00:00:01 Harry also worked with Shockley right at the start, and he also got a phone call to bring him into the industry.

00:00:10 He says at that time he was actually, he realized he was actually a real chemist at the time, he was working in the oil industry in catalysis,

00:00:20 and actually knew a little bit about oxide semiconductors.

00:00:23 But he didn't really understand what was going on with these newfangled semiconductors.

00:00:28 But since then he's had a tremendous experience, he's got a connection right back to the start of the industry,

00:00:37 and he's going to talk to us here about both what he's seen in the past and the connection to chemistry going forward.

00:00:44 So in this case, the chemists can give a warm welcome to one of their own.

00:01:14 Wow. What do you do now?

00:01:25 In a good ballgame, it occurred to me that it's the function of the leadoff man to do only one thing.

00:01:38 And Pat did it eminently well, to get on base.

00:01:42 Would you say that Pat got on base?

00:01:46 In fact, he's advanced over to third, and if the pusher onward who comes up second is supposed to advance him,

00:02:02 what do you do when the leadoff man goes all the way around and scores?

00:02:06 There's nothing left for the second in line to do.

00:02:10 Not quite here.

00:02:13 My duty is a pleasurable one, a very pleasurable one.

00:02:18 And that is to introduce to you a little bit of some of the prehistoric events that occurred in my lifetime,

00:02:31 which if you realize, looking at Gordon's wonderful work, are advances in one person's lifetime, starting from essentially zero.

00:02:45 Are there very many things you can think of that have gone as far and as fast as they have in semiconductors in one person's lifetime?

00:02:55 Almost unbelievable.

00:02:57 Of course, you have to get a little older to make it a lifetime, so I sneak up on it over Pat by being a little older.

00:03:04 But I brought along one of the first pieces of work that I had to work on, and it's right here in my hand.

00:03:11 You may not recognize this.

00:03:14 This is a piece of silicon, single crystal, vintage about, I would say, 1960, maybe a little earlier.

00:03:26 It's all of three-quarters of an inch wide.

00:03:30 May I take this over and put it alongside of Pat Gelsink's wonderful wafer?

00:03:37 Look at it.

00:03:39 That's a single crystal.

00:03:41 This is a single crystal, 300 millimeters versus just a few.

00:03:48 We were very proud of this.

00:03:52 I know Gordon was when I first met him, and he showed me a few of these to work with.

00:03:58 My mission will be, that I've been assigned, is to take you back a little bit more prehistorically and try to at least show you examples of some of the things that happened by the guys that I would like to say today had to do the work.

00:04:21 Pat's work is brilliant, unbeatable.

00:04:25 They're the designers, and they lay out what has to be done to meet our goals.

00:04:34 But now, it's a very simple matter.

00:04:37 Gordon Moore would say, okay, you know where you have to go?

00:04:41 Go out and do it.

00:04:43 Just like that.

00:04:45 Or maybe Bob Noyce would have said that in the same way.

00:04:49 Clear shot.

00:04:50 Bob would do it.

00:04:51 Well, let's see what we had to do in the early days to begin with in order to get to where Pat has brought us in story today by Gordon's law.

00:05:06 Chemistry and Moore's law from Shockley Semiconductor to today.

00:05:13 Is this the one I just pushed?

00:05:15 I'll just turn it off, and I can sit down now.

00:05:23 Okay, right.

00:05:25 Gotcha.

00:05:26 See how prehistoric you can get.

00:05:32 My lifetime starts, in real work, starts with a search by a gentleman who made a call one afternoon and said,

00:05:40 would you like to consider some new work over that which you were doing?

00:05:46 What I was doing was process development at Emeryville, California at Shell Chemical.

00:05:52 I said, well, may I ask who is this, please?

00:05:58 He said, well, I'm William Shockley.

00:06:00 Do you know William Shockley?

00:06:02 I said, well, yes, I do, sir.

00:06:04 Yes, I do.

00:06:06 Took the pipe out of my mouth, probably, to say that.

00:06:08 Took the pipe out of my mouth, probably, to say that.

00:06:13 We discussed what he wanted to do, and it sounded something good to me.

00:06:19 I'd only heard about the fact that a semiconductor at that time, as I knew it, was a funny-looking little bright piece of catalyst called cuprous oxide, which is stuck in a tube.

00:06:31 When you ran propylene over it with oxygen, you got acrolein.

00:06:36 That little device did magnificent things, and when it died, it changed color.

00:06:42 When Shockley said, do you know what a semiconductor is?

00:06:45 I said, well, I think so.

00:06:47 He said, how do you know?

00:06:49 He always got to the bottom of the question.

00:06:51 I said, well, cuprous oxide is nothing like that.

00:06:54 Nothing like that.

00:06:56 Why don't you come and we'll talk this over?

00:06:58 Well, I did.

00:07:00 The results are almost obvious.

00:07:02 Here are my projects that I was assigned to when I first walked into the Shockley Laboratory.

00:07:09 I just made a quick list of them to see how would you feel as a chemist.

00:07:14 Would you recognize any of this?

00:07:17 Of course, I had a little help.

00:07:19 There was another chemist in the group, and that chemist, thank goodness, was Dr. Moore.

00:07:24 I quickly got introduced to him, and he told me a few of the problems that he was working on, none of which I understood.

00:07:31 He said, well, there's an apron.

00:07:36 I talked about this last night at the dinner.

00:07:38 There's an apron, there's a coat, there's a pair of gloves.

00:07:41 There's HF and nitric acid over in the corner.

00:07:44 Just learn how to use those, and as you go, we'll help you along.

00:07:48 The only trouble with that was, having a good chemist around, was that he and a group of others elected to disappear shortly after I met them.

00:07:57 They left the Shockley Laboratory.

00:07:58 I wondered, did I drive them away because of the nature of the chemistry that I was bringing in?

00:08:05 I don't think so.

00:08:07 There were many other reasons why this could happen.

00:08:10 Look at the projects.

00:08:12 I think you'd recognize some of them.

00:08:14 Wax evaporation.

00:08:16 Carnauba wax evaporation.

00:08:18 Why would you want to work on Carnauba wax evaporation?

00:08:21 First, to find a wax that would evaporate at a reasonable temperature under 200 degrees.

00:08:26 Carbon-carbon bonds don't like that.

00:08:31 The reason we wanted to use wax evaporation, just as an example, is because Dr. Shockley wanted it.

00:08:38 He wanted to take this little one-inch piece of silicon and rig it so that we could cut the wafers and then evaporate little dots of wax on them.

00:08:50 The little dots of wax would be the spots where you could etch the crystal into its individual little chips.

00:09:01 The conventional way of doing this was to take a piece of old dirty black wax produced from the bottom of some kettle that shelled chemical and melt it and put it on a wafer with a little point.

00:09:14 Those were the dots.

00:09:16 Shockley thought, this is archaic.

00:09:18 Why can't you evaporate wax?

00:09:20 You can evaporate aluminum and a lot of metals.

00:09:23 I worked on this project a very short time.

00:09:26 You can evaporate wax.

00:09:29 The handy-dandy way of doing it, the old way, was much faster, much cheaper, and much better because the wax would die.

00:09:38 Then you had to replace the crucible full of the crud that was in there before, which you couldn't get out of there anyway after heating it so long at 200 degrees centigrade.

00:09:47 We went on to other projects.

00:09:49 There were resist problems.

00:09:51 I was introduced to a new resist, a present resist called KPR, Kodak photoresist, magic, light sensitive.

00:10:02 You could make patterns with it.

00:10:04 The only thing you couldn't do very well with it is it wouldn't resist HF and nitric acid very long.

00:10:11 HF and nitric acid had a tendency to creep under the photoresist.

00:10:16 We learned a new problem in our science of this kind of chemistry called lifting.

00:10:23 That lifting problem, to answer the advanced question, exists to today.

00:10:29 We have lifting problems.

00:10:32 There's where the chemists come in to try to solve it.

00:10:35 You see, Pat did a beautiful job of describing this wonderful palette of advanced designs and chips and what we look forward to.

00:10:44 How are you going to get there if resist keeps lifting?

00:10:49 Today, the most common resist used in the semiconductor industry is a positive resist.

00:10:55 This is a negative one.

00:10:57 We won't go into those details.

00:10:59 We'll hear about that later.

00:11:01 It's very successful, but has the same problems.

00:11:03 There are still lifting problems.

00:11:05 Let's see.

00:11:07 What else can I say?

00:11:09 Furnace development.

00:11:10 I embarked on the design of a furnace for Dr. Shockley's request to use for the purposes of diffusion of elements into silicon.

00:11:22 I think when it ended up, it was about half the size of one of these tables that I show you here.

00:11:29 It was about three times or four times as large as it need be.

00:11:34 I never really finished the project, but furnace development was important.

00:11:38 We needed to have these high temperature, constant temperature zones.

00:11:42 Well, so forth.

00:11:44 Of course, the real chemistry came in when I had to talk to engineers about dumping trichloroethylene down a sink where you had just dumped a whole bunch of hydrofluoric acid and nitric acid together.

00:11:58 It's not a good thing to do.

00:12:01 If you're standing too close to it, you may not get away from it at all.

00:12:05 Safety procedures were consistent of rushing over to the shower or to the sink and washing your hands because they got exposed to HF or something like that.

00:12:17 We had to work on safety procedures, a very tough chemical problem.

00:12:21 You know what I mean by that.

00:12:23 It has to be done, and chemists do it.

00:12:25 Finally, reagent chemical procurement.

00:12:29 Let me tell you, I thought that the essence of chemistry was illustrated by the term CP, chemically pure.

