Key applications for WBG power semiconductors and future outlook with Dr. Peter Friedrichs

Sound On. Power On. your power electronics podcast, powered by PCIM . Hello everybody and welcome to the second episode of Sound On. Power On., your power electronics podcast, powered by PCIM Europe. My name is Marco . I'm a professor for E-mobility and electrical infrastructure at the Bonn-Rhein-Sieg university of applied science at St. Augustine as well as head of department converter and electrical drives. At the Fraunhofer Institute for energy economics and energy system technology at Kassel. Both are located in Germany. Additionally, I'm the chairman, the IEEE joined IES IAS Pels German Chapter and today our technical topic is key applications for wide band gap, power, semiconductors, and future outlook.

And this, I will discuss with Dr. Peter Friedrichs from Infineon. Hi, Peter, how are you?

Hello, Marco. Yeah, I'm doing. well, how are you.

Fine. It's really great honor for me. And a pleasure to have a discussion about this because we know us several years and Peter is the Vice President of Silicon carbide. And as I know, Peter Silicon carbide, semiconductors have shaped your life.

Can you give us a short insight into this, about you and your Company?

Yeah, actually, you're right. This today is a very exciting topic. and was my first step into my professional career. After university times, I had the pleasure in the very early times of silicon carbide research to join the innovative Siemens team, working on this technology already in the early nineties. Yeah. And I also had a big pleasure to be a part of the journey from really basic research on material where nobody was really believing that this is going to work out commercially until the today's time where we run fully full fats based on SIC material. So very exciting time.

There's a lot of ups and downs, but in the end the success stands for it. Great really great. So I think you are the right person for the discussion today in this case. And for me let's say, I know it was in several years, so 10 years ago. And the funny thing for me is we met us the first time at the PCIM, I think it was in 2012 or 2013. And we discussed our first R&D project. And the acronym is, or was IFAS and first Silicon carbide power electronics for ups and ancillary power supply for trains, right. Peter do you know or can you remember about this project? Absolutely. One of the really first application oriented research projects we were in.

I mean nice results, very nice team. I think also part of the team continued in other projects. So you have some history of Marco. Yeah. And but very successful. As I say, it's also all the learnings we made on both sides, I think from our side regarding the components, how to improve them, but also from the application side, how to really use as a new technology, how to manage challenges.

So that was very fruitful, thanks to the ministry of education to found us great. This is a typical sentence we say. And for the audience, maybe as some information, it was a 100 kilo volt amp bidirectional b6 bridge and we checked or achieved a 48 kilohertz. So a as Peter says, yes, we have a great history together.

So we have some project in PV inverters with Silicon carbide, new projects with gallium nitride, let's say, and for hydrogen rectifiers. But I think let's come to our theme. In our first episode of Sound On. Power On. we discussed future trends in power electronics. We identified several future drivers or let's say mega drivers for power electronics.

So we discuss about green electrical energy supplies, or classical, renewable energies, PV, wind, and the new topics about mass production of hydrogen. We discussed about carbonization. Also transport sector means electric vehicle and charging infrastructure efficiency. Increasement in industrial applications like drives for, for production process.

For example, edge new electrical grid structures. DC grid a application and high voltage application in AC grid for example, Peter, from your question right now, do you miss. From the look outside of the power semiconductors. Yeah. Your list was already quite long and exhaustive. What I would like to emphasize a bit also in this context is the whole topic of storage, which is connected to many of those individual aspects you have mentioned whether it's energy transfer or renewable energy generation, what we experienced right now.

To all those let's say online applications the offline backup based on storage systems is adding, which is also a big driver for power semiconductors. And especially by wide bandgap.

And it's a driver Germany or in some other country in the whole world. Yeah, that's also an interesting view or we have developed here not necessarily to Germany or Europe simply due to the grid structure we have here.

So there are other regions in the world, where this Element of storage has a much higher importance in the energy flow compared to Europe. But of course, since Europe is always innovative, also here, a lot of projects, this is running but we thing that the real of implementation might happen in areas with more isolated energy grids and supply situations more,

so this means weak gridsor island grids, and so on.

