New Electron Microscope Shows Atoms in Color

Cornell's Duffield Hall has acquired a new electron microscope that is enabling scientists to see individual atoms in color for the very first time. While old electron microscopes can be compared to black and white cameras, this new scanning transmission electron microscope uses a new aberration-correction technology that is both more intense and allows for faster imaging speed. "The method also can show how atoms are bonded to one another in a crystal, because the bonding creates small shifts in the energy signatures. In earlier STEMs, many electrons from the beam, including those with changed energies, were scattered at wide angles by simple collisions with atoms. The new STEM includes magnetic lenses that collect emerging electrons over a wider angle. Previously, Silcox said, about 8 percent of the emerging electrons were collected, but the new detector collects about 80 percent, allowing more accurate readings of the small changes in energy levels that reveal bonding between atoms."

Not color, false color.

[oskay]

These atoms are color coded, not *seen* in color by the microscope.

Re:

[sakdoctor]

I'm not sure what the actual innovation is here. False colour, or colour coding atoms and other features is as old as electron microscopy itself.

The concept of colour doesn't really make sense at atomic scales anyway.

Re:Not color, false color.

[kebes]

I'm not sure what the actual innovation is here.

Using false-color in an image is certainly not the innovation. What is innovative is their use of corrective optics to achieve much higher signal (100-fold increase compared to conventional techniques), and the integration of energy-loss spectroscopy. This means that for each pixel in the image, they can determine what kind of atom is being measured. So they can generate false-color maps of atomic identity. Most electron microscopes simply measure electron density: you can guess which element is which based on density, but there can be ambiguities. Some microscopes include detectors for determining what elements are present, but not with high spatial resolution. This new refinement allows precise maps where individual atoms can be both localized, and elementally identified.

The image they show is impressive when you consider that each blob of color is actually an individual atom, and that they've identified exactly what kind of atom is at each position. In this case they were using it to analyze interdiffusion of atoms at an interface. As nanotechnology becomes more and more 'real' you can imagine how useful it will be to image nano-objects with atomic resolution and elemental discrimination.

Re:

[anastasd]

This means that for each pixel in the image, they can determine what kind of atom is being measured. So they can generate false-color maps of atomic identity.

That's interesting. I guess this microscope will have lots of applications. At first thought - in semiconductors production, carbon allotropes and God knows where else.

Re:

[kestasjk]

This means that for each pixel in the image, they can determine what kind of atom is being measured. So they can generate false-color maps of atomic identity.

That's interesting. I guess this microscope will have lots of applications. At first thought - in semiconductors production, carbon allotropes and God knows where else.

Just look at the images in the article; you can clearly distinguish lanthanum from titanium, manganese, and manganese-lanthanum. From that list alone the mind boggles with potential applications.

Re:

[Crypto Gnome]

you can clearly distinguish lanthanum from titanium, manganese, and manganese-lanthanum

All well and good, but unfortunately these false colors need a bit more care in the selection process.

Apparently you cannot "see" the difference between Krypton and Chlorine using this process.

Which, quite frankly, can be quite fatal for some [google.com].

Re:Not color, false color.

[koolguy442]

Not to get too technical here, but each blob is actually a column of atoms, as the specimen is wedge-shaped and certainly more than one atomic layer thick.

Electron energy-loss spectroscopy (EELS) has been combined with STEM imaging for several years at least, allowing similar sorts of images to be synthesized. The major contribution of this work is that they've modified the optics so that, even at 0.5 angstrom beam widths (and hence pixel sizes), they still get enough signal to the EELS detector to allow for EELS mapping and spectra acquisition for each of those pixels, giving direct bonding information about the particular portions of atoms probed by the beam. That means that the researchers can tell the difference between titanium atomic columns at different locations within the crystal, depending on the other atoms surrounding them.

Well, I suppose I did end up getting too technical.

IAATEL (I am a transmission electron microscopist)

Re:

[davros-too]

The scientists quoted are top notch. I used to work with David Muller, and you can trust this to be both scientifically sound and bleeding-edge technically.

I was *almost* doing this in the 1990s. I could have showed you a coloured image at atomic resolution with colours based on EELS spectra, but IIRC the contrast was mainly from electron-channeling and therefore bullshit. I'm confident that these guys have eliminated such effects.

The uses of this technology in materials science will be enormous.

Re:

[doom]

davros-too wrote:

I was *almost* doing this in the 1990s. I could have showed you a coloured image at atomic resolution with colours based on EELS spectra, but IIRC the contrast was mainly from electron-channeling and therefore bullshit. I'm confident that these guys have eliminated such effects.

