Myfab

Realize your nano vision

Clean Room Galore!

Last Updated Mar 2019
By: Jörgen Städje

Anything that can be put on a silicon chip is up for grabs to the researchers at the Electrum Laboratory at KTH! With those words, Associate Professor Per-Erik Hellström starts up his guided tour of the Myfab clean room in Kista, right north of Stockholm.

Space-people clad in white plastic suits, working behind hermetically sealed windows, seem like an immensely expensive experiment. It doesn’t look like something that would benefit society as a whole, nor the industry or the university’s sponsors. One can’t even make out what they’re doing. You only know it’s expensive.

Is the material exotic or strange? No. Most of the Earth’s crust is Silicon. The Silicon wafers consist mainly of refined beach sand.

It just happens to be the other way around. The Myfab clean room at KTH in Kista is used by both students and industry. The industry likes to run its production there, because of the high reliability. It makes for consistent, high-quality products, which makes the industry an eager sponsor. If they weren’t sure the results would always be good, no one would like to use the facilities. Also, the students working in the clean room will get world class education. In light of this, Sweden appears to be a leading edge nation, resulting in pure profits for both KTH and the industry.

But, a university, you say? That would make most of the workers students? Not at all. Roughly half of those working in the clean room on a daily basis represent various startups, producing essential components for Swedish industry and export. The advanced equipment is used for academic research as well as for startups and established companies’ series production. Truth is, half the work hours, equipment hours and revenue is generated by businesses.

– Myfab is, and will continue to be the best in Europe, especially concerning high-class lab equipment. High class equipment is reliable. No one beats us when it comes to reliability. The most important thing is that everything always works and always gives the same, predictable results. Down to the atomic level. It doesn’t matter how expensive machinery you have, if you can’t trust the whole process, it is useless to the researchers, Prof. Hellström says.

There are four Myfab labs spread around the country, in Gothenburg, Lund, Uppsala and Kista. Of course they do not have the same set of equipment, because that would mean useless duplication. However, there is some overlap, enabling the labs to help each other out. Anyone can apply for access to the machinery, whereuopon time and resources will be allocated after due academic consideration.

Prof. Per-Erik Hellström examining electron micrographs

The micro electronics developed in Kista is nowadays used in everyday electronics around us, such as mobile phones, TV sets, fibre optic communication networks, in hospitals, etc. Everything developed at Myfab is a real boon for society as a whole.

Take the IRnova Company, for example. They manufacture infrared sensors for heat and gas cameras, and sell their sensors to, among others FLIR, a major manufacturer of heat cameras worldwide.

Everything Must Work! Always!

The innovation lies in the combination of all the available equipment. It’s not one single machine that does all the work. A record-breaking electron microscope is in itself useless, if you don’t have working washing machines, dryers and inspection equipment, although the latter may not seem so advanced. For an outsider, this relationship may seem somewhat bewildering. Making an advanced product, could require some 45 process steps and the important thing is that all of these steps always work and always produce predictable results.

Completely un-amazing, low-tech equipment: a washer. But it too has to work, has to be updated and maintained for all the other processes to work properly.

Wide Field of Use

One might think that the clean room at KTH would be useful for one type of components only, such as transistors. That’s not true. Myfab is part of a lot of research projects simultaneously. The lab happens to be very popular with researchers in biological applications, micro electromechanical systems, silicon carbide semiconductors, heat camera sensors, medical nano-structures, fibre optics, graphene, and more. All of these use the same universal tool: the four-inch Silicon wafer.

Of course it’s expensive to run the Myfab. That’s why it’s important to spread the costs and have as many users as possible. And Myfab has succeeded brilliantly. An enormous amount of projects are running together. The users are virtually crowding and elbowing each other around the machines.

Myfab has decided not to use the very largest wafers, which would otherwise give best yields in mass production. Both ordinary Silicon and Silicon Carbide is available in four-inch sizes, making the same machines available to both areas of research. This means that the manufacturing processes can be standardised, again resulting in higher yields.

A Closer Look

Let’s look into one special project, the Silicon Carbide.

