They say it’s a picture of atoms, but what are the atoms: the glowing yellow balls or the entire meatball including the darker red? If it’s the meatballs, then why do some have apparently two nuclei?
Here’s the public press release: https://news.cornell.edu/stories/2021/05/cornell-researchers-see-atoms-record-resolution
Here’s the actual scientific article: https://www.science.org/doi/10.1126/science.abg2533
Not 100% comparable, but synchrotron XRD allows for real-imaging of solid state chemical reactions and can, in a sense, resolve the unit cell structure of the crystal. However, what you get from an XRD is nothing like this “photo-like” image, but a diffractogram. I think you could probably re-create an image like this from a 2D diffractogram though, but I’m not sure.
I could be wrong, but I think XRD requires very pure crystals of sufficiently large size. That can help you ascertain the structure and composition of something you can synthesize and crystalize, but I don’t believe xrd can image specific regions of interest like single doping sites like this article talks about.
Synchrotron XRD also has a major drawback in having a very significant equipment requirement that requires being able send samples away for analysis at a dedicated facility. That puts limits on sample preparation and stability time, as well as sharing beam time with lots of other groups.
It’s been a while since I’ve read up on synchrotron xray diffraction though, so there could be workarounds for some of those limitation or I could be misremembering details.
You’re definitely correct that getting sychrotron time is hard :(
On the first part though: Yes and no. XRD will tell you about things like strain and unit cell size distribution, so in that sense, you can’t resolve a single doping site. On the other hand, if you have a reaction going on, or some dopant diffusing into your sample, synchrotron XRD is powerful/fast enough that you can “film” how the crystal structure changes in real time. That “film” will be a kind of average of many sites, but can still be focused to a relatively small region (don’t remember exactly how small off the top of my head, but I believe we’re talking nm-scale).