Actually, you want to be very careful about getting involved with STEM work, because almost all of it is sample preparation, which is on the order of placing samples in a solvent and staring at them until you ca

Re:

[davros-too]

Sorry doom, just noticed your reply.

Yes, sample prep can be a huge pain - ranging from about 1 hour on prep to one on the microscope to litterally dozens of hours prep per hour on microscope. It really depends a lot on what you're looking at.

Beam damage is the other big problem. And that is where you might fall short on something like high-Tc superconductors. IIRC (and my knowledge is well out of date) these were fairly sensitive to beam damage.

EELS is great for low-Z elements. I did most of my work on carb

Re:

[Poromenos1]

IAATEL (I am a transmission electron microscopist)

Not to get too technical here, but that's an "ell", you need an "em".

Well, I suppose I did end up getting too technical! :P

Re:

[koolguy442]

Microscopy's what I do. I never said I was good at acronymization.

Though looking back at it, I feel really dumb because I don't know how I made such a blatant error, what with being capitalized and all!

Re:

[Poromenos1]

Heh, you just probably got confused with "ELectron", it's not that rare a mistake.

Re:

[JeanPaulBob]

Using false-color in an image is certainly not the innovation. What is innovative is their use of corrective optics to achieve much higher signal (100-fold increase compared to conventional techniques), and the integration of energy-loss spectroscopy. This means that for each pixel in the image, they can determine what kind of atom is being measured.

Almost. Energy-loss spectroscopy in SEMs isn't new. (And I don't think it's new in STEMs, either, AFAIK.) The innovation is in the corrective optics, as you

Re:

[ChrisMounce]

I'm guessing the major innovation is the new imaging technique (is it the "new aberration-correction technology"?) that allows them to gather enough information to false color things in the manner they did.

I suppose you could redefine "color" (what wavelengths will this atom emit), but it's still not going to be the color we know from the macro world.

Re:

[sveard]

(what wavelengths will this atom emit)

Shouldn't that be "reflect"?

Re:

[shawn(at)fsu]

Not necessarily, one way that atoms can emit light is if you bump its electrons in to a higher orbit with energy, when they return to their natural state the will emit energy sometimes in the form of visible light. This is how things that glow when exposed to ultraviolet light work. At least if I remeber my high school chemistry correctly.

Re:

[sentientbeing]

I always think of neutrons as black, protons as white and electrons as yellow.

I must have saw them coloured like that in a book early in my studies, and now I cant think of them any other way.

Re:

[cheater512]

Its significantly better false colour. :)

Actually, it *is* real color.

[CustomDesigned]

The color is based on the energy of the electrons, just like photon "color" is based on the energy of individual photons. The microscope is "color" because it can record the energy of the electrons as well as their density. Thus it is "color" just as much as your eyes - which measure photon energy (cone cells of 2 to 3 or in some cases 4 types) as well as photon density (rod cells). Note that your cone cells require more light to get a color signal. In dim light, you see black and white via your rod cel

Re:

[davros-too]

Sorry, no. The colours are atom types as inferred from the energy loss spectra - for example in one image lanthanum is coloured green.

Re:

[CustomDesigned]

Yes, energy loss spectra - as in electron energy. As in "color". Electron energy is "color". Just like photon energy.

Re:

[Spudtrooper]

Ted Turner, eat your heart out.

size of atoms wavelength of visible light

[vivin]

Good point. Also, atoms are much smaller than the wavelength of visible light. So they cannot reflect color.

Re:

[SamuelComeau]

"how atoms are bonded to one another in a crystal, because the bonding creates small shifts in the energy signatures " I may be a humble college chemisty student, but I can see you are all taking this way too literally. Obviously it isn't a visible color (Wavelength of about 3.9e-7m to about 7e-7m) but the point is that the spectrum goes anywhere from long radio waves (several meters) to gamma rays and shorter (1e-11m and less) The awesomely cool part is not the "color"; it is the fact that they can "see

Re:

[Let them eat cake]

The analogy in the article is technically correct. The atoms are actually "seen" in "color" by the microscope. Photons of light have an energy associated with them. For instance, blue light has a higher energy than red light. Sometimes when an electron scatters off of an atom in the sample, these electrons will lose a specific amount of energy which is related to the type of atom that they scattered from. If an electron looses allot of energy then it can be represented as blue while an electron which

Phew, what a relief!

[GameboyRMH]

I was imagining all the trouble involved in re-learning the atomic color schemes!

Ahh Color...

[clonan]

And I thought we were beyond Technocolor !!!

Re:

[l2718]

Perhaps you were thinking of Technicolor [wikipedia.org] ?