00:12:38 If you got chemically pure chemicals in glass bottles, you had what you really needed, the essence, far from the truth.

00:12:47 That's where we in semiconductor had to start.

00:12:51 We couldn't use anything other than chemically pure because that's all that was available.

00:12:55 Now, how do you convince a large chemistry outfit like Monsanto at that time that you may get up to using one liter bottles per week of this very special doubly, triply purified CP chemical that we want?

00:13:15 They looked at us like we were crazy.

00:13:18 When you get to buying railroad car quantities, we'll sell you some.

00:13:22 It just didn't exist.

00:13:23 Chemists had a few problems to solve.

00:13:26 Finally, I worked on a problem that Dr. Shockley was very interested in.

00:13:36 That had to do with the processing of silicon four-layer diodes, a device which Dr. Shockley invented and thought that one day would replace every electronic relay in every telephone system of the world.

00:13:50 By having this one switching device which could do the job, just a two terminal device, very simple, made out of one of these little silicon dots.

00:14:00 The fact of the matter was it just couldn't be made because we couldn't meet the specifications consistently and constantly and reliably.

00:14:11 It was a wonderful idea, but it was a paper idea.

00:14:15 Dr. Shockley, I don't think, liked that idea very much.

00:14:20 But such is what it was.

00:14:24 Things went on.

00:14:29 Now, the story became interesting.

00:14:32 Shortly after this little batch of work, I left Shockley Laboratories because I was very nicely invited by some of my old friends, whose friendship I wanted to renew, namely Bob Noyce and Gordon Moore,

00:14:47 to come and work at Fairchild.

00:14:51 Let me tell you, as a chemist, that was a beautiful explosion of light and sound and new things.

00:15:01 This is the first planar transistor, which starts at the bottom of Gordon Moore's curve, probably right at the intersection of zero, zero.

00:15:11 If you haven't seen this or heard about it, here is the invention by Jean Herny, one of our colleagues, both at Shockley and at Fairchild later.

00:15:22 It's a small chip made of silicon, of the very same silicon I just showed you right here.

00:15:32 It's made such that its junctions, PN junctions, fall underneath the protective oxide layer of silicon dioxide on top of silicon.

00:15:44 Let me tell you, the success of the integrated circuit industry is due to that wonderful, wonderful material that grows naturally on silicon and its silicon dioxide.

00:15:58 That's the world that we've been talking about.

00:16:01 If we didn't have that and we had only, say, germanium, for example, we would not even be close to where we are today.

00:16:10 I worked on this for all of two years.

00:16:13 Just when it was getting to be fun, in his inimitable fashion, Bob Noyce came running into our home one afternoon and said,

00:16:23 Cello, enough of what you've been doing. We need you to go to Italy.

00:16:28 Are you willing to go? What do you say?

00:16:33 Of course, with all kinds of quaking knees, but to do what?

00:16:39 We've just formed an alliance with a company called Società Generale Semiconduttori, SGS.

00:16:46 We're a partner now with Olivetti and Telletra, an instrument company.

00:16:52 Fairchild is now a third member in that triumvirate.

00:16:57 We're going to introduce silicon transistors into Europe because they're not making them at SGS and they would like to.

00:17:06 They are now making germanium devices.

00:17:09 I had never even barely heard of germanium, let alone seen it.

00:17:13 When I got to SGS, I'll summarize in one sentence.

00:17:18 They had something like, at that time, a million dollars worth of inventory of point contact diodes and simple junction transistors made out of germanium.

00:17:30 They were all unreliable.

00:17:33 They did not stand up to the rigorous treatment of life testing in the user's hands because germanium has problems.

00:17:43 The surface is sensitive. There is no protective, wonderful oxalide like there is in silicon.

00:17:50 Germanium went the way of the buggy whip and all of the old obsolete things.

00:17:57 We replaced that at SGS.

00:18:00 I was the operations manager at that point, manufacturing, at that time, assembling these devices.

00:18:08 SGS became very successful at it.

00:18:12 It was a point of entry into the market for Europe for Fairchild, who was then a new company trying to really grow into its market.

00:18:19 You've seen this one very recently.

00:18:28 Pat has that same picture.

00:18:31 I'll just repeat the story.

00:18:35 This is the first planar silicon integrated circuit.

00:18:40 It's a flip-flop.

00:18:43 It's one which was invented, by the way, by Bob Noyce.

00:18:48 To illustrate Bob Noyce's personality, I once asked him,

00:18:52 Bob, why didn't we copyright the name planar?

00:18:58 Such a thing became stupendous in the industry.

00:19:01 Everybody wanted to use it.

00:19:03 Bob said, no need.

00:19:05 They're all going to get to use it anyway.

00:19:07 Why don't we just encourage everyone to think planar?

00:19:10 We don't have to copyright it.

00:19:12 We've got the patent on the project anyway, on the device itself.

00:19:17 The royalties that one gets for this.

00:19:20 The planar name is free to be used by anyone.

00:19:23 That was Bob Noyce.

00:19:25 Very easy.

00:19:27 No problem.

00:19:29 We do it that way.

00:19:31 You'll notice that this is a circular chip, not a rectangular chip.

00:19:34 Of course, if you listen to my little story about drops of wax, this is why it's round.

00:19:44 You put a wax chip on top of an integrated circuit that you've completed on a wafer.

00:19:49 Then, instead of doing the modern way, which is to scribe and break that chip into rectangles,

00:19:57 you simply just etch your way around it.

00:19:59 You may get a circular chip because that's the pattern of wax that lays on it.

00:20:05 Well, you can see that whoever did this one, it wasn't me.

00:20:09 I doubt if it was Gordon.

00:20:10 It might have been some engineer who knew the designs beautifully, but didn't understand the chemistry.

00:20:17 You'll notice that it's missing a couple of pads, these bright rectangular pads up in the upper corner.

00:20:26 The reason those pads are missing is because the wax lifted.

00:20:32 We can't tolerate that kind of stuff, but we did in this particular instance.

00:20:37 It was early on in our existence.

00:20:38 However, things got better.

00:20:41 I would like to direct your attention to not the tail of the beautifully expanding curve of billions of transistors per chip,

00:20:54 but direct it down to the other end, close to zero.

00:20:58 This is a family of integrated circuits.

00:21:01 You'll recognize that we made it fair child in the period of 59 to 67, as shown.

00:21:06 You'll notice our little friend up there in the upper left-hand corner is the planar transistor.

00:21:12 Then you have a string of other chips.

00:21:15 These chips—I tried this at home, and if I did it right, counting the elements—

00:21:21 these chips plot out very beautifully at the beginning of Moore's curve.

00:21:28 Pats started in 1970, so we didn't look into what happened before 1970, but thank you for leaving us the space.

00:21:37 These plot right in very nicely according to the number of components right on the curve.

00:21:44 By the way, for your information, that big monstrous chip in the lower right-hand corner

00:21:52 had a total of 128 transistors on one chip.

00:21:58 It wouldn't even be a piece of a decimal point on anything we talk about today, 128 chips.

00:22:06 Not only that, it was with two layers of interconnected metal,

00:22:11 two layers of aluminum on top of the silicon and one on top of the silicon oxide on top of that.

00:22:18 We have here the real initiation of Gordon's curve.

00:22:27 Later, of course, when the work was published,

00:22:32 these came out to illustrate the beginnings of Moore's law,

00:22:37 which Gordon so brilliantly defined back in that period.

00:22:42 Well, pushing along, how do you describe, other than my little anecdotes,

00:22:50 what a chemist will do in silicon manufacture, even up to today?

00:22:58 Well, I have to take a rather narrow-minded, selfish view of it.

00:23:03 I have to say, you guys, you design them, we'll make them.

00:23:08 Well, there's a lot betwixt cup and lip on that one.

00:23:13 The designs only result when the design rules are put into them properly,

00:23:20 and the design rules result from what chemists and the few brilliant physicists

00:23:27 who understand chemistry can work with this in order to design those devices.

00:23:34 I'll show you a couple of the limitations,

00:23:38 just by examples of what we succeeded to do at Fairchild.

00:23:44 I was very fortunate at the work in SGS in Italy because I loved it.

00:23:51 I like working with all these wonderful Italian engineers, but I missed home.

00:23:57 I really did.

00:23:58 I was tired of trying to explain to a group of engineers

00:24:02 just what the meaning of teamwork is,

00:24:06 because every time I talked about teamwork, I got the answer back,

00:24:10 well, if we solve the problem, who gets the credit?

00:24:12 I said, the team.

00:24:14 But we have six people on this team.

00:24:16 Who's going to get the credit?

00:24:18 I said, don't worry about it.

00:24:20 You'll get the credit.

00:24:22 All of you will get the credit.

00:24:24 Well, that didn't wash.

00:24:25 It made me react to Gordon's invitation once again very happily

00:24:31 when he said, look, we've just really reorganized

00:24:34 and really started up our R&D department at Fairchild.

00:24:39 I want you to come back and become the department head of materials and processes,

00:24:44 stuff that you really should know a little bit about already.

00:24:48 We'll work a department with senior chemists, senior scientists,

00:24:56 experienced ones, and we had about 20 in our group at that time,

00:25:00 plus a lot of assistance to them.

00:25:06 We all have two missions, as he explained.

00:25:10 I remember these vividly.

00:25:11 One mission is you're going to develop and improve our existing processes,

00:25:19 develop new ones, and improve our processes.

00:25:23 The second mission will be one that I didn't quite appreciate,

00:25:28 but Gordon saw so far ahead that it really was becoming true.

