Table of Contents (click to expand)
- Can The Human Eye See Atoms, And Why Can't A Camera Photograph One?
- What Do Atoms Look Like Under A Microscope, And What Is The Image A Picture Of?
- Are Atoms Actually Spheres?
- Why Does Every Atom Come Out As A Round Blur?
- When Do Atoms Stop Looking Like Balls?
- So, Why Do Atoms Look Like Little Balls In Microscope Images?
Microscope images of atoms show round bumps because the instrument does not photograph the atom at all; it maps a quantity near the surface, such as the leaking electric current under a sharp tip or the push-back the tip feels, and renders that map in false color. Three things make the map round: the electron cloud of an atom with filled electron shells has no preferred direction, the tip's own round end is stamped onto every bump it crosses, and the atom wobbles with heat, so a long measurement averages it into a round blur. When the tip is sharp enough and the atom is bonded to its neighbors, atoms stop looking like balls, and one 2009 experiment showed a molecule's rings and bonds like a chemist's drawing.
Open a chemistry textbook to the page on the atom and you get two pictures that disagree. One shows a nucleus inside a fuzzy cloud. The caption explains that an electron has no fixed position, so the atom has no real edge. A few pages on comes a photograph from a scanning tunneling microscope. It shows rows of tidy round bumps, as neat as a tray of ball bearings.
Both pictures come from serious physicists. So which one is lying?
Neither, and the gap between them is the whole story. The round bumps are honest measurements. They are just not measurements of what most people assume. Once you know what the microscope records, the balls make sense. So does the strange moment when a better microscope makes them go away.

Can The Human Eye See Atoms, And Why Can't A Camera Photograph One?
Hold a hair up to a window. You can see it because light bounces off it and into your eye. The hair is about a tenth of a millimeter wide, big enough to block light. An atom cannot do that, and the reason is size.
Light travels as a wave, and a wave of visible light is roughly a thousand times wider than an atom. The wave rolls over the atom like an ocean swell over a pebble. Nothing bounces back that carries the pebble's shape. No lens can fix that: the smallest thing a light microscope can make out is about the size of a large virus. Scientists proved atoms exist long before anyone imaged one, and we tell that story in how we know so much about atoms when we can't see them.
They got around the problem by giving up on light. One route is a beam of electrons, whose waves are far shorter than an atom is wide. The other is a needle so sharp it ends in a single atom. The needle is where the balls come from.

What Do Atoms Look Like Under A Microscope, And What Is The Image A Picture Of?
The ball-bearing images come from a scanning tunneling microscope, or STM. Gerd Binnig and Heinrich Rohrer built it at IBM in Zurich, and it won them a share of the 1986 Nobel Prize in Physics. It is a microscope in name only. There is no lens and no light. It works like reading Braille with a needle, and that picture will carry the rest of this article.
The needle is a metal tip that ends in one single atom. It hovers about a nanometer above a conducting surface, roughly two atom widths, and never touches it. A small voltage sits between tip and surface. By everyday physics, no current should cross that gap. One does, because electrons can leak through a thin barrier they lack the energy to climb. The effect is quantum tunneling, and it gives the microscope its name.
That leak depends ferociously on distance, and that is what makes the machine work. As the tip sweeps across the surface, a feedback loop raises and lowers it to keep the current steady. Record that up-and-down motion line by line and you have a map, with a sideways resolution of about 0.2 nanometers, fine enough to pick out single atoms.
But a map of what? Theorists worked that out soon after: the tunneling current tracks where the surface has electrons ready to leak into the tip, at the energy the voltage selects. That is usually close to where the atoms are. It is not a picture of the atoms, and it is not always centered on them. The glowing colors are a choice, too. The data is a grid of heights, and someone picks a color scale, usually gold, for reasons physics cannot defend.

Are Atoms Actually Spheres?
The honest answer has two halves. A lone atom with full electron shells is round, on average. An atom bonded to its neighbors usually is not.
An electron in an atom has no fixed position, only a set of places it is likely to be. The map of those likelihoods is the electron cloud, and it comes in named shapes called orbitals. The simplest, the s orbital, is a sphere: its density depends only on distance from the nucleus, not on direction. The next shape, the p orbital, is a dumbbell along one axis, and there are three of them, one along each axis.
A single p electron is about as far from a ball as you can get. Fill all three p orbitals, though, and something tidy happens. A dumbbell along x, plus one along y, plus one along z adds up to a shape with no preferred direction: a sphere. A theorem from the 1920s proves that any atom with filled shells has a perfectly round charge cloud. Neon, argon and the other noble gases are round for this reason, and so are the inner shells of every heavier atom. That is the first reason the bumps are round: much of what an atom carries is already round. The other two reasons have nothing to do with the atom.

