Home / Courses / CSIR NET Physics / Practice / Dielectric and optical properties of solids

Dielectric and optical properties of solids

Solid state physics and devices · 77 questions and templates for CSIR NET Physics, in order from warm-up to top rank. Check each answer, then read the full solution.

New to this topic? Read the lesson first: Dielectric and optical properties of solids.

Loading questions… (they need JavaScript; nothing you type is sent anywhere)

Written questions with solutions

The same questions as a list, so you can print them or read without the practice box. Templates (fresh numbers) appear only in the practice box above.

  1. MCQWarm-up

    In a cubic crystal (or an isotropic medium) with macroscopic field E→ and polarisation P→, the Lorentz local field acting on an atom is
    1. E→
    2. E→+P→/ε0
    3. E→+P→/3ε0
    4. E→−P→/3ε0
    Answer and solution

    Answer: C

    Cut a small sphere round the atom. The rest of the medium acts through the macroscopic field plus the field of the charges on the sphere's surface, +P→/3ε0. In a cubic arrangement the dipoles inside the sphere give zero net field. So E→loc=E→+P→/3ε0.

    Why people get this wrong: −P→/3ε0 is the depolarising field inside a uniformly polarised sphere; the cavity field has the opposite sign.

  2. MCQWarm-up

    Which polarisation mechanism still responds to visible light (about 5×1014 Hz)?
    1. Orientational (permanent dipoles turning)
    2. Ionic (ions moving against each other)
    3. Electronic (electron clouds shifting against nuclei)
    4. Space charge (carriers piling up at interfaces)
    Answer and solution

    Answer: C

    Each mechanism can follow the field only below its own natural frequency. Orientation of molecules relaxes at about 10⁹ to 10¹¹ Hz (microwaves), ions vibrate at 10¹² to 10¹³ Hz (infrared), and electron clouds resonate in the ultraviolet (10¹⁵ to 10¹⁶ Hz). At optical frequencies only the electronic part is left, so ε(optical) = n².

    Why people get this wrong: Ionic polarisation dies out above the infrared phonon frequencies, well below visible light.

  3. MCQWarm-up

    For light of angular frequency ω above the plasma frequency ωp, a free-electron metal
    1. reflects almost all the light
    2. becomes transparent
    3. absorbs all the light within one wavelength
    4. becomes a superconductor
    Answer and solution

    Answer: B

    With negligible damping, ε(ω)=1−ωp2/ω2. Below ωp it is negative, n is imaginary and the wave is reflected. Above ωp, ε>0, n is real and the wave propagates: alkali metals become transparent in the ultraviolet.

    Why people get this wrong: Metals reflect visible light because visible frequencies lie below ωp, not because metals always reflect.

  4. MCQWarm-up

    The plasma frequency of a free-electron gas depends on the electron density n as
    1. n
    2. n
    3. n2
    4. n1/3
    Answer and solution

    Answer: B

    ωp2=ne2/ε0m: a slab of electrons displaced by x leaves surface charges that give a field nex/ε0 and a restoring force ne2x/ε0 per electron. So ωp∝n.

    Why people get this wrong: n1/3 is how the Fermi wave vector scales, not the plasma frequency.

  5. NATWarm-up

    A material has extinction coefficient κ=0.010 at a wavelength of 500 nm (in vacuum). Its intensity absorption coefficient α is ______ ×105 m−1 (two decimal places).
    Answer and solution

    Answer: 2.51 × 10⁵ m⁻¹ (±0.01)

    The field decays as e−κωz/c, so the intensity decays as e−2κωz/c: α=2κω/c=4πκ/λ=4π×0.010/(5.0×10−7)=2.51×105 m−1. The light falls to 1/e in about 4 μm.

    Why people get this wrong: Forgetting the factor 2 between field and intensity.

  6. MCQWarm-up

    Water has a static dielectric constant near 80, yet its refractive index for visible light is only 1.33. The reason is that
    1. the refractive index is measured at a different temperature
    2. water is magnetic at optical frequencies
    3. the electronic polarisability of water is negative
    4. the permanent dipoles of water molecules cannot turn fast enough to follow optical fields
    Answer and solution

    Answer: D

    n2=εr(ω) at the frequency of the light. The huge static value comes from orientational polarisation of the permanent dipoles, which relaxes at about 2×1010 Hz. At 1015 Hz only electronic polarisation remains, giving εr≈1.77=1.332.

