The Bohr Model of the Atom

A Brief Description: A model proposed by Niels Bohr to support his hypothesis about electrons in a hydrogen atom. This is a model of the atomic structure in which electrons travel around the nucleus in well-defined orbits determined by quantum conditions. A transition from a higher orbit to a lower orbit will release quantized energies of light, which would explain the light spectrum emitted by an element. Model: A tentative description of a system or theory that accounts for all of its known properties. Atom: The smallest indivisible (by chemical reaction) particle of an element that can either exist alone or enter into chemical combination. An atom consists of a central positively charged nucleus surrounded by a sufficient number of negatively charged electrons to balance the nucleus. Emission Spectrum: The spectrum of frequencies of radiation (light) emitted by an atom or a molecule in a transition from a higher state to a lower energy state. Electronic state: A condition or mode of being with regard to a set of circumstances; position. Bohr, Niels: The Danish physicist that attempted to account for the emission spectrum of atomic hydrogen by proposing that the single electron in the hydrogen atom travels in definite circular orbits around the central nucleus and that only certain orbits were possible at certain energies. This is what we call the Bohr model. As we all know, from our chemistry book, each element in the periodic table has its own characteristic emission spectrum. The spectrum is a result of applying some kind of energy, such as heat, to a sample and passing the light emitted through a slit, to obtain a narrow beam. Then the beam is sent through a glass prism to disperse the light into the spectrum. Each line in the spectrum corresponds to an electronic transition between energy levels associated with the element. These levels are called energy states. The lowest energy state an atom can be at is called its ground state. When an electron in an atom has absorbed energy it is said to be in an excited state. An excited atom is

unstable and tends to rearrange itself to return to its lowest energy state. When this happens, the electrons lose some or all of the excess energy by emitting light. Light is only emitted at certain frequencies, each corresponding to a particular electronic transition within the allowed states. The Bohr model is used to account for the spectrum of the hydrogen atom, but the basic idea is the same for all elements. The single electron in hydrogen revolves around the nucleus in one of a limited number of circular orbits. When it is in the orbit closest to the nucleus it is in its ground state, this electron is in the valence band of the element. When hydrogen is heated, or some other energy is being applied to it, the atom absorbs the energy and the electron becomes excited and "jumps" to an orbit farther from the nucleus. In other words, it goes up to a higher energy state. The more energy is applied to the atom, the higher in states the electron goes until the point where it becomes a free electron and no longer part of the atom. The lines in an emission spectrum occur when the electron loses energy, "falls back", from a higher energy state to a lower one emitting photons at different frequencies for different energy transitions.

The Energy Band Model

Description: Throughout the book, the energy band model, or band diagram, is used to describe the behavior of electrons and holes to externally applied forces like light, heat, and/or voltage; therefore, learning how to interpret them and sketch them is basic for this course. Here we are interested in the microscopic things, like individual or groups of electrons and their actions, so we have to use our imagination quite a bit here.

The band diagram models the allowed states in the semiconductor known as energy bands. The two bands of allowed states are called the conduction band and the valence band. The two are separated by an intervening forbidden gap, which we call the band gap. To review what the allowed states are take a look at the Bohr Model of the Atom. Band: A range of some physical variable, as of radiation wavelength or frequency. A range of very closely spaced electron energy levels in solids, the distribution and nature of which determine the electrical properties of a material. Conduction: The transmission or conveying of something through a medium or passage, esp. of electric charge or heat through a conducting medium without perceptible motion of the medium itself. Dopant: A small quantity of a substance, such as phosphorus, added to another substance, such as a semiconductor, to alter the latter's properties. Energy: The work that a physical system is capable of doing in changing from its actual state to a specified reference state, the total including, in general, contributions of potential energy, kinetic energy, and rest energy. Gap: A suspension of continuity; hiatus. A conspicuous difference; disparity. Model: A tentative description of a system or theory that accounts for all of its known properties. Valence: The capacity of an atom or group of atoms to combine in specific proportions with other atoms or groups of atoms. A valence electron is in the outer or next outer shell of an atom and can participate in forming chemical bonds with other atoms. The Conduction Band: The conduction band is the upper band of allowed states. When it is drawn it is represented by a line labeled by Ec which represents the lowest possible energy state in the conduction band. This band is usually empty, it contains few or no electrons since energy is required for them to get there from the valence band. Electrons in the conduction band are free to move about the crystal, thus the name conduction band.

If an electron does manage to get to the conduction band, it resides there for mere fractions of a second (an average lifetime). When it loses its energy it drops back down to the valence band emitting its energy as heat, light or by transferring it to another electron. The Valence Band: The valence band is the lower band of allowed states. In the drawings it is depicted by a line labeled by Ev which represents the highest energy state in the valence band. Since electrons have a tendency to fill the lowest available energy states,the valence band is always nearly completely filled with electrons, especially as the temperature falls toward 0K. As the temperature rises or light is introduced, electrons can absorb the energy and leave the valence band to rise up to the conduction band. When an electron gains enough energy, greater than the band gap energy, and gets to the upper band, it is free to move, becoming a carrier and therefore increasing the conductivity of the semiconductor. When electrons leave the valence band they leave behind a hole which can move about the crystal, also adding to the conductivity. The Band Gap Energy: The band gap energy is the energy needed to break a bond in the crystal. When a bond is broken, the electron has absorbed enough energy to leave the valence band and "jump" to the conduction band. The width of the band gap determines the type of material (conductor, semiconductor, insulator) you are working with. This is shown pictorially using a band diagram. Material Classification: Crystalline materials can be classified according to their bandgap. An insulator is a poor conductor since it requires a lot of energy, 5-8 eV, to excite the electrons enough to get to the conduction band. We can say that the width of the band gap is very large, since it requires that much energy to traverse the band gap, and draw the band diagram respectively. A metal is an excellent conductor because, at room temperature, it has electrons in its conduction band constantly, with little or no energy being applied to it. This may be because of its narrow or nonexistent band gap, the conduction band may be overlapping the valence band so they share the electrons. The band diagram would be drawn with Ec and Ev very close together, if not overlapping.

The reason semiconductors are so popular is because they are a medium between a metal and an insulator. The band gap is wide enough to where current is not going through it at all times, but narrow enough to where it does not take a lot of energy to have electrons in the conduction band creating a current. Introducing Dopants: When a semiconductor is doped, energy states are introduced in the band gap. If it is doped with donors, the energy states are called donor states. Because it takes very little energy, much less than the band gap energy, to free the electron that inhabits the donor state, the states are shown close to the conduction band. Adding donors, therefore, adds more electrons to the conduction band (without adding holes to the valence band) making the semiconductor more conductive. Acceptor states are introduced into the forbidden gap if the semiconductor is doped with acceptors. These initially empty states readily accept an electron to complete its bonds with the four nearest neighbors in the crystal. When an electron from the valence band transitions to an acceptor state, it leaves behind a hole. The energy required for an electron to move to an acceptor state is much less than the band gap energy so it is shown close to the valence band. Holes are created without creating electrons.

Carriers
A Brief Description Carriers are the objects of interest when dealing with semiconductors. If there is a lack of carriers, there is little or no current flow. If carriers are present, either an electron in the conduction band or a hole in the valence band, a negative or positive current, respectively, will result. A semiconductor is doped to increase the concentration of one or both carriers to create a larger current flow. If donors are added making the electron the majority carrier, the semiconductor is said to be ntype. For a p-type semiconductor, acceptors are added to make holes the majority carrier. Carrier: One that transports or conveys. A charge-carrying entity, especially an electron or a hole in a semiconductor. Electron:

