Tech Forums & Message Boards. Discuss Newest & Greatest Tech Trends. › Tech Forums & Message Boards. Discuss Newest & Greatest Tech Trends. › General Talk › Chemistry and crystallography, not aliens
MemberSeptember 9, 2021 at 4:54 pm
The researchers’ most surprising discovery was that the thinner the original gold layer, the faster the eutectic circles expanded. The reaction rate when the gold layers were only 20 nanometers thick was more than 20 times faster than when the layers were 300 nanometers thick. And while at first glance the dimensions of the gold and silicon squares inside the circular denuded zones seemed variable, there was in fact a strict relation between the size of the square and the size of the circle: the radius of the circle was always the length of the square raised to the power of 3/2.
How did the squares get there in the first place? They originated as weak spots that were the sources of the spreading eutectic gold-silicon circles; when the circular eutectic was ruptured the squares filled with the same eutectic, which remained at the centers of the denuded zones. As they cooled, the gold and silicon within the squares separated, leaving sharply defined edges that were pure silicon; the centers were more roughly outlined squares of pure gold.
By slicing through the silicon/silicon dioxide/gold layercake and looking sideways at the structures with an electron microscope, the researchers found that the surface squares were the bases of inverted pyramids, resembling teeth penetrating the thin silicon dioxide layer and embedded in the silicon wafer. The squares were square, in fact, because of the silicon’s orientation: the substrate had been cut along the crystal plane that defined the base. The four triangular sides of the pyramids lay along the low-energy planes of the crystal lattice and were defined by their intersections.
What began as a puzzling phenomenon reminiscent of “The X Files,” if on a considerably smaller scale than the cosmic, the mystery of the “nanoscale crop circles” eventually yielded to careful observation and theoretical analysis – despite the obstacles posed by high temperatures, nanoscale sizes, instabilities of the liquid state, and extremely rapid time scales.
“We found that the reaction rate in forming small-sized gold-silicon eutectic liquids – and perhaps in many other eutectics as well – is dominated by the thickness of the reacting layers,” says Wu. “This discovery may provide new routes for the engineering and processing of nanoscale materials.”