10. Optoelectronic and Photonic Semiconductor Devices

Introduction and context

Optoelectronics concerns the conversion and control of electrical and optical signals using semiconductor materials and structures. It forms the technological basis of fiber-optic communication, optical sensing, displays, imaging, industrial measurement, biomedical instrumentation, photovoltaics, and integrated photonic circuits. Whereas conventional semiconductor devices manipulate charge and electric fields, optoelectronic devices also exploit photons—the quantum particles associated with electromagnetic radiation.

The performance of an optoelectronic device depends on both semiconductor physics and optical engineering. Carrier generation, recombination, band structure, doping, junction fields, optical confinement, reflection, absorption, and thermal effects must be considered together. This makes optoelectronics especially important at postgraduate level: device selection is rarely based on one parameter alone. Wavelength, speed, sensitivity, noise, power consumption, coupling efficiency, reliability, fabrication complexity, and cost must be evaluated as a system.

This chapter develops the physical principles of optical absorption and emission and applies them to light-emitting diodes, laser diodes, photodiodes, avalanche photodetectors, solar cells, image sensors, and devices used in fiber-optic and integrated-photonic systems.

Learning objectives

After studying this chapter, the reader should be able to:

1. Explain photon absorption, spontaneous emission, stimulated emission, and radiative recombination in semiconductors.

2. Relate band structure and material properties to optical-device wavelength and efficiency.

3. Compare LEDs, laser diodes, PIN photodiodes, avalanche photodiodes, solar cells, and image sensors.

4. Analyze important figures of merit such as responsivity, quantum efficiency, dark current, gain, bandwidth, threshold current, and noise.

5. Select suitable optoelectronic devices for communication, sensing, imaging, and energy-conversion applications.

6. Identify packaging, coupling, thermal, reliability, safety, and system-integration issues.

7. Evaluate emerging developments in integrated photonics and intelligent optical systems.

Photons, bands, and semiconductor optical transitions

A photon of frequency f has energy

E_ph = hf = hc/λ,

where h is Planck’s constant, c is the velocity of light in vacuum, and λ is wavelength. In a semiconductor, an electron can absorb a photon and move from the valence band to the conduction band if the photon energy is sufficiently large. The minimum energy required is approximately the bandgap E_g. The corresponding wavelength is

λ_g ≈ 1.24/E_g( eV ) μm.

This relationship explains why material composition determines the operating wavelength of an optoelectronic device. Alloy semiconductors such as AlGaAs, InGaAsP, InP, GaN, and InGaN provide bandgap engineering across visible, near-infrared, and telecommunications bands.

In a direct-bandgap semiconductor, the minimum of the conduction band and the maximum of the valence band occur at nearly the same crystal momentum. Electron–hole recombination can therefore produce a photon efficiently. In an indirect-bandgap material such as silicon, momentum conservation generally requires phonon assistance, so radiative emission is weak. Silicon remains highly valuable for photodetection and integrated electronics, but efficient silicon light sources are more difficult to realize.

Optical absorption is described by an absorption coefficient α(λ), which determines how rapidly optical intensity decreases with distance:

I(z)=I(0)e^(-αz).

Absorption may be interband, involving valence- and conduction-band states; free-carrier absorption; or defect-related. In a photodetector, absorption is desirable within the active region. In a laser cavity, uncontrolled absorption reduces gain and efficiency.

Emission mechanisms and optical gain

When an electron and a hole recombine, energy may be released radiatively as a photon or nonradiatively through phonons, defects, and Auger processes. The internal quantum efficiency is the fraction of recombination events that produce photons. Temperature, crystal quality, carrier density, and confinement strongly affect this efficiency.

Spontaneous emission occurs when a carrier recombines without being triggered by an incoming photon. The emitted photons have random phase and direction, which gives LED light relatively broad spectral width and low coherence. Stimulated emission occurs when an incident photon induces a second photon with the same frequency, phase, polarization, and direction. This process is the basis of semiconductor lasers.

A useful gain condition for a semiconductor laser is

Γg(N) ≥ α_i + α_m,

where Γ is the optical confinement factor, g(N) is material gain, α_i represents internal loss, and α_m represents mirror loss. At threshold, optical gain balances total loss. Increasing the injection current above threshold produces a rapid increase in coherent output power.

The Einstein description of emission identifies three processes: absorption, spontaneous emission, and stimulated emission. Population inversion, or more precisely a carrier distribution producing positive optical gain, is required for net stimulated amplification. In a semiconductor laser this is produced by carrier injection into an active region rather than by the atomic-level pumping used in many gas or solid-state lasers.

Light-emitting diodes

An LED is a forward-biased p–n junction or heterostructure that converts electrical energy into mostly spontaneous optical emission. Modern LEDs commonly use a double-heterostructure or quantum-well active region. Wider-bandgap cladding layers confine carriers and photons within the active region, improving internal efficiency.

The emitted wavelength is primarily determined by the active-region bandgap, although alloy composition, strain, quantum confinement, temperature, and phosphor conversion may modify the spectrum. Visible LEDs may emit directly from materials such as InGaN or AlGaInP. White LEDs are commonly made using a blue or ultraviolet emitter with a phosphor, or by combining multiple colors.

Important LED performance measures include external quantum efficiency, wall-plug efficiency, luminous efficacy for visible devices, optical power, spectral width, modulation bandwidth, and lifetime. External efficiency includes losses caused by nonradiative recombination and imperfect extraction from the semiconductor. Because semiconductor refractive indices are high, total internal reflection can trap photons; surface texturing, shaped encapsulants, reflective contacts, and photonic structures improve extraction.

LEDs are robust, inexpensive, and suitable for illumination, indicators, displays, visible-light communication, optical encoders, and low-to-moderate-speed transmitters. Their principal limitations are broad linewidth, relatively low optical coherence, and lower modulation speed than laser diodes.

Laser diodes: structure, operation, and design trade-offs

A laser diode combines an electrically pumped active region with an optical resonator. Cleaved semiconductor facets, distributed Bragg reflectors, or distributed-feedback gratings provide wavelength-selective feedback. A ridge waveguide or buried heterostructure confines the optical mode laterally and vertically.

Below threshold, the device behaves approximately as an LED with amplified spontaneous emission. At threshold, modal gain equals cavity loss. Above threshold, stimulated emission dominates and the output becomes directional, narrowband, and coherent. The light-current characteristic therefore changes slope near the threshold current.

Laser-diode characteristics include threshold current, slope efficiency, differential quantum efficiency, linewidth, side-mode suppression ratio, wavelength stability, beam divergence, and modulation bandwidth. Distributed-feedback lasers are widely used in wavelength-division multiplexing because their grating selects a stable longitudinal mode. Vertical-cavity surface-emitting lasers emit perpendicular to the wafer and offer low threshold, circular beams, and convenient array fabrication; they are common in short-reach data links and sensing.

Laser diodes are sensitive to temperature. As temperature increases, threshold current generally rises and wavelength shifts. Thermal resistance, heat sinking, automatic power control, and wavelength stabilization are therefore important in practical transmitters. Optical feedback from connectors or reflections can destabilize a laser, so isolators or suitable termination may be needed.

Safety is a central professional consideration. A beam that appears dim can still damage the retina, particularly at near-infrared wavelengths. Laser products must be classified, labelled, enclosed, and operated according to applicable safety standards.

Photodetectors: converting photons into current

A photodetector absorbs photons and converts their energy into an electrical signal. In a reverse-biased p–n or p–i–n junction, photons generate electron–hole pairs. The electric field separates the carriers and produces photocurrent. The intrinsic layer in a PIN device is deliberately widened to increase absorption volume and reduce junction capacitance per unit active area.

The responsivity R is

R = I_ph/P_opt = ηq/(hf)=ηqλ/(hc),

where η is quantum efficiency, q is electronic charge, and P_opt is incident optical power. Responsivity increases with wavelength for constant quantum efficiency, until the photon energy approaches the material bandgap and absorption falls sharply.

A photodiode’s bandwidth is affected by carrier transit time and RC time constant. Increasing the depletion width reduces capacitance but may increase transit time; reducing device area lowers capacitance but can reduce optical collection and alignment tolerance. High-speed design is therefore a system-level compromise.

Dark current flows in the absence of light and contributes shot noise. Thermal generation, surface leakage, tunnelling, and defects can increase dark current. Cooling is useful in very sensitive receivers but adds complexity and power consumption.

Avalanche photodetectors

An avalanche photodiode operates with a strong reverse electric field. A photogenerated carrier gains enough energy to create additional electron–hole pairs through impact ionization. The resulting internal multiplication M increases photocurrent:

I_APD ≈ M I_primary.

Avalanche multiplication improves sensitivity but also increases excess noise, bias complexity, and temperature dependence. The device must be operated below breakdown under controlled bias. Automatic bias control is often required in communication receivers. APDs are useful when receiver sensitivity is more important than simplicity, whereas PIN photodiodes are preferred when low voltage, low noise, and linearity are priorities.

Solar cells and photovoltaic conversion

A solar cell is a large-area p–n junction designed to convert absorbed solar radiation into electrical power. Photons with energy above the bandgap generate electron–hole pairs, and the junction field separates them. Under illumination, the device produces a current–voltage characteristic commonly represented by the diode equation with a photocurrent term.

The short-circuit current I_SC measures current at zero terminal voltage, while open-circuit voltage V_OC is the voltage at zero current. The maximum-power point is defined by V_MP and I_MP. The fill factor is

FF = (V_MP I_MP)/(V_OC I_SC),

and conversion efficiency is

η_PV = (V_OC I_SC FF)/P_in.

Losses arise from reflection, incomplete absorption, recombination, series resistance, shunt leakage, spectral mismatch, and temperature. Increasing bandgap does not automatically improve efficiency: a high bandgap absorbs fewer photons, while a low bandgap loses more excess photon energy as heat. This spectral trade-off motivates multijunction cells, in which subcells with different bandgaps divide the solar spectrum.

Solar-cell engineering includes antireflection coatings, surface texturing, passivation, selective contacts, back reflectors, bypass diodes, and maximum-power-point tracking. Reliability concerns include moisture ingress, ultraviolet degradation, thermal cycling, mechanical stress, and hotspot formation.

Image sensors

Image sensors convert spatially varying optical intensity into an electronic image. Charge-coupled devices transfer accumulated charge through potential wells to a common readout, whereas complementary metal–oxide–semiconductor image sensors generally use pixel-level or column-level amplifiers and readout circuits.

A pixel typically includes a photodiode, charge-storage region, reset mechanism, source follower, and row-selection device. Important parameters include quantum efficiency, fill factor, full-well capacity, conversion gain, read noise, dark current, dynamic range, frame rate, fixed-pattern noise, and crosstalk.

Pixel size presents a design trade-off. Smaller pixels permit higher spatial resolution but collect fewer photons and may reduce full-well capacity. Microlenses improve fill factor by directing light toward the photosensitive region. Backside illumination places wiring behind the photodiode and improves optical collection. Global shutters expose all pixels simultaneously, which is essential for moving scenes; rolling shutters expose rows sequentially and can produce geometric distortion during motion.

Scientific and industrial sensors may use cooling, deep-depletion structures, pinned photodiodes, electron multiplication, or specialized materials for ultraviolet and infrared detection. Image processing—such as correlated double sampling, defect correction, demosaicing, high-dynamic-range combination, and machine-learning classification—can substantially improve system performance, but it cannot recover information lost through saturation or inadequate photon collection.

Devices for fiber-optic communication

A fiber-optic link contains a source, coupling structure, optical fiber, receiver, and often amplifiers, filters, modulators, and monitoring circuits. Sources are selected according to distance, data rate, wavelength, and cost. LEDs are appropriate for some short links, while laser diodes dominate high-speed and long-distance systems.

At the receiver, a PIN or APD detects the modulated optical signal. The receiver must provide adequate sensitivity, bandwidth, linearity, and dynamic range. A transimpedance amplifier converts the small photodiode current into voltage. The receiver noise budget includes shot noise, thermal noise, amplifier noise, dark-current noise, and relative-intensity noise from the source.

Optical coupling efficiency depends on mode-field matching, alignment, facet quality, numerical aperture, and connector losses. Single-mode systems demand precise alignment but support high bandwidth and long distance. Multimode systems tolerate greater alignment error and can be economical for shorter links, although modal dispersion limits distance–bandwidth performance.

Optical modulators provide an alternative to directly modulating the laser current. Electro-absorption modulators change absorption with applied voltage, while Mach–Zehnder modulators use electrically induced phase changes and interference. These devices support high-speed transmitters and can be integrated with lasers and waveguides.

Integrated photonic devices

Integrated photonics places waveguides, resonators, splitters, modulators, filters, detectors, and sometimes light sources on a chip. A waveguide confines light by total internal reflection or photonic-bandgap effects. Directional couplers divide or combine optical power. Ring resonators provide wavelength-selective filtering and compact sensing, although their response is sensitive to temperature and fabrication variation.

Silicon photonics combines low-loss silicon waveguides with mature electronic fabrication. Because silicon is an inefficient light emitter, lasers are often bonded, hybrid-integrated, or coupled from a different material platform. Heterogeneous integration with III–V semiconductors, silicon nitride, lithium niobate, and other materials expands the available functions.

Photonic integrated circuits are used in data-center interconnects, coherent communication, optical switching, microwave photonics, spectroscopy, biosensing, quantum technologies, and neuromorphic computing. Their advantages include compactness, reduced parasitics, repeatability, and the possibility of wafer-scale manufacturing. Design challenges include optical loss, thermal tuning power, packaging, fiber coupling, process variation, and testing of many optical ports.

Applied case study: selecting a receiver for a fiber link

Background

Consider an illustrative 10 Gb/s metropolitan fiber link operating near 1550 nm. The receiver must provide high sensitivity, moderate power consumption, and reliable operation over temperature.

Situation and analysis

A PIN photodiode offers simple biasing, low excess noise, and good linearity, but its sensitivity may be insufficient when the received optical power is low. An APD supplies internal gain and can improve sensitivity, but it requires a stable high-voltage bias and introduces multiplication noise and temperature dependence. A receiver using an APD must therefore include bias monitoring, temperature compensation, and protection against excessive optical power.

The design team estimates the required bandwidth from the bit rate, checks the photodiode transit-time and capacitance limits, and constructs a noise budget including the transimpedance amplifier. It also evaluates connector loss, fiber attenuation, dispersion, laser relative-intensity noise, and the optical power margin.

Approach and outcome

The team selects an APD because the link margin is limited and the additional receiver sensitivity outweighs the complexity. A temperature-compensated bias controller maintains the desired multiplication factor. An optical overload detector protects the APD, while a transimpedance amplifier is selected to match the detector capacitance and target bandwidth. Laboratory testing is performed across temperature, input power, and aging conditions.

Learning points

The case demonstrates that detector choice cannot be made from responsivity alone. System margin, noise, bias requirements, thermal behavior, reliability, and maintenance capability must be considered together. A PIN detector would be preferable if simplicity, low voltage, or high linearity dominated the design requirements.

Limitations, failure mechanisms, and professional practice

Common design mistakes include selecting a material without checking its cutoff wavelength, ignoring optical coupling losses, treating responsivity as equivalent to sensitivity, neglecting dark current and temperature dependence, and designing a high-speed detector without accounting for package capacitance. In photovoltaic systems, cell efficiency may be reported without considering module mismatch, inverter losses, shading, and temperature.

Packaging is often as important as the semiconductor structure. Thermal paths, hermetic sealing, fiber alignment, vibration, contamination, electrostatic discharge, and mechanical stress can determine field reliability. Qualification should include accelerated life testing, thermal cycling, humidity exposure, optical-power stress, and electrical overstress where appropriate.

Professional decisions also involve regulatory and ethical responsibilities. Laser safety, electrical isolation, eye protection, material handling, environmental compliance, and honest reporting of measurement uncertainty are essential. Photovoltaic and imaging systems may collect environmental or personal data; secure storage, privacy protection, and responsible use of automated image analysis must be addressed.

Current developments and best practices

Current research and industrial development emphasize high-efficiency micro-LED displays, quantum-dot emitters, silicon-photonic transceivers, heterogeneous laser integration, single-photon detectors, perovskite and tandem solar cells, short-wave infrared imaging, and photonic processors. Machine learning is increasingly used for optical-device calibration, fault prediction, image reconstruction, and adaptive control, but models require representative data and independent validation.

Best practice in optoelectronic design includes:

• Begin with wavelength, optical power, bandwidth, sensitivity, and environmental requirements.
• Use bandgap, absorption, gain, and carrier-transport data to screen candidate materials.
• Build an explicit optical, electrical, thermal, and noise budget.
• Include packaging and coupling losses from the earliest design stage.
• Characterize devices over temperature, bias, optical power, and aging conditions.
• Distinguish laboratory peak performance from guaranteed system performance.
• Provide safety controls, calibration procedures, failure detection, and maintainability.

Chapter Summary

Optoelectronic and photonic semiconductor devices exploit the interaction between photons and carriers. Absorption generates electron–hole pairs, while spontaneous and stimulated recombination produce optical emission. Direct-bandgap materials are preferred for efficient emitters, whereas carefully engineered junctions provide efficient detectors and photovoltaic conversion.

LEDs provide robust, spectrally broad spontaneous emission; laser diodes provide narrow, directional, stimulated emission; PIN photodiodes offer fast and low-noise detection; and avalanche photodiodes add internal gain at the cost of noise and bias complexity. Solar cells optimize photon-to-electric-power conversion, while image sensors combine photodetection with pixel-level readout and signal processing. Fiber-optic and integrated-photonic systems further require attention to coupling, waveguides, modulation, noise, thermal management, packaging, and safety. Effective engineering therefore requires simultaneous consideration of device physics, system budgets, manufacturing variation, reliability, and application constraints.

Key Terms

Photon: A quantum of electromagnetic radiation with energy hf.

Bandgap: Energy difference between the conduction-band minimum and valence-band maximum.

Direct-bandgap semiconductor: Material in which radiative transitions can occur without significant momentum assistance.

Absorption coefficient: Parameter describing exponential reduction of optical intensity in a material.

Spontaneous emission: Random photon emission during carrier recombination.

Stimulated emission: Emission triggered by an incident photon with matching phase, frequency, and direction.

Optical gain: Increase in optical power produced by stimulated emission exceeding absorption losses.

LED: Semiconductor emitter based primarily on spontaneous radiative recombination.

Laser diode: Injection semiconductor laser using an active region and optical resonator.

Threshold current: Minimum injection current at which laser gain balances cavity and internal losses.

Responsivity: Photocurrent produced per unit incident optical power.

Quantum efficiency: Ratio of collected charge carriers or emitted photons to incident photons or injected carriers, as defined for the device.

PIN photodiode: Photodiode containing a lightly doped or intrinsic absorption region between p and n layers.

Avalanche photodiode: Reverse-biased photodiode using impact ionization for internal gain.

Dark current: Current flowing through a detector in the absence of incident light.

Fill factor: Ratio of a solar cell’s maximum-power output to the product of open-circuit voltage and short-circuit current.

Photonic integrated circuit: Chip containing optical waveguides and functional photonic components.

Transimpedance amplifier: Amplifier that converts detector current into an output voltage.

Optical confinement factor: Fraction of optical mode power overlapping the active gain region.

Dynamic range: Ratio between the largest nonsaturating signal and the smallest reliably detectable signal.

Review Questions

Short Answer Questions

1. Why are direct-bandgap semiconductors generally preferred for LEDs and laser diodes?

2. Define responsivity and state its dependence on quantum efficiency and wavelength.

3. What is the purpose of the intrinsic region in a PIN photodiode?

4. Distinguish spontaneous emission from stimulated emission.

5. Name two major noise sources in an optical receiver.

Long Answer Questions

1. Explain the operating principles, structures, and performance trade-offs of LEDs and laser diodes.

2. Derive the principal factors affecting the bandwidth of a PIN photodiode.

3. Discuss avalanche multiplication, excess noise, temperature dependence, and applications of APDs.

4. Explain the current–voltage characteristics and efficiency parameters of a solar cell.

5. Describe the architecture and major design issues of a silicon-photonic communication link.

Analytical / Application Questions

1. A detector has quantum efficiency 0.80 and operates at 1550 nm. Calculate its ideal responsivity and identify practical factors that may reduce the measured value.

2. For a short-reach data link requiring low voltage and high linearity, compare a PIN photodiode with an APD and justify a device selection.

3. An image sensor shows motion distortion and excessive dark noise. Identify likely causes and propose device-level and system-level remedies.

22. Multiple Choice Questions with Answers

Question 1

Which material property most directly favors efficient light emission in an LED?

A. Indirect bandgap with strong phonon coupling

B. Direct bandgap permitting radiative recombination without major momentum assistance

C. Very high contact resistance

D. Large free-carrier absorption at the emission wavelength

Correct Answer: B. Direct bandgap permitting radiative recombination without major momentum assistance

Question 2

At laser threshold, the principal condition is that

A. the junction is at zero bias

B. mirror and internal losses exceed optical gain

C. optical gain balances total cavity and internal losses

D. all spontaneous emission ceases

Correct Answer: C. optical gain balances total cavity and internal losses

Question 3

For a photodiode with fixed quantum efficiency, responsivity generally increases with wavelength because

A. photon energy decreases as wavelength increases

B. dark current becomes zero

C. junction capacitance increases

D. the bandgap becomes larger

Correct Answer: A. photon energy decreases as wavelength increases

Question 4

The main function of the intrinsic region in a PIN photodiode is to

A. increase series resistance deliberately

B. provide an absorption and depletion region for carrier collection

C. produce stimulated emission

D. block all incident photons

Correct Answer: B. provide an absorption and depletion region for carrier collection

Question 5

An avalanche photodiode provides internal gain primarily through

A. thermionic emission

B. impact ionization in a strong reverse electric field

C. phosphor conversion

D. optical interference in a ring resonator

Correct Answer: B. impact ionization in a strong reverse electric field

Question 6

Which change generally improves solar-cell fill factor?

A. Increasing series resistance

B. Increasing shunt leakage

C. Reducing resistive losses while preserving junction quality

D. Increasing optical reflection

Correct Answer: C. Reducing resistive losses while preserving junction quality

Question 7

A major advantage of a laser diode over an LED in a long-distance fiber link is its

A. broader spectral width

B. lower coherence

C. narrower spectrum and higher modulation capability

D. inability to couple into single-mode fiber

Correct Answer: C. narrower spectrum and higher modulation capability

Question 8

If an APD is selected for a low-power optical receiver, the design must especially address

A. the absence of multiplication noise

B. high-voltage bias and temperature-dependent multiplication

C. phosphor aging only

D. the requirement for zero reverse bias

Correct Answer: B. high-voltage bias and temperature-dependent multiplication

Question 9

Rolling-shutter operation in an image sensor can produce

A. geometric distortion of moving objects

B. stimulated optical emission

C. higher avalanche multiplication

D. zero dark current

Correct Answer: A. geometric distortion of moving objects

Question 10

In an integrated photonic circuit, a ring resonator is commonly used for

A. wavelength-selective filtering or sensing

B. mechanical charge storage

C. eliminating all thermal effects

D. converting every photon into an electron with unity efficiency

Correct Answer: A. wavelength-selective filtering or sensing