Re:

[clonan]

Or in this case TECHNO-color

Re:

[l2718]

Are you under the impression I was referring to the photographic process? If so, follow the link.

Re:

[clonan]

Yes as was I until I did a typo ;)

Re:

[dido]

After all, an atom is smaller than a wavelength of visible light, so atoms are quite literally colorless.

Schrodinger's Fridge

[sm62704]

The summary didn't say, but the colors MUST be false color, since atoms are smaller than light wavelengths. But will it allow you to photograph atoms without destroying them? [angryflower.com] (yes the link is humorous, but the question I ask is serious)

Re:

[sconeu]

But do we know if Schroedinger has milk in his Fridge without looking?

Re:

[Anonymous Coward]

Let's hope he has milk. Otherwise his poor cat would starve to death.

Re:

[treeves]

Milk is the least of that cat's worries. The guy keeps the poor thing in a sealed box with a cyanide capsule for gosh sakes!

Re:

[sm62704]

There's no milk in my fridge but there are three cats in my house. The cats eat cat food.

What would a single man use milk for? At three dollars a gallon it would be cheaper to feed them gasoline. The only time I have milk in the fridge is when there's a woman living there. And it usually turns into stinky cottage cheese before it's half empty.

Befor you ask, they're my daughter's cats. I got stuck with them when she moved to Ohio with her fiancee.

Re:

[$RANDOMLUSER]

But do we know if Schroedinger has milk in his Fridge without looking?

And does the light go off when you close the door?

Re:

[smooth wombat]

And does the light go off when you close the door?

If you're this guy [olliesbargainoutlet.com], you never have to wonder about that question. (third paragraph)

And for the record, I worked with this guy for a time.

Re:

[Anonymous Coward]

but the colors MUST be false color, since atoms are smaller than light wavelengths.

People say that, but it's not true. How do atomic gases have colour? Oh yes, individual atoms can absorb and emit light with wavelengths many times their own size. Strange but true, E.M. is weirder than people realise.http://amasci.com/tesla/tesceive.html [amasci.com]

Re:

[nguy]

The summary didn't say, but the colors MUST be false color, since atoms are smaller than light wavelengths

It is false color, but it wouldn't have to be. It's possible to probe individual atoms with visible light of different wavelengths using STMs.

Re:Schrodinger's Fridge

[esocid]

I doubt that they still survive the process. Organic cells are destroyed due to the direct irradiation with electrons necessary to produce the "photograph" from the microscope. There are ways around this, such as only focusing the beam on a small part of a specimen or to use a deflection technique that minimally exposes the specimen and deflects the electron beam to the viewing stage. Others are preirradiating the specimens at low doses to stabilize them for increased irradiation. There are other complex techniques outside the realm of my understanding, but I think it still is really tough to preserve organic cells during electron microscopy.

Re:

[sploxx]

The summary didn't say, but the colors MUST be false color, since atoms are smaller than light wavelengths. But will it allow you to photograph atoms without destroying them? (yes the link is humorous, but the question I ask is serious)

No. Nothing says that a single atom can't send or receive single photons. The size of "the EM field belonging to the photon" may be much larger, but so what?
Look here [arxiv.org] for an example.

And, Of Course...

[Wandering Wombat]

The pink atoms won't let the black atoms share a molecule with them.

Proof at last...

[pushing-robot]

So we'll finally know for certain that carbon is black, oxygen is red, nitrogen is blue, and hydrogen atoms really are white.

Re:

[chibiace]

its amazing we can see in such an atmosphere!

Re:

[Actually, I do RTFA]

hydrogen atoms really are white.

So that is why the Hindenburg didn't use Helium.

Re:

[CityZen]

But I heard that the blue oxygen is better for you than the red oxygen.
Too much of the red stuff and your mind starts to close, so they say.

tHE nEW sKITTLES?

[webword]

Taste the Rainbow [imageshack.us] (of atoms)!

Sorry, couldn't help myself. Marketing controls my mind. And yours.

No native CMY support?

[140Mandak262Jamuna]

Unless it supports CYM color maps natively we will be forced to use Photoshop.

Re:

[Farmer Tim]

CMY: apartheid colour space, where Ks arent allowed.

Atoms don't have color!

[ramk13]

At least not how they are implying. Color as most people think of it has to do with absorbed, reflected and transmitted light. The arrangement of the atoms as much as the atoms themselves affect color. But individual atoms in a crystal don't have color, at least as most people understand. The headline makes it seems like you could come away saying, "So iron atoms really are red..." or something equivalently silly.

Re:

[porkThreeWays]

The quantum nature of light is a bitch!

