13.1 Boundary Conditions
When a plane wave hits a planar interface between media $(n_1, \mu_1)$ and $(n_2, \mu_2)$, we match tangential E and H across the boundary (from Maxwell's equations). By Snell's law: $n_1\sin\theta_i = n_2\sin\theta_t$.
Derivation: Snell's Law from Boundary Conditions (Phase Matching)
Starting from the requirement that boundary conditions hold at all points on the interface at all times.
Step 1: Write the incident, reflected, and transmitted waves
Place the interface at $z = 0$. The three waves are:
$$\mathbf{E}_I = E_{0I}\,e^{i(\mathbf{k}_I \cdot \mathbf{r} - \omega_I t)}, \quad \mathbf{E}_R = E_{0R}\,e^{i(\mathbf{k}_R \cdot \mathbf{r} - \omega_R t)}, \quad \mathbf{E}_T = E_{0T}\,e^{i(\mathbf{k}_T \cdot \mathbf{r} - \omega_T t)}$$
Step 2: Apply boundary conditions at the interface
The tangential E must be continuous at $z = 0$ for all $x$, $y$, and $t$:
$$E_{0I}\,e^{i(\mathbf{k}_I \cdot \mathbf{r} - \omega_I t)} + E_{0R}\,e^{i(\mathbf{k}_R \cdot \mathbf{r} - \omega_R t)} = E_{0T}\,e^{i(\mathbf{k}_T \cdot \mathbf{r} - \omega_T t)} \quad \text{at } z=0$$
Step 3: Phase matching in time
For the equation to hold at all times, the time dependence must be the same:
$$\omega_I = \omega_R = \omega_T \equiv \omega$$
The frequency does not change upon reflection or transmission.
Step 4: Phase matching along the interface
At $z = 0$, the spatial phases must match for all $x$ and $y$:
$$k_{Ix}x + k_{Iy}y = k_{Rx}x + k_{Ry}y = k_{Tx}x + k_{Ty}y$$
Step 5: Confine to the plane of incidence
Choose the plane of incidence as the $xz$-plane, so $k_{Iy} = 0$. Then $k_{Ry} = k_{Ty} = 0$ (all three waves lie in the same plane), and:
$$k_I\sin\theta_i = k_R\sin\theta_r = k_T\sin\theta_t$$
Step 6: Derive Snell's law
Since $k_I = k_R = n_1\omega/c$ (same medium) and $k_T = n_2\omega/c$:
$$\frac{n_1\omega}{c}\sin\theta_i = \frac{n_1\omega}{c}\sin\theta_r = \frac{n_2\omega}{c}\sin\theta_t$$
From the first equality: $\theta_i = \theta_r$ (law of reflection).
From the first and third: $n_1\sin\theta_i = n_2\sin\theta_t$ (Snell's law).
Derivation: Fresnel Equations for Normal Incidence
Starting from matching boundary conditions for E and B at normal incidence ($\theta_i = 0$).
Step 1: Write the fields at the interface
Incident: $\mathbf{E}_I = E_I\,\hat{x}$, $\mathbf{B}_I = (E_I/v_1)\,\hat{y}$
Reflected: $\mathbf{E}_R = E_R\,\hat{x}$, $\mathbf{B}_R = -(E_R/v_1)\,\hat{y}$ (propagates in $-\hat{z}$)
Transmitted: $\mathbf{E}_T = E_T\,\hat{x}$, $\mathbf{B}_T = (E_T/v_2)\,\hat{y}$
Step 2: Boundary condition on tangential E
$$E_I + E_R = E_T \quad \text{...(i)}$$
Step 3: Boundary condition on tangential H
For non-magnetic media ($\mu_1 = \mu_2 = \mu_0$), continuity of $H_y = B_y/\mu_0$:
$$\frac{1}{v_1}(E_I - E_R) = \frac{1}{v_2}E_T$$
Using $v = c/n$: $n_1(E_I - E_R) = n_2 E_T$ ...(ii)
Step 4: Solve for the reflection coefficient
From (i): $E_T = E_I + E_R$. Substitute into (ii):
$$n_1(E_I - E_R) = n_2(E_I + E_R)$$
$$n_1 E_I - n_1 E_R = n_2 E_I + n_2 E_R$$
$$E_R(n_1 + n_2) = E_I(n_1 - n_2)$$
$$r \equiv \frac{E_R}{E_I} = \frac{n_1 - n_2}{n_1 + n_2}$$
Step 5: Solve for the transmission coefficient
Substitute back into (i):
$$t \equiv \frac{E_T}{E_I} = 1 + r = 1 + \frac{n_1 - n_2}{n_1 + n_2} = \frac{2n_1}{n_1 + n_2}$$
Step 6: Physical interpretation
If $n_2 > n_1$: $r < 0$ (reflected E flips sign, $180Β°$ phase shift).