    Why people get this wrong: Maxwell's relation n=εr holds, but with εr taken at the same frequency.

  7. MCQWarm-up

    A transparent, non-magnetic crystal has relative permittivity εr=2.25 at the frequency of visible light. Its refractive index is
    1. 2.25
    2. 1.50
    3. 5.06
    4. 1.12
    Answer and solution

    Answer: B

    For a non-magnetic medium (μr=1), n=εrμr=2.25=1.50. The εr here must be the value at optical frequency; static values can be much larger (for water, 80 against n2=1.78).

    Why people get this wrong: n is the square root of εr, not equal to it.

  8. NATStandard

    For a cubic dielectric, Nα/3ε0=0.20, where N is the number density of atoms and α their polarisability. Using the Clausius-Mossotti relation, its dielectric constant is ______ (two decimal places).
    Answer and solution

    Answer: 1.75 (±0.01)

    εr−1εr+2=Nα3ε0=x, so εr=1+2x1−x=1.40.8=1.75. Without the local-field correction one would get 1+3x=1.60.

    Why people get this wrong: Using εr=1+Nα/ε0 ignores the Lorentz field, valid only for dilute gases.

  9. MCQStandard

    Molecules with permanent dipole moment p in a weak field at temperature T contribute an orientational polarisability (Langevin-Debye) of
    1. p2/3kBT
    2. p2/kBT
    3. p/3kBT
    4. 3p2/kBT
    Answer and solution

    Answer: A

    The mean dipole along the field is pL(pE/kBT) with the Langevin function L(x)=coth⁡x−1/x≈x/3 for small x. So ⟨pz⟩=p2E/3kBT and αor=p2/3kBT, falling as 1/T because heat randomises the dipoles.

    Why people get this wrong: The 1/3 comes from averaging cos²θ over all directions in 3D.

  10. NATStandard

    Copper has 8.5×1028 conduction electrons per m3. Treating them as free electrons, the plasmon energy ℏωp is ______ eV (one decimal place).
    Answer and solution

    Answer: 10.8 eV (±0.1)

    ωp=ne2/ε0m=8.5×1028×(1.602×10−19)2/(8.854×10−12×9.109×10−31)=1.64×1016 rad/s. ℏωp=1.055×10−34×1.64×1016/1.602×10−19=10.8 eV. (Copper's red colour comes from d-band transitions near 2 eV, not from the free-electron edge.)

    Why people get this wrong: Dividing by 2π: ℏω uses the angular frequency.

  11. MCQStandard

    In the Drude model with collision rate γ, the dielectric function of a metal is ε(ω)=1−ωp2/(ω2+iγω). In the limit ω≫γ it becomes
    1. 1+ωp2/ω2
    2. 1−ω2/ωp2
    3. 1−ωp2/ω2
    4. ωp2/ω2
    Answer and solution

    Answer: C

    Dropping iγω against ω2 leaves ε=1−ωp2/ω2: real, negative below ωp, zero at ωp (where longitudinal plasma oscillations live) and positive above.

    Why people get this wrong: The free-electron term lowers ε; electrons respond out of phase above their (zero) natural frequency.

  12. NATStandard

    Silicon has refractive index 3.4 and negligible absorption at 1.5 μm. The fraction of light reflected at normal incidence from air is ______ % (one decimal place).
    Answer and solution

    Answer: 29.8 % (±0.1)

    R=(n−1n+1)2=(2.4/4.4)2=0.2975: about 30%, which is why silicon solar cells need anti-reflection coatings.

    Why people get this wrong: Forgetting to square (n−1)/(n+1), which gives the amplitude ratio (55%).

  13. MCQStandard

    For Debye relaxation, ε(ω)=ε∞+εs−ε∞1+iωτ (sign convention aside). The imaginary part (dielectric loss) is largest when
    1. ωτ=1
    2. ω=0
    3. ωτ=2
    4. ω→∞
    Answer and solution

    Answer: A

    ε′′=(εs−ε∞)ωτ1+ω2τ2. With x=ωτ, d/dx[x/(1+x2)]=(1−x2)/(1+x2)2=0 at x=1. The peak value is (εs−ε∞)/2, and there ε′ is halfway between εs and ε∞.

