Scientists capture MRI scans of single atoms | Hacker News Reader

Scientists capture MRI scans of single atoms

nytimes.com

145 pointsdigital5544 comments

I love how amateur the physical machine looks, with all of that aluminum foil stuck to it and the wires falling about. This immediately reminded me of how my codebase looks whenever I am doing a proof-of-concept just to show someone that something is possible. "Let's get it working before we make it look pretty" -- what I imagined they said.

A bit like the first full-size MRI machine in fact: https://en.wikipedia.org/wiki/History_of_magnetic_resonance_...

This is the 2015 preproduction of the original machine: http://www.ghat-art.org.uk/mark-1-the-worlds-first-whole-bod...

Gosh, if they didn’t have the pretty shrouding no one would ever get inside.

Even includes what appears to be a can of orange pop. Seriously. Placed sideways, near-end of the table. Weird.

Actually, that's an Arizona Iced Tea: https://i1.wp.com/sanpedrosupermarket.com/wp-content/uploads...

Wow, good eye. Looks like Arizona iced tea can perhaps for emi shielding?

I saw that too... pretty funny looking.

More silly experimental physics hacks:

* Hand drill: https://youtu.be/FmmNRaKpBTI?t=1812

* Bridgeport mill: https://youtu.be/FmmNRaKpBTI?t=1852

many physics experiments look like this. I've always been bothered by the use of aluminum foil for insulation but it works. a lot of the parts would have been precision cut using CNC, etc.

Aluminum foil is used on a bunch of spacecraft as well. The Voyager spacecrafts use aluminum foil bought at grocery stores (and then laid out in sheets and thorough scrubbed with bleach).

ha, even better is Kapton tape. It's weird using it on my 3d printer then seeing it all over the moon lander https://www.quora.com/Why-does-the-Lunar-Landing-Module-look...

I spent years using the Department of Energy's giant X-ray generators, and foil wrappings are extremely common at the experimental endstations. My favorite however was the scientist who custom-built a robotic arm that for many years used wooden pencils as part of the gripping hand. (I'm still not sure why he used pencils, except they were exactly the right size and physical properties. The robot worked, anyway.)

Yessir. A lab I worked in had what had to be a $10M MBE, half covered in aluminium foil.

Just to clarify, the aluminum foil is for thermal insulation used for keeping the microscope warm?

Al-foil is ubiquitous in ultra-high vacuum labs because part of the process of reaching this vacuum level involves baking the whole chamber at 120-150dC for 24+ hours. The chambers are typically made of stainless steel which has a pretty mediocre thermal conductivity. Hence the Al-foil, to ensure even heating during the bake. Since you have to bake every time the chamber is opened, it’s easier to just leave the foil on.

It's amateur-looking because it's research - they're iterating on a new idea.

This is like a physical manifestation of what we come up with during an all-night Hackathon.

That was exactly my point.

It is basically an alpha test of new scientific instrumentation. It's obviously very common to see non-blackbox instrumentation in research labs.

Physicist here. True physical experiments are almost by definition one-off machinery, being built and hacked on as the experiments are conducted. The idea is, to measure something that hasn't been measured before. So basically there is no off-the shelf device for this. Designing an experiment means designing the machinery to conduct it. There are some standard components like oszilloscopes, at the core there is probably a custom machined part like a vacuum chamber, the rest is made out of duct tape, aluminium foil and hot glue :).

I worked with Dr. Lutz about 20yr ago, helping to make some of the early magnetically tipped cantilevers for the effort at single atom NMR. It's interesting to me that the NYT shifted the nomenclature to MRI... although with microscopic scanning that is what they can achieve. It's not that weird because (at the time) the tip magnetic field caused a torus to interact with the cantilever when the atom(s) resonate, so they do have to deconvolve the effect to localize the atom and I assume that they use force feedback control to maintain position much like STM (Tunneling Microscopy). I probably would have called it something like NMM (Nuclear Magnetic-Resonance Microscopy), but hey I didn't make it work!

I'd note that the same group headed by Lutz achieved single electron ESR with a similar scanning methodology at least 15 years back so this was follow on work. It's still amazing and incredibly difficult... just imagine how you would produce something (a test sample) that you know will have only 1 magnetically resonant atom in the affected scan area!

At the time the hope was to use single atom NMR to image DNA, but I think that has largely been accomplished by other methods.

edit: The paper in Nature for those with access. https://www.nature.com/articles/s41567-019-0573-x

That's fascinating, and I deeply admire the kind of science that involves developing new tools to find answers.

Can I ask about the rationale motivating attempts to image DNA specifically - was it mainly that we were attempting to get better spacial understanding of the different molecular conformations it enters at different stages of the cellular cycle?