Re:Atoms don't have color!

[Jesus_666]

They just use smaller light, duh!

Re:

[Atario]

But that requires very tiny photons! Have you priced those lately? I'm not made of money! Leave me alone!

Re:

[mikael]

I thought it was more to do with the orbitals of the electrons rather than the atomic number of the atom, and the orbitals of the electrons depend on the crystalline arrangement of atoms, and whether they have been ionised or not. Even different ions of the same atom will have a different
absorption spectrum [google.com] and emission spectrum [google.com]. So no atom has one unique color, but may have a series of wavelengths of light that it can emit [rochester.edu], which our sight would perceive as a mix of red, green or blue wavelengths [uc.edu],

Re:

[Rich0]

Well, it is all inter-related, but color absorption is all about electrons and energy levels. Photon hits electron, electron pops to higher level, electron falls to lower level, electron emits photon.

The energies of those orbitals have everything to do with the sizes/masses/etc of the atoms they're bound to, and the number of electrons around them.

The electrons that matter generally aren't bound to a single atom - they move in molecular orbitals around groups of atoms or larger. If you take a piece of ste

Re:

[raddan]

True, but no one who actually uses one of these would make that mistake. This is pretty cool. Our visual systems are keyed into color differentiation (well, most of us, anyway)-- so it only makes sense to take advantage of that additional visual processing ability to convey more information to the microscopist.

This thread is useless without pics!

[Trigun]

Pic, or it didn't happen!

Re:This thread is useless without pics!

[pushing-robot]

Picture here. [msu.edu]

Re:

[Trigun]

Man, that is so much better than I had expected!

Microsoft Interview

[slapout]

Sounds like a Microsoft interview question: Why are man hole covers round? What color are atoms?

Re:

[geekoid]

A: Because manholes are round!

B: Octarine

Correct answers they don't expect FTW!

Yow!

[$RANDOMLUSER]

A STEM shoots an electron beam through a thin-film sample and scans the beam across the sample in subatomic steps.

Holy crap! And we think 45nm is small!

Made in the USA

[treeves]

The instrument is a new type of scanning transmission electron microscope (STEM), built by the NION Company of Kirkland, Wash....

I lived there when I was in elementary school. More important, a certain warehouse store has its headquarters there. So I wanna know when I'll be able to pick up one of these STEMs at Costco!

Not really a breakthrough...

[achosler]

I've been able to see atoms in color for years, you just gotta light it, and remember to pull the slide out to clear it.

What do the electrons "reflect" off of?

[JoeGee]

Most of the space occupied by the atom is exactly that, space, nothing more. The electron cloud is a fuzzy region of probability, not a solid thing. The "side" of an atom must be defined by a force, not a particle?

Re:

[esocid]

It isn't so much a question of reflection, but more of capturing the excitation of electrons in the atoms that make up the sample by absorbing the irradiated energy. The electrons are excited into higher orbits, which gives off light that the "camera" on this microscope captures and resolves into a cleaner image. That is why organic samples are pretty much goners in EMs. They can't survive that much radiation.

Re:

[ruinevil]

However, the thin slices and embedded heavy metal "stains" are totally conducive to life.

Re:

[esocid]

I forgot to mention, the electrons are in the sample, not shot by the microscope. It uses EM radiation to excite the electrons in the sample.

Re:What do the electrons "reflect" off of?

[kebes]

Most of the space occupied by the atom is exactly that, space, nothing more. The electron cloud is a fuzzy region of probability, not a solid thing. The "side" of an atom must be defined by a force, not a particle?

You're right that an atom is mostly empty space, but that doesn't matter. An electron microscope works by shooting a beam of electrons at the sample, and measuring how many of those electrons are transmitted (this is called a TEM; an SEM works differently). The electrons that didn't go straight through the sample were scattered by the atoms of the material. Remember that electrons are charged: as the incident electrons travel through the atoms there will be very strong Coulomb forces. The incident electrons will be repelled by the electrons in the material. This interaction is 'long-range' by subatomic standards: even though the electrons themselves are vanishingly small, the Coulomb interaction distance is quite large.

To a first approximation, 'heavier' atoms (higher atomic number) will scatter electrons more strongly, since they have more electrons. On an electron micrograph, heavy atoms show up as dark (absorbed/scattered alot of electrons), whereas lighter atoms show up as being bright (most electrons were transmitted).

I'm glossing over many details, of course. The important thing to remember is that the incident charged electrons are interacting with the charged electron density surrounding the atoms in the material.

Re:

[Quadraginta]

The electron cloud is a fuzzy region of probability, not a solid thing.

