Principles of Solar Cells, LEDs and Related Devices E-Book

Adrian Kitai

• Herausgeber: John Wiley & Sons
• Kategorie: Fachliteratur
• Sprache: Englisch

Seitenzahl: 712

Veröffentlichungsjahr: 2018

Beschreibung

The second edition of the text that offers an introduction to the principles of solar cells and LEDs, revised and updated 

The revised and updated second edition of Principles of Solar Cells, LEDs and Related Devices offers an introduction to the physical concepts required for a comprehensive understanding of p-n junction devices, light emitting diodes and solar cells. The author – a noted expert in the field – presents information on the semiconductor and junction device fundamentals and extends it to the practical implementation of semiconductors in both photovoltaic and LED devices. In addition, the text offers information on the treatment of a range of important semiconductor materials and device structures including OLED devices and organic solar cells.  

This second edition contains a new chapter on the quantum mechanical description of the electron that will make the book accessible to students in any engineering discipline. The text also includes a new chapter on bipolar junction and junction field effect transistors as well as expanded chapters on solar cells and LEDs that include more detailed information on high efficiency devices. This important text:

• Offers an introduction to solar cells and LEDs, the two most important applications of semiconductor diodes
• Provides a solid theoretical basis for p-n junction devices
• Contains updated information and new chapters including better coverage of LED out-coupling design and performance and improvements in OLED efficiency
• Presents student problems at the end of each chapter and worked example problems throughout the text

Written for students in electrical engineering, physics and materials science and researchers in the electronics industry, Principles of Solar Cells, LEDs and Related Devices is the updated second edition that offers a guide to the physical concepts of p-n junction devices, light emitting diodes and solar cells.

Leseprobe

Table of Contents

Cover

Dedication

Introduction

Acknowledgements

Chapter 1: Introduction to Quantum Mechanics

1.1 Introduction

1.2 The Classical Electron

1.3 Two Slit Electron Experiment

1.4 The Photoelectric Effect

1.5 Wave Packets and Uncertainty

1.6 The Wavefunction

1.7 The Schrödinger Equation

1.8 The Electron in a One‐Dimensional Well

1.9 Electron Transmission and Reflection at Potential Energy Step

1.10 Expectation Values

1.11 Spin

1.12 The Pauli Exclusion Principle

1.13 Summary

Further Reading

Problems

Chapter 2: Semiconductor Physics

2.1 Introduction

2.2 The Band Theory of Solids

2.3 Bloch Functions

2.4 The Kronig–Penney Model

2.5 The Bragg Model

2.6 Effective Mass

2.7 Number of States in a Band

2.8 Band Filling

2.9 Fermi Energy and Holes

2.10 Carrier Concentration

2.11 Semiconductor Materials

2.12 Semiconductor Band Diagrams

2.13 Direct Gap and Indirect Gap Semiconductors

2.14 Extrinsic Semiconductors

2.15 Carrier Transport in Semiconductors

2.16 Equilibrium and Non‐Equilibrium Dynamics

2.17 Carrier Diffusion and the Einstein Relation

2.18 Quasi‐Fermi Energies

2.19 The Diffusion Equation

2.20 Traps and Carrier Lifetimes

2.21 Alloy Semiconductors

2.22 Summary

References

Further Reading

Problems

Chapter 3: The p–n Junction Diode

3.1 Introduction

3.2 Diode Current

3.3 Contact Potential

3.4 The Depletion Approximation

3.5 The Diode Equation

3.6 Reverse Breakdown and the Zener Diode

3.7 Tunnel Diodes

3.8 Generation/Recombination Currents

3.9 Metal–Semiconductor Junctions

3.10 Heterojunctions

3.11 Alternating Current (AC) and Transient Behaviour

3.12 Summary

Further Reading

Problems

Chapter 4: Photon Emission and Absorption

4.1 Introduction to Luminescence and Absorption

4.2 Physics of Light Emission

4.3 Simple Harmonic Radiator

4.4 Quantum Description

4.5 The Exciton

4.6 Two‐Electron Atoms

4.7 Molecular Excitons

4.8 Band‐to‐Band Transitions

4.9 Photometric Units

4.10 Summary

References

Further Reading

Problems

Chapter 5: p–n Junction Solar Cells

5.1 Introduction

5.2 Light Absorption

5.3 Solar Radiation

5.4 Solar Cell Design and Analysis

5.5 Thin Solar Cells,

G

 = 0

5.6 Thin Solar Cells,

G

 > 0

5.7 Solar Cell Generation as a Function of Depth

5.8 Surface Recombination Reduction

5.9 Solar Cell Efficiency

5.10 Silicon Solar Cell Technology: Wafer Preparation

5.11 Silicon Solar Cell Technology: Solar Cell Finishing

5.12 Silicon Solar Cell Technology: Advanced Production Methods

5.13 Thin‐Film Solar Cells: Amorphous Silicon

5.14 Telluride/Selenide/Sulphide Thin‐Film Solar Cells

5.15 High‐efficiency Multi‐junction Solar Cells

5.16 Concentrating Solar Systems

5.17 Summary

References

Further Reading

Problems

Chapter 6: Light‐Emitting Diodes

6.1 Introduction

6.2 LED Operation and Device Structures

6.3 Emission Spectrum

6.4 Non‐radiative Recombination

6.5 Optical Outcoupling

6.6 GaAs LEDs

6.7 GaAs

1−

x

P

x

LEDs

6.8 Double Heterojunction Al

x

Ga

1−

x

As LEDs

6.9 AlGaInP LEDs

6.10 Ga

1−

x

In

x

N LEDs

6.11 LED Structures for Enhanced Outcoupling and High Lumen Output

6.12 Summary

References

Further Reading

Problems

Chapter 7: Organic Semiconductors, OLEDs, and Solar Cells

7.1 Introduction to Organic Electronics

7.2 Conjugated Systems

7.3 Polymer OLEDs

7.4 Small‐Molecule OLEDs

7.5 Anode Materials

7.6 Cathode Materials

7.7 Hole Injection Layer

7.8 Electron Injection Layer

7.9 Hole Transport Layer

7.10 Electron Transport Layer

7.11 Light‐Emitting Material Processes

7.12 Host Materials

7.13 Fluorescent Dopants

7.14 Phosphorescent and Thermally Activated Delayed Fluorescence Dopants

7.15 Organic Solar Cells

7.16 Organic Solar Cell Materials

7.17 Summary

References

Further Reading

Problems

Chapter 8: Junction Transistors

8.1 Introduction

8.2 Bipolar Junction Transistor

8.3 Junction Field‐Effect Transistor

8.4 BJT and JFET Symbols and Applications

8.5 Summary

Further Reading

Problems

Appendix 1: Physical Constants

Appendix 2: Derivation of the Uncertainty Principle

Appendix 3: Derivation of Group Velocity

Appendix 4: The Boltzmann Distribution Function

Appendix 5: Properties of Semiconductor Materials

Index

End User License Agreement

List of Tables

Chapter 04

Table A4.1 A list of all possible divisions of total system energy among the atoms in the system

