Fastest Semiconductor May Also Be Most Expensive

Scientists have found what they think may be the fastest known semiconductor. Sounds great, right? But it happens to made from one of the rarest elements: rhenium. That rare element combines with selenium and chlorine to form a “superatom.” Unlike conventional semiconductor material, the superatom causes phonons to bind together and resist scattering. This should allow materials that can process signals in femtoseconds,

Rhenium was the last stable element to be found in 1925. It is primarily used in combination with nickel in parts of jet engines, although it is also known as a catalyst for certain reactions. It is very rare and has a high melting point, exceeded only by tungsten and carbon. When it was discovered, scientists extracted a single gram of the material by processing 660 kg of molybdenite. Because of its rarity, it is expensive, costing anywhere from $2,800 to $10,600 per kilogram.

So maybe we aren’t destined to have desktop computers with 100 terahertz processors in them. But maybe we will. While the structure now uses rhenium, it is possible that understanding this effect will lead to new high-speed semiconductors using more affordable materials.

This is one of those areas where we always hear about new materials, yet we rarely see any impact on the actual market. For example, where are our diamond transistors? GaN may be the exception that proves the rule.

26 thoughts on “Fastest Semiconductor May Also Be Most Expensive

      1. Only because there is much less demand for Rhenium. If something really comes from this, the price will shoot through the roof. There is far more gold on this planet than Rhenium

    1. I agree that a kilo goes a long way also. FYI. gold bond wires, gold platings are used throughout high performance semiconductors since inception. Nobody seems to be complaining about the cost of gold increasing the cost of semiconductors. Also, it is my understanding that rhenium will be used as a dopant or as an implant into the semiconductor in selected areas so rhenium cost would be insignificant.

      1. We have almost entirely moved to using copper bond wires specifically because of the increasing cost of gold. Copper is way more of a hassle for a number of reasons, but cost is driving the conversion.

      2. Silver was even “better”, but gold wasn’t vulnerable to oxidation.
        It’s a common misconception that gold is considered best for conductivity.
        In certain RF applications (waveguides), silver (silver plating) still has its place.

  1. “Rhenium was the last stable element to be found in 1925.”
    Oh, so others were found after 1925? Sometimes a comma makes all the difference to the meaning! “Help my uncle Jack, off his horse”. “Truly I tell you today, you will be with me in Paradise”, “Let’s eat, Grandma”.

  2. Not actually very useful.

    We could run current semiconductors at tens of GHz, but they’d melt into a puddle of slag due to a combination of the amount of charge needed to flip a gate, and more importantly the resistance of the interconnects of the chip. We’re capped by power consumption, not fundamental semiconductor speed.

    What we need isn’t a better semiconductor. It’s a better superconductor.

      1. “Electron tubes always were superior to transistors, always will.”

        Except for a few small details like; size, power consumption, ruggedness, automated construction
        B^)

    1. Speed of digital ICs is limited by power consumption in some contexts, and interconnect resistance is not always the worst contributor to heat. Heat is generated as a signal transitions, and some of that heat comes from the FET’s channel. Also, there’s off-state leakage, and a better semiconductor may reduce that.

      If a better semiconductor allows a lower gate capacitance or lower supply voltage, heat generation falls.

  3. Wouldn’t the combination of elements be a “super-molecule”? After all, it’s not undergoing spontaneous fusion.

    And what is this photon-pairing phenomenon? Is it related to relativistic effects seen in the electron shells of heavier elements?

    1. Guess it could be seen as a special case of a molecule. The wording is a bit confusing, but then I’m just a layman. From what I understand it’s something like this: a superatom has a common valence/outermost electron shell, which makes it (in some ways) chemically similar to ordinary atoms with the same number of valence electrons. A superatom with 1 valence electron makes it behave a bit like alkali metals, a 7 electron shell makes it similar to halogens etc. What they call a supermolecule is a bigger molecule that has at least one superatom in the structure.

      1. A quantum of sound or elastic vibration in solid matter.

        The elastic vibration modes of atoms in a lattice are quantized such that the vibrations behave like particles as they move along the lattice. It’s a group effect – a quasi-particle. You can think of it like a tight string that you pluck just so that exactly one wave appears to propagate down the line, except at the atomic level.

  4. There’s an old saying “Speed costs money. How fast do you want to go?” .

    That said, I bet now that they have a working model, other elements/compounds that work just as well.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.