Pepsi Big Blue
August 28, 2009 7:42 AM   Subscribe

Using this technology, obviously scaled up many times, we may eventually be able to photograph many molecules in aggregate. However, I predict that there is a world market for maybe five such machines.
posted by DU at 7:46 AM on August 28, 2009 [7 favorites]

Make that six. This looks pretty sweet.
posted by joe lisboa at 7:48 AM on August 28, 2009

Fantastic. Is there a reason the ends of the chain are so bright? Are the electrons hanging out there?
posted by East Manitoba Regional Junior Kabaddi Champion '94 at 7:49 AM on August 28, 2009

Actually, DU, scientists have used multiple electron micrographs to generate models of larger structures. Check out this electron tomography model of a mitochondrion (QuickTime).
posted by grouse at 7:50 AM on August 28, 2009

Holy shit. I just looked at a molecule... sometimes I feel like I truly am living in some glorious future.
posted by Never teh Bride at 7:52 AM on August 28, 2009

EMRJKC: i would hazard a guess the ends are the electrons floating around the hydrogens.
pentacene wiki
posted by Mach5 at 7:53 AM on August 28, 2009

There are a lot of applications of AFM in the life sciences as well.
posted by Premeditated Symmetry Breaking at 7:56 AM on August 28, 2009

Holy shit. I just looked at a molecule... sometimes I feel like I truly am living in some glorious future.

X-ray crystallography has allowed us to look at molecules for decades, but you have to grow a crystal of it for it to work. This king of complements it.
posted by dibblda at 8:15 AM on August 28, 2009

But do the molecules want us looking at them? Do the ends justify the means?
posted by blue_beetle at 8:23 AM on August 28, 2009 [1 favorite]

What amazes me is how much the actual molecules resemble the ball-and stick model. I have always assumed the models were only a rough approximation of the actual appearance of the molecule (and is "appearance" even the right word to describe something far too small to see with light?)
posted by TedW at 8:35 AM on August 28, 2009 [3 favorites]

Holy crap, those were benzene rings.
posted by haltingproblemsolved at 8:44 AM on August 28, 2009

Really? No one got DU's joke?
posted by Cookiebastard at 8:46 AM on August 28, 2009

Awesome! I am also amazed by the unexpected (why? dunno) correspondence between the image and the ball and stick models. Note that we don't see the the hydrogens hanging off the side like folks on a south american city bus. The benzene rings on the other hand...

(thanks Cookiebastard for making me click DU's link and then get the reference...)
posted by bumpkin at 8:51 AM on August 28, 2009

Really? No one got DU's joke?

Heh, yeah - I just looked at his link. We've come a long way, baybay.

Also. Where's my flying car??
posted by Devils Rancher at 8:53 AM on August 28, 2009

Make that six.

Joe, it looks like you're not aware of this quote and the inherent irony in the comment DU made.

# “I think there is a world market for maybe five computers.” [Thomas Watson, Chairman of IBM in 1943]

On preview: what Cookiebastard said.
posted by DreamerFi at 8:55 AM on August 28, 2009

I am sure Dick Cavett was joking in the page DreamerFi linked to. He could be very sarcastic at times.
posted by Xoebe at 9:04 AM on August 28, 2009

That is so f'ing cool.

Atomic Force Microscopy (AFM) is pretty f'ing cool, too. The articles talk about how it works: You put a specially-designed tip on the end of a cantilever. Specially designed? Oh yeah, they fashion it with tolerances around the scale of individual atoms, and these folks are placing a single carbon monoxide molecule right on the end, oriented with the oxygen atom facing out. I can usually put a piece of thread through the eye of a needle after a few tries. They put a single molecule exactly where they want it in the construction of this tip.

So they have this special tip on the cantilever. Then what? They bounce the cantilever up and down, like tapping a pencil on a desk, while moving it slowly over the surface they want to measure. Based on how far down the cantilever goes on each bounce, they can build a two-dimensional profile of the height of the surface at each point.