A subatomic particle in the lepton family having a rest mass of 9.1066e-28 grams and a unit negative charge of approximately 1.602e-19 coulomb. Hole: A vacant electron energy state that is manifested as a charge defect in a crystalline solid, the defect behaving as a positive charge carrier with charge magnitude equal to that of the electron. Dopant: A small quanitity of a substance, such as phosphorus, added to another substance, such as a semiconductor, to alter the latter's properites. Electrons & Holes Carriers are the objects of interest when dealing with semiconductors. When there is a lack of carriers, there is little or no current flow. This occurs when very few, if any, bonds are broken, which can be shown using the bonding model. This situation can also be shown as the valence band being full. On the other hand, a free electron and a hole give rise to the generic term carriers. When carriers are present there is a current flow, this only occurs if a bond is broken causing an electron to be in the conduction band and/or a hole being created in the valence band of a band diagram. It is fairly simple to visualize an electron "jumping" to the conduction band and moving in every direction, since we know it is a physical object. However, visualizing a hole is something new to almost, if not everyone, in the class. Since a hole is the absence of an electron, one way to look at it is by picturing a soft drink. When it is being poured, bubbles rise from the very bottom of the glass to the top and are released. As they rise to the top, the liquid is taking their place at the bottom. We can picture the glass to be the valence band and the atmosphere being the conduction band. The bubbles are the holes and the liquid is the electron. Since electrons like to be at the lowest energy state possible, they take the place of the hole in a lower state, making it rise to a higher state carrying the absence of an electron, a positive charge, with it; therefore, we get a positive current due to holes. When the soft drink goes flat, there aren't any bubbles moving around and the glass is full of liquid. Likewise, we can say the valence band is full and the conduction band is empty since positive charges do not occur in the conduction band. The number of carriers in an intrinsic semiconductor is one of the identifiable properties of a material. These are denoted n, for the number of electrons per cubic centimeter, and p, for the number of holes per cubic centimeter. In an intrinsic semiconductor and under equilibrium conditions, these two quantities are equal, since they occur in pairs. If a bond is broken, creating an electron, it also creates a hole in

the process. However, this does not make for an efficient semiconductor. The charge created by the free electron will be neutralized by the charge created by the hole; therefore, dopants are added to the semiconductor material to increase either the electron or hole concentration in the semiconductor. Doping A Semiconductor Doping is a routine process in fabricating semiconductor devices, making them extrinsic semiconductors. Donors, all from Column V of the periodic table, introduce energy states within the band gap close to the conduction band. These elements have a weakly bound 5th electron. This loose electron requires very little energy to leave the energy state it's in and enter the conduction band, becoming a carrier. Notice how holes are not created in this process, the valence band remains full. Free electrons are not created unless the energy exceeds the band gap energy. The resulting semiconductor is n-type, it can be associated with negative carriers. It is doped with donors, making the electron the majority carrier. The hole is the minority carrier, since there are very few, if any, present. A semiconductor can also be doped with acceptors. These come from Column III of the periodic table and introduce energy states in the forbidden gap close to the valence band. Acceptors need one more electron to complete a bond. In other words, they lack an electron. This situation creates holes by trapping electrons from the valence band in the acceptor states and leaving behind positively charged voids, holes. This occurs easily because the energy that is needed to get from the valence band to the acceptor states is much smaller than the band gap energy. Free electrons are not created in this process unless the energy exceeds the band gap energy. The semiconductor is now ptype, which can be associated withpositive carriers. It is doped with acceptors, making the hole the majority carrier. The minority carrier is the electron, since very few were created or even present. The bonding model can not be used to model these actions, but the band diagram can be used effectively. Density of States The number of allowed states in the energy bands grows as the number of atoms in a solid increases. When the number becomes very large, the energy states are so close together they "blend" and become a continuous band of allowed states. These bands become the valence band and the conduction band of the crystal. Electrons can only be found in these bands. No other energies are allowed. We use gc(E) and gv(E) to determine the density of states at energies close to the edges of the energy bands. These equations model the distribution of the energy states in the valence band and the conduction band.

E <= Ev

E >= Ec

The Parable of the Empty Seat An analogy that can be applied to carriers in a semiconductor can be found in a football stadium. No matter if anyone is playing or what day it is, the stadium has the same number of seats in a static arrangement. This is the stadium's capacity, and we can call this the "Density of Seats". Think of the rows of seats as being different energy levels. Now imagine a game is being played and some people are in the stands. Because people like to be close to the action, they tend to sit low in the stands. The lower seats, the rows lower in energy, are more likely to be filled. Electrons in a semiconductor are similar; they tend to "prefer" to occupy lower energy electronic states. Even so, one can always find some electrons at the higher energy states just as you will find some people choose seats in the higher rows of the stadium, even when some closer seats are available. Once the game has started and the people are settled, take a snapshot of where people are seated. Wait 15 minutes and take another. Chances are some people will have moved around since the first snapshot. Some may have left, some may have come in, some may have just switched seats. Now imagine that the stadium was sold out so all the seats are filled. Can anyone move? The answer is no, because in order to move, there must be an empty seat somewhere in the stadium. If a few people didn't show up, someone would be able to move to one of those seats, leaving an empty seat behind. Someone can move into this empty seat and leave another empty seat behind, and so on and so forth. This also happens with electrons. Electrons need an empty electronic state nearby in order to be able to move. Two electrons (people) cannot occupy the same state (seat) at the same time. For an electron to move, it must have an empty state nearby. Otherwise it will not move, no matter how much energy is applied. When a semiconductor is at zero Kelvin, all the states in the valence band are full and all the states in the conduction band are

empty. There is no conduction (movement) of electrons in the conduction band because it doesn't have any electrons. The valence band is full of electrons, but there aren't any empty states to where an electron can move. Therefore, at zero Kelvin a semiconductor does not conduct electricity. If this analogy really modeled a semiconductor in equilibrium: 1. The number of people in the stadium would essentially remain unchanged, but the people would continuously be moving. 2. The average number of people would be a function of temperature, their number increasing with increasing temperature (more people show up on a warm fall day than a cold winter day). 3. Even though people are continually moving, their distribution in the rows would always be similar and would be determinable mathematically with something like the Fermi function. 4. Using The Fermi Function 5. The Fermi function is a probability distribution function. It can only be used under equilibrium conditions. The Fermi function determines the probability that an energy state (E) is filled with an electron when the material we are working with is under equilibrium conditions. The Fermi level (EF) helps determine carrier distributions. To a first approximation, all the energy states above the Fermi level have a low probability of being filled with electrons and all the energy states below the Fermi level have a high probability of being filled with electrons. For an electronic state with energy the same as EF, the probability of that state being filled is 1/2, or 50%. 6. Some of the properties of the Fermi function are:

The probability an energy state is occupied: The probability an energy state is empty: The Fermi function at E = EF: The Fermi function is simply a mathematical function and has no units: f (EF) = 1/2 0 < f (E) < 1

7. In a band diagram, the position of the Fermi level determines which carrier dominates. If the semiconductor contains more electrons than holes, n-type material, the Fermi level is positioned above mid gap. If holes are more abundant than electrons, p-type material, EF is positioned below mid

gap. When the electron and hole concentrations are approximately equal, intrinsic material, EF is positioned at mid gap. The Fermi function, or level, also varies with temperature and carrier concentration. 8. Carrier Distribution 9. We need to know how many electrons and holes there are in a semiconductor. Sometimes we also need to know which energy levels they are typically occupying. Taking what we know of the density of states (the distribution of the energy states in the energy bands) and the Fermi function (the probability of an energy state being occupied by an electron) we can determine the distribution of the carriers. This is simply done by multiplying the two functions for the respective energy band. The distribution of electrons in the conduction band can be found by the product gc(E)f(E). The distribution of holes in the valence band can be found by the product gv(E)[1-f(E)]. 10. At the band edges g(E) dominates and the carrier distribution is zero. As we move up in energy, f(E) dominates in the conduction band and will approach zero as the probability of an energy state being occupied approaches zero. [1f(E)] dominates in the valence band and will approach zero as the probability of an energy state being empty approaches zero. Equilibrium Carrier Concentrations Once we know how to determine the carrier distribution, we can find the carrier concentration by integrating over all energies:
Electron concentration: Hole concentration:

After a lot of words and math we derive simple equations we can understand and use:
Electron concentration: Hole concentration:

and finally the nopo product relationship: These equations are only valid when the semiconductor is in equilibrium and nondegenerate . Another way to say a semiconductor is nondegenerate is that the Fermi level, EF, is more than 3kT from any of the states for which we are counting electrons. The nopo product relationship is one of the most useful equations because once you know one of the carrier concentrations (using the equations for no or po), the other can be easily calculated.

