On this page
NCERT's Class 12 wave optics chapter says the resolution of a microscope is limited by diffraction. For over a century that limit, about half the wavelength of light, was treated as final. Stefan Hell refused to accept that. This post explains the limit, how Hell came up with a way around it, how his method works, and a worked example of the resolution formula.
Who is Stefan Hell?
Hell was born in Arad, Romania, on 23 December 1962, according to his Nobel Prize facts page. In his Nobel autobiography he writes that his father was an engineer and his mother a primary school teacher, and that the family left for West Germany on 8 April 1978, when he was 15. He studied physics at Heidelberg from 1981 and received his doctorate in 1990. After posts at the European Molecular Biology Laboratory in Heidelberg, the University of Turku in Finland and Oxford, he joined the Max Planck Institute for Biophysical Chemistry in Göttingen in 1997. He shared the 2014 Nobel Prize in Chemistry with Eric Betzig and William Moerner "for the development of super-resolved fluorescence microscopy".
What is the diffraction limit?
The Nobel committee's scientific background explains that Ernst Abbe (1873) and Lord Rayleigh (1896) set out the limit. Two features closer than about half the wavelength cannot be resolved. More precisely, the smallest resolvable distance is about λ ÷ (2n sin α), where n is the refractive index of the medium and α is half the angle of the cone of light the lens collects. For visible light that works out to roughly 200 nm (blue) to 350 nm (red). A whole bacterium is only about 1,000 nm long, so its insides were a blur.
NCERT's Class 12 chapter Wave Optics makes the same point: the finite resolution of microscopes and telescopes is limited by diffraction. Our guide to single-slit diffraction covers the physics behind it.
How did Hell get the idea?
In his autobiography Hell says that as a PhD student he found optical microscopy dull, a 19th-century subject. The only part that interested him was the resolution limit: "So I figured that breaking this limit would be really new and exciting!"
After working on a two-lens design (the 4Pi microscope), he moved to Turku in 1993. One Saturday morning that autumn, reading a textbook on the quantum theory of light, he reached a chapter on stimulated emission. His thought: why excite the fluorescent molecules, why not de-excite them, keeping them dark so they can be told apart from their neighbours? A quick estimate suggested 30 to 35 nm resolution was possible.
With the student Jan Wichmann he published the proposal in 1994. Their paper in Optics Letters described a scanning fluorescence microscope "capable of resolving 35 nm in the far field". Many were sceptical. Back in Germany, his group gave the experimental proof in 2000: the PNAS paper reported a nearly spherical fluorescent spot 90 to 110 nm across, with the spot along the optical axis shrunk up to six times below the diffraction limit.
How does STED work?
STED stands for stimulated emission depletion. The Nobel background describes two laser beams:
A normal excitation beam makes molecules in a diffraction-limited spot ready to fluoresce.
A second, more intense beam of longer wavelength is shaped like a doughnut, with zero intensity at the centre. Wherever it is bright, it forces excited molecules to give up their energy by stimulated emission before they can fluoresce.
Only molecules at the doughnut's dark centre still glow. Scan that tiny spot across the sample and you build up a sharper image.
Worked example: how small can the spot get?
The Nobel background gives the STED resolution as Δ ≈ λ ÷ (2n sin α × √(1 + I₀/I_sat)), where I₀ is the peak intensity of the doughnut beam and I_sat is the saturation intensity of the fluorescent molecules.
Sample values: λ = 600 nm and n sin α = 1.4.
Ordinary limit (I₀ = 0): 600 ÷ (2 × 1.4) = 214 nm.
With I₀/I_sat = 24: √25 = 5, so Δ = 214 ÷ 5 ≈ 43 nm.
With I₀/I_sat = 99: √100 = 10, so Δ ≈ 21 nm.
Because of the square root, a ten-fold sharper spot needs about a hundred times the saturation intensity. The Nobel background notes that very intense light can damage living samples, which is why Hell also explored gentler switching methods.
Common slips:
Dropping the square root. I₀/I_sat = 99 improves resolution 10 times, not 100.
Forgetting the 1. With no STED beam, the formula must return the ordinary limit.
Mixing up n sin α. It is the numerical aperture, not the refractive index alone.
What can a student take from Hell's way of working?
Distil the idea. Hell writes that as a student he spent months of vacations picking textbooks apart, and "I firmly believed that everything could be boiled down to simple principles."
Question the accepted limit. Everyone knew about Abbe's limit; he asked what assumption it rested on.
Borrow from another field. The key idea came from a quantum optics book, not a microscopy one.
For how microscopes form an image in the first place, see our guide to optical instruments. If diffraction and interference still feel abstract, the Fermi Simulation Lab has interactive simulations for ideas that are easier to see than read.
Next step: with λ = 500 nm and n sin α = 1.4, find the I₀/I_sat needed to resolve 30 nm.