Cool! But how do they know how far down the tip goes? I mean, this is the machine that measures distances at that scale, and they can't use it to measure itself, right? I'll give you one guess. That's right, a LASER! (I'm both excited and using the proper capitalization of the acronym.) But not like a normal laser rangefinder, where you measure how long it takes a laser pulse to get to an object and back, using an equation from third grade [ Rate(speed of light)=Time(measured)*Distance(calculated) ]. Nope.

Attach a mirror to that pencil you're tapping on your desk. Shine a laser or a bright light on it. See how the reflection is dancing around on your ceiling or wall now? Same deal in AFM. See how the reflection is moving much larger distances than your pencil is? Greater distance amplifies the effect. So the minuscule movement of the AFM tip can be amplified by turning its rotation into a large deflection at a detector built at a normal scale (not atomic scale). Measure how far the laser reflection travels as it oscillates around on your detector, and you can calculate how far the cantilever tip is moving. The distance is too small to measure directly, but the rotation isn't.

And that's just clever.
posted by whatnotever at 9:06 AM on August 28, 2009 [19 favorites]

(I like that that quote page has Dumb, Really Dumb, Dan Quayle and W.)
posted by DU at 9:11 AM on August 28, 2009

So the minuscule movement of the AFM tip can be amplified by turning its rotation into a large deflection at a detector built at a normal scale...

That's a lot simpler (read: smaller) than what I was imagining, which was an interferometry setup.
posted by DU at 9:20 AM on August 28, 2009

simpler (read: smarter)
posted by DU at 9:20 AM on August 28, 2009

It's absolutely mindblowing to me that this looks exactly like it was thought to, like has been taught since Kekulé figured out this basic structure in 1856! Partly lucky guess, partly amazing deduction, figuring out the aromatic structure was one of the huge breakthoughs that enabled synthetic chemistry.

The history of aromatic compounds is like reading the greatest hits list of chemistry. It was discovered by Michael Farady in 1825. It was first synthesized in 1834 by Mitscherlich, who figured out the ratio of carbons to hydrogens (1:1). This ratio was so deeply weird, so unlike all the other oils containing only hydrogen and carbon, that nobody knew what to make of it. It wasn't until Kekulé that a theory was proposed. His idea of a ring of carbons surrounded by a crown of hydrogens revolutionized synthetic chemistry and was one of the key factors that allowed the production of many of the things we take for granted today: dyes, plastics, drugs to name but a tiny few.

And it was a hypothetical for a long time. Nobody understood how aromatic chemistry worked. We had 80 years of incredibly innovative chemistry based on Kekulé's guess. The structure was finally confirmed by Lonsdale in the 1930's. What aromaticity was and why it worked the way it did was the subject of much scrutiny as well, but, again, it wasn't until Hückel in 1931 that benzene and like compounds were finally put on even the most basic of theoretical footings.

The compound pentacene imaged in the Science article is a big brother of benzene, a polycyclic aromatic hydrocarbon (PAH). They're important chemicals mostly in a negative sense now; they're generally considered highly toxic and carcinogenic and we expend a lot of energy trying to keep their like out of water and food supplies. However, given the long history of benzene and it's relatives as important, unusual but basic building blocks of our present lives, it's choice as one of the first molecules to be so sharply imaged is somehow quite fitting.
posted by bonehead at 9:29 AM on August 28, 2009 [5 favorites]

For TedW and bumpkin and others who are marveling at how much this looks like the models from chemistry class: dibblda nailed it up there. Xray crystallography and NMR spectroscopy have been used as chemical-structure determining tools for a long time. The preceding link says that the symmetric hexagonal structure of benzene was affirmed by x-ray crystallography in 1928.

Chemistry has a very sophisticated understanding of bond lengths and angles. This forms the basis for understanding incredibly large and complicated biological molecules. (This page, which is related to the link in the preceding paragraph, has an interesting review of visualization of complex biomolecules, if anyone's interested. Yet another reason to be grateful for computers!)
posted by Sublimity at 10:39 AM on August 28, 2009

For the record: you could already pretty easily distinguish the hexagonal pattern on graphitic molecules using AFM; you just couldn't reliably say "this atom is right here". I've seen plenty of literature examples that look just like this picture if you ran it through a fairly small blur filter.