We typically deal with uniformly doped semiconductors and if they are at room temperature, we also assume total ionization of the dopant atoms. With these assumptions we can use the charge neutrality relationship and the nopo product relationship from above to derive equations for no and po that take into account the doping concentrations:

Charge neutrality relationship: po - no + ND - NA = 0 Electron concentration:

Hole concentration:

These equations can be simplified under a number of situations. Below are the most common: When a semiconductor is not doped, NA = 0 and ND = 0, the semiconductor is intrinsic and no = po = ni. This also occurs when NA and ND are approximately equal, or ni >> |ND - NA|. The equations for the carrier concentrations for a p-type semiconductor, NA >> ni and ND = 0, can be simplified. Since NA >> ni, we can neglect ni in the equation for po and obtain the carrier concentrations using the following equations:

The equations for the carrier concentrations for an n-type semiconductor, ND >> ni and NA = 0, can be simplified. Since ND >> ni, we can neglect ni in the equation for no and obtain the carrier concentrations using the following equations:

When a p-type semiconductor is compensated, doped with both acceptors and donors (NA - ND >> ni and ND is nonzero), the equations may be simplified similarly to Case 2 because we can still neglect ni in the equation for po. The nopo product relationship can then be used to solve for the electron concentration:

When an n-type semiconductor is compensated, doped with both acceptors and donors (ND - NA >> ni and NA is nonzero), the equations may be simplified similarly to Case 3 because we can still neglect ni in the equation for no. The nopo product relationship can then be used to solve for the hole concentration:

If the doping concentration, or the difference in doping concentrations if the semiconductor is compensated, is comparable to ni, we cannot simplify the equations. The full expression must be used. Charge Neutrality Charge neutrality occurs when all the charge in a volume adds to zero; it is neutral, neither positive or negative. The equation for charge density (Coulombs/cm3) is: = q(whatever has charge) where q = electronic charge. In a semiconductor, the most common and most prominent sources of charge are electrons holes, and ionized acceptors and ionized donors.

The zero net charge does not mean that the electrons, holes, ionized donors, and ionized acceptors are not present in the semiconductor. It simply states that in a uniformly doped semiconductor the negative charge associated with an electron or ionized acceptor would be canceled by the positive charge associated with a hole or ionized donor. This does not mean that the actual electrons, holes, and ionized impurities have ceased to exist in the semiconductor, it means that = 0. If = q(po - no + ND - NA) = 0, then the sum of the charges associated with the carriers must equal zero:
po - no + ND - NA = 0

This equation is useful in many areas, including computing po and no in equilibrium. We usually consider NA and ND to be known, so it gives us one equation to relate two unknowns, poand no. We also know that pono=ni2 and ni is considered known, so we can use the two equations to solve for po and no in terms of ND, NA, and ni Temperature Dependence Temperature is one of the most, if not the most, important parameter in semiconductors. A dictionary will tell you that temperature is the degree of hotness or coldness relative to a standardized scale. You should be accustomed to the Celsius scale by now; it was established with reference values set to the freezing and boiling of water. We will typically talk about the absolute scale, in degrees Kelvin, which is written as 300 K or 0 K, etc., without a degree symbol because it is absolute temperature. (need to review? check this out!) Why is temperature important? The average energy of a solid -- and its components (atoms, electrons, etc.) -- is measured by its temperature. The higher the temperature, the more (thermal) energy is available to be used by the atoms and electrons. Since we are talking about "electronic" devices, and temperature has a significant impact on the behavior of electron, temperature must have a significant impact on the behavior of electronic devices. Temperature directly impacts:

Carrier concentrations -- by impacting: o Fermi distribution o Band gap o Donor and acceptor ionization

o Intrinsic carrier concentration Carrier Energy

, or, in words, the average kinetic energy of an electron in a solid is linearly proportional to the temperature of the solid. For room temperature, we find that the average velocity of a free electron is approximately 107cm/sec. How does temperature affect the Fermi distribution? When we are trying to determine the position of the Fermi level we assume the semiconductor is in equilibrium. EF on the band diagram is a function of temperature and carrier concentration. This can be seen from the equations used to determine the position of the Fermi level.
Intrinsic semiconductor no = po

n-type semiconductor ND >> NA, ND >> ni

p-type semiconductor NA >> ND, NA >> ni

If the temperature is varied, the Fermi level will also vary. How does temperature affect the band gap? As temperature increases, the band gap energy decreases because the crystal lattice expands and the interatomic bonds are weakened. Weaker bonds means less energy is needed to break a bond and get an electron in the conduction band. The relationship between temperature and the band gap energy can be seen by the following equation:

EG(0) is the limiting value of the band gap at 0 K. and are constants chosen to obtain the best fit to experimental data. How does temperature affect donor and acceptor ionization? A donor is an element with typically one more valence electron than the element that forms the semiconductor. An acceptor is an element with typically one fewer valence electron than the element that forms the semiconductor. An ion is an electrically charged atom or group of atoms. Ionization occurs when an electron is removed from or added to an atom, in this case it would be the dopant atoms. The donor's extra electron is weakly bound and needs very little energy, thermal or otherwise, to break the bond and become a free electron. As temperature is increased, more and more of these bonds are broken, until all the donors are ionized, producing an increase in electron concentration. At room temperature, we assume all the donors are ionized. Figure 2.22(a) on page 66 illustrates the temperature dependence of the carrier concentration in a doped semiconductor. At 0 K the electrons at the donor sites do not receive enough energy to make it to the conduction band, so the ratio between the majority carrier concentration and the doping concentration n/ND = 0. As the temperature is increased , more donor electrons make it to the conduction band, but at temperatures below 150 K all the donor sites have not been ionized so the ratio n/ND < 1 and the majority carrier concentration is made up of the donor electrons that have made it to the conduction band. At room temperature, we assume all the donors are ionized, meaning all donor electrons are in the conduction band, so the ratio n/ND = 1 and the majority carrier concentration is made up of the donor electrons. At very high temperatures, above 500 K, electrons from the valence band receive enough energy to make it to the conduction band and out number the electrons from the donor sites, so the ratio n/ND > 1 and the majority carrier concentration is now made up of electrons from the valence band in the conduction band, as in an intrinsic semiconductor. Acceptors have one fewer valence electron than the neighboring semiconductor atoms and will readily accept an electron in order to completely bond. Very little energy is needed for an electron to travel to one of the acceptor sites and be trapped. As the temperature is increased, more electrons become trapped in acceptor sites, until all the acceptors are ionized, causing the hole concentration to increase. At room temperature, we assume all the acceptors are ionized. Figure 2.22(a) can be applied toward acceptors by substituting p/NA in place of n/ND. At 0 K the electrons in the valence band do not receive enough energy to

jump to an acceptor site and create a hole, so the ratio p/NA = 0. At temperatures below 150 K the ratio p/NA < 1 because all the acceptor sites have not been ionized, so the majority carrier concentration is made up of the holes left behind by trapped electrons in the acceptor sites. At room temperature we assume all the acceptors have been ionized. The ratio p/NA = 1 because the majority carrier is made up of the holes created by the ionized acceptors. At very high temperatures, above 500 K, the ratio p/NA > 1 because the holes created by electrons from the valence band jumping to the conduction band out number the holes that had been created by acceptors, so the majority carrier concentration is made up of holes created by valence electrons in the conduction band, as in an intrinsic semiconductor. How does temperature affect intrinsic carrier concentration? When we speak of an intrinsic semiconductor several factors come to mind: 1. It is extremely pure, containing an insignificant amount of impurities. 2. The properties of the material depend only on the element(s) the semiconductor is made of. 3. For every electron created, a hole is created also, no = po = ni. For an electron-hole pair to be created in an intrinsic semiconductor, a bond must be broken in the lattice, and this requires energy. An electron in the valence band must gain enough energy to jump to the conduction band and leave a hole behind. ni represents the intrinsic carrier concentration, or we can see it as the number of bonds broken in an intrinsic semiconductor. As the temperature is increased, the number of broken bonds (carriers) increases because there is more thermal energy available so more and more electrons gain enough energy to break free. Each electron that makes it to the conduction band leaves behind a hole in the valence band and there is an increase in both the electron and hole concentration. As the temperature is decreased, electrons do not receive enough energy to break a bond and remain in the valence band. If electrons are in the conduction band they will quickly lose energy and fall back to the valence band, annihilating a hole. Therefore, lowering the temperature causes a decrease in the intrinsic carrier concentration, while raising the temperature causes an increase in intrinsic carrier concentration.