Or imaging DNA transcription? Or DNA polymerase in action?

or was imaging DNA this way an attractive goal because we already anticipated some particular applications?

imaging dna in this way should give high resolution, but if I understand the experimental conditions, they're not physically relevant, so would not be able to look at the molecular conformation of cellular DNA.

The imaging depth that I was aware of at the time (um) would have allowed imaging with the DNA is solution... so ostensibly introns and other physical attributes could have been imaged. Also 20 years ago we didn't have good automated shotgun techniques so everyone was trying to solve that problem.

The Science History Institute in Philadelphia has some neat early lab equipment like this. It's a free museum and much better than the nearby Liberty Bell if you're into this kind of stuff. I highly recommend it. The focus is more chemistry, particularly with biomedical applications.

“Let’s get it working before we make it pretty”, up to the “before” defines experimental physics.

This reminds me of Feynman’s observation of the labs of physicists working on early cyclotrons and how they were an outward expression of the philosophy of the scientists working on them. I’m sure he would approve of this setup and approach.

This is a weird title, Nuclear Magnetic Resonance technology was first used in the field of Chemistry for identifying the structure and content of molecules. The medical industry was later able to adopt it, under the name MRI, for medical imaging. This should really be considered an extension of NMR.

Note that the medical industry chose MRI because it hides a word that is so taboo they don't even want to see it mentioned anywhere: "Nuclear", from Nuclear Magnetic Resonance, even though it has nothing to do with radiations in this particular case.

Not so sure medicine is afraid to use the word "nuclear". Nuclear medicine a field that patients deal with regularly from getting bone density scans for osteoporosis, thyroid and cardiac imaging, etc. In the US at least the name of the department is plastered on signs in hospitals next to everything else. As for why "nuclear" was dropped from I don't know the exact reason.

>> Note that the medical industry chose MRI because it hides a word that is so taboo they don't even want to see it mentioned anywhere: "Nuclear"

> Not so sure medicine is afraid to use the word "nuclear".

MRI was originally called NMRI (nuclear magnetic resonance imaging), but "nuclear" was dropped to avoid negative associations.

https://en.wikipedia.org/wiki/Magnetic_resonance_imaging

Probably because while nuclear medicine and most of radiology uses actual ionizing radiation for almost everything... NMR is ironically one mode that doesn’t involve it at all. Per informed consent, patients ought to know that XRay and CT and nuclear studies involve radiation, it would be unethical to hide that. But MR shouldn’t have an unnecessary image problem.

Nitpick: modern bone density scans use X-rays, and don't typically fall under nuclear medicine.

I mean RF is an important part of NMR and it's radiation.

I had an MRI scan last year, and not a single label or sign in the hospital said "MRI", they all sais "NMR". I find it hard to imagine that the use of the word 'nuclear' would deter hospitals from using it, given their heavy use of x-rays and various radioactive isotopes. But who knows, maybe my hospital is the exception here.

The technique described here differs a great deal from traditional chemical NMR. That NMR is a spectroscopy technique where the signal includes contributions from huge numbers of atoms. Molecular struture is inferred from the spectra measured by traditional NMR, not viewed in an image. This article describes an imaging technique where images of individual atoms are obtained by a reconstruction algorithm that depends on the imaged volume being in a magnetic gradient - somewhat like a hospital MRI scanner.

I'm not much in physics. Could someone explain what the picture actually shows ? I understand that there are 4 shots of an atom. I see that globally there are for spheres. But I don't understand the yellowish shapes...

What you're seeing is effectively a heat map of electromagnetic intensity, if I'm reading the figure correctly.

Does this mean we can increase the resolution of mris to the atomic level? That would be amazing for studying the brain.

Not an expert, but from my read they built this basically by combining an MRI sensor with a scanning tunneling probe - so it's set up kind of like atom-sensing atomic force microscopes (https://en.wikipedia.org/wiki/Atomic_force_microscopy). It doesn't seem like this technique would give you much depth penetration inside the body for one thing, and it'd be tough to process across human scales in a reasonable time. There are a lot of atoms in those 1mm-ish voxels of a medical MRI. It might be useful for purposes similar to AFMs or electron microscopes in the future, though.

Typo I think:

"disrupting the electrons (rather the protons, as a typical M.R.I. would) within each atom."

Should be:

"disrupting the electrons (rather THAN the protons, as a typical M.R.I. would) within each atom."

Couldn't read because of the paywall, then also couldn't read in private mode (because it detected it - fun). To be fair I don't even want to read it, I just want to see the picture.

Here's the picture (hopefully you can access it):

https://static01.nyt.com/images/2019/07/01/science/01TB-ATOM...

And bonus pic of the machine:

https://static01.nyt.com/images/2019/07/01/science/01TB-ATOM...

Funny thing; they hide the article after loading with javascript, so disabling it will allow you to read it normally.

Reply on news.ycombinator.com