A scanning electron microscope rasters a focused beam across the sample and images the flux of low-energy secondary electrons (SEs) it knocks loose. SE yield δ rises sharply at glancing incidence — near an edge or a steep facet the beam's path length inside the escape depth grows roughly as 1/cos θ, so edges and ridges glow brighter even though nothing about the material changed. That edge-brightening is real physics, not an artifact: it reports genuine topography.
A separate effect is not: on a poorly conductive sample, incident electrons that aren't re-emitted accumulate as a negative surface charge. That charge locally deflects the primary beam and modulates SE escape, producing bright/dark charging streaks that track the raster direction and drift or "bloom" over successive frames — they look like structure but are an instrumental artifact of insufficient conductive coating or grounding.
δ(θ) ≈ δ0 / cos θ (edge brightening — real)
Q(t+dt) = Q(t) + dose·(1−cond)·dt (charge buildup — artifact source)
streak_brightness ∝ artifactStrength · Q · f(scan_phase)
- Beam energy — higher keV drives deeper, more energetic charging on insulators; very low energies approach the "E2 crossover" where emitted and incident current balance and charging nearly vanishes.
- Sample conductivity — a grounded, conductive (or metal-coated) sample bleeds charge away as fast as it arrives; an insulator (ceramic, polymer, biological tissue) cannot.
- Charging-artifact strength — a demonstration gain on the streaking term, standing in for coating thickness / grounding quality in a real instrument.
Real-world relevance: "anomalous microstructures" reported from uncoated or poorly grounded SEM samples are a textbook false positive — the fix is a conductive coating, lower beam energy, or variable-pressure/low-vacuum imaging, not a new material phase.