(Which isn't to say that this isn't really cool; just that there's a lot of people doing almost this well pretty routinely, and they deserve recognition too)
posted by Dr.Enormous at 10:48 AM on August 28, 2009 [1 favorite]

The electrons on the end are bright, I'd guess, because while they are still locked in their various bonds, they spend more time (probably) at the ends due to repulsion from the electical charge all the way to the left or to the right.

It's a molecular-scale version of St. Elmo's Fire (not the movie), which created a ghostly light around the tips of masts, chimneys, etc., as the electrostatic repulsion of the bulk of a body forced some free electrons to the tips (or anything else with a tiny radius of curvature, far away from the main mass of the object). Just as a guess.

I went to a symposium held by the guys who invented the scanning tunneling microscope, many many years ago. One of the main problems, they said, was isolating the apparatus from all of the vibrations going around them. Even a truck down the street would throw it off, much less someone walking in the room. They had built an elaborate magnetic leviation device, involving liquid nitrogen in a moat, to pull it all off. And then they had to cool it, and protect the pumps, and on and on.

"But in the end," he concluded, "we found out that we could just suspend it all with the equivalent of rubber bands and bungee cords."
posted by adipocere at 10:58 AM on August 28, 2009

Sublimity, it's one thing to have known about this for 80 years or so (I linked to Kathleen Lonsdale's bio above), it's quite another to have seen this directly. I don't think anythone would claim that this one result is going to have a huge effect on the understanding of the structure of pentecene (though there is almost certainly some interesting surface chemistry going on here), but it's the immediacy of the image that's so impressive.

Both X-ray xtalography and NMR are indirect. It's an amazing thing to have a direct visualization of molecules compared to inference by transformation as with those other techniques. It's the difference between seeing a picture and reading about it in a book.
posted by bonehead at 10:59 AM on August 28, 2009

BTW, it's not electrons. One doesn't see the electronic structure (very well) with AFM. It measures the Van der Waal's force, essentially the electronic nuclear potential wells. That's why you see the atomic poistions in the image and not, say the π-cloud of the rings.
posted by bonehead at 11:01 AM on August 28, 2009 [1 favorite]

Wouldn't we have some Pauli exclusion mixed into all of this? How "close" is the scanning tip getting, anyway?
posted by adipocere at 11:06 AM on August 28, 2009

From the paper: The calculations take into account forces of three different physical origins, namely electrostatic forces, vdW forces, and Pauli repulsive forces. Comparing their contributions to the overall force, we found that the electrostatic forces are small (~10%) compared with the vdW forces. These two contributions to the force show little lateral corrugation on the atomic scale and yield a diffuse attractive potential above the entire molecule, giving rise to the observed dark halo surrounding the molecules in the Δf maps. The origin of the atomic contrast is the Pauli repulsion force, which becomes substantial when regions of high electron density overlap. These regions are concentrated to the atomic positions and to the C-C (and to a lesser extent also to the C-H) bonds in the pentacene molecule and are revealed for sufficiently small tip-sample distances (d ~ 5 Å).

We conclude that atomic resolution in NC-AFM imaging on molecules can only be achieved by entering the regime of repulsive forces because the vdW and electrostatic forces only contribute a diffuse attractive background with no atomic-scale contrast.

So, you're correct, it's mostly Pauli-exclusion repulsion that gives the sharp image here. VdW's is too diffuse!
posted by bonehead at 11:30 AM on August 28, 2009

Note that this would be the Pauli-exclusion from the tightly-packed nuclear orbitals where, as the authors say there are "regions of high electron density overlap". So the valence, low density orbitals are just background, part of the electrostatic "smear" in the image.
posted by bonehead at 11:34 AM on August 28, 2009

question: what surface is the pentacene itself on? why isn't the detector picking that up? is it just a regular structure that can be digitally removed or nullified in the final picture? is it floating in midvacuum?
posted by Mach5 at 11:53 AM on August 28, 2009

The substrate is a salt crystal coated on copper (NaCl(2ML)/Cu(111) in surface science lingo). You don't see it in the picture because it's too far away from the tip of the AFM probe to be detected.
posted by bonehead at 12:00 PM on August 28, 2009

Man, living in the future is so cool sometimes.
posted by lekvar at 12:04 PM on August 28, 2009

For the record: you could already pretty easily distinguish the hexagonal pattern on graphitic molecules using AFM; you just couldn't reliably say "this atom is right here".

