Hacker News new | past | comments | ask | show | jobs | submit login
Picture of a Single Atom Wins Science Photo Contest (petapixel.com)
355 points by gokhan on Feb 12, 2018 | hide | past | web | favorite | 85 comments

David Nadlinger, the photographer, is an avid contributor to the D programming language[1]. He works on LDC compiler[2], the LLVM backend for D.

[1] http://klickverbot.at/

[2] https://github.com/ldc-developers/ldc

This is an insanely cool picture.

Some quick napkin math (please let me know if I'm way off base here): in the second pic, the electrodes are ~74 pixels apart and the strontium atom is ~2 pixels wide.

The electrodes are ~0.078 inches apart, so these 2 pixels represent ~0.0021 inches (or ~0.05334 mm).

Google tells me that the Van der Waals radius of a single strontium atom is 255 pm.

So the diameter of the actual atom is something like 10,000x smaller than the space represented by these pixels. Crazy!

I guess it's basically a point source compared to the camera resolution, so what you're seeing as the "atom size" is the diffraction limit of the camera lens.

While the manual for my DSLR doesn't specifically refer to this use case, reading between the lines it sounds like using a jpg encoding for atomic-scale image capture isn't recommended.

I was just thinking that the atom has to be letting off (reflecting) some serious quantity of light in order for just one to appear with off the shelf camera equipment.

I still don't get it, 10,000 times more light than its size?

They say it was a long exposure.

The atom was illuminated and irradiated light for a long enough time to register on the sensor.

Even if the exposure were for a year, if the amount of light being emitted is less than the hardware can perceive, it wouldn't register - hence the need for electron-microscopes and such. There's a story here about the size of the light being emitted that doesn't seem satisfactorily answered (not by you, the article and stuff in general).

> Even if the exposure were for a year, if the amount of light being emitted is less than the hardware can perceive, it wouldn't register

That is not true. It might be noisy, but it would show up.

In point of comparison the human eye can see a single photon.

A single atom can emit a single photon, ergo you can see a single atom.

You won't get any detail out of it obviously, but you'll see it as a dot of light.

That's what's happening here.

> There's a story here about the size of the light being emitted that doesn't seem satisfactorily answered

You are misunderstanding the "size". The size "10,000" is the resolution needed to differentiate between two atoms sitting next to each other.

But in order to see the atom (detect the atom), you just need to be able to capture what it emits. And it emits photons, and it's not that hard to capture and record a single photon.

So you do, and it shows up as a single dot in the image.

If you had two atoms near each other, both emitting photons, you would not be able to distinguish them from each other (i.e. you couldn't say if there were one or two), unless they were farther apart.

If we're seeing the single photon emitted by a single atom, then how much larger is the spot of light in the image compared to the atom that emitted it?

Why isn't there a ruler somewhere in the picture? My brain is absolutely failing to make sense of the scales and sizes in this image.

> then how much larger is the spot of light in the image compared to the atom that emitted it?

The atom is about 255 picometers, the light is around 400 nanometers. That makes the light around 1500 times larger than the atom.

(Which also helps explain why visible light can not distinguish atoms placed close together - it's so much larger than them.)

> Why isn't there a ruler somewhere in the picture? My brain is absolutely failing to make sense of the scales and sizes in this image.

The two electrodes are about 2mm apart. About the width (not length) of a sesame seed. It's really small. The photo was taken through a microscope.

Another point of reference: the width between the electrodes is about 5,000 times the wavelength of the light being emitted.

In theory, if you had a good enough camera, the dot of light would be around 1/5000 of the width. (However the diffraction limit of your lens might enter into play, blurring the image.)

And finally, the width between the electrodes is about 8,000,000 times the size of the atom.

Thank you so much; that was exactly the explanation I needed =)

Maybe it's light reflected by the atom at multiple points? The article mentions that it's "held nearly motionless" but over a long exposure perhaps the degree of movement allowed created a composite large enough to show up in the photo.

No, that's not it. Just because it is small, doesn't make it invisible.

You can't see single atoms because they normally don't send any light toward you. But if they did, and this one is, you would see that light.

You aren't seeing the atom exactly, you are seeing what it sends toward you, and the human eye can see single photons. (And so can cameras.)

> No, that's not it.

That has to be at least partially it for the rest of your comment to be right (which it is!). To say the atom was "held nearly motionless" is to say it moved a little during the exposure. It sent the light toward the sensor from all of its positions.

> That has to be at least partially it

No, not at all. The wavelength of the light is about 1,500 times larger than the atom emitting it. So the atom would need to move at least 1,500 diameters to make any difference in the light.

It doesn't move that much.

> It doesn't move that much.


Think about it this way: every photon your eyes perceive came from a single atom.