Carrier Actions
A Brief Description

Now that you have a basic understanding of how carriers behave under equilibrium conditions, we can observe the behavior of carriers when the semiconductor environment is changed. These changes can come in the form of externally applied forces, like an electric field, a change in temperature, or a change in the degree of doping. Carriers have three primary actions in a semiconductor. These are drift, diffusion, and recombination-generation. Drift and diffusion currents make up the total current density in the semiconductor, these are not necessarily always present. Processes that are always present in a semiconductor to stabilize it and return it to equilibrium conditions are recombination and generation, where electrons and holes are destroyed if there is an excess of a carrier, or generated if there is a deficit. Band bending: Occurs when the conduction band and the valence band become a function of position because of an electric field that exists inside a material. The result is a variation of Ec andEv with position in the energy band diagram. Current density: Current is the charge per unit time crossing an arbitrarily chosen plane of observation oriented normal to the direction of current flow. Current density, J, is the current per unit area (J=I/A). Diffusion: The angular redistribution of radiation by a scattering, reflecting or refracting system, ideally producing an isotropic distribution of intensity. The gradual mixing of the molecules of two or more substances, as a result of random thermal motion. Drift: To be carried along by or as by currents of air or water. To vary from or oscillate randomly about a fixed setting, position, or mode of operation. To move leisurely or sporadically from place to place. Equilibrium: A condition in which all acting influences are canceled by others, resulting in a stable, balanced, or unchanging system. The condition of a system in which the resultant of all acting forces is zero and the sum of all torques about any axis is zero. The state of a reaction in which its forward and reverse reactions occur at equal rates so that the concentration of the reactants does not change with time. Generation: The act or process of generating, i.e. origination, production or procreation. We will use it as the process whereby electrons and holes are created. Lifetime: The period of time during which an individual is alive. The period of time during which an object, property, process, or phenomenon exists or functions.

Mobility: Capable of moving or of being moved from place to place. Moving quickly from one place to another; flowing freely. Perturbation: To disrupt greatly; to agitate. Variation in a designated orbit, as of an electron or planet, resulting from the influence of one or more external bodies. Radiation: The act or process of radiating. The emission and propagation of waves or particles. The propagating waves or particles, such as light, sound, radiant heat, or particles, emitted by radioactivity. Recombination: We will use it in the context of it being the process whereby electrons and holes (carriers) are annihilated, or destroyed. Resistivity: The capacity for or tendency toward resistance. The resistance per unit length of a substance with uniform cross-section. Resistance is the opposition to electric current characteristic of a medium, substance, or circuit element. Scattering: The act or process of separation or dispersing. The dispersal of a beam of particles or of radiation into a range of directions resulting from physical interactions. Thermal motion: Motion due to temperature. Thermal is of, pertaining to, using, producing, or caused by heat. Drift Current Drift is, by definition, charged particle motion in response to an applied electric field. When an electric field is applied across a semiconductor, the carriers start moving, producing a current. The positively charged holes move with the electric field, whereas the negatively charged electrons move against the electric field. The motion of each carrier can be described as a constant drift velocity, vd. This constant takes into consideration the collisions and setbacks each carrier has while moving from one place to another. It is considered a constant though, because the carriers will eventually go the direction they are supposed to go regardless of any setbacks, especially if you look at the direction of all the carriers, instead of each one individually. Drift current in a semiconductor is the resultant of carrier drift. Because we are talking about a semiconductor, or specific areas in a semiconductor, we are interested in the current density. When dealing with drift current, we are interested in the current density due to drift, and drift arises in response to an electric field. Drift current also

depends on the ability of the carriers to move around in the semiconductor, or the electron and hole mobility. Another parameter drift current depends on is the carrier concentration, because you have to have carriers in order for there to be current. Each one of these carriers has a charge, but in this case we will only take q as a magnitude. Finally, we have that the current density due to drift depends on four parameters: the electric field, the electron or hole concentration, the mobility constant, and the charge. The reason we use q for both electrons and holes when it's +q for holes and -q for electrons, is that the electric field takes care of the sign, or direction of the current. When a negative electric field is applied, the electrons will go opposite the electric field. The electron charge is -q, so the resulting electron drift current will be positive. On the other hand, when the electric field is negative, the holes will go the direction of the electric field. Their charge is +q, so the resulting hole drift current will be negative. Vice versa when the electric field is positive. Diffusion Current Diffusion is the process of particles distributing themselves from regions of high concentration to regions of low concentration. If this process is left unperturbed, there will eventually be a uniform distribution of particles. Diffusion does not need external forces to act upon a group of particles. The particles move about using only thermal motion. If we let the particles be carriers, so as they move around they take charge with them. The moving of charge will result in a current. We call this current due to diffusion. The difference between drift current and diffusion current is that drift current depends on the electric field applied: if there's no electric field, there's no drift current. Diffusion current occurs even though there isn't an electric field applied to the semiconductor. It does not have E as one of its parameters. The constants it does depend on are Dp and Dn, and +q and -q, for holes and electrons respectively. The first constants are called the diffusion coefficients, a proportionality factor. We don't worry too much about these because they are constants. We do worry about the gradient of the concentration of p and/or n, though. But, since we are talking about a one dimensional situation when we are solving for current densities, we only worry about the gradient (or derivative) with respect to the x-plane. The other difference between drift current and diffusion current, is that the direction of the diffusion current depends on the change in the carrier concentrations, not the concentrations themselves. In the equation, the signs are reversed as we are used to seeing them. We usually assign a +q to holes and -q to electrons. In the case of diffusion current, they are reversed to be opposite of the derivative of the

concentrations. This occurs because the carriers are diffusing from areas of high concentrations to areas of low concentrations. For example, if the derivative of p with respect to x is positive, then the concentration of holes is growing as you move towards the +x direction. Diffusion current will be the opposite of that, the holes will be diffusing in the -x direction to where there's a lower concentration of holes. If the derivative is negative, the opposite will occur. The concentration of holes is decreasing as you go from the -x to +x direction. Therefore, holes will diffuse to the +x direction where there's a lower concentration of holes. This is why the negative sign is needed in the equation for the hole diffusion current. The same goes for electrons, but in this case, the signs cancel for a positive derivative because the electrons, carrying -q, diffuse to the -x direction where there's less electrons. The sign remains if the derivative is negative, because electrons will be diffusing to the +x direction carrying a -q charge. For these reasons it's not included in the equation for the electron diffusion current. Both drift current and diffusion current make up the total current in a semiconductor. They may not be occurring at the same time, but the equation is still valid. Under equilibrium conditions, the current density should be zero because there shouldn't be any drastic changes occurring, like applying an electric field or changing the carrier concentrations by a large margin. Even so, if the doping is not completely uniform, there will be a change in concentration is some places in the semiconductor, resulting in a gradient. This gradient can in turn give rise to an electric field, which in turn can give rise to non-zero current densities. Mobility & Scattering In order for there to current, the electrons and holes must move. The ability for them to move around in a material and transport charge is called mobility, there's hole mobility and electron mobility. These are not necessarily constants inherent to electrons and holes. Several factors affect the mobility of a carrier. The most significant is scattering, the motion-impending collisions within the crystal. These collisions can be an electron bumping into another electron, or a hole or ionized impurities. Scattering may increase/decrease due to temperature and/or the addition of acceptors or donors. The higher the temperature the more excited the carriers are, so the more scattering there will be. The same occurs with dopants, because not only will there be more carriers to bump into and scatter, dopants will generate ionized impurities which also carry a charge.

When the doping concentration is low, the mobilities are pretty much independent of the doping concentration. The mobility of the carriers begins to monotonically decrease as the concentration of the dopants increases. Band Bending and Potential & Kinetic Energies Band bending occurs when an electric field is applied to a semiconductor. When there isn't an electric field being applied, the energy bands are not a function of position. Therefore, when an electric field is applied, energy is being supplied to the carriers in the material. Where band bending has occurred, we can measure the energy being given to the electrons or holes by taking a reference energy and measuring the distance between it and the energy the carrier is at. In essence, the potential energy of an electron is the distance the electron is from the reference energy to Ec, the lowest conduction band energy. The kinetic energy of an electron is the distance between Ec and the energy it is at. For holes, it's similar. A hole's potential energy is measured as being the distance between the reference energy and Ev, the highest valence band energy. The kinetic energy of a hole is measured by taking the difference between Ev and the energy the hole is at. Recombination Recombination is part of a process to restore equilibrium to a semiconductor that has been perturbed, or disturbed out of equilibrium. Perturbations can be in the form of an applied electric field, a change in temperature or exposure to light. Recombination occurs when there is an excess of carriers and they are destroyed, by recombining. Remember carriers are in the form of a free electron and the hole it leaves behind, or electrons or holes brought in from dopants. When they are destroyed, a negatively charged electron is attracted to a positively charged hole, and as they get together, their charges are canceled and the electron is part of a bond once again. There are three types of recombination that we cover in this course. These are band-to-band recombination, R-G center recombination and Auger recombination. Band-to-band recombination is the annihilation of a conduction band electron and a valence band hole simply by having the attraction of the negative charge of the electron and the positive charge of the hole cancel each other out as they move in the semiconductor lattice. The energy released during the process usually produces a photon and emits light. In R-G center recombination there are special locations within the semiconductor which are lattice defects or special impurity atoms that introduce allowed electronic levels near the center of the band gap. These levels are important because they will

not be mistaken for donor or acceptor levels. R-G center recombination involves a two step process. A carrier and electron, for example, strays into the vicinity of an RG center, an electronic level near mid-gap, and is trapped. Later comes a hole and the electron is attracted to the hole and loses more energy so it goes into the valence band annihilating the hole along with itself, this is called indirect recombination. This method releases thermal energy or produces lattice vibrations. Auger recombination is a type of band-to-band recombination that occurs when two carriers collide. The collision transfers the energy released from the recombining carrier to the surviving carrier. In other words, one carrier loses energy and the other gains it. The one that loses it is recombined, and the one that gains it goes to a higher energy level. Eventually, this highly energized carrier "thermalizes" - loses energy in small steps through heat producing collisions with the semiconductor lattice, until it eventually recombines or gains energy once more. The three recombination mechanisms are similar in the fact that they all annihilate an electron for each hole, and that they all produce some kind of energy, whether it be heat or light, in the process. They are all different in the fact that band-to-band recombination only uses the fact that the carriers of opposite charge happen to be in the same vicinity, no outside interference is needed. R-G center recombination uses defects in the crystal lattice or special impurities that are added to the semiconductor material to add energy levels to where the carriers can move up or down, closer to their opposite charge so the attraction can destroy them. Auger recombination uses collisions between electrons to transfer energy from one electron to another, sending one to a higher energy and another to a lower one where it is attracted to a hole and they are both annihilated. Generation Generation can be taken as the opposite of recombination. The generation of carriers is another part of the process to restore equilibrium to a semiconductor that has been perturbed, or disturbed out of equilibrium. Perturbations can be in the form of an applied electric field, a change in temperature or exposure to light. Generation occurs when there's a deficit in carrier concentration, when compared to equilibrium carrier concentration. Carriers then have to be created within the semiconductor to reach equilibrium. The three mechanisms of recombination can be reversed to create carriers. The three ways are band-to-band generation, R-G center generation, and Auger generation. Band-to-band generation can occur when an electron is exited directly into the conduction band. This can occur by obsorbing thermal energy or light. When generation occurs by absorbing thermal energy, this is called direct thermal generation

and when externally applied light is absorbed, it is called photogeneration. RG centers can also be used for generation. The electron would not need as much energy to make it to the conduction band as it would need for band-to-band generation. Auger generation occurs just as in recombination. This occurs more often when an electric field is applied, though.

Minority Carrier Diffusion Equations

A Brief Description Minority carrier diffusion equations (MCDEs) are equations used to model semiconductors under special circumstances. They are derived from the continuity equations after making a series of assumptions that therefore limit the applicability of the MCDEs. At the same time, MCDEs can be used in a large number of situations to derive closed-form, analytical solutions that describe carrier concentrations and currents. When semiconductors are being used (which is when they are most interesting) processes are constantly occurring within, caused by the outside influences like change in temperature, the presence or absence of light, or an applied voltage. These may cause carrier drift, diffusion, and recombination-generation and affect the carrier concentration within the semiconductor as a function of time or space as the semiconductor attempts to bring itself back to equilibrium. After making several assumptions, the continuity equations are simplified to the minority carrier diffusion equations (MCDEs).

Minority carrier diffusion equations:

The MCDEs can be further simplified to obtain a less complex differential equation, with respect to either time or space. The end result of using the MCDEs is to obtain the rate of change in minority carrier concentration with respect to time or space within a semiconductor.

Continuity Equations The continuity equations are "bookkeeping" equations that take into account all of the processes that occur within a semiconductor. Drift, diffusion, and recombinationgeneration are constantly occurring in a semiconductor. Although we have studied these processes individually, they take place at the same time. These carrier actions change the carrier concentration in the semiconductor as a function of time and space, and because carriers transport a charge, a current will result.

Continuity Equations:

The continuity equations are too complicated to use except on a computer as difference equations derived from an ordinary partial differential equation. We have two equations and three unknowns, n, p, and V. With Poisson's equation, we have the third equation we need to solve problems using a computer.
Poisson's Equation:

The minority carrier diffusion equations are derived from the continuity equations by making several assumptions. The assumptions allow us to model a semiconductor without using a computer. For a more visual explanation, check out the demo. You will need the shockwave plug-in for your browser in order for the movie to run. They should run properly on the computers in the Vectra lab using Internet Explorer, but they will not run on the Sun workstations. Once the page is loaded, the movie runs on its own. For better viewing, use the "full screen" option on your browser. Assumptions: Assume: To take for granted; suppose. Assumption: A statement accepted or supposed true without proof or demonstration. We make assumptions when modeling a semiconductor in various applications:

In using the continuity equations, In deriving the minority carrier diffusion equations, In analyzing diodes, In analyzing transistors.

Why do we make assumptions? We use the continuity equations to model the processes that are taking place inside a semiconductor. They are complicated equations containing partial differentials, one for each carrier action (drift, diffusion, thermal R-G, and other processes). Making assumptions about the area(s) we are studying allows us to simplify the mathematical equations into something we can solve by hand. However, all assumptions must be tested.

Assumptions made when using the continuity equations We do not make any assumptions when using the continuity equations because they take into account all the processes that are occurring within a semiconductor. The continuity equations can be solved only by computer since they are too mathematically complicated to allow a nice analytical solution. Assumptions are needed to obtain the minority carrier diffusion equations The continuity equations are mathematically complicated equations containing partial differentials that describe the change in carrier concentration due to individual processes; i.e., drift, diffusion, recombination, generation, and other processes. These equations can be solved only by computer. The minority carrier diffusion equations are derived from the continuity equations and are used to model semiconductors by obtaining the rate of change of the minority carrier concentration with respect to space and/or time.

The following assumptions are made in order to derive the minority carrier diffusion equations from the continuity equations, and these conditions must exist in order to use the MCDEs: 1. The semiconductor is one dimensional, all derivatives are with respect to one coordinate; usually called the x-coordinate. 2. We are only working with minority carriers. 3. An electric field does not exist in the semiconductor region we are analyzing. 4. Equilibrium minority carrier concentrations are not a function of position, meaning the doping is uniform. 5. We assume low-level injection conditions exist. 6. Thermal recombination-generation occurs indirectly through traps in the bandgap. 7. The only other process that may occur would be generation caused by shining light on the semiconductor. Assumptions made when analyzing pn-junction diodes When modeling a pn-junction diode, we are solving for the minority carrier current densities in the diode as a function of position and the applied voltage. In order to do this we must first solve the MCDE to obtain np and pn. Since we use the MCDEs to analyze diodes, all those assumptions are made and must be valid. We also assume: 1. Operation is under steady state conditions. 2. The doping profile is a nondegenerately doped step junction. 3. The diode is one dimensional, all derivatives are with respect to one coordinate; usually the x-coordinate. 4. Assume low-level injection in quasineutral regions. 5. Only drift, diffusion, and thermal recombination-generation occur within the diode. 6. The diode is in the dark, GL = 0. 7. No R-G in space charge region (depletion region). Assumptions made when analyzing BJTs Solving for BJT parameters and terminal currents involves assumptions that parallel those for the pn-junction diode, then we add:

1. Thermal recombination-generation is insignificant everywhere in both the emitter-base (EB) and collector-base (CB) depletion regions. 2. The widths of the emitter and collector are much greater than LN or LP. 2. Carrier Lifetimes & Diffusion Lengths 3. Minority carrier lifetimes, denoted n for electrons in a p-type material and p for holes in an n-type material, are the average time a typical minority carrier survives in a sea of majority carriers. Lifetimes vary from carrier to carrier and from sample to sample. An electron is considered to be "alive" when it is in the conduction band. It "recombines" after some amount of time (on average, that time is n) by returning to the valence band where it is no longer able to conduct/move because it is participating in a covalent bond to hold the crystal together. An electron can also be captured by a trap or recombination center. A hole is considered to be "alive" when it is in the valence band (remember the hole is simply the lack of an electron). It recombines when an electron annihilates it by falling from the conduction band. 4. Minority carrier diffusion lengths, denoted LN for electrons in a p-type material, and LP for holes in an n-type material, vary with the minority carrier lifetimes. They represent the average distance the excess minority carrier is from where it was created to when it is annihilated.