00:25:33 He said, the other part of your missions is that your senior scientists

00:25:38 are expected to go to work in the factory

00:25:41 and go to work in production.

00:25:43 Whenever there's a problem of something to do with making these devices

00:25:48 that the engineers, as sharp as they are in the factory, can't work with,

00:25:53 I want your guys to stop what they're doing,

00:25:56 and he made that very patently clear,

00:25:58 and get out there and go to work with them.

00:26:01 Process, handle wafers with them,

00:26:04 stick them into one end of the furnace and take them out of the other.

00:26:07 Work with the resist and all of that.

00:26:08 Now, I don't know if you have a full appreciation of the comment,

00:26:13 without running into the problem,

00:26:15 of telling a senior Ph.D. that he has to stop his research project

00:26:21 and go out and work in the factory on a plebeian problem,

00:26:24 that he comes back and says,

00:26:26 well, those guys, those production people don't know what they're doing.

00:26:30 They're mucking everything up, and it's a mess.

00:26:33 Get them out of there. Do something with it.

00:26:35 Well, that is not the proper attitude

00:26:38 for what we called at that time a type of sustaining engineering.

00:26:42 We expected our Ph.D.s to pick up wafers

00:26:47 and hand-carry them through the existing processes

00:26:51 to prove that everything can be done.

00:26:54 One of them uncovered a fresh new man.

00:27:00 I can't say he was an engineer or a chemist.

00:27:02 I mean, you don't know, so I don't want to accuse either one.

00:27:05 He was processing what looked like a bunch of brightly colored wafers.

00:27:11 I was called over by Bill Shepard, who was doing the job.

00:27:18 He said, guess what I found?

00:27:21 They're processing two colors of silicon.

00:27:24 I said, what do you mean two colors of silicon?

00:27:26 It's got oxide on it,

00:27:28 and oxide comes in various transparencies, et cetera, et cetera.

00:27:32 Well, this young chap, scientist, whatever you want to call him,

00:27:37 production man, had gone to Sears and gotten two cans of paint,

00:27:41 one red, one green.

00:27:43 He divided his batch of wafers,

00:27:46 which was about at that time something like 50 slices,

00:27:50 and he had painted one batch red and one batch green,

00:27:54 and when the paint sufficiently dried,

00:27:56 he washed it, and he put them through the furnace.

00:28:00 I said, why is he doing that?

00:28:02 I asked him, all three of us, why are you doing this?

00:28:06 Well, they asked us, they being the managers at the factory,

00:28:10 asked us to identify the wafers.

00:28:13 We want to know by code from which batch of wafers the highest yields came,

00:28:18 and we want to be able to have wafer history.

00:28:22 Well, this guy had a brilliant idea.

00:28:23 Of course, mark them.

00:28:25 So he painted them.

00:28:27 Well, of course, to some chemists in the crowd,

00:28:30 and even to some engineers,

00:28:32 even Pat Gelsinger would not have permitted that kind of thing,

00:28:36 or especially Pat wouldn't have committed that kind of thing.

00:28:39 Forgive me, Pat.

00:28:41 But there was no knowledge at the working level

00:28:47 of bachelor's level, master's level technicians

00:28:50 to exactly what was going on.

00:28:54 So it was a problem of the senior scientists

00:28:58 who came in there and just went through,

00:29:00 and they stopped all the furnaces,

00:29:02 and we spent days helping them clean up the affair.

00:29:05 This was a gross mistake, but there were many like that.

00:29:09 I want to jump ahead and show you three examples

00:29:13 of the types of projects that our senior scientists were working on

00:29:17 that were needed in the industry.

00:29:18 This is, as you see at the bottom here,

00:29:24 if I push the right one, shouldn't I get a pointer?

00:29:27 Well, at the very bottom, there it is.

00:29:29 I'm wiggling you too fast.

00:29:31 At the very bottom is a dual inline package.

00:29:35 It's a package which is composed of two level,

00:29:38 two pieces of ceramic like a sandwich.

00:29:40 Sandwich in between is a frame

00:29:42 which holds the integrated circuit at that period.

00:29:45 This one is a very small one.

00:29:46 It has 14 leads on it, seven on each side.

00:29:50 That package was invented at Fairchild,

00:29:54 and it was revolutionary at the time it was invented.

00:29:58 Memory products, microprocessors

00:30:00 were put into dual inline packages.

00:30:02 Others were invented.

00:30:04 At the top is little glock tops we used to call them,

00:30:10 little round beads of epoxy

00:30:12 which were put onto ceramic pieces,

00:30:14 and transistors were encapsulated

00:30:16 in them for only one reason.

00:30:18 That was to make something

00:30:20 that would cost under two cents a piece.

00:30:23 If we could do that,

00:30:25 we could sell it all the way

00:30:27 through the entertainment industry.

00:30:29 We asked Ralph Oldberg at that time,

00:30:31 bless his heart.

00:30:33 Also, he used to work at Shell Development.

00:30:35 Ralph came down, took a job,

00:30:37 and he took on the job of making

00:30:39 what we called the glock tops,

00:30:41 those little black beads up in the upper center of the lot.

00:30:45 The other products are also packages

00:30:47 worked on by the chemists

00:30:51 and the related ceramicists

00:30:54 or materials engineers

00:30:56 who are sort of half chemists in our group.

00:31:02 This was one of the most fascinating problems

00:31:06 found in our history.

00:31:09 It was called the electromigration of aluminum.

00:31:14 Aluminum is a wonderful metal,

00:31:17 and it revolutionized the industry

00:31:21 by allowing us to make complex integrated circuits

00:31:24 using aluminum as a conducting path.

00:31:27 When you look through an electron microscope

00:31:30 at 1500 magnification,

00:31:32 at a strip of aluminum conducting electricity on a chip,

00:31:35 you see this kind of strip.

00:31:38 It had a problem, however.

00:31:40 Under the influence of the flow of electrons

00:31:42 from left to right,

00:31:44 as they want to flow on a strip,

00:31:47 they caused, at high densities of current,

00:31:51 they caused voids at one end

00:31:53 and pills at the other end.

00:31:56 The aluminum was fluid.

00:31:58 It became crystalline.

00:32:00 The crystals moved.

00:32:02 That was a basic design characteristic

00:32:05 which we uncovered at Fairchild,

00:32:07 reported for the first time

00:32:09 as a reliability problem.

00:32:11 It was a basic nature of aluminum.

00:32:14 Today, there are still those same design rules

00:32:19 and those wonderfully complex structures

00:32:22 that Pat described.

00:32:24 Let me pick up the pace here a moment.

00:32:27 This was a tremendous, wonderful example

00:32:31 of the problem in the industry

00:32:33 worked on by Dr. Deal, Grove,

00:32:37 myself, Ed Snow, and others.

00:32:42 The problem, if I can get through this in time,

00:32:45 will illustrate itself.

00:32:47 At this time in industry,

00:32:49 we were making MOS transistors

00:32:51 and MOS integrated circuits.

00:32:53 They had a problem of instability.

00:32:56 You could apply a voltage to them

00:32:58 and where the voltage was supposed to stay,

00:33:00 it drifted.

00:33:02 Why did they drift?

00:33:03 They had here something like

00:33:06 seven or eight different explanations

00:33:09 of why of the same problem

00:33:11 which all others had seen in the industry.

00:33:14 They were walking around

00:33:16 like blind men exploring an elephant.

00:33:18 What are the reasons

00:33:20 why these MOS devices drifted

00:33:23 all the way from BTL at Baird

00:33:27 on the left-hand side

00:33:29 who said,

00:33:30 whatever that was, he didn't explain.

00:33:32 Up in the right-hand side,

00:33:36 we had Shockley and Hooper.

00:33:38 In the latter years of Shockley's work,

00:33:40 he said, well, they're due to surface ions.

00:33:42 Well, he was close.

00:33:44 We had several other theories,

00:33:46 oxygen vacancies,

00:33:48 Lindmeier at Sprague on the right-hand side,

00:33:50 there's a heterojunction being formed there

00:33:52 and this heterojunction is leaky

00:33:54 and therefore we have to clean that up.

00:33:56 The answer lay with those wonderful chemists

00:33:58 who worked with Snow, Grove and Diehl,

00:34:02 two of them chemists themselves,

00:34:05 and defined that we had a simple problem.

00:34:08 The devices were not clean enough.

00:34:10 They were susceptible to being contaminated

00:34:14 by odd garden variety sodium and potassium ions.

00:34:19 How to get that out of there?

00:34:21 Well, we worked out the process at Fairchild

00:34:23 and we made a unit

00:34:25 which we called planar two processing

00:34:29 sort of for an advertising gimmick

00:34:31 and we published the fact

00:34:34 that alkali-free processing in this ad

00:34:38 and of course the scientific establishment,

00:34:41 our own scientific establishment

00:34:43 said another one of these baloney types of ads

00:34:47 that semiconductor companies are putting out all the time

00:34:50 about what they can do,

00:34:52 but it was exactly the answer

00:34:53 that we gave to a conclave at Allentown to BTL

00:34:57 to explain to them

00:34:59 why we were making stable MOS devices

00:35:03 where others couldn't.

00:35:05 We just simply cleaned them up.

00:35:07 Planar two processing

00:35:10 and that was a move forward

00:35:12 that extended our capability

00:35:14 into the design area

00:35:17 and to this day

00:35:19 that factor has to be taken into consideration.

00:35:22 Well, that's the end of the story sort of.

00:35:26 That was the work of the chemists at Fairchild.

00:35:31 Unfortunately, some sad things happened.

00:35:34 Both Bob Noyce and Gordon Moore

00:35:36 for their own good reasons

00:35:38 left Fairchild to form Intel.