Exactly. Exactly. Yeah. So we have people simply are relying for instance, and PV, and they have to manage the supply for 24 hours a day. Then you have no other chance to go for a huge amount of storage in order to manage it. Great. I think there exist. A lot of big projects over the whole world to realize such solutions means reducing diesel and increasing more PV.

Great. So, and the list yes we have now one additional topic in our list and which application represents the greatest challenge for power semiconductors and why? Our experience is that all the emerging applications represent usually is the biggest challenge simply because we are facing completely new operating modes, sometimes new environmental condition.

So if you just take the example of PV so they, they operated in environments where basically before no power electronics was present. So. Let's say some, some farms or some, some ranches installations are made together with cows and pigs. Oh, that was from a, from a, let's say air, air pollution, point of view, a certain challenge will power electronics, and also then the outside installations with moisture exposure, completely different to other things.

So this are always, let's say something. Challenges to manage in those upcoming fields. As more, the more we address established field to easier to is also regarding stress modes, operating modes, you're all familiar with all the things which might happen, and we can address it in the design phase, but for new and emerging applications as is always, let's say a longer period of learning together with.

. So that means in harsh environments that are really yeah. hard or the hard conditions in this case. Correct. Plus the operating modes when sometimes also that's unexpected events failure or overstress from an electrical point of view sometimes there's also figured out only if the system is really up and running in real life.

So you can simulate and predict a lot, but not every. Great. Oh yeah. It was all know, understand a little bit more. And I think the audience, and let's go to the next question. I see a trend of increasing blocking voltage of Silicon carbide semiconductors in which business areas can these expected. We are having this discussion already.

As long as it's Silicon carbide exists since a potential to come to single die. Very high voltages is definitely. Demonstrations are done by many players in, in the field. But I have to be a little bit of bad guy here. Our observation in the industrial environment is rather, let's say a little bit hesitating regarding the move to single chip, higher voltages.

And so from an implementation point of view, we still don't see it at the moment. There's a lot of nice demonstrations. That's a, that's pretty clear, but at the moment, the focus from the Application site. And also from the development in the industrial world, we see our current view. Traditional segments may be up to six KV.

There might be something around 10 KV in the far future about not more. But the thinks all that is a, those are typical things. The last class, what you have in your mind at 3.3 kilovolts using. There's this blocking voltage. And I think so only to achieve a higher voltage with multilevel.

Inverters is a good point, but I see it to use it in public grid with 10 20 kilo volts and higher I think it's a little bit hard. But true only in high voltage DC application. Yes, it's clear. Okay, great. So we have a discussion right now about Silicon carbite at, but I know there exists some other technologies and I think in the future technologies like gallium nitride, but for me, the question about this, since more than 10 years, gallium, nitride, semiconductors were introduced in power electronics.

So. There's several other research products. And so on parts of technology has not achieved a breakthrough yet. Can you tell us a bit more why it is? Oh yeah. Well, if I look back to the certain competence that we experienced, the same situation in the nineties, so same, same story, a lot of hype around Silicon carbide storylines that we were completely eliminate Silicon.

And what happens always is that we need to sort out the real opportunities from things which sounded interesting, but cannot materialize. This is exactly what gallium nitride from my point of view was already going through in the last years, because right now, at least our internal Infineon gallium, nitride businesses.

is realy jumping up. So we think this settlement has taken place, so we know where, which technology find its application. Also sometimes the systems need to adopt a little bit to the new technologies. It's also a learning we have made if you take. Example of gallium nitride. So it is used in special topologies, which have not been possible before the Silicon.

That also means a learning site on the customer. And but after those, those painful periods the real fun begins, and this is also valid gallium nitride right now. So. I think that the next few years, will also proves that the potential of gallium nitride is there and that at least certain applications, will really gain from this new technology, it's a good point.

And yes, as advantage Silicon carbide and gallium nitride, more or less, it's clear. And you said, okay, each technology will have its place in the future. Can you give a little bit more look inside? Where do you think this business areas? Or that means, for example, it's an in about maybe in PVS is a period or in, in IGBTs is for large drives.