MCDEs are used to solve for minority carriers in:

p-type or n-type devices, pn junctions, BJTs

1. You have the problem, now what do you do with it? 2. Determine which equation you need. 3. Simplify the equation. o Steadystate o Diffusion current or change in concentration o Thermal R-G o Light 4. Obtain a general solution. 5. Boundary Conditions 6. Obtain exact solution. 7. Verifying your solution.

1. You have the problem, now what do you do with it? First of all, READ THE PROBLEM CAREFULLY! Take note of the information given. In one way or another, the type of semiconductor material, the temperature, time, presence or absence of light, doping, etc., will be given in the wording of the problem. Look for keywords like, "uniform", "infinite", "room temperature", "for a long time", "suddenly", "in the sun", or "in the dark". 2. Determine which equation you need. Once you understand the problem, you need to determine the type of material youre working with, whether it's p-type or n-type, and the minority carrier youre solving for. Remember, for a p-type material, electrons are the minority carrier, and for an ntype material, holes are the minority carrier. So if you're solving for electrons in a ptype material, you would use the following equation:

If you're solving for holes in an n-type material, you would use this equation:

3. Simplify the equation. Next, simplify the equation using the information you obtained in step 1. Also keep in mind the assumptions on pages 122, 241, and 389 in your book, depending on the problem you are solving. Steady State If the semiconductor has been a certain way for a long time, we can assume that it is in steady state. Remember that steady state means that nothing changes as a function of time, so time derivatives go to 0. So, in steady state:

Otherwise, if "suddenly at time t = ?" a condition changed, that term probably stays in the equation. If something like this occurs, you may have to solve for both instances, before the condition changed, and after because one represents the initial condition. Diffusion Current or Change in Concentration When you see the words "uniformly doped", we can assume that the semiconductor is doped equally throughout the region. In other words, the concentration does not vary as a function of position. This is a necessary condition to use the MCDEs. (Why?) The spatial derivatives go to zero when there is no variation in the concentration of n or p with position:

This one is hard to decide sometimes. Look for any indication that there are at least two points where the values are not the same. Note that one or both of the partial derivatives (time or space) must be eliminated to be able to find a "nice" solution to the problem. If the problem asks you to solve with respect to position, chances are it stays and the time derivative will be eliminated (steady state problem). If the problem asks for the change in concentration of the minority carrier with respect to time, the spatial derivative must go but, for full credit you must explain why. Look for clues in the problem statement. Thermal R-G Pay close attention to what is said about minority carrier lifetime or diffusion length and the width of the region of interest. This can give you clues as to whether or not the following is true:

We assume thermal R-G occurs when the minority carrier diffusion length is less than or equal to the width of the region we are analyzing. In the event that it is stated or assumed that the minority carrier diffusion length is much greater than the width of

the region, as in the base region of an ideal BJT, we can assume that there isn't any thermal R-G taking place in that region. The reasoning behind that is we assume the carriers won't have time to recombine before they cross the region and are removed. Light The presence of light causes photogeneration in the semiconductor, which means that carriers are being generated in the semiconductor. The clues to look for are phrases like "illuminated", "it has been outside", or "in the dark". If it is in the dark GL = 0. However, there are times when light is shining on the semiconductor, but GL = 0. Figure out what is happening to the light and whether it is penetrating the semiconductor. 4. Obtain a general solution. Once you have simplified the equation by taking out some of the terms, match the equation you have to the general differential equation solutions found on your cheat sheet given in class and at the exam. Write down the general solution corresponding to the differential equation remaining after simplifying. If you have a spatial partial derivative, you should have two unknowns to solve for because it is a second order differential equation. If the time derivative remained, you only need to solve for one unknown because it is a first order differential equation. 5. Boundary Conditions The hardest part of solving for the minority carrier concentration is determining the boundary conditions. Boundary conditions are used to solve for the unknowns and find the exact solution. First you must identify the location(s) of the boundary(ies). If you have a time varying general solution, you need an initial condition, the value of the quantity at t = 0. If you have a spatial problem, you need to know the boundaries of the region of interest. If the boundary is "a long way away", it is typically at x = +-infinity. Semi-infinite semiconductors have boundaries at x= 0 and x = infinity. pn junctions have boundaries at +-infinity, -xp, and xn. Next, determine the values at the boundary(ies). Sometimes these values are explicitly stated in the problem, but under most circumstances they must be inferred from the situation described. Here are some rules (assuming we are talking about holes, for electrons, substitute n for p where needed )

1. If the minority carrier concentration "is at its equilibrium value" at a boundary, p = p - p0 = 0 because p = p0. 2. If "all the minority carriers are removed" at a boundary, i.e. by an ohmic contact, p = p0 + p = 0, so p = -p0. 3. If the boundary is at the edge of a pn junction depletion region, the law of the junction holds, so

4. If a boundary is "a long way away" from things that are happening, but the region is "uniformly illuminated":

5. If a boundary is in the dark and "a long way away" from what we call the "perturbation", usually the minority carrier concentration has returned to its equilibrium value (see 1. above). 6. Obtain Exact Solution. Now youre ready to find the exact solution. You do this by "plugging in" the values of x or t that you used to determine your boundary conditions, i.e. x = 0, x = infinity, t = 0, or t = infinity, into your general equation and setting it equal to the value you obtained as your boundary condition. Solve for your constants. WATCH YOUR ALGEBRA! This is what gets you into trouble most of the time. 7. Verifying Your Solution. Finally! You've solved for the constants, you have the exact equation, but does it look right? The only way you can find out is to check it. "Plug in" the values you used to find your boundary conditions into your equation, the result should be the boundary condition you used in step 5 for that value. You can check the units of the result, as well.

PN Junction Diodes
Description: A pn junction is formed by diffusing a p-type material to an n-type wafer (and there are several other ways as well, see your text). At the point of the junction, which we call the metallurgical junction, is where N - N = 0. The doping profile is our starting point to analyze pn junction electrostatics, since it combines the information given for N and N . It gives us enough information to sketch the band diagram, the electrostatic potential V, the electric field E, and charge density , as a function of x. This is an exercise you should be able to do without consulting the book. Try it now!
D A D A

There are many different applications with pn junctions. Photodetectors, solar cells and LEDs, are optoelectronic devices that have a pn junction. BJTs, J-FETs and MESFETs, are devices that have more than one pn junction. Last, but not least, the pn junction and its characteristics are the basis of almost every semiconductor device made. Definitions: Bias: The fixed voltage applied to an electrode. Built-in potential: Also known as V , it is the change in potential across the depletion region of a pn junction under equilibrium conditions. Current: A flow of electric charge. The amount of electric charge flowing past a specified circuit point per unit time. Density: The amount of something per unit measure. Deplete: To reduce or lessen in quantity, value, or effectiveness; exhaust. To empty. Depletion region: The transition region from p-type semiconductor to n-type semiconductor in a pn junction. This region is characterized by a large electric field, huge changes in hole and electron concentrations, large amounts of fixed charge (can't move), and a varying potential. Depletion approximation: An idealization of the actual charge distribution in the depletion region that originates from the fact that the majority carriers have been removed. We say
bi

this region is "depleted" of majority carriers. It facilitates the use of Poisson's equation because we can obtain a closed-form solution. When using depletion approximation, we are assuming that the carrier concentration (n and p) is negligible compared to the net doping concentration (NA and ND) in the region straddling the metallurgical junction, otherwise known as the depletion region. Outside this region, it is assumed that the net charge density is zero.
Poisson's Equation: Charge Density: Simplified to one dimension by the depletion approximation:

Depletion width: The width of the region straddling the metallurgical junction of a pn diode where the diffusion of carriers has left a charge, otherwise known as the width of the depletion region, where W = xn +xp . Diode: An electronic device that restricts current flow chiefly to one direction. xn: The width of the depletion region on the n-side. xp: The width of the depletion region on the p-side. Equilibrium: A condition in which all acting influences are canceled by others, resulting in a stable, balanced, or unchanging system. The condition of a system in which the resultant of all acting forces is zero and the sum of all torques about any axis is zero. The state of a reaction in which its forward and reverse reactions occur at equal rates so that the concentration of the reactants does not change with time. Forward bias: A voltage V > 0 is applied to the diode terminals. The potential on the n-side of the junction is lowered relative to the p-side of the pn junction. V must be =< V since the voltage drop across the depletion region becomes V - V . Ideal: A conception of something in its absolute perfection. Linearly graded junction: A more realistic approximation than the step junction to model junctions formed by deep diffusions into moderate to heavily doped wafers. The linearly
A A bi bi A

graded profile uses a grading constant in the formula N - N = ax, where a's units are cm-4. pn-junction: The region that is formed by adjoining a p-type semiconductor to an n-type semiconductor. The line dividing the two regions, where N - N = 0, is the metallurgical junction. Region: A specified district or territory. Reverse bias: A voltage V < 0 is applied to the diode terminals. This lowers the potential on the p-side of the junction relative to the n-side of the pn junction. V must be =< V since the voltage drop across the depletion region becomes V - V . Space charge region: Antoher name for depletion region. Step junction: The doping profile we use to approximate the pn junction electrostatics of a lightly doped starting wafer. It means that is n-type with N = 0 next to a region that is p-type with N = 0. The doping profile looks like the figure below.
D A D A A A bi bi A A D

The Depletion Region: When analyzing the pn junction, we are most interested in the depletion region, or space charge region (SCR). This is where all the interesting stuff goes on. When we talked about band bending, one of the reasons that it occurred was because of nonuniform doping or an applied electric field. Well, an electric field can result because of non-uniform doping, as in a pn junction, and that is why the energy band diagram for a pn junction has band bending even when it is in equilibrium. As soon as you connect a p-type region with an n-type region, carriers will begin diffusing from regions of high concentration to regions of lower concentration. That

is, holes from the p region will diffuse to the n region, where there aren't as many holes, and electrons from the n region will diffuse to the p region where there aren't as many electrons. As they leave, they leave behind the ionized impurities (dopants) that created them. This migration of majority carriers stops when the electric field created by the ionized impurities causes carriers to drift back at the same rate as they diffulse away. The space charge region is the region around the metallurgical junction where all the ionized acceptors and donors were uncovered and remain. The number of ionized acceptors on the p-side equals the number of ionized donors on the n-side. When we look at the depletion region like this, we ae using the depletion approximation and we are assuming that the carrier concentrations (n or p) are negligible compared to the net doping concentration (N or N ). We are also assuming that the charge density is zero outside the space depletion region (see definitions).
D A

For a more visual explanation, take a look at the demo of a pn junction. You will need shockwave plug-in for your browser in order for the demo to run. It should run properly in the Vectras using Internet Explorer, but it will not run on the Suns. Use the "full screen" option for better viewing. In order to return to this page you must use the "back" button on your browser. Varying V :
A

Vbi is the built in voltage in a pn-junction diode in equilibrium. VA is the voltage that is applied to the diode terminals.

Applying VA > 0, the diode is forward biased. Reverse bias is when VA < 0. The diode is in equilibrium when VA = 0. VA is considered positive when the higher potential is applied to the p-side of the diode. The question is, how does varying VA affect the diode? The simple answer is that xp and xn change as VA changes, therefore changing the width of the depletion region. If you take a look at the equations for charge density, electric field and potential on pages 210-213, they all include the bounderies of the depletion

region, xp and xnso they all change as well. The width increases with VA < 0 and decreases with VA > 0. Let's take as an example a pn diode. When VA = 0, the Fermi level is constant and we can determine that the diode is in equilibrium. If we reverse bias the diode, apply VA < 0, we increase the potential hill by lowering the n-side of the diode with respect to the p-side, increasing the barrier to diffusion. If we forward bias the diode, apply VA > 0, we decrease the potential hill by lowering thep-side of the diode with respect to the n-side, making it almost flat. This can be seen with band diagrams. Electrostatic variables, such as potential, electric field, and charge density, all change with changing the applied bias in a similar manner since they are all a function of x. The easiest way to determine the electrostatic potential, V, is to follow the band diagram, since it can be drawn as an "upside down" conduction band, so as the band diagram changes, the potential changes by the same degree, take alook. The electric field, E, is also affected by varying the depletion region. There's a larger electric field when the diode is reverse biased since there is a wider depletion region. A smaller one with forward biasing corresponding to a narrower depletion region. Take a look at how the electric field is affected in this example. Charge density, , only depends on the doping and charge. What makes the graph vary is that it is graphed only between xp and xn, so as the depletion width changes, the graph for charge density changes, although the height doesn't. To see how charge density is affected by increasing the width of the depletion region of the pn diode with forward biasing and decreasing it with reverse biasing, see this example. Now, putting them all together for each VA:

Va = -10 Va = -5 Va = -2 Va = 0 Va = 0.4 Va = 0.75 You can also take a look at the demo for pn junction statics. You will need shockwave plug-in for your browser. It should run properly on the Vectras using Internet Explorer, but it will not run on the Suns. Use the "full screen" option for better viewing, and you will need to use the browser's "back" button to return to this page.

Diode Currents: Remember when drift current was first introduced? What about diffusion current? No? Well then, let's refresh your memory! Drift current is electrons and holes responding to an applied electric field. Holes move in the direction of the electric field while electrons move opposite the electric field. This occurs as long as there are carriers available. Diffusion current is holes and electrons moving from areas of high concentration, where they are the majority carrier, to areas of low concentration, where they become minority carriers. This occurs until they are uniformly distributed. Unlike drift, diffusion takes place without an electric field being present. Back then you thought, "When will I use this?" Well, here's your chance! Now we're talking about pn junction diodes. The band diagram that models a pn junction has band bending at the metallurgical junction. Remember that band bending is caused by an applied electric field and that non-uniform doping can cause an electric field. In a pn junction the electric field is produced when a p-type and an n-type semiconductor are brought together; therefore it is present even when it is in equilibrium. This means both drift current and diffusion current are present in equilibrium but the total current density J = 0, now how could that be? Keep in mind that the p-side of the diode contains many holes and few electrons and that the n-side has many electrons and few holes. When we draw the band diagram for a pn junction, the p-side is always at a higher level (potential), do you know why? (Hint: It has to do with the Fermi level.) This means there is a potential hill the majority carriers must climb to get to the other side, n-sideelectrons to get to the pside and p-side holes to get to the n-side. Do you recognize this activity? It's diffusion. This potential hill has a totally different meaning to the electrons on the p-side and holes on the n-side, otherwise known as the minority carriers. They don't see a hill, they see a slide! Remember electrons flow against the electric field and holes flow with it, therefore carriers near the depletion region get caught in the electric field and quickly drift to the other side. Do you recognize this activity? We call it drift. Applying a potential to the diode affects the current flowing through the diode. The potential hill varies with the applied voltage VA. As discussed earlier, if VA = 0 the diode is in equilibrium, if VA > 0 it is forward biased, and if VA < 0 it is reverse biased. Earlier we said that drift current and diffusion current are present even though J = 0 in equilibrium. This is because for every electron that diffuses from the n-side to the p-side there is an electron that drifts from the p-side to the n-

