YT YouTube2Text

YouTube2Text → Transcripts → Why It Was Almost Impossible to Make the Blue LED

Why It Was Almost Impossible to Make the Blue LED — Transcript

by Veritasium · 5,224 words · 759 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00- LEDs don't get their color from their plastic covers.
0:03And you can see that because here is a transparent LED
0:06that also glows the same red color.
0:09The color of the light comes
0:10from the electronics themselves.
0:12The casing just helps us tell different LEDs apart.
0:15In 1962, general Electric engineer Nick Holonyak
0:19created the first visible LED.
0:21It glowed a faint red.
0:23A few years after that, engineers at Monsanto
0:26created a green LED.
0:28But for decades, all we had were those two colors.
0:32So LEDs could only be used in things like indicators,
0:36calculators, and watches.
0:38If only we could make blue, then we could mix
0:41red, green, and blue to make white,
0:43and every other color,
0:45unlocking LEDs for every type of lighting in the world,
0:48from light bulbs, to phones, to computers,
0:51to TVs to billboards.
0:54But blue was almost impossible to make.
0:57(dramatic music)
1:00Throughout the 1960s,
1:01every big electronics company in the world,
1:04from IBM to GE, to Bell Labs,
1:07raced to create the blue LED.
1:10They knew it would be worth billions.
1:12Despite the efforts of thousands of researchers,
1:15nothing worked.
1:1810 years after Holonyak's original LED
1:20turned into 20, then 30,
1:22and the hope of ever using LEDs for light, faded away.
1:26According to a director at Monsanto,
1:29these won't ever replace the kitchen light.
1:32They'd only be used in appliances, car dashboards,
1:34and stereo sets to see if the stereo was on.
1:38This might still be true today, if not for one engineer
1:42who defied the entire industry
1:44and made three radical breakthroughs
1:46to create the world's first blue LED.
1:50(dramatic music)
1:52Shūji Nakamura was a researcher at a small Japanese chemical
1:55company named Nichia.
1:57They had recently expanded into the production
1:59of semiconductors to be used in the manufacture
2:02of red and green LEDs.
2:04But by the late 1980s,
2:05the semiconductor division was on its last legs.
2:08They were competing against far more established
2:10companies in a crowded market, and they were losing.
2:14Tensions started to run high.
2:17Younger employees begged Nakamura to create new products,
2:20while senior workers called his research a waste of money.
2:25And at Nichia, money was in short supply.
2:29Nakamura's lab mainly consisted of machinery
2:31he had scavenged and welded together himself.
2:34Phosphorus leaks in his lab created so many explosions,
2:37that his coworkers had stopped checking in on him.
2:40By 1988, Nakamura's supervisors were so disillusioned
2:44with his research that they told him to quit.
2:47So it was out of desperation
2:49that he brought a radical proposal to the company's founder
2:52and president Nobuo Ogawa.
2:55(dramatic music)
2:56The elusive blue LED,
2:58that the likes of of Sony, Toshiba and Panasonic
3:00had all failed at.
3:02What if Nichia could be the one to create it?
3:06After suffering loss after loss on their semiconductors
3:09for more than a decade,
3:10Ogawa took a gamble.
3:12He devoted 500 million yen or $3 million,
3:16likely around 15% of the company's annual profit,
3:19to Nakamura's moonshot Project.
3:24Everyone knew that LEDs have the potential
3:26to replace light bulbs,
3:28because light bulbs, the universal symbol for a bright idea,
3:33are actually terrible at making light.
3:35They work by running current through a tungsten filament,
3:38which gets so hot, it glows.
3:40But most of the electromagnetic radiation
3:43comes out as infrared, heat.
3:45Only a negligible fraction is visible light.
3:49In contrast, LED stands for light emitting diode.
3:53It's right there in the name.
3:55LEDs primarily create light, so they're far more efficient,
3:59and a diode is just a device with two electrodes,
4:02which only allows current to flow in one direction.
4:06So here's how an LED works.
4:09When you have an isolated atom,
4:11each electron in that atom occupies a discreet energy level.
4:15You can think of these energy levels like individual seats
4:17from a hockey stadium,
4:19and all atoms of the same element,
4:21when they are far apart from each other
4:22have identical available energy levels.
4:25But when you bring multiple atoms together to form a solid,
4:28something interesting happens.
4:30The outermost electrons now feel the pole,
4:33not only of their own nucleus,
4:35but of all the other nuclei as well.
4:37And as a result, their energy levels shift.
4:40So instead of being identical,
4:42they become a series of closely spaced,
4:45but separate energy levels.
4:47An energy band.
4:48The highest energy band with electrons in it,
4:51is known as the valence band,
4:53and the next higher energy band
4:55is called the conduction band.
4:56You can think of it like the balcony level.
5:00In conductors, the valence band is only partially filled.
5:03This means with a little bit of thermal energy,
5:05electrons can jump into nearby unfilled seats,
5:09and if an electric field is applied,
5:10they can jump from one unfilled seat to the next
5:13and conduct current through the material.
5:16In insulators, the valence band is full,
5:19and the difference in energy between the valence
5:21and conduction bands, the band gap, is large.
5:25So when an electric field is applied, no electrons can move.
5:29There are no available seats
5:31to move into in the valence band,
5:33and the band gap is too big for any electrons
5:35to jump into the conduction band,
5:38which brings us to semiconductors.
5:41Semiconductors are similar to insulators,
5:44except the band gap is much smaller.
5:46This means at room temperature,
5:48a few electrons will have sufficient energy
5:50to jump into the conduction band,
5:53and now they can easily access nearby empty
5:55seats and conduct current.
5:57Not only that, the empty seats they left
5:59behind in the valence band can also move.
6:02Well, really, it's the nearby electrons
6:04jumping into those empty seats.
6:06But if you look from afar,
6:07it's as though the empty seat or hole
6:10is moving like a positive charge in the opposite
6:13direction to the electrons in the conduction band.
6:17(soft music)
6:18By themselves, pure semiconductors are not that useful.
6:22To make them way more functional,
6:24you have to add impurity atoms into the lattice.
6:27This is known as doping.
6:29For example, in silicon,
6:30you can add a small number of phosphorus atoms.
6:33Phosphorus is similar to silicon,
6:34so it easily fits into the lattice,
6:36but it brings with it one extra valence electron.
6:40This electron exists in a donor level
6:42just beneath the conduction band.
6:44So with a bit of thermal energy,
6:46all these electrons can jump into the conduction band
6:48and conduct current.
6:50Since most of the charges that can move
6:53in this type of semiconductor are electrons,
6:54which are negative,
6:55this sort of semiconductor is called n-type,
6:58n for negative,
7:00but I should point out that the semiconductor
7:01itself is still neutral.
7:03It's just that most of the mobile charge
7:05carriers are negative.
7:06They're electrons.
7:08So there is also another type of semiconductor where most
7:10of the mobile charge carriers are positive,
7:12and it's called p-type.
7:16To make p-type silicon,
7:17you add a small number of atoms of, say, boron.
7:20Boron fits into the lattice,
7:21but brings with it one fewer valence electron than silicon.
7:25So it creates an empty acceptor level
7:27just above the valence band.
7:29And with a bit of thermal energy,
7:30electrons can jump outta the valence band,
7:32leaving behind holes.
7:34It is these positive holes which are mostly responsible
7:37for carrying current in the p-type semiconductor.
7:41Again, the material overall is uncharged,
7:43it's just that most of the mobile charge carriers
7:46are positive holes.
7:48Where things get interesting is when you put a piece
7:50of p-type and n-type together.
7:52Without even connecting this to a circuit,
7:55some electrons will diffuse from n to p
7:58and fall into the holes in the p-type.
8:01This makes the p-type a little negatively charged,
8:03and the n-type a little positively charged.
8:06So there is now an electric field
8:08inside an inert piece of material.
8:12Electrons keep diffusing
8:13until the electric field becomes so large,
8:16it prevents them from crossing over.
8:18And now we have established the depletion region,
8:21an area depleted of mobile charge carriers.
8:24There are no electrons in the conduction band
8:26and no holes in the valence band.
8:29If you connect a battery the wrong way to this diode,
8:31it simply expands the depletion region
8:34until its electric field perfectly opposes that
8:36of the battery and no current flows.
8:41But if you flip the polarity of the battery,
8:43then the depletion region shrinks,
8:45the electric field decreases,
8:47and electrons can flow from n to p.
8:50When an electron falls from the conduction band into a hole
8:54in the valence band, that band gap energy can be
8:57emitted as a photon.
8:59The energy change of the electron is emitted as light,
9:03and this is how a light emitting diode works.
9:06The size of the band gap determines the color
9:09of the light emitted.
9:10In pure silicon, the band gap is only 1.1 electron volts.
9:14So the photon released isn't visible, it's infrared light.
9:18These LEDs are actually used in remote controls
9:22for your TV, and you can capture them on camera.
9:25Moving up the spectrum, you can see why the first visible
9:28light LEDs were red and then green,
9:31and why blue was so hard.
9:33A photon of blue light requires more energy,
9:36and therefore a larger band gap.
9:39By the 1980s,
9:40after hundreds of millions of dollars had been spent hunting
9:44for the right material, every electronics company
9:46had come up empty handed.
9:48But researchers had at least figured out
9:50the first critical requirement, high quality crystal.
9:54No matter what material you used for the blue LED,
9:57it required a near perfect crystal structure.
10:00Any defects in the crystal lattice,
10:02disrupt the flow of electrons.
10:04So instead of emitting their energy as visible light,
10:06it is instead dissipated as heat.
10:09So the first step in Nakamura's proposal to Ogawa,
10:12was to disappear to Florida.
10:15He knew an old colleague there whose lab was beginning
10:17to use a new crystal making technology called
10:19Metal Organic Chemical Vapor Deposition,
10:22or MOCVD.
10:25An MOCVD reactor, essentially a giant oven,
10:29was and still is the best way to mass produce clean crystal.
10:33It works by injecting vapor molecules
10:36of your crystal into a hot chamber where they react
10:38with a base material called a substrate to form layers.
10:42It's important that the substrate lattice matches
10:44the crystal lattice being built on top of it
10:46to create a stable, smooth crystal.
10:49This is a precise art.
10:51The crystal layers often need to be as thin
10:54as just a couple of atoms.
10:56Nakamura joined the lab for a year to master MOCVD.
11:01But his time there was miserable.
11:04He wasn't allowed to use the working MOCVD,
11:07so he spent 10 of his 12 months assembling a new system,
11:10almost from scratch.
11:12Even worse, his lab mates shunned him
11:15because Nakamura didn't have a doctorate,
11:17nor any academic papers to his name,
11:19as Nichia didn't allow publishing.
11:21His lab mates, all PhD researchers,
11:24dismissed him as a lowly technician.
11:27This experience fueled him.
11:29Nakamura wrote, "I feel resentful
11:31when people looked down on me.
11:33I developed more fighting spirit.
11:35I would not allow myself to be beaten by such people."
11:42(inspirational music)
11:43He returned to Japan in 1989 with two things in hand.
11:46One, an order for a brand new MOCVD reactor for Nichia,
11:50and two, a fervent desire to get his PhD.
11:54At that time in Japan, you could earn a PhD
11:56without having to go to university,
11:58simply by publishing five papers.
12:02Nakamura had always known his chances
12:04of inventing the blue LED were low.
12:06But now he had a backup plan.
12:09Even if he didn't succeed, he could at least get his PhD.
12:13But now the question was with MOCVD under his belt,
12:17which material should he research?
12:21By this time, scientists had narrowed the options down
12:23to two main candidates, zinc selenide, and gallium nitride.
12:27These were both semiconductors with band gaps,
12:30theoretically, in the blue light range.
12:33Zinc selenide was the far more promising option.
12:36When grown in an MOCVD reactor,
12:38it had only a .3% lattice mismatch
12:41with its substrate, gallium arsenide.
12:44Therefore, zinc selenide crystal had about a thousand
12:46defects per square centimeter,
12:48within the upper limit for LED functioning.
12:51Its only issue was that while scientists
12:53had figured out multiple different
12:54ways to create n-type zinc selenide,
12:57no one knew how to create p-type.
13:00In contrast, gallium nitride had been abandoned
13:03by almost everybody for three reasons.
13:06First, it was much harder to make a high quality crystal.
13:10The best substrate for growing gallium nitride was sapphire,
13:13but its lattice mismatch was 16%.
13:16This resulted in higher defects,
13:19over 10 billion per square centimeter.
13:22The second problem was that like zinc selenide,
13:25scientists had only ever created
13:26n-type gallium nitride using silicon.
13:29P-type was elusive.
13:32And third, to be commercially viable,
13:34a blue LED would have to have a total light output power
13:37of at least a thousand microwatts.
13:40That's two orders
13:41of magnitude more than any prototype had ever achieved.
13:45So between the two candidates,
13:47almost all researchers were focused on zinc selenide.
13:51Nakamura surveyed the crowded field
13:53and decided that if he were going
13:55to publish five papers by himself,
13:57he'd better focus on gallium nitride,
13:59where the competition was much less fierce.
14:02This material's main claim
14:04to fame was one development back in 1972,
14:08when RCA engineer Herbert Maruska made a tiny
14:11gallium nitride blue LED, but it was dim and inefficient.
14:15So RCA slashed the project's budget, calling it a dead end.
14:2020 years later, scientific opinion hadn't changed.
14:23When Nakamura attended the biggest applied physics
14:25conference in Japan, the talks on zinc selenide
14:28had over 500 attendees.
14:30The talks on gallium nitride had five.
14:34(dramatic music)
14:36Two of those five attendees were the world experts
14:38on gallium nitride, Dr. Isamu Akasaki
14:41and his former grad student, Dr. Hiroshi Amano.
14:45In contrast to Nakamura's academic background,
14:47they were researchers at Nagoya University,
14:50one of Japan's best.
14:52A few years earlier, they had made a breakthrough
14:54on the first problem of high quality crystal.
14:57Instead of growing gallium nitride directly on sapphire,
15:01they first grew a buffer layer of aluminum nitride.
15:04This has a lattice spacing in between that
15:07of the other two materials, making it easier
15:09to grow a clean gallium nitride crystal on top.
15:12The only issue was that the aluminum caused problems
15:15for the MOCVD reactor,
15:17making the process hard to scale.
15:20But Nakamura wasn't even close at this stage.
15:24Back at Nichia, he couldn't get gallium nitride to even grow
15:27normally in his new MOCVD reactor.
15:31After six months, desperate for results,
15:33he decided to take the machine apart
15:35and build a better version himself.
15:39His 10 months spent putting together the reactor in Florida,
15:42were suddenly invaluable.
15:44He began following the same routine each day,
15:47arrive at the lab at 7:00 AM.
15:49Spend the first half
15:50of the day welding, cutting, and rewiring the reactor.
15:53Spend the rest of the day experimenting
15:55with the modified reactor to see what it can do.
15:58At 7:00 PM go home, eat dinner, wash and sleep.
16:04Nakamura repeated this routine every single day,
16:07taking no weekends
16:08and no holidays except for New Year's Day,
16:11the most important holiday in Japan.
16:14(soft music)
16:16After a year and a half of continuous work,
16:19he came into the lab on a winter day in late 1990.
16:23As usual, he tinkered around in the morning
16:25grew a gallium nitride sample in the afternoon,
16:28and tested it.
16:32But this time, the electron mobility was four times higher
16:36than any gallium nitride ever grown directly on sapphire.
16:40Nakamura called it the most exciting day of his life.
16:45His trick was to add a second nozzle
16:48to the MOCVD reactor.
16:50The gallium nitride reactant gases had been rising
16:53in the hot chamber,
16:54mixing in the air to form a powdery waste.
16:57But the second nozzle released a downward stream
17:00of inert gas, pinning the first flow to the substrate
17:03to form a uniform crystal.
17:06For years, scientists had avoided adding a second stream
17:09to MOCVD because they thought it would only
17:11introduce more turbulence.
17:13But Nakamura used a special nozzle
17:15so that even when the streams combined,
17:17they remained laminar.
17:19He called his invention the two-flow reactor.
17:23Now, he was ready to take on Akazaki and Amano,
17:26but instead of copying their aluminum nitride buffer layer,
17:29his two flow design allowed him to make gallium nitride
17:32so smooth and stable, it itself could be used
17:36as a buffer layer on the sapphire substrate.
17:38This in turn, yielded an even cleaner crystal
17:41of gallium nitride on top,
17:43without the issues of aluminum.
17:46Nakamura now had the highest quality
17:48gallium nitride crystals ever made.
17:51But just as he was getting started,
17:53things took a wrong turn.
17:55(dramatic music)
17:57While he had been in Florida,
17:58Nobuo Ogawa had stepped back from Nichia to become chairman.
18:02In his day, Nobuo had been a risk taking scientist,
18:05designing the company's first products.
18:08It's why he supported Nakamura's lofty plans all this time.
18:11But in his place, his son-in-law, Eji Ogawa,
18:14became CEO of the company,
18:17and the younger Ogawa had a much stricter outlook.
18:20One Nichia client said,
18:21"He has a mind of steel,
18:23and he remembers everything."
18:27In 1990, an executive at Matsushita,
18:29an LED manufacturer and Nichia's biggest customer,
18:33visited the company to give a talk on blue LEDs.
18:37In it, he claimed zinc selenide was the way forward,
18:40declaring "gallium nitride has no future."
18:44That very same day, Nakamura received a note from Eji,
18:47stop work on gallium nitride immediately.
18:51Eji had never supported the research
18:53and wanted to end what he saw as a colossal waste.
18:57But Nakamura crumpled up the note and threw it away,
19:01and he did so again, and again,
19:04when a succession of similar notes
19:06and phone calls came from company management.
19:09Out of spite, he published his work on the two-flow reactor
19:12without Nichia's knowledge.
19:14It was his first paper.
19:16One down, four to go.
19:20With crystal formation settled,
19:21he turned to the second obstacle,
19:24creating p-type gallium nitride.
19:26Here Akazaki and Amano had again beaten him to the punch.
19:31They had created a gallium nitride sample doped
19:33with magnesium, but at first,
19:35it didn't perform as a p-type as they expected.
19:38However, after exposing it to an electron beam,
19:41it did behave as a p-type,
19:44the world's first p-type gallium nitride,
19:46after 20 years of trying.
19:49The catch was that no one knew why it worked.
19:52And the process of irradiating each crystal
19:55with electrons was too slow for commercial production.
20:00At first, Nakamura copied Akazaki and Amano's approach,
20:03but he suspected the beam of electrons was overkill.
20:06Maybe all the crystal needed was energy.
20:09So he tried heating magnesium doped gallium nitride
20:12to 400 degrees Celsius in a process known as annealing.
20:16The result, a completely p-type sample.
20:20This worked even better than the shallow electron beam,
20:23which only made the surfaces of the samples p-type,
20:26and simply heating things up was a quick scalable process.
20:30His work also revealed why the p-type had been so difficult.
20:33To make gallium nitride
20:35with MOCVD, you supply the nitrogen from ammonia,
20:39but ammonia also contains hydrogen.
20:41Where there should have been holes in the magnesium
20:44doped gallium nitride,
20:45these hydrogen atoms were sneaking in
20:47and bonding with the magnesium, plugging all the holes.
20:51Adding energy to the system,
20:53released the hydrogen from the material,
20:55freeing up the holes again.
20:59(dramatic music)
21:00By now, Nakamura had all the ingredients
21:02to make a prototype blue LED,
21:05and he presented it at a workshop in St. Louis in 1992
21:09and received a standing ovation.
21:11He was beginning to make a name for himself,
21:14but even though he had created the best prototype to date,
21:18it was more of a blue violet color
21:20and still extremely inefficient,
21:22with a light output power
21:23of just 42 microwatts,
21:25well below the 1000 microwatt threshold for practical use.
21:30At Nichia, the new CEO's patience had run out.
21:34Eji sent written orders to Nakamura to stop tinkering
21:37and turn whatever he had into a product.
21:40His job was on the line,
21:42but in Nakamura's own words, "I kept ignoring his order.
21:46I had been successful because I didn't listen
21:49to company orders and trusted my own judgment."
21:52At this point, he only had the third hurdle left,
21:55getting his blue LED to a light output power
21:57of a thousand microwatts.
22:01(soft music)
22:02A known trick to increase the efficiency of LEDs
22:04was to create a well,
22:06a thin layer of material at the p-n junction
22:09called an active layer
22:11that shrinks the band gap just a bit.
22:14This encourages more electrons
22:16to fall from the end type conduction band into holes
22:18in the p-type valence band.
22:21The best active layer for gallium nitride was already known
22:24to be indium gallium nitride,
22:26which would not only make the band gap easier to cross,
22:29but also narrow it just the right amount
22:31to bring its blue violet gap down to true blue.
22:35This time, Akasaki and Amano didn't scoop Nakamura.
22:39They were stuck trying to grow
22:40indium gallium nitride in the first place.
22:43Amano recalled, "It was generally said that gallium nitride
22:46and indium nitride would not mix, like water and oil."
22:50But Nakamura had an advantage,
22:52his ability to customize his MOCVD reactor.
22:56This allowed him to use brute force,
22:58adjusting the reactor to pump as much indium
23:01as he could onto the gallium nitride,
23:03in the hopes that at least some would stick.
23:06To his surprise, the technique worked,
23:09giving him a clean indium gallium nitride crystal.
23:12He quickly incorporated this active layer into his LED,
23:16but the well worked a little too well
23:19and overflowed with electrons,
23:21leaking them back into the gallium nitride layers.
23:24Unfazed, within a few months, Nakamura had fixed this too
23:28by creating the opposite of a well, a hill.
23:31He returned to his reactor one more time
23:33to make aluminum gallium nitride,
23:36a compound with a larger band gap that could block
23:38electrons from escaping the well once inside.
23:46(dramatic music)
23:48The structure of the blue LED had become far more complex
23:52than anyone could have imagined, but it was complete.
23:56By 1992, Shūji Nakamura had this.
24:03- And I showed the chairman, I told him,
24:05"Please, hey chairman come to my office."
24:07I showed him the blue LED
24:08and he said, "ohh, this is great no?"
24:11I became so happy.
24:12I just became, out of my office, yeah.
24:16- [Derek] After 30 years of searching
24:18by countless scientists,
24:20Nakamura had done it.
24:21He had created a glorious, bright blue LED
24:24that could even be seen in daylight.
24:27It had a light output power of 1,500 microwatts
24:31and emitted a perfect blue at exactly 450 nanometers.
24:35It was over 100 times brighter
24:38than the previous pseudo-blue LEDs on the market.
24:41Nakamura wrote, "I felt like I had reached
24:44the top of Mount Fuji."
24:46Nichia called a press conference in Tokyo
24:48to announce the world's first true blue LED.
24:51The electronics industry was stunned.
24:54A researcher from Toshiba remarked,
24:56"Everyone was caught with their pants down."
24:59The effect on Nichia's fortunes was immediate and explosive.
25:03Orders flooded in,
25:05and by the end of 1994,
25:06they were manufacturing 1 million blue LEDs per month.
25:11Within three years,
25:12the company's revenue had nearly doubled.
25:15In 1996, they made the jump from blue to white,
25:19by placing a yellow phosphor over the LED.
25:22This chemical absorbs the blue photons
25:25and re-radiates them in a broad spectrum
25:27across the visible range.
25:29Soon enough, Nichia was selling the world's
25:31first white LED.
25:33At last, unlocking the final frontiers so many had doubted,
25:38LED lighting.
25:40Over the next four years, their sales doubled again.
25:44By 2001, their revenue was approaching $700 million a year.
25:49Over 60% came from blue LED products.
25:53Today, Nichia is one of the largest LED manufacturers
25:56in the world with an annual revenue in the billions.
26:01As for Nakamura, to whom Nichia owed
26:04the quadrupling of its fortunes?
26:08(dramatic music)
26:09- I increased my salary, $60,000.
26:12After doubling, yeah.
26:14- I heard you only got $170 bonus
26:17- Each patent.
26:18- So you got $170 bonus for the patent.
26:21- Yes, yes.
26:22- [Derek] This was all while the blue LED
26:24was generating hundreds of millions of dollars in sales.
26:28Eji Ogawa had always seen Nakamura's stubborn individuality
26:32as a liability, not a strength.
26:34The message was clear.
26:36In 2000, after more than 20 years at Nichia,
26:39Nakamura left the company for the US,
26:42where job offers had been pouring in.
26:44But his troubles with Nichia weren't over.
26:47He began consulting for Cree, another LED company.
26:51Nichia was furious and sued him for leaking company secrets.
26:55Nakamura responded by counter-suing Nichia
26:58for never properly compensating him for his invention,
27:01seeking $20 million.
27:05In 2001, the Japanese courts ruled with Nakamura
27:08and ordered Nichia to pay him 10 times his initial request.
27:12But Nichia appealed
27:14and the case was eventually settled
27:15with a payout of $8 million.
27:19In the end, this was only enough
27:21to cover Nakamura's legal fees.
27:24(soft music)
27:26This is all he got for an invention
27:29that now comprises an $80 billion industry,
27:33from house lights to streetlights.
27:36While you watch this video on a phone, computer or TV.
27:40If you're outside following traffic lights or displays,
27:43chances are you are relying on blue LEDs.
27:50We might even be getting too much of them.
27:53You may have heard warnings to avoid blue light from screens
27:56before bed because it can disrupt your circadian rhythm.
27:59That all comes from the gallium nitride blue LED.
28:05But as for lighting, there are virtually no downsides
28:08to an LED bulb.
28:10Compared to an incandescent or fluorescent bulb,
28:12they are far more efficient.
28:14They last many times longer, are safer to handle,
28:17and are completely customizable.
28:1930 years after the first white LED,
28:22high-end bulbs today
28:23allow you to choose between 50,000
28:26different shades of white.
28:28Most importantly, their price has come down to only a couple
28:31of dollars more than other types of bulbs.
28:34And at their efficiency, with average daily use
28:37and electricity pricing,
28:38you can recoup that cost in only two months
28:41and continue to save for years after that.
28:44The result is a lighting revolution.
28:47In 2010, just 1% of residential lighting sales
28:50in the world were LED.
28:52In 2022, it was over half.
28:56Experts estimate that within the next 10 years,
28:58nearly all lighting sales will be LED.
29:02(soft music)
29:03The energy savings will be enormous.
29:05Lighting accounts for 5% of all carbon emissions.
29:09A full switch to LEDs could save an estimated 1.4 billion
29:13tons of CO2,
29:14equivalent to taking almost half the cars
29:17in the world off the road.
29:21Today, Nakamura's research is on the next generation
29:24of LEDs, micro LEDs, and UV LEDs.
29:28- [Derek] So what are they making in there?
29:31- LEDs, lasers, power devices.
29:34This is one the best facility in the US.
29:37- And this is because of you?
29:41What's a standard LED size?
29:43- [Shūji] 300 times 200 microns.
29:46- [Derek] Okay.
29:47- [Shūji] Smallest is five microns.
29:49- [Derek] That is insanely tiny.
29:51- So basically you can use that for like
29:53near-eye display such as AR and VR.
29:55- You could have like a retina display
29:57that's like right up here?
29:58- Yep.
29:59- A human hair would be about that thick.
30:00- [Shūji] Yep.
30:01- And that's a really, really tiny LED.
30:04UV LEDs could be used to sterilize surfaces like
30:07in hospitals or kitchens.
30:08Just flick on the UV lights
30:10and pathogens would be dead in seconds.
30:12- COVID-19, you know,
30:14UV LED companies' stock prices were going,
30:17skyrocketed because everyone expected to be
30:19using these UV LEDS.
30:21We can sterilize all the COVID-19, no?
30:24For emitting diode, we use indium gallium nitride.
30:27For UV, we use aluminum gallium nitride.
30:30[Derek] Okay.
30:31- [Shūji] 'Cause the band gap is much bigger.
30:33- [Derek] Do you think this is what's coming?
30:35- [Shūji] It's okay, it work, but the problem is the cost.
30:38The efficiency is less than 10%.
30:40The cost is very high.
30:42But if the efficiency becomes more than 50%,
30:45cost is almost comparable to the mercury lamp.
30:47- [Derek] And you think it will happen, right?
30:48Like the efficiency will go up?
30:50- [Shūji] Yeah, yeah, I think so.
30:51- It's just a matter of time.
30:52- Yeah, I think so.
30:54- [Derek] And he's even tackling one
30:55of the biggest challenges of our time.
30:58- [Shūji] I'm interested in physics.
30:59- [Derek] Me too!
31:00- I'm still interested in nuclear fusion.
31:02So recently I started the company of nuclear fusion.
31:05- Really?
31:06- Oh yeah, last year.
31:07- No way. - No way, aha.
31:11(soft music)
31:11- In 2014, Nakamura, Akasaki and Amano
31:14were awarded the Nobel Prize in physics
31:17for creating the blue LED.
31:19Shortly afterwards, Nakamura publicly thanked Nichia
31:22for supporting his work,
31:23and he offered to visit and make amends,
31:26but they turned down his offer
31:28and today their relationship is still cold.
31:31But perhaps even more important than the Nobel Prize,
31:35By the time Nakamura released his blue LED in 1994,
31:38he had published over 15 papers,
31:40and he finally received his doctorate in engineering.
31:44Today he has published over 900 papers.
31:48Throughout his entire journey,
31:51one thing has never changed.
31:53What is your favorite color?
31:55- Oh, blue.
31:58- [Derek] Was it always blue?
31:59Or only after you made the LED?
32:01- I was born in a fishing village.
32:03Fishing village.
32:04In front of the house is awesome like, ocean.
32:07Blue always.
32:12- While I was learning about Nakamura's story,
32:15I realized that what set him apart from the thousands
32:17of researchers trying to unlock the blue LED,
32:20it wasn't necessarily his knowledge,
32:22but his determination, critical thinking,
32:24and problem solving skills.
32:26Where others saw dead ends,
32:27he saw potential solutions.
32:29So if you're looking for a free
32:31and easy way to start building these skills
32:33for yourself right now,
32:34look no further than today's sponsor, Brilliant.
32:37Brilliant will make you a better thinker
32:39and problem solver while helping you build real skills
32:41and everything from math
32:43and data science, to programming technology,
32:45and you name it.
32:47On Brilliant, you'll learn through discovery
32:49by trying things for yourself,
32:50and you'll not only gain knowledge
32:52of key concepts, but learn to apply them
32:54to real world situations, all while building your intuition,
32:58giving you the tools
32:59to solve whatever problems come your way.
33:01There is so much to learn on Brilliant.
33:03They have thousands
33:04of interactive lessons on just about
33:06anything that you're curious about.
33:07And because each lesson is bite-sized,
33:10you can do them in just minutes.
33:11So if you're like me and made a resolution
33:13to learn something new every day this year,
33:15Brilliant will help you actually do it.
33:18The best part is you can learn from anywhere,
33:20right on your phone.
33:21So whenever you have a few minutes,
33:23you can be building a quicker sharper mind.
33:26So to try everything Brilliant has to offer for free
33:28for 30 days, visit brilliant.org/veritasium
33:32or click that link down in the description,
33:34and the first 200 of you will get 20% off
33:36Brilliant's annual premium subscription.
33:38So I wanna thank Brilliant for sponsoring this video,
33:41and I want to thank you for watching.

About this transcript

This page contains the full transcript of Why It Was Almost Impossible to Make the Blue LED by Veritasium, generated from the public captions YouTube serves with the video. The transcript has 5,224 words across 759 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.