    Why people get this wrong: At very low or very high frequency there is no loss: the dipoles either follow perfectly or do not move at all.

  14. MSQStandard

    For a polar ionic solid, which statements about εr(ω) from DC to the ultraviolet are correct?
    1. εr rises in steps as the frequency increases.
    2. Each drop in ε′ comes with a peak in the loss ε′′.
    3. At visible frequencies, ε′ is close to n2.
    4. Orientational polarisation is the last mechanism to drop out.
    Answer and solution

    Answer: B, C

    As frequency rises, mechanisms drop out one by one (orientational near microwaves, ionic in the infrared, electronic in the ultraviolet), so ε′ falls in steps, each marked by absorption (Kramers-Kronig ties every change in ε′ to a loss feature). At optical frequencies only electrons respond, so ε′≈n2. Orientation is the slowest, so it is the first to go, not the last.

    Why people get this wrong: The steps go down, not up, as frequency rises.

  15. NATStandard

    In the simplest model an atom is a point nucleus inside a uniform electron cloud of radius R, which gives α=4πε0R3. For R=0.10 nm, α is ______ ×10−40 F m2 (two decimal places).
    Answer and solution

    Answer: 1.11 × 10⁻⁴⁰ F m² (±0.01)

    Shifting the nucleus by d inside a uniform sphere of charge −Ze gives a restoring field Zed/4πε0R3; balance with E: p=Zed=4πε0R3E. So α=4π×8.854×10−12×10−30=1.11×10−40 F m2.

    Why people get this wrong: Often quoted as the "polarisability volume" α/4πε0=R3; do not mix the two.

  16. MCQStandard

    Which statement about piezoelectric, pyroelectric and ferroelectric crystals is correct?
    1. Every piezoelectric crystal is ferroelectric.
    2. Every pyroelectric crystal is ferroelectric.
    3. A crystal with a centre of inversion can be piezoelectric.
    4. Every ferroelectric crystal is both pyroelectric and piezoelectric.
    Answer and solution

    Answer: D

    A ferroelectric has a spontaneous polarisation that can be reversed by a field. Spontaneous polarisation makes it pyroelectric (P changes with T), and a polar crystal lacks inversion symmetry, so it is piezoelectric. The reverse fails: quartz is piezoelectric but not pyroelectric, and tourmaline is pyroelectric but its polarisation cannot be switched.

    Why people get this wrong: The sets nest: ferroelectric ⊂ pyroelectric ⊂ piezoelectric ⊂ non-centrosymmetric.

  17. MCQStandard

    Just above the band gap Eg of a direct-gap semiconductor (allowed transitions), the absorption coefficient varies with photon energy as
    1. (ℏω−Eg)2
    2. (ℏω−Eg)1/2
    3. (ℏω−Eg)3/2
    4. e−(ℏω−Eg)/kBT
    Answer and solution

    Answer: B

    Vertical transitions connect states with the same k; the joint density of states of two parabolic bands grows as ℏω−Eg, like a 3D density of states. With a constant matrix element, α∝(ℏω−Eg)1/2.

    Why people get this wrong: (ℏω−Eg±ℏΩ)2 is the phonon-assisted indirect edge; 3/2 is the direct but forbidden case.

  18. MSQStandard

    A material has complex refractive index n~=n+iκ and is studied with light of vacuum wavelength λ. Which statements are correct?
    1. The reflectance at normal incidence from vacuum is (n−1)2+κ2(n+1)2+κ2.
    2. The intensity absorption coefficient is 4πκ/λ.
    3. ε′=n2−κ2 and ε′′=2nκ.
    4. A material with κ=0 reflects no light at all.
    Answer and solution

    Answer: A, B, C

    The Fresnel formula with complex n~ gives the first expression. The field decays as e−2πκz/λ, so the intensity as e−4πκz/λ. Squaring n~ gives ε=n2−κ2+2inκ. With κ=0 there is no absorption, but there is still reflection whenever n≠1: glass with n=1.5 reflects 4%.