Can you fix it more? Yeah. Yeah, absolutely. So that is one of the key questions we are discussing practically daily is many, many customers because Infineon is one of the few, if not the only company really having all three power technologies in the portfolio. So Silicon, Silicon carbide and gallium nitride

and of course the question is, well, we have to use what well I think Silicon it's, it's pretty clear as, as are the segments where . Cost is very interesting and where we do not have very strong requirements regarding efficiency, switching frequencies. Silicon carbidewe think is definitely the alternative or.

brings added value to Silicon. If we think in high power, high voltage, medium frequencies. So we don't fit it to certain applications. That's rather the requirements coming from the applications. Yeah. Gallium nitride benefits, or the possibility to switch very fast due to very low gate charge. That means extremely high frequencies integration.

So if you want to create a very compact system, some, let's say typologies already integrated on chips as, other things like galium nitride, very beneficial. And from that perspective, we see also the use cases, especially in power supplies, there will be low, low power motor drives, or very compact stuff around and enabling also the whole story of very compact charger.

So, so if you go to Amazon, for instance, you'll find a lot of USB charging. Equipment it's already labeled and powered by gallium nitride. So that's a typical playground. We see at least for the foreseeable future for gallium nitride, while, as I said, silicum carbite has its strengths and high power DCDC conversion main these and PV of course storage, but also drive s

so we see also a big pool at the moment from drives. Sowe separated from mostly from a point of frequency and power rating and. Yeah. So point is of course, since the quiet blocking voltage, and we spoke now about this technology, let's say of the advantages of the technology perspective, but I think never say less is the end of the customer.

The user let's say is, are the costs. And do you think gallium nitride has a possibility in the future? Get cheaper than. silicon carbiteas well as maybe silicone, IGBTQ. Is that as a possibility, that's a very difficult one. Because you know, all three technologies, we see progress. So even Silicon is not, not yet at the end of the roadmap.

So I have very smart colleagues working on future generations of IGBTs and I have a lot of respect. Because this is the products possible. And so Silicon carbide cost down is driven here by material improvements by larger diameters also by disruptive technologies like split technologies. So introduce this called split technology, which can release dramatacaly decrease the contribution of the substrate and gallium nitride also is going up with diameters

and here the key years of course, to improve also then the subsequent epiprocess on the already pretty cheap substride so the battle is open. I would say in the end cost performance will decide and. This might be an application specific. Oh, not necessarily. Maybe the comparison is fair on the pure component price.

Okay. So let's go back to the advantage or let's say, may I look in the technology side? I see, okay. Now the exists Silicon carbide, PV inverters, for example, that's a market up to 150 kilovolt Amps. And in research, let's say we increase it up to, let's say 500 kilovolt amps. So, and for me is, is the next step I think.

So it would be as, as the megawatt class. And that means large drives wind power generation, but let's say I see it as a portfolio from the whole manufacturer about this such tipicaly and using up in this high power using prime pack packaging. But I think using the advantages of Silicon carbide to let's say you need less or really minimized parasitic inductance.

Packaging form. Isn't the future? Not so, really useful for that application, but I think efficiency increasing a little bit higher switching frequencies or medium switching figuring is a really good advantage. Let's say in large drives, maybe more in wind power system because of the bouts, a grid integration perspective in this case.

Maybe there is some research or development more necessary. Let's say it's the packaging. Yeah. You mentioned prime pack Marko, well you're absolutely right. We discoreded this package for a silicon carbide already many, many years ago. So this is definitely not the right housing, even if you used it for some research projects.

Yeah our focus. here is on our new XHP power platform, which is a half bridge configuration. Exactly. Coming along with the things you mentioned, the low stray inductance scalability, et cetera. And this concept silicon carbide in combination with the XHP also is very interesting for the large drives and wind power installations, even if we expect, or it is a first deployment of those high power solutions actually in trains or that's a think, you might remember this first project, your IFAS also was dealing with a first step of silicon carbide in trains,

it was about power supplies. Meanwhile, we are stepping into the propulsion system with. Technologies and but it's also very similar to an large drive in the end. Therefore those as a direct link and director experience can be gained from source additional. The investigations, but absolutely clear. We need this new package which is well-developed already also for IGBTs, need some adoptions for SIC, but can be done and, and the field is open for high power applications based SIC.

