Tuesday, May 5, 2015

Perfect film for the chip of the future

Chip developers know that the time is not far off, as their raw material silicon will reach its limits. The probate recipe, more and more, always accommodate smaller components to the processors, then almost only deliver Committee. Alternative materials and alternative principles must therefore forth.

Particularly promising is considered a class of certain metal compounds with the best known representatives of molybdenum disulfide. Similar to the "miracle material" graphene They can be prepared in stable thickness layers only a few atoms. However, in contrast to carbon-based graphene they reveal a trait that she is predestined for use in computer chips: They are like silicon semiconductors, ie, their electrical conductivity can be changed within a split second. That's why they come, at least from a theoretical point of most likely for use in chip into consideration.

But their semiconductor property is just one of the features associated with these "transition metal dichalcogenides" points. Thanks to some exotic properties they exist also for alternative computer concepts, such as on the basis of electron spins (spintronics) or other atomic phenomena (Valleytronik, Orbitronik). Some scientists see their applications on the other hand more in optoelectronics, high-speed communications or alternatively as data or energy storage. And researchers have long not yet explored the full range of possibilities offered by these materials.

Nanolayers in Huge
One application was previously opposed mainly the disabled manufacturing process. He provided the scientists little more than small pieces of material. Such samples are usually wins by peeling method - of larger multi layer blocks most precise individual layers are removed. Using aggressive starting chemicals can also produce films by deposition from the gas phase. But experts say it is unlikely that with the hitherto existing methods pops useful for the computer industry.

Construction of molybdenum disulfide
Transition-metal dichalcogenides such as molybdenum disulfide are not true 2-D-materials, because their molecules form other than the planar graphs a three-dimensionally structured monolayer in which the Calkogen, here the sulfur, protrudes from the plane.
More promising sounds a new procedure that has a team to Jiwoong Park from Cornell University in Ithaca now presented. Transition metal dichalcogenides always consist of a metal such as molybdenum or tungsten, as well as one of the three elements sulfur, selenium or tellurium. Their molecules together form a three-layer structure in which the metal is flanked at the center of the overlying and underlying second element.

Park and colleagues have now succeeded to produce such precise three atoms thick layers of molybdenum disulfide (MoS2) and tungsten disulfide (WS2) - not in micron size, but in round slices of rich ten centimeters in diameter. They are therefore more than 100 million times wider than it is thick.

"There are a lot of people who are trying to breed such individual layers on a large scale, including me", the materials researcher Georg Duesberg says from Trinity College in Dublin against "Nature". "As it is, who it now really done."

Hot full steam process
The new method essentially corresponds to the conventional, ie the so-called chemical vapor deposition, however, includes a central gimmick that is to take other gaseous starting materials than previously customary, explained Park. The corresponding molecules (Mo (CO) 6 and W (CO) 6 and S (CH2CH3) 2 as sulfur supplier) each contain only one atom of the desired substance. With these gases, the layers are grown in 26 hours under 550 degrees Celsius on a silica-coated silicon wafer. As proved decisive, while continuously supplying hydrogen gas to prevent carbonaceous impurities in the final product.

Researchers used now gone one better and produced on this blank a total of 200 field-effect transistors, a standard component of modern electronic circuits. The review found that only two of them were not working as intended, what they gave a success rate of 99 percent. It shows that an error-free layer had grown to almost the entire surface.

The new computer era
According to researchers, it should be possible to cover a layer with a layer of silicon dioxide and again to undergo the process, so that a sandwich structure. One directed to the appropriate places a conductive links between the layers, to three-dimensional circuits would be realized, explained Park and colleagues. Also experiments with laminates from MoS2 and graphs or boron nitride have already delivered promising results.

Process too "inflexible"?
Before too much euphoria, however, warn Tobi Marks and Mark Hersam of Northwestern University in Evanston, Illinois, in their otherwise very positive accompanying Comment: The practical application still stood against a number of obstacles. Thus, the charge carriers in the three atoms thick layers move more slowly than in a conventional example silicon or even graphene, which theoretically limits the computer's speed. Also comply with their electronic properties not yet completely the wishes of chip manufacturers that indeed must, for example think of the energy consumption of their devices.

Due to the high production temperature, the layers may also not be created on flexible carriers. Here are the ultra-thin 2-D materials just play to their strengths here. However, there may be would find ways to create the MoS2 on silica and then transfer them to new surfaces, write Marks and Hersam.

And who knows, maybe the true "killer app" of these materials is so in an entirely different field. An exotic effects no shortage. According to Marks and Hersam to grain boundaries can be moved within the layer of electric shock, bringing the lead resistance can be deliberately manipulated - it creates a "memristor", and from there it is practically only a very short distance to the electronic nerve cell in the computer.

No comments:

Post a Comment