Question

  1. Explain why clean rooms are important for the production of semiconductor devices. What measures are used to keep a clean room clean?

Answer

Semiconductor features are nanoscale — a single particle on a wafer can bridge features, cause shorts, or block lithography patterns, killing devices and reducing yield. Measures: HEPA/ULPA filtered laminar airflow, positive pressure, airlocks and air showers, gowning (bunny suits, gloves, booties, face masks), restricted materials (no paper, pencils, cosmetics), temperature and humidity control, anti-static flooring, and strict protocols for wafer handling.

Question

  1. Why is the air pressure in a cleanroom usually higher than the outside air pressure?

Answer

Positive pressure ensures air flows outward through any openings (doors, gaps). This prevents unfiltered, particle-laden external air from entering the cleanroom when doors are opened or through any leaks.

Question

  1. What is meant by cleanroom classifications? What is the difference between a Class 10 and a Class 1000 cleanroom?

Answer

Classifications define maximum allowable particles (≥0.5µm) per cubic foot of air. Class 10: max 10 particles/ft³. Class 1000: max 1000 particles/ft³. Class 10 is 100× cleaner, used for critical lithography and processing steps. Class 1000 is suitable for less sensitive operations. ISO 14644 uses a similar scheme based on particles per m³.

Question

  1. Calculate the number of particles >10µm landing on a 1cm × 1cm solar cell in 5 min in a Class 1000 cleanroom (air stream 50m/min). Recalculate for normal room air.

Answer

From the graph, Class 1000 has ~1 particle/ft³ (≈35/m³) of size >10µm. Normal room air has ~1000 particles/ft³ (≈35,000/m³) at >10µm. Volume of air passing over the cell in 5 min: area × velocity × time = (1×10⁻⁴ m²) × (50 m/min) × (5 min) = 0.025 m³. Class 1000: 35 × 0.025 ≈ 0.9 particles (≈1 particle). Normal room: 35,000 × 0.025 ≈ 875 particles. This demonstrates why cleanrooms are essential — nearly 1000× fewer contaminants.

Question

  1. Describe the typical process steps involved in the production of a GaAs-based LED. Name and describe the key apparatus used in the fabrication process.

Answer

Steps: (1) Epitaxial growth of n- and p-type GaAs/AlGaAs layers on a GaAs substrate, (2) Photolithography to define device mesas, (3) Etching to isolate individual devices, (4) Metallisation for ohmic contacts (top and bottom), (5) Dicing and packaging. Key apparatus: MOCVD reactor (epitaxial growth of III-V layers with precise composition control), mask aligner (UV lithography for pattern transfer), wet/dry etcher (mesa formation), metal evaporator/sputterer (contact deposition), wire bonder (connecting die to package leads).

Question

  1. Why is UV light typically used in optical lithography? Why does electron beam lithography provide a higher spatial resolution? What are the advantages and disadvantages of each approach?

Answer

UV light: shorter wavelength than visible → smaller diffraction-limited features (resolution ∝ λ). DUV at 193nm enables sub-100nm features with immersion/multi-patterning. E-beam: electrons at typical accelerating voltages have de Broglie wavelengths of ~pm, far below UV, enabling sub-10nm resolution. UV advantages: parallel exposure (entire wafer at once), high throughput, low cost per wafer. Disadvantages: diffraction-limited resolution, expensive masks. E-beam advantages: highest resolution, maskless (direct write), flexible for prototyping. Disadvantages: serial process (very slow), low throughput, not viable for mass production.

Question

  1. What are the differences between a positive and negative resist?

Answer

Positive: exposed regions become soluble in developer (chain scission). Pattern on wafer matches the mask. Better resolution, used for fine features. Negative: exposed regions crosslink and become insoluble. Pattern on wafer is the inverse of the mask. More robust/chemically resistant but prone to swelling, limiting resolution. Generally used for less critical layers.

