Can We Fabricate FETs From Germanium & Tin?

While studying my end-of-the-year reports on new technologies, I came across one I'd completely overlooked.

Researchers in Belgium are working on a version of the MOSFET that uses a germanium-tin combination (an alloy?) deposited onto the silicon substrate. This work is being done by researchers from the Katholiek University of Leuven, the IMEC research institute at Leuven, and Japan's National Institute of Advanced Industrial Science and Technology (AIST). The reason it's of interest: This method should produce faster FETs (for logic gates) and opto devices (like phototransistors) for communications apps.

Manufacturers already use germanium with silicon in high-speed applications because it increases carrier mobility. But if you add tin to the germanium, you get even more improvement in carrier mobility. From Ruben Lieten of imec, “By adding tin to germanium, we improve both the electron and hole mobility and lower the difference between the direct and indirect transition, so the material gets closer to being a direct bandgap material.”

The folks at imec have figured out ways to improve solubility of tin in germanium and reduce compressive strain and fracturing. They are closing in on this. They are using special low temperature processes to deposit the metal vapors on the substrate as very thin (40nm) layers. The process they are using limits the diffusion of the Ge-Sn into the silicon and allows much higher amounts of tin to be introduced in germanium. The usual method (melting) allows around 1% tin. The imec researchers are getting around 5%, but other researchers have added up to 20% tin. The process results in a very homogeneous Ge-Sn layer.

Once the Ge-Sn is deposited, next is AL2 O3 followed by TaN. They also add NiGeSn metal for the source and drain.

(Source: imec)

(Source: imec)

What they are working towards is very fast p-channel depletion mode FETs. No speed data is available yet.

Other research teams are exploring use of these types of semiconductors for laser diodes and LEDs that operate in the IR region. These will see use in communications applications.

Again, from Ruben Lieten, “Because we have a direct bandgap, we can emit and absorb light much more efficiently. Some researchers are working on LEDs and even laser diodes. You can make the emission source and detector from GeSn and then integrate these together with GeSn CMOS transistors on a silicon substrate.”

Watch for these in the next year or two.

Related posts:

— Brad Albing, Editor-in-Chief, Planet Analog and Integration Nation Circle me on Google+

3 comments on “Can We Fabricate FETs From Germanium & Tin?

  1. RedDerek
    January 11, 2014

    This does sound interesting. I wonder if the cost would be reduced as well – using tin is a lower cost metal, and processing at lower temps would save on furnace operation costs.

    Then making a depletion P-FET sounds difference as there are few manufacturers of depetion N-FETs (no P-FETs).

    The real trick would be to make a depletion FET without a body diode that can take a few amperes of current.

  2. samicksha
    January 13, 2014

    Sounds intersting, tunneling efficiency in silicon is poor due to the relatively large band gap, please correct me in case i am wrong, what i understand from blog is, it may allow for high-performance operation at low voltages.

  3. jimfordbroadcom
    January 23, 2014

    I, for one, am salivating at the prospect of direct bandgap semiconductors processed on Si wafers!  Once the foundries get such processes, I think silicon photonics will explode like an atomic bomb, and I aim to be at ground zero!

Leave a Reply

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