    Why people get this wrong: Transparent does not mean non-reflecting; reflection comes from the index step.

  19. NATExam level

    A non-polar liquid has 2.0×1028 molecules per m3, each of polarisability 2.0×10−40 F m2. Using the Clausius-Mossotti relation, its dielectric constant is ______ (three decimal places).
    Answer and solution

    Answer: 1.532 (±0.002)

    x=Nα/3ε0=4.0×10−12/(2.656×10−11)=0.1506. εr=(1+2x)/(1−x)=1.3012/0.8494=1.532. Ignoring the local field would give 1+3x=1.452: the correction matters in dense media.

    Why people get this wrong: Using Nα/ε0 alone (εr=1.452) is the dilute-gas formula.

  20. NATExam level

    Sodium has 2.65×1028 free electrons per m3. In the free-electron model it becomes transparent for wavelengths shorter than ______ nm (nearest whole number).
    Answer and solution

    Answer: 205 nm (±2)

    ωp=ne2/ε0m=9.18×1015 rad/s, so λp=2πc/ωp=2π×3.00×108/9.18×1015=2.05×10−7 m =205 nm, close to the measured 210 nm: alkali metals are transparent in the ultraviolet.

    Why people get this wrong: Using c/ωp without the 2π gives 33 nm.

  21. MCQExam level

    An ionic crystal has dielectric function ε(ω)=ε∞ωL2−ω2ωT2−ω2 (no damping). For ωT<ω<ωL, infrared light at normal incidence is
    1. transmitted with no loss
    2. totally reflected, because ε is negative and the refractive index is purely imaginary
    3. absorbed only at exactly ω = ω_L
    4. reflected with R = ((√ε∞ − 1)/(√ε∞ + 1))²
    Answer and solution

    Answer: B

    Between the poles and zeros, numerator and denominator have opposite signs, so ε<0. Then n=iκ and r=(iκ−1)/(iκ+1) has |r|=1: total reflection. This "reststrahlen" band was used to make narrow-band infrared filters by multiple reflections.

    Why people get this wrong: At ωL, ε=0 (a longitudinal mode), not a peak in absorption; the absorption peak is at ωT.

  22. NATExam level

    GaAs has ε(0)=12.9, ε(∞)=10.9 and a transverse optical phonon at 8.00 THz. By the Lyddane-Sachs-Teller relation, the longitudinal optical phonon frequency is ______ THz (two decimal places).
    Answer and solution

    Answer: 8.7 THz (±0.01)

    ωL2/ωT2=ε(0)/ε(∞), so νL=8.0012.9/10.9=8.00×1.0879=8.70 THz. The split is small because GaAs is only weakly ionic.

    Why people get this wrong: The ratio of the dielectric constants equals the ratio of squared frequencies.

  23. MCQExam level

    The Kramers-Kronig relations connect ε′(ω) and ε′′(ω). They follow from
    1. energy conservation in the absorber
    2. time-reversal symmetry of the microscopic equations
    3. causality: the polarisation cannot respond before the field is applied
    4. the Lorentz local-field correction
    Answer and solution

    Answer: C

    Causality makes the response function χ(t) vanish for t<0, so χ(ω) is analytic in the upper half of the complex ω plane. Cauchy's theorem then gives each part as a principal-value integral of the other, e.g. ε′(ω)−1=2πP∫0∞ω′ε′′(ω′)ω′2−ω2dω′.

    Why people get this wrong: Time reversal gives the parity of each part (ε′ even, ε″ odd in ω), not the integral relations.

  24. NATExam level

    Copper has DC conductivity 5.8×107 S/m. At 1.0 GHz, where ωτ≪1, the skin depth is ______ μm (two decimal places).
    Answer and solution

    Answer: 2.09 μm (±0.01)

    δ=2/μ0σω. μ0σω=1.2566×10−6×5.8×107×6.283×109=4.58×1011, so δ=4.37×10−12=2.09×10−6 m.

    Why people get this wrong: Using f instead of ω = 2πf makes δ √(2π) times too large.

  25. NATExam level

    A polar liquid follows Debye relaxation with εs=80 and ε∞=5. The loss tangent tan⁡δ=ε′′/ε′ at the frequency where ε′′ is largest is ______ (two decimal places).
    Answer and solution

    Answer: 0.88 (±0.01)

    At ωτ=1: ε′=ε∞+(εs−ε∞)/2=42.5 and ε′′=(εs−ε∞)/2=37.5. tan⁡δ=37.5/42.5=0.88.

    Why people get this wrong: tan⁡δ is not largest at ωτ=1 (it peaks at ωτ=εs/ε∞); the question asks at the ε′′ peak.

  26. MCQExam level

    In Landau theory of a second-order ferroelectric transition, F=F0+12a(T−Tc)P2+14bP4−EP. At the same distance |T−Tc| below and above Tc, the ratio χbelow/χabove of the susceptibilities is
    1. 2
    2. 1/2
    3. 1
    4. 1/4
    Answer and solution

    Answer: B

    Above Tc: P=0, 1/χ=∂2F/∂P2=a(T−Tc). Below: P02=a(Tc−T)/b, and ∂2F/∂P2=a(T−Tc)+3bP02=2a(Tc−T). So 1/χ has twice the slope below Tc, and at equal distance χbelow/χabove=1/2.

    Why people get this wrong: The curvature below Tc includes 3bP02, not bP02.

  27. MSQExam level

    Of the 32 crystallographic point groups, which statements are correct?
    1. Every piezoelectric class is also ferroelectric.
    2. 21 lack a centre of inversion.
    3. 20 of these allow piezoelectricity.
    4. 10 are polar and allow pyroelectricity.
    Answer and solution

    Answer: B, C, D

    A centre of symmetry forbids any property described by an odd-rank tensor, such as piezoelectricity (rank 3) or a spontaneous polarisation (rank 1). Of the 21 non-centrosymmetric classes, 432 has piezoelectric coefficients that vanish by symmetry, leaving 20 piezoelectric classes; the 10 with a unique polar axis can be pyroelectric. Ferroelectricity needs, in addition, a switchable polarisation, which symmetry alone cannot decide.

    Why people get this wrong: Class 432 is non-centrosymmetric but not piezoelectric.

  28. NATExam level

    In a semiconductor with εr=12.5 and electron-hole reduced mass 0.060me, the hydrogen-like (Wannier) exciton ground-state binding energy is ______ meV (two decimal places).
    Answer and solution

    Answer: 5.22 meV (±0.02)

    Scale hydrogen: Eb=13.606 eV×μ/meεr2=13.606×0.060/156.25=5.22 meV. The Bohr radius grows by εrme/μ=208, to 11 nm, so the exciton spans many cells and the continuum model is consistent.

    Why people get this wrong: The dielectric constant enters squared in the energy (once from each factor of e2 in μe4/ε2).

  29. MCQExam level

    For a Lorentz oscillator medium, ε(ω)=1+ωp2ω02−ω2−iγω. In the narrow band around ω0 (of width about γ),
    1. the refractive index rises with frequency and absorption is negligible
    2. the medium is perfectly transparent
    3. the refractive index falls with increasing frequency (anomalous dispersion) and absorption is strong
    4. the group velocity equals c exactly
    Answer and solution

    Answer: C

    Away from resonance n rises with ω (normal dispersion). Within about γ of ω0, ε′ swings from its maximum down through its minimum as ω increases, so n falls with frequency, while ε′′ (absorption) peaks at ω0. Anomalous dispersion and absorption always go together, as Kramers-Kronig requires.

    Why people get this wrong: Normal dispersion (n rising with ω, as in glass in the visible) is the behaviour far below an ultraviolet resonance.

  30. NATExam level

    A cubic glass has refractive index 1.50 in the visible. Using the Lorentz-Lorenz form of the Clausius-Mossotti relation, Nαe/3ε0 for its electronic polarisability is ______ (three decimal places).
    Answer and solution

    Answer: 0.294 (±0.001)

    At optical frequencies εr=n2=2.25, so Nαe/3ε0=(n2−1)/(n2+2)=1.25/4.25=0.294.

    Why people get this wrong: Putting n instead of n2 into the relation.

← All topics · Timed mock test →