Question

  1. Explain the advantage of using a Phase-Shifting Mask compared with a Binary Mask.

Answer

Binary mask: chrome on glass, either blocks or transmits light. At feature edges, diffraction causes intensity spread, limiting resolution. Adjacent bright features can blur together. Phase-shifting mask: alternating transmission regions have a 180° phase shift (achieved by etching the glass to a specific depth). Where light from adjacent openings overlaps, destructive interference creates a sharp dark boundary. This significantly improves edge contrast and resolution, enabling smaller features than the diffraction limit of a binary mask would allow.

Question

  1. Describe the lift-off process for making metal contacts.

Answer

Steps: (1) Spin coat photoresist on wafer, (2) Expose and develop to create openings where contacts are needed — ideally with an undercut (re-entrant) profile, (3) Deposit metal over entire surface (evaporation preferred for directional deposition), (4) Dissolve resist in solvent — metal on the resist lifts off, metal in the openings remains bonded to the semiconductor. The undercut profile is critical: it creates a shadow during deposition ensuring a clean discontinuity between metal on the substrate and metal on the resist, enabling clean lift-off.

Question

  1. Describe the key differences between a Schottky and an Ohmic contact.

Answer

Schottky: metal-semiconductor junction with a potential barrier (Schottky barrier). Exhibits rectifying (diode-like) I-V behaviour — current flows easily in one direction. Used in Schottky diodes, gate contacts on FETs.

Ohmic: linear I-V characteristic (V = IR), no barrier to current flow in either direction. Achieved by heavily doping the semiconductor surface so carriers can tunnel through the thin barrier. Used for source/drain contacts, LED contacts — anywhere low-resistance bidirectional contact is needed.

Question

  1. What is the Damascene process and why is it used?

Answer

A metallisation technique where trenches/vias are first etched into a dielectric layer, then filled with metal (typically copper by electroplating), and excess metal is removed by chemical-mechanical polishing (CMP) to leave metal only in the trenches. Used because copper is difficult to etch by conventional plasma methods (non-volatile etch products). The Damascene process avoids the need to etch copper directly and produces planar surfaces essential for multilayer interconnects in modern ICs.

Question

  1. Why are test structures added to mask sets? What are the pros and cons? Give one example.

Answer

Purpose: monitor process parameters and quality control without sacrificing device area. Pros: enable in-line measurement of critical parameters (sheet resistance, contact resistance, line width, alignment accuracy), early detection of process drift, improved yield. Cons: consume wafer area (placed in scribe lanes to minimise this), add mask complexity, may not perfectly represent conditions in the active device area. Example: Transmission Line Model (TLM) structure — a series of metal contacts with varying spacing on a doped region. Measuring resistance vs gap length gives both contact resistance (y-intercept) and sheet resistance (gradient).

Question

  1. Describe three types of hazards in cleanrooms.

Answer

(1) Chemical: toxic, corrosive, and flammable chemicals (HF, H₂SO₄, organic solvents, toxic gases like arsine/phosphine). Require fume hoods, PPE, and gas monitoring. (2) Electrical: high-voltage equipment (plasma etchers, ion implanters, e-beam systems). Risk of electrocution and arc flash. (3) Radiation: UV exposure from lithography tools, laser radiation from alignment and characterisation systems. Require shielding, interlocks, and laser safety protocols.

Question

  1. Define “Threshold Limit Values (TLV)” and “Immediately Dangerous to Life and Health (IDLH)” in the context of chemical hazards.

Answer

TLV: the maximum airborne concentration of a substance that a worker can be exposed to repeatedly (typically 8-hour TWA) without adverse health effects. Set by occupational health bodies as safe working limits. IDLH: the concentration that poses an immediate threat to life, would cause irreversible health effects, or would impair a worker’s ability to escape. Exposure at IDLH levels requires immediate evacuation and use of self-contained breathing apparatus. IDLH values are always significantly higher than TLV values.