CCDs can detect single photons. It isn't much of a challenge. I would be suspicious that the surrounding apparatus is Photoshopped in from a separate exposure.

Nope, not photoshopped. It's a single exposure; the apparatus is illuminated using flashes.

The latter made lighting the shot in a controlled fashion a bit easier than if I had used continuous sources – you'd be looking at using a torch with a bunch of filters or a computer monitor on the lowest brightness otherwise.

View it here in actual life size on your display. http://beta.lifesizer.com/view?user=5&ref=_59&h=1080&w=1920

Edit: based on seeing it in actual size, I would say it would be hard to see it with the naked eye, that said as pointed out in other comments, I suppose not impossible if a human eye can detect a single photon.

The article did mention that the photo was long exposure, so it may not be possible to see it easily with the naked eye. A single photon emitted occassionally would be pretty hard to see, even though the human eye can detect a single photon.

Thanks for the link. Very cool. What are some use cases for this service?

Mostly dickpics

Photographer here. The amount of attention this has received has caught me a bit off guard – this is really just a somewhat pretty picture of what is a standard technique in physics by now.

I'm putting together a short post with answers to some of the most commonly asked questions, but in the meantime, check out this great comment by a well-informed Redditor:


Why does the atom look so big in the bicture? Atomic radius of Strontium is 219 pm, so that small spec there in the picture should be about 438 pm across.

I'm assuming the two ball-pen nib shaped structures on both sides of the spec in the picture are "two metal electrodes placed about 2mm (0.078in) apart". So, the space between the left tip of the electrode and left edge of the spec is about 1 mm. Based on some "visual calculation" (zooming in the picture and doing some approximation), the spec seems to be closer to about 0.03 mm across, which is orders of magnitude larger than 438 pm that it should be. What gives?

The atom can't be directly imaged by bouncing light off it, it's far too small to do that (with visible light).

Instead they're shining a laser at it, which excites some of its valence electrons to higher orbitals. When those electrons drop back to their ground states they emit (visible) light, some of which reaches the camera.

The atom is effectively acting as an isotropic radiator (radiating equally in all directions). The camera lens is much larger than the atom.

There are thus multiple paths from the atom to the lens and the sensor behind. Less light will reach the camera from greater angles, so the light that goes straight towards the lens or nearly straight will impact the sensor most, and that area will appear brightest. (This is my somewhat mangled attempt to explain abberation in optics...)

Even if the light were perfectly collimated in a beam the size of the atom it would still be unable to make a dot in the final image smaller than a single pixel of the sensor. If film were used instead of a digital sensor it would expose at least one pigment grain, again much larger than the atom itself.

The apparent size is an artifact of the imaging process.

You're not looking directly at the atom itself (which is impossible... atoms are smaller than light's wavelength).

You're looking at the photons emitted by an excited atom, as collected over an extremely long exposure by the camera's sensor. Which will resolve to ~1px in size... the smallest unit the camera can image.

It doesn't really mean anything to day "you aren't looking at it, you're looking at the reflected photons". You're always looking at reflected photons of everything.

Actually “reflected” doesn´t mean much for photons: they are absorbed and re-emitted.

Isn't that what "reflected" means?

Not exactly.


For most of the history of the word “reflect”, its meaning was definitely not “absorb and re-emit photons of the same frequency”.

It doesn't appear that the atom is 1 pixel, but is in fact quite a bit larger than this: there are lots of elements in the image which are significantly smaller than it. So what gives?

I used to work on trapped ion experiments, and the limitation of the atom size was always the diffraction limit, which is limited by the NA of the lens (f-stop in camera terms) and the wavelength of light. In this case, the optical system (I'm guessing a camera lens) is designed for multiple wavelengths, so it might not reach the diffraction limit at the emission wavelength, which is like 400nm. In that case, the limit would be the aberration of the camera lens, which can be wavelength-dependent. Most camera lenses aren't designed for 400nm light, which is marginally visible.

Possibly diffraction, possibly JPEG artifacts, possibly camera movement. Tough to say without more details.

Looking at it, my suspicion is diffraction. But I don't know enough optics to be sure.

This makes me think of Voyager 1's pale blue dot photograph. https://upload.wikimedia.org/wikipedia/commons/7/73/Pale_Blu...

I find the parallel awesome.

I did wonder if it was a Sagan reference: "... I was rewarded with this particular picture of a small, pale blue dot.”

Ref: https://en.wikipedia.org/wiki/Pale_Blue_Dot

Implying a pale, yellowish white dot depicting the Sun is next. (Or a pale red dot of a Tesla, more likely, given that negatives are probably already being post processed. You have to imagine that astronomy as a discipline took the last few nights off from science.)

