Low-Energy Electron Diffraction (LEED) Surface Crystallography Lab

The de Broglie wavelength of an electron scales inversely with the square root of its kinetic energy: \(\lambda = \sqrt{150.4 / E}\) Angstroms. As the incident beam energy increases, \(\lambda\) decreases, which expands the radius of the Ewald sphere (\(k = 2\pi/\lambda\)). Because the 2D surface reciprocal lattice spacing remains constant, the Bragg diffraction scattering angle \(\sin(\theta_{hk}) = |\mathbf{g}_{hk}| / k\) decreases proportionally to \(1/\sqrt{E}\). On the hemispherical phosphor screen, every diffraction spot smoothly migrates inward toward the central specular (0,0) beam position as energy increases.

In LEED, over 95% of backscattered electrons undergo inelastic collisions (exciting plasmons, phonons, or secondary valence cascades) and exit the sample with reduced kinetic energy. If allowed onto the phosphor screen, they would create an overwhelming diffuse green background haze that obliterates the faint diffraction spots. The 4 concentric hemispherical grids solve this: Grid 1 is at ground potential to ensure a field-free drift region around the sample; Grids 2 and 3 are biased negatively just below the primary gun potential (\(-V_{\text{primary}} + \Delta V\)), acting as an electrostatic high-pass energy filter that repels all inelastically scattered electrons; Grid 4 is grounded to prevent capacitive coupling; and the phosphor screen behind it is held at +5 to +6 kV to accelerate the transmitted elastic Bragg electrons, generating bright, high-contrast diffraction spots.

While a visual LEED pattern reveals 2D unit cell symmetry and surface reconstructions (Wood's notation), it cannot reveal vertical interlayer spacings (\(d_{12}\)) or atomic coordinates due to strong multiple-scattering (dynamical) effects. In quantitative \(I(V)\) LEED, the integrated intensity of specular and fractional-order diffraction spots is recorded as a function of incident beam energy (40 to 300 eV). Theoretical \(I(V)\) curves computed via full relativistic multiple-scattering dynamical theory are compared against experimental curves. By systematically adjusting atomic terrace relaxation, vertical rumpling, and bond angles, the Pendry reliability factor (\(R_p\)) is minimized. An \(R_p\) below 0.20 confirms the structural model, determining interlayer atomic positions with picometer precision (\(\pm 0.01\text{ \AA}\)).

A high-precision UHV LEED optics system (4-grid optics, low-current electron gun, digital CCD camera, and quantitative \(I(V)\) Tensor LEED analysis software) represents a capital investment of $320,000 to $540,000, with annual operating expenses of $20,000 to $25,000 for cathode replacement and UHV chamber bakeouts. In advanced gate oxide interfaces (sub-2nm gate stacks), 2D epitaxial graphene/semiconductor heterostructures, and catalytic wafer processing, in situ LEED validation of surface reconstruction, native oxide thermal desorption, and atomic terrace cleanliness saves $2.2M to $3.4M annually in avoided lot contamination and accelerated process qualification, achieving a discounted payback period of 1.1 to 1.4 years and a 20-year net present value exceeding $20M.