The Timpson Lyceum seal

The Timpson Lyceum · University Physics II (2026) – Electricity & Magnetism

Chapter 35 — Interference

Problems with worked solutions · 14 points · 3 extra credit

Chapter 35 — Interference

Core: 14 points · Extra credit: 3 points. Show your reasoning; partial credit is given.

  1. 1. (1 pt •) Two loudspeakers driven by the same amplifier emit identical waves of wavelength λ=480 nm\lambda=480\,\text{nm} (treat as coherent light sources for this problem). At a certain point the waves from the two sources have traveled path lengths that differ by Δr=1.44 μm\Delta r = 1.44\,\mu\text{m}.

    Is the interference at that point constructive or destructive? Justify with the ratio Δr/λ\Delta r/\lambda.

    Solution

    1.

    Δrλ=1.44×10−6480×10−9=3.0\dfrac{\Delta r}{\lambda}=\dfrac{1.44\times10^{-6}}{480\times10^{-9}}=3.0, a whole number of wavelengths. Since Δr=3λ\Delta r=3\lambda, the two waves arrive in phase, so the interference is constructive (this is the m=3m=3 maximum).

  2. 2. (2 pts ••) In a Young's double-slit experiment the slits are separated by d=0.20 mmd=0.20\,\text{mm} and the screen is L=1.8 mL=1.8\,\text{m} away. Monochromatic light of wavelength λ=550 nm\lambda=550\,\text{nm} is used.

    (a) Find the spacing between adjacent bright fringes on the screen.

    (b) How far from the central maximum is the tenth-order (m=10m=10) bright fringe?

    Solution

    2.

    (a) Fringe spacing Δy=λLd=(550×10−9)(1.8)0.20×10−3=4.95×10−3 m≈5.0 mm\Delta y=\dfrac{\lambda L}{d}=\dfrac{(550\times10^{-9})(1.8)}{0.20\times10^{-3}}=4.95\times10^{-3}\,\text{m}\approx 5.0\,\text{mm}.

    (b) The mm-th bright fringe is at ym=mλLd=m Δyy_m=m\dfrac{\lambda L}{d}=m\,\Delta y, so y10=10(4.95 mm)≈50 mm=5.0 cmy_{10}=10(4.95\,\text{mm})\approx 50\,\text{mm}=5.0\,\text{cm}.

  3. 3. (2 pts ••) Coherent light of wavelength λ=633 nm\lambda=633\,\text{nm} passes through two slits separated by d=0.12 mmd=0.12\,\text{mm}.

    (a) At what angle θ\theta (measured from the central axis) does the second-order bright fringe appear?

    (b) If the slit separation were instead d=2.5 μmd=2.5\,\mu\text{m}, how many bright fringes in total would appear on a distant screen? (Count all orders that satisfy ∣sin⁡θ∣≤1|\sin\theta|\le 1.)

    Solution

    3.

    (a) Bright fringes: dsin⁡θ=mλd\sin\theta=m\lambda. For m=2m=2: sin⁡θ=2λd=2(633×10−9)0.12×10−3=1.055×10−2\sin\theta=\dfrac{2\lambda}{d}=\dfrac{2(633\times10^{-9})}{0.12\times10^{-3}}=1.055\times10^{-2}, so θ=sin⁡−1(0.01055)≈0.60∘\theta=\sin^{-1}(0.01055)\approx 0.60^\circ.

    (b) The highest order satisfies sin⁡θ≤1\sin\theta\le1, i.e. mmax⁡=⌊d/λ⌋m_{\max}=\big\lfloor d/\lambda\big\rfloor. Here dλ=2.5×10−6633×10−9=3.95\dfrac{d}{\lambda}=\dfrac{2.5\times10^{-6}}{633\times10^{-9}}=3.95, so mmax⁡=3m_{\max}=3. Orders m=0,±1,±2,±3m=0,\pm1,\pm2,\pm3 appear: total =2(3)+1=7=2(3)+1=\mathbf{7} bright fringes.