How is this possible? For an atom to stand out, are all the atoms around the area cleared out using electricity so it’s pure, empty space? Is that possible? Also, as other people mentioned, I imagined atoms to be MUCH smaller in scale than anything like what seems to be visible here.

My guess is they made that atom very, very, very bright, among other things...But I'm no scientist

Awesome picture!

Could someone explain to me how a DSLR camera can capture an image of a single atop? I thought atoms were very very small, like so small only an electron microscope could "see" them. Maybe this device has been misrepresented and it is holding a small amount of atoms rather than a single one?

In exactly the same way that a camera can take a photograph of a star, which (for all but the Sun) is an infinitesimally small point of light. As long as the light intensity is higher than the background noise, it will render as a smeared out blob[1] due to diffraction.

[1] an Airy disk: https://en.wikipedia.org/wiki/Airy_disk

From the article:

> When illuminated by a laser of the right blue-violet color, the atom absorbs and re-emits light particles sufficiently quickly for an ordinary camera to capture it in a long exposure photograph.

That explanation doesn't really satisfy. Perhaps, based on a large body of experimental evidence, it is known that the number of atoms present can be determined by the apparent brightness. Perhaps it's something totally different. Regardless, it would be helpful if the author/photographer/somebody explained the rationale that makes them confident that this is only a single atom.

> the rationale that makes them confident that this is only a single atom

Nothing photographic, as I understand it. The atom is positively charged, and is suspended in ultra-high vacuum by electric fields. Effectively they're "weighing" it, by measuring the total charge of whatever is suspended, and knowing (from the ionization energy?) the charge on one atom.

It should be noted that this is similar to how Millikan first measured the electron charge back in 1909.

Isn't hard to figure out when you've only got one ion in the trap: https://www.physik.hu-berlin.de/de/nano/lehre/quantenoptik09...

Heck, you can line up ions in a row and take pictures of those if you'd like.

Just because my previous comment is now gone: It's nice to mark PDFs

It's nice to have PDFs mark themselves as such...

Basically the atom is glowing. Not very brightly, but enough so that the camera can see it.

The human eye after all can see a single photon. A single atom can emit a single photon, so it's not such a stretch.

They only put 1 atom there, and it's in vacuum. It can't be anything else because of that :)

I'm guessing there are at least two layers of glass between that atom and the sensor. Not really specific to this (excellent) picture, but I imagine that avoiding distortion due to the glass or smudges on the glass must be a huge part of high magnification photography. The whole light from a single atom part just makes it more amazing how clean and clear that window is...

The article makes it sounds like the photographer already knew he could trap one atom, before taking the picture.

My guess would be, as this is an electromagnetic trap, that two such ions would repel each other and the well is set to be shallow enough that at most one ion can remain inside.

When blasting light in every direction, does it really matter how small the source is? As long as those photons hit your eye, you can see the object.

Ah, missed that. Thanks :)

Years ago, someone created a similar picture with a single atomic nucleus. I'd like to find it again.

Normally, nuclear transitions are invisibly-high energies. X-ray and gamma. But a couple of nuclei have multi-step spin isomer decays (or was it shape isomer?). One step of which is visible. So someone trapped and fully stripped a nucleus, and bombarded it to visibility. Naked-eye visibility. And took a picture. Of vacuum vessel window, with a green(?) fluorescence dot.

I saw that picture years ago. Likely a cover photo on something.

I would very much like to find it again. For use in educational content. A lot of time has been spent looking for it, both by myself, and a couple of MIT science librarians. If anyone has a clue, I'd very much appreciate it. Here is a mockup[1] I gimped for user testing.

Why use it in education? To make the concept of atoms more concrete. Atomic electrons are too small (and slow and fragile) to see with your naked eye. But (a couple of) nuclei you can (under hard-to-contrive conditions). Absent concrete, students build their understanding of materials on muddy misconceptions. For example, one failure-mode in teaching high-school stoichiometry, is students not thinking of atoms as real physical objects.

And to preempt a common response... one professor complained "you aren't really seeing the nucleus - it's only a diffraction dot"... right before they headed out to a star party, to apparently "not see" stars. ;)

[1] http://www.clarifyscience.info/assets/2017-Atoms/assets0/wk/...

I love what I assume is a winking "pale blue dot" reference in the quote at the end.

It's the same idea, isn't it? A single pixel with a whole lot of meaning added to it.

The article is about seeing a single atom; it's possible to see a single photon too, as was previously discussed on HN:

People can sense single photons: Experiment suggests that humans are capable of perceiving even the feeblest flash of light [1][2]

[1] http://www.nature.com/news/people-can-sense-single-photons-1...

