Semiconductor characterization is the process of measuring a material's electrical properties—like current-voltage (I-V) response and sheet resistance—to determine its conductivity type, carrier mobility, and junction behavior. Whether you are analyzing epitaxial wafers at a high-throughput semiconductor lab in Mohali under the India Semiconductor Mission, or probing discrete MOSFETs on your home workbench, understanding these fundamental parameters dictates whether your circuit will switch cleanly or melt under load.
In this guide, we break down the core physics of sheet resistance versus bulk resistivity, walk through a benchtop four-point probe calculation, and explain exactly how these lab measurements translate to real-world PCB and power electronics design.
Core Parameters: Sheet Resistance vs. Bulk Resistivity
Before you can interpret an I-V curve, you must understand the physical geometry of the material you are testing. The most common point of failure for junior engineers and hobbyists is confusing bulk resistivity with sheet resistance.
Bulk resistivity (ρ) is an intrinsic material property, measured in ohm-centimeters (Ω·cm). It tells you how strongly a 3D block of material opposes current flow, regardless of its shape. Sheet resistance (Rs or R□), measured in ohms per square (Ω/□), applies specifically to thin films and 2D layers (like the doped channel in a MOSFET or the copper cladding on a PCB). It tells you the resistance between two opposite edges of any perfectly square piece of that film, regardless of the square's physical size.
Below is a reference table of common semiconductor materials you will encounter in modern testing facilities, detailing their intrinsic properties and typical thin-film behaviors.
| Material | Bandgap (eV @ 300K) | Electron Mobility (cm²/V·s) | Typical Bulk Resistivity (Ω·cm) | Sheet Resistance (1 μm film) |
|---|---|---|---|---|
| Silicon (Si) | 1.12 | ~1,400 | 0.001 - 10 (doped) | 10 - 100 kΩ/□ |
| Gallium Arsenide (GaAs) | 1.42 | ~8,500 | 0.01 - 100 | 100 - 1 kΩ/□ |
| Silicon Carbide (4H-SiC) | 3.26 | ~900 | 0.01 - 0.1 | 1 - 10 kΩ/□ |
| Gallium Nitride (GaN) | 3.39 | ~1,500 (2DEG) | N/A (heterostructure) | 300 - 800 Ω/□ (2DEG) |
Data sourced from standard All About Circuits semiconductor references and typical epitaxial layer specifications.
The Math: A Worked Four-Point Probe Calculation
If you walk into a semiconductor lab in Mohali—or any advanced characterization facility—you will see parameter analyzers like the Keysight B1500A or Keithley 4200A-SCS hooked up to probe stations. When measuring thin films, engineers use a four-point probe rather than a standard multimeter. Why? Because a standard 2-wire multimeter measures the resistance of the probes and the contact interface along with the material. The four-point probe forces current through the outer two pins and measures voltage across the inner two pins, effectively eliminating contact resistance from the equation.
For a standard inline four-point probe with equal spacing (s) between the needles, and assuming the sample width and length are much larger than the probe spacing (an 'infinite sheet' approximation), the sheet resistance formula is:
Rs = (π / ln 2) × (V / I) ≈ 4.532 × (V / I)
Worked Numeric Example
Suppose you are testing a doped polysilicon layer on a test wafer. You set your parameter analyzer to source a precise current (I) of 10.0 mA (0.01 A) through the outer probes. You measure a voltage drop (V) of 22.66 mV (0.02266 V) across the inner probes.
- Identify the variables: I = 0.01 A, V = 0.02266 V, Geometric correction factor = 4.532.
- Calculate the V/I ratio: 0.02266 V / 0.01 A = 2.266 Ω.
- Apply the geometric factor: Rs = 4.532 × 2.266 Ω.
- Final Result: Rs = 10.27 Ω/□.
This tells you that any square patch of this specific polysilicon film—whether it is 1 mm x 1 mm or 10 cm x 10 cm—will have a resistance of 10.27 ohms between opposite edges. If you need to calculate the actual resistance of a rectangular trace, you simply multiply Rs by the number of 'squares' (Length / Width) in that trace.
Where You Meet This in Practice (and Common Confusions)
Lab measurements do not stay in the lab. The parameters characterized on a probe station directly dictate the physical limitations of your final circuit.
What It Changes in a Real Circuit
- MOSFET RDS(on): The on-resistance of a power MOSFET is heavily dependent on the sheet resistance of its drift region and channel. If a lab characterizes a SiC wafer and finds the sheet resistance is 20% higher than spec due to doping variations, the final RDS(on) will increase, leading to higher I²R conduction losses and potential thermal runaway in your motor drive.
- PCB Copper Weight: Standard 1 oz/ft² copper cladding is about 35 μm thick and has a sheet resistance of roughly 0.5 mΩ/□. If you are designing a high-current 48V solar inverter and use 1 oz copper for a 50A trace that is 50 'squares' long, your trace resistance is 25 mΩ, dissipating 62.5W of heat. Upgrading to 2 oz copper (70 μm) halves the sheet resistance to 0.25 mΩ/□, cutting the heat to 31.25W.
- Solar Cell Busbars: In photovoltaics, the sheet resistance of the transparent conductive oxide (TCO) layer dictates how far apart you can space the silver busbars. Higher sheet resistance requires more busbars to prevent voltage drop, which in turn shadows more of the active silicon area.
Common Confusions to Avoid
The most frequent mistake hobbyists and junior engineers make is confusing the linear (ohmic) region of an I-V curve with the saturation region. When sweeping voltage across a MOSFET, the initial straight-line portion of the graph is the linear region, where the device acts like a simple resistor governed by the sheet resistance of the channel. As VDS increases, the channel 'pinches off' at the drain end, and the current flattens out (saturation). Designing a switch based on saturation-region math instead of linear-region RDS(on) will result in massive calculation errors.
Another trap is assuming a low bulk resistivity guarantees a low-resistance connection. You can have highly doped silicon (low bulk resistivity), but if the metal-to-semiconductor interface forms a Schottky barrier instead of an ohmic contact, your 2-wire multimeter will read massive resistance due to the contact junction, not the material itself.
FAQ: Semiconductor Lab Testing Nuances
A: A micro-ohmmeter still uses 2 wires (or 4 wires connected to 2 probes). The contact resistance between a metal probe and a semiconductor surface can range from ohms to kilo-ohms depending on surface oxidation and work function. The 4-point probe physically separates the current-forcing path from the voltage-sensing path, ensuring the voltmeter draws near-zero current and measures only the semiconductor's true voltage drop.
A: It depends on the material. For standard metals (like PCB copper), resistance increases linearly with temperature (positive temperature coefficient). For intrinsic semiconductors like pure silicon, heating the material generates more electron-hole pairs, causing resistance to drop drastically (negative temperature coefficient). This is why power MOSFETs exhibit a positive temperature coefficient for RDS(on)—it is a designed trade-off to allow for easy paralleling of devices.
A: For serious lab work, engineers use Source Measurement Units (SMUs). The Keysight Semiconductor Parameter Analyzers and Tektronix/Keithley SMUs are the industry standard. They can source picoamps of current while measuring microvolts, which is necessary for characterizing reverse-biased diode leakage and sub-threshold MOSFET behavior.
Whether you are setting up a university research station or just trying to understand the datasheet parameters of a new GaN FET for your next power supply build, mastering the relationship between physical geometry, sheet resistance, and I-V behavior is the bridge between abstract physics and working hardware.