Chapter 01

Table 1.1 The allowed quantum states for the one‐dimensional energy well of Example 1.9

Chapter 04

Table 4.1 Luminescence types, applications, and typical efficiencies

Table 4.2 Possible spin states for a two electron system

Table 4.3 Luminous efficiency values (in lm 

) for a variety of light emitters

List of Illustrations

Chapter 04

Figure A4.1 Plot of the probability of occurrence as a function of the energy for any atom in the system

Figure A4.2 Plot of the probability of occurrence as a function of energy together with an exponential function as an approximation of the data. The average energy at 0.75

is also shown

Chapter 01

Figure 1.1 Electron beam emitted by an electron source is incident on narrow slits with a screen situated behind the slits

Figure 1.2 Classically expected result of two‐slit experiment

Figure 1.3 Result of two‐slit experiment. Notice that a wave‐like electron is required to cause this pattern. If light waves rather than electrons were used, then a similar plot would result except the vertical axis would be a measure of the light intensity instead

Figure 1.4 Davisson–Germer experiment showing electrons reflected off adjacent crystalline planes. Path length difference is

Figure 1.5 The photoelectric experiment. The current

flowing through the external circuit is the same as the vacuum current

Figure 1.6 Wave packet. The envelope of the wave packet is also shown. More insight about the meaning of the wave amplitude for a particle such as an electron will become apparent in Section 1.6 in which the concept of a probability amplitude is introduced

Figure 1.7 Potential well with zero potential for

and a potential energy of

Φ

for

or

. In a hypothetical solid we can consider the wall height of the well to be equal to the workfunction of the solid

Figure 1.8 (a) A beam of electrons with kinetic energy

moving from left to right is incident upon a potential step of height

at

. A transmitted electron flux exists if

. (b) If

the condition

is valid for all values of

Figure 1.9 The Stern–Gerlach experiment in which silver atoms are vaporised in a furnace and sent through a converging magnetic field. The north and south poles of the magnet have distinct shapes that give the field lines a higher density near the north pole

Figure 1.10 A magnetic dipole formed by an orbiting electron passing through a converging magnetic field. The magnetic field causes a Lorentz force on the electron as shown in the detailed drawing on the right. For each electron position there is a force

that acts normal to the plane of electron orbit and towards the north pole of the magnet

Chapter 02

Figure 2.1 The energy levels of a single atom are shown on the left. Once these atoms form a crystal, their proximity causes energy level splitting. The resulting sets of closely spaced energy levels are known as energy bands shown on the right. The electrons in the crystal exist in a periodic potential energy

that has a period equal to the lattice constant

a

as shown

Figure 2.2 Periodic one‐dimensional potential energy

used in the Kronig–Penney model

Figure 2.3 Graph of right‐hand side of Eq. 2.14 as a function of

for

Figure 2.4 Plot of

versus

showing how

varies within each energy band and the existence of energy bands and energy gaps. The vertical lines at

are Brillouin zone boundaries. The first Brillouin zone extends from

, and the second Brillouin zone includes both

(negative wave numbers) and

(positive wave numbers)

Figure 2.5 Plot of

versus

comparing the result of the Kronig–Penney model to the free electron parabolic result

Figure 2.6 Plot of

versus

in reduced zone scheme taken from regions a, b, c, and d in Figure 2.5

Figure 2.7 The filling of the energy bands in (a) semiconductors, (b) insulators, and (c) metals at temperatures approaching 0 K. Available electron states in the hatched regions are filled with electrons and the energy states at higher energies are empty. In band gaps there are no energy states and therefore, although the filling of available energy states is shown hatched, no electrons can exist

Figure 2.8 Room temperature semiconductor showing the partial filling of the conduction band and partial emptying of the valence band. Valence band holes are formed due to electrons being promoted across the energy gap. The Fermi energy lies between the bands. Solid lines represent energy states that have a significant chance of being filled

Figure 2.9 Silicon atoms have four covalent bonds as shown. Although silicon bonds are tetrahedral, they are illustrated in two dimensions for simplicity. Each bond requires two electrons, and an electron may be excited across the energy gap to result in both a hole in the valence band and an electron in the conduction band that are free to move independently of each other

Figure 2.10 Plot of the Fermi–Dirac distribution function

, which gives the probability of occupancy by an electron of an energy state having energy

. The plot is shown for two temperatures

as well as for 0 K. At absolute zero, the function becomes a step function

Figure 2.11 A semiconductor band diagram is plotted along with the Fermi–Dirac distribution function. This shows the probability of occupancy of electron states in the conduction band as well as the valence band. Hole energies increase in the negative direction along the energy axis. The hole having the lowest possible energy occurs at the top of the valence band. This occurs because by convention, the energy axis represents electron energies and not hole energies. The origin of the energy axis is located at

for convenience

Figure 2.12 Reciprocal space lattice. A cell in this space is shown, which is the volume associated with one lattice point. The cell has dimensions

and volume

. The axes all have the same reciprocal length units, but the relative spacing between reciprocal space lattice points along each axis depends on the relative values of

,

, and

Figure 2.13 The positive octant of an sphere in reciprocal space corresponding to an equal energy surface. The number of electron states below this energy is twice the number of reciprocal lattice points inside the positive octant of the sphere

Figure 2.14 A portion of the periodic table containing some selected semiconductors composed of elements in groups II–VI

Figure 2.15 Plot of commonly accepted values of

as a function of

for intrinsic germanium (

), silicon (

eV), and gallium arsenide (

eV)

Figure 2.16 (a) The diamond unit cell of crystal structures of C, Si, and Ge. The cubic unit cell contains eight atoms. Each atom has four nearest neighbours in a tetrahedral arrangement. Within each unit cell, four atoms are arranged at the cube corners and at the face centres in a

face‐centred cubic

(

FCC

) sub‐lattice, and the other four atoms are arranged in another FCC sub‐lattice that is offset by a translation along one quarter of the body diagonal of the unit cell. (b) The zincblende unit cell contains four ‘A’ atoms (black) and four ‘B’ atoms (white). The ‘A’ atoms form an FCC sub‐lattice and the ‘B’ atoms form another FCC sub‐lattice that is offset by a translation along one quarter of the body diagonal of the unit cell. (c) The hexagonal wurtzite unit cell contains six ‘A’ atoms and six ‘B’ atoms. The ‘A’ atoms form a

hexagonal close‐packed

(

HCP

) sub‐lattice and the ‘B’ atoms form another HCP sub‐lattice that is offset by a translation along the vertical axis of the hexagonal unit cell. Each atom is tetrahedrally bonded to four nearest neighbours. A vertical axis in the unit cell is called the

‐axis

Figure 2.17 Band structures of selected semiconductors. (a) silicon, (b) germanium, (c) GaAs, (d) GaP, (e) cubic GaN, (f) CdTe, and (g) wurtzite GaN. Note that GaN is normally wurtzite. Cubic GaN is not an equilibrium phase at atmospheric pressure; however, it can be prepared at high pressure, and it is stable once grown. Note that symbols are used to describe various band features. Γ denotes the point where

. X and L denote the Brillouin zone boundaries in the 〈100〉 and 〈111〉 directions, respectively, in a cubic semiconductor. In (g)

and

denote the

and

directions, respectively, in a hexagonal semiconductor (see Figure 2.16c). Using the horizontal axes to depict two crystal directions saves drawing an additional figure; it is unnecessary to show the complete drawing for each

‐direction since the positive and negative

‐axes for a given

‐direction are symmetrical. There are also some direct energy gaps shown that are larger than the actual energy gap; the actual energy gap is the smallest gap. See Sections 2.13 and 5.2. These band diagrams are the result of both measurements and modelling results. In some cases, the energy gap values differ slightly from the values in Appendix 5.

Figure 2.18 The substitution of a phosphorus atom in silicon (donor atom) results in a weakly bound extra electron occupying new energy level

that is not required to complete the covalent bonds in the crystal. It requires only a small energy

to be excited into the conduction band, resulting in a positively charged donor ion and an extra electron in the conduction band

Figure 2.19 The substitution of an aluminium atom in silicon (acceptor atom) results in an incomplete valence bond for the aluminium atom. An extra electron may be transferred to fill this bond from another valence bond in the crystal. The spatially localised energy level now occupied by this extra electron at

is slightly higher in energy than the valence band. This transfer requires only a small energy

and results in a negatively charged acceptor ion and an extra hole in the valence band

Figure 2.20 The band diagrams for n‐type silicon with a donor doping concentration of 1 × 10

17

 cm

−3

and p‐type silicon with an acceptor doping concentration of 1 × 10

17

 cm

−3

. Note that the Fermi energy lies in the upper part of the energy gap for n‐type doping and lies in the lower part of the energy gap for p‐type doping

Figure 2.21 Carrier concentration as a function of temperature for an n‐type extrinsic semiconductor. In the high‐temperature range, carrier concentration is intrinsic‐like and the Fermi energy is approximately mid‐gap. In the intermediate‐temperature range, carrier concentration is controlled by the impurity concentration, virtually all the dopant atoms are ionised, and the Fermi energy is located above mid‐gap. At low temperatures, there is not enough thermal energy to completely ionise the dopant atoms, and the Fermi energy migrates to a position between

and

as temperature drops. A similar diagram could be prepared for a p‐type extrinsic semiconductor

Figure 2.22 Current (

) flows along a solid semiconductor rod of cross‐sectional area

Figure 2.23 Spatial dependence of energy bands in an intrinsic semiconductor. If there is no electric field, (a) the bands are horizontal and electron and hole energies are independent of location within the semiconductor. If an electric field

ε

is present inside the semiconductor, the bands tilt. For an electric field pointing to the right (b), electrons in the conduction band experience a force to the left, which decreases their potential energy. Holes in the valence band experience a force to the right, which decreases their potential energy. This reversed direction for hole energies is described in Figure 2.13

Figure 2.24 Dependence of drift velocity magnitude on electric field for a semiconductor

Figure 2.25 Plot of excess hole concentration as a function of time. A constant optical generation rate starts at

and continues indefinitely

Figure 2.26 The energy bands will tilt due to a doping gradient. Acceptor concentration increases from left to right in a semiconductor sample. This causes a built‐in electric field, and the hole concentration increases from left to right. The field causes hole drift from left to right, and there is also hole diffusion from right to left due to the concentration gradient

Figure 2.27 The quasi‐Fermi levels

and

for an n‐type semiconductor with excess carriers generated by illumination. Note the large change in

due to illumination and note that

is almost the same as the value of

before illumination

Figure 2.28 A solid semiconductor rod of cross‐sectional area

has a hole current

flowing in the positive

direction. Due to recombination, the hole current is dependent on

.

can be analysed due to the recombination that occurs within volume

Figure 2.29 Plot of excess hole concentration in a semiconductor as a function of

in a semiconductor rod where both diffusion and recombination occur simultaneously. The decay of the concentration is characterised by a diffusion length

Figure 2.30 Hole current density as a function of

for a semiconductor rod with excess carriers generated at

Figure 2.31 A trap level at the Fermi energy near mid‐gap

Figure 2.32 Surface traps at the surface of a p‐type semiconductor comprise electrons held in dangling bonds. The energy needed to release these electrons is approximately

. Since there are large numbers of dangling bond states, some being occupied and some not being occupied by electrons, the Fermi energy becomes pinned at this energy

Figure 2.33 Surface traps at the surface of an n‐type semiconductor causing the Fermi level to be trapped at approximately mid‐gap. An electric field opposite in direction to that of Figure 2.32 is formed in the semiconductor

Figure 2.34 Band gap versus lattice constant for (a) phosphide, arsenide, and antimonide III–V semiconductors; (b) nitride; and other III–V semiconductors neglecting bowing (see Figure 6.20); (c) sulfide, selenide, and telluride II–VI semiconductors and phosphide, arsenide, and antimonide III–V semiconductors.

Chapter 03

Figure 3.1 The p–n junction diode showing metal anode and cathode contacts connected to semiconductor p‐type and n‐type regions, respectively. There are two metal–semiconductor junctions in addition to the p–n semiconductor junction. The diode symbol and two examples of diode applications in circuit design are shown. The diode logic gate was used in early diode‐transistor logic solid‐state computers popular in the 1960s; however, diodes have been replaced by transistor‐based designs that consume much less power and switch faster

Figure 3.2 Band model of p–n junction in equilibrium showing constant Fermi energy and transition region to allow valence band and conduction band to be continuous

Figure 3.3 Flow directions of the four p–n junction currents. The two diffusion currents are driven by concentration gradients of electrons or holes across the junction and the two drift currents are driven by the electric field. Note that the electron currents flow in the direction opposite to the flux or flow of electrons. The electron diffusion flux is to the left and the electron drift flux is to the right

Figure 3.4 A p–n junction diode with external voltage source connected. The external bias voltage will modify the built‐in electric field

Figure 3.5 Diode band model with the application of a forward bias. The energy barrier across the transition region is smaller resulting in net diode current dominated by diffusion currents. In the depletion region,

ε

will be smaller and drift currents no longer fully compensate for diffusion currents. Note that the applied voltage

(in volts) must be multiplied by the electron charge

(in coulombs) to obtain energy (in joules)

Figure 3.6 Diode band model with the application of a reverse bias. Since the applied voltage

is negative, the energy barrier as well as electric field

ε

become larger across the transition region. A small net minority carrier drift current flows

Figure 3.7 Diode current as a function of applied voltage. The reverse drift current saturates at a small value called the reverse saturation current due to thermally generated minority carriers. When

the drift and diffusion currents are equal in magnitude and the net current is zero

