Tungsten provides a minimum resistance path for nanodiode

       As microchips become smaller and smaller, their size is interconnected with copper, resulting in an increase in resistivity at the nanoscale. Solving this upcoming technological bottleneck is a major problem in the semiconductor industry.



   It is desirable to solve this problem by changing the crystal orientation of the interconnect material to reduce the resistivity size. Two researchers from Rensselaer Polytechnic Institute conducted an electron transport measurement in epitaxial monocrystalline tungsten (W), a potential solution. They carried out the first-principles simulation and found the orientation-dependent effect. And they found that the anisotropic effect of resistivity is most pronounced between the layers with two specifically oriented lattice structures, namely W (001) and W (110). This work is also published this week in the Journal of Applied Physics.



   Author Zheng Peng-yuan pointed out that the 2013 and 2015 International Semiconductor Technology Roadmap (ITRS) have called for the use of new materials to replace copper as an interconnection material, thereby reducing the size of the scale to limit the increase in resistance and minimize power consumption and signal delay.



   In their study, Zheng He co-author Daniel Gall chose tungsten (ie, tungsten replaced copper) because it had an asymmetric Fermi surface (ie its electron energy structure). This makes it a good candidate to prove the anisotropic resistivity effect on a small scale. "The bulk material is completely isotropic, so its resistivity is the same in all directions, but if we have a thin film, the resistivity will change much," says Gall.



   In order to test the best direction, the researchers epitaxially grown W (001) and W (110) films on the substrate and immersed in liquid nitrogen at 77 K (about -196 ° C), room temperature or 295K, respectively Both resistometric measurements were made. "The difference in resistivity between the 001 orientation and the 110 orientation is about two times, but they find a smaller resistivity in the W (011) layer," Gall said.



   Although the measured anisotropic resistance effect is consistent with the calculated value, the effective average degree of freedom in the film experiment (the average distance that the electrons can move before the scattering to the boundary) is much greater than the theoretical value of the massive tungsten.



   "The electrons crossed the diagonal wires, bumped into a surface, spread out, and then proceeded before they hit something else," said Gall. "These things may be the other side of the wire or the lattice vibration. This pattern is wrong for small wires. "



   The experimenter believes that this can be explained by the quantum mechanics process that produces electrons on these limited scales. As the layer thickness decreases, electrons can contact both sides of the wire at the same time or undergo an increased electron-phonon (lattice vibration) coupling, which is likely to affect the search for another metal to replace the copper interconnect.



   "The advantages of rhodium, iridium and nickel may be smaller than predicted," says Mr. Cheng. "The results will become increasingly important because the demand for interconnection for quantum mechanics is becoming more common.



The research team is continuing to explore the anisotropic size effects of other metals, while the metals are to have aspheric fermions, such as molybdenum. They found that the orientation of the surface relative to the orientation and transport direction of the surface is critical because it determines the actual increase in resistivity as these dimensions decrease.