Researchers at the University of Alberta have created a new atomic-scale solid-state drive (SSD) that could potentially surpass the storage memory of conventional drives 1,000-fold. By developing a more efficient form of a procedure known as hydrogen lithography, the team was able to manufacture a room-temperature stable atomic structure capable of storing an estimated 1.1 petabytes per in2 of the material (1015 bytes). According to Roshan Achal, a Ph.D. student in the physics department and lead author on the research, the new technology would essentially allow one to store Itunes’ entire 45 million song library in an area the size of the surface of one quarter.
This incredibly tiny scale is a testament to the astounding evolution of computer storage systems through history. The first computer hard drives were created by IBM in 1956 for their RAMAC computer system. The RAMAC system was the size of two refrigerators and was so heavy that it required a forklift to be moved around. despite being able to store only a meager 5MB of data. The past 50 years have seen an exponential increase in the storage capacity of memory devices and a corresponding decrease in the size of said devices. Nowadays, consumer brand laptops routinely contain 1-6 terabyte memory hard drives and small flash drives with over 50 gigabytes of storage space can easily fit in your pocket. The new research seeks to take data storage to novel heights by creating memory devices that exploit atomic level micro-structures to encode even larger amounts of information in even smaller spaces.
The procedure developed by researchers is important in that it opens up new potential avenues for the efficient construction of atomic-scale devices. Since the first demonstrations of atomic manipulation, there has been much interest in creating practical atomic-scale devices. Past efforts, however, have been limited by 2 primary constraints: precision and stability. Constructing error-free atomic-scale structures requires an extremely high level of precision and many structures tend to not be stable enough at room temperature to be of practical use. Most techniques that exist require a lot of time and resources, and so are unfeasible for commercial manufacturing.
How to Make Atomic-Scale Devices
Efforts to create atomic memory devices typically utilize a technique known as hydrogen lithography. Hydrogen lithography is the process of removing hydrogen atoms from a hydrogen-passivated surface, such as silicon. Essentially, hydrogen lithography allows one to “etch” atomic-scale patterns into the lattice structure of passivated materials. By firing low energy electrons at an H-Si bond, the hydrogen atom can be removed, leaving behind a vacant valence of the silicon atom.
These vacant valences are known as “dangling bonds” as they consist of an exposed electron bond. By removing single hydrogen atoms one can create a complex dotted structure of dangling bonds on the surface of the silicon. These chains of dangling bonds act as transistors and can be used to store information, exactly like a traditional SSD. Because these chains of dangling bonds are extremely small, a large number of them can be present in a very small surface area.
Error-Correction in Atomic Manufacturing
One of the practical problems facing hydrogen lithography is the extremely small margin of error allowed. Even small imperfections in the dangling bond structure can compromise functionality and the removal of any incorrect hydrogen atoms can make a structure inoperable. What has been needed is an efficient way to correct any mistakes that an automated lithography tool may make. So, the team of researchers has developed a new state of the art technique called hydrogen re-passivation.
In the new study published July 23rd, researcher report that they have come up with a method for precise and efficient error-correction in constructing atomic structures. The described procedure is as follows: First, a hydrogen atom is attached to the end of an application tip via a process known as hydrogen functionalization. Next, the application tip is positioned over the desired site. The application tip is lowered and the hydrogen atom is transferred from the application tip to the silicon atom by small modulations of voltage. The device detects for a tunneling current to signify a successful re-passivation, then resets. Feedback algorithms ensure that the machine is able to reposition itself over the appropriate site if the needle is unintentionally moved.
Most importantly, the process is very efficient as the whole procedure takes approximately 1s and each application of hydrogen to the tip serves for 3-5 re-passivation events. Other existing methods of hydrogen re-passivation can take up to 10 seconds per bond and require one to reapply hydrogen after every re-passivation event.
Previous forms of atomic storage units have typically only worked at extremely low temperatures, but the dangling bond structures made by the team are fully functional and stable at room temperature. To prove the applicability of their findings, the researchers used their new technology to create rewritable 8-bit and 192-bit atomic memories, in which they encoded the entire ASCII English alphabet and the main theme song from Super Mario Brothers.
Current calibrations on the procedure allow the team to create dangling bond memory drives with an information density of 1.70 bits per nm2. For scale, this would be roughly the equivalent of inscribing 350,000 letters across a single grain of rice. The team also reported no changes in the atomic structure of the memory devices after storing them for 3 days at room temperature.
Although these procedures are still in their infancy, the team has demonstrated that their technology has readily available practical applications. The researchers are confident that their work has the potential to change the tech industry by paving the way for the manufacture of practical atomic-scale devices. Next for the team is to integrate a fully automated re-passivation process into automated passivation procedures and devise faster ways to read and write memory to the devices. Said Robert Wolkow, professor of physics at University of Alberta, “With this last piece of the puzzle in hand, atom-scale fabrication will become a commercial reality in the near future.”