Wow. That's really great. And I'm looking forward for this. And you mentioned now I think yes, for higher power application. So half-bridge topologies are really interesting, but I see it as a product lines and in some outcomes of research projects as, as the AmpC bridge in hybrid technologies, or it means IGBTs.

And Silicon carbide, MOSFET combined. Is this a future? For the, yeah, let's say different power and application. I think a general , the flexibility is very important. Maybe not for the very high power applications, but for every seeing in this range, let's say starting from 20 kilowatts up to 200 kilowatts.

So based on innovative typologies, you can gain a lot in your system. And since you can also then from a cost performance point of view, optimize the solution. If you not just thinking one chip technology are we experiences. All sorts of combination between Silicon power devices and silicon carbide power devices in certain typologies is very successful.

It attracts customers because they understand that. You need, you takes a technology. You need to know what the technology you have on hand right now. And if you then has also the right package technology around, which is provided by our press fit base plate free module technologies. Yeah, this is perfect.

To make Silicon carbide happen in large scale and also in combination with Silicon. But I want to come back to the cost. I, I want to stress it a little bit more. Let's say what does the next step for reducing the cost of Silicon carbide? Because I think so. Yeah, let's say is a target corridor must nearly come to the silicon IGBTs.

So what doesn't that's technology steps we need. Ah, for degreasing the costs. Yeah. Well, we have three big cost drivers at the moment. In the case of Silicon carbide, one is the base material cost itself. The second topic is defect density, which is still impacting the yield. And then of course the third big topic is the whole.

Processing effort. That's basically the technology itself to manufacture a device. At least on our side, we are addressing all three parts regarding the substrates we are trying to let's say Foster reuse of substrates by, by split technologies, where we take off a, an unused part of processing and reprocess basically as a wholesilicon carbite waiver defect density here.

We have programs in place also with our suppliers to work on dedicated improvement strategies where we are trying to address the important defects it's at once, which really kills a yield and different technology buyers. Of course, there's this, I think it's a typical approach to work on new generations, which then require less space on, on the material, but also helps us to, to come down this costs.

Several measures are in place and we are pretty confident that the cost down curve, we were able to show in the past. We can continue alsoin the future that sounds really, really great. So, Peter, in the last minutes we discussed a lot of the future of power electronics. And what is about your future? What will we see?

Well I hope you will still see myself also in the community around Silicon carbide. I hope also that we are going to meet in the future also again, in person in many research projects. Yeah. The biggest challenge we have in front of us also here in my direct environment, we are a growing team.

Oh, so many, many years we had been maybe a smarter group of 20, 30 people. We are hundrets of them right now as a company that also means we need to disseminate the knowledge we need to educate the people. That's one of my biggest tasks. So maybe something like an internal lecturer also. Yeah. In the company But I still look forward to a minimum 10 exciting years in the field of Silicon carbide.

There's a lot of new applications, new customers coming in. Yeah. And very nice academic exchange. Of course. Yeah, I think it's really important. Let's say the education about this to, to share the knowledge for how to use such Silicon carbide or gallium nitride semiconductor, because I think it's a, about the drive for this or the driver for this.

I think it's a, you need a lot of knowledge how to have. So EMC and so on in this case. So it would be great to work in the future together in this, like in R and D projects or in education. Maybe it's a good point in this case. So thank you, Peter. I enjoyed our conversation. Pleasure was on my side. To all the listeners, wherever you might be, thank you very much for listening.

We hope you have enjoyed today's episode and gained.some valuable insights, make sure to subscribe to our podcast, which is available on all major podcast platforms. If you would like to share your feedback with us, please do so via email to podcast-pcim@mesago.com . You can write to us also, if there is a specific topic you would like to be covered in future.

If you have a particular guest on your mind until then have a great time sound on power on your power electronics podcast, powered by PCIM.