00:35:40 Thank goodness they succeeded.

00:35:43 There were a raft of companies.

00:35:45 I counted over 150

00:35:47 in the first two years after they left Fairchild

00:35:49 150 that were formed in the valley

00:35:52 where we live and work.

00:35:54 We lost men.

00:35:56 LEGO was going out of style

00:35:58 to those new companies, to new jobs.

00:36:01 It was an extremely fluid time.

00:36:03 I was assigned to a new job

00:36:06 called Director of International Technology

00:36:09 and my job was to take our technology wherever we could

00:36:13 and plant it in a foreign country by an alliance

00:36:17 by which we could

00:36:20 penetrate the marketplace.

00:36:23 We had a very successful result

00:36:26 in the country of Hungary

00:36:28 with the Tungzum Company

00:36:30 where we wrote a 10-year contract

00:36:32 which was very profitable to Fairchild

00:36:34 which I succeeded in negotiating.

00:36:37 I only say that because of all the pain

00:36:40 that occurred at the time.

00:36:42 It was a thought that said for the future

00:36:45 if we want to get into marketplaces

00:36:47 we should not sell our technology

00:36:49 we should sell it for market share.

00:36:52 That is the vital project

00:36:54 that we need to do even today.

00:36:56 One word about the future

00:36:59 if I may just leak over a little bit.

00:37:02 It's only just beginning for the chemists.

00:37:06 The sort of thing that Pat talked about

00:37:09 should be a challenge to chemists

00:37:12 that they never saw before.

00:37:14 This is where it will lie.

00:37:16 Carbon nanotubes have a tremendous application

00:37:20 to semiconductor work

00:37:22 but it's going to take some chemists

00:37:24 along with the design engineers

00:37:26 bless their hearts.

00:37:28 It will take some chemists to define

00:37:30 how the heck can we make carbon nanotubes

00:37:32 into the proper conducting paths

00:37:34 and even semiconducting paths

00:37:36 for use in devices.

00:37:38 Down to LED technology.

00:37:40 We said only a word about that.

00:37:43 Light emitting devices

00:37:48 are devices that we never really worked on

00:37:52 in our past in silicon work

00:37:54 because they just weren't applicable

00:37:56 to the applications that exist today.

00:37:59 Today we're talking about organic

00:38:01 light emitting devices.

00:38:03 OLEDs and LEDs.

00:38:05 It's going to take chemists

00:38:07 to try to figure out

00:38:09 how do you take an organic molecule

00:38:10 like a phenolic chain

00:38:12 and stick it between two gold electrodes

00:38:14 and have it do a semiconducting job

00:38:17 and emit light at the same time.

00:38:19 We have quite a future for chemists

00:38:21 in our industry

00:38:23 and we're looking forward to it.

00:38:25 Thank you.

00:38:26 Thank you.

00:38:38 We have a 20-minute break now.

00:38:41 Maybe it's a 15-minute break.

00:38:43 If there's one burning question,

00:38:46 I will take it.

00:38:48 Otherwise, I suggest that you catch Harry

00:38:50 during the break.

00:38:52 I'd like to thank him again.

00:38:54 It was great to see.

00:38:56 It's great to see how so few people

00:38:58 can change the world.

00:39:00 There's just a connection

00:39:02 between a few key people

00:39:04 that really did it.

00:39:06 Thanks.

00:39:08 Thank you.

00:39:10 Thanks, John.

00:39:12 I kind of ran over there.

00:39:14 I got a little hung up in Shockley.

00:39:16 Sorry about that.

00:39:18 He was a guy that could wind you up too.

00:39:20 Does that seem...

00:39:30 Well, this morning,

00:39:32 we heard two very nice presentations

00:39:34 talking about the history

00:39:36 of Moore's Law,

00:39:38 talking about the extrapolation

00:39:40 of Moore's Law,

00:39:42 and talking about some of the problems

00:39:44 involved in processing

00:39:46 of microelectronic devices.

00:39:47 We're going to shift gears

00:39:49 just a little bit now,

00:39:51 and the next two talks

00:39:53 will deal with materials

00:39:55 and chemicals involved

00:39:57 in these processes.

00:39:59 So I have the pleasant task

00:40:01 of introducing Dr. Elsa Reichmanis.

00:40:03 As Elsa completed her PhD,

00:40:05 she had to decide

00:40:07 whether she would embark

00:40:09 on an academic career

00:40:11 or an industrial career.

00:40:13 Unfortunately for academia,

00:40:15 she chose the industrial route

00:40:17 to Bell Laboratories.

00:40:19 Her decision

00:40:21 to go into industrial research

00:40:23 was really based on her interest

00:40:25 in performing some research

00:40:27 that really mattered

00:40:29 and would have some clear applications,

00:40:31 would make a difference

00:40:33 in the world that she lived in.

00:40:35 Her investigations

00:40:37 into photoresist materials,

00:40:39 electronic paper,

00:40:41 and organic materials

00:40:43 for telecommunications

00:40:45 has certainly indicated

00:40:47 that she would be a good fit

00:40:49 for the postdoctoral program.

00:40:51 She earned her Ph.D.

00:40:53 and Ph.D. degrees

00:40:55 in chemistry from Syracuse University,

00:40:57 remained there for a few years

00:40:59 as a postdoc,

00:41:01 and then joined Bell Laboratories

00:41:03 in 1978,

00:41:05 and she's currently

00:41:07 Director of the Materials

00:41:09 Research Laboratory there.

00:41:11 There have been a number

00:41:13 of awards and recognitions

00:41:15 that Elsa has received.

00:41:17 For example,

00:41:19 she was awarded the

00:41:21 Nobel Peace Prize

00:41:23 and the Nobel Peace Prize

00:41:25 and the Nobel Peace Medal.

00:41:27 She's a member of the

00:41:29 National Academy of Engineering.

00:41:31 She also served as president

00:41:33 of the American Chemical Society

00:41:35 in 2003.

00:41:37 So I'm very pleased

00:41:39 to present Elsa Reichmanis,

00:41:41 who will present her talk

00:41:43 Resists Yesterday, Today,

00:41:45 Tomorrow.

00:41:47 It does work.

00:41:49 Every once in a while

00:41:51 I get technology challenged

00:41:53 with the projectors,

00:41:55 so hopefully I won't

00:41:57 inadvertently shut this off.

00:41:59 It's an honor for me

00:42:01 to be here first,

00:42:03 and this morning's presentations

00:42:05 are truly a tough act to follow.

00:42:07 What I'd like to do

00:42:09 is present a little bit

00:42:11 of a perspective

00:42:13 of where chemistry plays a role,

00:42:15 and this will be getting

00:42:17 a little bit more

00:42:19 chemical-oriented,

00:42:21 and really talking fundamentally

00:42:23 about the technology

00:42:25 that's used to pattern

00:42:27 integrated circuits today.

00:42:29 So it's really

00:42:31 the resist part,

00:42:33 the goop that is laid down

00:42:35 on silicon wafers

00:42:37 that leads to patterns

00:42:39 that are increasingly smaller.

00:42:41 And I think it's largely

00:42:43 the materials,

00:42:45 the chemistry,

00:42:47 that have allowed

00:42:49 integrated circuits

00:42:51 to go from the initial

00:42:53 single crystal transistor

00:42:55 that was invented at Bell Labs

00:42:57 now over 50 years ago,

00:42:59 from something which is

00:43:01 really a few inches in size

00:43:03 to devices where transistors

00:43:05 on this eight-inch silicon wafer

00:43:07 are on the order

00:43:09 of hundreds of nanometers,

00:43:11 100 nanometers.

00:43:13 Putting that size scale

00:43:15 in perspective,

00:43:16 we're getting a photoresist

00:43:18 in comparison

00:43:20 to an E. coli bacterium cell,

00:43:22 which is about one micron

00:43:24 by five microns across.

00:43:26 And a human hair

00:43:28 is approximately

00:43:30 100 microns in diameter.

00:43:32 So we're getting

00:43:34 really, really small,

00:43:36 and the size of those features

00:43:38 is shrinking even more

00:43:40 where we're talking about

00:43:42 truly nanometer-scale

00:43:44 transistor devices.

00:43:46 We're getting millions of these

00:43:48 into a single IC chip.

00:43:51 And what I won't be talking about

00:43:53 are efforts also concurrent

00:43:56 that are sort of going

00:43:58 larger in size,

00:44:00 where these are now

00:44:02 all organic or plastic

00:44:04 electronic devices,

00:44:06 where we're now looking

00:44:08 at printing large features

00:44:10 on very flexible plastic

00:44:12 substrates to allow

00:44:14 high-throughput,

00:44:15 where large material

00:44:17 is going to replace

00:44:19 sort of traditional

00:44:21 silicon-based semiconductor technology.

00:44:23 What they might do

00:44:25 is enable new technologies

00:44:27 and something different,

00:44:29 something silicon can't do.

00:44:31 But let's focus on

00:44:33 the silicon side

00:44:35 resists and lithography.

00:44:37 Certainly the driver

00:44:39 for getting from this

00:44:41 multi-inch

00:44:43 single crystal

00:44:45 to nanometer-scale devices

00:44:47 has been cost.

00:44:49 I like this particular

00:44:51 insert that came from Siemens

00:44:53 that really puts the cost

00:44:55 of a megabit of information

00:44:57 in perspective,

00:44:59 where in 1973

00:45:01 the cost of that information

00:45:03 was about equivalent

00:45:05 to the cost of buying

00:45:07 a house in Germany

00:45:09 or 150,000 Deutschmarks.