Well, to be fair, Heisenberg says we still can't reliably say "this atom is right here".
posted by Netzapper at 1:43 PM on August 28, 2009

Wow - we really do have cameras!
posted by nickmark at 2:18 PM on August 28, 2009

Note that this would be the Pauli-exclusion from the tightly-packed nuclear orbitals

Gawd, I love it when you talk dirty.
posted by Floydd at 2:20 PM on August 28, 2009

Heisenberg actually implies that we can say "the atom is right here," you just have to HOLY SHIT! DUCK!

I suppose this is true up to a point. There is the implication that the minimum uncertainty would be:

delta_x >= h_bar / (2 * mass_of_atom *c)

if you assume that the upper limit of speed of the atom whanging off into the distance is c as you measure it with a very, very high-intensity photon or something. Practically speaking, for an atom of carbon we should get:

delta_x >= (6.62606896×10−34 J·s) / (2 * 1.99x10−26 Kg * 2.9979x108 m/s)

delta_x >= 5.5533 x 10−17 meters, or around a millionth of an Angstrom.

We can measure this, but it will be going very fast by the time we're done. Let's build this. I think I smell a grant coming on!
posted by adipocere at 2:54 PM on August 28, 2009 [1 favorite]

Bonehead--man, that's a difficult pseud to type when you're trying to address someone respectably!--I absolutely agree that the AFM pictures are amazing.

I just wanted to clear up this idea that even though we haven't had direct, single-molecule observation before, that the models that chemists use (yes, even in freshman general chem) were somehow guesses or provisional or something.

There is a proverbial shit-ton of experimental evidence about bond lengths and angles and this information underlies a wide variety of sophisticated modern chemical applications.
posted by Sublimity at 4:07 PM on August 28, 2009

How do you control and monitor the movement of the tip relative to the sample so finely that you can know that "now it's moved 0.1 nanometers and vibrates this much more strongly..." which then allows you to draw another pixel? You don't just push it with your finger really carefully, I suppose?
posted by Anything at 5:07 PM on August 28, 2009

This made my jaw drop. Thanks!
posted by obiwanwasabi at 5:31 PM on August 28, 2009

X-ray crystallography has allowed us to look at molecules for decades, but you have to grow a crystal of it for it to work. This king of complements it.

I think it's great to see some chemistry on metafilter, but this image is taken at 5 K. Pentacene's mp is ~300C.

This is a crystal. Beautiful image, but meh, x-ray is cheaper.
posted by peppito at 5:59 PM on August 28, 2009 [2 favorites]

This is awesome.
posted by flippant at 6:20 PM on August 28, 2009

Anything: Certain materials are Piezoelectric, which means that they physically expand and contract based on an applied electric voltage. Watch alarms are usually made of piezoelectric buzzers, which vibrate strongly enough to become audible when an appropriate electric signal is applied.

The control over the expansion and contraction can be very fine, though, because the total change in size of an element depends on the strength of the effect and the size of the piezo element. Wikipedia says, "For example, lead zirconate titanate crystals will exhibit a maximum shape change of about 0.1% of the original dimension." So if you build a piezo element that is ~1000 atoms thick (probably not too hard, because there are technologies that can even grow crystals by depositing ~1 layer of atoms at a time) and control your voltage carefully enough to get 0.001% changes in dimension, you have movement control with a resolution of 1% of the diameter of an atom. Those numbers are just slightly-educated guesses, but they illustrate how we can get movement control at the desired scale.

That's one of the amazing things about electronics. Not only have we harnessed electricity to precisely manipulate information, but we have found ways to use it to precisely manipulate matter with a degree of control we could never hope to achieve with purely mechanical systems.
posted by whatnotever at 9:48 AM on August 29, 2009 [1 favorite]

There's a video press release from IBM labs.
posted by CrunchyFrog at 7:45 PM on August 30, 2009

« Older When this you see, think of me.   |   Moon turns out to be giant tree Newer »

This thread has been archived and is closed to new comments