Figure 3.8 The equilibrium p–n junction energy barrier height

may be obtained from

resulting in Eq. 3.5

Figure 3.9 Depletion occurs near the junction. In order to establish equilibrium conditions, electrons and holes recombine and the Fermi energy lies close to the middle of the bandgap in the strongly depleted region

Figure 3.10 A depletion region of width

is assumed at the junction. Charge density

ρ

is zero outside of the depletion region. Inside the depletion region a net charge density due to ionised dopants is established. The origin of the

‐axis is placed at the junction for convenience

Figure 3.11 A Gaussian surface having volume

(shaded) encloses the negative charge of magnitude

on the p‐side of the depletion region

Figure 3.12 The electric field directions for the two parts of the depletion region showing that the fields add at the junction

Figure 3.13 The equilibrium electric field

and potential

for the p–n junction follow from the application of Gauss's law to the fixed depletion charge. Note that

on the n‐side is higher compared to the p‐side, whereas in Figure 3.2 the energy levels on the n‐side are lower. This is the case because the energy scale in Figure 3.2 is for electron energy levels; however, the voltage scale in Figure 3.13 is established for a positive charge by convention

Figure 3.14 Coordinates

and

define distances into the p‐type and n‐type semiconductor regions starting from the depletion region edges

Figure 3.15 Quasi‐Fermi levels for a forward‐biased junction.

and

are horizontal because for low‐level injection the majority carrier concentrations are approximately fixed; however, minority carrier concentrations increase towards the depletion region due to carrier injection and therefore

and

are tilted. The separation between quasi‐Fermi levels at the depletion region edge is equal to

and we can write

Figure 3.16 Excess carrier concentration on either side of the p–n junction depletion region. For forward bias the excess concentration is positive and for reverse bias it is small and negative

Figure 3.17 (a) Minority currents

and

as well as majority currents

and

in a forward‐biased p–n junction. The sum of the majority and minority currents is always the total current

. Each majority current splits into two parts, one part supplying carriers to recombine with minority carriers and the other part being injected across the junction to supply the other side with minority carriers. Here we assume that each of majority hole and majority electron currents flowing towards the junction contribute equally to injection and recombination. (b) In the more general case, injection and recombination current contributions are not equal and an example of current flow for an

n

+

p

junction is shown

Figure 3.18 Measured current–voltage characteristics of a diode as well as predicted characteristics based on the diode equation. A very steep increase in current as applied voltage approaches the built‐in potential

is observed in practice as well as an abrupt onset of reverse breakdown current at

Figure 3.19 Increase in depletion region width and increase in junction field with the application of a reverse bias for the p–n junction of Figure 3.13. The equilibrium conditions with depletion width

and peak electric field

are shown with dotted lines. With the application of reverse bias

(

negative) the depletion width increases to

and the peak electric field increases to

Figure 3.20 Tunnelling of valence‐band electron from valence band on p‐side to conduction band on n‐side upon application of a small reverse‐bias voltage. Note that there is a large supply of valence‐band electrons on the p‐side. In comparison, there is only a small supply of thermally generated minority carrier electrons that result in current

. This explains how the reverse current can be much larger than

as shown in Figure 2.18, when

exceeds the breakdown voltage

Figure 3.21 In a tunnel diode, the depletion width is very narrow due to the use of degenerate p

+

and n

+

doping. In addition to the narrow depletion region, the Fermi level enters the bands on either side of the diode resulting in the alignment of electron energy states in the conduction band on the n‐side with valence electron states on the p‐side. Electron tunnelling occurs in either direction as shown

Figure 3.22 Current–voltage (

–

) characteristic of a tunnel diode. At low voltages, tunnelling currents result in significant current flow in both directions. At higher positive bias voltages, electrons in the conduction band on the n‐side will no longer be aligned with the valence band on the p‐side. This will prevent tunnelling and current flow will therefore decrease. Current flow will eventually rise upon further increase of forward bias since the potential barrier decreases as in a normal p–n junction

Figure 3.23 The quasi‐Fermi levels within the depletion region are shown. Although the depletion region is created by the recombination of charges in equilibrium, once injection takes place in forward bias, excess carriers must flow through this region

Figure 3.24 Metal–semiconductor contact for an n‐type semiconductor without any flow of charge between the two sides. The predicted barrier height

and the flat bands shown are not achieved in real devices due to charge flow and charges at the metal–semiconductor interface that cause band‐bending and associated electric fields at the interface and in the semiconductor

Figure 3.25 Metal–semiconductor contact energy band diagrams under various conditions. The subscripts ‘

’ refer to the fact that these are ‘built‐in’ and are present without the application of an external voltage. (a) If the interface is positively charged then the band‐bending will be as shown. This forms an ohmic contact provided that the semiconductor doping level is high enough to make the energy barrier

small. (b) If the interface is negatively charged then band‐bending will result in a large energy barrier

, which blocks electron flow from the metal to the semiconductor, as well as a depletion region in the semiconductor. A Schottky diode is formed

Figure 3.26 (a) Vacuum ‘box’ containing electrons completely surrounded by metal. Electrons in the vacuum are in thermal equilibrium with electrons in the metal. The vacuum box has dimensions (

a

,

b

,

c

). (b) Energy diagram showing the choice of origin for the energy axis

Figure 3.27 (a) Reciprocal space lattice showing positive standing wave solutions to Schrodinger's equation for electrons in an infinite‐walled box. (b) Momentum space lattice showing both positive and negative momentum values. Each standing wave solution is comprised of pairs of travelling waves proceeding in opposite directions along each axis

Figure 3.28 If the interface is negatively charged and the semiconductor is strongly doped to form an

region near the interface, then band‐bending can result in a very narrow energy barrier of height

, which permits electron flow by tunnelling through the barrier and an ohmic contact is formed, as well as an additional built‐in barrier

formed due to band‐bending in the semiconductor

Figure 3.29 Example of an ohmic contact between p‐type silicon and a metal. Electrons from the metal recombine with a high concentration of holes that accumulate near the surface of the p‐type semiconductor. Each hole that recombines allows another hole to take its place, resulting in continuous current flow

Figure 3.30 Example of heterojunction formed between p‐type GaAs and n‐type Ga

1−x

Al

x

As

Chapter 04

Figure 4.1 Lines of electric field

due to a point charge

Figure 4.2 Closed lines of magnetic field

due to a point charge

moving into the page with uniform velocity

Figure 4.3 Lines of electric field emanating from an accelerating charge

Figure 4.4 Direction of magnetic field

emanating from an accelerating charge.

is perpendicular to both acceleration and the radial direction

Figure 4.5 A time‐dependent plot of coefficients

and

is consistent with the time evolution of wave functions

and

. At

,

, and

. Next, a superposition state is formed during the transition such that

. Finally, after the transition is complete,

and

Figure 4.6 The exciton forms a series of closely spaced hydrogen‐like energy levels that extend inside the energy gap of a semiconductor. If an electron falls into the lowest energy state of the exciton corresponding to

, then the remaining energy available for a photon is

Figure 4.7 Low‐temperature transmission as a function of photon energy for Cu

2

O. The absorption of photons is caused through excitons, which are excited into higher energy levels as the absorption process takes place. Cu

2

O is a semiconductor with a band gap of 2.17 eV.