  4. 4. (2 pts ••) Two identical coherent sources illuminate a screen; at the center of the pattern (zero path difference) the combined intensity is I0I_0. At a point PP their waves arrive with a path difference Δr\Delta r, giving a combined intensity I=I0cos⁡2(ϕ/2)I=I_0\cos^2(\phi/2), where ϕ\phi is the phase difference.

    (a) Find the phase difference ϕ\phi (in radians and degrees) when Δr=0.10 λ\Delta r = 0.10\,\lambda, and evaluate I/I0I/I_0 there.

    (b) For what smallest nonzero path difference (as a fraction of λ\lambda) does the intensity fall to one-quarter of its central-maximum value?

    Solution

    4.

    (a) Phase difference ϕ=2πλΔr=2π(0.10)=0.63 rad=36∘\phi=\dfrac{2\pi}{\lambda}\Delta r=2\pi(0.10)=0.63\,\text{rad}=36^\circ. Then I/I0=cos⁡2(ϕ/2)=cos⁡2(18∘)=0.90I/I_0=\cos^2(\phi/2)=\cos^2(18^\circ)=0.90, so I≈0.90 I0I\approx0.90\,I_0.

    (b) Want cos⁡2(ϕ/2)=14⇒cos⁡(ϕ/2)=12⇒ϕ/2=60∘⇒ϕ=120∘=2π3\cos^2(\phi/2)=\tfrac14\Rightarrow\cos(\phi/2)=\tfrac12\Rightarrow \phi/2=60^\circ\Rightarrow\phi=120^\circ=\dfrac{2\pi}{3}. Since ϕ=2π Δr/λ\phi=2\pi\,\Delta r/\lambda, this is Δr=ϕ2πλ=λ3\Delta r=\dfrac{\phi}{2\pi}\lambda=\dfrac{\lambda}{3}.

  5. 5. (2 pts ••) A soap film (n=1.33n=1.33) of uniform thickness t=320 nmt=320\,\text{nm} is surrounded by air and illuminated at normal incidence with white light.

    (a) Explain why light reflected from the front surface undergoes a half-cycle (π\pi) phase shift while light reflected from the back surface does not.

    (b) Which visible wavelength (400400–700 nm700\,\text{nm}) is most strongly reflected? Identify its color.

    Solution

    5.

    (a) Reflection at a boundary where the light goes into a medium of higher refractive index produces a half-cycle (π\pi) phase shift. At the front surface light goes air (n=1n=1) → film (n=1.33n=1.33), higher index, so a π\pi shift occurs. At the back surface light goes film (n=1.33n=1.33) → air (n=1n=1), lower index, so there is no shift.

    (b) With one π\pi shift, the extra half-wavelength must be compensated for the two reflected rays to end up in phase (constructive/strong reflection): 2nt=(m+12)λ2nt=(m+\tfrac12)\lambda, i.e. λ=2ntm+12\lambda=\dfrac{2nt}{m+\frac12}. With 2nt=2(1.33)(320 nm)=851 nm2nt=2(1.33)(320\,\text{nm})=851\,\text{nm}: m=0: 1702 nmm=0:\ 1702\,\text{nm} (IR); m=1: 567 nmm=1:\ 567\,\text{nm}; m=2: 340 nmm=2:\ 340\,\text{nm} (UV). Only m=1m=1 lands in the visible: λ≈5.7×102 nm\lambda\approx 5.7\times10^{2}\,\text{nm} — green.

  6. 6. (2 pts ••) A lens is to be coated with a single layer of magnesium fluoride (n=1.38n=1.38) to minimize reflection of λ=550 nm\lambda=550\,\text{nm} light at normal incidence. The lens glass has n=1.52n=1.52.

    (a) Both reflections (air→coating and coating→glass) occur at a surface of higher index, so each reflected ray gets a π\pi phase shift. Explain whether these two shifts add or cancel, and use that to write the correct condition on 2nt2nt for destructive interference (a dark, non-reflecting result) of the reflected light.

    (b) Find the minimum (nonzero) coating thickness.

    Solution

    6.

    (a) Because each reflected ray suffers a π\pi shift, the two shifts are equal and cancel — they contribute no net relative phase. The only relative phase then comes from the extra path 2nt2nt traveled in the film. For the reflected light to interfere destructively (minimum reflection), that path must be a half-integer number of wavelengths:

    2nt=(m+12)λ,m=0,1,2,…2nt=\left(m+\tfrac12\right)\lambda,\qquad m=0,1,2,\dots

    (b) Minimum nonzero thickness is m=0m=0: t=λ4n=550 nm4(1.38)≈1.0×102 nmt=\dfrac{\lambda}{4n}=\dfrac{550\,\text{nm}}{4(1.38)}\approx 1.0\times10^{2}\,\text{nm} (about 100 nm100\,\text{nm}).

  7. 7. (3 pts •••) In a Newton's-rings apparatus a plano-convex lens of radius of curvature R=10.0 mR=10.0\,\text{m} rests on a flat glass plate, and the gap is illuminated from above with sodium light, λ=589 nm\lambda=589\,\text{nm}. Viewed in reflection, a pattern of concentric dark rings appears.

    (a) By considering the two reflected rays from the top and bottom of the air gap, show that the dark rings occur where the gap thickness satisfies 2t=mλ2t=m\lambda (m=0,1,2,…m=0,1,2,\dots), and hence that the radius of the mm-th dark ring is rm=mλRr_m=\sqrt{m\lambda R}. (Use t≈r2/2Rt \approx r^2/2R for the thin gap.)

    (b) Compute the radius of the fifth dark ring (m=5m=5).

    Solution

    7.

    (a) The top reflection (air→glass lens surface, but the ray of interest reflects off the bottom of the air gap: air→flat glass plate, higher index) gets a π\pi shift; the reflection off the top of the air gap (glass lens → air, lower index) gets none. So there is a net half-wavelength of relative phase from the reflections alone. The path difference from crossing the gap twice is 2t2t. Combining, the reflected rays interfere destructively (dark) when

    2t=mλ,m=0,1,2,…2t=m\lambda,\qquad m=0,1,2,\dots

    (the extra π\pi from reflection already supplies the "half," so the path term must be an integer number of wavelengths). Geometry of a sphere of radius RR over a flat plate gives t≈r22Rt\approx \dfrac{r^2}{2R} for r≪Rr\ll R. Substituting:

    2⋅r22R=mλ ⇒ rm=mλR.2\cdot\frac{r^2}{2R}=m\lambda\ \Rightarrow\ r_m=\sqrt{m\lambda R}.

    (b) r5=5(589×10−9)(10.0)=2.945×10−5 m=5.43×10−3 m≈5.4 mmr_5=\sqrt{5(589\times10^{-9})(10.0)}=\sqrt{2.945\times10^{-5}}\,\text{m}=5.43\times10^{-3}\,\text{m}\approx 5.4\,\text{mm}.

  8. Extra Credit (3 pts •••)

    Two slits act as identical coherent sources. Each alone produces at a point PP on a distant screen a field of amplitude E0E_0, i.e. E1(t)=E0cos⁡ωtE_1(t)=E_0\cos\omega t and E2(t)=E0cos⁡(ωt+ϕ)E_2(t)=E_0\cos(\omega t+\phi), where ϕ\phi is the phase difference from the path-length difference between the two slits and PP.

    (a) By adding the two waves — either with the trigonometric sum-to-product identity or by adding phasors (vectors of length E0E_0 separated by angle ϕ\phi) — show that the resultant field is a single cosine of amplitude

    EP=2E0cos⁡ ⁣(ϕ2).E_P = 2E_0\cos\!\left(\frac{\phi}{2}\right).

