energies. If the scattering time is strongly energy-dependent (as is the case for acoustic phonon or ionized impurity scattering), the averaging process for Hall and drift mobilities yields different results, and will deviate from 1. For silicon, it is typically in the range of 1.1-1.2 at room temperature.

In XRD, a diffraction peak occurs only when crystal planes are oriented at the correct Bragg angle to the incident beam. In a single crystal (Sample A), only a few sets of planes (e.g., the {400} family for a (100) wafer) will be properly aligned to diffract, resulting in a few sharp peaks. In a polycrystalline sample with random grain orientations (Sample B), there will be a statistical distribution of grains satisfying the Bragg condition for every possible set of crystal planes ({111}, {220}, {311}, etc.), resulting in a pattern showing all the allowed diffraction peaks for that crystal structure, though each may be less intense.

Zone refining, the principle behind the FZ method, works by segregating impurities based on their segregation coefficient (). The effectiveness of removal is proportional to . Impurities with very small values, like iron, are very effectively swept to the end of the ingot. However, boron's segregation coefficient in silicon is about 0.8. Since this is very close to 1, there is very little segregation of boron between the solid and liquid phases. Therefore, zone refining is highly inefficient at removing boron, which must be eliminated during the chemical purification of the precursor gas instead.

At low temperatures, ionized impurity scattering is the dominant factor limiting mobility. The scattering rate is proportional to the total number of ionized centers, regardless of whether they are positive (donors) or negative (acceptors). In a compensated n-type material, the total concentration of ionized centers is . At low T, all donors and acceptors are ionized, so . This is greater than the scattering centers that would be present in an uncompensated sample with the same free electron concentration. Therefore, compensation always reduces mobility at low temperatures.

In automated wafer handling and alignment systems, optical or mechanical sensors determine the wafer's center and rotational orientation. Finding the precise midpoint of a long, straight flat is subject to more measurement error than finding the well-defined point of a V-notch. For the extremely tight alignment tolerances (overlay) required in modern lithography, where successive layers must be aligned to within a few nanometers, the superior positional accuracy provided by the notch is a critical enabling factor.

Physically moving the crucible, even very slowly, can introduce subtle vibrations and perturbations into the molten material. These disturbances can disrupt the delicate diffusion boundary layer at the solid-liquid interface and induce unwanted convective flows, which can lead to striations (dopant variations) and structural defects like dislocations. By keeping the crucible stationary and moving the temperature gradient electronically, the growth environment is mechanically quiescent, which is highly conducive to growing high-perfection crystals.

Directional solidification is essentially a large-scale application of the segregation principle used in crystal growth. A large crucible of molten MGS is slowly cooled from the bottom up. Impurities with a segregation coefficient () less than one are preferentially rejected by the solidifying silicon and are pushed upwards into the remaining liquid. The last part of the ingot to solidify is thus highly contaminated and can be physically cut off. The major limitation is the same as for zone refining: it is very ineffective for removing key dopant impurities like Boron () and Phosphorus (), which significantly impact the final electrical properties of the solar cell.