Power electronics in the Swedish X60 rail cars is based entirly on Silicon Carbide to up the efficiency and effective use of energy. Image: Udo Schröter, CC BY-SA 3.0

Research at KTH isn’t something that might perhaps yield some result sometime in the distant future. On the contrary. The researchers have long been making components that are in daily use in the industry. Silicon Carbide transistors have meant profound change for the power component manufacturers, being used in high-voltage equipment for power transmission, power converters for wind generators, power components for electrical trains and cars, frequency converters for efficient motor drive in the industry and at home, built-in measurement equipment extending the life span of power generating gas turbines, and last but not least, computing equipment for spacecraft to be used in the balmy, 450 Celsius climate on planet Venus.

Part of a logic circuit in a processor for use on Venus. Venus cameras, memories, power converters and amplifiers work comfortably, despite glowing red.

Silicon Carbide is a type of semiconductor that withstands higher temperatures and voltages than ordinary Silicon, at the same time having a lower saturation voltage, meaning that the same equipment can be manufactured with fewer components, runs colder and has better efficiency.

Compared to Silicon, Silicon Carbide yields

  • More compact equipment
  • More efficient equipment
  • Increased life span for critical equipment

Everyone likes equipment with better efficiency, power-saving, cheaper price and more economic operation. Machinery costing billions will live longer thanks to KTH research. By being able to make continuous measurements inside 1000-degree hot gas turbines, breakdowns can be completely avoided.

The Fine Art of Semiconductor Processing

Most excitement in semiconductor manufacture happens inside closed machines, in vacuum, with electron beams, X-rays, ultraviolet rays and vicious acids. Many of the rooms also feature yellow lighting, so as not to harm the light sensitive coatings on the Silicon. It is true that many of the machines have features that may seem a bit like Science Fiction.

You need one atom layer of Silicon? My pleasure! A sprinkle of Gallium on top, perhaps? Just give me two seconds, please.

ALD, Atomic Layer Deposition is all about being able to deposit two atom layers of something on a silicon wafer. No more, no less. Myfab regularly deposits oxide layers of 20-30 Angstroms thickness. Guaranteed.

The metallization equipment may not look very exciting, but on the other hand the performance is outstanding. The machine is able to deposit a 25 nanometre conductive layer of aluminium. Everytime. And that’s a promise. 25 nm is about five atoms stacked on top of each other.

So, did it turn out well? Did you get your 25 nm, or is one atom layer missing?

It’s easy to find out, using the laser-based surface characterisation equipment.

You can have an even closer look, if you like.

Yes, one can see atoms! To be really sure the manufacturing process worked as intended, researchers like to make electron microscope images of their babies. It’s possible to image single atoms, and verify that they have been put in their proper places. Above is an image of a silicon transistor (bottom right). The little dots are single Silicon atoms nicely spaced in a honeycomb pattern. Atoms like to place themselves like that. Above the silicon surface is a messy area, an insulator, consisting of a mixture of Silicon atoms and Silicon Dioxide. This messes up the nicely ordered atoms. The limit region is only some 10 atoms, showing Myfab’s ability to control single atom placement.

Conclusion

Looking beyond this high-tech firework, any professional may note that Myfab in Kista does not possess the very latest machinery. Instead, the fleet consists of lithography equipment that was all the craze in 1998, some relatively un-exciting etching machines, ovens, spin coaters, washers and centrifuges. Their only advantage is stable operation! And they go on and on, just working. The companies paying for this, like it because they won’t have to pay for faulty processing and broken components, and they will always be able to deliver on time.

At the same time, researchers know that they can come up with new projects, that will work. They might be able to add another 20 steps to a process and be quite sure they will not have to waste time investigating processing faults.

Further Reading

Chip manufacturing: https://www.kth.se/ele/research/device-and-circuit-fabrication-1.830135

Embedded systems: https://www.kth.se/ele/research/embedded-systems-1.830155

New technology: https://www.kth.se/ele/research/emerging-technologies-1.830172

Integrated circuits: https://www.kth.se/ele/research/integrated-circuits-and-systems-1.830164

Silicon Carbide (in Swedish): https://techworld.idg.se/2.2524/1.563041/kretsar-som-tal-500-grader