Figure 4.8 A depiction of the symmetric and antisymmetric wave functions and spatial density functions of a two‐electron system. (a) Singlet state with electrons closer to each other on average. (b) Triplet state with electrons further apart on average

Figure 4.9 Energy levels associated with the ground singlet state and both excited singlet and triplet states of a two electron atom. The excited singlet state has the higher energy since the electrons are closer together on average which is less stable than the excited triplet state with electrons further apart on average

Figure 4.10 (a) Parabolic conduction and valence bands in a direct‐gap semiconductor showing two possible transitions. (b) Two ranges of energies

in the valence band and

in the conduction band determine the photon emission rate in a small energy range about a specific transition energy. Note that the two broken vertical lines in (b) show that the range of transition energies at

is the sum of

and

Figure 4.11 Photon emission rate as a function of energy for a direct‐gap transition of an LED. Note that at low energies, the emission drops off due to the decrease in the density of states term

and at high energies, the emission drops off due to the Boltzmann term

Figure 4.12 Absorption edge for direct‐gap semiconductor

Figure 4.13 The eye sensitivity function. The left scale is referenced to the peak of the human eye response at 555 nm. The right scale is in units of luminous efficacy. International Commission on Illumination (Commission Internationale de l'Eclairage, or CIE), 1931

Figure 4.14 Colour space chromaticity diagram showing colours perceptible to the human eye. Numbers on the boundaries are wavelengths in nanometres. Colour saturation or purity is maximum at the boundaries, and it decreases towards the centre of the diagram, eventually becoming white. International Commission on Illumination (Commission Internationale de l'Eclairage, or CIE), 1931

Chapter 05

Figure 5.1 Band diagram of a solar cell showing the directions of carrier flow. Generated electron–hole pairs drift across the depletion region. e, electron; h, hole

Figure 5.2 The

characteristic of a solar cell or photodiode without and with illumination. The increase in reverse current occurs due to optically generated electron–hole pairs that are swept across the depletion region to become majority carriers on either side of the diode

Figure 5.3 The absorption of a photon in a direct‐gap semiconductor proceeds in an almost vertical line since the photon momentum is very small on the scale of the band diagram. There are many possible vertical lines that may represent electron–hole generation by photon absorption as shown. These may exceed the bandgap energy

Figure 5.4 Indirect‐gap semiconductor showing that absorption near the energy gap is only possible if a process involving phonon momentum is available to permit momentum conservation. The indirect, two‐step absorption process involves a phonon to supply the momentum shift that is necessary to absorb photons. For higher energy photons, one of many possible direct absorption processes is shown.

Figure 5.5 Absorption coefficients covering the solar spectral range for a range of semiconductors. Note the absorption tails in silicon and germanium arising from two‐step absorption processes. Amorphous silicon is a non‐crystalline thin film that has different electron states and hence different absorption coefficients compared with single‐crystal silicon.

Figure 5.6 Solar radiation spectrum for a 5250 °C blackbody, which approximates the space spectrum of the sun, as well as a spectrum at the earth's surface that survives the absorption of molecules such as H

2

O and CO

2

in the earth's atmosphere. Note also the substantial ozone (O

3

) absorption in the UV part of the spectrum.

Figure 5.7 Cross‐section of a silicon solar cell showing the front contact metal grid that forms an ohmic contact to the n

+

‐layer. The depletion region at the junction has width

Figure 5.8 Concentrations are plotted on a log scale to allow details of the minority carrier concentrations as well as the majority carrier concentrations to be shown on the same plot. Note that the n

+

‐side has higher majority carrier concentration and lower equilibrium minority carrier concentration than the more lightly doped p‐side corresponding to Figure 5.7. Diffusion lengths are assumed to be much smaller than device dimensions. The p–n junction is shown in a short‐circuit condition with

Figure 5.9 Operating point of a solar cell. The fourth quadrant in Figure 5.2 is redrawn as a first quadrant for convenience. Open‐circuit voltage

and short‐circuit current

as well as current

and voltage

for maximum power are shown. Maximum power is obtained when the area of the shaded rectangle is maximised

Figure 5.10 Excess minority carrier concentrations for a solar cell having dimensions

and

that are small compared to the carrier diffusion lengths. Very high values of surface recombination velocity are present. Note that a forward bias is assumed, and there is no illumination. The straight dotted lines correspond to the first two terms in the square brackets of Eqs. 5.18a and 5.18b and the small parabolas that correspond to the second two terms in the square brackets are superimposed as solid lines

Figure 5.11 Excess minority carrier concentrations for a solar cell having dimensions

and

that are small compared to the carrier diffusion lengths. Zero surface recombination velocity is assumed, which means that there is little drop‐off of carrier concentrations towards the front and back surfaces. The graph shows the case for the approximation of Eq. 5.19a

Figure 5.12 Plot of

versus

using Eq. 5.22a with generation rate

. The boundary conditions are for short‐circuit conditions and a very high surface recombination velocity at

. The curve is a negative parabola

Figure 5.13 Generation rate as a function of depth showing zero generation except at a specific depth

Figure 5.14 Generation rate, excess carrier concentration, and magnitude of the diffusion current density as a function of position in the p‐side of the solar cell. Note that the diffusion current is positive for

and negative for

Figure 5.15 EHP generation rate is

at depth

and

and depth

Figure 5.16 Back surface field formed by a p

+

‐doped region near the back of the solar cell. A potential energy step that generates a built‐in electric field decreases the likelihood of electrons reaching the back surface of the silicon

Figure 5.17 Passivated emitter and rear contact (PERC) solar cell. Oxide/nitride layers may include aluminium oxide and silicon nitride layers in addition to a thin silicon oxide layer that forms at the silicon surface. The p‐type material is often called the emitter because in normal operation it emits minority electrons towards the depletion layer

Figure 5.18 Efficiency limit of solar cells based on a number of well‐known semiconductors. Note the increase in efficiency potentially available if the sunlight intensity is increased to 1000 times the normal sun intensity.

Figure 5.19 Best research solar cell efficiencies achieved in a given year.

Figure 5.20 Czochralski growth system.

Figure 5.21 A 10 × 10 cm

2

multicrystalline silicon wafer. Note the large single‐crystal grains that can be up to approximately 1 cm in dimension.

Figure 5.22 The possible paths of light beams reaching the solar cell surface are shown. At least two attempts to enter the silicon are achieved for each light beam. The angle of the sides of the pyramids may be calculated from the known crystal planes of silicon

Figure 5.23 Micrograph of pyramids selectively etched in the surface of a (100) silicon wafer. Sides of the pyramids are {111} planes, which form automatically in, for example, a dilute NaOH solution. Note the scale showing that the pyramids are approximately 10 μm in height although there is a range of pyramid heights. The pyramids are randomly placed on the surface; however, they are all orientated in the same direction due to the use of single‐crystal silicon.

Figure 5.24 Pattern of conductors used for the front contacts of a silicon solar cell showing bus bars as well as the narrow conducting fingers. The electrode was screen‐printed.

Figure 5.25 String ribbon growth of a silicon ribbon. Surface tension in liquid silicon forms a silicon sheet upon cooling. Thickness of the ribbon is controlled by pull rate and rate of cooling of the silicon web.

Figure 5.26 Atomic structure of a‐Si:H in which H atoms terminate dangling Si bonds, which are generally isolated but may also be clustered (two dangling bonds in one silicon atom).

Figure 5.27 Absorption edge of amorphous silicon compared to crystalline silicon. There is about a half electronvolt energy difference in absorption edge between the two materials.

Figure 5.28 Density of electron states in a‐Si:H. Note the bandtails as well as the mid‐gap states due to defects.

Figure 5.29 Structures of a‐Si:H solar cells (not to scale). (a) Glass substrate structure with illumination through the substrate. (b) Stainless‐steel substrate structure with direct illumination. TCO, transparent conductor layer

Figure 5.30 Tandem solar cell structure on glass substrate. The n–p junction formed at the interface between the first and second p–i–n diodes must be an effective tunnelling junction to allow carriers to flow to the next diode. The bandgap of the first p–i–n diode is higher than the bandgap of the second p–i–n diode. Similar tandem and triple‐junction structures may be formed on stainless‐steel substrates. TCO is a transparent conductive oxide layer

Figure 5.31 Band diagram of CdTe/CdS heterojunction. Note that the CdS is n‐type while the CdTe is p‐type. Since minority carriers are virtually all collected from the CdTe side as conduction band electrons, the large valence band offset does not play a significant role.

Figure 5.32 Grain structure of CdTe solar cell also showing the CdS layer and device structure. TCO, transparent conductor oxide layer. The term ‘superstrate’ is used since the glass faces the incoming sunlight.

Figure 5.33 Band diagram of basic CIGS cell with a CdS n‐type layer. Alternative materials for the CdS layer are used in CIGS solar cells currently in production yielding Cd‐free CIGS thin‐film solar cells.

Figure 5.34 Scanning electron microscope cross‐section of a CIGS solar cell. Note that unlike CdTe thin‐film cells, the glass substrate is at the rear of the cell and light enters through the ZnO transparent front electrode.

Figure 5.35 Structure of triple‐junction solar cell showing germanium wafer upon which two larger bandgap semiconductors are deposited epitaxially. The total output voltage is obtained between the metal grid and the rear electrode

Figure 5.36 Multi‐junction solar cell showing the portions of the spectrum that are covered by each material. Note that the infrared wavelengths above 2000 nm are not absorbed. The GaAs junction is modified and further optimised by the addition of a small indium component. Note that the three cell voltages add together; however, the cell current is controlled by the minimum current that is supplied by each of the cells. This means that current matching between the multiple cells in a multi‐junction cell is important.

Figure 5.37 Tunnelling between the InGaP and GaAs cells in a tandem solar cell shown in a band diagram.

Figure 5.38 Auger recombination process. Two electrons are involved. A first electron recombines with a hole but instead of producing a photon, energy is conserved by a second electron that is excited high up into the conduction band. This second electron then thermalises rapidly and heat is produced. Indirect‐gap transitions can also occur since the second electron can undergo a momentum change that allows for overall momentum conservation

Chapter 06

Figure 6.1 Efficiency of LEDs achieved as a function of time since the 1960s. The photometric quantity lm W

−1

(lumens per watt) was introduced in Chapter 4

Figure 6.2 Forward‐biased LED p–n junction. When one electron and one hole recombine near the junction one photon of light may be emitted. The achievement of high efficiency requires that there is a good chance that recombination events are radiative and that the generated photons are not reabsorbed or trapped in the device

Figure 6.3 LED packaging includes a transparent lens, which is usually made from an epoxy material, and a reflector cup into which the LED die is mounted. The radiation pattern is determined by the combination of the die emission pattern, the reflector cup design, and the shape and refractive index of the polymer lens.

Figure 6.4 LED die consists of a single‐crystal substrate on which a series of epitaxial layers is grown forming the active layers. (a) For conductive substrates a current spreading layer and the top and bottom contacts are shown as well as an n‐type current blocking layer discussed in Section 6.4. The notations DH and DBR refer to Double Heterostructure and Distributed Bragg Reflector, which will be discussed in Sections 6.8 and 6.11, respectively. (b) For LEDs grown on insulating substrates such as sapphire (see Section 6.10) the second contact is made via a buried current‐spreading layer as shown

Figure 6.5 Emission spectra of AlGaInP LEDs. The linewidth of the amber LED is measured as the full width at half maximum as shown and is 13.5 nm.

Figure 6.6 Observed luminescence at the junction of a GaInAs/GaAs LED near a cleaved surface showing that carrier diffusion lengths are in the range of a few microns in GaInAs. Non‐radiative recombination occurs for carriers near the cleaved surfaces

Figure 6.7 Light generated in the semiconductor will reach the surface and either reflect or be able to exit depending on the angle of incidence. The critical angle is

Figure 6.8 Light that escapes will be diffracted from an incident angle θ to an emitted angle Θ

Figure 6.9 Typical radiation pattern of intensity versus emission angle for an LED without an integral lens. This AlGaInP red LED has a radiation pattern with a 115° beam divergence determined at half maximum intensity. Source: Reproduced by permission of Osram. Data sheet is for GR PSLR31.13. Note the two ways of graphing the LED relative output intensity versus angle. The emission of the LED is symmetrical over positive and negative angles. Note the approximately lambertian emission profile in which

Figure 6.10 Photon energy plotted as a function of phosphorus mole fraction

in GaAs

1−

x

P

x

and GaAs

1−

x

P

x

:N.

Figure 6.11 Band gaps of double heterojunction using Al

x

Ga

1−

x

As layers grown epitaxially on a GaAs substrate. Carriers recombine in the active layer of width

. The cladding layers are doped such that one layer is n‐type and one layer is p‐type. Similar structures are used in GaInN LEDs. See Section 6.10

Figure 6.12 Electron and hole energy levels within the double heterojunction wells of height

in the conduction band and

in the valence band

Figure 6.13 Double heterojunction showing the exponential decay of excess carriers into the cladding layer. Only one side is shown for simplicity

Figure 6.14 (Al

x

Ga

1−

x

)

y

In

1−

y

P bandgap versus lattice constant graph showing the composition ranges in this quaternary system. By adjusting the two available parameters,

and

, a field of compositions is possible represented by the enclosed area. A range of energy gaps from 1.89 to 2.33 eV is available in the direct bandgap region while matching the GaAs lattice constant.

Figure 6.15 Radiative efficiency as a function of dislocation (etch pit) density for a variety of III–V semiconductors. Dislocation density is determined by etching the crystal surface and then counting the number of resulting etch pits per unit area. Etch pits form at the dislocations

Figure 6.16 Dislocations in GaN epitaxial layer grown on sapphire. In addition to these factors, the 12% lattice mismatch of GaN with respect to sapphire is effectively much less apparent since it turns out that a rotation about the

‐axis of GaN relative to the sapphire substrate allows a far better lattice match of the GaN system relative to the sapphire in the plane normal to the

‐axis. See Problem 6.18. Both sapphire and SiC are very stable substrate materials that may be heated to over 1000 °C during GaN growth, and both substrates are used in the high‐volume production of GaInN LEDs.

Figure 6.17 Emission spectra of blue, green, and red LEDs having the highest available efficiencies.

Figure 6.18 Output intensity versus ambient temperature for GaInN and AlGaInP LEDs. Note the decreased thermal quenching in GaInN.

Figure 6.19 Forward intensity versus current characteristics for GaInN and AlGaInP LEDs.

Figure 6.20 The nitride alloy semiconductor systems Al

x

Ga

1−

x

N, Ga

1−

x

In

x

N and Al

x

In

1−

x

N plotted to show energy gap as a function of lattice constant.

Figure 6.21 (a) Band diagram of a double heterostructure using In

x

Ga

1−

x

N active layer grown epitaxially. Carriers recombine in the active layer. The cladding layers are doped such that one layer is n‐type and one layer is p‐type. (b) Band diagram including the effect of polarisation. (c) Band diagram of multiple quantum well GaInN LED. Between 4 and 8 quantum wells are commonly employed for high‐efficiency nitride‐based LEDs

Figure 6.22 GaN planes are generally (0001) planes that are formed as a result of

‐axis growth; however, alternative growth directions and planes may also be achieved. This reduces or eliminates polarisation in the growth directions. (a)

‐planes

or

‐planes

as shown are non‐polar and (b) semi‐polar planes also exist. Growth of high‐quality GaN in directions resulting in non‐polar and semi‐polar quantum well structures is an area of LED research.

Figure 6.23 Emission spectrum of white‐emitting LED. Blue light from the LED die is down‐converted using YAG:Ce to produce broadband yellow fluorescence emission. The yellow fluorescence contains both green and red wavelengths. When combined with the blue LED emission white light results.

Figure 6.24 Three of the six possible escape cones for light emission from an LED die with vertical sidewalls

Figure 6.25 Light can outcouple more efficiently if the sidewalls of the die are tilted as shown. The tilted walls can be applied to a variety of substrates.

Figure 6.26 LED‐chip architectures that lead to light extraction efficiencies of 80–90%. (a) Light propagation through a textured surface. The light path shown exceeds the critical angle of the surface but the beam can pass through due to surface texturing. Up to 50% improvement in outcoupling has been achieved through the use of texturing. (b) Thin‐film‐flip‐chip GaInN LED. Note that this structure is grown on single‐crystal GaN to avoid sapphire–GaN interface optical reflection losses. GaN substrates are available but are not used for high‐volume LED production due to cost. Since the LED chip is mounted such that light emerges from the rear (substrate) side and contacts are made directly to the top of the chip, this is referred to as flip‐chip mounting. (d) Shaped transparent substrate GaInN LED. The triangular chip shape as well as surface texturing mitigate total internal reflection losses.

Figure 6.27 Structure of lighting‐grade white LED showing the LED chip mounted on a ceramic substrate suitable for heat removal due to its high thermal conductivity. The lens is made of a silicone polymer, which withstands higher optical flux without yellowing compared to epoxy lenses. Note the yellow‐emitting phosphor layer which, in combination with blue light from the LED chip, yields white light.

Figure 6.28 Achieved LED external quantum efficiency as a function of peak emission wavelength at room temperature. Note the limiting trends due to GaInN (short wavelength range) and (Al,Ga)InP (short wavelength range). The green gap refers to poor quantum efficiency in the green spectral range where the human eye response (solid curve) peaks.

Chapter 07

Figure 7.1 The molecular structure of polyethylene. Each carbon has four nearest neighbours and forms four bonds. Polyethylene is an insulator and has a wide energy gap in the ultraviolet energy range. Each carbon atom has an almost perfect tetrahedral bond symmetry even though it bonds to both carbon and hydrogen nearest neighbours

Figure 7.2 Polyacetylene is the simplest conjugated molecule. It is often thought of as a chain of single bonds alternating with double bonds although actually all the bonds are equal and are neither purely single nor purely double. It consists of a carbon chain with one hydrogen atom per carbon atom. Since only three of the four valence electrons of carbon are used for bonding, one π electron per carbon atom is available for electrical conduction and becomes delocalised along the carbon chain

Figure 7.3 Energy levels and bands in a two closely spaced organic molecules. Note the small energy barriers caused by the intramolecular bonding and the larger energy barriers caused by the intermolecular bonding

Figure 7.4 Molecular structures of well‐known conjugated polymers. Many molecules contain a combination of linear and ring‐type structures, the simplest example being poly para‐phenylene vinylene (PPV).

Figure 7.5 Poly para‐phenylene vinylene (PPV) derivative forming a silicon‐substituted soluble polymer

Figure 7.6 Absorption and emission of poly para‐phenylene vinylene (PPV) and PPV derivatives. The energy gap determines the upper wavelength range of absorption as well as the lower wavelength range of emission. Here energy gaps from 1.9 eV (≅640 nm) to 2.5 eV (≅500 nm) result in these spectra.

Figure 7.7 Structure of basic polymer OLED consisting of a glass substrate, a transparent ITO anode layer, an EL polymer layer and a low‐work‐function cathode layer

Figure 7.8 Energy diagram showing a high‐work‐function anode and a low‐work‐function cathode. With a constant vacuum energy the Fermi levels cannot be aligned (a) and this diagram is therefore not an equilibrium diagram, which invalidates the concept of Fermi energy. To resolve this problem an electric field forms between anode and cathode and an equilibrium diagram (b) having aligned Fermi energies is the result. This is the result of charge transfer

Figure 7.9 The upper π* band and lower π band in a polymer EL layer. (a) The equilibrium condition. (b) The flat‐band condition in which a positive voltage is applied to the anode. (c) Device in forward bias in which holes and electrons are injected and form molecular excitons, which annihilate to generate photons. The resulting hole injection barrier and electron injection barrier are shown

Figure 7.10 Typical luminance–voltage (L–V) and IV characteristics of a polymer OLED. A well‐defined threshold voltage is observed due to the sharp onset of carrier injection from the electrodes across the potential barriers at the electrode‐EL polymer interfaces. Note the similarity between the shapes of the current and luminance curves

Figure 7.11 Small organic molecules used for small‐molecule OLED devices. Hole‐transporting materials are TPD and NPD. Electron‐transporting materials are PBD and Alq3.

Figure 7.12 Small‐molecule OLED structure. The OLED includes a transparent substrate, transparent ITO anode, hole transport layer (HTL), electron transport layer (ETL), and cathode. HTL materials such as TPD or NPD and electron transport materials such as Alq3 or PBD are suitable. A popular cathode is a two‐layer Al/LiF structure as shown

Figure 7.13 A more optimised small‐molecule OLED structure includes an electron injection layer, a hole injection layer, and a light‐emitting material. The cathode includes the electron injection layer

Figure 7.14 Band diagram of a small‐molecule OLED showing LUMO and HOMO levels for the various layers of the device. The band diagram is drawn without a bias applied. The accepted work functions of anode (ITO) and cathode (LiF/Al,) which are 4.7 and 3.6 eV, respectively, are shown

Figure 7.15 OLED package includes front and back sheets, epoxy seal material on all edges, sacrificial desiccant or getter material, cathode and transparent anode having cathode contact areas for external connections. The rate of moisture penetration must be calculated to ensure a specified product life

Figure 7.16 Copper phthalocyanine, or CuPc, a widely used metal complex used for the HIL in small molecule OLEDs. Another HIL molecule is m‐MTDATA, also shown.

Figure 7.17 Organo‐metallic complexes may also be used for the electron injection layer. Examples are shown consisting of some lithium–quinolate complexes. Liq, LiMeq, Liph, and LiOXD.

Figure 7.18 Two further examples of hole‐conducting triarylamines include TPA (triphenylamine) and TPTE (a tetramer of TPA). TPTE enables high‐temperature OLED operation without crystallisation.

Figure 7.19 Phenylazomethines are formed by various arrangements of nitrogen‐terminated six‐carbon rings. These phenylazomethine molecules are thermally stable and are complexed with metal ions such as Sn ions introduced in the form of SnCl

2

molecules to form the HTL material.

Figure 7.20 TPBI, ATZL, and TPQ are members of imine‐based molecules which are candidate electron transport layer (ETL) materials as well as light‐emitting materials. Other candidate ETL materials include C60. See Section 7.17.

Figure 7.21 Host–guest energy transfer. The energy transfer can occur due to three possible processes, which may be Förster, Dexter or radiative energy transfer.

Figure 7.22 Electron transport hosts Alq3, BAlq, TPBI, and TAZ1.

Figure 7.23 Hole transport hosts CBP and CDBP.

Figure 7.24 Coumarin‐based green fluorescent dopant C‐545 TB and quinacridone‐based dopant DMQA.

Figure 7.25 The red fluorescent molecule DCJPP derived from the arylidene family of molecules and four variations of red fluorescent molecule DCDDC derived from the isophorone family of molecules.

Figure 7.26 Arylene family host DPVBI and dopant BCzVBI. Also shown are anthracene family host JBEM and dopant perylene.

Figure 7.27 Scheme illustrating internal quantum efficiencies that can be achieved under electroluminescence based on conventional fluorescent emitters (a) and organometallic phosphorescent emitters (b). With 75% triplet exciton generation, the internal efficiency of conventional fluorescence in OLEDs is limited to about 25%. On the right side, ISC (intersystem crossing) enables up to 100% quantum yield.

Figure 7.28 Phosphorescent emitters: (a) fac tris(2‐phenylpyridine) iridium (Ir(ppy)3), (b) bis(2‐phenylpyridine) iridium(III) acetylacetonate (Ir(ppy)2(acac)), (c) bis(2‐(3,5‐dimethylphenyl)‐4‐methylpyridine) iridium(III) (2,2,6,6‐tetramethylheptane‐3,5‐diketonate) (Ir(3′,5′,4‐mppy)2(tmd)), (d) tris(1‐phenylisoquinolinato‐C2,N)iridium(III) (Ir(piq)3), (e) bis[(4,6‐difluorophenyl)‐pyridinato‐N,C2′] iridium(III) picolinate (FIrpic), (f) (2,3,7,8,12,13,17,18‐octaethyl‐21H,23H‐porphine) platinum(II) (PtOEP), (g) bis[3‐trifluoromethyl‐5‐(2‐pyridyl)‐1,2‐pyrazolato] platinum(II) (Pt‐A), (h) [6‐(1,3‐dihydro‐3‐methyl‐2H‐imidazol‐2‐ylidene‐κC2)‐4‐tert‐butyl‐1,2‐phenylene‐κC1]oxy[9‐(4‐tert‐butyltpyridin‐2‐yl‐κN)‐9H‐carbazole‐1,2‐diyl‐κC1] platinum(II) (PtON7‐dtb).

Figure 7.29 (a) Thermally activated delayed fluorescence (TADF) emitter molecule. S1 and T1 denote the first excited singlet and triplet state, respectively. ki represents various rates. ISC = intersystem crossing, RISC = reverse ISC, r = radiative, nr = non‐radiative, F = fluorescence, P = phosphorescence.

indicates the singlet‐triplet splitting. (b) Qualitative representation of how RISC and the fluorescence rate depend on the singlet‐triplet splitting

.

Figure 7.30 Single‐layer organic solar cell consisting of a single organic semiconductor layer, a low work function cathode, and a transparent anode. Device efficiency is well below 1%

Figure 7.31 Energy level diagram for single‐layer organic solar cell. The absorption of light creates excitons through the promotion of molecular electrons from the HOMO level to the LUMO level. Electrode work functions are set to match the HOMO and LUMO levels to facilitate the collection of the electrons and holes as shown

Figure 7.32 Organic planar heterojunction solar cell structure showing donor and acceptor organic layers

Figure 7.33 Heterojunction solar cell showing donor and acceptor LUMO and HOMO levels. Excitons are generated throughout the donor layer, and these excitons are dissociated when they diffuse to the donor–acceptor interface. Finally, the separated holes and electrons can drift to their respective electrodes

Figure 7.34 Bulk heterojunction organic solar cell. A number of small (∼10 nm) donor regions are organised within the bulk heterojunction layer and optimised to absorb sunlight and allow exciton diffusion to a nearby junction

Figure 7.35 (a) Bulk heterojunction structure showing a typical random structure of donor and acceptor materials. The dimension of one region within the heterojunction is about 10 nm. The problem is the connectivity of these regions to their appropriate contact materials. (b) Bulk heterojunction of vertically oriented stripes of donor and acceptor materials that enables the donor material to be in contact with the ITO electrode and the acceptor layer to be in contact with the aluminium electrode. The acceptor layer could be made using vertically oriented carbon nanotubes

Figure 7.36 Molecular structures of (a) poly(3‐hexylthiophene) (or P3HT) and (b) PQT‐12.

Figure 7.37 Absorption spectra of P3HT and PQT‐12. Source: permission from Vemulamada et al. (2008). Copyright (2008). Adapted from Elsevier

Figure 7.38 Molecular structures of the fullerene C60 and its derivative [6,6]‐phenyl‐C61‐butyric acid methyl ester (PCBM).

Figure 7.39 Carbon nanotube.