How to Learn Off-Axis Electron Holography: From Electron-Wave Phase to Electrostatic Potentials, Magnetic Fields and Operando Nanoscale Field Mapping

## Wait, What? Electron Holography Can See a Potential Even When Ordinary TEM Intensity Looks Uniform A semiconductor junction can contain an electrostatic landscape that strongly controls device behaviour while producing little conventional TEM contrast. Electron holography uses a different receiver: the **phase of the electron wave**. > **Electron holography measures projected electron-wave phase; electrostatic or magnetic fields are inferred only after thickness, reference phase, diffraction and charging are controlled.** ## The One-Sentence Answer **Learn off-axis electron holography by tracing coherent electron wave → specimen phase shift → vacuum reference wave → biprism interference fringes → reconstructed phase, then add thickness, mean inner potential, charging, diffraction and magnetic/electric separation before turning a phase map into quantitative nanoscale fields.** # Beginner Layer — Electrons Have Phase ## Stage 1: A Fast Electron Has Amplitude and Phase Conventional intensity imaging does not preserve all exit-wave phase information directly. ## Stage 2: Electrostatic Potential Changes Electron Phase ## Stage 3: Magnetic Vector Potential Also Changes Phase This connects to the Aharonov–Bohm principle. ## Stage 4: Interference With a Reference Wave Makes Phase Observable # Biprism Layer ## Stage 5: A Fine Charged Wire Acts as an Electron Biprism It bends two electron waves toward one another. ## Stage 6: One Wave Crosses the Specimen—the Object Wave ## Stage 7: One Travels Through Reference Vacuum ## Stage 8: Their Overlap Produces Interference Fringes Fringe phase contains object-wave information. # Reconstruction Layer ## Stage 9: The Hologram Contains a Central Band and Two Sidebands in Fourier Space ## Stage 10: Select One Sideband and Shift It to the Origin ## Stage 11: Inverse Transform to Recover Complex Amplitude and Phase ## Stage 12: Sideband Aperture Trades Spatial Resolution Against Noise A wider aperture includes finer detail and more noise. # Phase Unwrapping ## Stage 13: Reconstructed Phase Is Naturally Modulo 2π Large shifts wrap. ## Stage 14: Phase Unwrapping Builds a Continuous Map Noise can cause incorrect unwrap paths. ## Stage 15: Smoothness Is Not Physical Proof An algorithm can produce a beautiful wrong phase map. # Reference-Hologram Layer ## Stage 16: The Microscope Adds Its Own Phase Distortion Biprism charging, lens distortion and Fresnel fringes contribute. ## Stage 17: A Reference Hologram Helps Remove Instrument Phase It should match lens and biprism conditions closely. ## Stage 18: A Poor Reference Leaves Ramps and False Long-Range Fields # Electrostatic Phase ## Stage 19: Projected Electrostatic Potential Contributes Phase A simplified form is **φE ∝ ∫V dz**. ## Stage 20: Mean Inner Potential Contributes Strongly Ordinary material has a nonzero average electrostatic potential relative to vacuum. ## Stage 21: Thickness and Composition Therefore Affect Phase A thicker region can shift phase without any device-field change. ## Stage 22: Thickness Must Be Measured or Constrained EELS, known wedges or tomography can help. # Magnetic Phase ## Stage 23: Magnetic Vector Potential Adds Another Phase Term Phase gradients can be related to projected magnetic induction. ## Stage 24: Electrostatic and Magnetic Contributions Add One hologram does not separate them automatically. ## Stage 25: Reverse Magnetization or Flip the Specimen Magnetic contribution can change sign while electrostatic contribution remains. ## Stage 26: Separation Requires Reproducible Geometry and State Changing the specimen during the reversal contaminates the calculation. # Diffraction and Mean Inner Potential ## Stage 27: Mean Inner Potential Depends on Composition and Structure ## Stage 28: Strong Dynamical Diffraction Can Bias Phase The projected-potential model becomes incomplete. ## Stage 29: Weakly Diffracting Conditions Reduce This Artifact Tilt control and simulation strengthen quantitative work. # Charging Layer ## Stage 30: Electron Irradiation Can Charge Insulators and Semiconductors ## Stage 31: Charging Produces Long-Range Phase Gradients These can masquerade as built-in device potentials. ## Stage 32: Conductive Coatings Can Reduce Charging but Alter Boundary Conditions Preparation can solve one problem while changing another. # FIB Preparation ## Stage 33: TEM Lamellae Are Commonly Prepared by Focused Ion Beam ## Stage 34: FIB Can Implant, Amorphize, Redeposit or Modify Surface Depletion The thin lamella is not electrostatically identical to the original bulk device. # Semiconductor Junctions ## Stage 35: Built-In Potential Can Produce a Measurable Phase Gradient Doped p and n regions become visible through phase. ## Stage 36: Phase Is Not Dopant Concentration Directly Thickness and mean-inner-potential contributions come first. ## Stage 37: Device Simulation Can Link Potential to Doping The semiconductor canonical owns device physics; holography owns the projected-potential measurement. # Magnetic Nanostructures ## Stage 38: Holography Can Map Magnetic Induction Around Wires and Particles Stray fields outside the specimen are scientifically useful. ## Stage 39: Domain Walls Produce Characteristic Phase Contours Micromagnetic simulations constrain magnetization models. # 2026 Operando Frontier ## Stage 40: Current Biasing Can Be Combined With Off-Axis Holography A June 2026 study/preprint combines 3D Fe nanobridges, micromagnetic simulation and operando current-driven domain-wall motion. ## Stage 41: Operando Validation Tests the Same Structure While the Drive Changes This is stronger than comparing separate before/after samples. ## Stage 42: Holder Leads and Currents Add Their Own Fields The operando apparatus belongs in the forward model. # 3D Holographic Tomography ## Stage 43: One Hologram Measures a Projection Phase integrates along the beam direction. ## Stage 44: Tilt Series Add Angular Information 3D electrostatic or magnetic fields can be reconstructed. ## Stage 45: The Missing Wedge Limits Tomography Finite tilt range creates anisotropic resolution. ## Stage 46: Vector Magnetic Reconstruction Is Especially Underdetermined Maxwell constraints and multiple orientations help. # Resolution and Dose ## Stage 47: Holographic Resolution Depends on Fringe Spacing, Sideband Aperture, Coherence and Dose It is not simply TEM point resolution. ## Stage 48: Phase Precision and Spatial Resolution Trade Off More averaging reduces noise but loses fine features. ## Stage 49: Better Precision Usually Costs More Electrons Beam-sensitive specimens can change during measurement. # DPC Comparison and Machine Learning ## Stage 50: Differential Phase Contrast Measures Beam Deflection It provides a different projected-field receiver. ## Stage 51: Agreement Between Holography, DPC and Simulation Is Powerful Different transfer functions fail differently. ## Stage 52: ML Can Assist Fringe Denoising and Phase Unwrapping But learned priors can smooth away real abrupt fields. ## Stage 53: Forward Re-Simulation Into the Hologram Domain Is the Final Test Recovered phase should reproduce the measured interference pattern. # Professional Layer ## Stage 54: Separate Four Objects 1. specimen electromagnetic potential; 2. transmitted electron wave; 3. recorded hologram; 4. reconstructed phase/field. ## Stage 55: Professional Electron Holography Is a Phase–Thickness–Field Inverse Problem > **Which electrostatic or magnetic field remains identifiable after mean inner potential, thickness, charging, diffraction, reference-wave distortion, FIB damage and phase-reconstruction choices are all allowed to explain the hologram?** # Evidence: What Makes an Electron-Holography Claim Strong? Stronger evidence combines reference holograms, thickness maps, diffraction control, charge-control experiments, magnetization reversal/specimen flip, repeated acquisitions, hologram-domain residuals, DPC/Lorentz comparison and device/micromagnetic simulation. # Misconceptions Worth Hunting – Electron holography directly photographs electric fields. – A phase map is automatically an electrostatic-potential map. – Vacuum is always a perfect reference. – Thickness does not affect phase. – FIB preserves device electrostatics perfectly. – Magnetic phase is automatically separated from electrostatic phase. – Better spatial resolution always means better phase precision. – ML denoising cannot alter quantitative field magnitude. # Transfer Check A junction phase gradient changes after conductive coating while doping is unchanged. Could charging have changed? **Yes.** A magnetic phase pattern reverses sign after magnetization reversal while electrostatic phase stays. Does that support magnetic origin? **Yes.** A phase step doubles where specimen thickness doubles. Did local device potential necessarily double? **No.** # Model Limits Electron holography measures projected phase, not direct local 3D potential, charge density or magnetization. Those require inversion and additional constraints. Professional holography keeps **beam energy + biprism/coherence + thickness + diffraction + charging + raw/reference holograms + phase algorithm + separation method + orthogonal validation** visible together. # Teaching Guide Teach in this order: **electron wave → phase → specimen potential → reference wave → biprism → fringes → Fourier reconstruction → phase unwrap → thickness/MIP → electrostatic phase → magnetic phase → charging → FIB → semiconductor junction → nanomagnetism → operando → tomography → DPC comparison → forward validation.** # Connect This to the eduKate Learning Estate – https://edukatesengkang.com/2026/08/28/how-to-learn-microscopy-scientific-imaging-super-resolution-image-evidence/ – https://edukatesengkang.com/2026/08/30/how-to-learn-ptychography-coherent-diffraction-imaging/ – https://edukatesengkang.com/2026/08/30/how-to-learn-spintronics-magnetic-memory/ – https://edukatesengkang.com/2026/08/28/how-to-learn-semiconductors-transistors-energy-bands-modern-electronics/ # The Quiet Ending The beginner asks, “Where did the fringes move?” The developing microscopist asks, “What phase shift did the specimen add?” The advanced learner asks, “Was it electrostatic, magnetic, thickness or charging?” And the professional asks: > **Which field survives after the reference wave, specimen preparation and phase reconstruction are all made explicit?**