side. The same goes for holes. The two currents balance each other and the total current density J = 0. When the diode is forward biased, VA > 0, the potential hill is still present, but is is less steep. A smaller potential hill means electrons from the n-side and holes from the p-side need less energy to climb the hill, therefore diffusion current increases. This current is positive, electrons flowing opposite the electric field produce a positive current and holes flowing with the electric field also produce a positive current. Under forward biasing, the applied voltage needs only to be increased by tenths of a volt to see a significant change in current. As the voltage is increased slightly, the current increases exponentially because of the increasing number of carriers that have enough energy to cross the junction. Remember, majority carrier concentrations are around the magnitude of 1017/cm3, so if even one third cross the junction it is still a lot. When the diode is reverse biased, VA < 0, the potential hill gets steeper. If the majority carriers had difficulty climbing the potential hill when the diode is in equilibrium, what do you think happens when the hill is even steeper? If you thought, "They won't be able to get to the other side!", then you're right! By setting VA just a few tenths of a volt less than zero, the potential hill gets steep enought that the majority carriers do not have enough energy to climb it and make it to the other side. This causes the diffusion current to become negiligible. Now you're thinking "But why does the ID vs. VAcurve show a negative current when VA = 0?" Well, we said that diffusion current goes to zero, we haven't mentioned what happens with drift current yet. Do you think VA affects drift current? If you thought "No!", you're right! As we mentioned before, the potential hill affects drift in a totally different way than diffusion. It doesn't matter how steep or flat the potential hill is, as long as there is an electric field and available carriers (ionized dopants don't count) drift will remain the same. The current produced is negative, electrons are flowing with the electric field producing a negative current and holes are flowing against the electric field also producing a negative current. The reverse bias current is expected to be small, unlike forward bias current, because it depends on the minority carrier drift. When the diode is forward biased drift current is present, but because diffusion current grows exponentially, it dominates. The total current flowing through the depletion region under forward biasing is made up of mostly majority carrier diffusion. When the diode is reverse biased diffusion is negligible, but drift remains constant. The total current flowing through the depletion region under reverse biasing is made up of mostly of minority carrier drift.

The Ideal Diode Equation: Diodes should be familiar to us by now. We use them when we need current to flow in only one direction. The question is, just how much current is there? Of course, we can just give you the equation to figure it out, but what good will that do? Instead let's derive it! You should know by now that when we derive something, we must first make a few assumptions. These are similar to those we make when we use the MCDEs. As a matter of fact, we will be using the MCDEs to derive the ideal diode equation.

Next, we have to recognize that we have to consider three regions (instead of just one like we're used to), the quasineutral p-region, the quasineutral n-region, and the depletion region. Are you wondering what quasineutral is? Take a detour here. The quasineutral p-region is from the edge of the depletion region, which we will call xp, to the edge of the diode, which we assume is an infinite distance away from xp. The quasineutral n-region is from the edge of the depletion region, which we will call xn, to the edge of the diode, which we assume is an infinite distance away from xn. In the quasineutral regions there is no electric field. This allows us to use the MCDEs to find the current densities in these regions. In order to solve the MCDE for the quasineutral regions we must first determine the boundaries and the boundary conditions. (Hint: You were just given the boundaries for solving the MCDEs in these regions.) We assume the edges of the diode are an infinite distance away from any actions taking place in the depletion region. This means that there isn't any variation of carrier concentrations as we get to the edges of the diode:

At the edges of the depletion region, -xp and xn, equilibrium conditions do not prevail so we must use the "law of the junction".
The Law of the Junction:

To find the boundary conditions at -xp and xn we use the law of the junction and solve for the minority carrier in each region to obtain:

Using the assumptions we made, the MCDE and J in the quasineutral regions simplify to:

On the n-side: x >= xn On the p-side: x <= -xp

Next, using the boundary conditions, we solve the MCDEs for each quasineutral region. You can do the math and come up with:

Then, to find the current densities of the quasineutral regions, we take the derivatives of np and pn and plug them into the equations for JP and JN. Next, we evaluate JP and JN at the depletion region edges, -xp and xn respectively, to obtain the current density in each region. By adding them together, we obtain the current density

in the depletion region.

Derivatives of np and pn:

JP and JNat the depletion region edges:

The current density in the depletion region:

Did you already forget what we're deriving? Don't worry, we're almost done. We have solved for the current densities in the quasineutral region to obtain the current density in the depletion region, but what we're looking for is current through the diode. If you recall, current is charge crossing an area, therefore we multiply (you can do this) the current density (J) by the area (A) to obtain the ideal diode equation (emphasis on ideal):

Frequently Asked Questions

Energy states, Energy Levels, and Energy Bands

What is an electronic state? The statement on page 27 states that bringing the atoms closer and closer together forces the spread of allowed energies, how does this occur? How does an electron get to the conduction orbital and move around (in the Si model)?

What is an electronic state? Electronic states, also known as energy levels or energies, are sites in an atom or a crystal that an electron can reside in. In a semiconductor, these sites are found in the valence band and in the conduction band and vary from semiconductor to semiconductor. They are like a bookshelf, with shelves at certain heights. You can place objects on one shelf or another shelf, but not in floating in the air or with the shelf in the middle of the book. In any crystal, these states organize themselves into bands. Semiconductors have valence and conduction bands. An electronic state has associated with it a specific electron energy, electron momentum, and location. See also: The Bohr Model of the Atom The Energy Band Model

The statement on page 27 states that bringing the atoms closer and closer together forces the spread of allowed energies, how does this occur? Each isolated atom has its own electrons in discrete energy levels. Atoms in a solid are so close together that the energy levels of each atom split into many energy levels so closely spaced that they become bands of energies. We call these bands the valence band and the conduction band. See also: Carriers - Density of states Isolated silicon (director movie) Many atoms (director movie)

How does an electron get to the conduction orbital and move around (in the Si model)? Silicon crystal is made up of silicon atoms forming covalent bonds. This means that the atoms share pairs of electrons. These electrons are locked in the valence band until they obtain enough energy to jump to the conduction band. Electrons in the conduction band are able to move within the crystal lattice.

Material Classification

How is the transmission of light used for material classification?

How is the transmission of light used for material classification? The wider the band gap, more light can pass through. Conductors have no bandgap, so they are opaque. Insulators have the widest bandgaps, so they tend to be translucent. Semiconductors are somewhere in between and tend to be opaque

The Fermi Level

What do I need to know in order to understand the Fermi function? How is the Fermi energy (level) useful in determining whether a semiconductor is p-type or n-type?

What do I need to know in order to understand the Fermi function? The fundamental principles needed to understand the Fermi function are: 1. Electrons tend to occupy lower energy states. 2. Temperature is a measure of the energy available to electrons. 3. Electrons must absorb energy from somewhere to occupy higher energy states. 4. The more energy there is around (more at higher temperatures), the more likely it is that we will see electrons at higher energy states.

5. It is a statistical probability function. 6. It is only valid under equilibrium conditions.

How is the Fermi energy (level) useful in determining whether a semiconductor is p-type or n-type? In an energy band diagram if EF is above EI, it is n-type. If EF is below EI, it is ptype.

Carrier Concentrations

When is it correct to use the simplified equations for n and p? What is meant by total ionization of dopant atoms? What does the graph in Figure 2.22(a) on page 66 represent?

When is it correct to use the simplified equations for n and p? The equations for n and p are:

These can be used all the time; it happens to be that if one of the doping concentrations is much greater than the other, , they will be simplified to equations 2.30 and 2.31 on page 60 of your textbook. On the other hand, if the doping concentrations are comparable to ni, you should not use the simplified equations, both terms should be kept.

What is meant by total ionization of dopant atoms?

Atoms used as donors have one more valence electron than silicon. When the atom loses the electron, a negative charge, the atom is ionized and is positively charged. Acceptors have one fewer valence electron than silicon, so they do not bond completely with the rest of the silicon atoms. When it takes an electron from a neighboring silicon atom to complete a bond it becomes ionized and negatively charged with the extra electron, and a hole is created. At room temperature, dopants are assumed to be fully ionized. At 0 K, no dopants are ionized.

What does the graph in Figure 2.22(a) on page 66 represent? It is many things. It is a plot of carrier concentration versus temperature. The graph illustrates how dopants are ionized as a function of temperature. It depicts how the ratio of the majority carrier concentration (n) to the doping concentration (ND) changes with temperature T(K). In the freeze out region, below 100K, the ratio is less than 1, meaning few dopants are ionized and the majority carrier is mostly due to electrons from ionized donors. In the extrinsic T region, the ratio is approximately 1, meaning that the majority carrier is made mostly of donor electrons and all of them are ionized. Increasing the tempreature no longer increases the electron concentration because there are no unionized donors to ionize. Above 500K the ratio is greater than 1 so the carrier concentration is greater than the doping concentration. At this point, electrons coming from the valence band begin to outnumber the electrons coming from donor sites. This graph can be applied for holes also, substituting NA for ND and holes for electrons. Can you plot the minority carrier concentration that would correspond to this plot?