A photograph of a silicon singlecrystal wafer with fabricated computer processors. The wafer is a thin slice from a silicon single crystal, and the wafer is processed to contain many computer processors. Each square is approximately a half a centimeter across and is an individual processor. The processors on this wafer use copper as the conductor. The silicon single-crystal wafer is approximately 30 cm (12 in.) in diameter. Courtesy of Intel Corporation The goals of this chapter are to understand ●● The relationship of electrical resistivity and conductivity to electrical resistance ●● The variation of electrical conductivity across material classes ●● The electrical conductivity and resistivity of metals ●● The relationship among electrical resistivity, defects, and temperature in metals ●● The distribution of free-electron energies in metals ●● Thermionic emission of electrons from metals ●● Superconductivity in metals and compounds ●● The electrical conductivity of semiconductors ●● The contact potential of junctions ●● The diode equation for pn-junctions ●● The function and applications of dielectric, ferroelectric, and piezoelectric materials ●● The electrical conductivity of ionic materials ●● The procedures for producing electronic integrated circuits Chapter 16 Electrical Properties of Materials 16.1 Electrical Conductivity and Resistivity of Materials The resistance (R) of a specimen is determined by measuring the current (I ) in amperes (A) with an applied voltage (V ) in volts (V) and using Ohm’s law, as shown in Equation 16.1. V 5 IR 16.1 The resistance of a specimen is not a material property, since it depends upon both the material and the dimensions of the specimen. The electrical resistivity () is a material property independent of specimen dimensions, as given by Equation 16.2 for a specimen with a uniform cross-sectional area A and length L. 5R A L 16.2 W-4 CHAPTER 16 The electrical conductivity () is also a material property, and the electrical conductivity is inversely related to the electrical resistivity (), as shown in Equation 16.3. 5 1 16.3 Table 16.1 presents the electrical conductivity at room temperature of some materials with metallic, covalent, and ionic bonding. FCC metals in Group IB, such as copper and silver, have the highest electrical conductivity or lowest resistivity of the metals. These metals are good conductors of electricity, and we will discuss their conductivity in Section 16.2. Pure silicon (Si) and germanium (Ge) have an electrical conductivity that is approximately eight to eleven orders of magnitude less than those of silver and copper, and these materials (Si and Ge) are called electrical semiconductors. We will discuss the Table 16.1 The Electrical Conductivity of a Variety of Materials Conductivity Conductivity (ohm21 ? cm21) Material Material (ohm21 ? cm21) SuperconductorsSemiconductors Hg, Nb3Sn Infinite below Group 4B elements YBa2Cu3O72x a critical Si 4 3 1026 MgB2 temperature Ge 0.02 Metals Compound semiconductors Alkali metals GaAs Na 2.13 3 105 AlAs K 1.64 3 10 SiC 2.5 3 1029 0.1 10210 5 Alkali earth metals Insulators, Linear, and Nonlinear Dielectrics Mg 2.25 3 105 Ca 3.16 3 105 Polyethylene Polymers 10215 Group 3B metals Polytetrafluoroethylene 10218 Al 3.77 3 10 Polystyrene 10 Ga 0.66 3 105 Epoxy 10212 to 10217 5 217 to 10219 Transition metals Ceramics Fe 1.00 3 105 Alumina (Al2O3)10214 Ni 1.46 3 105 Silicate glasses Group 1B metals Boron nitride (BN) 10217 10213 Cu 5.98 3 105 Barium titanate (BaTiO3)10214 Ag 6.80 3 105 C (diamond) Au 4.26 3 105 , 10218 Unless specified otherwise, assumes high-purity material. * Based on data from Askeland, D.R., Fulay, P.P., and Wright, W.J., The Science and Engineering of Materials, 6th ed., Cengage Learning, Stamford, CT. (2011), p. 723. Electrical Properties of Materials electrical conductivity of semiconductors in Section 16.5. Diamond and PE have an electrical conductivity nearly twenty-four orders of magnitude less than those of silver and copper, and these materials are electrical insulators. Also listed in Table 16.1 are superconductors. Superconductors have an infinite electrical conductivity or zero electrical resistivity below a critical temperature, as we will discuss in Section 16.4. Currently the critical temperatures for all superconductors are well below room temperature. In this chapter we discuss why materials have such different electrical conductivities. Figure 16.1 is a photograph of an individual computer processor. Most processors are fabricated on a thin single-crystal wafer of silicon, as shown in the photograph at the beginning of this chapter. A computer processor is produced with a combination of metal (M) conductors, oxide (O) insulators, and semiconductors (S). The computer processor is an integrated circuit (IC). In an IC, all of the devices needed for operation of the processor are built into the wafer. The most common fabrication architecture to produce ICs is called CMOS, an abbreviation for complementary metal-oxide semiconductor. Complementary refers to the way in which the semiconductors are designed into the circuit. The metal aluminum was used as the conductor in older processors; however, some newer processors use copper, as shown in the photograph at the beginning of this chapter. The insulating oxide is usually silica (SiO2), although other oxides are also utilized. The insulating oxide separates the metal conductor from the semiconductor in areas where contact is not desired, and oxides are used for capacitors and for memory. Figure 16.1 A photograph of an Intel® Pentium® 4 computer processor. (Courtesy of Intel Corporation ) W-5 W-6 CHAPTER 16 16.2 Electrical Conductivity of Metals and Metal Alloys Most materials with a high electrical conductivity and a low resistivity are metals. Table 16.2 presents the room-temperature electrical resistivity of some metals. In solid and liquid metals, the valence electrons are free of the ion cores. The free electrons can move anywhere in the metal. When a voltage is applied to a metal specimen, the free electrons are accelerated toward the positive terminal. The core electrons on the atoms do not contribute to the current. Since the free electrons can be found anywhere in the metal, each free electron has a probability of interacting with every other free electron. The Pauli exclusion principle states that no two interacting electrons with the same quantum numbers can occupy the same space. Therefore, in metals the Pauli exclusion principle applies to all of the free electrons, and no two free electrons can have the same quantum numbers in the entire metal specimen. The free electrons propagate through the metal with momentum components in the three directions x, y, and z. Each free electron has a unique set of quantum numbers (nx, ny, and nz) and a spin quantum number of 112 or 212 . If there are n free electrons, then there are ne/2 different sets of quantum numbers (nx, ny, and nz). In a cubic meter of metal, the number of free electrons (n) is typically 1028, and there would be 0.5 3 1028 different sets of quantum numbers (nx, ny, and nz) for 1028 free electrons. The set of quantum numbers (nx, ny, and nz) for a free electron is called a quantum state. The lowest set of quantum numbers is 1, 1, 1, and Table 16.2 Room-Temperature Electrical Resistivity for Some Metals Room-Temperature Resistivity (ohm ? cm) Metal Temperature Resistivity Coefficient (R) [ohm/(ohm ? °C)] Be4.0 3 10260.0250 Mg4.45 3 10260.0037 Ca3.91 3 10260.0042 Al2.65 3 10260.0043 Cr12.90 3 1026 (0°C) 0.0030 Fe9.71 3 10 0.0065 26 Co6.24 3 10260.0053 Ni6.84 3 10260.0069 Cu1.67 3 10260.0043 Ag1.59 3 10260.0041 Au2.35 3 10260.0035 Pd10.8 3 10260.0037 W5.3 3 1026(27°C)0.0045 Pt9.85 3 10260.0039 Based on data from HANDBOOK OF ELECTROMAGNETIC MATERIALS: MONOLITHIC AND COMPOSITE VERSIONS AND THEIR APPLICATIONS by P. S. Neelkanta. TAYLOR & FRANCIS GROUP LLC—BOOKS (1995). W-7 Electrical Properties of Materials Table 16.3 The Crystal Structure, Lattice Parameter (a), Free-Electron Density (n), and Fermi Energy (EF) for Several Monovalent Metals Metal Crystal Structure a nm 1028 electrons/m3 EF eV Cs BCC 0.6130.868 1.53 K BCC 0.5341.32 2.14 Na BCC 0.4292.55 3.16 Li BCC 0.3514.66 4.72 Ag FCC 0.4095.85 5.51 Cu FCC 0.3628.43 7.04 the quantum numbers increase until all electrons are assigned unique sets of quantum numbers. The kinetic energy of each free electron is proportional to the sum of the squares of the quantum numbers (n2x 1 n2y 1 n2z ). Figure 16.2a shows that the energy of a valence electron of a single isolated atom is a single energy. However, the free-electron energy levels in a metal are spread into a band of energy called the conduction band, as shown in Figure 16.2b. The band indicates that individual energy levels are so close that they form a nearly continuous band of energies. The maximum kinetic energy of free electrons at 0 K is called the Fermi energy (EF), as shown in Figure 16.2b. The Fermi energy of a metal is typically in the range of 1.5 to 7 eV, as shown in Table 16.3. The Fermi energy is measured with experimental techniques, such as low-energy X-ray spectroscopy, which is discussed in Section 15.3.1. Experimental measurements have confirmed the results of the quantum mechanical theory of electrons. Below the energy of the conduction band in Figure 16.2b are the core electrons; however, these electron energy levels are not normally indicated because they do not contribute to electrical conductivity. Ohm’s law can be written in a form that contains the electrical conductivity, as shown in Equation 16.4. 16.4 J 5 sEd J is the current density in amperes per square meter (A/m2), and E is the magnitude of the applied electric field. The magnitude of the electric field (E) in this book is printed as a capital in Roman type to differentiate it from internal energy (E) that is written in italics. The vectors, such as the electric field (E), Energy –EF (a) (b) Figure 16.2 (a) The discrete energy level of a single 3s electron on an isolated sodium atom. (b) The conduction band of energy levels resulting from the free 3s electrons in sodium metal and the Fermi energy (EF). W-8 CHAPTER 16 are printed in bold Roman type. Equation 16.5 gives the magnitude of the electric field (E ) for a metal specimen of length L in the x direction with a uniform cross section and an applied voltage V. E5 V dV 5 L dx 16.5 The electric field is a vector that points from the positive terminal to the negative terminal of the applied voltage. The current density in Equation 16.4 can also be written as the product of the free-electron density (n), the charge on an electron (e), and the drift velocity of the free electrons (vd), as shown by Equation 16.6. J electrons m 5 n1 v 5 E 1coulombs 2 e 1coulombs s?m 2 m electron 2 1 s 2 2 d 3 16.6 The drift velocity of the free electrons is the electron velocity resulting from the applied electric field, and the drift velocity for all free electrons is the same for a specified electric field applied to a specific metal. Equation 16.6 is solved for the electrical conductivity in Equation 16.7: 5 ne vd E 5 nen 16.7 where n is the electron mobility. Table 16.3 gives values for the free-electron density (n) for some metals, and the free-electron density (n) can be calculated as demonstrated in Example Problem 16.1. Equation 16.4 is then rewritten in Equation 16.8 with the electrical conductivity from Equation 16.7 and the magnitude of the electric field as the voltage gradient (dV/dx) from Equation 16.5. J 5 nen dV dx 16.8 Equation 16.8 is similar to the steady-state diffusion Equation 4.37. The sign difference between Equations 16.8 and 4.37 occurs because electrons go in the negative direction if the electric field is in the positive direction. Example Problem 16.1 Determine the free-electron density in sodium metal. Sodium is BCC and in Group I of the periodic table, with a lattice parameter of 0.428 nm. Solution The number of conduction electrons per unit volume (n) for sodium is calculated by knowing that each atom in the solid contributes one electron to the free-electron density. Sodium is in Group I of the periodic table; therefore, there is one electron per atom contributed to the free-electron density. Sodium is BCC, there are two atoms per unit cell, and the lattice parameter (a) is 0.428 nm. Since there is one free electron contributed by each sodium atom, the number of atoms per unit volume is equal to the number of electrons per unit volume. n5 2 atoms 1 electron 2 electrons electrons 3 5 5 2.55 3 1028 a3 atom m3 s0.428 3 1029 md3 In Equation 16.7 the mobility of the electrons is the primary variable for the electrical conductivity of a metal. The charge on an electron (e) is a constant, and for a given metal, such as pure copper, the free-electron density is a constant. The electron mobility depends upon a number of factors, including the type of metal, the various possible defects in the metal, and the temperature. For example, silver and gold W-9 Electrical Properties of Materials are both in Group IB of the periodic table, and both have one valence electron per atom. Also, both have FCC lattices with lattice parameters given in Appendix B of 0.40786 nm for gold and 0.40862 nm for silver. The free-electron density for silver is slightly less than that for gold. However, the electrical conductivity of the two elements is significantly different. Silver has an electrical conductivity calculated from the inverse of the resistivity in Table 16.2 of 6.3 3 107 (V ? m)21, and gold has an electrical conductivity of 4.3 3 107 (V ? m)21. The only term in Equation 16.7 that can account for the difference in electrical conductivity is the mobility of the free electrons. The electrons in silver are more mobile than the electrons in gold. The mobility of free electrons and the electrical conductivity of a given metal, such as pure silver, can be changed by altering the temperature, adding impurities, deforming the metal, or heat treating the metal. Equation 16.7 shows that the electron mobility is the drift velocity of the electrons divided by the magnitude of the electric field. Since the electrical conductivity of a particular piece of metal, such as annealed pure silver at room temperature, is a constant, and if the electric field is constant, the drift velocity of the free electrons must also be a constant. However, there is a constant force on each free electron equal to Ee, resulting in a constant acceleration. What keeps the electron drift velocity from continually increasing as long as the force due to the electric field is applied? Electron scattering provides a limit to the drift velocity. The electrical conductivity of a metal is decreased by the addition of defects, as demonstrated in Figure 16.3 for copper additions to nickel. Defects in a crystal scatter free conduction electrons, and conduction electrons are only accelerated for the time () between scattering events. The shorter the time between scattering events (), the lower the drift velocity. Conduction electrons are scattered by all of the possible structural defects in a crystal and by lattice vibrations. According to Mathiessen’s rule in Equation 16.9, the resistivity is the sum of resistivity components, where 0 is the resistivity due to all of the point (zero-dimension) defects, 1 is the resistivity due to all the one-dimensional defects, 2 is the resistivity due to all of the two-dimensional defects, 3 is the resistivity due to all three-dimensional defects, and th is the resistivity contribution of the thermal vibrations of the atoms. 5 th 1 0 1 1 1 2 1 3 5 1 16.9 600 Resistivity (nano-ohm-m) 500 400 300 200 100 0 0 20 40 60 Atomic percent copper 80 100 Figure 16.3 The electrical resistivity in 1029 V ? m of copper-nickel alloys as a function of atomic-percent copper. (Based on Ho, C.Y. et al. J. Phys. Chem. Ref. Data 12(2) (1983), p. 183.) W-10 CHAPTER 16 The point-defect contribution to resistivity is the easiest relation to demonstrate. The electrical resistivity of alloys is more than that of pure metals, as demonstrated by the increase in electrical resistivity for copper-nickel alloys in Figure 16.3. Impurity atoms have an atomic size and electronegativity that differ from those of the host atom, and impurity atoms scatter the free conduction electrons. Using substitutional B-type atoms as an example, the distance between the substitutional atoms (lB) decreases as the chemical composition of substitutional B-type atoms in atomic fraction (CB) increases, according to the relationship in Equation 16.10: lB 5 dA CB 16.10 where dA is the shortest distance between the host A-type atoms. We can assume that each substitutional B-type atom is a scattering center. The resistivity contribution for substitutional solid-solution B-type atoms ( 0B ) is inversely proportional to the scattering distance. A short scattering distance results in a high resistivity, and a large scattering distance results in a low resistivity, as shown in Equation 16.11: 0B 5 A1 lB 5 A2CB 16.11 where A1 and A2 are constants for a given metal. Equation 16.11 demonstrates Nordheim’s rule, that the electrical resistivity of a metal alloy linearly increases with increases in the atomic fraction of impurity atoms for dilute solutions. Equations 16.10 and 16.11 are applied to other point defects such as vacancies or self-interstitials. Equations similar to Equation 16.10 relating defect densities to scattering distance are also developed for all other defects in materials. In general, defects decrease the distance between scattering events for conduction electrons and increase the resistivity. For example, annealed oxygen-free high-conductivity (OFHC) copper has a low defect density, and it has a resistivity at room temperature of 0.167 3 1029 V ? m. Drawing copper into a wire increases the number of defects, including interstitial copper atoms, vacancies, dislocations, and grain boundaries. The terms 0, 1, 2, and 3 in Equation 16.9 are increased, and the electrical resistivity at room temperature for drawn OFHC copper wire is 0.33 3 1027 V ? m. The electrical resistivity of drawn OFHC copper wire is more than two orders of magnitude higher than that of annealed OFHC copper. Equation 16.10 applies to dilute solid solutions of B-type atoms into an A-type metal host. As can be seen in Figure 16.3, when the concentration of nickel atoms in copper becomes large, the electrical resistivity reaches a maximum and then starts to decrease. For large concentrations of B-type atoms in an A-type host, Equation 16.10 is replaced by Equation 16.12: 0B 5 A3CBs1 2 CBd 16.12 where A3 is a constant for a given metal host and impurity-atom addition. Equations 16.10 and 16.11 apply when the impurity atom is in the form of a solid solution. Some specimens have two phases: and . Two phases can result from an alloy of metals with limited solid solubility in each other, or from a composite material. The electrical resistivity of a material with a mixture of and phases ( ) depends upon the volume fractions of each phase present (v ,v ) and the electrical resistivity of each phase ( , ), as shown in Equation 16.13. 5 v 1 v 16.13 Equation 16.13 is the rule of mixtures. The rule of mixtures also applies to other physical properties of two-phase materials, including thermal conductivity and some mechanical properties. However, the rule of mixtures does not apply if the second phase is oriented in layers that are perpendicular to the direction of the current flow, heat flow, or applied force. Chapter 12 covers the rule of mixtures and other rules that govern the properties of two-phase materials. W-11 Electrical Properties of Materials As the temperature increases, the electrical resistivity of a metal increases, as shown in Figure 16.4. The increase in electrical resistivity in excess of the value at 0 K is due to the thermal component of electrical resistivity (th). As the temperature approaches 0 K for a metal that is not a superconductor, there is a residual electrical resistivity (s) that is due to zero-, one-, two-, and three-dimensional structural defects, as shown in Figure 16.4. We can assume that the thermal component of electrical resistivity (th) in Equation 16.9 at 0 K is equal to 0. At any temperature (T ) above 0 K, the total resistivity is the sum of the structural component of the resistivity at 0 K (s) plus the thermal component of resistivity (th), as shown in Figure 16.4 and in Equation 16.14. 