Questions & explanations
1. What is the main job of an erbium-doped fiber amplifier (EDFA) in a fiber optic system?
An EDFA boosts the power of light signals traveling through optical fibers without turning them into electrical signals first. It uses a short length of fiber doped with erbium ions, which are atoms that can store energy. When a weak signal and a pump laser (a strong light at a different wavelength) pass through, the erbium ions release their stored energy to amplify the signal. The gain is how much the signal power increases, usually measured in decibels (dB). However, the amplifier also adds noise, mainly from spontaneous emission, which is random light released by excited erbium ions. The noise figure tells how much the signal-to-noise ratio gets worse after amplification. A good EDFA design balances high gain with low noise figure.
2. Compare the propagation length and confinement of an SPP on a silver-air interface versus a silver-glass interface at the same wavelength.
At a given wavelength (e.g., 633 nm), the SPP on a silver-air interface has a longer propagation length but weaker confinement than on a silver-glass interface. This is because glass has a higher refractive index (≈1.5) than air (≈1), which pulls the SPP field more into the dielectric, increasing confinement (smaller decay length into the dielectric) but also increasing the field in the metal, leading to higher absorption and shorter propagation length. Typically, on silver-air, propagation length is about 20-40 μm, while on silver-glass it is about 5-10 μm. The confinement (1/e decay into the dielectric) is about 200 nm for air and 100 nm for glass. So there is a trade-off: tighter confinement comes at the cost of shorter range.
3. Compare transformation optics with conventional lens design.
Conventional lens design uses curved surfaces of homogeneous materials to bend light, based on Snell's law. It is limited to smooth variations and cannot achieve extreme control like bending light around an object. Transformation optics, on the other hand, allows arbitrary control of wave propagation by using inhomogeneous and anisotropic materials. For example, a conventional lens focuses light but cannot make an object invisible. Transformation optics can design a 'perfect lens' that beats the diffraction limit, but it requires metamaterials with negative refractive index, which are hard to make. Conventional lenses are simpler and broadband, while transformation optics devices are narrowband but offer unprecedented control.
4. How does the confinement of an SPP compare to that of light in a conventional dielectric waveguide?
An SPP can confine light to dimensions much smaller than the wavelength, down to tens of nanometers, because the field is concentrated at the metal surface. In contrast, a dielectric waveguide (like an optical fiber) confines light to regions about the size of the wavelength (micrometers). This strong confinement is the main advantage of plasmonics for miniaturization. However, the trade-off is higher loss: the metal absorbs light, so SPPs cannot travel as far as light in low-loss dielectrics. Also, the confinement is not uniform: the field decays exponentially into both materials, so some energy is still in the metal. For applications needing both small size and low loss, hybrid plasmonic-dielectric waveguides are used.
5. What is chirp in an MZI modulator and how does it affect the optical signal?
Chirp is a change in the instantaneous frequency of the light during modulation. In an MZI, chirp happens when the phase change in the two arms is not symmetric. For example, if only one arm is modulated, the output light's frequency shifts slightly as the phase changes. This frequency shift can cause the light pulse to spread out in optical fiber due to dispersion, leading to signal distortion. Chirp is measured by the chirp parameter α, which is the ratio of phase modulation to intensity modulation. Push-pull operation minimizes chirp, while single-arm drive produces positive or negative chirp. Sometimes a small chirp is intentionally added to compensate for fiber dispersion, but usually chirp is unwanted.
6. How does the pump power affect the gain and noise figure of an EDFA?
Increasing the pump power generally raises the gain because more erbium ions get excited and can amplify the signal. But if the pump is too weak, the gain is low and the noise figure is high because the signal doesn't get enough amplification relative to the spontaneous emission. At very high pump power, the gain saturates (stops increasing) because almost all erbium ions are already excited. In saturation, the noise figure can actually improve slightly because the amplifier is more efficient. However, too much pump power can also cause unwanted effects like nonlinearities. The key is to choose a pump power that gives enough gain while keeping the noise figure low, typically around 4-6 dB for a good EDFA.
7. What is the propagation length of an SPP and what factors limit it?
The propagation length is the distance an SPP travels before its intensity drops to 1/e (about 37%) of its initial value. It is limited by ohmic losses in the metal: the oscillating electrons generate heat. Metals like silver and gold have relatively low loss, giving propagation lengths of tens to hundreds of micrometers at visible and near-infrared wavelengths. The propagation length also depends on the dielectric: a higher index dielectric confines the SPP more, but also increases loss. Shorter wavelengths have higher loss because the metal's absorption increases. For practical devices, the propagation length must be long enough to travel between components, which limits the size of plasmonic circuits.
8. Explain how a ring resonator can be used as an add-drop filter.
An add-drop filter uses a ring resonator coupled to two bus waveguides: an input bus and a drop bus. At resonance, light from the input bus couples into the ring and then drops into the drop bus, while off-resonance light passes straight through the input bus. This allows selecting one wavelength from a multi-wavelength signal (drop function). Conversely, you can add a new wavelength by injecting it into the ring from the drop bus, and it will couple into the input bus (add function). The ring's resonance wavelength can be tuned by heating or applying voltage to change the refractive index. This is useful in wavelength-division multiplexing (WDM) systems for routing individual channels.
9. Explain how a prism can be used to excite an SPP on a metal film (Kretschmann configuration).
In the Kretschmann configuration, a thin metal film is deposited on a glass prism. Light from a laser shines through the prism onto the metal at an angle greater than the critical angle, creating an evanescent wave that penetrates the metal. When the wavevector of the evanescent wave matches that of the SPP on the outer metal surface, energy transfers into the SPP, causing a dip in the reflected light intensity. The angle at which this happens is very sensitive to the refractive index of the material on the other side of the metal (the analyte). This is used in SPR sensors to detect binding events, like biomolecules, because the resonance angle shifts with changes in refractive index.
10. Why does a nanoscale copper wire heat up less than a bulk copper wire when carrying the same current?
At the nanoscale, heat is carried by phonons—vibrations of the atomic lattice. In a bulk material, phonons can scatter off each other, reducing thermal conductivity. In a very thin wire, the surface scatters phonons differently, and the mean free path of phonons is limited by the wire's dimensions. This can actually increase thermal conductivity for some materials because surface scattering reduces the number of phonon-phonon collisions. However, for copper, the thermal conductivity decreases at small scales due to increased electron scattering at surfaces. The key is that nanoscale dimensions change how heat carriers move, so the wire's ability to dissipate heat depends on its size.
11. Compare a simple isotropic material with a bianisotropic one.
In an isotropic material, the electric displacement D is parallel to the electric field E, and the magnetic induction B is parallel to the magnetic field H. The material is described by just two scalar constants: permittivity and permeability. In a bianisotropic material, D depends on both E and H, and B depends on both H and E. This means the material has four tensors: permittivity, permeability, and two magnetoelectric tensors. For example, a simple isotropic material like glass has no cross-coupling, while a bianisotropic metamaterial can have strong cross-coupling, leading to exotic properties like negative refraction without needing both negative permittivity and permeability.
12. Explain how a two-stage EDFA design can achieve a lower noise figure than a single-stage design.
A two-stage EDFA splits the amplification into two separate erbium fiber coils with a filter or isolator in between. The first stage is designed for low noise: it uses a low pump power and short fiber to amplify the signal just enough to overcome the noise of the second stage. The second stage then boosts the signal to the final high power. Because the first stage adds little noise, the overall noise figure is lower than a single stage that does all the amplification at once. The isolator between stages prevents backward ASE from the second stage entering the first. This design is common in high-performance systems where low noise is critical, such as in long-haul undersea cables.