talks on Blue LEDs. As you can see, and some very wonderful title references by War Rhythm. We love them. What are LEDs? Essentially, what they are is from, okay, let's start from cs132, 'cause we've all done that hopefully. They’re probably just something that turns on, and you get either happy or confused because you probably have to decode your whole little y that you made on a silly breadboard. LEDs are made of two separate semiconductor film layers of the same base material, one of which is n-type and the other of which is p-type. N-type has an extra electron doped in through just essentially adding impurities to a silicon base, which makes it just have an extra free electron. P-type has three electron holes or just has a material which has one less electron doped in. Essentially, here's an example of n-type material, which, as you can see, it's a silicon lattice doped with extra phosphorus, and going through, you just see extra electrons kind of sprinkled.

all over the place similarly with the p-type ex except phosphorus would replace with material like with something with three valence electrons that's usually Boron I think yeah it's usually Boron because that's got three valence electrons and it's similar inside to Silicon putting an N-type and a p-type semiconductor together we get a conductor that conducts one way and not the other it's known as a pn or an MP Junction and graphically it looks a bit like that the first bit is saying hey what is it under no potential difference whatsoever so essentially what happens at the start before the material before anything happens to the material before we put a voltage through it is that electrons kind of diffuse from the N-type to the P-type material because they want to fill in the electron holes that are in the p-type material this makes the N-type very slightly positively charged and the P-type very slightly negatively charged um and this region in which this shift occurs is called a depletion

region you can see that in WD when we connect a battery to that NP Junction in the right direction. This depletion region would shrink because you'd actually add electrons into the N type again and um making it more neutral and allows electrons to flow through a diode essentially one way. But if we go the other way because we're adding more electrons to an already slightly negative P type material um it's just going to increase depletion region and this depletion region will actually increase to oppose the voltage and therefore no current will flow. In short though how does this kind of thing create light when electrons flow under it. There is a band called a conduction band and when conduction band electrons fall down into electron holes in the P type material which is called a valence band. That fall and energy creates essentially creates a photon uh because of wave particle duality uh E=hf whatever. Essentially when you connect it in the right way uh conduction electrons fall down into valence electron holes and that jump creates a band wavelength of light and

the next bit was why was blue so hard to make. So, for a bit of history and a bit of reference on this, I believe the red LED and the green LEDs were made in like the late '70s, early '80s, and the blue LED wasn't made until 1992. And essentially, the whole reason was blue is really hard to make from a band gap because its photon energy is so high. If you want 650 nanometer red light, it's 1.9 eV. If you want 510 nanometer green light, it's about 2.4 eV. And I believe 510 was about as easy of a material to make as possible because when you got to 2.7, it's really difficult to make n-type materials with a band gap of about 2.7 or 2.8 electron volts. So, there were two types of materials that could be used when they were doing research for this. It was gallium nitride and zinc.

selenide. These materials were the only ones that could find, could produce a blue or, in Gallum nitrides case, a blue violet light. Okay, Shuji Nakamura was essentially the inventor of this. He was working under Nishia Corporation and was researching how to make these blue LEDs for, I think, like 14-15 years before he actually made it, which is kind of crazy. There was a lot of research done on zinc selenide and its possibilities of making the blue LED that everyone was chasing for, but he believed that gallium nitride was the way forward. Now, there's a problem making Gallum nitride materials. The problem was twofold. It's met organic chemical vapor deposition or whatever those acronyms stand for. And what they do is they take a vaporized material or like a vaporized inorganic material and pass it out onto a crystalline substrate layer. As long as they ensure lamina flow and essentially as long as there's no turbulence in the gases that are being flow, that's

flowing over the substrate um uniform and very flat crystals without any um impurities or dense wood form start to grow on the substrate. And because you want as smooth of an electron flow as possible, this is probably one of the only ways to make a good uniform perfect, well, as perfect as possible crystal lattice. There were two downsides to this rather unfortunate of gallium nitride thought, one of which was essentially Gallum crystals were horribly imperfect to zinc selenide. They had about 10 billion imperfections per square cm when grown on a substrate of sapphire and compared to zinc, which was perfected to about a th000 per square cm. So imperfections obviously led to lower efficiency and like being released that essentially wasn't blue. There was another thing of research on ptype Gan and ptype Gan can be made by doping with magnesium, but it was ineffective as a p type material um. We'll talk about this a little bit later because it's really funny about how he kind of just did it.