5 s 1 th 5 1 16.14 The thermal component of resistivity increases linearly with increasing temperature, because the amplitude of the atom vibrations increases, and the conduction electrons are scattered more frequently by the vibrating atoms. The distance between scattering events due to thermal vibrations (lT) is inversely related to temperature (T ), as shown in Equation 16.15: lT 5 b T 16.15 where b is a constant for a given material that has units of m∙kelvin. Since electrical resistivity is inversely proportional to the distance between scattering events, as expressed in Equation 16.11, the thermal component of the electrical resistivity is directly proportional to the temperature, as expressed in Equation 16.16: 16.16 5 BT th where B is a constant for a given material. Equation 16.16 predicts that the thermal component of electrical resistivity increases linearly with temperature (T ), and experimentally we can observe this in Figure 16.4. Inserting Equation 16.16 into Equation 16.14 results in the electrical resistivity as a function of temperature, as expressed in Equation 16.17. 5 s 1 BT 5 1 16.17 160 Resistivity (nano-ohm-m) 140 120 100 80 60 40 20 0 0 200 400 600 800 Temperature (K) 1000 1200 Figure 16.4 The electrical resistivity of copper with 5 weight percent nickel (5.39 atom percent) as a function of temperature. (Based on Ho, C.Y. et al. J. Phys. Chem. Ref. Data 12(2) (1983), p. 183.) W-12 CHAPTER 16 Equation 16.17 predicts that at temperatures approaching 0 K, the electrical resistivity is equal to the structural component of the electrical resistivity (s), as we can observe in Figure 16.4. Then, as temperature is increased, the electrical resistivity increases linearly. Because the electrical resistivity is not perfectly linear over large temperature ranges, and because the residual electrical resistivity at 0 K is small in comparison to the room-temperature electrical resistivity, handbooks of electrical properties of materials list the temperature coefficient of electrical resistivity (R) at reference temperatures. The relationship between the temperature coefficient of electrical resistivity and the electrical resistivity at a temperature different than the reference temperature is given by Equation 16.18, where DT is the temperature difference (T 2 TR) between the actual temperature and the reference temperature, and R is the electrical resistivity at the reference temperature. 5 Rs1 1 R DT d 16.18 Table 16.2 presents values of R for some metals at room temperature. 16.3 The Effect of Temperature on Electron Energies in Metals The source of electrons in free space necessary for the operation of many devices, such as X-ray tubes, LEED systems, and electron microscopes, results from the temperature dependence of the free-electron energies in metals. The schematic of the free electrons in the conduction band of a metal, shown in Figure 16.2, is for 0 K. In a metal at 0 K, the electron energy levels are filled with a probability of 1 for energies up to the Fermi energy (EF); and for energies above the Fermi energy the probability of occupation is 0. The probability P(E) that a free electron has the energy E at temperature T in a metal is given by the Fermi-Dirac statistical distribution, shown in Equation 16.19. 1 PsE d 5 1 1 exp E 2 EF 16.19 kT The Fermi-Dirac statistical distribution at 0 K is a step function that is equal to 1 for free-electron energies less than the Fermi energy, and equal to 0 for free-electron energies greater than the Fermi energy, as shown in Figure 16.5 and in Equations 16.20 and 16.21. T 5 0 K, E , EF , PsE d 5 1 1 5 51 1 1 exp 2 ` 1 16.20 T 5 0 K, EF . EF , PsE d 5 1 1 5 50 1 1 exp ` ` 16.21 For any finite temperature, the value of the Fermi-Dirac statistical distribution is 0.5 for free-electron energies equal to the Fermi energy, as shown in Equation 16.22 and in Figure 16.5. T . 0 K, E 5 EF , PsE d 5 1 1 1 5 5 1 1 exp 0 1 1 1 2 16.22 W-13 Electrical Properties of Materials T=0K Fermi-Dirac probability P (E) 1.0 T = T1 T = T2> T1 0 Energy (E ) F EF Figure 16.5 A schematic of Fermi-Dirac probability distributions P(E ) plotted as a function of energy at 0 K (dashed line), at a higher temperature (T1), and at an even higher temperature (T2). On the energy scale the Fermi energy (EF) and the work function (F) are indicated. The effect of increasing the temperature on the Fermi-Dirac statistical distribution is shown in Figure 16.5. Some of the electrons that were at energies less than the Fermi energy are excited by thermal vibrations to energies greater than the Fermi energy, as shown for temperature T1 and the higher-temperature T2. Since P(E ) is the probability that free-electron energies are filled, then 1 2 P(E ) is the probability that an electron energy level E is unfilled. In Figure 16.5, the distance from the plot of P(E ) to 1 is equal to 1 2 P(E ). P(E ) and 1 2 P(E ) are symmetric about P(E ) equal to 12 when plotted as a function of energy. 16.3.1 Thermionic Emission of Electrons from Metals Free electrons outside a metal are developed by heating a filament with a current to a high temperature; electrons are then emitted by thermionic emission. One material for filaments is tungsten, which has a melting temperature of 3410°C (3683 K). To understand thermionic emission of electrons by metals requires a further development to the model of a metal. The band of conduction electron energy levels, shown in Figure 16.2b, does not include the possibility of electron emission from the metal. Electron emission is included in Figure 16.6 by assuming that electrons are in a finite potential well. The top of the potential well is the energy of free space, and the bottom of the potential well is a free electron with the minimum kinetic energy inside the metal. If the electron has the energy of free space, it is free of the metal. If the electron has less than the energy of free space, it is bound inside the metal. The energy required to excite a Fermi-energy electron to the energy of free space is the work function (F). Table 16.4 presents values of the work function for several metals. To remove any other energy free electron inside the metal to the energy of free space requires the work function plus the energy of the free electron relative to the Fermi energy. Free electrons with energies greater than the Fermi energy require less energy than the work function for removal to the energy of free space, and electrons with energies less than the Fermi energy require more energy than the work function for removal to the energy of free space. The potential well in Figure 16.6 does not include temperature; this is a representation of the energies at 0 K. At higher temperatures, it is possible to have electrons with energies greater than the Fermi energy. W-14 CHAPTER 16 Free space Energy F = Work function EF E=0 Figure 16.6 The potential well model of a metal showing the conduction band (shaded), the Fermi energy (EF), the work function (F), and the energy of free space. Table 16.4 Values of the Work Function and the Parameter ARD Metal F (eV) ARD (A/m2 ? K2) W4.5 75 3 104 Ta4.2 55 3 104 Ni4.6 30 3 104 Cs1.8 160 3 104 Pt5.3 32 3 104 Cr4.6 48 3 104 Based on data from Herring, C. and Nichols, M.H., Revs. Mod. Phys. 21, (1949), p. 185. Example Problem 16.2 Calculate the probability that an electron in tungsten at 3000 K has sufficient energy to be excited to the energy of free space. Solution From Table 16.4, the work function of tungsten is 4.50 eV. The energy of free space is equal to EF 1 F. Putting this energy and temperature into Equation 16.19. PsEF 1 Fd 5 1 1 exp 1 5 EF 1 F 2 E F PsEF 1 Fd 5 kT 1 F 1 1 exp kT 1 4.5 eV 1 1 exp 0.259 eV 5 1 5 1 1 exp 4.5 eV eV 8.62 3 1025 s3000 Kd K 1 5 2.8 3 1028 1 1 exp 17.4 W-15 Electrical Properties of Materials The free electrons that have sufficient energy for thermionic emission are those that have an energy greater than Fermi energy plus the work function (EF 1 F), as shown in Figure 16.5. The value of 1 in the denominator of the Fermi-Dirac statistical distribution is insignificant in comparison to the exponential term in Example Problem 16.2, and little accuracy is lost by approximating P(EF 1 F), as in Equation 16.23. PsEF 1 Fd < exp 2F kT 16.23 The thermionic current density (J ) in A/m2 through the surface of a metal at a temperature T is given by the Richardson-Dushman equation, shown in Equation 16.24: J < ARDT 2 exp 2F kT 16.24 where ARD is a material constant whose value is given for some metals in Table 16.4. 16.4 Superconductors Log of resistivity In a superconductor, the electrical resistivity is 0 and the electrical conductivity is infinite below a temperature called the critical temperature (Tc ). Superconductivity was first observed in mercury by Heike Kamerlingh Onnes in 1908. As shown in Figure 16.7, the electrical resistivity of a metallic superconductor as a function of temperature behaves like a normal metal for temperatures above the critical temperature (Tc ). However, for temperatures below Tc, the electrical resistivity of the superconductor is 0. The resistivity of a normal metal conductor, such as copper, does not go to zero as shown in Figures 16.4 and 16.7. Since its discovery in mercury, superconductivity has been found in many other materials, and Table 16.5 lists the critical temperatures (Tc ) of some of these materials. Superconductor Metal conductor Tc 0 Temperature (K) Figure 16.7 A schematic of the electrical resistivity as a function of temperature for a normal metal conductor, such as copper, and for a superconducting metal. Below Tc, the superconductor has 0 electrical resistivity, and above Tc the superconducting metal behaves as a normal metal. The normal conductor, such as copper, has a residual electrical resistivity at 0 K. W-16 CHAPTER 16 Table 16.5 Critical Temperatures for Selected Superconducting Materials Material Critical Temperature (K) Elements V5.4 Nb9.3 Ta4.5 Al1.2 Sn3.7 Pb7.2 Ti0.4 W 0.02 Compounds AlV311.8 AlNb.318 MgB239 GaV316.8 Nb3Sn18 Oxide compounds YBa2Cu3O72 90 Ba2 CaCu2O8Tl2120 Bi2CaCu2O8Sr2110 Based on data from Haynes, W.M., ed., Handbook of Chemistry and Physics, 92nd ed., CRC Press, Boca Raton, FL. (2011-2012), p. 12–60. The superconducting elements and compounds and their values of Tc in Table 16.5 demonstrate some of the interesting aspects of superconductivity. Although over 30 elements have been found to be superconductors, the metals with the highest conductivities, such as copper, silver, and gold, are not superconductors. Also, the only metal alloys that are superconductors are compounds. Solid solutions are never superconductors. Once the model of superconductivity in metals is presented, we can explain why solid solutions with randomly placed substitutional or interstitial atoms are not superconductors. The materials with the highest critical temperatures are not metals; they are compound ceramic oxides. Superconductivity has also been found in some polymers at temperatures up to approximately 12 K. A theory for superconductivity in metals was developed by Bardeen, Cooper, and Schrieffer (BCS) in 1957. In the BCS theory, pairs of negative free electrons with spin up and spin down are coupled together through interaction with the crystal positive ion cores. Figure 16.8 is a schematic of the interaction. As an electron, such as the spin-up electron in Figure 16.8, moves through a solid that has strong electron-ion coupling, there is an attraction between the positive ion cores and the negative electron. As a result of the electron passing through the solid, the positive ions are displaced in the direction of the electron. The electron has a much higher velocity than the positive ion cores do, and as a result the ion cores are displaced in the direction of the spin-up electron after the electron has already passed. The displaced positive ion cores attract the spin-down electron. The spin-up electron coupled with spindown electron through the displacement of the positive ion cores is called a Cooper pair of electrons. At temperatures below the critical temperature, all of the free electrons in the crystal are coupled together through the lattice of positive ion cores into Cooper pairs of electrons and the metal is a superconductor. W-17 Electrical Properties of Materials s = 1/2 + + ve – + + + + + + + + ve – + + s = –1/2 Figure 16.8 A schematic of the coupling of a Cooper pair of electrons (2) of opposite spin (s) with velocity ve by interaction with the crystal lattice of positive ion cores (1) in a superconductor. At temperatures below the critical temperature, the binding energy of the Cooper pair of electrons through the displaced ion cores is greater than the thermal energy of the vibrating ions, which is equal to approximately 3kT/2, and the Cooper pairs of electrons are not scattered by the ions vibrating in the crystal. The stronger the coupling between the Cooper pair of electrons, the higher is the critical temperature. The material becomes a normal conductor upon heating to temperatures above the critical temperature when thermal vibrations of the atoms have sufficient energy to break the coupling between the Cooper pair of electrons. The coupled pairs of electrons are like the coupling of automobile or bike racers, where the second racer gets in the draft of the first racer, and this creates coupled pair. All of the racers become coupled into one group where all of the racers travel at the same speed. When the electrons of spin 112 and 212 form a Cooper pair, they no longer behave as individual particles with respective spins of 112 and 212. The Pauli exclusion principle applies to individual electrons with spins of 112 and 212; it does not apply to the Cooper pair. Therefore, in a superconductor, Cooper pairs of electrons can have the same set of quantum numbers and move with the same velocity. Now with the BCS theory we can explain some of the observations about superconducting materials. Metals with a high electrical conductivity, such as copper, silver, and gold, are not superconductors, because the coupling between the electrons and positive-ion cores is very weak. Superconductivity requires a strong coupling between electrons and positive-ion cores. The superconductors are materials such as compounds that have strong ion-electron coupling. Why are alloys with random substitutional or interstitial atoms not superconductors? It is because the substitutional or interstitial atoms scatter the Cooper pairs of atoms at any temperature, destroying the coupling of pairs of electrons. When an ordered compound is in the superconducting state, the distance between the Cooper pair of electrons adjusts to the regular spacing of the compound atoms and the Cooper pair is not scattered. The higher the critical temperature, the easier it is to make superconducting products. Room temperature is a desirable critical temperature; however, this has not been achieved to date. Superconductivity has been achieved in some of the oxides at temperatures above the liquid nitrogen equilibrium vaporization temperature of 77 K. The oxide superconductors are commercially viable in some applications because liquid nitrogen is of reasonable cost. High-temperature superconductors have not achieved widespread use, because it is difficult to make a durable wire of ceramic material that can carry a high current density. Ceramic polycrystalline superconductor wires are brittle and subject to cracking; therefore, they are encased in a metal such as silver. If the superconducting wire breaks, the silver casing can carry current. Also, polycrystalline superconductor wires carry a lower current density than single-crystal superconductors, because it is difficult for the electrons to cross the nonsuperconducting grain boundaries. Because of the defects at the grain boundary of a superconductor, grain boundaries are not normally superconducting. In spite of these problems, the city of Detroit, Michigan, in the United States has installed 400 feet of high-temperature superconducting cable made of YBa2Cu3O72 that can deliver 100 million watts of power. Superconducting cable has the potential of delivering more watts per unit area of cable without any loss of power over the length of the cable. In YBa2Cu3O72 the indicates that there must be some vacant oxygen sites for there to be superconductivity. W-18 CHAPTER 16 Single-crystal thin films of superconductors can carry very high current densities of approximately 107 A/cm2. It is possible that an early production application of high-temperature superconductors will be for single-crystal, thin-film interconnection applications, such as in supercomputers, that require a high current density and can be cooled to liquid nitrogen temperatures. The production of magnetic fields with superconductors provides the interesting phenomena of magnetic levitation (maglev). In maglev a magnetic field is used to lift an object. We can observe maglev when two similar poles of two magnets are brought together; they repel each other. Maglev trains are lifted and propelled by magnetic fields. Maglev trains use superconducting coils to produce the magnetic fields. Chapter 17 covers superconductors in relation to magnetic fields and maglev trains, and the introductory photograph to Chapter 17 is of a maglev train. 16.5 Electrical Conductivity of Semiconductors and Insulators 16.5.1 Room-Temperature Electrical Conductivity of Intrinsic Semiconductors The electrical conductivity of an intrinsic semiconductor is the same as the electrical conductivity of a pure semiconductor. At high temperatures, semiconductors with impurity atoms have the same electrical conductivity as a pure semiconductor, and these semiconductors are then considered to be intrinsic semiconductors, as we will discuss in Section 16.7.3. As shown in Table 16.1, semiconductors have an electrical conductivity that is typically from 1026 to 10215 times the electrical conductivity of metals such as copper. Table 16.6 lists some of the different semiconductors and their intrinsic electrical conductivities at room temperature. Table 16.6 The Energy Gap, Electrical Conductivity, Electron and Hole Mobility, and Density of Carriers at Room Temperature for Some Typical Intrinsic Semiconductors Energy Electrical Electron Hole Intrinsic Gap Conductivity Mobility Mobility Carrier Material (eV) (V ? m)21(m2/V ? s) (m2/V ? s) Density (m23) Elements Si 1.114 3 1024 Ge 0.672.2 0.14 0.05 1.0 3 1016 0.38 0.18 2.4 3 1019 III-V Compounds GaAs1.42 1026 0.85 0.452 3 1012 InSb 0.172 3 10 7.7 0.07— 4 II-VI Compounds CdTe 1.44 — 0.12 0.005— ZnTe2.26 — 0.03 0.01 — The Data is from a Variety of Sources Electrical Properties of Materials The electrical conductivity of intrinsic semiconductors is related to breaking the covalent bonds, because if the covalent bonding is perfect, the covalently bonded electrons do not migrate and the electrical conductivity is 0. Figure 16.9a shows the covalent bonding in Group IV elements, such as silicon. Two electrons are localized between each pair of silicon atoms in a covalent electron pair bond, as we discussed in Chapter 2. Each silicon atom is surrounded by four other silicon atoms, and between each of these four atoms and the central silicon atom is a pair of covalently bonded electrons, resulting in a total of eight electrons surrounding each silicon atom. Figure 16.9 is a planar view of the bonding; the atoms are actually in a tetrahedral arrangement. The electrical conductivity of the semiconductor in Figure 16.9a is 0. However, as shown in Figure 16.9b, if an electron is excited out of the covalent bond, a free electron similar to the free electrons in a metal is created. All of the discussion of the electrical properties of free electrons in metals applies to a free electron in a semiconductor. The free electron can be conducted because it is free of the covalent bond, but it is bound inside the semiconductor. To maintain charge neutrality, a positive charge called a hole remains on the atom where the negative electron is excited from the covalent bond. The excitation of the electron out of the covalent bond can be due to thermal vibrations, irradiation with photons, energetic electrons, or other energetic particles. +4 +4 +4 +4 +4 +4 +4 +4 +4 (a) T = 0 K +4 +4 +4 +4 + +4 +4 +4 +4 – +4 (b) T > 0 K E +4 +4 +4 + – +4 +4 +4 +4 +4 +4 (c) T > 0 K Figure 16.9 (a) A two-dimensional schematic of the covalently bonded electrons (a pair of dashes) in silicon. (b) The creation of a free conduction electron (red circle with 2 sign) by thermal excitation and the resulting hole (yellow circle with 1 sign). (c) The application of an electric field (E) directed to the right moves the free electron to the left, and the hole migrates to the right by diffusion through covalent bonds. W-19 W-20 CHAPTER 16 This chapter considers excitation only by thermal vibrations, and Chapter 18 discusses how photons can excite covalently bonded electrons. An electric field (E) applied to the silicon produces a force on both the free conduction electron and the hole. The conduction electron is accelerated in a direction opposite to the electric field, and the hole is accelerated in the direction of the electric field. The conduction mechanism for the free electron in a semiconductor is the same as for a free electron in a metal. The velocity of the electron is increased by the applied electric field, creating a drift velocity of the free electrons. The mechanisms for electrical resistivity for the free conduction electrons in a semiconductor are the same as for the conduction electrons in metals; it is the scattering of conduction electrons by defects and lattice vibrations. The conduction mechanism for the hole is similar to vacancy diffusion. The hole moves when an electron from a neighboring covalent bond jumps or diffuses to the hole. This neutralizes the hole, and the jumping electron leaves behind another hole. In intrinsic semiconductors the electrical conductivity is much lower than is observed in a metal. For example, at room temperature the electrical conductivity of OFHC copper is 6.0 3 109 (V ? m)21 and the electrical conductivity of pure silicon is 4.0 3 1024 (V ? m)21. This is a difference of 13 orders of magnitude in electrical conductivity. Elemental semiconductors come from Group IV of the periodic table. Silicon and germanium are intrinsic elemental semiconductors, as shown in Table 16.6. In addition, there are compound semiconductors, such as the III-V (GaAs) and II-VI (CdTe) compounds. The Roman numbers III-V and II-VI are the group numbers of the component atoms in the compound semiconductor, and the sum of the group numbers is 8 or the average group number is 4. Eight is the same number of valence electrons in the stable inert gas atoms. Compound semiconductors are utilized because of their unique properties. For example, GaAs has a very high electron mobility, as shown in Table 16.6. Figure 16.10 presents the band structure model of the conduction process in a pure elemental semiconductor. In semiconductors and insulators, the covalently bonded electrons form the valence band (VB). At 0 K the VB is filled with the covalently bonded electrons, as indicated by the blue shading in Figure 16.10a, and the maximum energy electron in the VB is Ev. In the energy band model, there is an energy gap (Eg) between Ev and the lowest possible energy conduction electron at energy Ec. The energy gap is the minimum energy required to excite a valence electron into the conduction band. Table 16.6 presents the energy gap for some semiconductors. The conduction electrons are in the conduction band of energies. The conduction band (CB) contains the possible energy levels for conduction electrons; however, the energy levels are not necessarily occupied with conduction electrons. For example, at 0 K there are no conduction electrons in the CB, as indicated by the white background in Figure 16.10a. The intrinsic semiconductor is an insulator at 0 K, because there is no charge that can be conducted. As the E Eg Energy – – Ec + (a) (b) Conduction Band (CB) + Ev Valence Band (VB) (c) Figure 16.10 (a) Electron band structure of an intrinsic semiconductor at 0 K. The VB is full as indicated by the full blue shading. (b) Electron band structure at a temperature above 0 kelvin showing an electron (red circle with a 2 sign) excited into the CB from the VB, leaving behind a hole (yellow circle with a 1 sign). (c) The conduction electron and the hole both are both accelerated by an applied electric field (E), but in opposite directions. Electrons in the CB are at a minimum energy Ec; electrons in the VB have a maximum energy of Ev, and the energy gap is equal to Eg. W-21 Electrical Properties of Materials temperature increases, there is an increasing probability that electrons are thermally excited from the covalent bonds in the VB into the CB, as shown in Figure 16.10b by the red circle with a minus 2sign, and a hole is created in the valence band as indicated by the yellow circle with the 1 sign. The free conduction electrons have a probability of being found anywhere in the semiconductor, and all of the free electrons have a probability of interacting with each other. Therefore, the conduction electrons obey the Pauli exclusion principle and Fermi-Dirac statistics. In semiconductors, the CB is similar to the band of energies for metals shown in Figure 16.2. The negative electron in the CB and the positive hole in the VB are both accelerated by an applied electric field (E). Equation 16.7 calculates the electrical conductivity due to electrons for an intrinsic semiconductor; however, a term for the hole conductivity must be added, as shown in Equation 16.25: 5 nen 1 pep 16.25 where p is the hole density and p is the hole mobility. For intrinsic semiconductors, the number of conduction electrons and holes is equal because the electrons and holes are created in the same event. Table 16.6 presents the density of intrinsic electrons and holes (n 5 p) at room temperature for some semiconductors. In summing the electron and hole conductivities in Equation 16.25, the value of e is the magnitude of the charge on an electron, and the current resulting from an electron going in the opposite direction to the electric field adds positively to the current resulting from a hole going in the direction of the electric field. The electron and hole mobility (n and p) are material properties that are dependent upon the type of semiconductor, temperature, and defect density. Table 16.6 presents typical room-temperature electron and hole mobility. If a semiconductor has a high number of defects, such as impurity oxygen atoms or dislocations as a result of poor crystal growth procedures, the mobility of the electrons and holes in the semiconductor is reduced. For conduction electrons in a semiconductor, the distance between scattering events is reduced by defects, as we discussed for metals in Section 16.2, and holes are trapped at negatively charged defects. Different semiconductors have different intrinsic mobility. For example, GaAs is a compound semiconductor with an electron mobility of 0.85 m2/V ? s in comparison to the electron mobility of 0.14 m2/V ? s for silicon. Electrons are six times as mobile in GaAs as in silicon. This makes GaAs a desirable semiconductor for high-speed electronic applications, such as fast computers; however, GaAs is more expensive than silicon because it is more difficult to produce. Example Problem 16.3 With the values of mobility and carrier density given in Table 16.6, calculate the electrical conductivity of pure germanium at room temperature. Solution The electrical conductivity is calculated with Equation 16.25 for elecrons (n) and holes (p). 5 nen 1 pep 5 2.4 3 1019 1 n C 1.60 3 10219 3 m n 210.38V ? s2 1 12.4 3 10 m 211.60 3 10 Cp210.18 Vm? s2 m2 19 p 3 5 1.5(V ? m)21 1 0.7(V ? m)21 5 2.2(V ? m)21 This is in agreement with the measured value of 2.2(V ? m)21 in Table 16.6. 2 219 W-22 CHAPTER 16 16.5.2 The Effect of Temperature upon the Electrical Conductivity of Intrinsic Semiconductors As is indicated in Figure 16.10, at 0 K there are no conduction electrons or holes in a covalently bonded material. However, as temperature is increased there is an increasing probability, given by the FermiDirac statistical distribution, that an electron is excited from the VB into the CB, creating a negative conduction electron and a positive hole. The Fermi-Dirac statistical distribution given in Equation 16.18 shows that this probability increases exponentially with temperature; therefore, the density of both holes and electrons increases exponentially in an intrinsic semiconductor, and the electrical conductivity given in Equation 16.25 increases exponentially. The temperature dependence of the electrical conductivity of an intrinsic semiconductor is the opposite to that of a metal whose electrical conductivity decreases with increases in temperature. In a metal, there is no change in the conduction electron density with temperature, and the mobility of the electrons in a metal decreases with increases in temperature. The mobility of electrons in a semiconductor also decrease with increases in temperature for the same reason that it does in a metal. Intrinsic semiconductors are not utilized in electronic devices. Electronic devices utilize semiconductors that are modified with impurity atoms, as we will discuss in Sections 16.5.4 and 16.5.5. 16.5.3 Electrical Conductivity of Insulators An insulator has a high electrical resistivity and a low electrical conductivity because of the small density of conduction electrons and holes. The density of conduction electrons and holes is small at room temperature if the energy gap is large, because then there is a low probability for electrons in the VB to be thermally excited into the CB. For a pure insulator with a large energy gap (Eg ), the number of conduction electrons and holes is smaller than for an insulator with a small energy gap. This is demonstrated in Table 16.6, where the density of intrinsic carriers at room temperature in germanium with a band gap of 0.67 eV is 2.4 3 1019 m23; and for GaAs with a band gap of 1.42 eV the intrinsic carrier density is 2.0 3 1012 m23. The term carrier refers to the particles that transport charge. In an intrinsic semiconductor, the carriers are both electrons and holes of equal number. The difference in carrier density between Ge and GaAs is then reflected in the conductivities that are 2.2 (V ? m)21 and 1026(V ? m)21, respectively. The electrical conductivity of an insulator also obeys Equation 16.25. Insulators have a large energy gap. For example, diamond has the same crystal structure and bonding type as silicon, but diamond has a very large energy gap of 5.5 eV in comparison to the energy gap of 1.1 eV for Si and other semiconductors shown in Table 16.6. This results in pure diamond having one of the lowest conductivities of all materials at room temperature. Diamond is an excellent insulator. The electrical conductivity of insulators increases exponentially as temperature increases, as predicted by the Fermi-Dirac statistical distribution. 16.5.4 The Electrical Conductivity of n-Type Extrinsic Semiconductors Integrated circuits are not made from intrinsic semiconductors; they are made from extrinsic semiconductors. An extrinsic semiconductor is one whose electrical properties are dependent upon the addition of impurity atoms (doping) to a pure host crystal. An extrinsic n-type semiconductor is one that has more negative charge carriers (free electrons) than holes. An n-type semiconductor is made, for Electrical Properties of Materials example, with silicon from Group IV by adding impurity atoms from a higher-number group, such as phosphorus or arsenic from Group V. In the bond model of an n-type semiconductor, shown in Figure 16.11a, a Group V substitutional atom contributes four electrons to the covalent pair bonds of the host silicon crystal, but the Group V atom has an additional fifth valence electron. This fifth electron is not in a covalent bond, because all of the eight bonding orbitals in the host silicon crystal are satisfied by four of the five valence electrons of the Group V atom. In a host silicon crystal with dopant atoms of phosphorus, the fifth valence electron is bonded to the phosphorus atom by a weak electrostatic bond of 0.045 eV that maintains electrical neutrality, as shown in Table 16.7. If the fifth electron is removed from the Group V atom by thermal energy, it is a free conduction electron, as shown in Figure 16.11b, and the Group V atom becomes a positively charged ion. The positive charge on the Group V ion is not conducted unless the ion diffuses. If an electric field (E) is applied to the n-type semiconductor, as shown in +4 +4 +4 +4 +4 +5 +4 +4 +4 – Bound electron (a) T = 0 K +4 +4 +4 +4 +4 +5 +4 +4 +4 – Free electron (b) T > 0 K E +4 +4 +4 +4 +4 +5 +4 +4 +4 – (c) T > 0 K Figure 16.11 The electron bond model of an n-type semiconductor. (a) When a Group V atom, such as phosphorous, with five valence electrons substitutes for a silicon (Si) atom, only four of the group V atom’s valence electrons are in covalent bonds (pairs of dashes) in the silicon lattice. The fifth valence electron (white circle with a 2 sign) on the Group V atom has a weak electrostatic bond. (b) The fifth valence electron on the Group V atom is excited by thermal vibrations to become a free electron (red circle with a 2 sign) in the CB. (c) An applied electric field (E) accelerates the conduction electron in a direction opposite to the direction of the applied electric field. W-23 W-24 CHAPTER 16 Table 16.7 Donor and Acceptor Impurity Levels in Silicon and Germanium Silicon Germanium Dopant Ec 2 Ed Ea Ec 2 Ed P0.045 0.0120 As0.049 0.0127 Sb0.039 0.0096 Ea B 0.0450.0104 Al0.0570.0102 Ga0.065 0.0108 In 0.1600.0112 Based on data from Askeland, D.R., Fulay, P.P., and Wright, W.J., The Science and Engineering of Materials, 6th ed., Cengage Learning, Stamford, CT. (2011), p. 738. Figure 16.11c, the free electron is accelerated by the electric field in a direction opposite to the direction of the electric field. Figure 16.12 shows the band model of an n-type semiconductor. The energies associated with the VB and the CB are the same as in the intrinsic semiconductor. The energy level of the fifth valence electron bound to the Group V atom is the donor level (Ed). The electron bound to the Group V atom has a probability of being excited (donated) into the CB by thermal vibrations to provide conduction electrons, as shown in Figure 16.12b. The conduction electrons are accelerated by an applied electric field, as shown in Figures 16.11c and 16.12c. Some values of the donor level relative to the CB (Ec 2 Ed) are given in Table 16.7 for some different donor atoms in silicon and germanium. It is possible for an electron in the CB to move back to a donor atom or to recombine. When recombination happens, the positive and negative charges of the ionized donor atom and the conduction electron are eliminated. Recombination is an ongoing process, since this lowers the system energy by releasing an amount of energy equal to Ec 2 Ed. However, a donor atom is again ionized, and the electron is excited into the CB by thermal vibrations to produce an equilibrium number of conduction electrons. Energy E – – – – – – – + – CB – – – + – Ec Ed Ev VB (a) (b) (c) Figure 16.12 (a) The occupation of electron energy levels for an n-type semiconductor at 0 K. The VB is filled up to the energy Ev, as indicted by the blue shading. The donor level at energy Ed is full of electrons, as indicated by the symbols ⊝. The CB beginning at energy Ec is empty. (b) At a temperature above 0 K, donor level electrons are excited into the CB, leaving behind a positive charge (∙) on the donor atom. (c) An electric field (E) accelerates the free electron in the CB (red circle with a 2 sign). W-25 Electrical Properties of Materials Mobility (m2/ V-s) 100 10–1 n-GaAs 10–2 n-Si p-GaAs p-Si –3 10 10–4 20 10 1022 1024 Impurity density (/m3) 1026 Figure 16.13 Mobility of electrons (n) and holes (p) in silicon and GaAs, as a function of impurity atom density at 295 K. (Based on Kwok, H.L., Electronic Materials, PWS Publishing (1997), p. 111.) The n-type semiconductors are normally designed to operate in the extrinsic temperature range, where the density of conduction electrons is constant. In the extrinsic temperature range, all of the donor atoms are ionized, because the energy difference between the donor level and the CB (Ec 2 Ed) is very small, such as 0.01 eV. Also, there is an insignificant density of electrons excited from the VB into the CB across the energy gap, because the energy gap is large (approximately 1 eV) in comparison to 0.01 eV. For example, in germanium doped with 6 3 1021 arsenic atoms per cubic meter, the extrinsic range of operation extends from approximately 50 to 300 K. At temperatures higher than 300 K, a significant number of electrons are excited from the valence band into the conduction band. In the extrinsic range of operation of an n-type semiconductor made from an element in Group IV with donor atoms from Group V, the concentration of electrons in the CB (n) is equal to the concentration of donor atoms (Nd), as shown in Equation 16.26. n 5 Nd 16.26 In Equation 16.26, we assume that each donor atom is singly ionized and that there are no holes in the VB ( p 5 0). The mobility of electrons at the temperature and the donor concentration of the semiconductor are needed for the electrical conductivity calculation in Equation 16.25. Figure 16.13 shows that the mobility of electrons in n-type semiconductors at room temperature decreases as the impurity concentration increases. This is because the ionized donor atoms scatter the conduction electrons, thereby decreasing the electron mobility. The electrical conductivity of an n-type semiconductor is calculated with Equation 16.25 at any temperature within the extrinsic range of operation where n 5 Nd and p 5 0, if the electron mobility is known for that temperature and donor concentration (Nd). Compound semiconductors, such as GaAs, are also made n-type. Gallium is in Group III and arsenic is in Group V. If there are equal numbers of gallium and arsenic atoms, then each atom shares eight electrons, and the semiconductor is intrinsic. GaAs is made n-type by replacing As from Group V with an element such as sulfur from Group VI. Example Problem 16.4 An n-type semiconductor is produced by doping silicon with 10 parts per billion phosphorous, which is 10 atoms of phosphorous for 109 atoms of silicon. Calculate the electrical conductivity for this semiconductor at room temperature using the data in Figure 16.13. Assume that the semiconductor is operating in the extrinsic range. W-26 CHAPTER 16 Solution Phosphorous is in Group V of the periodic table; thus each donor atom contributes one electron to the CB of silicon. Assume that the phosphorous atoms are all ionized at room temperature, which allows us to calculate the electron density from the impurity composition. Nd 5 1 2 1 2 2.33 3 103 kg of Si 10 P atoms 1 n 6.02 3 1023 Si atoms 1 mole of Si 23 9 m3 10 Si atoms P atom 1 mole of Si atoms 28.09 3 10 kg of Si n 5 Nd 5 4.99 3 1020 n m3 Taking the value of the electron mobility from Figure 16.13 to be 0.1 m2/V ? s results in 5 nen 5 4.99 3 1020 1 C n 1.60 3 10 219 m3 n 21 0.1 2 m2 5 8 sV ? md 21 V?s 16.5.5 The Electrical Conductivity of p-Type Extrinsic Semiconductors Semiconductor devices are made by combining n-type and p-type extrinsic semiconductors. A p-type extrinsic semiconductor is one that has more positive charge carriers (holes) than negative charge carriers (electrons). An example of a p-type extrinsic semiconductor is Group IV silicon doped with an element of a group number less than Group IV, such as the Group III elements boron, aluminum, or gallium. A p-type semiconductor operating in the extrinsic temperature range has a constant density of holes in the VB equal to the density of dopant atoms. As schematically shown in Figure 16.14a, when a Group III element substitutes for a silicon atom, the Group III atom provides only three covalentbonding electrons (dashes) to share with the four silicon atoms that surround the Group III atom in the silicon crystal. This leaves one vacant electron orbital (missing dash) in the covalent bonds surrounding the Group III atom; however, the Group III atom in Figure 16.14a is electrically neutral. At temperatures above 0 K, it is possible for a covalently bonded electron from a neighboring silicon atom to diffuse into the previously vacant electron orbital at the Group III atom, as shown in Figure 16.14b. This results in the Group III atom having an extra covalently bonded electron charging it negative. A positively charged hole (yellow circle with a 1 sign) is associated with the silicon atom that is the source of the extra electron on the Group III atom. The covalent molecular orbitals characteristic of the silicon crystal exist at every atom site in the silicon crystal, even if an atom site is occupied by a Group III atom. The extra electron associated with the negatively charged Group III ion is a covalently bonded electron and not a mobile electron. In Figure 16.14c, an electric field (E) is applied to the p-type semiconductor, and the hole migrates in the direction of the electric field by diffusion through covalent bonds. The energy of the fourth covalently bonded electron on the Group III atom is the acceptor level (Ea), as shown in the band model of Figure 16.15. Figure 16.15a shows that at 0 K, all of the electrons in the p-type semiconductors are in the VB. In Figure 16.15b, at temperatures above 0 K it is possible for an electron to be excited into the acceptor energy level indicated by the symbol ⊝, and a hole is created in the VB, as indicated by the yellow circle with the 1 sign. The acceptor level (Ea) is a little higher in energy than the VB of the silicon atoms, because the fourth electron on the Group III ion is covalently bonded, Electrical Properties of Materials +4 +4 +4 +4 +4 +3 +4 +4 +4 (a) T = 0 K +4 +4 +4 +4 +4 +3 +4 +4 +4 + (b) T > 0 K E +4 +4 +4 +4 +4 +3 +4 +4 +4 + (c) T > 0 K Figure 16.14 The electron bonding model of a p-type semiconductor. (a) A Group III atom with three valence electrons substitutes for a silicon (Si) atom, and there is one missing covalently bonded electron in the silicon crystal lattice, indicated by the missing dash. The Group III atom is neutral in charge. (b) A covalently bonded electron from a neighboring silicon atom diffuses to the Group III atom, creating a positive charge (hole) indicated by the yellow circle with a 1 sign where the electron is missing from the silicon atom. There is a negative charge at the Group III atom (acceptor atom), where there are now eight covalently shared electrons. (c) The hole with a positive charge is accelerated by the applied electric field (E) in the direction of E, but the covalently bonded negative charge on the Group III atom is not mobile. but this fourth electron charges the Group III atom negatively rather than being neutral. Table 16.7 presents some acceptor levels in silicon and germanium semiconductors. It is possible for the electron on the acceptor atom to move into a hole in the VB; when this recombination happens, the positive and negative charges of the hole and acceptor electron cancel, and the two charges are eliminated. Recombination in p-type semiconductors is an ongoing process, since this lowers the system energy by releasing an amount of energy equal to the energy of the acceptor level above the top edge of the VB. However, other holes in the VB are created by thermal vibrations exciting VB electrons to the acceptor level to produce an equilibrium number of holes. An applied electric field (E) produces a force (Ee) on the hole, and the positively charged hole moves in the direction of the electric field by diffusion, as shown in Figure 16.14c and 16.15c. W-27 W-28 CHAPTER 16 E Eg CB Energy Ec (a) – + – (b) (c) Ea + Ev VB Figure 16.15 (a) The occupation electron energy levels in a p-type semiconductor at 0 K. (b) At a temperature above absolute zero an electron is excited from the VB to the acceptor level (Ea) indicated by ⊝ associated with a group three atom leaving a hole in the VB (yellow circle with a 1 sign). (c) An applied electric field (E) accelerates the hole in the VB in the direction of the electric filed, but there is no conductivity of the covalently bonded electron (⊝) in the acceptor level. Equation 16.25 is used to calculate the electrical conductivity of a p-type semiconductor, and simplifications are possible in the extrinsic range of operation. The energy difference between the acceptor level and the VB (Ea 2 Ev) is very small, such as 0.01 eV, and the energy gap is large in comparison (approximately 1 eV). Normally in semiconductor analysis the energy Ev is set equal to 0. In the extrinsic temperature range of operation of a p-type semiconductor, all of the acceptor energy states are filled, and there is an insignificant density of electrons excited from the VB into the CB across the energy gap. For example, in silicon doped with 6.5 3 1023 boron atoms per cubic meter, the extrinsic range of operation extends from approximately 200 to 700 K. In the extrinsic range of operation of a p-type semiconductor made from a doped Group IV element, the concentration of holes in the VB (p) is equal to the concentration of acceptor atoms from Group III (Na), as shown in Equation 16.27. p 5 Na 16.27 In Equation 16.27, we assume that each acceptor atom is singly charged for Group III substitutional atoms in silicon. The mobility of holes at the temperature and impurity concentration of the semiconductor is needed for the electrical conductivity calculation in Equation 16.25. At any temperature within the extrinsic range of operation, p 5 Na and n 5 0. The hole mobility must be known for that temperature and acceptor concentration (Na). Figure 16.13 shows that the mobility of holes in p-type semiconductors at room temperature decreases as the impurity concentration increases. A p-type semiconductor can be produced in compound semiconductors such as GaAs by replacing Ga from Group III by an element such as Be, Mg, or Zn from Group II. The Group II element provides the acceptor atoms. Example Problem 16.5 A semiconductor is produced by doping Group IV silicon with 1023 atoms per cubic meter of Group III gallium. Assume that at room temperature this semiconductor is in the extrinsic range of operation. (a) What type of semiconductor is this? (b) Calculate the room-temperature electrical conductivity of this semiconductor. (c) State any assumptions you made in calculating the electrical conductivity in part (b). Electrical Properties of Materials Solution a) Since the dopant is Group III gallium, the semiconductor is p-type. b) In the extrinsic range of operation, the number of holes is equal to the number of acceptor atoms, resulting in p 5 1023 holes/m3 and n 5 0. From Figure 16.13, the mobility of holes in p-type silicon is approximately 1.5 3 1022 m2/V ? s. From Equation 16.25, is the electrical conductivity of a p-type semiconductor is 5 pep 5 1 3 1023 1 2 p C m2 5 2.4 3 102 sV ? md 21 1.60 3 10 219 1.5 3 10 22 3 m p V?s c) Assume n 5 0 and p 5 1023 atoms/m3. 16.6 Polymer Insulators, Semiconductors, and Conductors Most polymers such as polyethylene (PE) are insulators; PE has one of the highest values of electrical resistivity, as shown in Table 16.1. The model of insulators, such as diamond, that we discussed in Section 16.5.3 also applies to polymers. The covalently bonded electrons are in the VB, and electrons must be excited out of covalent bonds to be conducted. PE has a large energy gap of 9 eV. Therefore, at room temperature there is an insignificant density of conduction electrons and holes in pure PE. PE and other polymers with high electrical resistivity are combined with conductive powders, such as graphite or silver, to produce composite materials with increased electrical conductivity that are utilized in resistors. The conductive powder is a second phase; it does not chemically combine with the polymer. The polymer binds the conductive powder particles together into a composite solid. A current can pass from conductive particle to conductive particle if the particles touch, in which case the electrical resistivity is given by the rule of mixtures in Equation 16.13. However, if the particles do not touch, conduction electrons must penetrate (tunnel) through the polymer in the gap between the conductive particles to provide electrical conductivity. The gaps between particles result in a much higher electrical resistivity than when the conductive particles touch. When the particles do not touch, the electrical resistivity of the polymer-conductor composite material increases exponentially with increase in distance between the conductive particles. A conductive composite polymer with metal particles has recently been developed that has potential biomedical applications as a skin material. This composite polymer is electrically conductive, it heals if torn or cut, and it is sensitive to pressure. The polymer is produced by combining an olgiomer with urea [CO(NH2)2]. An oligomer has only a few units, in contrast to a polymer, which has many units. Oligos in Greek means “a few.” Urea links the oligomer into a network polymer in a manner similar to how formaldehyde reacts with phenol molecules to form phenolformaldehyde (PF), as shown in Figure 7.28. If the bond formed by the urea between the olgiomers is broken by cutting or tearing, the bond can reform. However, broken PF bonds do not re-form. Micron-sized particles of nickel (Ni) are added to the network polymer to provide electrical conductivity. There is pressure sensitivity because compression forces the μNi particles closer together. The shorter distance between μNi particles increases the electrical conductivity. An electrically conductive and pressure sensitive skin material can detect damage, and self-healing can repair the damage. Recently polymers such as polyacetylene have been found to be semiconductors and conductors. The electrical conductivity of undoped polyacetylene is approximately 1026 (V ? m)21, and for doped W-29 W-30 CHAPTER 16 polyacetylene it is approximately 104 (V ? m)21, depending upon the type and amount of dopant. As shown in Table 16.6, the electrical conductivity of intrinsic silicon is 4 3 1024 (V ? m)21. Polyacetylene has an energy gap of 1.8 eV in comparison to 1.1 eV for silicon. Figure 16.16a shows the molecular structure of polyacetylene with perfect bonding. PE has two hydrogen atoms per carbon atom. However, in polyacetylene each carbon atom has only one attached hydrogen atom. Because there is only one hydrogen atom per carbon atom, every other carbon atom along the polyacetylene chain forms a double bond with an adjacent carbon atom, giving the carbon atoms a total of eight shared electrons. Four electrons are shared between two carbon atoms in a double bond. If the bonds on polyacetylene are perfect, as shown in Figure 16.16a, polyacetylene is an insulator. If there are defects in the bonding, polyacetylene conducts electricity. In polyacetylene, defects occur in the bonding, as shown schematically in Figure 16.16b, where the alternation of single and double bonds is incorrect, and two single bonds repeat to form a soliton. The carbon atom at the soliton is electrically neutral if there is also an electron in a non–covalently bonded state indicated by the dot ( ) in Figure 16.16b. The non–covalently bonded electron can be moved to the right along the polyacetylene chain if one electron from the double bond to the right of the soliton moves to the left one spacing along the chain; this moves the soliton along with the non–covalently bonded electron one spacing to the right. The activation energy for this electron motion is estimated to be only 0.01 eV, and the current density is calculated with an Arrhenius-type rate equation. ? – C C C C H N C H C H C C C C H H H C C H C H H H H H C C N C C C H H H H C H C H H C C C H C C H H (d) H C C C C C H H H H C C C C H C C H (c) H + H H C C H H H C C C H (b) C C C C H H H C C C H H H H H H (a) H Figure 16.16 (a) The atomic structure of perfect polyacetylene. (b) The structure of neutral polyacetylene with a defect in the bonding (soliton); and a single electron in a nonbonding state, indicated by ( ), resulting in an electrically neutral chain. (c) A positive charge on polyacetylene (1) located at a soliton is created by the removal of the nonbonded electron. (d) Negatively charged polyacetylene has two nonbonding electrons, indicted by ( ), and a negative charge, indicated by (2). (Based on http://en.wikipedia.org/wiki/File:Trans-(CH)n.png) ? ?? Electrical Properties of Materials Dopants of iodine, chlorine, or fluorine oxidize polyacetylene and produce a p-type semiconductor with a positive charge by removing the non–covalently bonded electron from the soliton, as shown in Figure 16.16c. The dopant does not substitute for an atom along the chain, as in the case of doped silicon semiconductors. Rather, the dopant forms a second phase that is mixed with the polyacetylene. The non– covalently bonded electron is transferred from the polyacetylene to the second phase. The positive charge on the polyacetylene moves in the same way as described for the neutral soliton above. Dopants such as sodium and lithium from Group IA mixed with polyacetylene produce an n-type semiconductor by contributing electrons to the polyacetylene, as shown schematically in Figure 16.16d by the two dots ( ), which represent two electrons. The second non–covalently bonded electron at the soliton results in a negative charge, as shown in Figure 16.16d. The negative charge on the polyacetylene moves in the same way as is described for the neutral soliton above. The dopant is mixed with the polymer by several different processes. A solid polymer can be exposed to a gas of the dopant, and the gas diffuses into the polymer. Liquid dopant is also mixed with liquid polymer. If the dopant and the polymer are not soluble in each other, a two-phase mixture of polymer and dopant results upon solidification. Some possible advantages of polymer conductors and semiconductors relative to single-crystal silicon are sustainable supplies, reduced costs, mechanical flexibility, and transparency. There have been predictions that polymers with conductivities greater than that of copper can be developed. At very high levels of doping, greater than 6 atom percent, polyacetylene has an electrical conductivity comparable to that of some metals. There are conducting polymers other than polyacetylene; however, the simple bonding in polyacetylene allows for demonstration of the conduction mechanism in polymers. Electronic devices have been made from polymer semiconductors, and polymer semiconductors have already found applications such as cell phone display screens. ?? 16.7 Junctions and Electronic Devices 16.7.1 Introduction Crystalline semiconductor devices such as diodes, light-emitting diodes (LEDs), diode lasers, and photovoltaic solar cells are made from pn-junctions. The transistors that amplify a signal in an electronic circuit are npn-junctions. Diodes control the direction of the current flow in an electrical circuit. Diodes are also made from metal-polymer-metal (MPM) junctions; several uses of MPM-junctions are as LEDs and solar cells. 16.7.2 Metal-Metal Junctions Metal-metal junctions provide a foundation to understand pn-junctions and MPM-junctions. A metalmetal junction is formed when two different metals are in electrical contact. Two metals with different work functions (FA and FB) are shown separated in Figure 16.17a. The work functions of some metals are presented in Table 16.4. If metals A and B are brought into electrical contact, as in Figure 16.17b, electrons from metal B flow into metal A, as indicated by the arrow in Figure 16.17a, because there are lower unoccupied electron energy levels in metal A. A positive charge results on metal B, and a negative charge results on metal A. W-31 W-32 CHAPTER 16 Free space (a) FB FA EFB Electrons EFA Electrons FA – FB = eVc Free space (b) EF –+ –+ –+ –+ Metal A Metal B Figure 16.17 (a) The band structure of two separated metals A and B with work functions FA and FB. (b) The band structure of metals A and B in electrical contact forming a junction and producing the contact potential (VC). The horizontal dashed line is the energy of free space after contact. The buildup of charge on metals A and B produces a contact voltage (VC ) between metals A and B. Metals A and B are in equilibrium with each other when the Fermi energies of metals A and B are equal, as shown in Figure 16.17b. To remove an electron from the Fermi energy of metal B to the energy of free space requires an additional amount of energy equal to 0.5eVC relative to what is required before contact. In metal A, to remove an electron from the Fermi energy to the free-space energy level requires an amount of energy equal to 0.5 eVC less than is required before contact. Equation 16.28 is deduced from Figure 16.17b. FA 2 FB 5 eVC 16.28 Equation 16.29 then gives the contact voltage developed between metals A and B. VC 5 FA 2 FB e 16.29 Because the work function is the energy of free space minus the Fermi energy, Equation 16.30 is deduced from Figure 16.17a: FA 2 FB 5 EFB 2 EFA 16.30 where EFA and EFB are the Fermi energies of metals A and B. The contact voltage of the junction in terms of the Fermi energy of metals A and B is then given in Equation 16.31. VC 5 EFB 2 EFA e 16.31 The contact voltage between the two metals A and B is not a voltage in the sense of a battery. If a voltmeter is connected to a junction of two metals, the voltage reading is 0, because the electrons at the Fermi energy on either side of the junction are at the same energy. The contact voltage is an internal voltage associated with the junction interface. W-33 Electrical Properties of Materials Anyone who has metal fillings in their teeth has probably experienced the contact voltage when a fork has touched one of the fillings and they received a tingle in the tooth. This is the result of the electron transfer between the silver fork and the metal filling. The metal filling is a mercury-based amalgam of metals. Example Problem 16.6 Mag wheels for automobiles are an alloy of magnesium and aluminum. Magnesium has a work function of 3.68 eV. Mag wheels are bolted to the axle of a car that is made of an iron alloy that has a work function of 4.50 eV. (a) What contact voltage is developed where the iron alloy contacts the magnesium in the wheels? (b) Which metal is negative and which is positive? Solution The contact voltage is calculated with Equation 16.29 because the work functions are known. VC 5 FA 2 FB e 5 4.50 eV 2 3.68 eV 5 0.82 V 1e The value of e in this case is 1, since the energy unit of the work function is electron volts. If the work function is in joules, then the value of e is 1.6 3 10219 C. (b) Because the magnesium alloy has the smaller work function, electrons flow from this alloy to the iron alloy at contact. The magnesium alloy becomes positive and the iron alloy takes on a negative charge. The positive charge on the magnesium alloy can contribute to corrosion by Mg21 ions dissolving in an electrolytic solution, such as salt water. 16.7.3 pn-Junctions A pn-junction is produced by joining p-type and n-type semiconductors in electrical contact. Figure 16.18a shows the energy levels for separated p-type and n-type semiconductors, and Figure 16.18b shows the energy levels for a pn-junction. p-Type Fp Ea – – – – + + + + n-Type Fn – – – – + + + + Ec EF Eg EF Ev Free space Ecp Eo p-Type M Eg Donors in SCL Ed EF Evp n-Type – – – – E cn EF Acceptors in SCL ++ ––– – + + + + eVc ++ Evn (a) Bulk SCL Bulk (b) Figure 16.18 (a) Separated p-type and n-type semiconductors at a temperature in the extrinsic range of operation. (b) The pn-junction after contact, showing that the Fermi energy is continuous across the junction. The original location of the junction is called the metallurgical junction, and it is the vertical dashed line (M). W-34 CHAPTER 16 The values of the Fermi energy of the p-type and n-type semiconductors are necessary to calculate the contact potential of a pn-junction. The value of the Fermi energy in p-type and n-type semiconductors at 0 K is deduced by recalling that the Fermi energy has a probability of occupation of 12. In an n-type semiconductor at 0 K, the electrons are all on the donor atoms and none of the electrons are in the CB. Therefore, the donor energy level has a probability of 1 of being occupied, and the CB has a probability of 0 of being occupied. The Fermi energy (EF) is halfway between the donor energy (Ed) and the conduction-band energy (Ec), and EF is equal to (Ed 1 Ec )/2. In a p-type semiconductor at 0 K, the valence electrons are all in the VB in covalent bonds, and no electrons are excited to acceptor atoms. Therefore, the valence-band energy (Ev) has a probability of 1 of being occupied, and the acceptor energy (Ea) has a probability of 0 of being occupied. EF is halfway between the valence-band energy (Ev) and the acceptor energy (Ea), and EF is equal to (Ev 1 Ea)/2. If the p-type and n-type semiconductors are raised to a sufficiently high temperature, eventually there are many more electrons in the CB that are excited from the VB than there are acceptor atoms in the p-type semiconductor or donor atoms in the n-type semiconductor. At these high temperatures the p-type and n-type semiconductors have essentially the same number of electrons in the CB as a pure (intrinsic) semiconductor, and the p-type and n-type semiconductors both behave as an intrinsic semiconductor. In an intrinsic semiconductor the Fermi energy is always in the middle of the energy gap and equal to Eg/2. This is because at 0 K the VB is full of electrons and the CB has zero electrons. At 0 K in an intrinsic semiconductor, the probability of occupation of the energy level Ev is 1; the probability of occupation of the energy level Ec is 0, and the probability of 12 is halfway between Ev and Ec, or in the middle of the energy gap at Eg/2. As an intrinsic semiconductor is heated, any electron that is in the CB comes from the VB, any increase in probability at Ec is balanced by a decrease in probability at Ev, and the probability of 1 2 occupation remains at Eg/2. In an n-type semiconductor the Fermi energy is at (Ed 1 Ec)/2 at 0 K and decreases toward Eg/2 as temperature is increased, because at sufficiently high temperature the n-type semiconductor behaves as an intrinsic semiconductor. In a p-type semiconductor, the Fermi energy is at (Ev 1 Ea)/2 at 0 K and increases toward Eg/2 as temperature is increased, because at sufficiently high temperature the p-type semiconductor behaves as an intrinsic semiconductor. We assume in Figure 16.18 that the temperature is in the extrinsic range of operation for the p-type and n-type semiconductors; therefore, the Fermi energy of the p-type semiconductor is more than Ea/2 but less than Eg/2, and the Fermi energy of the n-type semiconductor is less than (Ed 1 Ec)/2 but more than Eg/2. After contact the Fermi energy, as shown in Figure 16.18b, is constant across the junction, as it is for metals A and B in Figure 16.17b. The energy bands in the n-type and p-type semiconductors shift with the Fermi energy. We can calculate the contact voltage for a pn-junction by using a modification of Equation 16.31, as presented in Equation 16.32. VC 5 Fp 2 Fn e 5 EFn 2 EFp 16.32 e The contact voltage results from electrons in donor levels and in the CB of the n-type semiconductor in the junction region moving across the junction to lower-energy unoccupied states in the VB of the p-type semiconductor in the junction region. This electron movement in the junction region results in a positive charge (∙) in the n-type semiconductor and a negative charge in the p-type semiconductor (⊝), as shown in Figure 16.18b. The separation of charge in the junction region is called the space charge layer (SCL). The SCL creates an internal electric field (E0 ) pointing from the positively charged n-type semiconductor to the negatively charged p-type semiconductor. The internal electric field is what bends the electron energy bands, resulting in a constant Fermi energy across the junction. Because the charges that produce the electric field are spread over a distance across the junction, the bending of the electron energy bands occurs over a distance at the junction, W-35 Electrical Properties of Materials as shown in Figure 16.18b, resulting in a width to the pn-junction. Away from the SCL, the p-type semiconductor has a hole density in the VB and the n-type semiconductor has an electron density in the CB characteristic of bulk semiconductors. 16.7.4 The Diode Equation for a pn-Junction A pn-junction in an electrical circuit is a diode. It allows current to flow in one direction in response to an applied voltage, but not in the other direction. Why the pn- junction behaves as a diode is explained by observing the changes that occur in the semiconductor electron energy band structure when a forward and a reverse bias (voltage) is applied to the pn-junction, as shown in Figures 16.19a, 16.19b, and 16.19c. In the upper part of Figure 16.19a, the band structure of a pn-junction is shown with no applied voltage, and the lower part of Figure 16.19a shows schematically the location of conduction electrons (red circles with a 2 sign) and holes (yellow circles with a 1 sign) in the pn-junction. When a forward-bias voltage (Va) is applied to the pn-junction, the p-type end of the junction is connected to the positive terminal of a battery, and the n-type end is connected to the negative terminal, as shown in Figure 16.19b. The forward bias forces electrons in the n-type semiconductor and holes in the p-type semiconductor toward the junction interface, as shown in the lower part of Figure 16.19b. The forward bias creates a positive current flow of holes from the p-type to the n-type semiconductor, and of electrons from the n-type to the p-type semiconductor. As shown in the upper part of Figure 16.19b, p n eVc Ecp – Ecp e(Vc-Va) – – – – – Ecn E1 E1 – – – –E cn EF Evp + + + + + EF eVa EFp Evp + + + + + + + EFn e(Vc-Va) Ecp E1 EFp Evp + + + + + – – – – Ecn EFn Evn + Evn Evn p p + + + + + + + + + n – – – – – – – + + + n + + + + + – – – – – – – + n – – – – – – – – I Va (a) p + + + + + + + + (b) Va + (c) Figure 16.19 (a) The band structure of a pn-junction with no applied voltage, showing the contact potential (Vc ) and the location of conduction electrons (red circles with 2 sign) and holes (yellow circles with 1 sign) in the band structure. The schematic below shows the physical location of conduction electrons holes. (b) The band structure of a pn-junction with a forward bias (Va), showing the displacement of the energy bands and the electron and hole motion at the junction. In the schematic below, the electrons and holes are shown driven toward the junction by the forward bias creating a positive current across the junction from left to right. (c) The band structure of a pn-junction with a reverse bias (Va) showing the displacement of the energy bands. In the schematic below, the electrons and holes are driven away from the junction, resulting in no current crossing the junction. (Adapted from Kasap, S. O., Principles of Electronic Materials and Devices, 3rd ed., McGraw-Hill, Boston MA. (2006), p. 496.) W-36 CHAPTER 16 the applied forward bias (Va) lowers the energy of the electrons in the p-type semiconductor, and it raises the energy of the electrons in the n-type semiconductor. This reduces the height of the energy barrier for electrons to migrate from the n-type to the p-type semiconductor at the pn-junction by an energy equal to Vae. The reduction in barrier height results in more electrons in the n-type semiconductor with sufficient energy to overcome the energy barrier that can flow to the p-type semiconductor. The electrons flowing from the n-type to p-type semiconductor and holes flowing from the p-type to n-type creates a positive current from left to right across the pn-junction. If a reverse bias is applied, as in Figure 16.19c, the p-type semiconductor is charged negative and the n-type semiconductor is charged positive. The reverse bias results in both the holes and electrons being forced away from the junction, as shown in the bottom part of Figure 16.19c. The applied voltage does not drive current through the junction. The electron current across the junction with a reverse bias depends upon the rate at which electrons are thermally excited from the Fermi energy to the CB of the p-type semiconductor (E1). The value of E1 is not affected by the application of the forward or reverse bias, because all of the electron energy levels in the p-type semiconductor shift together as a result of an applied voltage. If electrons are excited into the CB of the p-type semiconductor, they will flow down in energy to the n-type semiconductor, with or without an applied voltage. It is the rate of thermal excitation of electrons in the VB or the acceptor level by energy E1 into the CB of the p-type semiconductor that controls the reverse electron current flow, and the magnitude of the applied reverse voltage has no effect. A pn-junction allows current to flow with forward bias that is an exponential function of the applied voltage; however, with a reverse bias the current does not increase with voltage, as shown in Figure 16.20. The pn-junction functions as a diode; it allows current to flow in one direction in an electronic circuit but not in the reverse direction. The current (I ) through a pn-junction is expressed in Equation 16.33a, where I0 is the current with a large reverse bias. The current density through a pn-junction, in Equation 16.33b, has the same form as Equation 16.33a. Vae 16.33a Vae 16.33b 3 1 kT 2 2 14 I 5 I0 exp 3 1 kT 2 2 14 J 5 J0 exp An npn-junction is a transistor that amplifies an electrical current. In solid-state electronics the transistor has replaced the vacuum-tube amplifier. The operation of transistors is discussed in books that focus on the electronic properties of materials, such as those by Kasap and Kwok. I I0 –1.0 –0.5 12 10 8 6 4 2 –2 0.5 Voltage 1.0 Figure 16.20 A schematic of the current (I ) divided by the current with a large reverse bias I0, as a function of applied voltage for a diode (pn-junction). W-37 Electrical Properties of Materials Example Problem 16.7 A pn-junction diode at 300 K produces a current density of 0.50 3 104 A/m2 with a forward bias of 0.100 V. Calculate the current density for a reverse bias of 0.100 V. Solution We use Equation 16.33b, the diode equation, to solve this. Everything in Equation 16.33b is known except J0. Inserting the known values into Equation 16.33b for the forward bias allows solution of J0. Vae 3 1 kT 2 214 5 0.5 3 10 mA 5 J 3exp 18.62 3 10 0.1eVeV? K J 5 J0 exp 0.5 3 104 4 2 3 1 0 25 s300 Kd 21 2 214 2 4 A 0.1 eV 5 J0 exp 21 5 J0 [exp s3.86d 21g 5 J0 f47.5 2 1g 5 J0 f46.5g m2 0.026 eV Solving for J0, J0 5 0.5 3 104 A/m2 A 5 1.08 3 102 2 47.5 m The reverse current should be approximately 1.08 3 102 A/m2, and independent of applied voltage; however, we can check this with the diode equation by inserting the reverse voltage of 20.1 eV. 3 1 2 4 J 5 1.08 3 102 A 20.1 eV A exp 21 5 1.08 3 102 2 fexp s23.86d 21g m2 0.026 eV m J 5 1.08 3 102 A A A f0.02 2 1g 5 1.08 3 102 2 f20.98g 5 21.05 3 102 2 2 m m m The current with an applied voltage of 20.1 V is indeed approximately equal to 2J0. Example Problem 16.8 A pn-junction diode at 300 K produces a current density of 0.50 3 104A/m2 with a forward bias of 0.100 V. Calculate the current density for a forward bias of 0.200 V. Solution We use Equation 16.33b, the diode equation, to solve this. We have two unknowns, J2 and J0. We could solve for J0 as we did in Example Problem 16.7; however, because J0 occurs in the equations for both J1 and J2, if we take the ratio of J1 /J2, then J0 is eliminated, and we solve for J2 in terms of J1. First determine the equations for J1 and J2. Vae 3 1 kT 2 214 5 0.5 3 10 mA 5 J 3exp 18.62 3 10 0.1eVeV? K J1 5 J0 exp 4 2 3 18.62 3 10 J2 5 J0 exp 0 25 2 4 s300 Kd 21 3 1 2 214 5 J f46.5g 0 2 4 0.2 eV 0.2 eV 21 5 J0 exp 21 0.026 eV eV ? K21s300 Kd 25 J2 5 J0 f2191 2 1g 5 J0 f2190g W-38 CHAPTER 16 Now take the ratio J1/J2. J1 J2 5 J0 f46.5g J0 f2190g 5 0.5 3 104A/m2 J2 Now cancel J0 and solve for J2. J2 5 0.5 3 104 1 2 A 2190 A 5 23.6 3 104 2 m2 46.5 m The exponential dependence of the diode current density on the applied voltage is demonstrated by this calculation. 16.7.5 Metal-Polymer-Metal Junctions and Electronic Devices Electronic devices such as LEDs, solar cells and batteries, and visual displays have been produced from MPM-junctions. Chapter 18 covers the photonic uses of MPM-junctions. Several advantages of MPM-junctions over single-crystal semiconductor pn-junctions are that MPM-junctions are flexible, and they have the potential of being relatively inexpensive. Figure 16.21 shows a schematic of the energy levels of a MPM-junction. Figure 16.21a shows the metal-1 (M1), polymer, and metal-2 (M2) before contact. The work function of M1 (F1) is larger than that of M2 (F2), and the lowest energy in the polymer CB (Ec ) is above the Fermi energy of M1 and M2. After the junction is made and equilibrium established, as shown in Figure 16.21b, the Fermi energies of the two metals are equal. The equality of Fermi energies creates a contact voltage (Vc ) that results from a positive charge on M2 and a negative charge on M1, due to electron flow that goes from M2 to M1. The polymer is a thin layer, and sufficient charge is able to cross the polymer from M2 to M1 to establish equilibrium. As in the semiconductor pn-junction, an internal electric field is created across the polymer due to the positive charge on M2 and the negative charge on M1. The electric field bends the energy bands of the polymer, as shown in Figure 16.21b. The negative charge on M1 results in a lower energy for electrons in M1 to be excited to free space, and the positive charge on M2 results in a higher energy for electrons in M2 to be excited to free space. After equilibrium is established, electrons in M2 must overcome the energy barrier eVc to flow to M1. We can observe the diode action of this device by applying a forward bias voltage (Va ) to the MPM-junction such that M2 is charged negative and M1 is charged positive, as schematically shown in Figure 16.21c. The applied voltage reduces the energy barrier for electron flow from M2 to M1 provided by the polymer to e(Vc 2Va ). The Fermi energy of M2 indicated by the top of the purple band is the probability of occupation of 12 , and there are electrons with greater energy than the Fermi energy that enter the polymer conduction band. The positive current flow is in the opposite direction to the electron flow. With a reverse bias, M1 is charged negative and M2 is charged positive. The reverse bias has no effect on the reverse current flowing from M1 to M2, because the energy level of the CB of the polymer (Ec) increases, along with the Fermi energy of the electrons in M1 as the reverse bias voltage (Va) is increased. Therefore, the probability of an electron flowing from M1 to M2 is not increased by the reverse bias. Electrical Properties of Materials Free space Ec F2 F1 EF2 Ev EF1 Polymer Metal 2 Metal 1 (a) Before Contact Energy Ec eVc ––– +++ Polymer Metal 1 Metal 2 (b) After Contact – –E – – c Polymer Metal 1 + Va – Electrons Metal 2 I (c) Applied Bias = Va Figure 16.21 (a) Energy-level diagrams of separated metal-1 (M1) and metal-2 (M2) and a polymer, showing the metal Fermi energies EF1 and EF2, the work functions F1 and F2, and the polymer valence-band and conduction-band energies Ev and Ec. (b) The energy bands after contact and the contact potential Vc. (c) The energy bands after the application of a forward bias Va and the flow of electrons from M2 to M1 and the positive current I in the reverse direction. 16.8 Dielectric Materials What happens in a material when an electric field or voltage is applied and there is no conduction of electrons or holes? The center of the positive charge is displaced from the center of the negative charge. These materials are called dielectrics, with dia meaning “through” or “across.” A dielectric material allows penetration of an electric field through the material. Electric fields cannot penetrate metals. Supersecret electronic systems are enclosed in metal-lined rooms to prevent electric fields from entering or escaping. Examples of dielectric materials are diamond, sapphire, glass, quartz, salt, water, and polymers such as PE. Dielectric materials are utilized as insulators, capacitors, dynamic random-access memory (DRAM), permanent memory, electro-optical light switches, micro electromechanical actuators, sound and ultrasound transducers and detectors, thermistors, and sensors. W-39 W-40 CHAPTER 16 16.8.1 Capacitance Dielectrics are characterized by capacitance (C ), which is measured in farads (F), and 1 F is equal to 1 C/V. For the parallel-plate capacitor shown in Figure 16.22a, which has a vacuum between the plates, an applied voltage (V ), and a total charge (Q0) on the plates measured in coulombs (C ), the capacitance (C ) is given by Equation 16.34. Q0 C5 16.34 V If the vacuum capacitor has plates of area (A) with a distance between the plates of (l ), as shown in Figure 16.22a, the capacitance is also given by Equation 16.35: Q0 C5 V 5 0A 16.35 l where 0 is the permittivity of a vacuum (8.85 3 10212 F/m). From the equality on the right side of Equation 16.35, the charge density (D0) on the plates is calculated from Equation 16.36. D0 5 Q0 5 A 0V 16.36 l The term V/l is the magnitude of the electric field (E ) in volts/m across the capacitor. Inserting the magnitude of the electric field into Equation 16.36 results in Equation 16.37. 16.37 D0 5 0E In Figure16.23a, the electric field vector (E) is shown penetrating across the capacitor, because vacuum is a dielectric. + D0 = 0E + l + + + – – Vacuum – – – – – – – E – – + V + – – – – – – (a) D = 0E + P + + + + + + + + + + – – – – – – – – – – Dielectric P E + + + + + + + + + + – – – – – – – – – – + + – + – + – + + – + – – + – + – + – – V (b) Figure 16.22 (a) A parallel-plate capacitor with a vacuum, showing the surface charge density D0 and the electric field vector E. (b) A parallel-plate capacitor with a dielectric material present, showing the polarization vector P, the surface charge density D, and the electric field vector E. W-41 Electrical Properties of Materials Example Problem 16.9 A vacuum parallel-plate capacitor with an area of 1 cm2 and a gap of 1 mm has 1 V applied. Calculate (a) the capacitance and (b) the surface charge density. Solution a) The capacitance is calculated from Equation 16.35. C5 0 A l 5 8.85 3 10212 F/ms1 3 1022 md2 5 8.85 3 10213 F 1.00 3 1023m b) The total charge is calculated from the capacitance in Equation 16.35, and the surface charge density is calculated from the surface charge, with Equation 16.36. Q0 5 C V 5 8.85 3 10213 Fs1 Vd 5 8.85 3 10213 C D0 5 Q0 A 5 8.85 3 10 213 C C 5 8.85 3 10 29 2 m 1 3 10 24 m2 If a material replaces the vacuum in a capacitor with an applied voltage (V ), in analogy with Equation 16.35, the capacitance is given by Equation 16.38: C5 Q A 5 V l 16.38 where Q is the charge on the plates with a dielectric material and is the permittivity of the material. It follows from Equation 16.36 that the surface charge density on the plates (D) with a dielectric material is given by Equation 16.39. D5 Q V 5 5 E A l 16.39 The ratio of the permittivity of the dielectric material to the permittivity of a vacuum is the relative permittivity (r ), as shown in Equation 16.40. r 5 0 16.40 The relative permittivity is also called the dielectric constant. Table 16.8 presents the values of the relative permittivity or dielectric constant of some dielectric materials at 60 Hz and at 106 Hz, along with values of the dielectric strength and the electrical resistivity. As shown later in this section, the dielectric constant is a function of frequency. Since the dielectric constant of these materials is greater than 1, the capacitance with these dielectric materials for a constant voltage is greater than with a vacuum. Some materials, such as barium titanate (BaTiO3), have dielectric constants approaching 104. The dielectric strength, listed in Table 16.8, is the maximum static electric field that the material can withstand without dielectric breakdown. The dielectric strength limits how much electrical energy can be stored in a capacitor. Why does the capacitance increase when a dielectric material is inserted between the plates of a capacitor? To answer this question, we must look at what happens to the dielectric material in the capacitor when subjected to an electric field. As indicated in Figure 16.22b, the dielectric material develops a surface negative charge on the side adjacent to the capacitor plate with a positive charge, and a surface positive charge develops on the side of the material adjacent to the negative plate of the capacitor. When W-42 CHAPTER 16 Table 16.8 The Dielectric Constant, Dielectric Strength, and Electrical Resistivity of Some Dielectric Materials Material Dielectric Constant (at 60 Hz) (at 106 Hz) Dielectric Strength (106 V ? m) Resistivity (ohm ? cm) Polyethylene2.3 2.3 20 . 1016 Teflon 2.12.1 20 1018 Polystyrene 2.52.5 20 1018 PVC 3.53.2 40 1012 Nylon 4.03.6 20 1015 Rubber 4.03.2 24 Phenolic 7.04.9 12 1012 Epoxy 4.03.6 18 1015 Paraffin wax 2.3 10 1013–1019 Fused silica 3.8 3.8 10 1011–1012 Soda-lime glass 7.0 7.0 10 1015 Al2O3 9.0 6.5 6 10 –1013 14–110 8 1013–1018 TiO2 Mica7.0 40 BaTiO32000–5000 12 Water78.3 11 1013 10 –1015 8 1014 Based on data from Askeland, D.R., Fulay, P.P., and Wright, W.J., The Science and Engineering of Materials, 6th ed. Cengage Learning, Stamford, CT. (2011), p. 754. there is a separation of charge in the material, it is polarized. The magnitude of the polarization (P) of the material is equal to the surface-charge density on the dielectric material, in coulombs per square meter. The surface-charge density on the plates with a material inserted (D) is equal to the surface-charge density with a vacuum (D0) plus the magnitude of the polarization (P), as shown in Equation 16.41. D 5 D0 1 P 5 E 16.41 The polarization vector points from the negative charge in the dielectric material to the positive charge. The charge density that is established on the plates of the capacitor with a dielectric material inserted compensates for the opposite-charge density that develops on the dielectric material as a result of polarization. The capacitor response to the polarization of the dielectric material results in the increased capacitance. Example Problem 16.10 The vacuum space in the capacitor in Example Problem 16.9 is filled with fused silica (SiO2). Calculate the following for this capacitor: (a) the capacitance, (b) the surface-charge density on the capacitor plates, and (c) the magnitude of the polarization of the fused silica. W-43 Electrical Properties of Materials Solution a) The capacitance is calculated from Equations 16.38 and 16.40. The value of the dielectric constant for fused silica is 3.8. Since the applied voltage is constant, the lowest frequency value of the dielectric constant is used. C5 A r0 A 3.8s8.85 3 10212F/mds1 3 1024 m2d 5 5 5 33.6 3 10213 F l l 1.00 3 1023 m b) The surface charge is then calculated from the capacitance and the applied voltage using Equation 16.38. Q 5 CV 5 (33.6 3 10213 F)(1.00 V) 5 33.6 3 10213 C The surface-charge density is then calculated from Equation 16.39 Q 33.6 3 10213 C C 5 5 33.6 3 1029 2 A m 1 3 1024 m2 D5 c) The magnitude of the polarization is calculated from Equation 16.41. P 5 D 2 D0 5 (33.6 2 8.85) 3 1029 C/m2 5 24.8 3 1029 C/m2 Polarization in a material results when it is possible to displace positive and negative charges relative to each other. Polarization occurs when there is no flow of current. The types of polarization are shown in Figure 16.23; they are (a) electronic, (b) ionic, and (c) orientational. Electronic polarization occurs when the atomic centers of charge for the electrons and the nucleus are displaced from each other, as E E =0 + – – + (a) – – – – + + – – – – (b) – + – – + + + – – + – + (c) Figure 16.23 Schematics of (a) electronic, (b) ionic, and (c) permanent polarization. The left side of the figure shows the charge configurations without an electric field, and the right side shows the polarization in an electric field (E). W-44 CHAPTER 16 shown in Figure 16.23a. Electronic polarization is the only polarization type in a covalently bonded element, such as silicon. In ionic polarization, an electric field displaces anions and cations relative to their normal lattice positions, as shown in Figure 16.23b. Ionic polarization is present in ionic materials, such as NaCl and MgO. H2O is a molecule with a permanent orientational dipole moment, as shown in Figure 16.24. In a water molecule, the two hydrogen atoms attached to the oxygen are separated by 105°. The end of the hydrogen atoms away from the oxygen atom is a positively charged nucleus, because the hydrogen electron is localized between the hydrogen and the oxygen in a covalent bond. The end of the molecule toward the oxygen is negatively charged, because oxygen is more electronegative (3.5) than hydrogen (2.1), and the electrons are more attracted to the oxygen. The positive and negative charges on H2O are permanently displaced from each other. In H2O, if there is no applied electric field, the orientation of the H2O molecules is random, and there is no net polarization, as shown in Figure 16.23c on the left. However, in an applied electric field the H2O molecules are oriented as shown in Figures 16.23c and 16.24. Polarization results when a charge qi is displaced by a vector distance di from a position of zero dipole moment to create the vector dipole moment pi, given by Equation 16.42. pi 5 qi di 16.42 Both qi and di can be positive or negative, depending upon the charge and the direction of displacement. Assume that the positive direction is the direction of movement of the positive charge. The polarization vector (P) is the sum of the dipole moments per unit volume, as shown in Equation 16.43. P5 pi oV 16.43 i The polarization vector (P) shown in Figure 16.22b is pointing in the direction of the displacement of the positive charge. The units of polarization, as shown in Equation 16.43, are dipole moment per unit volume or the equivalent coulombs per square meter. In Table 16.8 the dielectric constant is given for an electric field oscillating at 60 Hz and 106 Hz. The dielectric constant is a function of the frequency of the electric field, as shown in Figure 16.25, because the displacement of atoms and molecules requires a certain response time. If the time for the electric field cycle is less than the response time for a particular mode of polarization, then that mode of polarization does not occur. As shown in Figure 16.25, at low frequency all of the modes of polarization are present: electronic, ionic, and orientational. At frequencies above 108 Hz, there is insufficient time for the molecules that produce orientational polarization to respond to the alternating electric field, and only ionic and electronic polarization are present. An applied electric field of frequency above 1014 Hz E + P H Oxygen + – H Figure 16.24 The permanent dipole of an H2O molecule oriented in an electric field (E) producing polarization (P). (Based on Askeland, D.R., Fulay, P.P., and Wright, W.J., The Science and Engineering of Materials, 6th ed., Cengage Learning, Stamford, CT. (2011), p. 39.) W-45 Dielectric constant Electrical Properties of Materials Orientationpermanent Ionic Electronic 104 108 1012 Frequency (Hz) 1016 Figure 16.25 A schematic of the dielectric constant (r) as a function of frequency for a dielectric material with orientational, ionic, and electronic polarization. is higher than the lattice vibration frequencies (~1013 Hz), and the ions in a crystal cannot be displaced by applied electric fields of frequency higher than approximately 1013 Hz. The ionic component of the dielectric constant disappears at frequencies above approximately 1013 Hz. Electrons have a much shorter response time than atoms do, and the electronic component of dielectric constant continues to be present to frequencies of 1017 Hz, but at frequencies above 1017 Hz even electronic polarization disappears. In a normal dielectric material, the polarization is small, positive, and reversible. Polarization is reversible if the polarization returns to 0 when the applied electric field returns to 0, as shown in Figure 16.26. The magnitude of polarization (P) relative to the magnitude of the electric field (E ) is proportional to the dimensionless dielectric susceptibility (d), given by Equation 16.44. The dimensionless dielectric susceptibility is also the magnitude of the polarization of the dielectric material divided by the surface-charge density of a vacuum capacitor, as shown by Equation 16.44. d 5 P P 5 0 E D0 16.44 The dielectric constant (r ) and the dielectric susceptibility are related through Equation 16.45. r 5 D0 1 P DE 5 5 5 1 1 d 0 D0E D0 16.45 In Figure 16.26, the slope of the P versus E plot is equal to d 0, or 0(r 2 1), as shown by Equations 16.44 and 16.45. P E Figure 16.26 A schematic of the polarization (P) as a function of electric field (E) for a normal dielectric material. W-46 CHAPTER 16 The material for DRAM in computers is the dielectric silica (SiO2). This is the memory that is active only when the computer power is on. Computer memory is binary with values of 1 and 0 that represent information. A value of 1 indicates a region of material with one direction of polarization, and a 0 represents a material polarized in the reverse direction. When the computer power is turned off, there is no electric field; the polarization of all the material returns to 0, and all of the information is lost. Dielectric materials that can permanently store information when the power is off are covered next. 16.8.2 Ferroelectric Materials In ferroelectric materials, after a large electric field is applied and then reduced to 0, the polarization does not return to 0, as it did in a normal dielectric material. The term ferroelectric comes from the shape of the plot of the polarization as a function of an applied electric field, as schematically shown in Figure 16.27. The shape is similar to the shape to the applied magnetic field-magnetization plot for ferromagnetic materials, such as iron, shown in Figure 17.9. The shape of the P-vs.-E plot in Figure 16.27 results from ferroelectric materials having a permanent ionic dipole. Ferroelectric materials, such as barium titanate (BaTiO3), shown in Figure 16.28, and lead zirconate (PbZrO3) have ionic polarization. At temperatures above 130°C, BaTiO3 is a cubic crystal with no permanent dipoles. At 130°C, the lattice parameter is 0.401 nm. However, below 130°C, the crystal is tetragonal with the lattice parameters shown in Figure 16.28a. Figure 16.28b shows the ion displacements relative to the barium ions at the corner positions. The ion displacements result in permanent polarization. The Curie temperature is the temperature where the material changes from ferroelectric to paraelectric upon heating. For barium titanate, the Curie temperature is approximately 130°C. In paraelectric materials, the polarization returns to 0 when the electric field is reduced to 0; however, the relationship between P and E can be nonlinear. P Ps Pr –Ec 0 Ec E Pr Ps Figure 16.27 The polarization (P) as a function of an electric field (E) for a ferroelectric material. W-47 Electrical Properties of Materials 0.009 nm 0.403 nm 0.006 nm 0.398 nm Ti4+ Ba2+ 0.006 nm 0.398 nm O2– (a) (b) Figure 16.28 The barium titanate unit cell at room temperature. (a) An isometric view. (b) A front view showing the displacements of the Ti41 and O22 ions relative to the barium atoms. (Based on Callister W.D., Materials Science and Engineering, An Introduction 6th ed. John Wiley &Sons NY (2003), p. 645.) The properties of a ferroelectric material are demonstrated by applying a large electric field in one direction and then reversing the direction, as demonstrated in Figure 16.27. The origin in Figure 16.27 is a nonpolarized material (P 5 0) in a zero electric field (E 5 0). An increase in the electric field causes an increase in polarization. Once all of the possible electrical dipoles are oriented in the direction of the applied electric field, the polarization is saturated at Ps. In a ferroelectric material, when the applied electric field is then reduced to 0, there remains a residual polarization (Pr) in the direction of the former electric field. The polarization does not return to 0, because the internal electric field created by the dipoles is sufficient to maintain the dipoles oriented in the direction of the original electric field. Equation 16.41 shows that if the applied electric field is 0, the sum of the polarization and vacuum surface charge density is 0, but the polarization is not 0. The electric field in the reverse direction that reduces the polarization to 0 is the coercive electric field (2Ec). When P 5 0 in a ferroelectric material, it does not mean that there are no longer dipole moments in the material. In a ferroelectric material there are domains that are regions with the polarization in the same direction. When P 5 0, there are as many domains oriented in the original direction of the electric field as in the reversed field direction (2Ec ), as shown schematically in Figure 16.29a. When the total polarization of the ferroelectric material changes, it is because the polarization direction within individual domains changes. Figure 16.29b shows ferroelectric domains in a crystal of barium titanate. The polarization of a ferroelectric material is saturated in the reverse direction (P 5 2Ps) by application of a sufficient electric field in the reverse direction. If the applied electric field is then reduced to 0, the polarization is the residual polarization 2Pr. The net material polarization is forced to 0 by applying the coercive field (Ec) in the forward direction. At the coercive field, there are as many domains with polarization in the forward direction as there are domains with polarization in the reverse direction. The polarization is saturated in the forward direction by application of sufficient electric field in the forward direction to orient all of the domains in the direction of E. The nonlinear cycle of P as a function of E from 1Ps to 2Ps to 1Ps shown in Figure 16.27 is a hysteresis loop. Hystersis loops are encountered in other cyclic plots such as stress and strain, and magnetic field and intensity of magnetization as demonstrated in Chapter 17. If the electric field is cycled from saturation in the forward direction to saturation in the reverse direction, the hysteresis loop is repeated. One way to return the material to the origin (P 5 0, E 5 0) is to heat the material above the Curie temperature to disorder W-48 CHAPTER 16 + – + – – + – + (a) Figure 16.29 (a) A schematic of ferroelectric domains, showing the polarization direction. (b) An image of ferroelectric domains in barium titanate produced with a variant of atomic force microscopy (AFM) that applies a voltage to the tip. Piezoelectric deformation in the ferroelectric domains resulting from the applied voltage is detected by AFM. The width of the portion of the specimen in the image is 47 m. http://en.m.wikimedia.org/ wiki/File:Bulk_BTO_PFM_scan.png. (Tertib64) the dipole moments, and then to cool the material to below the Curie temperature in the absence of an electric field. With this procedure the domains are in random orientations, and the polarization is 0. The polarized-unpolarized transformation in dielectric materials can be thermodynamically analyzed. The internal energy change (dE) when the polarization changes by dP in an electric field (E ) is given by dE 5 2EdP, and the entropy (S) is expressed in terms of the degree of polarization. The change in the Gibbs free energy is then calculated from the internal energy change, the entropy change, and the temperature. Ferroelectric crystals are utilized in applications such as microdimensioned capacitors and permanent memory for computers. Microcapacitors made from high dielectric constant materials, such as barium titanate, have a high capacitance. In the permanent memory of a computer, polarization in the positive direction corresponds to a 1 and polarization in the reverse direction corresponds to a 0. Writing is accomplished with an electric field. If the power goes off and the electric field (E) is reduced to 0, the data remains intact because of the residual polarization, and the data is preserved. Example Problem 16.11 (a) Calculate the dipole moment in a unit cell of barium titanate with the room-temperature atom positions shown in Figure 16.28b. (b) Calculate the polarization of barium titanate at saturation. Solution a) To solve this problem, we apply Equation 16.42 and sum the dipole moments over the unit cell. The charges and displacements are as follows: 1. Two oxygen atoms of charge 22 move down 0.006 nm. The oxygen atoms in the face-centered positions are shared between two unit cells. The positive direction is up. Electrical Properties of Materials q1 5 22(2e electrons) 5 24e electrons d1 5 20.006 nm 5 26 3 10212 m p1 5 q1d1 5 24e electrons(1.602 3 10219 C/electron)(26 3 10212 m) p1 5 38.45 3 10231 C ? m 2. One oxygen ion of charge 22 moves down 0.009 nm. q2 5 22e electrons d2 5 20.009 nm 5 29 3 10212 m p2 5 q2d2 5 22e electrons (1.602 3 10219 C/electron)(29 3 10212 m) p2 5 28.84 3 10231 C ? m 3. One Ti atom of charge 5 14 moves up 10.006 nm. q3 5 4e electrons d3 5 0.006 nm 5 6 3 10212 m p3 5 q3d3 5 4e electrons (1.602 3 10219 C/electron)(6 3 10212 m) p3 5 38.45 3 10231 C ? m The dipole moment per volume of a unit cell is the sum of calculations 1, 2, and 3. p 5 (p1 1 p2 1 p3) 5 (38.45 1 28.84 1 38.45) 3 10231 C ? m 5 105.7 3 10231 C ? m b) The dipole moment at saturation polarization corresponds to when all of the dipole moments are oriented in the direction of the applied electric field. The polarization P is the sum of the dipole moments per unit volume (V ). P5 pi o V 5 s0.403 3 10 i 1.057 3 10229C ? m 1.057 3 10229C ? m 5 29 29 0.638 3 10228 m3 mds0.398 3 10 mds0.398 3 10 md 29 P 5 0.165 C?m m3 16.8.3 Piezoelectric Materials Piezoelectric materials produce a voltage when deformed, and an applied voltage deforms the material. The barium titanate crystal, shown in Figure 16.28 is a piezoelectric crystal. Other piezoelectric materials include lead zirconate, quartz, and polymers such as polyvinylidene fluoride. The word piezo comes from the Greek word piezein, meaning “to press or squeeze.” Piezoelectric crystals are utilized to convert the bumps in the grooves of a record into a voltage that is amplified to produce music. Piezoelectric crystals are also utilized to produce sound and ultrasound waves from a cyclic applied voltage, and they are detectors of sound. Piezoelectric crystals are also utilized in micro-electromechanical systems (MEMS) W-49 W-50 CHAPTER 16 to produce displacement, and they are utilized as the displacement control in scanning tunneling and atomic-force microscopes, as discussed in Chapter 15. Figure 16.28 shows the crystal structure of the piezoelectric crystal barium titanate at room temperature. With the displacements shown in Figure 16.28b, the top of the crystal is charged positive, and bottom is charged negative because the Ti14 has moved up and all of the O22 ions have moved down. If a voltage is applied to the crystal such that the top of the crystal is charged positive and the bottom charged negative, this causes the dipole displacement vectors d1, d2, and d3 in Example Problem 16.11 to decrease. This decreases the 0.403-nm lattice parameter. Conversely, if the top of the crystal in Figure 16.28b is charged negative and the bottom positive, this causes the displacement vectors (d1, d2, and d3) in Example Problem 16.11 to increase. The increase in the displacement vectors causes the vertical lattice parameter (c) to increase to more than 0.403 nm. The crystal is stretched or compressed by applying a voltage to the crystal. The piezoelectric-strain coefficient (kp ) is the longitudinal strain () per unit of applied electric field (E ), as shown in Equation 16.46. kp 5 E 16.46 Table 16.9 presents values of kp for some materials. The piezoelectric-strain coefficient multiplied by the magnitude of the electric field in V/m yields the dimensionless strain. Piezoelectric-strain coefficients as high as 6.8 3 1029 m/V have been obtained for lead-lanthanum-zirconium-titanate. As long as the electric field is less than the dielectric breakdown strength of the piezoelectric crystal, the displacement produced by the voltage is very accurate and the response times are short. To see how the voltage is generated by straining the crystal, we can look at the equations that relate dipole moments to polarization and capacitance. When the crystal is mechanically strained in tension, this increases the dipole displacements (d1, d2, and d3) and increases the dipole moments (p1, p2, and p3) and the total polarization by DP according to Equations 16.42 and 16.43. The increase in the magnitude of polarization results in an increase in the total surface charge density (D) given by Equation 16.41. A larger surface-charge density produces a larger electric field (E) according to Equation 16.41, and a larger voltage. Compressing the crystal causes a decrease in the dipole displacements, the polarization, surfacecharge density, and voltage. The piezoelectric materials allow for a direct conversion of electrical voltages to mechanical displacements, or from mechanical displacements to electrical voltage. There are also polymer ferroelectrics and piezoelectrics, such as polyvinylidene fluoride (PVDF). Figure 16.30 shows the molecular structure of the trans form of PVDF. The structure of PVDF is like that of PE, except that every other carbon atom is bonded to fluorine atoms rather than hydrogen. Because of the strong affinity of fluorine for electrons, the fluorine atoms on the molecule are negatively charged and hydrogen atoms are positively charged, creating a permanent dipole. The piezioelectric strain coefficient for PVDF is typically 23 3 10212 m/V in comparison with 149 3 10212 m/V for BaTiO3, as shown in Table 16.9. Table 16.9 The Piezoelectric Strain Coefficient (kp) for Some Materials Material kp (m/V) BaTiO3149 3 10212 PbZrO6 (PZT-4) 285 3 10212 PbNb2O6 85 3 10212 Quartz 2.3 3 10212 Based on data from the CRC Handbook of Tables for Applied Engineering Science, CRC Press, Boca Raton, Fl. (1973), p. 983. W-51 Electrical Properties of Materials Figure 16.30 The molecular structure of the piezoelectric polymer trans polyvinylidene fluoride. The red atoms are carbon; the blue atoms are hydrogen, and the gold atoms are fluorine. (Based on http://en.wikipedia.org/wiki /Ferroelectric_polymers) Example Problem 16.12 A piezoelectric actuator for a micro electromechanical system of length 1 3 1024 m is produced from barium titanate, which has a dielectric breakdown strength of 6 3 106 V/m. The piezoelectric strain coefficient of the barium titanate is 1.49 3 10210 m/V. Assume that the maximum electric field applied to the piezoelectric actuator is 1 3 106 V/m. What is the maximum displacement obtained from the actuator? Solution Since the maximum electric field is 1 3 106 V/m, then the strain () is calculated from the coefficient of piezoelectric strain. kp 5 1 E Dl 5 l0 kp E 5 1 3 1024 m 1.49 3 10210 m V 2 11 3 10 m2 5 1.49 3 10 6 V 28 m 16.9 Electrical Conductivity in Ionic Materials Ionic conducting materials are utilized for devices such as thermistors and chemical sensors. Thermistors use the electrical resistivity of a material to determine its temperature. Both thermistors and chemical sensors utilize the electrical conductivity of ionic materials to measure properties of a system. Figure 16.31 shows the electrical conductivity of various mole percentages of Fe3O4 in solid solutions with MgCr2O4, as a function of 1000/T. As the temperature increases, the electrical resistivity decreases. Most thermistors are made from Mn, Ni, Fe, and Co oxides. In ionic materials the electrical conductivity results primarily from the diffusion of ions. The ionic contribution to the electrical conductivity (ion) is expressed in Equation 16.47, which is similar to Equation 16.25. ion 5 o Z en i i i i 16.47 W-52 CHAPTER 16 –8 90% 100% –7 80% 66.5% 49.4% 55% –6 log10 conductivity () –5 39.4% –4 25.5% –3 –2 8.6% 5.4% –1 1% 0 1 2 3 0 1 2 3 4 5 6 7 8 9 10 11 1000 T Figure 16.31 The log of the electrical conductivity of various mole percentages of Fe3O4 in solid solution with MgCr2O4, as a function of 1000/T. (Based on Verwey, E.J. et al., J. Chem. Phys., 15 (1947), p. 181.) The ionic electrical conductivity is a sum of the electrical conductivity over all of the ions per unit volume (ni ) with electron charge Zi e that have a mobility μi . The mobility of the ions is given by the Nernst-Einstein Equation 16.48: i 5 ZieDi 16.48 kT where Di is the diffusion coefficient (diffusivity) of the i-type ions. Ionic electrical conductivity is directly proportional to the diffusivity of the ions, and diffusivity of the ions is exponentially dependent upon temperature. For example, in the temperature range from 1100°C to 1600°C, the diffusion coefficient for oxygen ion diffusion in Ca0.15Zr0.85O2 is given by Equation 16.49. 1 DsO21 d 5 1 3 109 exp 2 2 0.84 eV m2/s kT 16.49 The electrical conductivity from ions, such as oxygen, is high in partially stabilized zirconia-calcia, because of a high concentration of oxygen vacancies. The high oxygen-vacancy concentration results because calcia (CaO) contributes only one oxygen atom for every calcium atom in a substitutional solid solution in the zirconia structure (ZrO2) that requires two oxygen atoms for every zirconium atom. Materials such as zirconia-calcia are used in devices such as oxygen sensors. The oxygen sensor in an exhaust system of a gasoline powered engine operates on the principle that the concentration of oxygen in a zirconia electrolyte is proportional to the 12 power of the partial pressure of the oxygen in the atmosphere in contact with the zirconia electrolyte. Example Problem 4.12 discusses this principle in the purification of hydrogen by diffusion through palladium. For oxygen sensors in automobile exhaust systems called Nernst cells, the exhaust is on one side of the sensor and air is on the Electrical Properties of Materials Electrodes (gas permeable) Zirconia membrane Exhaust Gas Nernst cell Heater (option) Reference air 0.2–0.8 V Heater (option) Figure 16.32 A schematic of an oxygen sensor showing exhaust gas on one side of the electrolyte made of partially stabilized zirconia, and air on the other side. Electrodes, normally porous platinum, plated on the electrolyte are electrically connected to a voltage sensor. (Based on http://en.wikipedia.org/wiki/Oxygen_sensor) other side, as shown in Figure 16.32. The different concentrations of oxygen on either side of the zirconia electrolyte sensor create an electrical potential across the zirconia, because the Gibbs free energy of the zirconia depends upon the oxygen concentration. The difference in the Gibbs free energy is converted into a voltage, as given in Equation 10.12. The voltage from the sensor controls the air/fuel mixture going into the engine for optimal combustion and minimal exhaust pollution. An output voltage of 0.2 V indicates a lean fuel to air mixture, and the concentration of oxygen is sufficient to produce CO2 rather than CO. An output voltage of 0.8 V indicates a high fuel-to-air mixture, and there is insufficient oxygen to fully burn the fuel and CO is produced. An ideal voltage output is approximately 0.45 V. A feedback system adjusts the fuel input to obtain the ideal fuel-to-air mixture. 16.10 Fabrication of Crystalline Semiconductor Devices Silicon is the most common elemental semiconductor in electronic devices. To have reproducible electronic properties, the semiconductor material must be of very high purity. Silicon for semiconductor applications is purified by zone refining until the impurity level is a few impurity atoms per billion (109) silicon atoms. In zone refining, a zone of liquid is moved from one end of the crystal to the other, as shown schematically in Figure 16.33. The silicon is purified, because the solubility of the impurities in the liquid is greater than in the solid. Figure 16.34 shows a hypothetical phase diagram of some impurity in silicon. The original impurity concentration in the silicon is C0. When the solid silicon is heated in a zone to form a liquid, the liquid impurity concentration remains at C0. If the liquid is then cooled to form a W-53 W-54 CHAPTER 16 Connection to positive high voltage Specimen Focusing shield Motion of molten zone Filament Molten zone Flux of electrons Figure 16.33 A schematic of an electron-beam zone-refining furnace. (Based on Guy, A.G., Introduction to Materials Science, McGraw-Hill, New York (1972), p. 322.) solid at Ts, the impurity concentration in the solid is Cs, and since Cs is much smaller than C0, the material is purified. In semiconductor production, many liquid-solid zones simultaneously pass over a silicon rod by having a series of heating elements. In solid-state electronic systems, the semiconductor is normally a silicon single crystal, called an ingot. The ingot is doped to produce a p-type semiconductor. The doped silicon ingots are grown a Liquid (L) + L Ts Temperature a+L Solid a Cs C0 Impurity concentration (Ci ) Figure 16.34 Part of the phase diagram near the melting temperature of silicon, with an impurity showing the original composition (C0) and the composition of the solid (Cs ) that forms from the liquid of composition C0. W-55 Electrical Properties of Materials Rotary chuck Seed crystal Growing crystal Silicon ingot Molten silicon Crucible Silicon IC fabrication wafer on wafer surface Chip (die) Packaged chip Heating coils (a) SiO2 Exposed SiO2 is etched away p-type silicon substrate p-type silicon substrate (i) (iv) Ultraviolet radiation Photomask Photoresist SiO2 n-type dopant n-type n-type p-type silicon substrate (v) SiO2 p-type silicon substrate n-type n-type p-type silicon substrate (ii) (vi) Exposed photoresist is dissolved SiO2 p-type silicon substrate (iii) (b) Figure 16.35 (a) A schematic of the overall process of producing an integrated circuit (IC). Crystal growth by the Czochralski technique produces a silicon ingot. The ingot is cut into wafers; followed by fabricating the metal, oxide, and semiconductor materials into an IC (chip or die). Finally, the chip is packaged. (b) A schematic of the steps in producing n-type and p-type regions in a semiconductor by a photolithographic process. ((a) Based on Microchip Fabrication, Third Edition, by P. VanZant, The McGraw-Hill Companies. (1997), Fig. 3.7. (b) Based on Fundamentals of Modern Manufacturing, by M.P. Groover, John Wiley, p. 849, Fig. 34-3.) using the Czochralski technique, which we discussed in Section 13.2.7, and is shown schematically in Figure 16.35a. At present, single crystals of silicon with a diameter up to 0.3m can be produced. The p-type crystal is sliced into wafers with a diamond saw, and then the surface is polished for processing. The insulating and n-type regions are produced by the following photolithographic process, shown in Figure 16.35b. The p-type silicon is oxidized to produce an insulating layer of SiO2, as shown in W-56 CHAPTER 16 Figure 16.35b(i). The oxide is then coated with a photoresist polymer that resists acid etching under normal conditions; however, the photoresist can be etched by acid after exposure to UV light. In Figure 16.35b(ii), a mask allows UV light to impinge upon areas of the photoresist that are to be etched. In Figure 16.35b(iii), the exposed photoresist polymer material is etched, and in Figure 16.35b(iv) the insulating layer of SiO2 is etched to expose the p-type silicon. The unexposed photoresist is removed with a solvent. In Figure 16.35b(v), donor atoms are added to the exposed p-type silicon to make it n-type. If sufficient donor atoms are added to the p-type semiconductor, all of the acceptor levels and holes in the VB are filled by electrons from the donor level, and there are still sufficient electrons to be excited into the CB, producing an n-type semiconductor. The n-type regions of the semiconductor can be produced by diffusion of donor atoms, as shown in Figure 16.36a. However, now the preferred process is to use particle accelerators to implant donor ions, as shown in Figure 16.36b. The location of the donor atoms is controlled with masks that allow atoms to be diffused or implanted only in the desired locations. In ion implantation, the penetration of the implanted ions is ballistic. Just as with shooting bullets into a soft wooden target, there are no ions (bullets) that are stopped at the surface. Figure 16.35b(vi) shows the npn-junction of a transistor. Conductors, such as aluminum and more recently copper, are deposited through masks from a vapor in a vacuum system. Other dielectric oxides, such as iron oxide or zirconiumcalcia oxide (Zr-CaO2), are produced by vapor-depositing the metal and then oxidizing the metal. Through the formation of the silicon semiconductor crystal (S) with dopant atoms, the insulator oxide (O), and conductive metal (M), an integrated circuit (IC) is produced of MOS architecture. An IC integrates all of the electronic devices into the silicon chip or die. The die is then packaged by embedding it into a protective material, such as epoxy, as shown in Figure 16.35a. The packaged die is then electrically connected to other devices in a product by metal conductors. Gas of dopant atoms High velocity dopant ions Mask Distance O O Semiconductor t O Dopant concentration Dopant concentration t Distance (a) t Distance (b) t Figure 16.36 (a) A schematic of diffusion of dopant atoms into a silicon crystal through a mask. The lower figure on the left is a schematic of the distribution of dopant atoms resulting from diffusion. (b) Ion implantation of dopant atoms through a mask. The lower figure on the right is a schematic of the distribution of dopant atoms resulting from ion implantation. Electrical Properties of Materials Heater GaAs crystal atomic or molecular beams Shutters Furnaces Ga As Figure 16.37 A schematic of molecular-beam epitaxial growth of a GaAs crystal. Compound semiconductor crystals, such as GaAs, are produced by both the Bridgeman-Stockbarger and Czochralski techniques, and by molecular-beam epitaxy (MBE), shown schematically in Figure 16.37. In the MBE growth of a GaAs crystal, the Ga and the As are each vaporized from different furnaces. The vaporized atoms are deposited on an existing GaAs crystal surface where they attach in the proper locations to continue the epitaxial growth of the crystal. Shutters allow the furnaces to be closed. Other furnaces can be added to deposit doping atoms to produce p-type and n-type semiconductors. Epitaxial growth is discussed in Section 13.2.7. 16.11 Present and Future Electronic Systems Moore’s law states that the density of transistors in an integrated circuit will double every two years. This is due to the reduction in the size of components in the MOS architecture. In 1971, the average size of a state-of-the-art transistor was 10 m; in 1985 it was 1 m, in 2002 it was 90 nm, in 2007 it was 45 nm, and 22-nm devices were introduced in 2011. In the 45-nm transistor circuits, some of the insulators are as thin as 1.2 nm, and in 22-nm transistors some of the insulators are so thin that they are porous. In the 45-nm transistors, 30 million transistors fit on the head of a pin. Because of this small size, a 45-nm transistor can switch on and off 300 billion times a second. With silicon-based integrated circuits, it is difficult to produce circuits of further reduced size. At some small size the silicon-based circuits may not perform well, and a change of technology may be necessary. The integrated circuits of the future might be produced from carbon nanotubes or from graphene. Conductor and semiconductor nanotubes are produced by changing the diameter of the nanotube and by changing the orientation of the graphite structure relative to the axis of the nanotube. Carbon nanotubes are also doped to produce p-type and n-type semiconductors. Electronic components, such as diodes, transistors, and interconnections, have been produced from carbon nanotubes. Transistors have also been made from graphene. Polymer semiconductors will replace crystalline silicon semiconductors in applications that do not require high-speed switching, because of their low cost. Organic light-emitting diodes (OLEDs) are already utilized in the display screen in cell phones, smart phones, and digital cameras. OLEDs are discussed in Chapter 18. W-57 W-58 CHAPTER 16 Summary ●● ●● ●● ●● ●● ●● ●● ●● ●● Electrical resistivity and electrical conductivity are material properties independent of the specimen geometry, and they are inversely related to each other. Metals, such as copper and silver, have the highest electrical conductivity of all materials at room temperature. Semiconductors, such as pure silicon and germanium, have electrical conductivities that are approximately eight to eleven orders of magnitude less than that of silver and copper at room temperature. Insulators, such as diamond and polyethylene, have an electrical conductivity nearly twenty-four orders of magnitude less than that of silver and copper at room temperature. Superconductors have an infinite electrical conductivity, and zero electrical resistivity, below a critical temperature that is well below room temperature. In metals with an applied voltage, the movement of free electrons produces the current. Because each free electron has a probability of being found anywhere in a metal, each free electron has a probability of interacting with every other free electron. The Pauli exclusion principle applies to interacting electrons and therefore to all of the free electrons. No two free electrons can have the same set of quantum numbers in the entire metal specimen. Each free electron has a unique set of quantum numbers (nx, ny, and nz) and a spin quantum number of 112 or –12. The kinetic energy of each free electron is proportional to the sum of the squares of the quantum numbers. The free-electron energy levels in a metal are spread into a band of energy called the conduction band. The maximum-kinetic-energy free electron in the conduction band at 0 K is called the Fermi energy. In a metal specimen subject to an electric field, the current density is equal to the product of the free-electron density, the charge on an electron, and the drift velocity of the electrons. The drift velocity is assumed constant for all free electrons in a specimen with a specified electric field. Electron scattering provides a limit to the drift velocity. Conduction electrons are scattered by the structural defects in a crystal and by lattice vibrations. For dilute concentrations of point defects, the scattering distance is inversely related to the concentration of point defects. Nordheim’s rule states that for a dilute solution, the electrical resistivity of a metal alloy linearly increases with increases in the atomic fraction of impurity atoms. The electrical resistivity of a composite material follows the rule of mixtures, unless the composite has layers perpendicular to the current flow. In a metal that is not a superconductor, as the temperature approaches 0 K, there is a residual electrical resistivity due to structural defects, and the thermal component of electrical resistivity approaches 0. The thermal component of electrical resistivity increases linearly with temperature, because the amplitude of the atom vibrations increases. The electrons are scattered more frequently by the vibrating atoms, and the distance between scattering events decreases. The probability that a free electron has the energy E at temperature T is given by the Fermi-Dirac statistical distribution. Thermionic emission is when a material is heated to a high temperature and electrons are emitted. The energy required to excite a Fermi energy electron to the energy of free space is called the work function. In a superconductor below the critical temperature, the electrical resistivity is 0 and the electrical conductivity is infinite. In the BCS theory of superconductivity of metals, at temperatures below the critical temperature electrons are coupled into pairs through interactions with the crystal lattice. Electrical Properties of Materials ●● ●● ●● ●● ●● ●● ●● ●● The electrical conductivity of an intrinsic semiconductor is the same as the electrical conductivity of a pure semiconductor, and the electrical conductivity is due to free electrons in the conduction band and holes in the valence band. An extrinsic semiconductor is one whose electrical conductivity depends upon the concentration of foreign atoms. For example an extrinsic n-type semiconductor is made, with silicon from Group IV by adding impurity atoms from a group of a higher number, such as phosphorus (P) from Group V. In the extrinsic range of operation of an n-type semiconductor, such as Si doped with P, the concentration of electrons in the conduction band is constant and equal to the concentration of P. An example of a p-type extrinsic semiconductor is Group IV silicon doped with an element of a group number less than Group IV, such as the Group III elements boron, aluminum, or gallium. In the extrinsic range of operation of a p-type semiconductor made from Si and Group III atoms, the concentration of holes in the valence band (p) is equal to the concentration of the Group III atoms. Most polymers such as polyethylene are insulators. Insulating polymers are combined with conductive powders, such as graphite or silver, to produce composite materials with increased electrical conductivities that are utilized in resistors. Polymers such as polyacetylene can be semiconductors and conductors. Polyacetylene conducts electricity if there are defects in the bonding called solitons. Doping polyacetylene with iodine, chlorine, or fluorine oxidizes polyacetylene and produces a p-type semiconductor by removing electrons from the solitons. Dopants such as sodium and lithium mixed with polyacetylene produce an n-type semiconductor by adding electrons to the solitons. At very high levels of doping, polyacetylene behaves like a metal. Some possible advantages of polymer conductors and semiconductors relative to metals and single-crystal silicon are sustainable supplies, reduced costs, mechanical flexibility, and transparency. When two metals with different work functions are brought into electrical contact, a few electrons flow from the metal with the smaller work function to the metal with the larger work function, producing a constant Fermi energy in the two metals. The electron flow produces a contact voltage at the interface of the two metals, equal to the difference in the Fermi energy of the two metals divided by the charge of an electron. A pn-junction is produced by joining p-type and n-type material. In an electronic circuit, a pn-junction functions as a diode that allows current to flow in one direction but not in the reverse direction. An npn-junction is a transistor that acts as an amplifier in an electronic circuit. A metal-polymer-metal junction is also a diode in an electronic circuit. Dielectric materials are characterized by their capacitance. The capacitance of a parallel-plate capacitor is the total electrical charge on the plates per applied volt. For a parallel-plate capacitor, the capacitance is also the permittivity of the dielectric material between the plates times the area of the