Ah, the evil remnants of a flawed basic chemistry and/or atomic physics class.

Just FYI -- not that it relates to this article -- this is wrong. So far as we know, an electron is a point particle, and the electrons in an atom aren't any different from a free electron. They are a collection of little points located at various definite positions. There's no "fuzziness" and they aren't "smeared out" in any sense at all. The "fuzzy cloud"

Re:

[Scott Carnahan]

So far as we know, an electron is a point particle, and the electrons in an atom aren't any different from a free electron. They are a collection of little points located at various definite positions. There's no "fuzziness" and they aren't "smeared out" in any sense at all.

I agree that electrons are point particles to the best of our knowledge. However, they are smeared out in the sense that they don't admit position eigenstates, so they are not located at definite positions. If you want to calculate a

Re:

[Quadraginta]

Yes, pretty much. There are some subtleties, of course, since you're working in the quantum regime.

Next up

[Fnord666]

Next thing you know they will have photos showing charm and spin as well! Will wonders never cease.

The FIRST time???

[dos4who]

"...enabling scientists to see individual atoms in color for the very first time."...

Actually, I'm guessing the folks over at NION (the company who built the thing) were the first... Somebody had to test it out, right?

Re:

[$RANDOMLUSER]

Beta-testing is for sissies.

Real Harmonic Color

[Doc Ruby]

I'd like to see these atoms rendered in necessarily false color (they're smaller than visible light wavelengths) that is at least the color corresponding to their size. They're smaller than visible wavelengths, but their actual size is a specific fraction of a visible wavelength. Let's see the atoms colored with the color that's a harmonic multiple.

Or maybe the color should be derived from the "texture" of the atom, just like the actual color of macroscopic materials. If a carbon atom has 12 electrons evenly distributed around a sphere in shells (2, 8 and another 2 in valence), let's see it get colored accordingly. Maybe the inner shell's diameter harmonic color in the visible range, divided by 2 and scaled back into the visible, overlapped with the same algorithm for the outer 8 in the second shell, then again for the 2 in the outermost shell.

The point is that these colors can mean something. And since the number and combination of electrons is so important to the characteristics of the electron, as well as offering the femtoscopic equivalent to macroscopic colored surfaces, I'd like to finally see what I've been imagining since high school chemistry class.

Re:

[Scott Carnahan]

They're smaller than visible wavelengths, but their actual size is a specific fraction of a visible wavelength. Let's see the atoms colored with the color that's a harmonic multiple.

I think you have a failure of uniqueness. If you have an atom with diameter 100pm, there are about three thousand wavelengths of visible light that are integer multiples of that. Which one will you choose?

Or maybe the color should be derived from the "texture" of the atom, just like the actual color of macroscopic materials

Re:

[Doc Ruby]

The multiples of the wavelengths have to be harmonic, 2^n, not just simple multiples. That's how harmonics work. They will be unique: a given femtoscopic distance will have only a single frequency harmonic in the visible spectrum, because the visible spectrum fits within a single "octave", eg from f to (f*2).

I am saying that the false color should indeed represent the electronic structure of the atom being colored. I did make a mistake substituting the atomic weight for atomic number (and therefore nuclear

snorkfud?

[Walking The Walk]

Who tagged this snorkfud, and what on earth does it mean? A google search just hits this slashdot article and a dummy website.

Re:

[Faylone]

Perhaps The Snorks [wikipedia.org] now work for Microsoft?

Screenshot

[peterpi]

Screenshot:

                .

There's a nice AFM technique which does this too

[Sockatume]

Atomic force microscopy (AFM) uses the weak Van der Waals-type interactions between the atoms in a probe, and the surface itself, to measure the locations of atoms. They also developed a qualitative way of identifying the atoms, by measuring the variation of the strength of interaction with probe height. It's not as neat as being able to read real-life energy level information out of atoms, mind you.

Coming to a store near you...

[mojotooth]

The instrument is a new type of scanning transmission electron microscope (STEM), built by the NION Company of Kirkland, Wash

Kirkland? Awesome, that means it should be available at Costco real soon now.

Atoms/molecules really might have colour?

[Twinbee]

General science question: If wavelengths of light are too large to find out the colour of atom or molecule (or 100 molecules), then why can't you use much finer wavelengths to measure, and scale the results up to the range we can see?

DIY STM

[the_kanzure]

$100 - put together an STM [heybryan.org] (or another instrument of your desire; scroll down for the relevant links and text).

Is it true ...

[PPH]

... that the green ones are aphrodisiacs?

Sounds like

[BluenoseJake]

They are using a Heisenberg compensator.