    (b) Intensity is proportional to the square of the field amplitude. Using the fact that a single source alone gives I0∝E02I_0 \propto E_0^2, derive

    I=I0cos⁡2 ⁣(ϕ2),I = I_0\cos^2\!\left(\frac{\phi}{2}\right),

    and identify the value of II at the maxima in terms of I0I_0. Explain in one sentence why this is not a violation of energy conservation.

    (c) Relate ϕ\phi to the path difference Δr\Delta r and, for small angles in a double-slit geometry (Δr=dsin⁡θ≈d y/L\Delta r = d\sin\theta \approx d\,y/L), show that this intensity formula reproduces bright fringes at ym=mλL/dy_m=m\lambda L/d.

    Solution

    Extra Credit

    (a) Add the two fields:

    EP=E1+E2=E0cos⁡ωt+E0cos⁡(ωt+ϕ).E_P=E_1+E_2=E_0\cos\omega t+E_0\cos(\omega t+\phi).

    Using cos⁡A+cos⁡B=2cos⁡ ⁣A+B2cos⁡ ⁣A−B2\cos A+\cos B=2\cos\!\frac{A+B}{2}\cos\!\frac{A-B}{2} with A=ωtA=\omega t, B=ωt+ϕB=\omega t+\phi:

    EP=2E0cos⁡ ⁣(ϕ2)cos⁡ ⁣(ωt+ϕ2).E_P=2E_0\cos\!\left(\frac{\phi}{2}\right)\cos\!\left(\omega t+\frac{\phi}{2}\right).

    The time-varying factor cos⁡(ωt+ϕ/2)\cos(\omega t+\phi/2) is a single oscillation at the original frequency; its amplitude is

    EP(amp)=2E0cos⁡ ⁣(ϕ2).E_P^{\text{(amp)}}=2E_0\cos\!\left(\frac{\phi}{2}\right).

    (Phasor picture: two vectors of length E0E_0 with an angle ϕ\phi between them form an isosceles triangle; the resultant bisects the angle and has length 2E0cos⁡(ϕ/2)2E_0\cos(\phi/2) — the same result.)

    (b) Intensity ∝\propto (amplitude)2^2. Write I=C [2E0cos⁡(ϕ/2)]2=4CE02cos⁡2(ϕ/2)I=C\,\big[2E_0\cos(\phi/2)\big]^2=4CE_0^2\cos^2(\phi/2). A single source alone gives amplitude E0E_0 and intensity I0=CE02I_0=CE_0^2. Therefore

    I=4I0cos⁡2 ⁣(ϕ2).I=4I_0\cos^2\!\left(\frac{\phi}{2}\right).

    Writing Imax⁡≡4I0I_{\max}\equiv 4I_0 (the value at ϕ=0\phi=0, where the amplitudes fully add), this is I=Imax⁡cos⁡2(ϕ/2)I=I_{\max}\cos^2(\phi/2), the form quoted in the problems (with I0I_0 there meaning the maximum two-source intensity). At the maxima I=4I0I=4I_0 — four times what one slit gives, not twice.

    Energy is still conserved: the extra intensity at the bright fringes is exactly balanced by the zero intensity at the dark fringes. Averaging cos⁡2(ϕ/2)\cos^2(\phi/2) over the screen gives ⟨I⟩=12(4I0)=2I0\langle I\rangle=\tfrac12(4I_0)=2I_0, the sum of the two independent sources' powers — interference merely redistributes the same total energy.

    (c) A path difference Δr\Delta r corresponds to phase ϕ=2πλΔr\phi=\dfrac{2\pi}{\lambda}\Delta r. Maxima occur when cos⁡2(ϕ/2)=1\cos^2(\phi/2)=1, i.e. ϕ=2πm\phi=2\pi m, giving Δr=mλ\Delta r=m\lambda. In the double-slit geometry Δr=dsin⁡θ≈d y/L\Delta r=d\sin\theta\approx d\,y/L for small angles, so

    d yL=mλ ⇒ ym=mλLd,\frac{d\,y}{L}=m\lambda\ \Rightarrow\ y_m=\frac{m\lambda L}{d},

    the standard bright-fringe positions.