[2] https://news.ycombinator.com/item?id=12163109

Is the light refracting to make the atom appear bigger than it actually is?

The laser light is absorbed by the atom and is released at a different wavelength. While the radius of the atom is 200 x 10^-12 meters, the wavelength of the light being emitted is in the visible spectrum, so between 380-500 x 10^-9 M which is why it is visible at the scales you can see.

A scanning electron microscope, or an x-ray crystallography based machine, uses items with much smaller wavelengths, .01 - 10 nm in the case of x-rays and so they're able to peer inside an atomic structure.

I'm imaging that the process of "magnifying" objects through photography has to degrade the true representation of the object. Kind of like expanding 1 pixel to cover a 500x500px space. In that sense I wonder how clearly we can ever hope to see an atom.

Obviously the technology and intent is impressive, but will we ever be able to see an atom more clearly or is it simply too small to be seen without Photoshopping in the details?

The internal structure of an atom is vastly different from the impression that most science books give you with their illustrations to the point that the question of us being able to clearly seeing an atom is almost nonsensical/doesn't make sense.

Atoms are much more nebulous than the concept of small indivisible balls of matter.

Machines like the scanning electron microscope or an x-ray crystallography machine allows us to peer inside atomic structures to some extent.

The problem is that of scale: visible light has a wavelength of 380-750 nm (10^-9 meters) while the atomic radius of even large atoms are still in the ~200 pm (10^-12m) which means that visible light still has a wavelength 1000 times that of an atomic radius. To peer more closely at an atom, you would need to be able to use something moving at a much smaller wavelength, like an x-ray or a focused beam of electrons.

I don't think that is completely true... if you are using lenses to magnify, you aren't doing the same thing as expanding 1 pixel to 500x500. The light 'resolution' is already present, you are just spreading it out so you can see each bit more clearly.

I'm curious to know how they managed to get a single atom there in the first place. I'm assuming it's not with atomic tweezers?

I don't know in this particular case, but in the late 80s, IBM figured out how to move individual atoms (see https://en.wikipedia.org/wiki/IBM_(atoms)).

For the folks wondering how a single atom is showing up on a DSLR CMOS, try taking a photograph of a head of a metal pin with a very bright light shining on it. You'll see lots of glint and reflection, and it'll look larger on the camera - basically a single point reflecting light into multiple pixels of the camera sensor.

How do they know it's a single atom?

I followed ion and atom trapping a long time ago, when it was still fairly new. (i.e., a couple decades ago) That was a valid question from the git-go, and there were a number of techniques. One way is that the trap is less stable when there are multiple atoms, so the extra atoms will eventually evaporate away, and you can measure the diminishing intensity of the scattered laser light. A paired atom will also interact with the light differently, for instance having a different spectral signature.

This is just great. Two great photographs in such a short span. SpaceX launching a Tesla car and landing the rockets back for reuse on the macro level. An atom on the nano level. Evidence that humanity’s progress in the field of physical sciences is well balanced.

How does this not violate Heisenberg's uncertainty principle? Unless the explanation is that the atom moves around a lot in that tiny space, so we haven't actually determined its exact position and momentum.

this takes the cake for great desktop backgrounds.

For a while I used the "opposite" of a single atom for my desktop background. Quite a few atoms in that picture! 5500 galaxies worth.

https://upload.wikimedia.org/wikipedia/commons/2/22/Hubble_E... https://en.wikipedia.org/wiki/Hubble_Ultra-Deep_Field#Hubble...

Gorgeous photo in so many ways. I'm unable to dig up a high-res version. Anyone else having more luck?

Awesome, thanks! Do you know where I can find the other submissions? Particularly "In a kitchen far far away..."? It looks astounding!

Scroll to the bottom and click around the slideshow. It's in with a number of others not mentioned:


Sadly those aren't the high-resolution photos. If you look at the link abainbridge shared you'll notice the image is a LOT higher resolution than the ones in the link you shared.

Thank you so much!

Awesome pic - and just as mind blowing. Is there a 4k version of the pic?

Are they using a Rydberg atom?

Lame. This is an optical illusion. Call me when you actually image an atom at the atomic scale.

That's what I was expecting, when I read the title description: Picture of a Single Atom Wins Science Photo Contest

Like, the helical structure of DNA. I want to see how an atom looks like in reality, on the atomic scale. Along with the protons, neutrons, and maybe the electrons flying around it.

Zzz.. going back to sleep now..

> Along with the protons, neutrons, and maybe the electrons flying around it.

That is not how reality works. Stuff is quantum mechanical and has no analog in the macroscopic world.

All one can ever see are consequences of interactions. This is what we are seeing.

Registration is open for Startup School 2019. Classes start July 22nd.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact