Electrical Properties

Energy bands and the classification of solids, conduction in metals and its resistivity, non-metallic conductors, intrinsic and extrinsic semiconductors and their applications, and insulating, dielectric, ferroelectric and piezoelectric materials.

📖 Source: Universidad Carlos III de Madrid · Session 13 T6 Electrical properties.pdf

Central idea: the electrical behaviour of a solid is decided by how its electrons fill the energy bands that the atoms form when they come together. A half-filled (or overlapping) band gives a conductor; a full valence band separated by a small gap $E_g$ from an empty conduction band gives a semiconductor; a large gap gives an insulator. In every case the conductivity is carriers times charge times mobility, $\sigma = n\,\lvert q\rvert\,\mu$, and the whole topic is about what controls $n$ (temperature, doping, band gap) and $\mu$ (collisions with vibrations, impurities and defects). Insulators that can be polarised (dielectrics, ferroelectrics, piezoelectrics) close the unit.

This page follows the single official session of Topic 6 (Session 13) slide by slide. Slide 1 is the title and agenda (band structure and electrical properties; conducting materials, metallic and non-metallic; semiconducting materials, intrinsic and extrinsic of types n and p; insulating and dielectric materials). Slides 9, 24 and 50 repeat this agenda as section dividers. Every section names the slides it comes from. Figures that appear on the slides are copied from them as PNG crops, and each caption names its slide. The few drawings that have no figure on the slides (the conductivity scale built from the table of Slide 5, the bar of Ohm's law, the band-gap trends and the cubic BaTiO3 cell) are drawn for this page, and their captions say so. The questions printed on the slides (Q1 to Q5 and Q10) are problems 1 to 5 and 10 of the official sheet, and they are solved on the problems page of this topic.

§ 1 Learning objectives and classification of materials [Session 13 Slides 2, 5]

The slides set three objectives [Session 13 Slide 2]:

Materials are classified by their electrical conductivity $\sigma$, measured in $(\Omega\cdot\text{m})^{-1}$ [Session 13 Slide 5]:

ClassConductivity $\sigma$ in $(\Omega\cdot\text{m})^{-1}$Example
Insulators$10^{-20}$ to $10^{-10}$Polyethylene (PVC coating on the slide)
Semiconductors$10^{-6}$ to $10^{4}$Si, Ge (microprocessors)
Conductorsabout $10^{7}$Metals (copper wire)

Counting from the smallest to the largest value of the table, the scale spans about 27 orders of magnitude (the author's count, not printed on the slide).

10−20 10−16 10−12 10−8 10−4 1 104 108 σ in (Ω·m)⁻¹ (logarithmic scale) Insulators: 10⁻²⁰ to 10⁻¹⁰, e.g. polyethylene Semiconductors: 10⁻⁶ to 10⁴, e.g. Si, Ge Conductors: about 10⁷, e.g. metals
Figure 1. Insulators, semiconductors and conductors on a logarithmic conductivity scale, drawn for this page from the table of Session 13 Slide 5 (the slide gives the ranges only as a table; schematic, not to scale).

The reciprocal of the conductivity is the resistivity $\rho = 1/\sigma$, in $\Omega\cdot\text{m}$ [Session 13 Slide 13].

§ 2 Band structures of solids [Session 13 Slides 3-4]

2.1 From atomic levels to bands [Session 13 Slide 3]

An isolated atom has discrete energy levels. When atoms come within close proximity, their orbitals interact and each atomic level splits: two atoms give two levels, a few atoms give a few close levels, and a large number of atoms (a solid) gives so many levels, so close in energy, that they form a continuous energy band. The slide plots the electron energy against the interatomic separation for an aggregate of atoms, with the core-level electrons staying almost unchanged and the outer levels broadening into bands. Between bands a band gap $E_g$ may remain, where no electron states exist.

Formation of energy bands
Figure 2. Energy levels broaden into bands as the number of atoms grouped together increases: one atom, two atoms, a small number of atoms and a solid, with the band gap Eg and the core-level electrons (Figure 19-2 of the slide, courtesy of John Bravman), copied from Session 13 Slide 3.

2.2 Valence band, conduction band and gap [Session 13 Slide 4]

The electron energy band structure of a solid is given by its interatomic separation: the bands broaden as the atoms approach and the solid sits at the equilibrium interatomic distance, where the bands have the width and the gap shown in the figure. Two bands matter:

Energy bands against inter-atomic distance
Figure 3. Conduction and valence bands widening as the inter-atomic distance decreases, with the equilibrium inter-atomic distance marked, copied from Session 13 Slide 4.

§ 3 Band structure of conductors, semiconductors and insulators [Session 13 Slides 5-8]

3.1 Conductors [Session 13 Slides 6-7]

A material conducts when electrons can find empty states just above the filled ones, so that a small excitation (an electric field) lets them move. There are two ways [Session 13 Slide 6]:

The slide lists the electron configurations used to decide this:

ElementAtomic numberConfiguration
Mg12$1s^2\,2s^2\,2p^6\,3s^2$
Al13$1s^2\,2s^2\,2p^6\,3s^2\,3p^1$
Cu29$1s^2\,2s^2\,2p^6\,3s^2\,3p^6\,3d^{10}\,4s^1$
Printed configuration of aluminium

The slide's table gives aluminium as $3s^2\,3p^1$ (one electron in the $3p$ level), while its caption says that Na, Cu and Al have a half-filled band "since there is only one $s$ electron". The table is kept as printed; Na is not in the table.

For a metal, the occupancy of the electron states changes by an electron excitation as in the figure: before the excitation the states are filled up to $E_f$, and after it an electron has been promoted to an empty state just above $E_f$ [Session 13 Slide 7].

3.2 Semiconductors and insulators [Session 13 Slide 8]

Band structure of conductors
Figure 4. Band structure of conductors: A1) half-filled valence band (one s electron: Na, Cu, Al) and A2) overlapping filled 3s and empty 3p bands (Mg), with the Fermi energy Ef, copied from Session 13 Slide 6.
Band structure of semiconductors and insulators
Figure 5. Band structure of semiconductors (narrow gap, below 2 eV: Si 1.1 eV, Ge 0.67 eV) and insulators (large gap, above 2 eV: diamond 6-7 eV), copied from Session 13 Slide 8.

§ 4 Metallic conductors: Ohm's law and the free-electron model [Session 13 Slides 10-13]

4.1 Electrical and ionic conduction [Session 13 Slide 10]

The macroscopic law is Ohm's law, $I = V/R$, with $I$ the electric current (A), $V$ the potential difference (V) and $R$ the resistance ($\Omega$). The charge can be carried in two ways: by ionic conductivity (aqueous electrolytes, some ceramics) or by electrical (electronic) conductivity (metals). In a conductive metal there is a cloud of electrons around the nuclei (the positive ions).

For a bar of length $l$ and cross-section $A$ the microscopic form is $\mathbf J = \sigma\mathbf E$. Writing $E = V/l$ and $J = I/A$ in it gives $V = (l/\sigma A)\,I$, so $R = \rho\,l/A$ (derived here: the slide only writes $I = V/R$; this is the link with the geometry of the specimens of the sheet).

I I E J V = E l l A J = σ E, σ = 1/ρ, R = V / I = ρ l / A (I = J A and V = E l give R = l / (σ A))
Figure 6. A bar of length l and cross-section A carries a current I under a voltage V: E = V/l, J = I/A, so R = ρ l / A. Drawn for this page: the slide only writes I = V/R (schematic).

4.2 Acceleration of the electrons [Session 13 Slide 10]

When a uniform electric field $E$ is applied, the electrons accelerate. Newton's second law for one electron of charge $q_e$ ($1.6\times10^{-19}$ C) and mass $m_e$ ($9.109\times10^{-31}$ kg) is

$$F = q_e E = m_e a \quad\Rightarrow\quad a = \frac{q_e E}{m_e}$$
Printed units

The slides write the field $E$ in V/m² (and in A/m² on Slide 12). The field strength is in V/m (volt per metre); the units printed are typographical slips and V/m is used here.

4.3 Collisions, drift velocity and mobility [Session 13 Slide 11]

If the electrons only accelerated, the current would grow without limit. For a given $E$ the current $I$ is constant, so the speed of the electrons is constant: there are "friction forces", due to collisions with the ions, that counteract the acceleration. This is the origin of the resistivity. Between collisions the drift velocity grows linearly with time; at each collision it falls back (sawtooth of the figure), so the average drift velocity $v_d$ is the average speed of the electrons in the direction of the force generated by the field:

$$v_d = \frac{\tau\,\lvert q_e\rvert\,E}{m_e} = \mu\,E \qquad\text{with}\qquad \mu = \frac{\tau\,\lvert q_e\rvert}{m_e}$$

$\tau$ is the relaxation time (the mean time between collisions) and $\mu$ is the electron mobility, in m²/(V·s).

Electron path with scattering events
Figure 7. Path of a conduction electron under a field E, with the scattering events and the net electron motion, copied from Session 13 Slide 10.
Drift velocity against time
Figure 8. Electron drift velocity against time in the classical model of a free electron in a metal: it grows between collisions and falls back at each one, with the maximum and average values vmax and vav, copied from Session 13 Slide 11.

4.4 Current density, conductivity and resistivity [Session 13 Slides 12-13]

For a material with $n$ free conducting electrons per unit volume, the current density is the charge that crosses a unit area per second:

$$J = n\,\lvert q_e\rvert\,v_d = n\,\lvert q_e\rvert\,\mu\,E \qquad\text{and, since}\quad J = \sigma E:\qquad \sigma = n\,\lvert q_e\rvert\,\mu$$

In metals the carriers are the non-bonded electrons, so the conductivity depends on the number of electrons per unit volume and on their mobility. The resistivity is $\rho = 1/\sigma$, in $\Omega\cdot\text{m}$. The slide adds that the resistivity grows with the disorder in the metal structure and with the number of collisions between the non-bonded electrons and the metallic ions [Session 13 Slide 13].

§ 5 Electrical resistivity of metals [Session 13 Slides 14-15]

5.1 Matthiessen's rule [Session 13 Slide 14]

The total resistivity of a metal is the sum of a thermal part and a residual part:

$$\rho_{total} = \rho_T + \rho_r$$

$\rho_T$ is the thermal resistivity, produced by the vibrations of the lattice, and $\rho_r$ is the residual resistivity, produced by impurities and defects (next section). The slide's plot of resistivity against temperature shows $\rho_r$ as a constant floor (the curve tends to it at low temperature) and $\rho_T$ as the temperature component added on top.

Matthiessen's rule
Figure 9. Electrical resistivity against temperature: the temperature component ρT on top of the residual component ρr, copied from Session 13 Slide 14.

5.2 Thermal resistivity [Session 13 Slide 14]

The thermal resistivity grows linearly with the temperature:

$$\rho_T = \rho_0\,(1 + \alpha_T\,\Delta T)$$

$\rho_0$ is the resistivity at 0 °C (in $\Omega\cdot\text{m}$) and $\alpha_T$ is the temperature coefficient of resistivity (in ${}^{\circ}\text{C}^{-1}$). Because $\rho_0$ is the value at 0 °C, $\Delta T$ is the temperature in degrees Celsius. The slide's table:

MetalResistivity at 0 °C in $\mu\Omega\cdot\text{cm}$$\alpha_T$ in ${}^{\circ}\text{C}^{-1}$
Aluminium2.70.0039
Copper1.60.0039
Gold2.30.0034
Iron90.0045
Silver1.470.0038
Resistivity at 0 degrees C and temperature coefficient of selected metals
Figure 10. Resistivity at 0 °C and temperature coefficient of resistivity of aluminium, copper, gold, iron and silver, copied from Session 13 Slide 14.

The second chart of the slide shows the effect of temperature on the resistivity of selected metals (Fe, Rb, Os, Co, Na, Ir, Rh, Mg, Au, Cu, Ag): almost straight lines above about −100 °C.

Resistivity against temperature of selected metals
Figure 11. Effect of temperature on the electrical resistivity of selected metals, copied from Session 13 Slide 14.

5.3 Question 1: equal resistivity of silver and aluminium [Session 13 Slide 15]

Q1 asks at which temperature a silver wire has the same resistivity as an aluminium wire at 45 °C, with $\rho_0 = 1.47\times10^{-6}$ and $2.70\times10^{-6}\ \Omega\cdot\text{cm}$ and $\alpha_T = 0.0038$ and $0.0039\,{}^{\circ}\text{C}^{-1}$ for silver and aluminium. The slide first evaluates the aluminium resistivity with the law above, $\rho_{Al,45} = 2.7\times10^{-6}(1 + 0.0039\cdot45) = 3.173\times10^{-6}\ \Omega\cdot\text{cm}$, and then solves the law for the temperature of silver, $T = (\rho_T - \rho_{0,Ag})/(\rho_{0,Ag}\,\alpha_{T,Ag}) = 304.86\,{}^{\circ}\text{C}$. This is problem 1 of the sheet, and its page redoes the calculation with the units in SI and the unrounded intermediate value (the key of the sheet is 305 °C).

§ 6 Residual resistivity: impurities and defects [Session 13 Slides 16-17]

The residual resistivity $\rho_r$ does not depend on temperature. It depends on [Session 13 Slide 16]:

So $\rho_r = \rho_i + \rho_d$ and, with Matthiessen's rule, $\rho_{total} = \rho_T + \rho_i + \rho_d$. $\rho_r$ increases with the degree of distortion of the metal lattice: a foreign atom, a large one or a small one, disturbs the periodicity around it. The slide's chart shows how small additions of various elements raise the room-temperature resistivity of copper: for each element the resistivity grows linearly with the weight percent of impurity, with a slope that depends on the element (the steepest are P, Fe, Si and As; Ni, Pb and Ag are the gentlest among those plotted).

Effect of impurities on the resistivity of copper
Figure 12. Effect of small additions of various elements on the room-temperature electrical resistivity of copper, copied from Session 13 Slide 16.

6.1 Scattering of the electrons [Session 13 Slide 17]

A conduction electron moves through a perfect crystal without scattering. In a crystal heated to a high temperature the atoms vibrate and displace from their sites, and in a crystal containing atomic-level defects (foreign atoms, vacancies, dislocations) the periodicity is broken too. In both cases the electron is scattered, which reduces its mobility and therefore the conductivity. The chart of the slide shows copper with different nickel contents and deformed copper: the curves are parallel, separated by the constant residual terms $\rho_i$ and $\rho_d$, while the temperature component $\rho_T$ is the same.

Movement of an electron through a crystal
Figure 13. Movement of an electron through (a) a perfect crystal, (b) a crystal heated to a high temperature and (c) a crystal containing atomic-level defects, copied from Session 13 Slide 17.
Resistivity of copper and Cu-Ni alloys against temperature
Figure 14. Resistivity against temperature of pure, alloyed (Cu + Ni) and deformed copper, showing the thermal, impurity and deformation parts, copied from Session 13 Slide 17.

§ 7 Mechanical properties and resistivity [Session 13 Slides 18-19]

All hardening methods cause, to a greater or lesser extent, a decrease of the conductivity, [Session 13 Slide 18]. The reason, in the author's reading (not printed on the slide), is that every hardening method of Topic 5 works by putting obstacles in the way of the dislocations, and the same obstacles scatter the electrons. The slide plots yield strength and conductivity against the percent of cold work and against the percent of zinc (a solid solution):

Yield strength and conductivity against cold work and zinc content
Figure 15. Yield strength and electrical conductivity against percent cold work (left) and percent Zn (right), copied from Session 13 Slide 18.

Q2 on Slide 19 asks what happens to the conductivity and the tensile strength of an aluminium alloy wire when it is alloyed with copper, when its working temperature is lowered, when its grain size is increased, when intermetallic precipitates are formed and when it is cold worked, and which strategy gives the lightest wire with the same strength and conductivity. It is problem 2 of the sheet.

§ 8 Non-metallic conductors [Session 13 Slides 20-23]

8.1 Conductive polymers [Session 13 Slide 20]

A few thermoset polymers, when they are specifically manufactured, can become conductive. Their advantages against metals are flexibility and easy formability. The main mechanism for transmitting the charge is the movement of delocalised electrons: an exchange reaction of electrons between neighbouring redox sites.

8.2 Graphene [Session 13 Slide 21]

Graphene is a single layer of a hexagonal lattice formed by hexagons with an atom at each vertex (graphite is a stack of these layers). It has a huge $\pi$ cloud with delocalised electrons that can move all over the surface at a speed higher than in any other material (about $1000$ km/s). The slide compares

MaterialConductivity in $(\Omega\cdot\text{m})^{-1}$
Graphene$0.96\times10^{8}$
Cu$0.60\times10^{8}$
Si$4.5\times10^{-4}$
Printed symbol

The slide writes these values with the symbol $\rho$, but the unit $(\Omega\cdot\text{m})^{-1}$ is the unit of conductivity, so they are $\sigma$.

Graphene lattice
Figure 16. Graphene: a single layer of carbon atoms on a hexagonal lattice, copied from Session 13 Slide 21.
σ bonds and π cloud of graphene
Figure 17. The σ bonds in the plane and the π cloud of delocalised electrons above and below it, copied from Session 13 Slide 21.

8.3 Conductive ceramics [Session 13 Slide 22]

Ceramics are traditionally insulators, but some advanced ceramics are good electrical conductors. The main mechanism for transmitting the charge is

Movement of ions or of vacancies
Figure 18. Ionic conduction by the movement of ions or, equivalently, of vacancies, copied from Session 13 Slide 22.
Interstitial cation
Figure 19. Ionic conduction by a small cation (ri smaller than ratom) moving through the interstitial voids, copied from Session 13 Slide 22.

Q3 on Slide 23 compares the effect of vacancies, interstitial cations, interstitial anions, temperature, disorder of the lattice and doping with other valences on an ionic ceramic and on a metal. It is problem 3 of the sheet.

§ 9 Semiconductors: definition and types [Session 13 Slides 25-26]

Semiconductors are materials whose electrical conductivity is approximately midway between the values for good conductors and insulators. In the energy-band model the electronic band structure consists of a full valence band and an empty conduction band, separated by an energy gap $E_g < 2$ eV [Session 13 Slide 25]. The materials are mediocre conductors, but their conductivity is easily controllable, and this became useful in the twentieth century with the manufacture of materials of extremely high chemical and structural purity.

Two types depending on the origin of the carriers [Session 13 Slide 26]:

§ 10 Intrinsic semiconductors [Session 13 Slides 27-30]

10.1 Excitation and the two kinds of carriers [Session 13 Slides 27-28]

In an intrinsic semiconductor an electric or thermal excitation makes an electron pass from the valence band to the conduction band. This generates a hole in the valence band, and both the free electron and the hole are charge carriers. In the covalent crystal a bonding valence electron cannot move unless it receives enough energy to leave its bonding position; once a critical amount of energy has been supplied, it becomes a free conduction electron and leaves a positively charged hole in the lattice.

Excitation in an intrinsic semiconductor
Figure 20. Intrinsic semiconductor before and after excitation: an electron excited across the gap Eg becomes a free electron in the conduction band and leaves a hole in the valence band, so both a free electron and a hole are generated, copied from Session 13 Slide 27.

There are two types of carriers: electrons in the conduction band (charge $-1.6\times10^{-19}$ C, density $n_e$) and holes in the valence band (charge $+1.6\times10^{-19}$ C, density $n_H$). The conductivity is the sum of both contributions:

$$\sigma = n_e\,\lvert q_e\rvert\,\mu_e + n_H\,\lvert q_H\rvert\,\mu_H = n\,\lvert q_e\rvert\,(\mu_e + \mu_H) \qquad\text{since}\quad n_e = n_H = n,\ \ \lvert q_e\rvert = \lvert q_H\rvert$$

$\mu_e$ and $\mu_H$ are the mobilities of electrons and holes, and $\lvert q_e\rvert$ is the absolute value of the electron or hole charge. In an intrinsic semiconductor $\mu_e > \mu_H$ [Session 13 Slide 33].

10.2 Effect of temperature [Session 13 Slide 29]

For an intrinsic semiconductor both $n$ and $p$ increase with temperature:

$$n_i \propto \exp\!\left(-\frac{E_g}{2kT}\right)$$

$k$ is Boltzmann's constant (the sheet gives $k_B = 8.62\times10^{-5}$ eV/K), $T$ the temperature in kelvin and $E_g$ the band gap; $n_i = n = p$ is the number of excited electrons (and holes) per unit volume. Since $\sigma \propto n_i$,

$$\sigma = \sigma_0\exp\!\left(-\frac{E_g}{2kT}\right) \qquad\Rightarrow\qquad \ln\sigma = \ln\sigma_0 - \frac{E_g}{2kT}$$

Plotted as $\ln\sigma$ against $1/T$ the law is a straight line of slope $-E_g/2k$, as the slide's chart shows. Temperature also disorders the crystal and lowers the mobility, but this effect is masked by the increase in the number of carriers [Session 13 Slide 29]. Compare with a metal, where $n$ is fixed and only the mobility changes, so that $\sigma$ falls with temperature.

Conductivity of an intrinsic semiconductor against temperature
Figure 21. Electrical conductivity against temperature (left) and against 1/T with the slope −Eg/2k (right) of an intrinsic semiconductor, copied from Session 13 Slide 29.

Q4 on Slide 30 asks to justify the different effect of a temperature increase on a metal and on an intrinsic semiconductor. It is problem 4 of the sheet, and problems 7 (germanium at 75 °C) and 8 (carrier concentrations of silicon) apply the formulas of this section.

§ 11 Intrinsic semiconductors: elemental and compound materials [Session 13 Slides 31-34]

11.1 Elemental semiconductors [Session 13 Slide 31]

The elemental semiconductors are the pure elements of group IVA, silicon and germanium, bonded by covalent bonds. As the atom size increases, $E_g$ decreases, and the conductivity of the intrinsic semiconductor increases. The figure of the slide shows the bands of Si and Ge at $0$ K (no electrons in the conduction band, gap printed as $1.09$ eV for Si and $0.72$ eV for Ge) and at $300$ K, when a few electrons have crossed the gap. The data table of the slide:

PropertySiliconGermanium
Energy gap, eV1.10.67
Electron mobility $\mu_n$, m²/(V·s)0.1350.39
Hole mobility $\mu_p$, m²/(V·s)0.0480.19
Intrinsic carrier density $n_i$, carriers/m³$1.5\times10^{16}$$2.4\times10^{19}$
Intrinsic resistivity $\rho$, $\Omega\cdot$m23000.46
Density, g/m³$2.33\times10^{6}$$5.32\times10^{6}$
Properties of silicon and germanium
Figure 22. Energy gap, mobilities, intrinsic carrier density, intrinsic resistivity and density of Si and Ge, copied from Session 13 Slide 31.
Si Eg = 1.1 eV atom size increases Ge Eg = 0.67 eV Eg decreases, σ increases Group 14, covalent bonds more covalent more ionic Group 14 Si, Ge Groups 13-15 GaAs, InSb Groups 12-16 CdS, ZnTe Eg increases σ decreases (intrinsic semiconductors)
Figure 23. Trends of the band gap: it falls as the atom size grows in group 14 and rises as the bonding becomes less covalent and more ionic; the conductivity follows the opposite trend. Drawn for this page from the statements of Session 13 Slides 31 and 32, which give these trends in words (schematic).

11.2 Compound semiconductors [Session 13 Slides 32-33]

Compounds of elements of groups IIIA and VA (GaAs, InSb) and of groups IIB and VIA (CdS, ZnTe) are also intrinsic semiconductors. Going from the group-14 elemental semiconductors (purely covalent) to the 13-15 and then to the 12-16 compounds, the covalent character decreases and the ionic character increases: $E_g$ increases and, for an intrinsic semiconductor, $\sigma$ decreases. The slide also classifies semiconductors as simple (Si, Ge) and compound (III-V: GaAs, InP, GaN; II-VI: ZnS, ZnSe; IV-IV: SiC, SiGe).

The conductivity of an intrinsic semiconductor is determined by $n$: the number of charge carriers is small, but the high chemical and structural purity favours high mobilities. The table of the slide (it prints slightly different Si and Ge mobilities from the previous one):

Material$E_g$ in eV$\sigma$ in $(\Omega\cdot\text{m})^{-1}$$\mu_e$ in m²/(V·s)$\mu_H$ in m²/(V·s)
Si1.11$4\times10^{-4}$0.140.05
Ge0.672.20.380.18
GaP2.25n/a0.030.015
GaAs1.42$10^{-6}$0.850.04
InSb0.17$2\times10^{4}$7.70.07
CdS2.40n/a0.03n/a
ZnTe2.26n/a0.030.01
Band gap, conductivity and mobilities of intrinsic semiconductors
Figure 24. Band gap, electrical conductivity and mobilities of Si, Ge, GaP, GaAs, InSb, CdS and ZnTe, copied from Session 13 Slide 33.

Q10 on Slides 34 and 41 asks to order semiconductors by conductivity at the same temperature and to explain the criterion; Slide 34 poses AlP, GaAs and InSb, and Slide 41 the doped pairs (Ge with As or Sb, Si with B or In). It is problem 10 of the sheet.

§ 12 Extrinsic semiconductors: type n and type p [Session 13 Slides 35-39]

12.1 Doping [Session 13 Slide 35]

Extrinsic semiconductors are the most usual type: the dopants add charge carriers, raise the conductivity and make it possible to manufacture devices with special properties (transistors and others). Two types, depending on the doping atoms:

12.2 Type n [Session 13 Slide 36]

The impurity atoms have more valence electrons than the atoms of the network, so they behave as electron donors (group VA: P, As, Sb; the periodic-table fragment of Slide 35 also lists N). In silicon ($4$ valence electrons) a phosphorus atom ($5$ valence electrons) uses four of them in the covalent bonds and has one excess electron, loosely bound to it. This electron needs only a small energy to jump into the conduction band. In band language the donor atoms create a donor level $E_d$ inside the gap, very close to the conduction band, so the excitation energy is $\Delta E = E_c - E_d$, small compared with $E_g$. The distance from the donor level to the conduction band determines $n_e$. The electrons in the conduction band are the majority carriers, $n_e \gg n_H$, and

$$\sigma \approx n_e\,\lvert q_e\rvert\,\mu_e$$
Donor level in an n-type semiconductor
Figure 25. n-type semiconductor: the donor level Ed just below the conduction band Ec, with ΔE = Ec − Ed, and the table of groups 4A and 5A of the periodic table, copied from Session 13 Slide 36.

12.3 Type p [Session 13 Slide 37]

The impurity atoms have fewer valence electrons than the atoms of the network, so they behave as acceptors (B, Al, Ga and In of group IIIA; in silicon, boron has $3$ valence electrons). There is a bond with a missing electron: an excess of holes. The acceptor creates an acceptor level $E_a$ in the gap, just above the top of the valence band, so a small energy $\Delta E = E_a - E_v$ lifts an electron from the valence band into the acceptor level, leaving a hole in the valence band. The holes moving through the valence band are the majority carriers, $p \gg n$, and

$$\sigma \cong p\,\lvert q_e\rvert\,\mu_H$$
Printed group

Slide 37 writes "B, Al, Ga (group VA)", but the table printed on the same slide places B, Al, Ga and In in group 3A (IIIA); group VA is the one of the donors. The acceptors are IIIA.

Acceptor level in a p-type semiconductor
Figure 26. p-type semiconductor: the acceptor level Ea just above the valence band Ev, with ΔE = Ea − Ev, the electron and the positive hole, and the table of groups 3A and 4A of the periodic table, copied from Session 13 Slide 37.

12.4 Excitation and majority and minority carriers [Session 13 Slide 38]

In a donor semiconductor the excitation lifts an electron from the donor state to the conduction band, generating a free electron. In an acceptor semiconductor the excitation lifts an electron from the valence band into the acceptor level, generating a hole in the valence band. The slide ends with the table of carrier names:

SemiconductorMajority-carrier concentrationMinority-carrier concentration
n-type$n_n$ (electrons in n-type material)$p_n$ (holes in n-type material)
p-type$p_p$ (holes in p-type material)$n_p$ (electrons in p-type material)

12.5 Question 5 [Session 13 Slide 39]

Q5 asks whether silicon doped with antimony, germanium doped with arsenic and silicon doped with gallium (substitutional solid solutions) are n-type or p-type and whether the impurities act as donors or acceptors. The version of the sheet changes the second case to silicon doped with arsenic. It is problem 5 of the sheet, and problem 6 (phosphorus in silicon) applies $\sigma \approx n_e\lvert q_e\rvert\mu_e$.

§ 13 Extrinsic semiconductors: dopants and mobility [Session 13 Slides 40-42]

13.1 Donor and acceptor levels [Session 13 Slides 40-41]

The slide gives the donor and acceptor energy gaps (in electron volts) when silicon and germanium are doped, that is, the energy $E_c - E_d$ or $E_a - E_v$ that an electron needs, and the maximum solid solubility of each dopant:

DopantSi: $E_d$Si: $E_a$Ge: $E_d$Ge: $E_a$
P0.0450.0120
As0.0490.0127
Sb0.0390.0096
B0.0450.0104
Al0.0570.0102
Ga0.0650.0108
In0.1600.0112
Donor and acceptor energy gaps in silicon and germanium
Figure 27. Donor and acceptor energy gaps (in eV) when silicon and germanium are doped, copied from Session 13 Slide 40.

The arrows of the slide say that, as the ionic radius of the dopant increases, the conductivity increases for the donors (P, As, Sb: Si and Ge) and decreases for the acceptors (B, Al, Ga, In), in agreement with the table, where the acceptor energies grow with the size (B $<$ Al $<$ Ga $<$ In). The author's rule of thumb for comparing dopants (not printed on the slide) is to compare $E_d$ or $E_a$: the smaller the energy, the more carriers at a given temperature. The lower table of the slide gives the maximum solid solubility of the dopants, in atoms/m³:

HostDopantGroupMaximum solubility in atoms/m³
SiBIIIA$600\times10^{24}$
SiAlIIIA$20\times10^{24}$
SiGaIIIA$40\times10^{24}$
SiPVA$1000\times10^{24}$
SiAsVA$2000\times10^{24}$
SiSbVA$70\times10^{24}$
GeAlIIIA$400\times10^{24}$
GeGaIIIA$500\times10^{24}$
GeInIIIA$4\times10^{24}$
GeAsVA$80\times10^{24}$
GeSbVA$10\times10^{24}$
Maximum solid solubility of dopants in silicon and germanium
Figure 28. Maximum solid solubility of some dopants in Si and Ge, copied from Session 13 Slide 40.

Slide 41 repeats the same two tables with the second part of Q10: germanium with 5 ppm of As or of Sb, silicon with 10 ppm of B or of In (the slide prints the germanium pair twice).

13.2 Mobility of the carriers [Session 13 Slide 42]

In extrinsic semiconductors also $\mu_e > \mu_H$ (the ratio $\mu_e/\mu_H$ is $2.81$ for Si and $2.05$ for Ge). The distortion of the crystal structure caused by the doping is negligible, so $\sigma$ is controlled by the number of charge carriers. The chart shows the room-temperature electron and hole mobilities of silicon against the dopant concentration: both are almost constant (about $0.13$ and $0.05$ m²/(V·s)) up to about $10^{21}$ m⁻³ and fall for high doping (electrons about $0.10$ m²/(V·s) at $10^{22}$ m⁻³ and below $0.01$ m²/(V·s) at $10^{25}$ m⁻³; read from the chart).

Mobility of electrons and holes in silicon against dopant concentration
Figure 29. Dependence of the room-temperature electron and hole mobilities on the dopant concentration for Si, copied from Session 13 Slide 42.

§ 14 Extrinsic semiconductors: effect of temperature [Session 13 Slides 43-44]

At low temperature the number of ionised impurities per unit volume determines $\sigma$. As $T$ increases, the number of ionised atoms increases and so does $\sigma$: this is the extrinsic range, whose slope in the $\ln\sigma$ against $1/T$ plot depends on the type of semiconductor. For the n-type and p-type the slide writes

$$\sigma = \sigma_0\exp\!\left(-\frac{E_c - E_d}{2kT}\right) \qquad\text{and}\qquad \sigma = \sigma_0\exp\!\left(-\frac{E_a - E_v}{2kT}\right)$$

so $\ln\sigma = \ln\sigma_0 - \dfrac{E_c-E_d}{2kT}$ for the n-type and the same with $E_a - E_v$ for the p-type: the slope of the line is $-(E_c-E_d)/2k$ (n-type) or $-(E_a-E_v)/2k$ (p-type).

Printed energy for the p-type

Slide 43 prints $E_a - E_c$ for the p-type. The ionisation energy of an acceptor is the distance from the acceptor level to the top of the valence band, $E_a - E_v$ (Slide 37), and that is the form used here.

Slope label of the chart

The chart on the right of Slide 43 labels the extrinsic slope as $-(E_c - E_d)/k$, without the factor 2 that the equations on the left of the same slide carry, $-(E_c - E_d)/2k$. The factor 2 is the one that follows from $\sigma \propto \exp[-(E_c - E_d)/2kT]$, and it is the one used on this page.

At intermediate temperature all the impurities have been ionised and $\sigma$ is almost constant: the exhaustion range for the n-type (all the donor electrons have been lost) and the saturation range for the p-type (all the atoms are ionised). At high temperature the intrinsic generation across the gap takes over, with slope $-E_g/2k$: the intrinsic range [Session 13 Slide 43]. Slide 44 summarises it: at low and moderate $T$ the material behaves as an extrinsic conductor and $\sigma$ is governed by the number of ionised atoms; at high $T$ the conductivity slightly decreases, because the number of carriers stays constant and the disordering effect of the temperature (lower mobility) becomes visible; at very high $T$ the material behaves as an intrinsic conductor.

Effect of temperature on an extrinsic semiconductor
Figure 30. ln σ against 1/T of an extrinsic semiconductor: intrinsic range at high temperature (slope −Eg/2k), exhaustion range and extrinsic range at low temperature (slope printed −(Ec − Ed)/k), copied from Session 13 Slide 43.

§ 15 Applications of semiconductors [Session 13 Slides 45-49]

15.1 The p-n junction [Session 13 Slide 45]

A p-n junction is a boundary or interface between two types of semiconductor material, p-type and n-type, inside a single crystal. It is the basis of the diode, the photovoltaic cell and the LED of the next sections (the slide links two videos, listed at the end of this page).

The slide draws the junction in three steps: the carriers on both sides, the depletion layer (fixed negative charge on the p side and positive on the n side, with the internal electric field) and the band diagram with the diffusion potential $V_{bi}$ between the two sides.

The p-n junction: carriers, depletion layer and energy bands
Figure 31. The p-n junction as drawn on Session 13 Slide 45: (1) holes and electrons on both sides, (2) the depletion layer with its internal electric field, (3) the band bending with the diffusion potential, copied from the slide.

15.2 Photoconductivity and photovoltaic cells [Session 13 Slide 46]

Electrons jump from the valence to the conduction band due to thermal excitation, due to intense electric fields or due to exposure to radiation of a given wavelength. Photoconductivity is the phenomenon in which a material becomes more conductive due to the absorption of electromagnetic radiation, such as visible light, whose photon energy must be at least $E_g$. The same effect used to generate electricity gives the photovoltaic cells; solar light has $E_g \sim 1.5$ eV. The slide's absorption figure marks that silicon absorbs photons above $1.1$ eV, gallium arsenide above $1.43$ eV and aluminium gallium arsenide above $1.7$ eV, while photons of lower energy go through.

Photoconductivity and the photovoltaic cell
Figure 32. Photoconductivity: a photon hν lifts an electron across the gap Eg, and the photovoltaic cell (emitter, base, contacts, external load) in which sunlight creates electron-hole pairs, copied from Session 13 Slide 46.
Absorption thresholds of Si, GaAs and AlGaAs
Figure 33. Silicon absorbs photons above 1.1 eV, gallium arsenide above 1.43 eV and aluminium gallium arsenide above 1.7 eV; photons of lower energy go through, copied from Session 13 Slide 46.

15.3 LED [Session 13 Slide 47]

A light-emitting diode is a semiconductor doped to create a p-n junction that emits monochromatic light when it is directly polarised (forward bias): electrons from the n side and holes from the p side recombine at the junction and release the energy $E_g$ as a photon. The colour depends on the distance between the valence and the conduction band, as in the slide's table (the band gap increases from top to bottom and the wavelength increases from bottom to top):

MaterialRadiation
AlGaAsred and infra-red
GaNgreen
GaPgreen
ZnSeblue
InGaNblue
SiCblue
Diamondultra-violet
Light-emitting diode
Figure 34. LED: forward-biased p-n junction and its band diagram, where electrons and holes recombine and emit light, copied from Session 13 Slide 47.

15.4 Thermistors [Session 13 Slide 48]

Thermistors are devices based on the change of the electrical conductivity of semiconductors with temperature. The chart compares aluminium, whose conductivity decreases with temperature (reducing mobility), with germanium, whose conductivity increases by several orders of magnitude between $0$ and 400 °C (increasing charge carriers).

Conductivity of aluminium and germanium against temperature
Figure 35. Electrical conductivity against temperature of aluminium (reducing mobility) and germanium (increasing charge carriers), the basis of thermistors, copied from Session 13 Slide 48.

15.5 Transistors [Session 13 Slide 49]

A transistor is a semiconductor device that can be used to amplify electrical signals. The slide shows the downscaling of the transistor and of its interconnects through the generations of $180$ nm (1999), $130$ nm (2001), $90$ nm (2003), $65$ nm (2005) and $45$ nm (2007), down to prototypes of $30$ nm and $20$ nm, with micrographs and a plot of the size against the year (source: Intel).

§ 16 Insulating and dielectric materials [Session 13 Slides 51-54]

16.1 Dipoles and polarisation [Session 13 Slide 51]

An electrically insulating material (large $E_g$) that exhibits, or can be made to exhibit, a dipolar structure is a dielectric. The dipole moment of two charges $+q$ and $-q$ separated by a distance $d$ is

$$p = q\cdot d$$

Without field the dipoles of the material are randomly oriented and the net polarisation is zero (unpolarised); an applied electric field polarises the material by aligning the dipoles, and opposite charges appear on the faces next to the plates.

Dipole moment
Figure 36. Dipole moment p of two charges −q and +q a distance d apart, copied from Session 13 Slide 51.
Unpolarised and polarised dielectric
Figure 37. Dielectric unpolarised (random dipoles) and polarised by an applied electric field, copied from Session 13 Slide 51.

16.2 Insulators, dielectrics and capacitance [Session 13 Slide 52]

Highly resistive materials have two applications, depending on their response to the electric field:

The capacitance of two parallel plates of area $A$ separated by a distance $d$ is

$$C = \varepsilon_r\,\varepsilon_0\,\frac{A}{d} \qquad\text{and}\qquad C = \varepsilon_r\,C_0 \quad\text{with}\quad C_0 = \varepsilon_0\frac{A}{d}$$

$\varepsilon_0 = 8.85\times10^{-12}$ F/m is the permittivity of vacuum, $\varepsilon_r$ the dielectric constant (relative permittivity) and $C_0$ the capacitance with vacuum between the plates. The dielectric polarises, so the net charge at its surfaces ($-Q'$ and $+Q'$) lets the plates store more charge, $Q_0 + Q'$, for the same voltage.

Parallel-plate capacitor with a dielectric
Figure 38. Parallel-plate capacitor with vacuum between the plates (charges ±Q0) and with a dielectric, whose surface charges ∓Q' let the plates hold Q0 + Q', copied from Session 13 Slide 52.

16.3 Loss factor and dielectric strength [Session 13 Slide 53]

Capacitors are normally used in alternating-current circuits, so the dipoles must realign fast when the field changes. The loss factor $\tan\delta$ is the amount of electrical energy lost, as heat, by a capacitor in an alternating circuit, due to the friction among dipoles when they try to reorient. It increases with moisture, temperature and the frequency of the current, and these also lower the resistivity.

The dielectric strength is the maximum electric field that the dielectric can sustain without electrical breakdown:

$$E_{break} = \frac{U_{break}}{d}$$

$U_{break}$ is the breakdown voltage and $d$ the separation between the conductors (in mm on the slide).

Dipoles in an alternating field
Figure 39. Dipoles of a dielectric realigning when the field between the plates changes sign, the origin of the loss factor tan δ, copied from Session 13 Slide 53.

16.4 What an insulator should have [Session 13 Slide 54]

RequirementWhy
High resistivity $\rho$to avoid current leaks
High dielectric strength $E_{break}$to avoid breakdowns of the insulation
Small loss factor $\tan\delta$(no reason printed; see the loss factor above: energy lost as heat)
Low permittivity $\varepsilon$to avoid polarisation

The slide closes with a table of typical electrical insulators ("..." marks a value the slide does not give):

MaterialVolume resistivity in $\Omega\cdot$mDielectric strength in V/milDielectric strength in kV/mmDielectric constant $K$ at 60 Hz$K$ at $10^6$ HzLoss factor at 60 HzLoss factor at $10^6$ Hz
Electrical porcelain insulators$10^{11}$ to $10^{13}$55-3002-126...0.06...
Steatite insulators$>10^{12}$145-2806-11660.008-0.0900.007-0.025
Fosterite insulators$>10^{12}$2509.8...6...0.001-0.002
Alumina insulators$>10^{12}$2509.8...9...0.0008-0.009
Soda-lime glass............7.2...0.009
Fused silica...8......3.8...0.00004

§ 17 Ferroelectric materials [Session 13 Slides 55-57]

17.1 Spontaneous polarisation [Session 13 Slide 55]

Ferroelectric materials have permanent dipoles that can remain aligned in the absence of an electric field: they show spontaneous polarisation with no external field. If a ferroelectric is slowly cooled to a temperature below the critical temperature in the presence of a strong electric field, the anions are displaced towards the positive side of the field and the cations towards the negative side, which produces a permanent electric dipole (the slide draws a perovskite cell with $\text{Ti}^{4+}$ or $\text{Zr}^{4+}$ at the centre, $\text{O}^{2-}$ at the faces and $\text{Pb}^{2+}$ at the corners, non-polarised and polarised, for $T < T_{Curie}$).

17.2 Hysteresis and Curie temperature [Session 13 Slide 56]

The electric-field hysteresis cycles are similar to those of ferromagnetic materials with the magnetic field: the polarisation $P$ (or $D$) against the field $E$ shows a saturation polarisation, a remanent polarisation $P_r$ that stays when the field is removed, a non-remanent part, and a coercive field $E_c$ that must be applied in the opposite direction to bring $P$ to zero. The area under the hysteresis curve is related to the energy required to change the orientation of the dipoles.

The Curie temperature $T_c$ is the temperature above which the material loses its ferroelectric properties and becomes a plain dielectric; typically $-200\,{}^{\circ}\text{C} < T_c < 640\,{}^{\circ}\text{C}$. Example: BaTiO₃ has $T_c = 120\,{}^{\circ}\text{C}$. Below 120 °C the structure is tetragonal; above 120 °C the crystal structure of BaTiO₃ is cubic and the permanent dipoles disappear.

Ferroelectric hysteresis loop
Figure 40. Polarisation against electric field of a ferroelectric: saturation polarisation Ps, remanent polarisation Pr, non-remanent polarisation and coercive field Ec, copied from Session 13 Slide 56.
T > 120 °C: cubic no permanent dipole Ba²⁺ Ti⁴⁺ O²⁻ BaTiO3 above the Curie temperature
Figure 41. BaTiO3 above the Curie temperature (120 °C): cubic cell with the Ti4+ ion at the centre, so there is no permanent dipole. Drawn for this page: the slide states the cubic structure in words and draws only the tetragonal cell of the next figure (schematic).

The slide gives the dimensions of the cell, 0.403 nm for the vertical edge and 0.398 nm for the horizontal ones, and the displacements of the ions in the tetragonal phase (0.006 nm and 0.009 nm).

BaTiO3 cell dimensions and ion displacements
Figure 42. BaTiO3 unit cell with its dimensions (0.403 nm by 0.398 nm) and, on the right, the displacements of the ions in the tetragonal phase (0.006 nm and 0.009 nm), copied from Session 13 Slide 56.

17.3 Dielectric constants [Session 13 Slide 57]

MaterialClass$\varepsilon_r$
BaTiO₃Ferroelectric4000
KNbO₃Ferroelectric700
Rochelle saltFerroelectric170
TiO₂Dielectric100 to 50
Al₂O₃Dielectric10
PorcelainDielectric6.0
PETDielectric3.2
PVCDielectric3.5
PolyethyleneDielectric2.3
AirDielectric1.0

$\varepsilon_r(\text{ferroelectric}) \gg \varepsilon_r(\text{common dielectric})$, so ferroelectrics give small and very efficient capacitors (for example, for computer memories). Since $C = \varepsilon_r C_0$, a capacitor with BaTiO₃ stores about $4000$ times the charge of the same capacitor with vacuum.

§ 18 Piezoelectric materials [Session 13 Slides 58-60]

Piezoelectric materials are ferroelectric materials that can be polarised when an external mechanical stress is applied. They occur in ceramics whose crystal structure has a low degree of symmetry [Session 13 Slide 58]. Two effects:

Direct piezoelectric effectReverse piezoelectric effect (electrostriction)
External stressExternal electric field
$\downarrow$ crystalline structure deformation$\downarrow$ dipole alteration
$\downarrow$ dipole alteration$\downarrow$ crystalline structure deformation
Electric field (voltage) generationDimensional change of the material
Direct piezoelectric effect
Figure 43. Direct piezoelectric effect: an external stress deforms the crystal structure, alters the dipoles and generates an electric field, copied from Session 13 Slide 58.
Reverse piezoelectric effect
Figure 44. Reverse piezoelectric effect (electrostriction): an external electric field alters the dipoles and deforms the crystal structure (tension or compression), copied from Session 13 Slide 58.

They are used to transform electrical energy into mechanical energy and vice versa [Session 13 Slide 59], for example in lighters (a mechanical shock converted into a spark), transducers in speakers, inkjet printers (a piezoelectric actuator pushes the ink droplet out of the nozzle with a pulsed voltage), and microphones and hydrophones [Session 13 Slide 60].

§ 19 Collected formulas of the topic

Everything below is taken from the sections above; it is a quick reference, not new content.

QuantityFormulaWhere
Resistance of a bar$R = \rho\,l/A$ (derived from $J = \sigma E$)Section 4
Drift velocity and mobility$v_d = \mu E$, $\ \mu = \tau\lvert q_e\rvert/m_e$Section 4
Conductivity of a conductor$\sigma = n\lvert q_e\rvert\mu = 1/\rho$Section 4
Matthiessen's rule$\rho_{total} = \rho_T + \rho_r = \rho_T + \rho_i + \rho_d$Sections 5, 6
Thermal resistivity of a metal$\rho_T = \rho_0(1+\alpha_T\Delta T)$, $\rho_0$ at 0 °CSection 5
Intrinsic semiconductor$\sigma = n\lvert q_e\rvert(\mu_e+\mu_H)$, $\ \sigma = \sigma_0 e^{-E_g/2kT}$Section 10
Extrinsic n-type, p-type$\sigma \approx n_e\lvert q_e\rvert\mu_e$, $\ \sigma \cong p\lvert q_e\rvert\mu_H$Section 12
Extrinsic range$\sigma = \sigma_0 e^{-(E_c-E_d)/2kT}$ or $e^{-(E_a-E_v)/2kT}$Section 14
Dipole moment$p = q\,d$Section 16
Capacitance with dielectric$C = \varepsilon_r\varepsilon_0 A/d = \varepsilon_r C_0$Section 16
Dielectric strength$E_{break} = U_{break}/d$Section 16

Videos linked in the slides

These are references printed on the slides, not content of the portal: