A conventional lens cannot resolve two points closer than the diffraction limit, because the high-spatial-frequency (evanescent) light they scatter decays exponentially within a fraction of a wavelength and never reaches a far-field detector:
d_min = λ / (2n)
A hyperlens is a cylindrical (or spherical) shell of alternating nm-thick metal and dielectric layers. Effective-medium theory gives it a hyperbolic dispersion relation — its radial and tangential permittivities have opposite sign:
k_r² / ε_θ + k_θ² / ε_r = (ω/c)², ε_r · ε_θ < 0
Because the equation is hyperbolic rather than elliptical, it has no upper cutoff in k_θ — even very high spatial-frequency (sub-wavelength) components propagate radially instead of decaying. As they travel from the small inner radius r₁ to the large outer radius r₂, the same angular slice Δθ is stretched over a larger circumference, so the absolute separation grows linearly with radius:
Δθ = s / r₁ (object features preserve their angle)
Δ_image = r₂ · Δθ = s · M M = r₂ / r₁
Once Δ_image exceeds a normal microscope's diffraction limit, an ordinary lens behind the hyperlens can capture and resolve detail that was physically impossible to image directly. This is the mechanism behind demonstrated silver/alumina hyperlenses used for real-time, label-free sub-diffraction imaging of cell membranes and early sub-cellular pathology.
- Feature spacing s — the true separation of two points on the imaged surface (a membrane protein cluster, say).
- Wavelength λ — sets the classical diffraction limit at the object plane.
- Magnification M — the shell's outer/inner radius ratio; larger M pulls sub-diffraction features further apart by the time they reach open air.
- Layer pairs — more, thinner alternating layers approximate the ideal hyperbolic medium more closely (drawn as the concentric ring structure).