plates, divided by the spacing between the plates. The ratio of the permittivity of a dielectric material to the permittivity of a vacuum is called the relative permittivity. The relative permittivity is also called the dielectric constant. The dielectric strength is the maximum static electric field that the material can withstand without dielectric breakdown. The capacitance of a capacitor with a dielectric material is increased relative to that of a vacuum capacitor, due to polarization of the dielectric material. Polarization in a material results when positive and negative charges are displaced relative to each other, but there is no flow of current. The types of polarization are electronic, ionic, and orientational. In dielectric materials the polarization is small, positive, and reversible; and when the applied electric field is 0, the polarization is 0. In ferroelectric materials after a large electric field is applied and then reduced to 0, a residual polarization is present because of ion displacements. W-59 W-60 CHAPTER 16 ●● ●● ●● Piezoelectric materials produce a voltage when deformed, and an applied voltage deforms a piezoelectric. Piezoelectric crystals are utilized to produce music from records, produce and detect sound and ultrasound waves, and produce controlled displacement in micro electromechanical systems, scanning tunneling microscopes, and atomic force microscopes. In ionic materials electrical conductivity results primarily from the diffusion of ions. Ionic conducting materials are utilized for devices such as thermistors and chemical sensors. Silicon for electronic device applications is purified by zone-refining until the impurity level is a few impurity atoms per billion (109) silicon atoms. Silicon single crystals for integrated circuits are grown using the Czochralski technique. The silicon single crystals are cut into wafers, polished, and processed into integrated circuits by photolithography. Metal conductors are deposited by vapor deposition. Metal oxides are produced by vapor-depositing the metal, and then oxidizing the metal. Compound semiconductor crystals, such as GaAs, are produced by crystal growth techniques and by molecular-beam epitaxy (MBE). Supplemental Reading: Subjects and Authors Full references are listed at the end of the book. General: Askeland, Fulay, and Wright Electronic properties of materials: Gersten and Smith; Kasap; Kwok Electronic properties of metals:Wilkes Electronic properties of semiconductors: Mayer and Lau; Kwok: Sze; Sze and Kwok Electronic properties of superconductors: Fossheim and Subdφ, Bardeen, Cooper, and Schrieffer Electronic properties of ceramics: Chiang, Birnie, and Kingery Nano-electronic properties: Dresselhaus, Dresselhaus, and Avouris; Turton Electronic properties of polymers: Hadziioannou and van Hutten; Heeger, Sariciftci, and Namdas Integrated Circuit Technology: Beadle, Tsai, and Plummer Bioelectric composites:Benjamin Homework Concept Questions 1. Electrical resistivity and electrical ________________ are both material properties independent of the specimen geometry, and they are inversely related to each other. 2. Materials with a room-temperature electrical conductivity eight to eleven orders of magnitude less than metals are called __________________. 3. The maximum-kinetic-energy electron in the conduction band of a metal at 0 K is the ____________ energy. Electrical Properties of Materials 4. According to Mathiessen’s rule, the electrical resistivity of a metal is the _________ of electrical resistivity of zero-, one-, two-, and three-dimensional defects and thermal vibrations. 5. For dilute concentrations of point defects, the scattering distance for free electrons is _______________ related to the concentration of point defects. 6. Nordheim’s rule states that the electrical resistivity of a metal alloy increases ___________ with increases in the atomic fraction of impurity atoms for a dilute solution. 7. _________________ emission occurs when a material is heated to a high temperature and electrons are emitted. 8. The energy required to excite a(n) ___________ energy electron to the energy of free space is called the work function. 9. In a superconductor, the electrical conductivity is infinite below the ______________ temperature. 10. In the BCS theory of superconductivity of metals, at temperatures below the critical temperature electrons are coupled into ______________ pairs through interactions with the crystal lattice. 11. The electrical conductivity of a(n) _______________ semiconductor is the same as the electrical conductivity of a pure semiconductor. 12. The electrical conductivity of an intrinsic semiconductor is due to the creation of free electrons in the conduction band and __________ in the valence band. 13. A(n) _________________ semiconductor is one whose electrical conductivity depends upon the concentration of foreign atoms added to the semiconductor. 14. The thermionic emission current density from a hot filament is in the form of a rate equation, with the energy term equal to the ____________ __________. 15. Polyacetylene conducts electricity if it has bonding defects called ____________. 16. A mechanical advantage of polymer semiconductors relative to single-crystal silicon is that polymers are _____________. 17. A metal with a work function of 5 eV and a metal with a work function of 3 eV are brought together in electrical contact. They produce a contact voltage equal to ___________ volts. 18. A pn-junction in an electronic circuit functions as a(n) ____________. 19. Silicon for electronic device applications is purified by ____________ refining. 20. Silicon single crystals are grown using the ______________ technique. 21. Silicon single-crystal polished wafers are processed into integrated circuits with insulators, p-type and n-type materials, and conductors by _______________________. 22. For a parallel-plate capacitor with a vacuum between the plates, the capacitance (C ) is the total charge on the plates divided by the applied _______________. 23. The polarization in a vacuum capacitor is equal to __________. 24. The relative permittivity is also called the _______________ constant. 25. The dielectric _____________ is the maximum static electric field that the material can withstand without dielectric breakdown. 26. ______________________ in a material results when it is possible to displace positive and negative charges relative to each other, but there is no flow of current. W-61 W-62 CHAPTER 16 27. In dielectric materials the polarization is small, _______________, and reversible. 28. The dimensionless dielectric ___________________ is the magnitude of the polarization of the dielectric material divided by the surface-charge density of a vacuum capacitor. 29. In ____________________ materials, after an electric field is applied to saturation and then reduced to 0, there is a residual polarization. 30. In a ferroelectric material there are ____________ that are regions with the polarization in the same direction. 31. In a ferroelectric material, the cycle from saturation in the positive direction, to saturation in the negative direction, to saturation in the positive direction again is a(n) _______________ loop. 32. In a ferroelectric material, the electric field in the reverse direction that reduces the polarization to 0 after saturation in the forward direction is called the ______________ electric field. 33. Below the Curie temperature barium titanate is ferroelectric, and above the Curie temperature barium titanate is _________________. 34. __________________ are used to determine temperature based upon the electrical resistivity of a material. Engineer in Training–Style Questions 1. In a metal that is not a superconductor, as the temperature approaches 0 K the electrical resistivity is due to free-electron scattering by: (a) Structural defects (b) Thermal vibrations (c) Core electrons (d) Nuclei 2. In a composite material, the electrical resistivity for current parallel to the axis of the 2nd phase follows the rule of (a) Mathiessen (b) Nordheim (c) Mixtures (d) Fermi-Dirac 3. In a metal specimen subjected to a constant electric field, the drift velocity of the free electrons is: (a) The same for all free electrons (b) A linearly increasing function of time for all free electrons (c) Dependent upon the electron quantum numbers (d) Dependent upon the electron energy 4. _________________ have the highest electrical conductivity at room temperature. (a) Metals (b) Semiconductors (c) Insulators (d) Superconductors 5. If a metal contains 1028 free electrons, the number of different quantum states (nx, ny, and nz) is equal to: (a) 1028 (b) 2 3 1028 (c) 0.5 3 1028 (d) 1 Electrical Properties of Materials 6. The electrical conductivity of a metal is not proportional to which of the following? (a) Free-electron density (b) Electron mobility (c) Electron charge (d) Atomic number 7. In a metal that is not a superconductor, as the temperature increases which of the following is true? (a) Free electrons are scattered more frequently by structural defects. (b) Free electrons are scattered more frequently thermal vibrations. (c) Free-electron drift velocity is increased because of scattering by vibrating atoms. (d) Free-electron drift velocity is a constant. 8. At any finite temperature, the probability that the Fermi energy is occupied with free electrons is equal to: (a) 1 (b) 0 (c) 12 (d) A number less than 1 that is dependent upon the temperature. 9. An extrinsic n-type semiconductor can be made from Group IV germanium by adding impurity atoms of: (a) Group IV silicon (b) Group III aluminum (c) Gorup II magnesium (d) Group V arsenic 10. In the extrinsic range of operation of a p-type semiconductor made from silicon with acceptor atoms of gallium, the concentration of holes in the valence band (p) is equal to the concentration of: (a) Gallium atoms (b) Silicon atoms (c) Donor atoms (d) Electrons in the conduction band 11. A p-type semiconductor is produced in polyacetylene by doping with: (a) Boron (b) Aluminum (c) Fluorine (d) Lithium 12. An n-type semiconductor is produced in polyacetylene by doping with: (a) Arsenic (b) Phosphorus (c) Fluorine (d) Lithium 13. Which of the following materials is a dielectric? (a) Copper (b) n-type silicon (c) Diamond (d) Lithium-doped polyacetylene 14. Which of the following is not a type of polarization? (a) Electron current flow (b) Electronic (c) Ionic (d) Orientational W-63 W-64 CHAPTER 16 15. If the applied electric field has a frequency of 1016 Hz, the type of polarization present in a dielectric is: (a) Electron current flow (b) Electronic (c) Ionic (d) Orientational 16. What type of material produces a voltage when deformed, and an applied voltage deforms the material? (a) Dielectric (b) Paraelectric (c) Ferroelectric (d) Piezoelectric 17. In a partially stabilized zirconia-calcia chemical sensor, the electrical conductivity is due primarily to: (a) Electron flow in the conduction band (b) Hole diffusion in the valence band (c) Oxygen ion diffusion (d) Zirconium ion diffusion 18. In an n-type semiconductor, the conduction-band energy is 1.1 eV and the donor energy level is 1.0 eV. At temperatures approaching 0 K the Fermi energy is equal to: (a) 1.1 eV (b) 1.0 eV (c) 1.05 eV (d) 0.55 eV 19. In a semiconductor doped with donor atoms, the conduction-band energy is 1.1 eV and the donor energy level is 1.0 eV. At a high temperature there are approximately as many electrons in the conduction band as there are holes in the valence band. For this semiconductor, the Fermi energy is approximately equal to: (a) 1.1 eV (b) 1.0 eV (c) 1.05 eV (d) 0.55 eV Problems Problem 16.1 Copper is in Group IB of the periodic table, is FCC, and its room-temperature lattice parameter is 0.362 nm. (a) Calculate the free-electron density in Cu. (b) The electrical conductivity of Cu at room temperature is 6.0 3 107 (V ? m)21. Calculate the mobility of free electrons in Cu. Problem 16.2 The electrical resistivity of pure copper metal at 0°C is 1.6 3 1028 V ? m. An alloy of 1.12 atom percent nickel, which is substitutional in copper, produced with this copper has an electrical resistivity at 0°C of 3 3 1028 V ? m. What is the electrical resistivity of an alloy produced from this copper with 3.32 atom percent nickel in copper at 0°C? Electrical Properties of Materials Problem 16.3 The electrical conductivity of copper and nickel material at room temperature are 6.0 3 107 (V ? m)21 and 2.0 3 107 (V ? m)21, respectively. Copper and nickel form a solid solution at low concentrations at room temperature. If an alloy of nickel plus 5 atom percent copper produced from these materials has an electrical resistivity of 11 3 1028 V ? m, what would be the electrical resistivity of an alloy of nickel plus 10 atom percent copper? You may assume that these are dilute solutions. Problem 16.4 The room-temperature electrical conductivity of pure silver is 6.8 3 107 (V ? m)21, and for pure copper the electrical conductivity is 6.0 3 107 (V ? m)21. An alloy is made of 40 volume percent copper and 60 volume percent silver. Copper and silver form a eutectic phase diagram, with no solubility of copper in silver or of silver in copper at room temperature. Predict the roomtemperature electrical resistivity of an alloy made from these materials. Problem 16.5 The electrical resistivity of copper at 123 K is 0.5 3 1028 V ? m, and at 173 K it is 0.88 3 1028 V ? m. Evaluate the B term in Equation 16.17. Problem 16.6 The electrical resistivity of tantalum (Ta) at temperatures approaching 0 K is 1 3 1028 V ? m, and at 150 K it is 7 3 1028 V ? m. Analytically predict the electrical resistivity of Ta at 295 K. Problem 16.7 (a) For copper, determine the energies at which the probability of electron occupation is 0.9 and 0.1 at 300 K. (b) Are these energies symmetric about the Fermi energy of 7.04 eV? Problem 16.8 A filament of tungsten wire is 0.001 m in diameter, and the filament is heated to 2500 K. The work function of tungsten is 4.5 eV. (a) Calculate the probability of electrons having sufficient energy to be emitted from the filament. (b) Calculate the emission current density from the tungsten wire. (c) What length of wire can produce a thermionic electron current of 1 A? Problem 16.9 With the values of mobility and carrier density given in Table 16.6, calculate the intrinsic electrical conductivity of GaAs at 296 K, and compare the result with the listed value. Problem 16.10 A semiconductor is produced by doping Group IV silicon with 1023 atoms per cubic meter of Group III gallium. Assume that at room temperature this semiconductor is in the extrinsic range of operation. (a) What type of semiconductor is this? (b) Calculate the room-temperature electrical conductivity of this semiconductor. (c) Clearly state any assumptions you made in calculating the electrical conductivity in part b. Problem 16.11 A semiconductor with an electrical conductivity of 300 (V ? m)21 at room temperature is produced by doping Group IV silicon with Group V antimony. Assume that the mobility of electrons in silicon is equal to 0.10 m2/V ? s and that the mobility of holes is equal to 0.001 m2/V ? s. You can assume that this semiconductor is in the extrinsic range of operation at this temperature. (a) What type of semiconductor is this? (b) What doping concentration of antimony in atoms/m3 is necessary for this electrical conductivity? (c) State clearly any assumptions you make in this calculation. W-65 W-66 CHAPTER 16 Problem 16.12 It is desired to produce a p-type semiconductor out of Group IV silicon that will have an electrical conductivity of 100 (V ? m)21 at room temperature. Assume that the mobility of electrons in silicon is equal to 0.10 m2/V ? s, and the mobility of holes is equal to 0.03 m2/V ? s. You may assume that this semiconductor is in the extrinsic range of operation at this temperature. (a) What is the required density of holes in the valence band? (b) What concentration of substitutional acceptor atoms will produce the desired electrical conductivity, assuming that an element from Group III is substituted for the silicon atoms? Problem 16.13An n-type semiconductor is produced by doping Group IV silicon with 1 3 1022 atoms/m3 of Group V antimony. Assume that the semiconductor is at room temperature, and that this semiconductor is in the extrinsic range of operation. (a) What is the density of electrons in the conduction band? (b) Calculate the room-temperature electrical conductivity for this semiconductor. Problem 16.14 (a) For the metals with work functions listed in Table 16.4, what combination of metals produces the largest contact voltage, and what is its magnitude? (b) In this couple, which metal is charged positive and which is charged negative? Problem 16.15A pn-junction is produced by doping silicon that has an energy gap of 1.11 eV with arsenic (Group V) to produce the n-type material, and with boron (Group III) to produce the p-type material. (a) At temperatures approaching absolute zero, what is the value of the Fermi energy in the n-type and p-type material? (b) At temperatures approaching absolute zero, what would be the value of the contact potential at a pn-junction of these two materials? (c) At temperatures in the intrinsic range of operation for both of these materials, what is the value of the contact potential? Problem 16.16A pn-junction diode at 300 K produces a current density 1.00 3 104 A/m2, with a forward bias of 0.100 V. Calculate the current density for a forward bias of 0.300 V. Problem 16.17 A parallel-plate vacuum capacitor has an area of 1 3 1024 m2, a spacing between the plates of 1 3 1023 m, and a voltage of 1 V. (a) Calculate the capacitance of this capacitor. (b) A microcapacitor is replacing the capacitor in part (a). It must have the same capacitance; however, the new area is 1 3 10210 m2, and the spacing between the plates is now 1 3 1026 m, with 1 V applied. (b1) What is the dielectric constant and dielectric strength required for the material in the microcapacitor? (b2) Are there any materials available that can meet these requirements, and, if so, what are they? (b3) What is the polarization (P) in this microcapacitor, and how does this polarization compare to the surface-charge density if this were a vacuum capacitor? Electrical Properties of Materials Problem 16.18 The dynamic random access memory (DRAM) in computers is produced by oxidizing a silicon wafer to produce a capacitor with SiO2 as the dielectric, and the memory corresponds to the charge stored on the capacitor. Assume that the thickness of the dielectric is 1 3 1026 m, the plates of the capacitor are square with an edge length of 10 3 1026 m, the dielectric constant or relative permeability of SiO2 is 4, and the applied voltage is 10 V. (a) What is the capacitance? (b) What total charge is stored in this capacitor? (c) What is the charge density on the capacitor plates? (d) What is the polarization of the SiO2 in the capacitor? Problem 16.19 A new polymer is developed, and to characterize the electrical properties of the polymer, the capacitance is determined to be 20.36 3 10213 F in a parallel-plate capacitor that has square plates with a side length of 1022 m, and the thickness of the polymer sheet and the spacing between the plates is 1023 m. (a) What is the dielectric constant or relative permittivity of this new material? (b) If the capacitance measurement is made with an applied voltage of 10 V, what is the magnitude of the polarization in the new polymer? Problem 16.20A piezoelectric actuator is being produced for a scanning tunneling microscope. The piezoelectric actuator is to be 1 3 1023 m in length and produced from barium titanate, which has a dielectric breakdown strength of 6 3 106 V/m. The piezoelectric strain coefficient of the barium titanate is 1.9 3 10210 m/V. Assume that the maximum electric field that can be applied to the piezoelectric material is 1 3 106 V/m. (a) What is the required voltage to produce an atomic-scale displacement of 0.1 nm? (b) What is the displacement produced by 1000 V? (c) With 1000 V applied, is the piezoelectric safe from dielectric breakdown? Problem 16.21 Cubic zirconia stabilized with CaO is used as an oxygen sensor material because of the high electrical conductivity resulting from the fast diffusion of oxygen ions. (a) Calculate the mobility of O21 ions in zirconia at 1373 K if the diffusivity of the oxygen ions is given by 1 DsO 21 d 5 1 3 109 exp 2 2 0.84 eV m2/s kT (b) What is the ratio of the electrical conductivity of the zirconia at 1373 K to that of an n-type silicon semiconductor at room temperature with the same density of Group V charge carriers as oxygen ions? W-67