00:45:11 By the mid-1990s,

00:45:13 the cost of that same

00:45:15 number of Deutschmarks

00:45:17 and we've continued

00:45:19 to decrease in cost.

00:45:21 This looks at it another way,

00:45:23 where again,

00:45:25 we've got Moore's Law

00:45:27 where we're seeing

00:45:29 a doubling of transistors

00:45:31 approximately every 18 months

00:45:33 to two years.

00:45:35 And that's been enabled

00:45:37 by a number of elements,

00:45:39 feature size,

00:45:41 increase in wafer size,

00:45:43 certainly yield improvements

00:45:45 over time.

00:45:47 Orthography is a process

00:45:49 where we take

00:45:51 this silicon wafer,

00:45:53 now 300 millimeters across,

00:45:55 12 inches,

00:45:57 the size of a small pizza,

00:45:59 coat it with a resist,

00:46:01 and it's typically

00:46:03 a photoactive polymer-based material,

00:46:05 expose that resist

00:46:07 with a pattern,

00:46:09 develop that material

00:46:11 in a suitable developer,

00:46:13 transfer the pattern

00:46:15 to a substrate,

00:46:17 and then continue

00:46:19 to coat layers of the device

00:46:21 onto that substrate.

00:46:23 How do we do that lithography?

00:46:25 Certainly traditionally

00:46:27 we've been using photolithography

00:46:29 where we've seen a reduction

00:46:31 in the wavelength of light

00:46:33 that's been used over the years.

00:46:35 What we're also seeing

00:46:37 is that if we look

00:46:39 at the theoretical resolution limit

00:46:41 based on enhancement technologies,

00:46:43 we're seeing a drastic difference

00:46:45 in the perspective

00:46:47 of what that theoretical resolution limit

00:46:49 is going to be,

00:46:51 where what we heard this morning

00:46:53 is that we're printing images

00:46:55 that are about a quarter

00:46:57 of the size of the wavelength of light

00:46:59 that we're actually using

00:47:01 to image the resist.

00:47:03 We're really seeing

00:47:05 a drastic reduction

00:47:07 in the exposure wavelength,

00:47:09 a drastic reduction

00:47:11 in the technology

00:47:13 that's used to fabricate

00:47:15 those images.

00:47:17 The wafer tool costs

00:47:19 are increasing

00:47:21 rather dramatically,

00:47:23 but what is also interesting

00:47:25 is that as we heard this morning

00:47:27 is even with this

00:47:29 increase in lithography cost,

00:47:31 the cost of the individual devices

00:47:33 is not going up astronomically.

00:47:35 It's going to be interesting

00:47:37 to see where the technology goes

00:47:39 and where ultimately

00:47:41 we're going to end up.

00:47:43 Advances in the technology

00:47:45 are being driven

00:47:47 by the Semiconductor Industry Association,

00:47:49 a group of companies

00:47:51 interested in the industry

00:47:53 get together

00:47:55 and think about options.

00:47:57 What are the problems

00:47:59 that need to be solved

00:48:01 near-term, long-term?

00:48:03 Where are innovations needed?

00:48:05 Where do we need true invention

00:48:07 where we need

00:48:09 a revolutionary change

00:48:11 in how we're doing things?

00:48:13 When we look at lithography

00:48:15 currently we're looking at

00:48:17 lithographic technologies,

00:48:19 we're looking at photoresist materials,

00:48:21 optical enhancement technologies.

00:48:24 How can we get this resolution limit

00:48:27 for optical lithography

00:48:29 still lower?

00:48:31 How do we integrate

00:48:33 the various enhancements

00:48:35 in order to more effectively

00:48:37 fabricate

00:48:39 the devices that we're interested in?

00:48:41 We're looking at new

00:48:43 future lithographic technologies

00:48:45 whether it's 157 nanometer imaging,

00:48:48 EUV, projection E-beam,

00:48:50 X-ray, maskless approaches,

00:48:52 and also other

00:48:54 quote-unquote novel ideas.

00:48:56 When we're developing

00:48:59 a resist technology

00:49:01 and a lithographic technology,

00:49:03 while chemistry plays

00:49:05 a very important role,

00:49:07 chemistry can't do it alone.

00:49:09 We really need to be very interactive

00:49:11 working with process engineers,

00:49:13 working with device designers,

00:49:15 working with tool designers

00:49:17 in order to be able to develop materials

00:49:19 that are really going to be effective

00:49:21 for device fabrication

00:49:23 within the current environment.

00:49:25 Some of the things we need to address

00:49:27 are the radiation response.

00:49:29 Is the material going to be sufficiently sensitive

00:49:32 to be able to have

00:49:34 a sufficient throughput through a tool?

00:49:36 Is it going to respond

00:49:38 to incident or radiation?

00:49:40 What's the contrast

00:49:42 of that material going to be?

00:49:43 Are we going to be able to get

00:49:45 a sufficiently high resolution image?

00:49:47 What is that resolution?

00:49:50 What sort of line width control do we need

00:49:52 when we're printing that image

00:49:54 and transferring it into the substrate?

00:49:56 In terms of defect densities,

00:49:58 what's tolerable?

00:50:00 Today, we basically don't want any defects

00:50:02 across this pizza pie wafer.

00:50:05 What's the etching resistance

00:50:07 of those materials?

00:50:09 Is the resist going to withstand

00:50:11 the pattern transfer processes

00:50:13 of transferring that resist image

00:50:16 into the device substrate?

00:50:18 Then there are other little things

00:50:20 that we need to worry about like adhesion.

00:50:22 The stuff has to stick

00:50:24 or we're not going to get any kind of a yield.

00:50:26 The resist itself

00:50:28 has to be the same.

00:50:30 We have to worry about the supply.

00:50:32 We have to worry about the quality assurance.

00:50:34 Every time we get a new batch of material,

00:50:37 it's got to behave exactly the same way

00:50:40 because without that,

00:50:41 we're not going to have a reliable process.

00:50:44 It's got to stay around

00:50:46 for some reasonable period of time

00:50:48 without changing to something else

00:50:50 so that we, again,

00:50:52 have a reasonable process.

00:50:54 It's got to be sufficiently low cost

00:50:56 to make it worthwhile.

00:51:02 Looking at the evolution

00:51:04 of resist materials,

00:51:06 I'm starting here in 1970

00:51:08 or in that time frame

00:51:09 which was the transition point

00:51:11 of going from the negative resists

00:51:13 that we heard about this morning

00:51:15 to positive tone resists.

00:51:17 The first positive tone resist

00:51:19 was a Novolac-based material.

00:51:21 These are basically

00:51:23 what form conventional photoresists.

00:51:26 It was introduced for proximity printing

00:51:28 of 10-micron design rule devices.

00:51:32 Within about a decade,

00:51:34 G-line lithography

00:51:36 using photoresist

00:51:37 and G-line lithography

00:51:39 using 436-nanometer light

00:51:42 began to dominate

00:51:44 and Novolac resists

00:51:46 were continuing to dominate

00:51:48 fabrication lines.

00:51:50 It was at about this time

00:51:52 that deep UV resist materials research

00:51:54 was initiated

00:51:56 where deep UV at the time

00:51:58 was using a mercury lamp

00:52:00 with wavelengths of anywhere

00:52:02 from about 220 out to 250 nanometers.

00:52:04 At this time,

00:52:05 it was thought that optical lithography

00:52:07 and this is about

00:52:09 when I joined Bell Labs

00:52:11 that optical lithography

00:52:13 was going to hit a roadblock

00:52:15 when we got to about

00:52:17 2-micron design rule devices

00:52:19 that we're never going to be able

00:52:21 to use conventional resists

00:52:23 and conventional lithographic processes

00:52:25 to get smaller than

00:52:27 somewhere between a micron

00:52:29 and 2-micron size devices.

00:52:31 At that time,

00:52:32 there was interest in

00:52:33 two different lithographic options

00:52:35 one of which was

00:52:37 reducing the wavelength of light

00:52:39 that you're using to the UV range.

00:52:41 All right, what are the chemistries?

00:52:43 This shows conventional photoresist chemistry

00:52:45 in Novolac resin

00:52:47 that's soluble in an aqueous-based developer.

00:52:49 You add to it

00:52:51 anethoquinone dizide

00:52:53 as a dissolution inhibitor.

00:52:55 This combined material

00:52:57 is insoluble in aqueous-based,

00:52:59 shine light on it

00:53:01 in a pattern-wise manner.

00:53:03 It's converted to an indene carboxylic acid,

00:53:05 base-soluble polymer,

00:53:07 base-soluble product.

00:53:09 You can develop an image.

00:53:11 So you coat the goop

00:53:13 to resist onto the silicon wafer,

00:53:15 expose it, and develop it.

00:53:17 You get a nice image.

00:53:19 All right, well,

00:53:21 you can't use this

00:53:23 if you're going down to UV wavelengths

00:53:25 because of the aromatic materials,

00:53:27 benzene rings, naphthalene, et cetera,

00:53:29 are too absorbent.

00:53:31 If you now try to use this material

00:53:33 at 220 to 250 nanometer exposure,

00:53:35 you're going to get

00:53:37 all the light absorbed

00:53:39 at this resist-air interface.

00:53:41 You're not going to get

00:53:43 the reaction chemistry occurring

00:53:45 through the thickness of the film,

00:53:47 so you're not going to get

00:53:49 this nice, clean profile.

00:53:51 All right, so we were looking

00:53:53 at some early approaches

00:53:55 to short-wavelength resists.

00:53:57 This is work from IBM and Grant Wilson

00:53:59 where he looked at naldrin's acid

00:54:01 as a dissolution inhibitor.

00:54:04 It was nice,

00:54:06 but not appropriate

00:54:08 for a manufacturing line

00:54:10 just that it was too slow.

00:54:12 It involved a single-photon process.

00:54:14 At the same time,

00:54:16 there was interest in EB materials,

00:54:18 and these are some quarter-micron images

00:54:20 that were developed

00:54:22 over a decade ago now

00:54:24 using Inovolac

00:54:26 with a methylpentene sulfone

00:54:28 dissolution inhibitor.

00:54:30 It's amazing that at the time,

00:54:31 getting quarter-micron line

00:54:33 and space images

00:54:35 was a real achievement.

00:54:37 We compare this to where today

00:54:39 we're getting 25 to 45 nanometer images

00:54:44 that actually look better than this.

00:54:47 So that's how far we've gone

00:54:49 from over a decade ago to now.

00:54:52 We've looked at negative E-beam resists

00:54:55 both for mask-making

00:54:57 and for imaging applications.

00:54:59 And we've looked at alternate processes

00:55:02 where we're trying to do dry-developed materials

00:55:05 where we're introducing

00:55:07 a silicon-containing material

00:55:09 that can be dry-developed

00:55:11 where you can use plasma etching

00:55:13 to transfer the pattern

00:55:15 through a thicker organic layer

00:55:17 in order to provide for planarization.

00:55:19 But ultimately,

00:55:21 none of these approaches

00:55:23 really have borne fruit.

00:55:25 So we continue to look at

00:55:26 DPV chemistry.

00:55:28 And what we really focused on then

00:55:31 was the first revolutionary change

00:55:34 in chemistry of resists,

00:55:37 which was the introduction

00:55:39 of the chemical amplification mechanism

00:55:41 by Ito, Wilson, and Frechet

00:55:43 working at IBM,

00:55:45 where what they did

00:55:47 was they took a hydrophobic matrix resin

00:55:49 that had a protective group here.

00:55:51 In the presence of a photogenerated acid,

00:55:53 you could cleave this group,

00:55:54 generating, you know,

00:55:56 isobutylene and carbon dioxide,

00:55:58 which would be evolved from the film,

00:56:00 and you generate a phenol,

00:56:02 aqueous-based insoluble,

00:56:04 aqueous-based soluble.

00:56:06 You have to add an extra process step

00:56:08 of doing a bake,

00:56:10 but a very clean,

00:56:12 effective way to develop images,

00:56:14 and one that was catalytic

00:56:16 because of this acid-generation process

00:56:18 so that you can get to

00:56:20 effectively effective quantum yields

00:56:22 that were greater than one.

00:56:24 So we had a way of now

00:56:26 having a very fast

00:56:28 deep UV resist material.

00:56:30 But there are a lot of problems

00:56:32 related to implementing

00:56:34 chemically amplified resists,

00:56:36 and certainly surface inhibition

00:56:38 and substrate contamination

00:56:40 were key factors,

00:56:42 where there is acid deactivation

00:56:44 at the polymer surface,

00:56:46 at the air interface,

00:56:48 or at the substrate interface

00:56:50 that interfered with

00:56:52 imaging.

00:56:54 And then processing in

00:56:56 base-free environments

00:56:58 or adding a weakly acidic overcoat

00:57:00 were ways to solve that problem.

00:57:02 The original materials

00:57:04 had poor etching resistance.

00:57:06 The protective group

00:57:08 would be removed during etch.

00:57:10 If you look at this material,

00:57:12 about 40% of the volume

00:57:14 of that material

00:57:16 belongs to these groups

00:57:18 that are volatile,

00:57:20 that emanate from the film.

00:57:22 They have poor etching resistance.

00:57:24 So we really had to look at

00:57:26 different chemistries

00:57:28 to solve that problem.

00:57:30 And then there was a large change

00:57:32 in the critical dimension

00:57:34 with post-exposure bake temperature.

00:57:36 Again, we needed to look at

00:57:38 the acid catalysis,

00:57:40 decreasing the catalytic chain

00:57:42 length of the reaction,

00:57:44 and lowering that activation energy.

00:57:46 And the bottom line is

00:57:48 from the original invention

00:57:50 until when we saw insertion

00:57:52 of these materials,

00:57:54 then came a drive

00:57:56 to go to even still

00:57:58 smaller features.

00:58:00 And what technology do we use?

00:58:02 Is it going to be optical?

00:58:04 Is it going to be x-ray?

00:58:06 Is it going to be e-beam?

00:58:08 Well, 193 nanometer

00:58:10 technology research

00:58:12 was initiated

00:58:14 about in the early 1990 timeframe.

00:58:16 Well, that resulted

00:58:18 in sort of a second paradigm

00:58:20 shift in resist materials

00:58:22 where we really wanted

00:58:24 to address the problem

00:58:26 of aromatic and olefinic moieties

00:58:28 that were really too absorptive

00:58:30 at 193 nanometers.

00:58:32 So what we needed to do

00:58:34 was design resist materials

00:58:36 that were structurally different

00:58:38 from the 248 nanometer resist,

00:58:40 structurally different

00:58:42 from NOVOLAC resist,

00:58:44 but functionally superior.

00:58:46 And the way we could do that

00:58:48 was through using

00:58:50 what's called

00:58:52 a high carbon content,

00:58:54 which is using a high cyclic carbon content

00:58:56 to get enhanced

00:58:58 plasma etching performance

00:59:00 coupled with the presence

00:59:02 of structural oxygen.

00:59:04 And from when we started research

00:59:06 to when we had the initial insertion,

00:59:08 we're talking about six years

00:59:10 or somewhat more than that.

00:59:12 But by doing that,

00:59:14 we're able to use

00:59:16 what's called a traditional

00:59:18 single-layer resist design.

00:59:20 And the advantage here

00:59:22 is that we have an existing

00:59:24 manufacturing process knowledge base.

00:59:26 And I think that's very key

00:59:28 to being able to implement

00:59:30 a new technology

00:59:32 because there is

00:59:34 a cost associated

00:59:36 with taking what you know

00:59:38 and starting

00:59:40 from a completely different point.

00:59:42 From the chemistry's perspective,

00:59:44 though,

00:59:45 it required a paradigm shift

00:59:47 in the materials chemistry

00:59:49 in order to effect

00:59:50 high resistance

00:59:52 and not losing anything

00:59:54 in lithographic performance.

00:59:56 And we'd come a long way

00:59:58 in what we needed to do

01:00:00 and what we understood

01:00:02 about the chemistry of the materials

01:00:04 and how we go about

01:00:06 designing materials

01:00:08 so that for each lithographic parameter

01:00:10 that we could identify,

01:00:12 we now knew enough

01:00:14 about the chemistry

01:00:16 and the fundamental

01:00:17 interaction of materials

01:00:18 and characteristics

01:00:20 associated with that

01:00:22 lithographic parameter.

01:00:24 And just to chew the couple,

01:00:26 you know, absorption,

01:00:28 clearly we didn't want

01:00:30 any olefinic or aromatic moieties.

01:00:32 If we're thinking about

01:00:34 low metal ion content,

01:00:36 we wanted to identify

01:00:38 synthesis and scale-up

01:00:40 methodology of the resists

01:00:42 that avoided the use of metals

01:00:44 so that we didn't want

01:00:46 any metals left

01:00:47 in the material.

01:00:49 So in terms of materials design,

01:00:51 one of the approaches

01:00:53 that we took at the labs

01:00:55 was to start with a matrix resin

01:00:57 that had L-cyclic moieties

01:00:59 for etching resistance.

01:01:01 This is sort of a

01:01:02 norbornian component.

01:01:04 We added maleic anhydrides

01:01:06 that allowed us to do

01:01:08 a metal ion pre-synthesis.

01:01:10 And then we added

01:01:12 acrylate functionalities

01:01:14 for differential solubility.

01:01:15 So this is a solubility

01:01:17 derivative that occupies

01:01:19 a very large molecular volume.

01:01:21 And if we convert this

01:01:23 colic acid from

01:01:25 aqueous base insoluble

01:01:27 to aqueous base soluble,

01:01:29 we'd have a very large

01:01:31 differential solubility.

01:01:33 It's miscible with polar

01:01:35 matrix resins.

01:01:37 It's transparent at 193

01:01:39 nanometers, and it's readily

01:01:41 available.

01:01:42 Colic acid is a bile acid.

01:01:43 And then we had the

01:01:44 photoacid generator,

01:01:45 where for the initial

01:01:47 formulations we used

01:01:48 a diphenyliodonium

01:01:49 nonaflate.

01:01:50 Certainly the fluorinated

01:01:52 materials are something

01:01:53 we want to get away from

01:01:54 now because of environmental

01:01:55 issues.

01:01:57 But it was miscible with

01:01:58 the resist components.

01:01:59 It afforded a strong acid,

01:02:01 and it generated

01:02:02 nonvolatile byproducts

01:02:04 upon exposure.

01:02:06 So that chemistry

01:02:08 allowed us to have

01:02:09 good transparency

01:02:10 at the exposing wavelength.

01:02:11 It was compatible

01:02:12 with the standard

01:02:14 developer used in a

01:02:15 process line, 0.262

01:02:17 normal tetramethyl

01:02:18 ammonium hydroxide.

01:02:20 Had a high Tg.

01:02:22 It was compatible with

01:02:23 using the solution

01:02:24 inhibitor approaches.

01:02:25 It had good adhesion

01:02:27 and good etching

01:02:28 performance.

01:02:29 This shows

01:02:30 the chemistry for the

01:02:32 chemists in the audience

01:02:33 where we've got the

01:02:34 matrix resin that has

01:02:36 a t-butyl ester

01:02:37 protective group.

01:02:38 Again, a similarly

01:02:39 protected t-butyl ester

01:02:41 collate derivative

01:02:43 and the photo acid

01:02:44 generator chemistry.

01:02:46 We go from a material

01:02:47 that's aqueous based

01:02:48 insoluble to one

01:02:50 that is aqueous based

01:02:51 soluble.

01:02:53 These were some early

01:02:54 images that we were able

01:02:55 to obtain on 193

01:02:57 exposure showing

01:02:58 100 nanometer

01:02:59 or tenth micron

01:03:00 line in space images.

01:03:03 We were also able to use

01:03:04 phase shift mass

01:03:05 techniques to generate

01:03:07 nanometer scale images

01:03:08 where these are

01:03:09 60 nanometer

01:03:10 images patterned

01:03:12 using an alternating

01:03:13 phase shift mass.

01:03:15 Down here is

01:03:17 the first flash memory

01:03:18 device, an 80 nanometer

01:03:19 device that was

01:03:21 prepared using

01:03:22 193 nanometer

01:03:24 lithography.

01:03:26 What I have in the

01:03:27 inset here is

01:03:29 a comparison of

01:03:30 what the very early

01:03:32 chemically amplified

01:03:33 resist images

01:03:34 looked like.

01:03:36 Because of the

01:03:37 acid that's

01:03:38 generated,

01:03:39 a means in

01:03:41 an exposure environment,

01:03:43 we got neutralization

01:03:44 of acid at this

01:03:45 resist air interface

01:03:47 and all we could get

01:03:48 were tunnels.

01:03:50 Unfortunately,

01:03:51 we could generate

01:03:52 really great tunnels

01:03:54 but nobody had

01:03:55 any use for them

01:03:56 at least

01:03:57 up until now.

01:04:00 A lot of effort went

01:04:01 into understanding

01:04:02 this process

01:04:03 and then really being

01:04:04 able to

01:04:05 evolve

01:04:06 and generate

01:04:07 nanometer scale

01:04:08 devices

01:04:10 where these were

01:04:11 80 nanometers

01:04:12 and smaller.

01:04:14 What about

01:04:15 the future?

01:04:17 We're using

01:04:18 193 nanometer

01:04:19 lithography today.

01:04:21 The options for the

01:04:22 perceivable future

01:04:23 that are being

01:04:24 discussed

01:04:25 are immersion

01:04:26 lithography,

01:04:27 UV,

01:04:28 157,

01:04:29 along with others.

01:04:31 What is going to be

01:04:32 key to making

01:04:33 those technologies

01:04:34 feasible

01:04:35 are

01:04:37 the materials

01:04:38 and the processes

01:04:39 that are associated

01:04:40 with

01:04:41 these technologies.

01:04:43 In terms of

01:04:44 invention to insertion,

01:04:45 when we're going

01:04:46 to see that happen

01:04:47 really is going to

01:04:48 depend on

01:04:49 the need

01:04:50 and how much

01:04:51 of a change

01:04:52 we're going to be

01:04:53 seeing

01:04:54 in the fundamental

01:04:55 technology that's

01:04:56 going to be used

01:04:57 to fabricate

01:04:58 silicon devices.

01:04:59 We really want to

01:05:00 build on

01:05:01 what we know.

01:05:03 That has led to

01:05:04 a lot of interest

01:05:05 in what is called

01:05:06 immersion lithography,

01:05:08 where we're now

01:05:09 using projection optics,

01:05:10 focusing that beam

01:05:12 through a liquid,

01:05:13 which is water,

01:05:15 developing resists

01:05:16 that now work

01:05:17 in this

01:05:18 water environment.

01:05:20 That provides for

01:05:22 an improvement

01:05:23 in the resolution

01:05:24 that we can get

01:05:25 at

01:05:26 the wafer plane.

01:05:29 Some examples of that,

01:05:30 these are 45

01:05:31 nanometer lines

01:05:33 that were imaged

01:05:34 with

01:05:35 193 nanometer

01:05:36 light

01:05:37 using immersion

01:05:38 lithography,

01:05:39 and this is

01:05:40 courtesy of

01:05:41 Nikon and Clarion.

01:05:43 This is demonstration

01:05:45 that by using

01:05:46 this technique,

01:05:47 we can

01:05:48 fundamentally stay

01:05:49 with the same

01:05:50 resist technology

01:05:52 and be able to get

01:05:53 images that are now

01:05:54 a quarter

01:05:55 of the wavelength

01:05:56 of light

01:05:57 that is being used

01:05:58 to expose

01:05:59 those materials.

01:06:01 If we look at

01:06:02 the wavelength,

01:06:03 whether it's

01:06:04 193 nanometer

01:06:06 or 157 nanometer

01:06:07 light

01:06:08 with fluorine

01:06:09 lasers,

01:06:10 by using

01:06:11 immersion lithography,

01:06:13 the effective

01:06:14 wavelength is down

01:06:15 to either 134

01:06:16 nanometers

01:06:17 or 115

01:06:18 nanometers.

01:06:20 By using

01:06:21 argon fluoride

01:06:22 immersion,

01:06:23 we're effectively

01:06:24 doing 134

01:06:25 nanometer

01:06:26 lithography.

01:06:28 What are some

01:06:29 future options?

01:06:30 What are we going

01:06:31 to do when

01:06:32 those images

01:06:33 get to be

01:06:34 one or two

01:06:35 or three or

01:06:36 four nanometers

01:06:37 across?

01:06:38 How are we going

01:06:39 to be able to

01:06:40 fabricate those

01:06:41 devices?

01:06:42 Well, there are a

01:06:43 number of options

01:06:44 that are being

01:06:45 talked about.

01:06:46 One of them is

01:06:47 nano imprint

01:06:48 lithography,

01:06:49 which largely

01:06:50 involves

01:06:51 hot embossing

01:06:52 or stamping

01:06:53 onto a substrate.

01:06:54 It can be done

01:06:55 die by die

01:06:56 or by full

01:06:57 wafer patterning,

01:06:58 but there are

01:06:59 some advantages

01:07:00 to that

01:07:01 process.

01:07:02 Certainly, it's a

01:07:03 high resolution

01:07:04 process,

01:07:05 but you've got

01:07:06 the disadvantages

01:07:07 effectively of

01:07:08 contact lithography,

01:07:09 namely,

01:07:10 defects

01:07:11 and distortions.

01:07:12 And at least

01:07:13 to date,

01:07:14 there's no clear

01:07:15 way to be able

01:07:16 to align

01:07:17 one layer

01:07:18 to the next.

01:07:19 This is just

01:07:20 an example

01:07:21 of what can

01:07:22 be done

01:07:23 with

01:07:24 step-and-flash

01:07:25 imprint

01:07:26 lithography,

01:07:27 where these

01:07:28 are 40

01:07:29 nanometers

01:07:30 across

01:07:31 and

01:07:32 contained

01:07:33 with this

01:07:34 kind of

01:07:35 process.

01:07:36 Similarly,

01:07:37 there have

01:07:38 been six

01:07:39 nanometer lines

01:07:40 here,

01:07:41 six nanometers

01:07:42 across

01:07:43 and five

01:07:44 nanometers

01:07:45 across,

01:07:46 done by

01:07:47 nano imprint

01:07:48 lithography,

01:07:49 again,

01:07:50 by pressing

01:07:51 a template

01:07:52 into a

01:07:53 thermoplastic

01:07:54 material.

01:07:55 But you're

01:07:56 going to have

01:07:57 problems

01:07:58 with

01:07:59 this

01:08:00 because

01:08:01 you're

01:08:02 not

01:08:03 going to

01:08:04 be able

01:08:05 to

01:08:06 do

01:08:07 the

01:08:08 same

01:08:09 thing

01:08:10 over and

01:08:11 over

01:08:12 again.

01:08:13 And so

01:08:14 we're

01:08:15 coming to

01:08:16 a lot of

01:08:17 discussion

01:08:18 about what

01:08:19 is going

01:08:20 to be

01:08:21 the next

01:08:22 lithographic

01:08:23 technology

01:08:24 and when

01:08:25 is it

01:08:26 going to

01:08:27 be

01:08:28 the next

01:08:29 lithographic

01:08:30 technology.

01:08:31 And

01:08:32 we're

01:08:33 going to

01:08:34 have

01:08:35 a lot

01:08:36 of

01:08:37 discussion

01:08:38 about

01:08:39 that.

01:08:40 And

01:08:41 we're

01:08:42 going to

01:08:43 have

01:08:44 a lot

01:08:45 of

01:08:46 discussion

01:08:47 about

01:08:48 that.

01:08:49 And

01:08:50 we're

01:08:51 going to

01:08:52 have

01:08:53 a lot

01:08:54 of

01:08:55 discussion

01:08:56 about

01:08:57 what

01:08:58 is going

01:08:59 to be

01:09:00 the next

01:09:01 lithographic

01:09:02 technology.

01:09:03 And

01:09:04 we're

01:09:05 going to

01:09:06 have

01:09:07 a lot

01:09:08 of

01:09:09 discussion

01:09:10 about

01:09:11 that.

01:09:12 And

01:09:13 we're

01:09:14 going to

01:09:15 have

01:09:16 a lot

01:09:17 of

01:09:18 discussion

01:09:19 about

01:09:20 what

01:09:21 is going

01:09:22 to be

01:09:23 the next

01:09:24 lithographic

01:09:25 technology.

01:09:26 And

01:09:27 we're

01:09:28 going to

01:09:29 have a

01:09:30 lot

01:09:31 of

01:09:32 discussion

01:09:33 about

01:09:34 what

01:09:35 is going

01:09:36 to be

01:09:37 the next

01:09:38 lithographic

01:09:39 technology.

01:09:40 And

01:09:41 we're

01:09:42 going to

01:09:43 have a

01:09:44 lot

01:09:45 of

01:09:46 discussion

01:09:47 about

01:09:48 what

01:09:49 is going

01:09:50 to be

01:09:51 the next

01:09:52 lithographic

01:09:53 technology.

01:09:54 And

01:09:55 we're

01:09:56 going to

01:09:57 have a

01:09:58 lot

01:09:59 of

01:10:00 discussion

01:10:01 about

01:10:02 what

01:10:03 is going

01:10:04 to be

01:10:05 the next

01:10:06 lithographic

01:10:07 technology.

01:10:08 And

01:10:09 we're

01:10:10 going to

01:10:11 have a

01:10:12 lot

01:10:13 of

01:10:14 discussion

01:10:15 about

01:10:16 what

01:10:17 is going

01:10:18 to be

01:10:19 the next

01:10:20 lithographic

01:10:21 technology.

01:10:23 And

01:10:24 we're

01:10:25 going to

01:10:26 have a

01:10:27 lot

01:10:28 of

01:10:29 discussion

01:10:30 about

01:10:31 what

01:10:32 is going

01:10:33 to be

01:10:34 the next

01:10:35 lithographic

01:10:36 technology.

01:10:37 And

01:10:38 we're

01:10:39 going to

01:10:40 have a

01:10:41 lot

01:10:42 of

01:10:43 discussion

01:10:44 about

01:10:45 what

01:10:46 is going

01:10:47 to be

01:10:48 the next

01:10:49 lithographic

01:10:50 technology.

01:10:51 And

01:10:52 we're

01:10:53 going to

01:10:54 have a

01:10:55 lot

01:10:56 of

01:10:57 discussion

01:10:58 about

01:10:59 what

01:11:00 is going

01:11:01 to be

01:11:02 the next

01:11:03 lithographic

01:11:04 technology.

01:11:05 And

01:11:06 we're

01:11:07 going to

01:11:08 have a

01:11:09 lot

01:11:10 of

01:11:11 discussion

01:11:12 about

01:11:13 what

01:11:14 is going

01:11:15 to be

01:11:16 the next

01:11:17 lithographic

01:11:18 technology.

01:11:19 And

01:11:20 we're

01:11:21 going to

01:11:22 have a

01:11:23 lot

01:11:24 of

01:11:25 discussion

01:11:26 about

01:11:27 what

01:11:28 is going

01:11:29 to be

01:11:30 the next

01:11:31 lithographic

01:11:32 technology.

01:11:33 And

01:11:34 we're

01:11:35 going to

01:11:36 have a

01:11:37 lot

01:11:38 of

01:11:39 discussion

01:11:40 about

01:11:41 what

01:11:42 is going

01:11:43 to be

01:11:44 the next

01:11:45 lithographic

01:11:46 technology.

01:11:47 And

01:11:48 we're

01:11:49 going to

01:11:50 have a

01:11:51 lot

01:11:52 of

01:11:53 discussion

01:11:54 about

01:11:55 what

01:11:56 is going

01:11:57 to be

01:11:58 the next

01:11:59 lithographic

01:12:00 technology.

01:12:01 And

01:12:02 we're

01:12:03 going to

01:12:04 have a

01:12:05 lot

01:12:06 of

01:12:07 discussion

01:12:08 about

01:12:09 what

01:12:10 is going

01:12:11 to be

01:12:12 the next

01:12:13 lithographic

01:12:14 technology.

01:12:15 And

01:12:16 we're

01:12:17 going to

01:12:18 have a

01:12:19 lot

01:12:20 of

01:12:21 discussion

01:12:22 about

01:12:23 what

01:12:24 is going

01:12:25 to be

01:12:26 the next

01:12:27 lithographic

01:12:28 technology.

01:12:29 And

01:12:30 we're

01:12:31 going to

01:12:32 have a

01:12:33 lot

01:12:34 of

01:12:35 discussion

01:12:36 about

01:12:37 what

01:12:38 is going

01:12:39 to be

01:12:40 the next

01:12:41 lithographic

01:12:42 technology.

01:12:43 And

01:12:44 we're

01:12:45 going to

01:12:46 have a

01:12:47 lot

01:12:48 of

01:12:49 discussion

01:12:50 about

01:12:51 what

01:12:52 is going

01:12:53 to be

01:12:54 the next

01:12:55 lithographic

01:12:56 technology.

01:12:57 And

01:12:58 we're

01:12:59 going to

01:13:00 have a

01:13:01 lot

01:13:02 of

01:13:03 discussion

01:13:04 about

01:13:05 what

01:13:06 is going

01:13:07 to be

01:13:08 the next

01:13:09 lithographic

01:13:10 technology.

01:13:11 And

01:13:12 we're

01:13:13 going to

01:13:14 have a

01:13:15 lot

01:13:16 of

01:13:17 discussion

01:13:18 about

01:13:19 what

01:13:20 is going

01:13:21 to be

01:13:22 the next

01:13:23 lithographic

01:13:24 technology.

01:13:25 And

01:13:26 we're

01:13:27 going to

01:13:28 have a

01:13:29 lot

01:13:30 of

01:13:31 discussion

01:13:32 about

01:13:33 what

01:13:34 is going

01:13:35 to be

01:13:36 the next

01:13:37 lithographic

01:13:38 technology.

01:13:39 And

01:13:40 we're

01:13:41 going to

01:13:42 have a

01:13:43 lot

01:13:44 of

01:13:45 discussion

01:13:46 about

01:13:47 what

01:13:48 is going

01:13:49 to be

01:13:50 the next

01:13:51 lithographic

01:13:52 technology.

01:13:53 And

01:13:54 we're

01:13:55 going to

01:13:56 have a

01:13:57 lot

01:13:58 of

01:13:59 discussion

01:14:00 about

01:14:01 what

01:14:02 is going

01:14:03 to be

01:14:04 the next

01:14:05 lithographic

01:14:06 technology.

01:14:07 And

01:14:08 we're

01:14:09 going to

01:14:10 have a

01:14:11 lot

01:14:12 of

01:14:13 discussion

01:14:14 about

01:14:15 what

01:14:16 is going

01:14:17 to be

01:14:18 the next

01:14:19 lithographic

01:14:20 technology.

01:14:21 And

01:14:22 we're

01:14:23 going to

01:14:24 have a

01:14:25 lot

01:14:26 of

01:14:27 discussion

01:14:28 about

01:14:29 what

01:14:30 is going

01:14:31 to be

01:14:32 the next

01:14:33 lithographic

01:14:34 technology.

01:14:35 And

01:14:36 we're

01:14:37 going to

01:14:38 have a

01:14:39 lot

01:14:40 of

01:14:41 discussion

01:14:42 about

01:14:43 what

01:14:44 is going

01:14:45 to be

01:14:46 the next

01:14:47 lithographic

01:14:48 technology.

01:14:49 And

01:14:50 we're

01:14:51 going to

01:14:52 have a

01:14:53 lot

01:14:54 of

01:14:55 discussion

01:14:56 about

01:14:57 what

01:14:58 is going

01:14:59 to be

01:15:00 the next

01:15:01 lithographic

01:15:02 technology.

01:15:03 And

01:15:04 we're

01:15:05 going to

01:15:06 have a

01:15:07 lot

01:15:08 of

01:15:09 discussion

01:15:10 about

01:15:11 what

01:15:12 is going

01:15:13 to be

01:15:14 the next

01:15:15 lithographic

01:15:16 technology.

01:15:17 And

01:15:18 we're

01:15:19 going to

01:15:20 have a

01:15:21 lot

01:15:22 of

01:15:23 discussion

01:15:24 about

01:15:25 what

01:15:26 is going

01:15:27 to be

01:15:28 the next

01:15:29 lithographic

01:15:30 technology.

01:15:31 And

01:15:32 we're

01:15:33 going to

01:15:34 have a

01:15:35 lot

01:15:36 of

01:15:37 discussion

01:15:38 about

01:15:39 what

01:15:40 is going

01:15:41 to be

01:15:42 the next

01:15:43 lithographic

01:15:44 technology.

01:15:45 And

01:15:46 we're

01:15:47 going to

01:15:48 have a

01:15:49 lot

01:15:50 of

01:15:51 discussion

01:15:52 about

01:15:53 what is

01:15:54 going

01:15:55 to be

01:15:56 the next

01:15:57 lithographic

01:15:58 technology.

01:15:59 And

01:16:00 we're

01:16:01 going to

01:16:02 have a

01:16:03 lot

01:16:04 of

01:16:05 discussion

01:16:06 about

01:16:07 what is

01:16:08 going

01:16:09 to be

01:16:10 the next

01:16:11 lithographic

01:16:12 technology.

01:16:13 And

01:16:14 we're

01:16:15 going to

01:16:16 have a

01:16:17 lot

01:16:18 of

01:16:19 discussion

01:16:20 about

01:16:21 what is

01:16:22 going

01:16:23 to be

01:16:24 the next

01:16:25 lithographic

01:16:26 technology.

01:16:27 And

01:16:28 we're

01:16:29 going to

01:16:30 have a

01:16:31 lot

01:16:32 of

01:16:33 discussion

01:16:34 about

01:16:35 what is

01:16:36 going

01:16:37 to be

01:16:38 the next