How to Learn Electron Beam-Induced Current (EBIC): From Electron–Hole Generation and Junction Collection to Diffusion Length, Defect Imaging and Ultrafast Semiconductor Metrology

## Wait, What? An SEM Can Measure Electrical Collection Efficiency Without Putting a Probe on Every Pixel A focused electron beam creates electron–hole pairs inside a semiconductor. If those carriers are generated near a junction or electric field, some are separated and collected as current through external contacts. Scan the beam and record the current. > **EBIC is not a direct image of doping or defect density. It is a map of how beam-generated carriers are created, transported, recombined and collected by the device electrostatics.** ## The One-Sentence Answer **Learn EBIC by tracing electron beam → carrier generation → diffusion/drift → recombination → junction collection → current map, then add beam-energy generation volume, surface recombination, high injection, junction geometry and contact resistance before converting contrast into diffusion length, depletion width or defect activity.** # Beginner Layer — Beam-Generated Carriers ## Stage 1: The SEM Beam Deposits Energy Energetic electrons create many electron–hole pairs. ## Stage 2: The Generation Volume Extends Below the Surface Its size depends strongly on beam energy and material density. ## Stage 3: Carriers Move by Diffusion and Drift Built-in fields can separate them. ## Stage 4: External Contacts Collect the Surviving Charge That collected current is EBIC. # Junction Layer ## Stage 5: p–n Junctions Have Built-In Electric Fields Carriers generated near the depletion region can be collected efficiently. ## Stage 6: EBIC Peaks Near Collection-Active Regions ## Stage 7: A Bright EBIC Region Is Not Simply “more carriers” Generation and collection both matter. # Diffusion-Length Layer ## Stage 8: Carriers Generated Outside the Depletion Region Can Diffuse Toward It ## Stage 9: EBIC Decay With Distance Can Constrain Minority-Carrier Diffusion Length ## Stage 10: Geometry and Surface Recombination Affect the Decay A simple exponential is not universal. # Beam-Energy Layer ## Stage 11: Higher Beam Energy Generates Carriers Deeper and Over a Larger Volume ## Stage 12: Spatial Resolution Can Worsen as Excitation Volume Grows ## Stage 13: Beam-Energy Series Add Depth Information But depth sensitivity is broad rather than layer exact. # Defects and Recombination ## Stage 14: Grain Boundaries and Dislocations Can Reduce EBIC They act as recombination centres. ## Stage 15: Dark Contrast Is Not Automatically a Structural Defect Local doping, field direction, contact resistance or surface recombination can also reduce collection. ## Stage 16: Correlative SEM/TEM/EBSD Improves Identification # Surface Recombination ## Stage 17: Surfaces Can Remove Minority Carriers Before Junction Collection ## Stage 18: Passivation Changes EBIC Without Changing Bulk Lifetime ## Stage 19: Surface and bulk recombination must be separated in thin specimens # High-Injection and Beam Damage ## Stage 20: A Strong Beam Creates a High Carrier Density The device can leave the small-signal regime. ## Stage 21: High Injection Can Screen Built-In Fields ## Stage 22: Beam Damage and Charging Can Evolve During Imaging Dose dependence belongs in the evidence. # Quantitative Collection Efficiency ## Stage 23: Absolute EBIC Requires Current Calibration ## Stage 24: Beam current and pair-creation energy set the generation rate ## Stage 25: Collection efficiency is inferred from collected current relative to generated charge # Solar Cells and Power Devices ## Stage 26: EBIC Maps Junction Uniformity in Photovoltaics ## Stage 27: Local recombination at grain boundaries can be linked to device loss ## Stage 28: Wide-bandgap power devices use EBIC to localize depletion and defect activity # Cross-Sectional EBIC ## Stage 29: Cleave or prepare a cross section ## Stage 30: Scan across the junction in depth Depletion width, diffusion and recombination become spatially separable. # Frequency and Time-Resolved EBIC ## Stage 31: Modulated beam methods add a timescale dimension ## Stage 32: Ultrafast EBIC pushes into picosecond carrier transport Recent work combines pulsed electron beams and high-bandwidth detection. ## Stage 33: Fast response can show sign reversals or multiple collection pathways A transient sign does not automatically mean carrier type changed. # STEM-EBIC Frontier ## Stage 34: EBIC Can Be Performed in a Transmission Electron Microscope The excitation volume becomes far smaller. ## Stage 35: Nanoscale STEM-EBIC can correlate electrical collection with atomic-scale structure ## Stage 36: Thin-foil surfaces and contacts become part of the device physics # 2026 Frontier ## Stage 37: Current EBIC work is becoming more quantitative and time resolved Recent studies address ultrafast collection, contrast reversal and calibrated STEM-EBIC in nanoscale junctions. # Machine-Learning Layer ## Stage 38: ML can segment electrically active defects ## Stage 39: It can also learn SEM topography or beam-current drift ## Stage 40: Raw current plus beam and bias metadata remain primary evidence # Professional Layer ## Stage 41: Separate Five Objects 1. true semiconductor state; 2. electron-beam generation volume; 3. carrier transport/recombination; 4. junction/contact collection; 5. measured EBIC map. ## Stage 42: Professional EBIC Is a Generation–Transport–Collection Inverse Problem > **Which diffusion length, recombination centre or junction field remains identifiable after beam-energy generation volume, surface recombination, high injection, contact resistance and device geometry are all allowed to explain the same EBIC contrast?** # Evidence: What Makes an EBIC Claim Strong? Strong evidence combines calibrated beam current, beam-energy series, bias series, current linearity tests, surface-passivation controls, cross-sectional geometry, repeat doses and PL/DLTS/transport/TEM/EBSD comparison. # Misconceptions Worth Hunting – EBIC directly images doping concentration. – A bright region has more carriers. – Dark EBIC contrast uniquely proves a defect. – Beam energy only changes signal strength. – Surface recombination is irrelevant in bulk devices. – Higher beam current only improves signal-to-noise. – EBIC diffusion length always follows one exponential. – STEM-EBIC automatically gives atomic-resolution electrical truth. # Transfer Check A grain boundary becomes much brighter after surface passivation. Did the grain-boundary crystal structure change? **Not necessarily. Surface recombination may have been suppressed.** EBIC width increases strongly with beam energy. Did depletion width grow? **Not necessarily. The generation volume expanded.** # Model Limits EBIC requires electrically contacted semiconductor structures that can collect beam-generated carriers. It is an electrical collection map, not a universal defect microscope. Professional EBIC keeps **beam energy/current + generation volume + surface recombination + junction electrostatics + bias + contact resistance + injection regime + dose + current calibration + orthogonal semiconductor evidence** visible together. # Teaching Guide Teach in this order: **electron beam → carrier generation → diffusion/drift → junction collection → EBIC map → diffusion length → surface recombination → beam energy → high injection → defects → cross-sectional EBIC → ultrafast/STEM-EBIC → validation.** # Connect This to the eduKate Learning Estate – Semiconductors and Transistors — device physics. – DLTS — deep electrically active traps. – KPFM — electrostatic potential/band bending. – EBSD/TEM — structural defects and orientation. – Photoluminescence — radiative recombination mapping. # The Quiet Ending The beginner asks, “Where did the beam create carriers?” The developing device scientist asks, “Which carriers reached the junction before recombining?” The advanced learner asks, “How much of the contrast belongs to diffusion, surface loss or device field?” And the professional asks: > **Which electrically active structure survives after excitation volume, carrier transport and junction collection are all treated as one measurement chain?**