Electromagnetic Optics and Light-Wave Propagation

Introduction and Context

Optical communication systems convey information by controlling electromagnetic radiation, usually in the near-infrared region around 850, 1310, or 1550 nm. A rigorous understanding of electromagnetic optics is therefore essential for analysing optical sources, fibres, connectors, photodetectors, and free-space optical links. The subject connects Maxwell’s equations with geometric-ray models and wave phenomena such as interference, diffraction, polarization, coherence, and Gaussian-beam propagation.

Ray optics is highly useful when the wavelength is much smaller than the dimensions of the structures involved. Wave optics becomes necessary when phase, interference, diffraction, polarization, or modal behaviour determines system performance. Optical-fibre theory uses both viewpoints: rays provide an intuitive description of guidance, whereas electromagnetic mode theory gives the accurate description of field distributions, propagation constants, and dispersion.

Learning Objectives

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

1. State and interpret Maxwell’s equations for optical media.

2. Derive the electromagnetic wave equation and explain plane-wave propagation.

3. Apply reflection, refraction, Snell’s law, and Fresnel equations.

4. Explain interference, diffraction, polarization, and coherence in communication systems.

5. Analyse total internal reflection and the numerical aperture of an optical fibre.

6. Describe Gaussian beams and calculate their propagation parameters.

7. Evaluate practical limitations and select suitable optical models for engineering problems.

Maxwell’s Equations and Constitutive Relations

In differential form, Maxwell’s equations are

∇ · D = ρ,

∇ · B = 0,

∇ × E = −∂B/∂t,

∇ × H = J + ∂D/∂t.

Here E is electric field intensity (V m−1), H is magnetic field intensity (A m−1), D is electric flux density (C m−2), B is magnetic flux density (T), ρ is free-charge density (C m−3), and J is free-current density (A m−2).

For a linear, isotropic medium,

D = εE, B = μH, J = σE,

where ε is permittivity, μ is permeability, and σ is electrical conductivity. In a homogeneous, source-free, lossless dielectric, ρ = 0, J = 0, and σ ≈ 0. For most nonmagnetic optical materials, μ ≈ μ0 and ε = ε0 n², where n is the refractive index.

The refractive index may depend on wavelength. This material dispersion is important in optical communication because different spectral components of a pulse may propagate at different velocities. In an absorbing medium, ε is generally complex; the imaginary part represents attenuation associated with energy dissipation.

Derivation of the Electromagnetic Wave Equation

Consider a homogeneous, source-free, nonmagnetic dielectric. Maxwell’s curl equations reduce to

∇ × E = −μ ∂H/∂t,

∇ × H = ε ∂E/∂t.

Taking the curl of the first equation gives

∇ × (∇ × E) = −μ ∂(∇ × H)/∂t = −με ∂²E/∂t².

Using the vector identity

∇ × (∇ × E) = ∇(∇ · E) − ∇²E

and ∇ · E = 0 in a homogeneous source-free medium, one obtains

∇²E − με ∂²E/∂t² = 0.

Similarly,

∇²H − με ∂²H/∂t² = 0.

The wave velocity is

v = 1/√(με) = c/n,

where c is the speed of light in vacuum. The derivation assumes linearity, homogeneity, isotropy, negligible conductivity, and absence of free charge and current. It does not apply without modification to strongly nonlinear, anisotropic, dispersive, or conducting media.

For a plane wave travelling in the +z direction, a harmonic electric field can be written as

E(z,t) = E0 cos(ωt − βz + φ),

where E0 is amplitude, ω is angular frequency in rad s−1, β is phase constant in rad m−1, and φ is phase. The phase constant is

β = nω/c = 2πn/λ0,

where λ0 is the vacuum wavelength. The wavelength inside the medium is λ = λ0/n. The electric and magnetic fields are transverse to the direction of propagation and mutually perpendicular. Their amplitude ratio is the intrinsic impedance,

η = |E|/|H| = √(μ/ε).

For a nonmagnetic dielectric, η ≈ η0/n, where η0 ≈ 377 Ω is the impedance of free space.

The time-averaged power density of a uniform plane wave is represented by the Poynting vector,

S = E × H.

For sinusoidal fields in a lossless medium, the average magnitude is

⟨S⟩ = |E0|²/(2η) = η|H0|²/2.

This relation is useful for connecting optical field amplitude with irradiance and transmitted power.

Phase, Group Velocity, and Dispersion

The phase velocity is

v_p = ω/β = c/n.

A modulated optical signal contains a range of frequencies. The envelope travels approximately at the group velocity,

v_g = dω/dβ.

The group index is defined by

n_g = c/v_g = n + ω(dn/dω) = n − λ0(dn/dλ0).

The group velocity determines the arrival time of a pulse, whereas phase velocity determines the motion of a constant phase point. If n varies with wavelength, different spectral components experience different group velocities. This produces chromatic or material dispersion and broadens pulses in fibre links.

A useful engineering quantity is the group-velocity-dispersion parameter β2 = d²β/dω². For a sufficiently narrowband pulse, the sign and magnitude of β2 determine whether the pulse broadens and whether phase modulation can be converted into amplitude variation. In practical fibres, material dispersion and waveguide dispersion jointly determine the total dispersion.

Boundary Conditions, Reflection, and Refraction

At an interface between two media, Maxwell’s equations imply the boundary conditions

E1t = E2t,

H1t − H2t = K_s,

D2n − D1n = ρ_s,

B1n = B2n,

where subscripts t and n denote tangential and normal components, while K_s and ρ_s are surface current and charge densities. For an ordinary dielectric interface with no free surface charge or current, tangential E and H are continuous, and normal D and B are continuous.

When a plane wave enters a second dielectric, the tangential phase must remain continuous. This gives Snell’s law:

n1 sin θ1 = n2 sin θ2.

Angles are measured from the normal to the interface. If n2 > n1, the ray bends towards the normal; if n2 < n1, it bends away from the normal.

The critical angle exists when light travels from a higher-index medium to a lower-index medium. It is defined by θ2 = 90°:

sin θc = n2/n1, n1 > n2.

For incidence angles θ1 > θc, total internal reflection occurs, provided the interface is sufficiently smooth and the second medium does not introduce significant loss.

Fresnel Reflection Coefficients

For nonmagnetic media, the electric-field amplitude reflection coefficients are

r_s = (n1 cos θ1 − n2 cos θ2)/(n1 cos θ1 + n2 cos θ2),

r_p = (n2 cos θ1 − n1 cos θ2)/(n2 cos θ1 + n1 cos θ2),

for perpendicular (s) and parallel (p) polarization, respectively. The reflected power fraction is R = |r|². The transmitted power fraction must be calculated from the normal components of the Poynting vector; at a lossless interface,

T = (n2 cos θ2)/(n1 cos θ1)|t|²,

and R + T = 1.

At normal incidence, polarization distinctions disappear and

r = (n1 − n2)/(n1 + n2),

R = [(n1 − n2)/(n1 + n2)]².

The Brewster angle for p-polarized light satisfies

tan θB = n2/n1.

At this incidence angle, the reflected p-polarized field vanishes for ideal lossless isotropic dielectrics. This principle is used in polarization control and reduces reflection in some optical components.

Worked Example 1: Fibre Critical Angle and Numerical Aperture

An ideal step-index fibre has core index n1 = 1.48 and cladding index n2 = 1.46. Find the critical angle at the core–cladding interface and the numerical aperture for air outside the fibre.

The critical angle is

θc = sin−1(n2/n1) = sin−1(1.46/1.48) ≈ sin−1(0.98649) ≈ 80.6°.

The numerical aperture in air is

NA = √(n1² − n2²).

Thus,

NA = √(1.48² − 1.46²) = √(2.1904 − 2.1316)

= √0.0588 ≈ 0.2425.

The acceptance half-angle in air is

θa = sin−1(NA) ≈ sin−1(0.2425) ≈ 14.0°.

The critical angle is measured inside the core from the normal to the core–cladding boundary, whereas the acceptance angle is measured at the fibre entrance from the fibre axis. They are different geometric quantities.

Optical-Fibre Guidance: Ray and Wave Descriptions

In a step-index fibre, the core has approximately constant index n1 and the cladding has lower index n2. The ray picture explains guidance through repeated total internal reflection. A meridional ray entering within the acceptance cone is guided if its incidence at the core–cladding interface exceeds the critical angle.

For a fibre in an external medium of refractive index n0, the maximum acceptance angle θa satisfies

n0 sin θa = √(n1² − n2²).

Therefore,

NA = n0 sin θa = √(n1² − n2²).

This expression assumes a circular, step-index, weakly absorbing fibre and ray-optics conditions. For a graded-index fibre, the index changes continuously with radius and the ray description must be modified.

The normalized frequency or V-number is

V = (2πa/λ0) NA,

where a is core radius. For a conventional step-index circular fibre, single-mode operation is obtained approximately when

V < 2.405.

This criterion applies to the ideal weakly guiding step-index fibre and refers to the cutoff of the first higher-order mode. The fundamental mode has no cutoff in the same ideal model. The approximate number of guided modes in a highly multimode step-index fibre is

M ≈ V²/2,

when V is sufficiently large. This is an approximation and is not appropriate near cutoff or for detailed mode counting.

The wave description treats the fibre as a dielectric waveguide. Only field distributions satisfying Maxwell’s equations and the boundary conditions can propagate. Each mode has a propagation constant β satisfying, for guided modes,

n2k0 < β < n1k0,

where k0 = 2π/λ0. Modes possess distinct propagation constants and group velocities, producing intermodal dispersion in multimode fibres. Single-mode fibres substantially reduce intermodal dispersion but still experience material and waveguide dispersion.

Interference and Coherence

When two optical fields overlap, the total intensity is not generally the sum of their individual intensities. If

E1 = A1 cos(ωt + φ1),

E2 = A2 cos(ωt + φ2),

then, for equal frequencies and time averaging,

I = I1 + I2 + 2√(I1I2) cos Δφ,

where Δφ = φ1 − φ2. Constructive interference occurs for Δφ = 2mπ, and destructive interference for Δφ = (2m + 1)π, with integer m.

Interference requires an appropriate degree of coherence. Temporal coherence describes phase correlation over time and is associated with spectral linewidth. A rough coherence time is τc ≈ 1/Δν for a source with spectral width Δν, with the precise numerical factor depending on the spectral shape and definition. The coherence length is approximately

Lc ≈ v_g τc,

or in air approximately cτc. Spatial coherence describes correlation between different points across a wavefront. Lasers generally have much greater temporal and spatial coherence than LEDs.

In communication systems, coherence influences interferometric receivers, optical frequency-domain measurements, coherent detection, fibre sensors, and interference noise. Incoherent sources can still be useful because low coherence reduces unwanted interference and speckle.

Diffraction and Apertures

Diffraction is the spreading and restructuring of a wave when it encounters an aperture, obstacle, or waveguide boundary. It becomes important when the aperture dimension is comparable to the wavelength or when long propagation distances magnify angular divergence.

For a circular aperture of diameter D, the first dark ring of the Airy pattern occurs at an angular radius approximately

θ ≈ 1.22 λ/D,

where λ is the wavelength in the propagation medium. This result assumes uniform illumination and the Fraunhofer far-field condition. For a rectangular single slit of width a, minima satisfy

a sin θ = mλ,

where m = ±1, ±2, ... .

Diffraction limits the focusing of optical beams, affects coupling into fibre cores, and determines the divergence of semiconductor-laser beams. It also explains why an optical system cannot confine a beam arbitrarily tightly without increasing angular spread.

Polarization

Polarization specifies the trajectory and orientation of the electric-field vector transverse to propagation. Linear polarization has a fixed direction. Circular polarization has two orthogonal equal-amplitude components with a ±90° phase difference. Elliptical polarization is the general case.

A linearly polarized wave can be represented by Jones vectors. For example, horizontal and vertical polarization are

[1, 0]ᵀ and [0, 1]ᵀ.

A quarter-wave plate introduces a relative phase shift of π/2 between orthogonal components. A half-wave plate introduces a phase shift of π and rotates the direction of linear polarization by twice the angle between the incident polarization and the plate’s principal axis.

For ideal linear polarizers, Malus’s law gives

I = I0 cos² θ,

where θ is the angle between the incident linear polarization and the transmission axis. For unpolarized light passing through an ideal polarizer, the average transmitted intensity is I0/2.

Fibre polarization is affected by birefringence, stress, bending, temperature, and imperfections. Polarization-mode dispersion occurs because two orthogonal polarization modes may have different group velocities. Polarization-maintaining fibre deliberately introduces strong birefringence so that launched polarization remains aligned with a principal axis.

Gaussian-Beam Propagation

Many laser beams are well approximated by the fundamental Gaussian mode. In the paraxial approximation, the beam radius, defined as the radial distance at which intensity falls to 1/e² of its axial value, is

w(z) = w0 √[1 + (z/zR)²],

where w0 is the waist radius and z is measured from the waist. The Rayleigh range is

zR = πw0²/λ,

where λ is the wavelength in the propagation medium.

At z = zR, the beam radius is √2 w0 and the on-axis intensity has decreased because the beam has begun to diverge. Far from the waist, |z| >> zR,

w(z) ≈ w0 |z|/zR,

and the half-angle divergence is approximately

θ ≈ λ/(πw0)

for a diffraction-limited beam in a medium where λ is the wavelength in that medium. The product w0θ ≈ λ/π expresses the beam-quality trade-off: tighter focusing produces larger divergence.

The wavefront radius of curvature is

R(z) = z[1 + (zR/z)²],

and the Gouy phase is

ψ(z) = tan−1(z/zR).

Real beams are described using the beam-quality factor M². For an ideal Gaussian beam M² = 1. A practical beam has

θ ≈ M² λ/(πw0),

and a larger M² indicates poorer focusability or departure from the fundamental mode.

Worked Example 2: Gaussian-Beam Rayleigh Range

A laser beam has waist radius w0 = 2.0 μm and wavelength λ0 = 1550 nm in air. Find its Rayleigh range and approximate far-field divergence.

Using λ ≈ 1.55 × 10−6 m,

zR = π(2.0 × 10−6)²/(1.55 × 10−6)

= π(4.0 × 10−12)/(1.55 × 10−6)

≈ 8.11 × 10−6 m.

Thus zR ≈ 8.1 μm. The divergence half-angle is

θ ≈ λ/(πw0)

= 1.55 × 10−6/[π(2.0 × 10−6)]

≈ 0.246 rad ≈ 14.1°.

The very small waist produces a comparatively large divergence, illustrating the diffraction trade-off.

From Maxwell Fields to Communication-System Models

A communication engineer must choose a model appropriate to the physical scale and performance question. The plane-wave model is suitable for reflection, transmission, impedance, and power-flow calculations at planar interfaces. Ray optics is useful for acceptance cones, connector alignment, and intuitive analysis of multimode guidance. Full waveguide analysis is required for mode cutoffs, field distributions, polarization effects, and dispersion. Gaussian optics is appropriate for many laser beams, lens systems, and free-space links under paraxial conditions.

The hierarchy is not a set of competing descriptions; rather, each model is an approximation derived from, or consistent with, Maxwell’s equations. Errors occur when a model is used beyond its assumptions. For example, ray optics cannot predict modal interference in a single-mode fibre, while a full electromagnetic solution may be unnecessary for estimating the acceptance angle of a large multimode link.

Applied Case: Coupling a 1550-nm Laser into a Single-Mode Fibre

Background

Consider an illustrative design case in which a 1550-nm distributed-feedback laser must be coupled into a standard single-mode fibre for a metropolitan optical link. The coupling lens must provide adequate power transfer while tolerating small lateral and angular alignment errors.

Problem

The laser emits an approximately Gaussian beam with waist radius 2.5 μm. The fibre supports a fundamental mode whose field radius is approximately 5.2 μm at the coupling plane. Determine whether a simple direct waist match is possible and identify the dominant design concerns.

Analysis

For maximum ideal Gaussian coupling, the incident beam should match the fibre mode in size, phase curvature, polarization, and transverse position. A beam waist of 2.5 μm is substantially smaller than the fibre-mode radius of 5.2 μm. If the beam is focused directly to the laser waist, it will underfill the fibre mode and produce mode mismatch loss.

A first design step is to select a lens and beam-expansion arrangement that produces a waist close to 5.2 μm at the fibre facet. The required magnification of the beam radius is approximately

5.2/2.5 = 2.08.

This is only a geometric starting point; the complete design must include the input beam’s curvature, lens aberration, wavelength, and the actual complex field of both source and fibre mode.

The fibre’s numerical aperture determines the angular acceptance, but matching NA alone does not guarantee high coupling. Longitudinal displacement changes phase curvature, lateral displacement reduces field overlap, and angular tilt introduces phase variation across the mode. Polarization mismatch can also reduce useful coupling, particularly when polarization-sensitive components are present.

Approach and Expected Outcome

An optical engineer would use an aspheric or compound coupling lens, adjust the beam magnification, and optimize the six alignment degrees of freedom: three translations and three rotations. The final metric should be measured coupled power or insertion loss, not merely visual alignment. Active alignment using a photodetector is often preferred for production packaging.

Learning Points

1. The fibre mode is an electromagnetic field, not simply a geometric spot.

2. Maximum coupling requires overlap of amplitude, phase, and polarization.

3. Numerical aperture describes angular acceptance but does not fully characterize mode matching.

4. Mechanical tolerances, thermal drift, contamination, and laser ageing must be included in a production design.

5. Reliable designs use calibrated power measurements and statistical tolerance analysis rather than nominal alignment alone.

Challenges, Limitations, and Contemporary Developments

Real optical systems depart from ideal assumptions. Material absorption, surface roughness, scattering, nonlinear response, anisotropy, and temperature variation can alter propagation. Connectors introduce Fresnel reflection and lateral or angular misalignment. Fibre bends cause radiation loss and can modify polarization. High optical powers may produce self-phase modulation, cross-phase modulation, stimulated scattering, or other nonlinear effects.

Coherent systems require accurate management of laser linewidth, phase noise, polarization evolution, and digital signal processing. Space-division multiplexing uses multiple cores or modes, but mode coupling and differential modal delay complicate receiver design. Photonic integrated circuits increasingly combine waveguides, resonators, modulators, and detectors on compact platforms; electromagnetic simulation is essential because subwavelength geometry determines effective index and loss.

AI and machine learning are relevant when used responsibly for tasks such as fibre fault classification, channel equalization, beam alignment, and predictive maintenance. They do not replace physical validation. Training data may be biased toward particular equipment or environmental conditions, and a model should not be deployed without monitoring, explainable performance criteria, and fail-safe operating limits.

Best practice includes documenting model assumptions, using SI units consistently, distinguishing phase and group quantities, checking energy conservation at interfaces, validating approximations against numerical or experimental results, and reporting uncertainty. Optical safety is also essential: invisible infrared radiation can cause retinal injury even when no bright visible warning is present. Designs and laboratory procedures should comply with applicable laser-safety standards and institutional controls.

Chapter Summary

Maxwell’s equations provide the fundamental description of electromagnetic propagation. In a homogeneous, source-free dielectric they lead to the wave equation, with phase velocity c/n and impedance determined by the material properties. Boundary conditions produce reflection, transmission, Snell’s law, critical angle, total internal reflection, Fresnel coefficients, and Brewster-angle behaviour.

Ray optics explains fibre acceptance and total internal reflection, while waveguide theory explains modes, propagation constants, cutoff, and dispersion. The numerical aperture and V-number are useful engineering parameters, but their assumptions must be recognized. Interference depends on phase and coherence; diffraction limits focusing and beam confinement; polarization influences components, fibres, and system penalties. Gaussian-beam equations describe waist evolution, Rayleigh range, divergence, wavefront curvature, and beam quality.

Effective optical-communication analysis requires selecting the correct model, checking its assumptions, and relating field behaviour to link-level metrics such as power, coupling loss, bandwidth, dispersion, and reliability. Contemporary systems extend these principles into coherent detection, photonic integration, space-division multiplexing, and intelligent monitoring, while maintaining requirements for safety and engineering validation.

Key Terms

Maxwell’s equations: Fundamental differential equations governing electric and magnetic fields.

Permittivity: A material parameter relating electric flux density to electric field.

Refractive index: Ratio of the speed of light in vacuum to phase velocity in a medium.

Plane wave: An ideal wave whose constant-phase surfaces are planes.

Poynting vector: Vector representing electromagnetic power-flow density.

Phase velocity: Speed at which a constant phase point propagates.

Group velocity: Speed at which a narrowband signal envelope propagates.

Snell’s law: Relation between incident and refracted angles at an interface.

Critical angle: Incidence angle at which the refracted ray is tangent to the interface.

Total internal reflection: Complete reflection at a higher- to lower-index interface above the critical angle.

Fresnel coefficient: Amplitude or power ratio describing reflection or transmission at an interface.

Numerical aperture: Measure of the light-acceptance capability of a fibre.

V-number: Dimensionless fibre parameter involving core radius, wavelength, and numerical aperture.

Mode: An allowed electromagnetic field distribution of a waveguide.

Coherence: Degree of phase correlation between optical fields.

Diffraction: Wave spreading caused by apertures, obstacles, or confinement.

Polarization: Orientation and evolution of the transverse electric-field vector.

Rayleigh range: Distance from a Gaussian waist over which the beam radius increases by √2.

Beam-quality factor: M², a measure of departure from an ideal Gaussian beam.

Review Questions

Short Answer Questions

1. State Maxwell’s four equations in differential form and identify the physical meaning of each.

2. Distinguish phase velocity from group velocity.

3. Define numerical aperture and state the assumptions behind the step-index expression.

4. What is meant by temporal coherence?

5. Define the Rayleigh range of a Gaussian beam.

Long Answer Questions

1. Derive the electromagnetic wave equation in a homogeneous, source-free, lossless dielectric and state all assumptions.

2. Explain reflection, refraction, critical angle, and total internal reflection using boundary conditions and Snell’s law.

3. Compare ray-optics and waveguide descriptions of optical-fibre propagation.

4. Explain interference, coherence, diffraction, and polarization as they apply to optical communication systems.

5. Derive and interpret the principal Gaussian-beam propagation relations.

Analytical / Application Questions

1. A step-index fibre has n1 = 1.50, n2 = 1.47, and core radius 4 μm. At λ0 = 1310 nm, calculate its numerical aperture and V-number. Assess whether the ideal single-mode criterion is satisfied.

2. A plane wave travels from glass with n1 = 1.50 into air. Calculate the critical angle and explain what happens for incidence angles below and above this value.

3. A laser beam with wavelength 1550 nm is focused to a waist radius of 4 μm. Calculate its Rayleigh range and far-field divergence, then discuss the practical alignment implications.

Multiple Choice Questions with Answers

Question 1

In a homogeneous, source-free, lossless dielectric, the electromagnetic wave velocity is

A. √(με)

B. 1/√(με)

C. μ/ε

D. ε/μ

Correct Answer: B. 1/√(με)

Question 2

For light incident normally from a medium of refractive index n1 to n2, the reflected power fraction is

A. (n1+n2)²

B. (n1−n2)/(n1+n2)

C. [(n1−n2)/(n1+n2)]²

D. n2/n1

Correct Answer: C. [(n1−n2)/(n1+n2)]²

Question 3

Total internal reflection requires light to travel

A. From lower index to higher index at any angle

B. From higher index to lower index with incidence above the critical angle

C. From vacuum to glass at normal incidence

D. Between equal-index media

Correct Answer: B. From higher index to lower index with incidence above the critical angle

Question 4

The numerical aperture of an ideal step-index fibre in air is

A. n1+n2

B. n1/n2

C. √(n1²−n2²)

D. 2πa/λ

Correct Answer: C. √(n1²−n2²)

Question 5

The approximate single-mode condition for an ideal step-index circular fibre is

A. V > 10

B. V = 0

C. V < 2.405

D. V < 0.405 only

Correct Answer: C. V < 2.405

Question 6

The group velocity of a pulse is defined by

A. dβ/dω

B. ωβ

C. dω/dβ

D. c/n² only

Correct Answer: C. dω/dβ

Question 7

For two coherent fields, destructive interference occurs when their phase difference is

A. 2mπ

B. mπ/2 only

C. (2m+1)π

D. π/4

Correct Answer: C. (2m+1)π

Question 8

For an ideal Gaussian beam, the Rayleigh range is

A. λ/(πw0²)

B. πw0²/λ

C. πλw0²

D. w0/λ

Correct Answer: B. πw0²/λ

Question 9

The first Airy minimum for a circular aperture of diameter D occurs approximately at

A. θ = D/λ

B. θ = λ/D

C. θ = 1.22λ/D

D. θ = 2D/λ

Correct Answer: C. θ = 1.22λ/D

Question 10

Malus’s law for linearly polarized light transmitted through an ideal polarizer is

A. I = I0 sin θ

B. I = I0 cos² θ

C. I = I0 tan θ

D. I = I0/θ²

Correct Answer: B. I = I0 cos² θ