An excerpt from Charles Darwin’s On the Origin of Species—atom by atom. (Image: Sander Otte et al., 2016/Nature Nanoscience)

Researchers working in the Netherlands have developed an atomic-scale rewritable data-storage device capable of packing 500 terabits onto a single square inch. Incredibly, that’s enough to store every book written by humans on a surface the size of a postage stamp. Holy shit.

This atomic hard drive, developed by Sander Otte and his colleagues at Delft University, features a storage density that’s 500 times larger than state-of-the-art hard disk drives. At 500 terabits per square inch, it has the potential to store the entire contents of the US Library of Congress in a 0.1-mm wide cube. The new system, described in the latest issue of Nature Nanotechnology, still requires considerable work before it’s ready for prime time, but it’s an important proof-of-principle that lays the groundwork for the development of useable atomic-scale data storage devices.


This isn’t the first time scientists have positioned individual atoms at will. Researchers have been moving atoms using scanning tunneling microscopes since the early 1990s, but current methods are tedious and show, requiring tremendous patience and persistence. The new system, while still a bit slow, is a huge improvement in user friendliness.

To make it work, Otte and team placed chlorine atoms on a copper surface, resulting in a perfect square grid. Importantly, a hole appears on this grid whenever an atom is missing. As we all know, this type of on/off type configuration lends itself well to binary switching—the foundation of digital data storage. Using the sharp needle of a scanning tunneling microscope, the researchers were able to probe the atoms one by one, and even drag individual atoms towards a hole.

“The combination of chlorine atoms and supporting copper crystal surface that we found now, combined with the fact that we manipulate ‘holes’—just as in a sliding puzzle—makes for a much more reliable, reproducible and scalable manipulation technique that can easily be automated,” explained Otte to Gizmodo. “It is as if we have invented the atomic scale printing press.”


The comparison to a sliding puzzle is apt. Every bit consists of two positions on a surface of copper atoms, and one chlorine atom that can be adjusted back and forth between these two positions. When a chlorine atom is in the top position, and there’s a hole beneath it, it’s a 1. Reversed, the bit is a 0. Voila, instant hard drive.

Each chlorine atom is surrounded by other chlorine atoms, which helps keep them in place, except near the holes. This method makes it much more stable than methods that use loose atoms. Using this technique, the researchers were able to perform write, read-out, and re-write operations in a one-kilobyte device comprising 8,000 atomic bits. It is by far the largest atomic structure ever constructed by humans.

Image: Sander Otte et al., 2016/Nature Nanoscience

The researchers organized memory into blocks of 8 bytes (64 bits). Each of these blocks was assigned a marker, constructed from the same types of “holes” as the raster of chlorine atoms. Similar to QR codes, these markers function like miniature barcodes that carry information about the precise location of their block on the copper layer. These markers can also indicate when a block is damaged, say on account of a local contaminant or an error on the surface. This means that memory can be easily scaled up to a big size, even if there are physical deficiencies in the copper surface.

An excerpt from Richard Feynman’s lecture, “There’s Plenty of Room at the Bottom.” (Image: Sander Otte et al., 2016/Nature Nanoscience)

During the experiment, the researchers preserved the positions of more than 8,000 chlorine “vacancies,” or missing atoms, for more than 40 hours at 77 kelvin (more on this in just a bit). After developing a binary alphabet based on the positions of the holes, the researchers stored various texts, including physicist Richard Feynman’s seminal lecture, “There’s Plenty of Room at the Bottom,” and Charles Darwin’s On the Origin of Species. This data was stored atom by atom, bit by bit, on the surface of the copper sheet. The ensuing write/re-write speed was relatively slow—on the scale of minutes—but the demonstration showed that it’s possible to reliably write, store, and read data at the atomic scale.



“While the memory outperforms existing media by far in terms of capacity, it still stays far behind in terms of read/write speed,” Otte told Gizmodo. “However, I foresee no physical boundaries that will prevent us from speeding up these processes to similar speeds that are currently seen in [hard disk drives]. It will be a technological challenge for sure, but in terms of physics it should work.”

An important caveat, however, is that this system cannot function in an everyday environment. In its current form, the atomic hard drive can only operate in clean vacuum conditions and at liquid nitrogen temperatures, which is -346°F (-210°C). As Otte admits, “the actual storage of data on an atomic scale is still some way off,” but “through this achievement we have certainly come a big step closer.”


The fact that this system only works at liquid nitrogen temperatures may seem like a deal breaker, but Otte disagrees. Now that his team has found that this particular combination of chlorine and copper provides a balance in terms of stability and manipulability, a logical next step would be to look for similar but different atoms such as iodine or bromine, and see if that enhances stability and thereby operating temperature.

“But even if that would not work, it would not be inconceivable to have data storage solutions at nitrogen temperatures in larger data centers,” Otte said. “Many MRI scanners in hospitals are kept at helium temperature permanently, so in that sense it is not as extreme as it may sound at first.”

Otte also sees this as a general breakthrough in the field of nanotechnology, saying it’s really just a demonstration of our new abilities to engineer the world on the smallest possible length scale. “I cannot at this point foresee where this will lead, but I am convinced that it will be much more exciting than just data storage,” he said.


[Nature Nanotechnology]