Why Does Every Atom Come Out As A Round Blur?
The tip is not a mathematical point. It is a lump of metal that ends in an atom, and that atom has its own electron cloud. Every bump the tip passes over comes out stamped with the shape of the tip, the way every dot of Braille feels round and fat under a fat fingertip. A sharp tip adds a little width and a blunt tip adds a lot, and either way the addition is round, because the end of the tip is round. The STM's cousin, the atomic force microscope, or AFM, feels the surface with a tip on a flexible arm instead of tunneling into it, and it has the same problem.
Even a perfect tip has a floor. The current falls off so steeply with distance that the brightest spot is always the nearest point of a feature, and the nearest point of anything is its rounded top. The microscope sees each atom's summit and little of its slopes.

Then there is the atom itself, which is terrible at sitting for portraits. Every atom in a solid wobbles about its resting place, and the hotter the solid, the wider the wobble. A scan takes time, so what the microscope records is the atom's average position, and that average is a round blur centered on where the atom spends most of its time.
For decades this did not matter, because the instrument's own blur was far larger than the wobble. That changed in 2021, when a team reconstructing a thin crystal from a moving electron beam pushed the instrument's blur under 20 picometers, a picometer being a thousandth of a nanometer. At that point the width of each atom in the image was no longer set by the microscope but by the wobble of the atom itself. Cooling would shrink that blur, but not to nothing, because even at absolute zero atoms still have quantum fluctuations. Even the best image of an atom ever made holds a round blur that no engineering can remove. It is the atom's own motion, exposed for the length of the scan.

When Do Atoms Stop Looking Like Balls?
If the balls come from the measurement, a better instrument should change their shape. It does, and the first clean demonstration came in 1987, on a crystal of gallium arsenide. The crystal has two kinds of atoms in a strict alternating pattern. Flip the sign of the voltage on the tip and half the atoms vanish while the other half appear. The bumps mark the electrons on offer at a chosen energy, and the two elements offer theirs at different energies. The atoms did not move; the question put to them changed.

The second demonstration went around the world. In 2009, a team at IBM imaged a single molecule of pentacene, a flat chain of five carbon rings, with an atomic force microscope whose tip ended in a single carbon monoxide molecule. With that one molecule as the needle's point, the image showed the five rings and the bonds between the atoms, the way a chemist draws them. It was the first time all the atom positions and bonds of a single molecule had been seen.
Two details answer two more questions people ask. First, the tip felt the molecule through Pauli repulsion, the refusal of two electron clouds to overlap, the same push that stops your hand passing through a table. The AFM touches the molecule exactly as much as you touch anything: electron clouds pushing on each other, never solid on solid. Second, pentacene's atoms do not look like balls because they are bonded. Shared electrons pile up between the carbons, and the round shells of a lone atom give way to the ring-and-stick shape of a molecule.
So, Why Do Atoms Look Like Little Balls In Microscope Images?
Because three separate things make them round, and none of the three is a camera.
The instrument measures a quantity near the atom, not the atom. An STM maps where electrons are available to tunnel; an AFM maps the push the tip feels. Both peak near the top of each atom and fall away in every direction, so each atom becomes one bump. The tip adds its own round shape to every bump it crosses, like a fat fingertip on Braille. A measurement that takes time averages the atom's wobble into a round blur. Underneath all that, a lone atom with filled shells is round on average, because its dumbbell orbitals add up to a sphere.
Better instruments change the balls, which is the proof that the instrument is partly to blame. Switch the voltage on gallium arsenide and half the balls disappear. Put a carbon monoxide molecule on the tip and pentacene shows its rings and bonds. Push the blur below the atoms' own wobble and a blur the size of that wobble remains, and no cooling can fully remove it.
So the ball-bearing pictures are honest data read with a fat fingertip. The fuzzy-cloud drawing is the theory the fingertip cannot feel. Neither picture shows an edge. Even in the sharpest images ever made, the smallest blur on record is about 20 picometers wide, and it belongs to the atom.
References (click to expand)
- Visible Light — NASA Science
- Atom — Britannica
- The Nobel Prize in Physics 1986, Summary — NobelPrize.org
- The Nobel Prize in Physics 1986, Press Release — NobelPrize.org
- Tersoff, J. & Hamann, D. R. Theory of the scanning tunneling microscope. Physical Review B 31, 805 (1985)
- 4.07: s-orbitals are Spherically Symmetric — Chemistry LibreTexts (Lebanon Valley College)
- Rosu, H. C. Elementary Quantum Mechanics (arXiv:physics/0004072), problem on Unsöld's theorem
- Mirror effect in atomic force microscopy profiles enables tip reconstruction. Scientific Reports (2020), PMC7641199
- Chen, Z. et al. Electron ptychography achieves atomic-resolution limits set by lattice vibrations. Science 372, 826 (2021) — NSF Public Access copy
- Chen, Z. et al. (2021) — arXiv:2101.00465 (author version)
- Cornell researchers see atoms at record resolution — Cornell Chronicle, May 2021
- Feenstra, R. M., Stroscio, J. A., Tersoff, J. & Fein, A. P. Atom-selective imaging of the GaAs(110) surface. Physical Review Letters 58, 1192 (1987)
- Gross, L. et al. The Chemical Structure of a Molecule Resolved by Atomic Force Microscopy. Science 325, 1110 (2009)
- Molecule's Atoms, Bonds Visualized — Chemical & Engineering News, 2009