um, but on the question of n-type GaN, I believe they doped with silicon to add that extra electron. And basically, silicon doping was very well researched as a method of making n-type materials at that point. As Nakamura overcame these issues, he made a very, very smart change to MOCVD as a method. I can't really go into it, but essentially what he did was he made it flow instead of just one material onto the substrate. He managed to flow two materials as gases on top of two layers and just flowed it through like, um, essentially they grew on two layers, and it's really cool. I can't really go into it because it will take this presentation. It's too long, but do check out two-flow MOCVD. Um, it's really, really fun to take a look at that. Um, the crystals basically grown from this because they had a GaN buffer on the bottom. The crystals grown on top of this was much more perfect, and um, it basically had on the level of, yeah, it

had on the level of zinc sides, level of perfection of about a th000 imperfections per square cm. And when making p-type GaN, hydrogen was theorized to bond to the areas where there should have been an electron hole, which obviously leads to massive inefficiencies because you can't really push electrons through electron-like non-existent electron holes. And basically, what Nakamura did was literally cook the material. He shoved it in an oven for about, I think it was like 24 hours and at 600°. And basically, that just gave the material enough energy to break all of the hydrogen bonds and just release the hydrogen as is. So yeah, he stuck it in an oven. Previous methods were shooting electrons at it. I don't know why, but they found that shooting electrons at it worked. But no, Nakamura just stuck it in the oven. I think this was one of the big things as to why blue LEDs are so popular right now because of the fact that there weren't many hard

processes to perfect after two flow MBD. You literally just had to cook it, and then there was light. It glowed a brilliant blue-violet. Nakamura presented this light at St. Louis in a workshop in St. Louis in 1992. He received a very nice standing ovation, apparently. And now there was the problem of kind of turning this into a commercially viable product. Because right now, it glowed an alpha power of about 42 microwatts, which was well below the micro threshold. But you know, it was something we loved. But, and here's a nice little image showing essentially the ANE layer of p-type GaN on top of n-type, and then you have your buffer which was grown on the substrate. So essentially, what you do here is literally flow every single layer as it went on, and then kneel the top layer. How do we perfect this for how do you go from 42 microwatts to 1,000 microwatts? How do

you make this thing almost 20 times brighter. There were a few issues with this current design, essentially one of which was electrons being blocked at the MP Junction. And essentially what the blocking does is that it forces the electrons to flow from conduction band to conduction band and not from conduction band to valence band, which obviously wouldn't release this blue light. And also it's blue violet, so it's really difficult to see with the naked eye. So that he wanted to make it more blue and less and brighter. What he did with the brighter thing was use a thing called aluminium Gallium nitride or AlGaN or whatever. It's used as a hill material because aluminium Gallium nitride has a larger band gap than gallium nitride itself, so it blocks any electrons from flowing from conduction band to conduction band. Essentially just acts as a bigger resistor to make it more blue. It Indian Gallium Nitro was used as a well material.

to sandwich between the NP layers of aluminum gallium nitride to bring the wavelength down from a very Blue Violet to a nice 450 NM blue. And now that's our little structure. It's a lot more complex than your average LED, where you just have an NP Junction. But essentially, what you do, yeah, you have a large hill material, aluminum gallium nitride, and indium gallium nitride sandwiched in the middle to make a nice... Well, the result was that surprisingly, if you gave this to someone about 40 years ago, they'd be in shock because no projector screen, no LED display, no phone screen for that matter, could display a nice brilliant color of blue. It glowed in on a perfect 1500 microwatts. It was unveiled in 1992, the same year as the first blue LED, actually. And by 1994, NIIA were manufacturing a million units per month. It quadrupled their income in a

few years and made itself a staple because white LEDs were now possible and uh lighting. I think this is Led. Yeah, it's LED. Uh yeah, lighting like stuff in here. It's literally perfect for any sort of application like that. And yeah, that's the end. Um does anyone have questions? No. Okay. Uh very very quick adding them. Veritasium posted this video like two weeks ago. I'm pretty sure some of you might have realized that that's kind of the format I kind of followed. If he didn't make this video, I wouldn't have made it on here because all of the blue light uh blue light stuff kind of just was all cated very well into this one video. I. It was really easy to find the references. Do check out this video. Uh Shameless promotion for this video I guess but it's very tting so we all love him. Anyways, thank you.

⚡The LED that was thought impossible - Tingyi L

Scripts