If $n_2 < n_1$: $r > 0$ (no phase shift on reflection).
Note $t > 0$ always (no phase shift on transmission at normal incidence).
13.2 Fresnel Equations
Separating into s-polarization (TE: E β₯ plane of incidence) and p-polarization (TM: E β₯ plane):
s-polarization (TE)
p-polarization (TM)
Reflectance $R = r^2$, Transmittance $T = (n_2\cos\theta_t)/(n_1\cos\theta_i)\,t^2$, and $R + T = 1$ (energy conservation).
Derivation: Fresnel Equations for Oblique Incidence
Starting from boundary conditions at a planar interface for oblique incidence, treating s-polarization (TE) and p-polarization (TM) separately.
Step 1: Set up the geometry
Interface at $z = 0$, plane of incidence is $xz$. The incident wave arrives at angle $\theta_i$. By Snell's law, $n_1\sin\theta_i = n_2\sin\theta_t$.
Step 2: s-polarization (TE) -- E perpendicular to plane of incidence
$\mathbf{E}$ is along $\hat{y}$ for all three waves. Boundary conditions (tangential E and tangential H continuous):
$$E_I + E_R = E_T \quad \text{(tangential E)}$$
$$\frac{n_1}{\mu_1}(E_I - E_R)\cos\theta_i = \frac{n_2}{\mu_2}E_T\cos\theta_t \quad \text{(tangential H)}$$
Step 3: Solve for s-polarization coefficients
For non-magnetic media ($\mu_1 = \mu_2$), substitute $E_T = E_I + E_R$ into the H equation:
$$n_1\cos\theta_i(E_I - E_R) = n_2\cos\theta_t(E_I + E_R)$$
$$r_s = \frac{E_R}{E_I} = \frac{n_1\cos\theta_i - n_2\cos\theta_t}{n_1\cos\theta_i + n_2\cos\theta_t}$$
$$t_s = \frac{E_T}{E_I} = \frac{2n_1\cos\theta_i}{n_1\cos\theta_i + n_2\cos\theta_t}$$
Step 4: p-polarization (TM) -- E parallel to plane of incidence
$\mathbf{E}$ lies in the $xz$-plane. The boundary conditions become:
$$(E_I - E_R)\cos\theta_i = E_T\cos\theta_t \quad \text{(tangential E: x-component)}$$
$$\frac{n_1}{\mu_1}(E_I + E_R) = \frac{n_2}{\mu_2}E_T \quad \text{(tangential H: y-component)}$$
Step 5: Solve for p-polarization coefficients
For non-magnetic media, from the H condition: $E_T = n_1(E_I + E_R)/n_2$. Substitute into tangential E:
$$(E_I - E_R)\cos\theta_i = \frac{n_1}{n_2}(E_I + E_R)\cos\theta_t$$
$$r_p = \frac{E_R}{E_I} = \frac{n_2\cos\theta_i - n_1\cos\theta_t}{n_2\cos\theta_i + n_1\cos\theta_t}$$
$$t_p = \frac{E_T}{E_I} = \frac{2n_1\cos\theta_i}{n_2\cos\theta_i + n_1\cos\theta_t}$$
Step 6: Verify normal incidence limit
At $\theta_i = \theta_t = 0$: $r_s = r_p = (n_1 - n_2)/(n_1 + n_2)$ and $t_s = t_p = 2n_1/(n_1 + n_2)$, recovering the normal-incidence result.
Derivation: Reflectance, Transmittance, and Energy Conservation
Starting from the Poynting vector to derive the reflected and transmitted power fractions.
Step 1: Define reflectance and transmittance
Reflectance $R$ is the fraction of incident power reflected; transmittance $T$ is the fraction transmitted.
Step 2: Power flux normal to the interface
The time-averaged Poynting vector component normal to the interface for each wave:
$$\langle S_z \rangle_I = \frac{n_1}{2\mu_0 c}E_I^2\cos\theta_i, \quad \langle S_z \rangle_R = \frac{n_1}{2\mu_0 c}E_R^2\cos\theta_i, \quad \langle S_z \rangle_T = \frac{n_2}{2\mu_0 c}E_T^2\cos\theta_t$$
Step 3: Compute R
$$R = \frac{\langle S_z \rangle_R}{\langle S_z \rangle_I} = \frac{E_R^2}{E_I^2} = |r|^2$$
The reflectance is simply the square of the Fresnel reflection coefficient.
Step 4: Compute T
$$T = \frac{\langle S_z \rangle_T}{\langle S_z \rangle_I} = \frac{n_2\cos\theta_t}{n_1\cos\theta_i}\frac{E_T^2}{E_I^2} = \frac{n_2\cos\theta_t}{n_1\cos\theta_i}|t|^2$$
Note that $T \neq |t|^2$ in general due to the beam width change and impedance mismatch.
Step 5: Verify energy conservation R + T = 1
For s-polarization, substituting $r_s$ and $t_s$:
$$R_s + T_s = \left(\frac{n_1\cos\theta_i - n_2\cos\theta_t}{n_1\cos\theta_i + n_2\cos\theta_t}\right)^2 + \frac{n_2\cos\theta_t}{n_1\cos\theta_i}\left(\frac{2n_1\cos\theta_i}{n_1\cos\theta_i + n_2\cos\theta_t}\right)^2$$
Let $a = n_1\cos\theta_i$ and $b = n_2\cos\theta_t$:
$$= \frac{(a-b)^2}{(a+b)^2} + \frac{4ab}{(a+b)^2} = \frac{(a-b)^2 + 4ab}{(a+b)^2} = \frac{(a+b)^2}{(a+b)^2} = 1 \;\checkmark$$
13.3 Special Angles
Brewster's Angle
p-polarized light is completely transmitted at Brewster's angle. Used in laser Brewster windows.
Critical Angle (TIR)
For $\theta_i > \theta_c$, total internal reflection: $R = 1$. Basis of optical fibers.
Derivation: Brewster's Angle
Starting from the p-polarization Fresnel coefficient, find the angle at which reflection vanishes.
Step 1: Set the p-polarization reflection coefficient to zero
$$r_p = \frac{n_2\cos\theta_i - n_1\cos\theta_t}{n_2\cos\theta_i + n_1\cos\theta_t} = 0$$
This requires the numerator to vanish:
$$n_2\cos\theta_B = n_1\cos\theta_t$$
Step 2: Use Snell's law to eliminate the transmitted angle
From Snell's law: $\sin\theta_t = (n_1/n_2)\sin\theta_B$, so:
$$\cos\theta_t = \sqrt{1 - \frac{n_1^2}{n_2^2}\sin^2\theta_B}$$
Step 3: Substitute and square both sides
$$n_2^2\cos^2\theta_B = n_1^2\cos^2\theta_t = n_1^2\left(1 - \frac{n_1^2}{n_2^2}\sin^2\theta_B\right)$$
$$n_2^2\cos^2\theta_B = n_1^2 - \frac{n_1^4}{n_2^2}\sin^2\theta_B$$
Step 4: Use $\cos^2\theta = 1 - \sin^2\theta$ and solve
$$n_2^2(1 - \sin^2\theta_B) = n_1^2 - \frac{n_1^4}{n_2^2}\sin^2\theta_B$$
$$n_2^2 - n_2^2\sin^2\theta_B = n_1^2 - \frac{n_1^4}{n_2^2}\sin^2\theta_B$$
$$\sin^2\theta_B\left(\frac{n_1^4}{n_2^2} - n_2^2\right) = n_1^2 - n_2^2$$
$$\sin^2\theta_B\cdot\frac{n_1^4 - n_2^4}{n_2^2} = n_1^2 - n_2^2$$
Step 5: Factor and simplify
Since $n_1^4 - n_2^4 = (n_1^2 - n_2^2)(n_1^2 + n_2^2)$:
$$\sin^2\theta_B = \frac{n_2^2}{n_1^2 + n_2^2}, \qquad \cos^2\theta_B = \frac{n_1^2}{n_1^2 + n_2^2}$$
$$\tan\theta_B = \frac{\sin\theta_B}{\cos\theta_B} = \frac{n_2}{n_1}$$
Step 6: Note that the reflected and transmitted rays are perpendicular
At Brewster's angle: $\theta_B + \theta_t = 90Β°$. Proof: $n_1\sin\theta_B = n_2\sin\theta_t = n_2\cos\theta_B$ gives $\tan\theta_B = n_2/n_1$ and $\sin\theta_t = \cos\theta_B = \sin(90Β°-\theta_B)$.
Physically, the reflected wave would need to oscillate along its own propagation direction, which is impossible for a transverse wave.
Derivation: Total Internal Reflection and Evanescent Waves
Starting from Snell's law for the case $n_1 > n_2$, derive the critical angle and evanescent wave.
Step 1: Find the critical angle
From Snell's law: $\sin\theta_t = (n_1/n_2)\sin\theta_i$. Since $n_1 > n_2$, $\sin\theta_t > \sin\theta_i$.
When $\sin\theta_t = 1$ (maximum), we get the critical angle:
$$\sin\theta_c = \frac{n_2}{n_1}$$
Step 2: What happens beyond the critical angle?
For $\theta_i > \theta_c$: $\sin\theta_t = (n_1/n_2)\sin\theta_i > 1$. Then:
$$\cos\theta_t = \sqrt{1 - \sin^2\theta_t} = i\sqrt{\sin^2\theta_t - 1} \equiv i\beta$$
where $\beta = \sqrt{(n_1/n_2)^2\sin^2\theta_i - 1}$ is real and positive.
Step 3: The transmitted wave becomes evanescent
The transmitted wave's $z$-dependence: $e^{ik_{Tz}z}$ where $k_{Tz} = k_T\cos\theta_t = (n_2\omega/c)(i\beta)$:
$$\mathbf{E}_T \propto e^{ik_{Tx}x}\,e^{-\kappa z}\,e^{-i\omega t}$$
where $\kappa = (n_2\omega/c)\beta = (\omega/c)\sqrt{n_1^2\sin^2\theta_i - n_2^2}$.
Step 4: The field decays exponentially into medium 2
$$|\mathbf{E}_T| \propto e^{-\kappa z} = e^{-z/d}$$
The penetration depth is:
$$d = \frac{1}{\kappa} = \frac{c/\omega}{\sqrt{n_1^2\sin^2\theta_i - n_2^2}} = \frac{\lambda/(2\pi)}{\sqrt{n_1^2\sin^2\theta_i - n_2^2}}$$
Typically $d \sim \lambda$, so the evanescent wave extends about one wavelength into medium 2.
Step 5: Show R = 1 (total reflection)
Substituting $\cos\theta_t = i\beta$ into the Fresnel coefficient for s-polarization:
$$r_s = \frac{n_1\cos\theta_i - n_2(i\beta)}{n_1\cos\theta_i + n_2(i\beta)} = \frac{a - ib'}{a + ib'}$$
where $a = n_1\cos\theta_i$ and $b' = n_2\beta$ are both real. Therefore:
$$|r_s|^2 = \frac{a^2 + b'^2}{a^2 + b'^2} = 1 \implies R_s = 1$$
Similarly $R_p = 1$. All incident energy is reflected -- no time-averaged power flows into medium 2.
Step 6: Phase shift on total internal reflection
Although $|r| = 1$, the reflection coefficient acquires a phase:
$$r_s = e^{i\phi_s}, \quad \tan(\phi_s/2) = \frac{\sqrt{\sin^2\theta_i - (n_2/n_1)^2}}{\cos\theta_i}$$
This phase shift is exploited in the Fresnel rhomb to convert linear to circular polarization.
Simulation: Fresnel Equations
Fresnel Equations & TIR
Plots Rs, Rp, Ts, Tp vs angle for air-glass interface, shows Brewster's angle and total internal reflection.
Click Run to execute the Python code
First run will download Python environment (~15MB)
Video Lectures & Demonstrations
MIT β Fresnel equations derivation, Brewster's angle, and total internal reflection with demonstrations.
Thin film interference and anti-reflection coatings β how optical coatings use wave interference.
Fortran Implementation
Transfer matrix method for thin-film reflectance β computes the reflectance of a multi-layer anti-reflection coating as a function of wavelength.
Thin Film AR Coating
FortranTransfer matrix method for single-layer anti-reflection coating reflectance vs wavelength
Click Run to execute the Fortran code
Code will be compiled with gfortran and executed on the server
Griffiths Problem Solutions
Video walkthroughs of Griffiths problems on boundary conditions, image charges, and separation of variables.
Problem 3.25
Problem 3.26
Problem 3.27
Problem 3.28
Problem 3.29
Problem 3.30