WEBVTT 00:00:00.000 --> 00:00:01.973 align:middle line:90% [SQUEAKING] 00:00:01.973 --> 00:00:03.937 align:middle line:90% [RUSTLING] 00:00:03.937 --> 00:00:05.901 align:middle line:90% [CLICKING] 00:00:05.901 --> 00:00:16.540 align:middle line:90% 00:00:16.540 --> 00:00:20.220 align:middle line:84% PROFESSOR: OK, so let's actually talk about optical fibers 00:00:20.220 --> 00:00:21.600 align:middle line:90% and waveguides. 00:00:21.600 --> 00:00:25.278 align:middle line:84% And again, we have already mentioned 00:00:25.278 --> 00:00:27.820 align:middle line:84% about fiber, why they are so important, why we care about it. 00:00:27.820 --> 00:00:31.320 align:middle line:84% What I'm quoting here, this is actually from the citation 00:00:31.320 --> 00:00:35.183 align:middle line:84% for the 2009 Nobel Prize in physics awarded to Charles Kao, 00:00:35.183 --> 00:00:37.600 align:middle line:84% which basically says, "If we were to unravel all the glass 00:00:37.600 --> 00:00:39.380 align:middle line:84% fibers that went around the globe, 00:00:39.380 --> 00:00:42.700 align:middle line:84% we would get a single thread over 1 billion kilometers long, 00:00:42.700 --> 00:00:46.620 align:middle line:84% which is enough to encircle the globe more than 25,000 times, 00:00:46.620 --> 00:00:49.480 align:middle line:84% and it's increasing by thousands of kilometers every hour." 00:00:49.480 --> 00:00:50.920 align:middle line:90% This is 2009. 00:00:50.920 --> 00:00:53.992 align:middle line:84% We get a lot more optical fibers, obviously, 00:00:53.992 --> 00:00:56.700 align:middle line:84% because driven by the increasing internet traffic and, of course, 00:00:56.700 --> 00:00:58.980 align:middle line:84% recently by AI, by the emergence of AI. 00:00:58.980 --> 00:01:03.020 align:middle line:84% So needless to say, this is very, very important. 00:01:03.020 --> 00:01:05.900 align:middle line:84% So how does the light propagate in fibers? 00:01:05.900 --> 00:01:09.020 align:middle line:84% You are probably aware that essentially light 00:01:09.020 --> 00:01:11.520 align:middle line:84% gets trapped in a fiber through a total internal reflection. 00:01:11.520 --> 00:01:13.800 align:middle line:90% You have two materials. 00:01:13.800 --> 00:01:16.300 align:middle line:84% One is the high index that we call the core. 00:01:16.300 --> 00:01:17.980 align:middle line:84% Surrounding it is the low index material 00:01:17.980 --> 00:01:19.500 align:middle line:90% that we call the cladding. 00:01:19.500 --> 00:01:23.407 align:middle line:84% So then when the light essentially stays in the core 00:01:23.407 --> 00:01:25.740 align:middle line:84% and they get reflected at the interface between the core 00:01:25.740 --> 00:01:29.220 align:middle line:84% and the cladding, if the total internal reflection condition 00:01:29.220 --> 00:01:32.220 align:middle line:84% is satisfied, means that the angle theta 00:01:32.220 --> 00:01:36.500 align:middle line:84% greater than some critical angle, then 100% of light 00:01:36.500 --> 00:01:38.520 align:middle line:84% will get reflected at every bounce. 00:01:38.520 --> 00:01:40.660 align:middle line:84% So you get a very low loss propagation 00:01:40.660 --> 00:01:43.260 align:middle line:90% through this mechanism here. 00:01:43.260 --> 00:01:45.700 align:middle line:84% I can actually do a really quick, simple demo here 00:01:45.700 --> 00:01:49.620 align:middle line:84% to show you how optical fiber works. 00:01:49.620 --> 00:01:53.740 align:middle line:90% I have my little laser pointer. 00:01:53.740 --> 00:01:55.620 align:middle line:90% So light goes in straight lines. 00:01:55.620 --> 00:01:57.930 align:middle line:84% And what I have here, this is just 00:01:57.930 --> 00:02:05.650 align:middle line:84% a little plastic pipe, tube, whatever, just 00:02:05.650 --> 00:02:07.170 align:middle line:90% raw, actually, plastic raw. 00:02:07.170 --> 00:02:17.170 align:middle line:84% So if I just insert it in here and then I light it up, 00:02:17.170 --> 00:02:18.750 align:middle line:90% we will see is-- 00:02:18.750 --> 00:02:22.645 align:middle line:90% 00:02:22.645 --> 00:02:24.990 align:middle line:84% I don't know if you can see the laser spot coming up. 00:02:24.990 --> 00:02:26.450 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:02:26.450 --> 00:02:27.095 align:middle line:90% 00:02:27.095 --> 00:02:28.970 align:middle line:84% PROFESSOR: You can see in the beginning here, 00:02:28.970 --> 00:02:31.770 align:middle line:84% the light is kind of getting in there. 00:02:31.770 --> 00:02:34.230 align:middle line:90% But then do you see the spot? 00:02:34.230 --> 00:02:34.750 align:middle line:90% All good? 00:02:34.750 --> 00:02:35.430 align:middle line:90% OK, good. 00:02:35.430 --> 00:02:38.130 align:middle line:84% So you can clearly see the laser spot coming out here. 00:02:38.130 --> 00:02:40.290 align:middle line:84% And you can clearly see how the light actually 00:02:40.290 --> 00:02:43.370 align:middle line:84% propagates and attenuates as it goes along this little pipeline 00:02:43.370 --> 00:02:43.930 align:middle line:90% here. 00:02:43.930 --> 00:02:46.890 align:middle line:84% So this is just a simple rule of total internal reflection 00:02:46.890 --> 00:02:49.650 align:middle line:84% at the plastic and air interface. 00:02:49.650 --> 00:02:53.450 align:middle line:84% You can consider this as a very large core kind of optical fiber 00:02:53.450 --> 00:02:56.230 align:middle line:90% or light guide. 00:02:56.230 --> 00:02:57.190 align:middle line:90% All right. 00:02:57.190 --> 00:03:02.950 align:middle line:84% So next, I'm going to talk a little bit about the terminology 00:03:02.950 --> 00:03:06.990 align:middle line:84% we're going to use here, how to describe waveguides in general. 00:03:06.990 --> 00:03:09.312 align:middle line:84% Fibers can be considered as a waveguide. 00:03:09.312 --> 00:03:11.270 align:middle line:84% Waveguide is just a device that can essentially 00:03:11.270 --> 00:03:12.870 align:middle line:84% trap light and allow light propagate 00:03:12.870 --> 00:03:14.370 align:middle line:84% through a total internal reflection. 00:03:14.370 --> 00:03:17.110 align:middle line:84% So if you have a cylindrical cross-section, 00:03:17.110 --> 00:03:18.670 align:middle line:90% these are the fibers we see. 00:03:18.670 --> 00:03:21.070 align:middle line:84% There are two common kinds of fibers people use. 00:03:21.070 --> 00:03:23.025 align:middle line:90% One is called step-index fiber. 00:03:23.025 --> 00:03:25.150 align:middle line:84% Really very simple, you just have core and cladding 00:03:25.150 --> 00:03:27.400 align:middle line:84% made of two materials with different indices, 00:03:27.400 --> 00:03:29.650 align:middle line:84% and you have a well-defined interface in between them. 00:03:29.650 --> 00:03:33.910 align:middle line:84% You also have so-called graded-index fiber, 00:03:33.910 --> 00:03:37.070 align:middle line:84% where you have a gradual change of refractive index 00:03:37.070 --> 00:03:38.770 align:middle line:90% from the core to the cladding. 00:03:38.770 --> 00:03:40.630 align:middle line:90% You have a diffused interface. 00:03:40.630 --> 00:03:42.990 align:middle line:84% So there's no well-defined separation 00:03:42.990 --> 00:03:44.210 align:middle line:90% between the two materials. 00:03:44.210 --> 00:03:48.030 align:middle line:84% You can similarly trap light inside here 00:03:48.030 --> 00:03:50.630 align:middle line:90% with a similar mechanism. 00:03:50.630 --> 00:03:53.350 align:middle line:84% They are useful for various applications, 00:03:53.350 --> 00:03:55.570 align:middle line:84% such as, for example, engineering, the dispersion 00:03:55.570 --> 00:04:00.050 align:middle line:84% behavior of light propagating in the fibers. 00:04:00.050 --> 00:04:03.850 align:middle line:84% Now, turning into what we call the planar waveguides, 00:04:03.850 --> 00:04:08.210 align:middle line:84% these are essentially basically photonic wires 00:04:08.210 --> 00:04:11.110 align:middle line:84% that can trap photons, but not in along the fiber, 00:04:11.110 --> 00:04:14.890 align:middle line:84% but usually on a chip, small silicon chip. 00:04:14.890 --> 00:04:16.810 align:middle line:84% And they are fabricated essentially 00:04:16.810 --> 00:04:18.892 align:middle line:84% using a standard semiconductor microfabrication 00:04:18.892 --> 00:04:21.350 align:middle line:84% techniques, much like how you actually make computer chips. 00:04:21.350 --> 00:04:24.490 align:middle line:84% So it usually involves deposition 00:04:24.490 --> 00:04:26.330 align:middle line:84% of multilayer thin films, and then you 00:04:26.330 --> 00:04:28.970 align:middle line:84% do photolithography to then define 00:04:28.970 --> 00:04:31.170 align:middle line:90% those waveguide structures. 00:04:31.170 --> 00:04:34.090 align:middle line:84% So there are three different geometries for these waveguides 00:04:34.090 --> 00:04:34.690 align:middle line:90% here. 00:04:34.690 --> 00:04:37.850 align:middle line:84% The first one, this is called a slab waveguide. 00:04:37.850 --> 00:04:39.890 align:middle line:84% It's basically nothing more than a layer 00:04:39.890 --> 00:04:43.450 align:middle line:84% of high-index film sandwiched between two low-index cladding 00:04:43.450 --> 00:04:43.990 align:middle line:90% layers. 00:04:43.990 --> 00:04:46.870 align:middle line:90% 00:04:46.870 --> 00:04:50.170 align:middle line:84% In this configuration in, this slab waveguide configuration, 00:04:50.170 --> 00:04:52.130 align:middle line:84% light is essentially only confined 00:04:52.130 --> 00:04:55.222 align:middle line:90% in this out-of-plane direction. 00:04:55.222 --> 00:04:56.680 align:middle line:84% In the in-plane direction, you have 00:04:56.680 --> 00:04:59.600 align:middle line:84% perfect translational symmetry, so light is actually 00:04:59.600 --> 00:05:02.880 align:middle line:84% going to spread laterally as a propagating plane. 00:05:02.880 --> 00:05:06.460 align:middle line:84% So that makes slab waveguide actually not very useful. 00:05:06.460 --> 00:05:08.200 align:middle line:84% Most of the time people don't use them 00:05:08.200 --> 00:05:10.400 align:middle line:84% because you don't have any control over where 00:05:10.400 --> 00:05:12.040 align:middle line:90% light propagates in plane. 00:05:12.040 --> 00:05:16.520 align:middle line:84% So to enable this kind of two-dimensional confinement 00:05:16.520 --> 00:05:18.880 align:middle line:84% also in plane as well, you can either 00:05:18.880 --> 00:05:20.818 align:middle line:84% try to pattern this high-index film, 00:05:20.818 --> 00:05:23.360 align:middle line:84% so you get essentially this kind of rectangular cross-section 00:05:23.360 --> 00:05:24.343 align:middle line:90% waveguides. 00:05:24.343 --> 00:05:25.760 align:middle line:84% This is what's called a, depending 00:05:25.760 --> 00:05:27.680 align:middle line:84% on who you talk to, strip waveguide, 00:05:27.680 --> 00:05:30.860 align:middle line:84% channel waveguide, photonic wire waveguide, whatever. 00:05:30.860 --> 00:05:32.400 align:middle line:84% But essentially, you can trap light 00:05:32.400 --> 00:05:33.880 align:middle line:84% inside this [INAUDIBLE] core here 00:05:33.880 --> 00:05:36.360 align:middle line:90% via total internal reflection. 00:05:36.360 --> 00:05:39.340 align:middle line:84% Alternatively, you can also, instead of doing a full etch, 00:05:39.340 --> 00:05:41.240 align:middle line:84% you do a partial etch of the film. 00:05:41.240 --> 00:05:43.600 align:middle line:84% You create a height difference here. 00:05:43.600 --> 00:05:45.640 align:middle line:84% Similarly, you can also trap light, 00:05:45.640 --> 00:05:48.747 align:middle line:84% essentially in the middle here, because effectively what's 00:05:48.747 --> 00:05:51.080 align:middle line:84% on the left side here and what's on the right side here, 00:05:51.080 --> 00:05:53.455 align:middle line:84% you can think of it as you have less high-index material. 00:05:53.455 --> 00:05:56.840 align:middle line:84% So effectively, you have lower average refractive index. 00:05:56.840 --> 00:06:00.120 align:middle line:84% So light tend to get trapped right in the middle here. 00:06:00.120 --> 00:06:04.017 align:middle line:84% So this is called rib or ridge waveguide. 00:06:04.017 --> 00:06:06.100 align:middle line:84% So those are actually the waveguide people we use, 00:06:06.100 --> 00:06:07.000 align:middle line:90% channel or ridge. 00:06:07.000 --> 00:06:08.780 align:middle line:84% These are the waveguide people use to actually draw light, 00:06:08.780 --> 00:06:11.405 align:middle line:84% for example, on a silicon chip, which we're going to talk about 00:06:11.405 --> 00:06:12.682 align:middle line:90% in the next lecture. 00:06:12.682 --> 00:06:14.140 align:middle line:84% For this lecture, we're going to be 00:06:14.140 --> 00:06:17.940 align:middle line:84% focusing on this cylindrical waveguide or optical fibers. 00:06:17.940 --> 00:06:22.020 align:middle line:84% So in general, there are two kinds of optical fibers 00:06:22.020 --> 00:06:25.460 align:middle line:84% here, called multimode and single mode. 00:06:25.460 --> 00:06:29.900 align:middle line:84% So the concept of mode, it refers to a specific way 00:06:29.900 --> 00:06:33.300 align:middle line:84% that light propagates inside a fiber or inside waveguide. 00:06:33.300 --> 00:06:37.140 align:middle line:84% The easiest way to understand the mode 00:06:37.140 --> 00:06:42.900 align:middle line:84% is you can think of it as it's a horizontal specific angle where 00:06:42.900 --> 00:06:45.120 align:middle line:84% the light is propagating in a waveguide or fiber. 00:06:45.120 --> 00:06:50.880 align:middle line:84% So for example, consider this is a core for fiber. 00:06:50.880 --> 00:06:57.160 align:middle line:84% And we look at this angle, this angle theta. 00:06:57.160 --> 00:07:00.400 align:middle line:84% Each mode corresponds to one particular angle 00:07:00.400 --> 00:07:02.760 align:middle line:84% where the light actually propagates inside this fiber 00:07:02.760 --> 00:07:03.880 align:middle line:90% here. 00:07:03.880 --> 00:07:08.520 align:middle line:84% What we argue is that this angle is actually discretized, 00:07:08.520 --> 00:07:10.600 align:middle line:84% if you have a waveguide with a finite thickness 00:07:10.600 --> 00:07:11.600 align:middle line:90% or finite diameter. 00:07:11.600 --> 00:07:15.960 align:middle line:90% 00:07:15.960 --> 00:07:17.800 align:middle line:84% There are a few ways to understand it. 00:07:17.800 --> 00:07:21.360 align:middle line:84% If you take my class 346, I actually give a quantum 00:07:21.360 --> 00:07:22.920 align:middle line:90% mechanical analogy. 00:07:22.920 --> 00:07:27.160 align:middle line:84% Turns out, mathematically, the Maxwell equation 00:07:27.160 --> 00:07:29.640 align:middle line:84% that governs light propagation in this kind of media 00:07:29.640 --> 00:07:33.360 align:middle line:84% is exactly similar to a Schrodinger equation. 00:07:33.360 --> 00:07:35.760 align:middle line:84% And you can essentially see this kind of waveguide 00:07:35.760 --> 00:07:38.972 align:middle line:84% as a particle in a potential well. 00:07:38.972 --> 00:07:40.680 align:middle line:84% So you have particle in a potential well. 00:07:40.680 --> 00:07:42.597 align:middle line:84% The energy levels actually becomes quantized. 00:07:42.597 --> 00:07:44.180 align:middle line:84% You have a few discrete energy levels. 00:07:44.180 --> 00:07:46.600 align:middle line:84% Energy cannot access arbitrary values. 00:07:46.600 --> 00:07:50.750 align:middle line:84% Similarly, this angle cannot access arbitrary values in that 00:07:50.750 --> 00:07:51.870 align:middle line:90% regard. 00:07:51.870 --> 00:07:53.770 align:middle line:84% So that's from wave optics perspective. 00:07:53.770 --> 00:07:54.270 align:middle line:90% Now, 00:07:54.270 --> 00:07:56.470 align:middle line:84% More simply, you can also think of there's also 00:07:56.470 --> 00:07:57.950 align:middle line:90% a simple explanation. 00:07:57.950 --> 00:07:59.390 align:middle line:84% You can think of light propagation 00:07:59.390 --> 00:08:03.030 align:middle line:84% in this kind of a fiber as the superposition of two plane 00:08:03.030 --> 00:08:05.367 align:middle line:90% waves in a waveguide. 00:08:05.367 --> 00:08:06.950 align:middle line:84% So you have essentially one plane wave 00:08:06.950 --> 00:08:08.830 align:middle line:90% that's going in this direction. 00:08:08.830 --> 00:08:12.990 align:middle line:84% You have another plane wave that's going in this direction. 00:08:12.990 --> 00:08:16.673 align:middle line:84% So now, essentially, if you look at these two light rays here, 00:08:16.673 --> 00:08:18.590 align:middle line:84% they both belong essentially to the same plane 00:08:18.590 --> 00:08:21.350 align:middle line:84% wave that propagate along the same direction. 00:08:21.350 --> 00:08:24.610 align:middle line:84% So in order for light to propagate over long distances, 00:08:24.610 --> 00:08:27.390 align:middle line:84% it's important that all of these light rays that 00:08:27.390 --> 00:08:32.030 align:middle line:84% goes in one direction, they have to interfere constructively. 00:08:32.030 --> 00:08:35.350 align:middle line:84% So that condition specifies that only at specific angles 00:08:35.350 --> 00:08:38.130 align:middle line:84% you can fulfill the constructive interference condition, 00:08:38.130 --> 00:08:40.970 align:middle line:84% and therefore this angle has to be discretized. 00:08:40.970 --> 00:08:42.470 align:middle line:90% It cannot be arbitrary values. 00:08:42.470 --> 00:08:49.450 align:middle line:90% 00:08:49.450 --> 00:08:51.030 align:middle line:84% And similar to quantum mechanics, 00:08:51.030 --> 00:08:52.530 align:middle line:84% if you recall, in quantum mechanics, 00:08:52.530 --> 00:08:56.650 align:middle line:84% if you have a potential well, if the potential well gets smaller 00:08:56.650 --> 00:09:02.970 align:middle line:84% and it gets shallower, then you have a less confined state. 00:09:02.970 --> 00:09:03.810 align:middle line:90% Same thing here. 00:09:03.810 --> 00:09:07.130 align:middle line:84% For waveguides or for fibers, either way, 00:09:07.130 --> 00:09:10.450 align:middle line:84% if you make the core smaller, you also 00:09:10.450 --> 00:09:12.930 align:middle line:90% supports less number of modes. 00:09:12.930 --> 00:09:14.650 align:middle line:84% Means there's less number of angles, 00:09:14.650 --> 00:09:18.570 align:middle line:84% options that light can propagate inside a small core fiber. 00:09:18.570 --> 00:09:21.750 align:middle line:84% Ultimately, if the core size gets really small, 00:09:21.750 --> 00:09:23.930 align:middle line:84% once it's below some kind of threshold value, 00:09:23.930 --> 00:09:26.770 align:middle line:84% you have the so-called single-mode fiber, 00:09:26.770 --> 00:09:30.270 align:middle line:84% where light can only propagate along one specific angle, 00:09:30.270 --> 00:09:35.870 align:middle line:84% not in any other angles, versus multimode fiber, 00:09:35.870 --> 00:09:39.050 align:middle line:84% light can propagate along multiple different angles there. 00:09:39.050 --> 00:09:42.330 align:middle line:90% So why is that important? 00:09:42.330 --> 00:09:44.840 align:middle line:84% Because for fiber, obviously, one big application 00:09:44.840 --> 00:09:46.120 align:middle line:90% is for data communications. 00:09:46.120 --> 00:09:48.800 align:middle line:84% We want to be able to send data over long distances. 00:09:48.800 --> 00:09:50.780 align:middle line:84% If you think about different orders of modes, 00:09:50.780 --> 00:09:52.660 align:middle line:84% just by looking at that simple graph on top 00:09:52.660 --> 00:09:54.640 align:middle line:84% there, different orders of modes, 00:09:54.640 --> 00:09:56.440 align:middle line:84% because they travel at different angles, 00:09:56.440 --> 00:09:59.920 align:middle line:90% the path length is different. 00:09:59.920 --> 00:10:03.840 align:middle line:84% So if you're trying to send signal over long distances 00:10:03.840 --> 00:10:05.360 align:middle line:84% and you have a multimode fiber, it 00:10:05.360 --> 00:10:07.000 align:middle line:84% means that your signal is essentially 00:10:07.000 --> 00:10:09.280 align:middle line:84% getting distributed in these different modes. 00:10:09.280 --> 00:10:12.520 align:middle line:84% At the other end, the receiver will actually see your signal 00:10:12.520 --> 00:10:15.240 align:middle line:90% coming at different times. 00:10:15.240 --> 00:10:17.760 align:middle line:84% This will be problematic because the simple way 00:10:17.760 --> 00:10:21.265 align:middle line:84% to transmit a signal is you have to essentially modulate 00:10:21.265 --> 00:10:23.140 align:middle line:84% the light to encode some digital information. 00:10:23.140 --> 00:10:25.300 align:middle line:84% Simple way, for example, you can turn light on. 00:10:25.300 --> 00:10:26.380 align:middle line:90% It's a digital 1. 00:10:26.380 --> 00:10:28.200 align:middle line:84% You turn light off, it's digital 0. 00:10:28.200 --> 00:10:30.080 align:middle line:84% You do this very, very, very, very 00:10:30.080 --> 00:10:33.643 align:middle line:84% quickly, so that you can do high speed data communication. 00:10:33.643 --> 00:10:35.560 align:middle line:84% But you can imagine that if the light comes in 00:10:35.560 --> 00:10:37.600 align:middle line:84% at a different time, it will completely 00:10:37.600 --> 00:10:41.680 align:middle line:84% smear out this kind of on/off tempo for the data stream. 00:10:41.680 --> 00:10:46.080 align:middle line:84% So that's why, for long-distance communication, 00:10:46.080 --> 00:10:47.560 align:middle line:84% they call long-haul communications, 00:10:47.560 --> 00:10:50.960 align:middle line:84% people tend to use a single mode instead of a multimode. 00:10:50.960 --> 00:10:56.840 align:middle line:90% 00:10:56.840 --> 00:10:58.940 align:middle line:84% Now, you may wonder, why do people still 00:10:58.940 --> 00:11:00.380 align:middle line:90% make multimode waveguides? 00:11:00.380 --> 00:11:02.300 align:middle line:84% So this really depends on the application. 00:11:02.300 --> 00:11:04.660 align:middle line:84% So even in communications, if you're not 00:11:04.660 --> 00:11:07.225 align:middle line:84% trying to send a signal over very long distances, 00:11:07.225 --> 00:11:08.600 align:middle line:84% not like kilometers, for example, 00:11:08.600 --> 00:11:10.700 align:middle line:84% you only need to send it for a few meters, 00:11:10.700 --> 00:11:13.460 align:middle line:84% like between, for example, essentially 00:11:13.460 --> 00:11:18.780 align:middle line:84% the modem at your house to maybe high-definition digital TV, 00:11:18.780 --> 00:11:20.280 align:middle line:90% that's a pretty short distance. 00:11:20.280 --> 00:11:23.220 align:middle line:84% In that case, the velocity difference 00:11:23.220 --> 00:11:26.700 align:middle line:84% between the different modes is actually negligible. 00:11:26.700 --> 00:11:29.157 align:middle line:84% And having this kind of multimode fiber 00:11:29.157 --> 00:11:30.740 align:middle line:84% will be actually a lot easier in terms 00:11:30.740 --> 00:11:34.900 align:middle line:84% of assembly and integration because the core is actually 00:11:34.900 --> 00:11:36.380 align:middle line:90% much larger. 00:11:36.380 --> 00:11:38.320 align:middle line:84% So in terms of alignment and everything, 00:11:38.320 --> 00:11:42.335 align:middle line:84% assembly is all straightforward compared to doing a single mode. 00:11:42.335 --> 00:11:43.460 align:middle line:90% So there's one application. 00:11:43.460 --> 00:11:44.835 align:middle line:84% It's a lot easier to couple light 00:11:44.835 --> 00:11:47.040 align:middle line:84% into a multimode fiber compared to a single mode. 00:11:47.040 --> 00:11:48.957 align:middle line:84% The other application where people like to use 00:11:48.957 --> 00:11:52.560 align:middle line:84% multimode if they are doing something for high power, 00:11:52.560 --> 00:11:54.640 align:middle line:84% like they are trying to do laser welding, 00:11:54.640 --> 00:11:57.360 align:middle line:84% or sometimes they do some laser surgery 00:11:57.360 --> 00:12:00.080 align:middle line:84% or infrared countermeasures, whatever. 00:12:00.080 --> 00:12:05.080 align:middle line:84% In those cases, multimode fiber offer you much higher power 00:12:05.080 --> 00:12:06.780 align:middle line:84% handling capability for obvious reasons, 00:12:06.780 --> 00:12:08.280 align:middle line:84% because you get a lot more material. 00:12:08.280 --> 00:12:10.113 align:middle line:84% Power density is actually lower, so any kind 00:12:10.113 --> 00:12:11.800 align:middle line:84% of nonlinear effect or damage effect 00:12:11.800 --> 00:12:13.360 align:middle line:90% is going to be a lot less. 00:12:13.360 --> 00:12:16.000 align:middle line:84% So for those reasons, that's why people still 00:12:16.000 --> 00:12:17.720 align:middle line:84% use a lot of these multimode fibers, 00:12:17.720 --> 00:12:19.100 align:middle line:90% for those kind of applications. 00:12:19.100 --> 00:12:22.360 align:middle line:90% 00:12:22.360 --> 00:12:22.980 align:middle line:90% All right. 00:12:22.980 --> 00:12:25.600 align:middle line:84% So the main difference between multimode and single mode 00:12:25.600 --> 00:12:27.925 align:middle line:84% practically is really just the core size. 00:12:27.925 --> 00:12:29.800 align:middle line:84% If you made the core size sufficiently small, 00:12:29.800 --> 00:12:31.383 align:middle line:84% you will support only one single mode. 00:12:31.383 --> 00:12:32.840 align:middle line:90% You get single-mode fiber. 00:12:32.840 --> 00:12:37.600 align:middle line:84% For the standard fiber people use for data communications, 00:12:37.600 --> 00:12:40.150 align:middle line:84% this is the standard configuration here. 00:12:40.150 --> 00:12:43.030 align:middle line:84% It has a core diameter of about 9 microns, 00:12:43.030 --> 00:12:45.627 align:middle line:84% and it has a cladding diameter of 125 microns. 00:12:45.627 --> 00:12:47.210 align:middle line:84% These are all international standards. 00:12:47.210 --> 00:12:49.230 align:middle line:84% This is the same kind of fiber that people use. 00:12:49.230 --> 00:12:50.610 align:middle line:84% And it's all made of-- basically, 00:12:50.610 --> 00:12:52.110 align:middle line:90% this is silica glass fiber. 00:12:52.110 --> 00:12:57.610 align:middle line:90% 00:12:57.610 --> 00:13:00.190 align:middle line:84% So to make such fibers, you would 00:13:00.190 --> 00:13:03.630 align:middle line:84% need to essentially have two glass materials, two silica 00:13:03.630 --> 00:13:06.625 align:middle line:84% variants with slightly different refractive indices. 00:13:06.625 --> 00:13:08.750 align:middle line:84% So what people will do is then they can essentially 00:13:08.750 --> 00:13:11.750 align:middle line:84% dope different kind of elements into the fiber 00:13:11.750 --> 00:13:14.250 align:middle line:84% to essentially get the core and cladding in this difference. 00:13:14.250 --> 00:13:17.990 align:middle line:84% So for silica, which is the common material that people 00:13:17.990 --> 00:13:22.030 align:middle line:84% use for essentially all kind of data communication applications, 00:13:22.030 --> 00:13:24.510 align:middle line:84% you can either try to increase the refractive index 00:13:24.510 --> 00:13:27.370 align:middle line:84% by adding heavier elements with a higher probability, 00:13:27.370 --> 00:13:28.890 align:middle line:90% so they increase index. 00:13:28.890 --> 00:13:31.990 align:middle line:84% You can add germanium oxide, titanium oxide, phosphorus 00:13:31.990 --> 00:13:32.830 align:middle line:90% oxide. 00:13:32.830 --> 00:13:34.550 align:middle line:84% You can increase index of the core. 00:13:34.550 --> 00:13:38.930 align:middle line:84% Or you can also choose to decrease index of cladding. 00:13:38.930 --> 00:13:41.410 align:middle line:84% In this case, you essentially add small atoms 00:13:41.410 --> 00:13:43.950 align:middle line:84% with small polarizability, like boron, 00:13:43.950 --> 00:13:46.527 align:middle line:84% which is much smaller compared to silicon. 00:13:46.527 --> 00:13:48.610 align:middle line:84% So there are two different ways to actually create 00:13:48.610 --> 00:13:50.530 align:middle line:90% index contrast. 00:13:50.530 --> 00:13:54.330 align:middle line:84% You can also add other kind of functional elements, 00:13:54.330 --> 00:13:57.350 align:middle line:84% such as erbium, which we discussed about in last lecture. 00:13:57.350 --> 00:13:59.730 align:middle line:84% They are important for erbium-doped optical fiber 00:13:59.730 --> 00:14:01.127 align:middle line:90% amplifiers. 00:14:01.127 --> 00:14:03.210 align:middle line:84% You can also add other kind of rare earth elements 00:14:03.210 --> 00:14:05.730 align:middle line:84% to provide amplification at different wavebands 00:14:05.730 --> 00:14:07.830 align:middle line:84% across your communication bandwidth. 00:14:07.830 --> 00:14:12.358 align:middle line:84% So silica is the most commonly used material for making fibers, 00:14:12.358 --> 00:14:14.650 align:middle line:84% but I want to mention that you can also make fibers out 00:14:14.650 --> 00:14:17.330 align:middle line:84% of essentially any kind of glass materials. 00:14:17.330 --> 00:14:18.950 align:middle line:84% You can make out of chalcogenides, 00:14:18.950 --> 00:14:20.890 align:middle line:90% halides, polymers. 00:14:20.890 --> 00:14:22.690 align:middle line:90% People do that all the time. 00:14:22.690 --> 00:14:25.150 align:middle line:84% More recently, in the past couple of decades, 00:14:25.150 --> 00:14:28.810 align:middle line:84% people have also developed so-called multi-material fibers. 00:14:28.810 --> 00:14:31.310 align:middle line:84% Means that you not only incorporate one single material, 00:14:31.310 --> 00:14:32.922 align:middle line:84% you can also incorporate materials 00:14:32.922 --> 00:14:35.630 align:middle line:84% that conventionally people think are difficult to make fibers out 00:14:35.630 --> 00:14:38.750 align:middle line:84% of, such as the crystalline semiconductors, 00:14:38.750 --> 00:14:41.693 align:middle line:90% crystalline oxides, metals. 00:14:41.693 --> 00:14:43.110 align:middle line:84% In fact, people, for example, they 00:14:43.110 --> 00:14:47.267 align:middle line:84% can even make fibers out of crystalline silicon. 00:14:47.267 --> 00:14:49.350 align:middle line:84% They are still a little more tricky to make fibers 00:14:49.350 --> 00:14:51.092 align:middle line:84% out of compared to glass materials, 00:14:51.092 --> 00:14:52.550 align:middle line:84% but that's a possibility if you get 00:14:52.550 --> 00:14:54.570 align:middle line:84% all the thermomechanical conditions correctly. 00:14:54.570 --> 00:14:58.753 align:middle line:90% 00:14:58.753 --> 00:15:00.670 align:middle line:84% Now, before I talk about manufacturing process 00:15:00.670 --> 00:15:02.470 align:middle line:84% and why glass is preferred, I just 00:15:02.470 --> 00:15:04.590 align:middle line:84% want to quickly mention here that this 00:15:04.590 --> 00:15:07.270 align:middle line:84% is the standard configuration, if you look at the silica glass 00:15:07.270 --> 00:15:08.310 align:middle line:90% fiber. 00:15:08.310 --> 00:15:10.830 align:middle line:84% As I've said, you have a core with a diameter 00:15:10.830 --> 00:15:12.750 align:middle line:90% of about 9 microns. 00:15:12.750 --> 00:15:15.770 align:middle line:84% You have a cladding with a diameter of 125 microns. 00:15:15.770 --> 00:15:17.750 align:middle line:84% Both are made of silica glass with some kind 00:15:17.750 --> 00:15:18.890 align:middle line:90% of different dopants. 00:15:18.890 --> 00:15:24.430 align:middle line:84% Then outside, people usually put a polymer protective layers. 00:15:24.430 --> 00:15:27.670 align:middle line:84% That's called a buffer and the jacket. 00:15:27.670 --> 00:15:30.670 align:middle line:84% So the reason they need to put polymer around it essentially 00:15:30.670 --> 00:15:32.550 align:middle line:84% is to protect the glass against all kinds 00:15:32.550 --> 00:15:34.222 align:middle line:90% of environmental damages. 00:15:34.222 --> 00:15:36.180 align:middle line:84% This can come from some kind of optical damage, 00:15:36.180 --> 00:15:37.740 align:middle line:90% like UV radiation. 00:15:37.740 --> 00:15:39.860 align:middle line:84% So if you have exposed glass core, 00:15:39.860 --> 00:15:41.460 align:middle line:84% ultraviolet radiation can actually 00:15:41.460 --> 00:15:42.935 align:middle line:84% get absorbed in the glass material, 00:15:42.935 --> 00:15:45.060 align:middle line:84% and that can actually create, essentially, damages, 00:15:45.060 --> 00:15:47.900 align:middle line:84% like color centers, which increases optical loss 00:15:47.900 --> 00:15:48.920 align:middle line:90% in the glass material. 00:15:48.920 --> 00:15:50.780 align:middle line:84% So they want to absorb all those UV 00:15:50.780 --> 00:15:52.300 align:middle line:90% light in the polymer cladding. 00:15:52.300 --> 00:15:55.220 align:middle line:84% Also, we have learned that glasses, 00:15:55.220 --> 00:15:57.580 align:middle line:84% if you have any kind of surface defects, that 00:15:57.580 --> 00:15:58.640 align:middle line:90% tend to be a weak point. 00:15:58.640 --> 00:16:00.780 align:middle line:84% They eventually lead to a catastrophic failure. 00:16:00.780 --> 00:16:03.820 align:middle line:84% So having this polymer protective coatings 00:16:03.820 --> 00:16:06.540 align:middle line:84% also prevent the formation of any kind of surface damage 00:16:06.540 --> 00:16:10.100 align:middle line:84% that compromise the mechanical robustness of optical fibers. 00:16:10.100 --> 00:16:13.160 align:middle line:84% So that's why this is the fiber you will normally get, 00:16:13.160 --> 00:16:13.660 align:middle line:90% if you see. 00:16:13.660 --> 00:16:16.100 align:middle line:84% If you have fiber, like fiber to the home, 00:16:16.100 --> 00:16:20.060 align:middle line:84% this is kind of fiber cable you would normally see. 00:16:20.060 --> 00:16:23.780 align:middle line:84% Now, comes back to the fiber manufacturing process. 00:16:23.780 --> 00:16:25.300 align:middle line:84% I have alluded to the fact that it's 00:16:25.300 --> 00:16:27.980 align:middle line:84% a lot easier to make fibers out of glass 00:16:27.980 --> 00:16:30.020 align:middle line:84% than out of crystalline material, basically 00:16:30.020 --> 00:16:33.348 align:middle line:84% because of the glass materials, as we have known, 00:16:33.348 --> 00:16:35.640 align:middle line:84% they have this kind of continuous viscosity temperature 00:16:35.640 --> 00:16:36.480 align:middle line:90% behavior. 00:16:36.480 --> 00:16:38.360 align:middle line:84% By simply changing temperature, you 00:16:38.360 --> 00:16:40.580 align:middle line:84% can precisely control the viscosity 00:16:40.580 --> 00:16:43.060 align:middle line:84% so that during a fiber formation process, 00:16:43.060 --> 00:16:45.480 align:middle line:84% you can essentially get sufficient deformation 00:16:45.480 --> 00:16:48.000 align:middle line:84% while also ensure that it doesn't break up easily. 00:16:48.000 --> 00:16:52.120 align:middle line:84% So specifically, the way to actually fabricate fiber 00:16:52.120 --> 00:16:54.260 align:middle line:84% is so-called using a fiber drawing process. 00:16:54.260 --> 00:16:57.360 align:middle line:84% So what you will do, you start from essentially 00:16:57.360 --> 00:16:59.160 align:middle line:90% what's called a preform. 00:16:59.160 --> 00:17:04.780 align:middle line:84% A preform is basically a glass rod with varying diameter, 00:17:04.780 --> 00:17:06.540 align:middle line:84% sometimes as small as millimeters. 00:17:06.540 --> 00:17:09.680 align:middle line:90% Big ones can be centimeters. 00:17:09.680 --> 00:17:16.720 align:middle line:84% The key construction for preform is its cross-sectional structure 00:17:16.720 --> 00:17:19.579 align:middle line:84% has to exactly map to what you want in the end. 00:17:19.579 --> 00:17:23.839 align:middle line:84% So for example, if you're trying to do a simple step-index fiber, 00:17:23.839 --> 00:17:26.640 align:middle line:84% you have core and cladding, then what you do, 00:17:26.640 --> 00:17:30.800 align:middle line:84% you essentially have a preform with a diameter ratio of core 00:17:30.800 --> 00:17:33.060 align:middle line:84% and cladding that's exactly the same as what you 00:17:33.060 --> 00:17:34.440 align:middle line:90% want to get in the final fiber. 00:17:34.440 --> 00:17:39.220 align:middle line:84% But this is a much bigger version, a scaled version. 00:17:39.220 --> 00:17:41.440 align:middle line:84% So during the fiber drawing process, 00:17:41.440 --> 00:17:44.660 align:middle line:84% essentially you transform this macroscopic cross-section 00:17:44.660 --> 00:17:46.940 align:middle line:84% into something that's a lot smaller, 00:17:46.940 --> 00:17:51.630 align:middle line:84% but still maintain exactly the size ratio between this core 00:17:51.630 --> 00:17:53.100 align:middle line:90% and cladding part. 00:17:53.100 --> 00:17:56.420 align:middle line:84% And you do this not just for this kind of very simple 00:17:56.420 --> 00:17:57.480 align:middle line:90% cross-section here. 00:17:57.480 --> 00:18:00.020 align:middle line:84% Later on, we going to talk about so-called microstructured 00:18:00.020 --> 00:18:01.437 align:middle line:84% optical fiber, where you can have, 00:18:01.437 --> 00:18:03.270 align:middle line:84% for example, you can have not just one core, 00:18:03.270 --> 00:18:04.620 align:middle line:90% you can have multiple cores. 00:18:04.620 --> 00:18:06.440 align:middle line:84% You can even have some air holes. 00:18:06.440 --> 00:18:09.240 align:middle line:84% You can have some additional materials distributed around it. 00:18:09.240 --> 00:18:13.020 align:middle line:84% Whatever you have in the macroscopic preform, when 00:18:13.020 --> 00:18:16.660 align:middle line:84% you draw the fiber, it all gets scaled down proportionally 00:18:16.660 --> 00:18:22.820 align:middle line:84% into this thin strand of, say, 100 micron thick fiber there. 00:18:22.820 --> 00:18:24.925 align:middle line:84% So this is actually quite remarkable. 00:18:24.925 --> 00:18:26.300 align:middle line:84% It's one way to essentially allow 00:18:26.300 --> 00:18:29.170 align:middle line:84% you to use any kind of ingenious techniques 00:18:29.170 --> 00:18:32.950 align:middle line:84% to create a very complicated cross-sectional geometry 00:18:32.950 --> 00:18:34.980 align:middle line:84% and then transfer it into the micro domain 00:18:34.980 --> 00:18:36.730 align:middle line:84% to actually get the fiber structure there. 00:18:36.730 --> 00:18:38.980 align:middle line:84% We're going to show some examples of how people can do 00:18:38.980 --> 00:18:40.830 align:middle line:90% that in a very clever manner. 00:18:40.830 --> 00:18:45.250 align:middle line:84% But now, what you do then you have the preform is a glass, 00:18:45.250 --> 00:18:47.730 align:middle line:84% you then lower it into the furnace. 00:18:47.730 --> 00:18:50.130 align:middle line:84% In the furnace, gets heat up, essentially 00:18:50.130 --> 00:18:51.570 align:middle line:90% decrease the viscosity. 00:18:51.570 --> 00:18:53.690 align:middle line:84% You want to control the viscosity so they're 00:18:53.690 --> 00:18:56.300 align:middle line:84% forced within this window of 10 to the four to 10 00:18:56.300 --> 00:18:59.030 align:middle line:84% to the six parts per second for the reason I mentioned there. 00:18:59.030 --> 00:19:01.210 align:middle line:84% Number one, you want the viscosity sufficiently low 00:19:01.210 --> 00:19:02.668 align:middle line:90% so you can do the joint process. 00:19:02.668 --> 00:19:04.210 align:middle line:84% You also don't want it to be too low, 00:19:04.210 --> 00:19:05.793 align:middle line:84% otherwise they break up into droplets. 00:19:05.793 --> 00:19:11.010 align:middle line:90% 00:19:11.010 --> 00:19:13.470 align:middle line:84% So controlling this viscosity is very important here. 00:19:13.470 --> 00:19:15.178 align:middle line:84% You want to make this is how you actually 00:19:15.178 --> 00:19:20.130 align:middle line:84% ensure you have a reproducible and scalable process. 00:19:20.130 --> 00:19:22.970 align:middle line:84% So after that, after you exit from the furnace, 00:19:22.970 --> 00:19:25.150 align:middle line:84% you then go through essentially a diameter gauge. 00:19:25.150 --> 00:19:27.590 align:middle line:84% This is usually essentially measuring the fiber diameter 00:19:27.590 --> 00:19:29.950 align:middle line:84% using a laser, using a laser beam. 00:19:29.950 --> 00:19:33.150 align:middle line:84% And then after that you go through a coater, 00:19:33.150 --> 00:19:35.430 align:middle line:84% where essentially the polymer jackets 00:19:35.430 --> 00:19:38.150 align:middle line:84% and the protective coatings, they essentially 00:19:38.150 --> 00:19:42.310 align:middle line:84% are coated onto fiber, initially in a liquid monomer form. 00:19:42.310 --> 00:19:46.230 align:middle line:84% This liquid monomer then goes through this UV curing lamp, 00:19:46.230 --> 00:19:49.470 align:middle line:84% which induces cross-linking of the monomers 00:19:49.470 --> 00:19:52.710 align:middle line:84% to form the polymer jacket or the polymer coatings. 00:19:52.710 --> 00:19:55.710 align:middle line:84% Finally, the fiber then goes into this 00:19:55.710 --> 00:19:57.870 align:middle line:84% winding drum so that you get essentially 00:19:57.870 --> 00:19:59.310 align:middle line:90% a large spools of fiber. 00:19:59.310 --> 00:20:00.450 align:middle line:90% That's very, very low. 00:20:00.450 --> 00:20:03.618 align:middle line:90% 00:20:03.618 --> 00:20:05.910 align:middle line:84% So I want to mention that this manufacturing process is 00:20:05.910 --> 00:20:08.270 align:middle line:90% actually very, very efficient. 00:20:08.270 --> 00:20:10.950 align:middle line:84% Typical drawing speed of fibers is of the order 00:20:10.950 --> 00:20:13.790 align:middle line:90% of several meters per second. 00:20:13.790 --> 00:20:17.110 align:middle line:84% So within a day, you can get tens, hundreds of kilometers 00:20:17.110 --> 00:20:18.350 align:middle line:90% of optical fiber there. 00:20:18.350 --> 00:20:20.170 align:middle line:84% So it's a very, very efficient process. 00:20:20.170 --> 00:20:23.670 align:middle line:84% In fact, by the way, there's actually a fiber drawing tower 00:20:23.670 --> 00:20:24.470 align:middle line:90% on campus. 00:20:24.470 --> 00:20:25.947 align:middle line:90% Do you know that? 00:20:25.947 --> 00:20:27.530 align:middle line:84% It's actually in the building starting 00:20:27.530 --> 00:20:30.368 align:middle line:84% in the basement from Professor Yoel Fink's group. 00:20:30.368 --> 00:20:32.410 align:middle line:84% So if you ever want to see a fiber drawing tower, 00:20:32.410 --> 00:20:35.490 align:middle line:90% that's what they have there. 00:20:35.490 --> 00:20:41.610 align:middle line:84% So it's a pretty impressive two-story high structure. 00:20:41.610 --> 00:20:43.890 align:middle line:84% Lots of fun to look at how they actually make fibers. 00:20:43.890 --> 00:20:44.390 align:middle line:90% All right. 00:20:44.390 --> 00:20:46.870 align:middle line:84% So now, come back about the glass lines here. 00:20:46.870 --> 00:20:50.890 align:middle line:84% So what glass materials are easier to make fibers? 00:20:50.890 --> 00:20:54.050 align:middle line:84% This has to do with fragility, same as any kind 00:20:54.050 --> 00:20:57.750 align:middle line:84% of thermal plastic-based kind of glass processing. 00:20:57.750 --> 00:21:03.130 align:middle line:84% So a few things that actually impact the ease of fiber 00:21:03.130 --> 00:21:03.907 align:middle line:90% processing here. 00:21:03.907 --> 00:21:06.490 align:middle line:84% I have quoted two things here, liquid fragility and also glass 00:21:06.490 --> 00:21:07.530 align:middle line:90% forming ability. 00:21:07.530 --> 00:21:11.250 align:middle line:84% For glass forming ability, it's straightforward to understand. 00:21:11.250 --> 00:21:13.830 align:middle line:84% When you actually are trying to draw glass fibers, 00:21:13.830 --> 00:21:16.050 align:middle line:84% the viscosity window is much lower than 10 00:21:16.050 --> 00:21:18.170 align:middle line:90% to the 12 pascal second. 00:21:18.170 --> 00:21:21.170 align:middle line:84% 10 to 12 pascal second is basically Tg. 00:21:21.170 --> 00:21:23.680 align:middle line:84% So it means that fiber drawing should take place 00:21:23.680 --> 00:21:27.320 align:middle line:84% at a temperature much higher than Tg. 00:21:27.320 --> 00:21:30.480 align:middle line:84% So if it's much higher than Tg, glass materials 00:21:30.480 --> 00:21:33.200 align:middle line:90% will risk crystallization. 00:21:33.200 --> 00:21:35.040 align:middle line:84% The general figure of merit we use 00:21:35.040 --> 00:21:38.840 align:middle line:84% to gauge the glass stability or its ability 00:21:38.840 --> 00:21:41.960 align:middle line:84% against crystallization is this delta T here, 00:21:41.960 --> 00:21:43.700 align:middle line:84% where T is crystallization temperature. 00:21:43.700 --> 00:21:45.800 align:middle line:84% Tg is glass transition temperature. 00:21:45.800 --> 00:21:47.813 align:middle line:84% So you have a large window large delta T, 00:21:47.813 --> 00:21:49.480 align:middle line:84% it means that the glass material is more 00:21:49.480 --> 00:21:51.440 align:middle line:90% resistant to crystallization. 00:21:51.440 --> 00:21:54.560 align:middle line:84% So you want a glass that does not crystallize, 00:21:54.560 --> 00:21:56.560 align:middle line:84% because if you crystallize, you then 00:21:56.560 --> 00:21:58.920 align:middle line:84% introducing all this crystalline precipitate. 00:21:58.920 --> 00:22:01.018 align:middle line:84% They are going to scatter your light like crazy. 00:22:01.018 --> 00:22:02.560 align:middle line:84% You're not going to get a good fiber. 00:22:02.560 --> 00:22:06.300 align:middle line:84% It has to be amorphous, normally. 00:22:06.300 --> 00:22:09.270 align:middle line:90% 00:22:09.270 --> 00:22:12.960 align:middle line:84% So that's why materials like silica, which never crystallize 00:22:12.960 --> 00:22:15.520 align:middle line:84% this, is really a lot more straightforward to join 00:22:15.520 --> 00:22:18.440 align:middle line:84% to fibers compared to other glasses, especially 00:22:18.440 --> 00:22:20.560 align:middle line:84% things like halite that crystallize 00:22:20.560 --> 00:22:22.892 align:middle line:84% a lot easier compared to silica glasses. 00:22:22.892 --> 00:22:25.100 align:middle line:84% And that's one of the reasons I mentioned previously. 00:22:25.100 --> 00:22:26.500 align:middle line:84% Despite the halides, people predict 00:22:26.500 --> 00:22:28.250 align:middle line:84% that they may have lower loss than silica. 00:22:28.250 --> 00:22:30.680 align:middle line:84% People still stick to silica glasses, 00:22:30.680 --> 00:22:32.260 align:middle line:84% because of the ease of manufacturing. 00:22:32.260 --> 00:22:36.605 align:middle line:84% Second thing is liquid fragility. 00:22:36.605 --> 00:22:38.980 align:middle line:84% We talked about this earlier when we talk about viscosity 00:22:38.980 --> 00:22:40.085 align:middle line:90% of glass materials. 00:22:40.085 --> 00:22:41.460 align:middle line:84% You have this fragility parameter 00:22:41.460 --> 00:22:45.060 align:middle line:84% that gives you essentially the slope of the line at the angle 00:22:45.060 --> 00:22:45.900 align:middle line:90% plot here. 00:22:45.900 --> 00:22:48.880 align:middle line:84% So if you have a steep line, you have a large slope, 00:22:48.880 --> 00:22:50.058 align:middle line:90% you have a large fragility. 00:22:50.058 --> 00:22:52.600 align:middle line:84% Then it essentially means that if you look at this part here, 00:22:52.600 --> 00:22:54.433 align:middle line:84% this is where the fiber drawing takes place, 00:22:54.433 --> 00:22:55.920 align:middle line:90% in this viscosity window. 00:22:55.920 --> 00:23:00.780 align:middle line:84% So this large slope translates to a much higher temperature 00:23:00.780 --> 00:23:01.660 align:middle line:90% sensitivity. 00:23:01.660 --> 00:23:03.460 align:middle line:84% Any small variation of temperature 00:23:03.460 --> 00:23:06.460 align:middle line:84% leads to a very large change of viscosity there, 00:23:06.460 --> 00:23:08.280 align:middle line:90% and that's not preferable. 00:23:08.280 --> 00:23:10.080 align:middle line:90% You want a stable process. 00:23:10.080 --> 00:23:15.380 align:middle line:84% So that's why glass materials are strong in the glass forming 00:23:15.380 --> 00:23:17.780 align:middle line:84% liquid phase with silica, again why 00:23:17.780 --> 00:23:19.940 align:middle line:84% it's preferable for drawing optical fibers. 00:23:19.940 --> 00:23:25.160 align:middle line:90% 00:23:25.160 --> 00:23:25.700 align:middle line:90% All right. 00:23:25.700 --> 00:23:28.880 align:middle line:84% Next, what I want to do is then we 00:23:28.880 --> 00:23:32.360 align:middle line:84% can talk a little bit about the fiber drawing process 00:23:32.360 --> 00:23:33.760 align:middle line:90% specific parameters. 00:23:33.760 --> 00:23:38.600 align:middle line:84% So what I'm showing here on this plot, 00:23:38.600 --> 00:23:42.340 align:middle line:84% vertical axis is basically the fiber drawing temperature, 00:23:42.340 --> 00:23:44.360 align:middle line:90% this temperature in the furnace. 00:23:44.360 --> 00:23:47.240 align:middle line:84% Horizontal axis is the fiber drawing speed. 00:23:47.240 --> 00:23:52.000 align:middle line:84% I have two plots here where each of this line on the top panel 00:23:52.000 --> 00:23:54.440 align:middle line:84% correspond to one constant fiber radius. 00:23:54.440 --> 00:23:57.560 align:middle line:84% Bottom, each line correspond to constant drawing tension, 00:23:57.560 --> 00:24:02.280 align:middle line:84% the tension you apply along the fiber in the axial direction. 00:24:02.280 --> 00:24:04.840 align:middle line:84% If you look at these two parts here, 00:24:04.840 --> 00:24:09.840 align:middle line:84% this on the top panel, these lines are more or less vertical. 00:24:09.840 --> 00:24:14.360 align:middle line:84% So what it says is that if you want to control a fiber radius, 00:24:14.360 --> 00:24:18.430 align:middle line:84% it is a lot more effective to actually change the drawing 00:24:18.430 --> 00:24:18.970 align:middle line:90% speed. 00:24:18.970 --> 00:24:21.637 align:middle line:84% If you change the drawing speed, you can change the fiber radius 00:24:21.637 --> 00:24:22.630 align:middle line:90% very effectively. 00:24:22.630 --> 00:24:26.870 align:middle line:84% It's a lot less sensitive to the fiber drawing temperature. 00:24:26.870 --> 00:24:30.190 align:middle line:84% So that's why, if you look at this kind of fiber drawing 00:24:30.190 --> 00:24:33.170 align:middle line:84% process here, to control the fiber diameter, 00:24:33.170 --> 00:24:34.810 align:middle line:84% essentially you change the speed. 00:24:34.810 --> 00:24:37.830 align:middle line:84% This winding drum is actually rotating. 00:24:37.830 --> 00:24:41.187 align:middle line:84% And usually people do of closed feedback loop control. 00:24:41.187 --> 00:24:42.770 align:middle line:84% They have this kind of diameter gauge. 00:24:42.770 --> 00:24:44.370 align:middle line:84% They measure the fiber diameter in real time, 00:24:44.370 --> 00:24:46.245 align:middle line:84% and then they use it to essentially determine 00:24:46.245 --> 00:24:49.230 align:middle line:84% what kind of winding speed they need to have on the winding job. 00:24:49.230 --> 00:24:51.030 align:middle line:84% That's how they ensure the fiber, 00:24:51.030 --> 00:24:53.090 align:middle line:84% you always have a very uniform diameter. 00:24:53.090 --> 00:24:56.110 align:middle line:90% 00:24:56.110 --> 00:24:57.650 align:middle line:84% Now, come back to this plot here. 00:24:57.650 --> 00:25:02.090 align:middle line:84% If look at the bottom panel, the draw tension, 00:25:02.090 --> 00:25:04.190 align:middle line:84% these lines are pretty much flat. 00:25:04.190 --> 00:25:08.710 align:middle line:84% So what that says is that this draw tension will quickly 00:25:08.710 --> 00:25:12.310 align:middle line:84% determine by the fiber drawing temperature, which essentially 00:25:12.310 --> 00:25:13.670 align:middle line:90% translates to viscosity. 00:25:13.670 --> 00:25:17.750 align:middle line:84% But it has a much weaker dependence on the draw speed, 00:25:17.750 --> 00:25:21.170 align:middle line:84% so it's pretty much constant draw tension for given fiber 00:25:21.170 --> 00:25:22.290 align:middle line:90% drawing temperature. 00:25:22.290 --> 00:25:24.650 align:middle line:84% You can explain that, quantitatively, 00:25:24.650 --> 00:25:28.090 align:middle line:84% by looking at this equation here where f is basically the drawing 00:25:28.090 --> 00:25:29.810 align:middle line:84% tension, and then on the right-hand side 00:25:29.810 --> 00:25:33.010 align:middle line:84% here, eta is viscosity, A is the cross-section 00:25:33.010 --> 00:25:36.530 align:middle line:84% at one particular location, and this partial derivative here, 00:25:36.530 --> 00:25:40.410 align:middle line:84% vz, is essentially the axial velocity of fiber 00:25:40.410 --> 00:25:42.010 align:middle line:90% at that particular location. 00:25:42.010 --> 00:25:44.650 align:middle line:84% And z is, of course, the vertical direction. 00:25:44.650 --> 00:25:48.330 align:middle line:84% So what it says, if you have a higher viscosity, 00:25:48.330 --> 00:25:51.750 align:middle line:84% you need a higher drawing tension, which makes sense. 00:25:51.750 --> 00:25:54.810 align:middle line:84% You have higher viscosity, you need a larger force 00:25:54.810 --> 00:25:57.310 align:middle line:84% to essentially introduce in this viscous flow. 00:25:57.310 --> 00:25:59.328 align:middle line:84% And that's why you see this dependence here. 00:25:59.328 --> 00:26:00.870 align:middle line:84% As you go to higher high temperature, 00:26:00.870 --> 00:26:03.270 align:middle line:84% the tension drops pretty rapidly, 00:26:03.270 --> 00:26:05.770 align:middle line:90% because of reduced viscosity. 00:26:05.770 --> 00:26:09.330 align:middle line:84% You also see that this parameter here the velocity 00:26:09.330 --> 00:26:11.250 align:middle line:90% gradient along the z direction. 00:26:11.250 --> 00:26:14.050 align:middle line:84% So what this characterizes, if you have, say, 00:26:14.050 --> 00:26:17.840 align:middle line:84% essentially, a cylinder with uniform cross-section, 00:26:17.840 --> 00:26:21.253 align:middle line:84% then, due to mass conservation, if you pull it, 00:26:21.253 --> 00:26:23.420 align:middle line:84% then all the parts should move at the same velocity, 00:26:23.420 --> 00:26:26.040 align:middle line:90% if it's uniform cross-section. 00:26:26.040 --> 00:26:29.480 align:middle line:84% You have a velocity gradient, because the diameter 00:26:29.480 --> 00:26:32.040 align:middle line:84% during this fiber drawing process actually not constant. 00:26:32.040 --> 00:26:34.280 align:middle line:84% On top, you have much wider, of course, the preform. 00:26:34.280 --> 00:26:37.760 align:middle line:84% At the bottom, you have a thin strand of fiber. 00:26:37.760 --> 00:26:40.860 align:middle line:84% So because of this diameter non-uniformity, 00:26:40.860 --> 00:26:43.720 align:middle line:84% it means that the vertical translational distance 00:26:43.720 --> 00:26:46.300 align:middle line:84% for each of the material, each cross-section is different. 00:26:46.300 --> 00:26:48.920 align:middle line:84% On top, the deformation is a lot smaller. 00:26:48.920 --> 00:26:51.720 align:middle line:84% And then at the bottom, because of the small amount of material 00:26:51.720 --> 00:26:54.840 align:middle line:84% at the cross-section, you need to sustain a much higher drawing 00:26:54.840 --> 00:26:58.240 align:middle line:84% speed for mass conservation, effectively. 00:26:58.240 --> 00:27:03.137 align:middle line:84% So this term, basically, essentially it 00:27:03.137 --> 00:27:05.720 align:middle line:84% would reach maximum somewhere right in between here, where you 00:27:05.720 --> 00:27:09.400 align:middle line:90% have largest diameter change. 00:27:09.400 --> 00:27:09.900 align:middle line:90% All right. 00:27:09.900 --> 00:27:13.220 align:middle line:84% So now, if you now go back come back to this equation here, 00:27:13.220 --> 00:27:16.620 align:middle line:84% let's again see F, fiber drawing tension. 00:27:16.620 --> 00:27:19.700 align:middle line:84% This is supposed to be constant throughout the entire fiber 00:27:19.700 --> 00:27:22.200 align:middle line:84% here, just due to a mechanical balance. 00:27:22.200 --> 00:27:25.560 align:middle line:90% 00:27:25.560 --> 00:27:28.540 align:middle line:84% So if eta is constant, then what it 00:27:28.540 --> 00:27:33.320 align:middle line:84% says is that this term, the velocity gradient, 00:27:33.320 --> 00:27:34.940 align:middle line:84% it should be inversely proportional 00:27:34.940 --> 00:27:38.100 align:middle line:90% to the cross-sectional area. 00:27:38.100 --> 00:27:39.780 align:middle line:84% This also makes sense because you 00:27:39.780 --> 00:27:44.660 align:middle line:84% can imagine that when you have large cross-sectional area here, 00:27:44.660 --> 00:27:48.580 align:middle line:84% the extrusion of material is small compared 00:27:48.580 --> 00:27:49.940 align:middle line:90% to the cross-sectional area. 00:27:49.940 --> 00:27:55.820 align:middle line:84% So the change in the deformation is relatively small, versus, 00:27:55.820 --> 00:27:59.900 align:middle line:84% if you move down here, you have a smaller diameter. 00:27:59.900 --> 00:28:03.840 align:middle line:84% The shape deformation becomes a lot larger. 00:28:03.840 --> 00:28:07.740 align:middle line:84% This also corresponds to a give rise to a much larger velocity 00:28:07.740 --> 00:28:09.220 align:middle line:90% gradient. 00:28:09.220 --> 00:28:12.080 align:middle line:84% So if you plot this term here, this velocity gradient, 00:28:12.080 --> 00:28:15.320 align:middle line:84% along the z-axis, what you'll see is that initially, 00:28:15.320 --> 00:28:18.300 align:middle line:84% when you have a uniform diameter, it's basically 0. 00:28:18.300 --> 00:28:21.240 align:middle line:84% It will then increase as you move down here because you 00:28:21.240 --> 00:28:24.680 align:middle line:84% start to get shape deformation, and then it will decrease again. 00:28:24.680 --> 00:28:28.940 align:middle line:84% It decreases again because as you leave the drawing furnace, 00:28:28.940 --> 00:28:30.520 align:middle line:90% the temperature starts to drop. 00:28:30.520 --> 00:28:33.980 align:middle line:84% When the temperature drops, viscosity goes up. 00:28:33.980 --> 00:28:36.680 align:middle line:84% So viscosity goes up, it means that this term will actually 00:28:36.680 --> 00:28:37.600 align:middle line:90% decrease. 00:28:37.600 --> 00:28:39.880 align:middle line:84% At the end, when you are far away from the fiber 00:28:39.880 --> 00:28:42.840 align:middle line:84% drawing furnace, you essentially goes back to room temperature, 00:28:42.840 --> 00:28:44.100 align:middle line:90% and glass becomes a solid. 00:28:44.100 --> 00:28:46.017 align:middle line:84% It doesn't want to flow anymore, and therefore 00:28:46.017 --> 00:28:48.440 align:middle line:84% deformation essentially goes to 0. 00:28:48.440 --> 00:28:51.400 align:middle line:84% So basically, if you're trying to plot the strain rate of glass 00:28:51.400 --> 00:28:53.700 align:middle line:84% materials along the axial direction, 00:28:53.700 --> 00:28:55.560 align:middle line:90% it will be essentially 0 here. 00:28:55.560 --> 00:28:58.020 align:middle line:84% It goes to some kind of maximum value somewhere in between, 00:28:58.020 --> 00:29:01.583 align:middle line:90% and then again drop back to 0. 00:29:01.583 --> 00:29:03.000 align:middle line:84% So most of the deformation process 00:29:03.000 --> 00:29:05.940 align:middle line:84% essentially takes place along of necking region. 00:29:05.940 --> 00:29:12.510 align:middle line:90% 00:29:12.510 --> 00:29:13.150 align:middle line:90% All right. 00:29:13.150 --> 00:29:16.890 align:middle line:84% So this is basically about fiber drawing process. 00:29:16.890 --> 00:29:19.790 align:middle line:84% So what I want to do-- you can already probably guess 00:29:19.790 --> 00:29:22.150 align:middle line:90% what that gadget is for. 00:29:22.150 --> 00:29:23.110 align:middle line:90% Oops. 00:29:23.110 --> 00:29:28.630 align:middle line:84% Let me just open up the separate slide deck 00:29:28.630 --> 00:29:31.390 align:middle line:84% to explain what I'm actually going to do here. 00:29:31.390 --> 00:29:36.790 align:middle line:84% I'm going to try to draw fibers using a candy glass. 00:29:36.790 --> 00:29:38.470 align:middle line:84% This is exactly the same as actually 00:29:38.470 --> 00:29:40.275 align:middle line:90% a real fiber drawing process. 00:29:40.275 --> 00:29:41.150 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:29:41.150 --> 00:29:42.070 align:middle line:90% PROFESSOR: Oh. 00:29:42.070 --> 00:29:42.910 align:middle line:90% OK. 00:29:42.910 --> 00:29:43.510 align:middle line:90% Sorry. 00:29:43.510 --> 00:29:44.050 align:middle line:90% All right. 00:29:44.050 --> 00:29:48.590 align:middle line:84% OK I just need to make sure the wire is long enough. 00:29:48.590 --> 00:29:50.530 align:middle line:90% Or I can plug in here. 00:29:50.530 --> 00:30:17.570 align:middle line:90% 00:30:17.570 --> 00:30:18.150 align:middle line:90% All right. 00:30:18.150 --> 00:30:20.970 align:middle line:90% 00:30:20.970 --> 00:30:22.890 align:middle line:90% The contact is not very good. 00:30:22.890 --> 00:30:25.890 align:middle line:90% And let me put it here. 00:30:25.890 --> 00:30:28.110 align:middle line:84% Let me turn around and adjust it a little bit. 00:30:28.110 --> 00:30:43.170 align:middle line:90% 00:30:43.170 --> 00:30:47.290 align:middle line:84% So I will leave it to heat up for a little bit. 00:30:47.290 --> 00:30:49.650 align:middle line:84% Now, while it's heating up, let me talk a little bit 00:30:49.650 --> 00:30:52.310 align:middle line:90% about what we are doing here. 00:30:52.310 --> 00:30:56.130 align:middle line:84% So you can probably guess why I'm not 00:30:56.130 --> 00:30:58.767 align:middle line:84% using real glass, because glass, of course, 00:30:58.767 --> 00:31:00.350 align:middle line:84% would require such a high temperature. 00:31:00.350 --> 00:31:02.610 align:middle line:84% You would think EHS would really frown upon it if it 00:31:02.610 --> 00:31:05.050 align:middle line:84% I do a real glass fiber drawing here. 00:31:05.050 --> 00:31:11.710 align:middle line:84% But that's not the case if we dial back 70 years, go to 1950s. 00:31:11.710 --> 00:31:13.910 align:middle line:84% What I found here, this is actually 00:31:13.910 --> 00:31:20.060 align:middle line:84% a toy that was sold by this company, Gilbert, and for kids. 00:31:20.060 --> 00:31:22.310 align:middle line:84% So you can actually draw a fiber at home regular time. 00:31:22.310 --> 00:31:25.030 align:middle line:84% They actually including a high temperature lamp, 00:31:25.030 --> 00:31:27.490 align:middle line:84% and then you can have all the glass and everything. 00:31:27.490 --> 00:31:29.365 align:middle line:84% So you can actually do glass blowing at home. 00:31:29.365 --> 00:31:32.070 align:middle line:84% You can essentially also draw fibers, obviously. 00:31:32.070 --> 00:31:38.790 align:middle line:84% If you think that's bad, well, I like to open up this. 00:31:38.790 --> 00:31:39.290 align:middle line:90% Let's see. 00:31:39.290 --> 00:31:42.230 align:middle line:90% Hopefully the internet works. 00:31:42.230 --> 00:31:45.630 align:middle line:84% Bottom line, I also found the glass blowing manual. 00:31:45.630 --> 00:31:47.630 align:middle line:84% They actually have a digital version for it now. 00:31:47.630 --> 00:31:48.790 align:middle line:90% People digitize it. 00:31:48.790 --> 00:31:51.290 align:middle line:84% And you would think they would talk about essentially safety 00:31:51.290 --> 00:31:52.310 align:middle line:90% precautions. 00:31:52.310 --> 00:31:53.350 align:middle line:90% Nothing. 00:31:53.350 --> 00:31:54.030 align:middle line:90% Nothing. 00:31:54.030 --> 00:31:55.510 align:middle line:84% The only thing they actually mentioned there is, 00:31:55.510 --> 00:31:57.050 align:middle line:84% try to not do this in your house. 00:31:57.050 --> 00:31:58.770 align:middle line:90% It's going to be a lot of smoke. 00:31:58.770 --> 00:32:00.490 align:middle line:84% It's going to make everything very black, 00:32:00.490 --> 00:32:02.633 align:middle line:84% and your mom is not going to be happy about it. 00:32:02.633 --> 00:32:04.550 align:middle line:84% Literally, that's actually what it says there. 00:32:04.550 --> 00:32:07.100 align:middle line:84% Do it in your backyard, something like that. 00:32:07.100 --> 00:32:09.180 align:middle line:84% Well, as a father of three kids, I 00:32:09.180 --> 00:32:11.520 align:middle line:84% feel like that doesn't read like a safety precaution. 00:32:11.520 --> 00:32:12.580 align:middle line:84% That's more like instruction, how 00:32:12.580 --> 00:32:15.080 align:middle line:84% to actually screw up your house and make your parents mad. 00:32:15.080 --> 00:32:16.880 align:middle line:84% I'm sure they will just follow that direction, 00:32:16.880 --> 00:32:18.047 align:middle line:90% actually do it in the house. 00:32:18.047 --> 00:32:19.380 align:middle line:90% So anyway. 00:32:19.380 --> 00:32:22.080 align:middle line:84% Now, if you think that's bad enough, let me show you this. 00:32:22.080 --> 00:32:26.680 align:middle line:84% The same company, they also have a kit for-- 00:32:26.680 --> 00:32:27.180 align:middle line:90% it's not. 00:32:27.180 --> 00:32:30.660 align:middle line:90% 00:32:30.660 --> 00:32:32.540 align:middle line:90% Let me see. 00:32:32.540 --> 00:32:34.780 align:middle line:90% Oh, [INAUDIBLE]. 00:32:34.780 --> 00:32:35.280 align:middle line:90% Come on. 00:32:35.280 --> 00:32:36.900 align:middle line:90% I mean, not this one. 00:32:36.900 --> 00:32:37.843 align:middle line:90% It's very slow. 00:32:37.843 --> 00:32:38.760 align:middle line:90% It's like we bring up. 00:32:38.760 --> 00:32:40.060 align:middle line:84% Anyway, you can read the manual there. 00:32:40.060 --> 00:32:41.720 align:middle line:84% It says, it's lots of fun actually. 00:32:41.720 --> 00:32:42.980 align:middle line:90% OK, see this. 00:32:42.980 --> 00:32:46.300 align:middle line:84% You also get a cast kit where you can play around 00:32:46.300 --> 00:32:47.520 align:middle line:90% with liquid metal. 00:32:47.520 --> 00:32:49.420 align:middle line:84% Liquid metal thing, actually cast 00:32:49.420 --> 00:32:51.340 align:middle line:90% figurines and all that stuff. 00:32:51.340 --> 00:32:53.160 align:middle line:84% So if you buy both pieces together, 00:32:53.160 --> 00:32:55.720 align:middle line:84% you can transform your home into basement of building 10. 00:32:55.720 --> 00:32:59.380 align:middle line:84% You get glass lab and metal lab all in your kitchen. 00:32:59.380 --> 00:33:00.400 align:middle line:90% You think that's nice? 00:33:00.400 --> 00:33:02.580 align:middle line:90% I have something more. 00:33:02.580 --> 00:33:04.773 align:middle line:84% AUDIENCE: Oh, it is the same one [INAUDIBLE]. 00:33:04.773 --> 00:33:05.440 align:middle line:90% PROFESSOR: Yeah. 00:33:05.440 --> 00:33:06.460 align:middle line:90% You have seen this? 00:33:06.460 --> 00:33:10.880 align:middle line:84% AUDIENCE: Yeah, because I mentioned to Alice, like, 00:33:10.880 --> 00:33:12.210 align:middle line:90% the toy uranium. 00:33:12.210 --> 00:33:13.960 align:middle line:84% PROFESSOR: No, I also want-- they actually 00:33:13.960 --> 00:33:15.880 align:middle line:84% include a little slip there in the box, 00:33:15.880 --> 00:33:17.400 align:middle line:90% if you get the original version. 00:33:17.400 --> 00:33:21.460 align:middle line:84% If you use up your radioactive isotope, you can ask for refuel. 00:33:21.460 --> 00:33:26.120 align:middle line:90% 00:33:26.120 --> 00:33:26.620 align:middle line:90% All right. 00:33:26.620 --> 00:33:31.400 align:middle line:84% Good that we have EHS today, right? 00:33:31.400 --> 00:33:34.280 align:middle line:90% Is it high enough temperature? 00:33:34.280 --> 00:33:35.840 align:middle line:90% Yeah. 00:33:35.840 --> 00:33:38.340 align:middle line:84% So anyway, we're not going to try to do that. 00:33:38.340 --> 00:33:40.020 align:middle line:84% We are going to be good citizens. 00:33:40.020 --> 00:33:42.560 align:middle line:84% We are on good terms with EHS, so that's 00:33:42.560 --> 00:33:44.360 align:middle line:90% why we want to do it with candy. 00:33:44.360 --> 00:33:47.240 align:middle line:84% So what I'm showing you on the left-hand side here, 00:33:47.240 --> 00:33:51.320 align:middle line:84% this is not a phase diagram, first thing. 00:33:51.320 --> 00:33:53.920 align:middle line:84% A lot of books mistakenly call the phase diagram. 00:33:53.920 --> 00:33:54.720 align:middle line:90% It's really not. 00:33:54.720 --> 00:33:56.800 align:middle line:84% Number one thing, you don't see glass on phase diagram. 00:33:56.800 --> 00:33:58.160 align:middle line:90% It's thermodynamically unstable. 00:33:58.160 --> 00:34:00.140 align:middle line:84% Also, you don't see this kind of crossings on phase diagram. 00:34:00.140 --> 00:34:01.280 align:middle line:90% That should not happen. 00:34:01.280 --> 00:34:03.108 align:middle line:84% This should more correctly called-- 00:34:03.108 --> 00:34:04.400 align:middle line:90% this is called a state diagram. 00:34:04.400 --> 00:34:07.520 align:middle line:84% The way it works is that in a manufacturing plant, 00:34:07.520 --> 00:34:11.179 align:middle line:84% people use this diagram to predict the concentration 00:34:11.179 --> 00:34:13.699 align:middle line:84% or composition of sugar [INAUDIBLE] solutions. 00:34:13.699 --> 00:34:15.739 align:middle line:84% So if you heat up a solution, it's 00:34:15.739 --> 00:34:18.313 align:middle line:84% going to essentially follow this line as it evolves, 00:34:18.313 --> 00:34:19.480 align:middle line:90% as you go up to temperature. 00:34:19.480 --> 00:34:21.000 align:middle line:90% That's how they actually use it. 00:34:21.000 --> 00:34:25.100 align:middle line:84% But what I want you to focus on is actually this part here. 00:34:25.100 --> 00:34:27.699 align:middle line:84% So in this region here, where you have essentially pretty 00:34:27.699 --> 00:34:33.260 align:middle line:84% much pure sugar, you can actually form a glassy material. 00:34:33.260 --> 00:34:35.179 align:middle line:84% And the way to do it is essentially 00:34:35.179 --> 00:34:38.542 align:middle line:84% you heat up a sugar solution, you completely saturate it, 00:34:38.542 --> 00:34:41.000 align:middle line:84% and then you quickly cool it down to avoid crystallization, 00:34:41.000 --> 00:34:42.659 align:middle line:90% and then you get your glass. 00:34:42.659 --> 00:34:44.580 align:middle line:84% It is a glass, so it behaves exactly 00:34:44.580 --> 00:34:47.800 align:middle line:84% in the same way in many, many aspects as oxide glasses. 00:34:47.800 --> 00:34:49.875 align:middle line:90% You can see that they are clear. 00:34:49.875 --> 00:34:51.000 align:middle line:90% You see some bubbles there. 00:34:51.000 --> 00:34:56.340 align:middle line:84% That's because this is a poor manufacturing process. 00:34:56.340 --> 00:34:58.340 align:middle line:90% I think I can pass this along. 00:34:58.340 --> 00:35:02.450 align:middle line:84% I have a much more clear glass that's bubble free. 00:35:02.450 --> 00:35:03.150 align:middle line:90% I did not do it. 00:35:03.150 --> 00:35:08.130 align:middle line:84% My student did it, so credit should not come to me. 00:35:08.130 --> 00:35:11.650 align:middle line:84% It's gray because we overcook it, 00:35:11.650 --> 00:35:15.330 align:middle line:84% so it probably decompose a little bit. 00:35:15.330 --> 00:35:17.130 align:middle line:90% But yeah, it's basically sugar. 00:35:17.130 --> 00:35:19.130 align:middle line:84% AUDIENCE: It just really feels like [INAUDIBLE]. 00:35:19.130 --> 00:35:20.250 align:middle line:90% PROFESSOR: Yeah. 00:35:20.250 --> 00:35:21.470 align:middle line:90% They can crystallize. 00:35:21.470 --> 00:35:23.487 align:middle line:84% If you expose them, especially to humidity, 00:35:23.487 --> 00:35:24.570 align:middle line:90% then they can crystallize. 00:35:24.570 --> 00:35:27.330 align:middle line:84% And you see this kind of foggy kind of precipitation? 00:35:27.330 --> 00:35:31.690 align:middle line:84% That's the crystalline materials coming out. 00:35:31.690 --> 00:35:35.470 align:middle line:84% I just want to mention that this is also a really sweet material, 00:35:35.470 --> 00:35:36.550 align:middle line:90% as you would expect. 00:35:36.550 --> 00:35:40.930 align:middle line:90% 00:35:40.930 --> 00:35:46.890 align:middle line:84% Practically all the hard candy is made of amorphous glass. 00:35:46.890 --> 00:35:50.210 align:middle line:84% I have to distinguish between another kind of candy, 00:35:50.210 --> 00:35:51.850 align:middle line:90% though, called rock candy. 00:35:51.850 --> 00:35:53.490 align:middle line:84% So in the food industry, rock candy 00:35:53.490 --> 00:35:55.530 align:middle line:84% actually specifically refers to crystalline form 00:35:55.530 --> 00:35:57.090 align:middle line:90% of sugar or sucrose. 00:35:57.090 --> 00:36:00.190 align:middle line:84% The way they actually make it is they essentially dip 00:36:00.190 --> 00:36:03.070 align:middle line:84% this kind of sticks into a solution that's 00:36:03.070 --> 00:36:04.605 align:middle line:90% supersaturated with sugar. 00:36:04.605 --> 00:36:06.230 align:middle line:84% And then they actually essentially then 00:36:06.230 --> 00:36:09.028 align:middle line:84% precipitate onto the stick to get this crystalline. 00:36:09.028 --> 00:36:10.570 align:middle line:84% You see the crystalline facets there. 00:36:10.570 --> 00:36:11.778 align:middle line:90% So this is crystalline sugar. 00:36:11.778 --> 00:36:14.530 align:middle line:84% So hard candy and rock candy are two different things, 00:36:14.530 --> 00:36:16.030 align:middle line:84% from a material science perspective. 00:36:16.030 --> 00:36:20.630 align:middle line:84% Don't mix the two, even though they may taste similar. 00:36:20.630 --> 00:36:23.630 align:middle line:84% Now, last thing, I also want to acknowledge that I actually 00:36:23.630 --> 00:36:25.930 align:middle line:90% get my recipe from a journal. 00:36:25.930 --> 00:36:28.270 align:middle line:90% Let's talk about Cuisine. 00:36:28.270 --> 00:36:30.370 align:middle line:84% I normally don't read that kind of journals, 00:36:30.370 --> 00:36:32.750 align:middle line:84% but this is actually where I get my recipe. 00:36:32.750 --> 00:36:34.990 align:middle line:84% And it turns out this kind of fibers they actually 00:36:34.990 --> 00:36:36.232 align:middle line:90% use in Chinese cuisine. 00:36:36.232 --> 00:36:38.190 align:middle line:84% They actually use it for essentially decoration 00:36:38.190 --> 00:36:40.510 align:middle line:84% and, of course, for flavoring purposes. 00:36:40.510 --> 00:36:45.230 align:middle line:84% So I copy exactly that recipe when we are making this. 00:36:45.230 --> 00:36:46.950 align:middle line:90% All right. 00:36:46.950 --> 00:36:48.965 align:middle line:84% And also, because sugar is a glass, 00:36:48.965 --> 00:36:51.590 align:middle line:84% essentially means that all the processing methods we talk about 00:36:51.590 --> 00:36:53.110 align:middle line:84% for frosted glasses can also be used 00:36:53.110 --> 00:36:54.508 align:middle line:90% to essentially structure sugar. 00:36:54.508 --> 00:36:57.050 align:middle line:84% So this is a very interesting study where people actually put 00:36:57.050 --> 00:37:00.640 align:middle line:84% together a 3D printer in very much the same way as how you 00:37:00.640 --> 00:37:02.890 align:middle line:84% actually 3D print glass materials that we talked about 00:37:02.890 --> 00:37:03.530 align:middle line:90% earlier. 00:37:03.530 --> 00:37:05.630 align:middle line:84% You have basically a little furnace here. 00:37:05.630 --> 00:37:06.830 align:middle line:90% Heat up the sugar glass. 00:37:06.830 --> 00:37:09.530 align:middle line:84% They come out from this nozzle, and by translating the nozzle 00:37:09.530 --> 00:37:11.830 align:middle line:84% in 3D space, you can get all kinds of 3D patterns. 00:37:11.830 --> 00:37:16.410 align:middle line:84% So you can actually 3D print sugar in this case. 00:37:16.410 --> 00:37:17.370 align:middle line:90% All right. 00:37:17.370 --> 00:37:20.330 align:middle line:84% So before I go to my experiment, I just want to acknowledge, 00:37:20.330 --> 00:37:22.610 align:middle line:84% if you want to get a bit more information 00:37:22.610 --> 00:37:25.670 align:middle line:84% about this experiment, this is where I got my inspiration from. 00:37:25.670 --> 00:37:27.490 align:middle line:84% And then Yifei is one of my former students 00:37:27.490 --> 00:37:29.050 align:middle line:84% who had already graduated, but he 00:37:29.050 --> 00:37:32.930 align:middle line:84% made those glasses, which I still use after like five years. 00:37:32.930 --> 00:37:36.110 align:middle line:84% So now let me turn to this guy here. 00:37:36.110 --> 00:37:40.890 align:middle line:90% I hope this is warm enough. 00:37:40.890 --> 00:37:42.210 align:middle line:90% Let me see. 00:37:42.210 --> 00:37:45.533 align:middle line:84% I can try to put it a bit closer. 00:37:45.533 --> 00:37:47.450 align:middle line:84% Sometimes it takes a little bit of time for it 00:37:47.450 --> 00:37:49.530 align:middle line:90% to just basically heat up. 00:37:49.530 --> 00:37:54.800 align:middle line:84% I think it's still not flowing yet, so we need a bit of time. 00:37:54.800 --> 00:37:56.360 align:middle line:90% So you see what I'm doing here? 00:37:56.360 --> 00:37:59.660 align:middle line:84% I'm trying to push this rod close to the surface of lamp, 00:37:59.660 --> 00:38:00.860 align:middle line:90% so heat up a bit quickly. 00:38:00.860 --> 00:38:04.640 align:middle line:90% 00:38:04.640 --> 00:38:07.260 align:middle line:84% Now you also see what happens before I came to class. 00:38:07.260 --> 00:38:10.500 align:middle line:84% This sugar basically melt and stick [INAUDIBLE] 00:38:10.500 --> 00:38:14.160 align:middle line:90% trying to pull it off. 00:38:14.160 --> 00:38:18.420 align:middle line:84% OK, just let it melt. Just be patient. 00:38:18.420 --> 00:38:22.640 align:middle line:90% 00:38:22.640 --> 00:38:25.640 align:middle line:84% AUDIENCE: The fiber drawing earlier lecture 00:38:25.640 --> 00:38:29.200 align:middle line:84% reminds me of the candy stores where 00:38:29.200 --> 00:38:33.760 align:middle line:84% they get really complicated patterns and [INAUDIBLE]. 00:38:33.760 --> 00:38:35.920 align:middle line:84% PROFESSOR: You get can essentially pull 00:38:35.920 --> 00:38:37.220 align:middle line:90% candy and that kind of thing. 00:38:37.220 --> 00:38:39.180 align:middle line:84% It's basically of like a thin fiber. 00:38:39.180 --> 00:38:39.680 align:middle line:90% Yeah. 00:38:39.680 --> 00:38:42.300 align:middle line:84% Especially when you can solidify, you can pull out, 00:38:42.300 --> 00:38:43.840 align:middle line:90% you see long fibers coming out. 00:38:43.840 --> 00:38:47.340 align:middle line:84% Yeah, that's basically elasticity, if you remember. 00:38:47.340 --> 00:38:52.600 align:middle line:90% 00:38:52.600 --> 00:38:53.960 align:middle line:90% So let me see. 00:38:53.960 --> 00:38:56.580 align:middle line:90% 00:38:56.580 --> 00:38:59.540 align:middle line:84% Maybe one of you can hold this for me for a second. 00:38:59.540 --> 00:39:03.060 align:middle line:90% I will try to do-- 00:39:03.060 --> 00:39:05.500 align:middle line:84% just hold it like that for a second. 00:39:05.500 --> 00:39:07.900 align:middle line:90% All right. 00:39:07.900 --> 00:39:11.140 align:middle line:84% I'm actually interested to see if this candy glass is 00:39:11.140 --> 00:39:14.360 align:middle line:90% transparent to this laser beam. 00:39:14.360 --> 00:39:16.540 align:middle line:84% I can see [INAUDIBLE] a couple lines to it. 00:39:16.540 --> 00:39:20.260 align:middle line:90% 00:39:20.260 --> 00:39:23.280 align:middle line:84% Yeah, somewhat, not very transparent. 00:39:23.280 --> 00:39:26.020 align:middle line:90% 00:39:26.020 --> 00:39:28.040 align:middle line:84% It absorbs a little bit because of the color, 00:39:28.040 --> 00:39:30.680 align:middle line:84% but maybe once we draw the fiber, 00:39:30.680 --> 00:39:33.480 align:middle line:84% I can see if we can actually send the light through it. 00:39:33.480 --> 00:39:34.880 align:middle line:90% I've never tried that. 00:39:34.880 --> 00:39:37.580 align:middle line:90% It may not work, but. 00:39:37.580 --> 00:39:39.880 align:middle line:90% OK, where do we stand now? 00:39:39.880 --> 00:39:44.000 align:middle line:84% It's still-- oh, it's melting, it's melting. 00:39:44.000 --> 00:39:49.860 align:middle line:90% 00:39:49.860 --> 00:39:52.595 align:middle line:84% You can see the meniscus, or it or had already deformed, 00:39:52.595 --> 00:39:53.970 align:middle line:84% so shifted towards that direction 00:39:53.970 --> 00:39:56.850 align:middle line:84% because it's heading a bit lower there. 00:39:56.850 --> 00:39:59.390 align:middle line:90% We may still need a few minutes. 00:39:59.390 --> 00:40:01.390 align:middle line:84% I think I should have heated up a bit earlier. 00:40:01.390 --> 00:40:03.920 align:middle line:90% 00:40:03.920 --> 00:40:05.170 align:middle line:90% Oh, actually, it starts there. 00:40:05.170 --> 00:40:06.670 align:middle line:90% OK, nice. 00:40:06.670 --> 00:40:09.450 align:middle line:90% 00:40:09.450 --> 00:40:11.490 align:middle line:84% Yeah, so then what we'll do, once the viscosity 00:40:11.490 --> 00:40:13.698 align:middle line:84% is about right, we're going to try to pull the fiber. 00:40:13.698 --> 00:40:16.950 align:middle line:84% And then this is basically the drum, winding drum. 00:40:16.950 --> 00:40:18.410 align:middle line:84% So what we have here, the two lamps 00:40:18.410 --> 00:40:23.530 align:middle line:84% essentially approximates the fiber drawing furnace. 00:40:23.530 --> 00:40:26.670 align:middle line:84% This is our preform, effectively, 00:40:26.670 --> 00:40:28.570 align:middle line:90% which contains the sugar glass. 00:40:28.570 --> 00:40:30.290 align:middle line:84% And then at the bottom here, we have 00:40:30.290 --> 00:40:34.090 align:middle line:84% this a very high-tech version of fiber drum, where 00:40:34.090 --> 00:40:37.570 align:middle line:84% we can have long spools of fibers, meters in length. 00:40:37.570 --> 00:40:41.090 align:middle line:84% And as you will see, the key parameter 00:40:41.090 --> 00:40:42.570 align:middle line:90% here is actually viscosity. 00:40:42.570 --> 00:40:45.563 align:middle line:84% If viscosity is too high, then fiber will break, 00:40:45.563 --> 00:40:47.230 align:middle line:84% essentially will break up in the middle. 00:40:47.230 --> 00:40:50.530 align:middle line:84% If it's too low, you get large chunks of sugar coming out. 00:40:50.530 --> 00:40:52.170 align:middle line:84% Then you don't get uniform diameter. 00:40:52.170 --> 00:40:56.070 align:middle line:90% So let's see how-- 00:40:56.070 --> 00:40:57.490 align:middle line:90% let's just wait a little bit. 00:40:57.490 --> 00:41:04.550 align:middle line:90% 00:41:04.550 --> 00:41:06.890 align:middle line:84% You can clearly see it's already flowing. 00:41:06.890 --> 00:41:09.650 align:middle line:84% You see a big bubble coming out here [INAUDIBLE] gravity. 00:41:09.650 --> 00:41:10.588 align:middle line:90% It's going to-- 00:41:10.588 --> 00:41:12.630 align:middle line:84% I'm a bit concerned it's going to flow out a bit. 00:41:12.630 --> 00:41:14.550 align:middle line:90% You see it's flowing? 00:41:14.550 --> 00:41:15.810 align:middle line:90% It's flowing a bit too fast. 00:41:15.810 --> 00:41:18.227 align:middle line:84% I think I need to move it away a little bit away from the. 00:41:18.227 --> 00:41:20.470 align:middle line:90% 00:41:20.470 --> 00:41:23.350 align:middle line:84% I think it's viscosity gets a bit too low. 00:41:23.350 --> 00:41:24.910 align:middle line:90% This is probably going to work. 00:41:24.910 --> 00:41:27.710 align:middle line:90% What I'll do is maybe a bit-- 00:41:27.710 --> 00:41:31.350 align:middle line:84% I will essentially start to draw a fiber, a thin channel fiber. 00:41:31.350 --> 00:41:32.710 align:middle line:90% I will stick it to [INAUDIBLE]. 00:41:32.710 --> 00:41:34.810 align:middle line:84% Now I need to maintain the viscosity here. 00:41:34.810 --> 00:41:41.550 align:middle line:90% 00:41:41.550 --> 00:41:45.100 align:middle line:84% So now let me try to draw a channel fiber. 00:41:45.100 --> 00:41:47.350 align:middle line:84% You can come a little bit closer, if you want to look. 00:41:47.350 --> 00:41:49.770 align:middle line:84% And then I stick to a surface, and then I 00:41:49.770 --> 00:41:52.730 align:middle line:84% can start to do my fiber drawing. 00:41:52.730 --> 00:41:54.885 align:middle line:90% Oh, it breaks. 00:41:54.885 --> 00:41:56.510 align:middle line:84% I think the viscosity is a bit too low. 00:41:56.510 --> 00:42:09.730 align:middle line:90% 00:42:09.730 --> 00:42:14.010 align:middle line:84% OK, so I don't know if the video can actually see it. 00:42:14.010 --> 00:42:16.330 align:middle line:90% It's actually coming. 00:42:16.330 --> 00:42:18.290 align:middle line:84% You may have to come close to see it. 00:42:18.290 --> 00:42:19.670 align:middle line:84% The fiber is actually coming out. 00:42:19.670 --> 00:42:25.650 align:middle line:90% 00:42:25.650 --> 00:42:28.490 align:middle line:84% You can draw it at pretty high speed, actually. 00:42:28.490 --> 00:42:30.970 align:middle line:84% If it's not flowing even fast enough. 00:42:30.970 --> 00:42:32.610 align:middle line:90% Ow! 00:42:32.610 --> 00:42:33.690 align:middle line:90% Oops. 00:42:33.690 --> 00:42:34.230 align:middle line:90% All right. 00:42:34.230 --> 00:42:35.290 align:middle line:90% That's too much. 00:42:35.290 --> 00:42:36.550 align:middle line:90% Let's try to do it again. 00:42:36.550 --> 00:42:39.700 align:middle line:90% 00:42:39.700 --> 00:42:41.950 align:middle line:84% OK, now you can clearly see the fiber drawing process. 00:42:41.950 --> 00:42:45.170 align:middle line:84% You see how it actually deforms as it comes out? 00:42:45.170 --> 00:42:47.340 align:middle line:84% And then you get a long spool of wires. 00:42:47.340 --> 00:42:52.920 align:middle line:90% 00:42:52.920 --> 00:42:56.740 align:middle line:84% Once in a while, there's a big piece comes out, 00:42:56.740 --> 00:42:58.280 align:middle line:90% but that's fine. 00:42:58.280 --> 00:43:00.000 align:middle line:84% That's why we are not a professional. 00:43:00.000 --> 00:43:02.120 align:middle line:84% AUDIENCE: Is that like the droplet formation, when 00:43:02.120 --> 00:43:03.663 align:middle line:90% it's too high viscosity? 00:43:03.663 --> 00:43:05.580 align:middle line:84% PROFESSOR: No, it's probably too low viscosity 00:43:05.580 --> 00:43:07.080 align:middle line:90% we are putting out here. 00:43:07.080 --> 00:43:09.000 align:middle line:84% I should maybe lift this a little bit 00:43:09.000 --> 00:43:12.480 align:middle line:84% higher so that you illuminate the other tip. 00:43:12.480 --> 00:43:14.440 align:middle line:84% That's where you want it to have low viscosity. 00:43:14.440 --> 00:43:16.857 align:middle line:84% If viscosity gets too low there, then the whole thing just 00:43:16.857 --> 00:43:18.218 align:middle line:90% falls off because of gravity. 00:43:18.218 --> 00:43:19.260 align:middle line:90% That's just what happens. 00:43:19.260 --> 00:43:22.860 align:middle line:84% Yeah, we can see, again, very stably draw long fibers. 00:43:22.860 --> 00:43:23.360 align:middle line:90% Oops. 00:43:23.360 --> 00:43:28.360 align:middle line:90% 00:43:28.360 --> 00:43:30.280 align:middle line:84% This can go on and on almost forever 00:43:30.280 --> 00:43:34.160 align:middle line:90% until that big chunk falls off. 00:43:34.160 --> 00:43:34.660 align:middle line:90% Take it off. 00:43:34.660 --> 00:43:37.240 align:middle line:84% Any of you want to try it yourself? 00:43:37.240 --> 00:43:38.380 align:middle line:90% I have another fiber drum. 00:43:38.380 --> 00:43:39.300 align:middle line:90% You can take a look. 00:43:39.300 --> 00:43:43.160 align:middle line:84% By the way, this actually comes off. 00:43:43.160 --> 00:43:44.220 align:middle line:90% This is the fibers. 00:43:44.220 --> 00:43:47.980 align:middle line:84% They actually come up very nicely. 00:43:47.980 --> 00:43:50.140 align:middle line:84% These are essentially the chains of fibers. 00:43:50.140 --> 00:43:52.260 align:middle line:90% They all stick together now. 00:43:52.260 --> 00:43:55.220 align:middle line:90% But yeah. 00:43:55.220 --> 00:43:58.820 align:middle line:84% You can see these are basically just sugar fibers. 00:43:58.820 --> 00:43:59.440 align:middle line:90% Eat it? 00:43:59.440 --> 00:43:59.940 align:middle line:90% No. 00:43:59.940 --> 00:44:02.940 align:middle line:90% 00:44:02.940 --> 00:44:06.780 align:middle line:84% That's not a good recommendation. 00:44:06.780 --> 00:44:08.000 align:middle line:90% Any of you want to try it? 00:44:08.000 --> 00:44:12.360 align:middle line:90% 00:44:12.360 --> 00:44:14.880 align:middle line:84% So the idea is that you try to pull a little bit, stick it 00:44:14.880 --> 00:44:16.880 align:middle line:84% to the surface, and then you can start rotating. 00:44:16.880 --> 00:44:19.860 align:middle line:90% 00:44:19.860 --> 00:44:21.380 align:middle line:90% Make sure the stick rotates. 00:44:21.380 --> 00:44:22.360 align:middle line:90% Just rotate. 00:44:22.360 --> 00:44:24.882 align:middle line:84% It's going to continuously come out. 00:44:24.882 --> 00:44:27.340 align:middle line:84% You can also laterally translate it a little bit around it, 00:44:27.340 --> 00:44:29.500 align:middle line:84% along it, so that you get essentially that there. 00:44:29.500 --> 00:44:35.740 align:middle line:90% 00:44:35.740 --> 00:44:36.240 align:middle line:90% All right. 00:44:36.240 --> 00:44:38.283 align:middle line:84% So now, this comes up a little bit. 00:44:38.283 --> 00:44:39.700 align:middle line:84% Take a look at how people actually 00:44:39.700 --> 00:44:42.000 align:middle line:90% make real silica glass fibers. 00:44:42.000 --> 00:44:45.160 align:middle line:84% So one of the critical step in making this kind of fiber here 00:44:45.160 --> 00:44:47.400 align:middle line:84% is how to actually make the fiber preform. 00:44:47.400 --> 00:44:50.220 align:middle line:84% This is really critical because, as I mentioned, 00:44:50.220 --> 00:44:53.200 align:middle line:84% in order to reach this remarkably low loss of 0.2 dB 00:44:53.200 --> 00:44:55.440 align:middle line:84% per kilometer at 1,550 nanometer, 00:44:55.440 --> 00:44:58.680 align:middle line:84% you really need to get rid of all the impurities to ensure 00:44:58.680 --> 00:45:01.760 align:middle line:84% that you can achieve this very low loss. 00:45:01.760 --> 00:45:03.880 align:middle line:84% So the way people actually use to make 00:45:03.880 --> 00:45:07.200 align:middle line:84% the preform is to use a so-called Modified Chemical 00:45:07.200 --> 00:45:11.480 align:middle line:84% Vapor Deposition process, or MCVD. 00:45:11.480 --> 00:45:13.880 align:middle line:84% It is different from conventional CVD 00:45:13.880 --> 00:45:17.880 align:middle line:84% because in conventional CVD, if you want to deposit SiO2, 00:45:17.880 --> 00:45:21.660 align:middle line:84% people typically use a silane, SiH4 gas. 00:45:21.660 --> 00:45:24.440 align:middle line:90% 00:45:24.440 --> 00:45:27.520 align:middle line:84% Now, the reason people don't use SiH4 in MCVD 00:45:27.520 --> 00:45:31.120 align:middle line:84% is because SiH4 contains lots of hydrogen. 00:45:31.120 --> 00:45:33.200 align:middle line:84% So it's very, very difficult to completely get 00:45:33.200 --> 00:45:35.280 align:middle line:84% rid of hydrogen in silica deposited 00:45:35.280 --> 00:45:37.240 align:middle line:90% using conventional CVD process. 00:45:37.240 --> 00:45:40.700 align:middle line:84% So that's why they have to use, instead of using a silane, 00:45:40.700 --> 00:45:44.270 align:middle line:84% they use a silicon chloride as a precursor. 00:45:44.270 --> 00:45:46.690 align:middle line:84% It's also convenient because silicon chloride, turns out, 00:45:46.690 --> 00:45:48.398 align:middle line:84% is actually a liquid at room temperature. 00:45:48.398 --> 00:45:50.950 align:middle line:84% It has a boiling point of 58 degrees Celsius. 00:45:50.950 --> 00:45:52.590 align:middle line:84% So it's really a very easy material 00:45:52.590 --> 00:45:55.790 align:middle line:84% that can be subjected to a fractional distillation process 00:45:55.790 --> 00:45:58.410 align:middle line:84% to get very high purity of silicon chloride. 00:45:58.410 --> 00:46:01.750 align:middle line:84% So this silicon dichloride is then actually 00:46:01.750 --> 00:46:04.970 align:middle line:84% oxidized to form essentially this kind of SiO2. 00:46:04.970 --> 00:46:08.370 align:middle line:84% They are deposited in the form of very fine powders. 00:46:08.370 --> 00:46:10.110 align:middle line:90% That's called a soot. 00:46:10.110 --> 00:46:14.270 align:middle line:84% The soot eventually condenses with high temperature 00:46:14.270 --> 00:46:16.990 align:middle line:84% to form a dense layer of silica glass. 00:46:16.990 --> 00:46:20.270 align:middle line:84% Similarly, you can also dope the glass material 00:46:20.270 --> 00:46:23.310 align:middle line:84% by essentially adding a little bit of germanium tetrachloride, 00:46:23.310 --> 00:46:25.990 align:middle line:84% and then it will decompose to form germanium oxide that forms 00:46:25.990 --> 00:46:27.630 align:middle line:90% a dopant in the glass material. 00:46:27.630 --> 00:46:34.470 align:middle line:84% Now, specifically, the way you actually make the preform 00:46:34.470 --> 00:46:39.150 align:middle line:84% is start out with essentially a tube, a silica tube that's 00:46:39.150 --> 00:46:40.410 align:middle line:90% hollow in the center. 00:46:40.410 --> 00:46:43.930 align:middle line:84% You flow the reaction gases through a tube here. 00:46:43.930 --> 00:46:45.970 align:middle line:84% You also have a torch that essentially 00:46:45.970 --> 00:46:49.330 align:middle line:84% moving in the axial direction of the tube here. 00:46:49.330 --> 00:46:51.292 align:middle line:84% So where you have the torch heating up, 00:46:51.292 --> 00:46:52.750 align:middle line:84% you have a high enough temperature. 00:46:52.750 --> 00:46:55.230 align:middle line:84% You induce the decomposition of the reactants, 00:46:55.230 --> 00:46:57.250 align:middle line:84% and that's where the soot is getting deposited. 00:46:57.250 --> 00:47:00.290 align:middle line:84% So essentially you're depositing soot 00:47:00.290 --> 00:47:02.530 align:middle line:84% as you actually translate this little torch 00:47:02.530 --> 00:47:04.730 align:middle line:90% along the axial direction here. 00:47:04.730 --> 00:47:08.690 align:middle line:84% So the process continues until the whole tube 00:47:08.690 --> 00:47:10.490 align:middle line:90% becomes completely filled. 00:47:10.490 --> 00:47:12.930 align:middle line:84% Then you actually get a high-quality preform. 00:47:12.930 --> 00:47:16.110 align:middle line:84% The process is also quite convenient to control here, 00:47:16.110 --> 00:47:18.450 align:middle line:84% because you can always change the refractive index 00:47:18.450 --> 00:47:21.570 align:middle line:84% distribution as a function of position 00:47:21.570 --> 00:47:24.130 align:middle line:84% along the radial position by just simply changing 00:47:24.130 --> 00:47:26.490 align:middle line:84% the reactant gas during this process here. 00:47:26.490 --> 00:47:28.290 align:middle line:84% Essentially, it's done in a layer 00:47:28.290 --> 00:47:33.130 align:middle line:84% by layer manner along the radial direction. 00:47:33.130 --> 00:47:35.930 align:middle line:84% So I would like to actually show you 00:47:35.930 --> 00:47:38.910 align:middle line:84% just one quick video that actually 00:47:38.910 --> 00:47:40.310 align:middle line:90% illustrates this kind of a-- 00:47:40.310 --> 00:47:41.150 align:middle line:90% [VIDEO PLAYBACK] 00:47:41.150 --> 00:47:44.830 align:middle line:84% - This three-story [INAUDIBLE] at the Thorlabs headquarters 00:47:44.830 --> 00:47:48.430 align:middle line:84% in Newton, New Jersey, houses our fiber draw towers. 00:47:48.430 --> 00:47:50.750 align:middle line:90% [MUSIC PLAYING] 00:47:50.750 --> 00:47:54.470 align:middle line:90% 00:47:54.470 --> 00:47:57.270 align:middle line:84% These towers produce many of the optical fibers 00:47:57.270 --> 00:48:00.670 align:middle line:84% we offer in our catalog, and also gives us the ability 00:48:00.670 --> 00:48:02.910 align:middle line:84% to draw high-quality custom fibers 00:48:02.910 --> 00:48:07.150 align:middle line:90% for a variety of applications. 00:48:07.150 --> 00:48:10.190 align:middle line:84% The draw process begins by cleaning all surface 00:48:10.190 --> 00:48:12.990 align:middle line:84% contaminants from the silica preform. 00:48:12.990 --> 00:48:16.350 align:middle line:84% The internal structure of an optical fiber is already present 00:48:16.350 --> 00:48:21.230 align:middle line:84% in this 1-meter-long glass cylinder. 00:48:21.230 --> 00:48:24.770 align:middle line:84% One quartz handle is welded onto each end of the preform, 00:48:24.770 --> 00:48:28.510 align:middle line:84% using an oxygen-hydrogen torch that heats the glass to 2,000 00:48:28.510 --> 00:48:29.730 align:middle line:90% degrees Celsius. 00:48:29.730 --> 00:48:36.460 align:middle line:90% 00:48:36.460 --> 00:48:39.660 align:middle line:84% Eye protection is required to shield the engineer's eyes 00:48:39.660 --> 00:48:41.160 align:middle line:84% from the brightness of the torch. 00:48:41.160 --> 00:48:55.940 align:middle line:90% 00:48:55.940 --> 00:48:58.020 align:middle line:84% Once both handles are attached, they 00:48:58.020 --> 00:49:00.740 align:middle line:84% are mounted or chucked into a lathe, 00:49:00.740 --> 00:49:06.460 align:middle line:84% suspending the preform over the water-cooled half-moon burner. 00:49:06.460 --> 00:49:08.900 align:middle line:84% As the lathe rotates the preform, 00:49:08.900 --> 00:49:11.560 align:middle line:84% the burner slowly passes along its length, 00:49:11.560 --> 00:49:13.700 align:middle line:84% removing all surface contaminants. 00:49:13.700 --> 00:49:16.660 align:middle line:84% This process is called fire polishing. 00:49:16.660 --> 00:49:20.060 align:middle line:84% The moving burner mills off the outermost layer of glass 00:49:20.060 --> 00:49:24.020 align:middle line:84% and removes all organics that may be present on the preform. 00:49:24.020 --> 00:49:26.220 align:middle line:84% The more pristine the preform, the higher 00:49:26.220 --> 00:49:29.820 align:middle line:90% quality fiber it yields. 00:49:29.820 --> 00:49:33.340 align:middle line:84% Once the fire polishing is complete, one of the two handles 00:49:33.340 --> 00:49:35.440 align:middle line:84% is removed, leaving a pointed drop-in. 00:49:35.440 --> 00:49:40.800 align:middle line:90% 00:49:40.800 --> 00:49:43.600 align:middle line:84% The other handle remains welded to the preform 00:49:43.600 --> 00:49:45.840 align:middle line:90% and serves as the support end. 00:49:45.840 --> 00:49:48.160 align:middle line:84% The support end is removed from the lathe 00:49:48.160 --> 00:49:50.120 align:middle line:84% and chopped into the top feed unit 00:49:50.120 --> 00:49:51.820 align:middle line:90% at the top of the draw tower. 00:49:51.820 --> 00:49:54.800 align:middle line:90% 00:49:54.800 --> 00:49:57.080 align:middle line:84% The unit lowers the preform into the furnace, 00:49:57.080 --> 00:50:00.520 align:middle line:84% where it is heated to 2,000 degrees Celsius. 00:50:00.520 --> 00:50:02.920 align:middle line:84% This graphite resistance furnace has 00:50:02.920 --> 00:50:06.880 align:middle line:84% an element that surrounds but does not touch the preform. 00:50:06.880 --> 00:50:09.600 align:middle line:84% The graphite allows for an even distribution of heat 00:50:09.600 --> 00:50:11.440 align:middle line:90% around the glass rod. 00:50:11.440 --> 00:50:14.120 align:middle line:84% The fiber is pushed by the top feed unit 00:50:14.120 --> 00:50:17.720 align:middle line:84% and pulled, first by hand and then by capstan, 00:50:17.720 --> 00:50:19.260 align:middle line:90% down the length of the tower. 00:50:19.260 --> 00:50:23.120 align:middle line:90% 00:50:23.120 --> 00:50:25.560 align:middle line:84% After exiting the furnace, it travels through 00:50:25.560 --> 00:50:29.080 align:middle line:84% a 150-centimeter-long cooling chamber. 00:50:29.080 --> 00:50:31.760 align:middle line:84% Chilled water and helium gas keep this chamber 00:50:31.760 --> 00:50:35.540 align:middle line:84% at an internal temperature of 12 degrees Celsius. 00:50:35.540 --> 00:50:37.380 align:middle line:84% The fiber exiting the cooling chamber 00:50:37.380 --> 00:50:38.980 align:middle line:84% has been cooled to room temperature 00:50:38.980 --> 00:50:42.820 align:middle line:90% and is ready for coating. 00:50:42.820 --> 00:50:46.020 align:middle line:84% A variety of polymer coatings can be applied to our fiber, 00:50:46.020 --> 00:50:49.060 align:middle line:90% depending on the application. 00:50:49.060 --> 00:50:52.300 align:middle line:84% A carbide die at the bottom of the conical coating cup 00:50:52.300 --> 00:50:56.100 align:middle line:84% ensures that the coating is applied evenly around the fiber. 00:50:56.100 --> 00:50:59.220 align:middle line:84% The coated fiber then passes through a set of UV lamps, 00:50:59.220 --> 00:51:01.380 align:middle line:90% where the coating is cured. 00:51:01.380 --> 00:51:04.740 align:middle line:84% Finally, the fiber receives a blue tassel jacket 00:51:04.740 --> 00:51:08.080 align:middle line:84% to protect it from mechanical damage and is then smoothed. 00:51:08.080 --> 00:51:10.820 align:middle line:90% 00:51:10.820 --> 00:51:14.580 align:middle line:84% The fiber is closely monitored as it makes the 36-foot journey 00:51:14.580 --> 00:51:18.700 align:middle line:84% from the top of the tower to the capstan below. 00:51:18.700 --> 00:51:21.700 align:middle line:84% Three meters provide real-time diameter measurements 00:51:21.700 --> 00:51:24.620 align:middle line:84% to computers controlling both speed of the top feed 00:51:24.620 --> 00:51:27.860 align:middle line:90% and the capstan. 00:51:27.860 --> 00:51:29.540 align:middle line:84% These speeds are constantly being 00:51:29.540 --> 00:51:32.730 align:middle line:84% adjusted to ensure a consistent diameter throughout the length 00:51:32.730 --> 00:51:35.490 align:middle line:90% of the fiber. 00:51:35.490 --> 00:51:38.570 align:middle line:84% Our engineers also monitor the location of the fiber 00:51:38.570 --> 00:51:40.790 align:middle line:84% as it travels through the many devices, 00:51:40.790 --> 00:51:43.570 align:middle line:84% and make adjustments when necessary. 00:51:43.570 --> 00:51:46.430 align:middle line:84% It is important that the fiber is perfectly centered, 00:51:46.430 --> 00:51:48.530 align:middle line:84% because any rubbing will negatively affect 00:51:48.530 --> 00:51:50.430 align:middle line:90% the quality of the end product. 00:51:50.430 --> 00:51:53.730 align:middle line:90% 00:51:53.730 --> 00:51:58.210 align:middle line:84% This tower can produce fiber as thin as 50 microns in diameter, 00:51:58.210 --> 00:52:00.590 align:middle line:84% which is about the thickness of human hair. 00:52:00.590 --> 00:52:03.250 align:middle line:90% 00:52:03.250 --> 00:52:06.730 align:middle line:84% The finished fiber undergoes a four-axis proof test 00:52:06.730 --> 00:52:09.410 align:middle line:90% to measure its strength. 00:52:09.410 --> 00:52:11.890 align:middle line:84% This thin optical fiber is stronger 00:52:11.890 --> 00:52:14.210 align:middle line:84% than stainless steel of the same thickness 00:52:14.210 --> 00:52:17.410 align:middle line:90% when pulled along its length. 00:52:17.410 --> 00:52:20.570 align:middle line:84% Engineers also measure the attenuation and centration 00:52:20.570 --> 00:52:23.090 align:middle line:90% of the fiber. 00:52:23.090 --> 00:52:28.410 align:middle line:84% Any fiber that does not meet spec is not sold. 00:52:28.410 --> 00:52:30.750 align:middle line:84% The fiber drawn here is sold bare 00:52:30.750 --> 00:52:33.530 align:middle line:84% and is also made into connectorized patch cables, 00:52:33.530 --> 00:52:35.270 align:middle line:84% or incorporated into many products 00:52:35.270 --> 00:52:37.607 align:middle line:84% in our ever-growing line of photonics equipment. 00:52:37.607 --> 00:52:38.190 align:middle line:90% [END PLAYBACK] 00:52:38.190 --> 00:52:40.482 align:middle line:84% PROFESSOR: This is how actually our fiber manufacturing 00:52:40.482 --> 00:52:42.910 align:middle line:84% is actually executed in industry setting. 00:52:42.910 --> 00:52:43.990 align:middle line:90% Now, comes back here. 00:52:43.990 --> 00:52:46.150 align:middle line:84% In the remaining 10 minutes, I will just also 00:52:46.150 --> 00:52:50.750 align:middle line:84% quickly talk about some of the more recent developments 00:52:50.750 --> 00:52:54.350 align:middle line:84% in constructing fibers with more complex geometry. 00:52:54.350 --> 00:52:57.230 align:middle line:84% I have already mentioned that if you can construct essentially 00:52:57.230 --> 00:52:59.390 align:middle line:84% a preform with complex structures containing 00:52:59.390 --> 00:53:01.710 align:middle line:84% different materials, you can use this fiber drawing 00:53:01.710 --> 00:53:03.550 align:middle line:84% process to essentially draw down to create 00:53:03.550 --> 00:53:05.310 align:middle line:84% a scale version of optical fibers 00:53:05.310 --> 00:53:06.870 align:middle line:90% with different cross sections. 00:53:06.870 --> 00:53:10.450 align:middle line:84% So one such example is so-called Microstructured Optical Fiber. 00:53:10.450 --> 00:53:11.750 align:middle line:90% Some people call it MOF. 00:53:11.750 --> 00:53:14.110 align:middle line:84% Some people call it photonic crystal fiber. 00:53:14.110 --> 00:53:16.790 align:middle line:84% They refer to basically the same thing. 00:53:16.790 --> 00:53:18.183 align:middle line:84% You can see in this diagram here, 00:53:18.183 --> 00:53:20.350 align:middle line:84% this is actually cross-section of some of the fibers 00:53:20.350 --> 00:53:22.430 align:middle line:84% that's shown in this geometry here. 00:53:22.430 --> 00:53:24.510 align:middle line:84% You see this kind of hexagonal arrangement 00:53:24.510 --> 00:53:28.570 align:middle line:90% of periodic holes, air holes. 00:53:28.570 --> 00:53:31.770 align:middle line:84% And you see that in this optical micrograph image here 00:53:31.770 --> 00:53:34.290 align:middle line:84% they have different colors due to the different purity 00:53:34.290 --> 00:53:35.238 align:middle line:90% of the holes. 00:53:35.238 --> 00:53:37.030 align:middle line:84% They essentially refract light differently. 00:53:37.030 --> 00:53:38.270 align:middle line:84% That's why they have different colors there. 00:53:38.270 --> 00:53:39.930 align:middle line:84% But they are basically colorless. 00:53:39.930 --> 00:53:41.930 align:middle line:84% This is silica glass in this case. 00:53:41.930 --> 00:53:45.795 align:middle line:84% So this kind of a fiber can be quite useful here. 00:53:45.795 --> 00:53:47.170 align:middle line:84% If you look at the cross-section, 00:53:47.170 --> 00:53:48.690 align:middle line:90% this is the way it looks like. 00:53:48.690 --> 00:53:53.410 align:middle line:84% They can be used, for example, for sensing applications. 00:53:53.410 --> 00:53:55.970 align:middle line:84% You can essentially use this hollow core 00:53:55.970 --> 00:53:58.570 align:middle line:90% as a conduit for liquid or air. 00:53:58.570 --> 00:54:02.210 align:middle line:84% So when air flow through it, you can also pass light through it 00:54:02.210 --> 00:54:04.610 align:middle line:84% so that you can actually create a very strong interaction 00:54:04.610 --> 00:54:06.850 align:middle line:84% between light and the material or the gas or species 00:54:06.850 --> 00:54:09.770 align:middle line:84% you want to detect to make it more sensitive. 00:54:09.770 --> 00:54:12.490 align:middle line:84% This is also useful for high-power handling 00:54:12.490 --> 00:54:14.130 align:middle line:90% applications. 00:54:14.130 --> 00:54:16.643 align:middle line:84% So even though we mentioned that glass is basically 00:54:16.643 --> 00:54:18.810 align:middle line:84% very low loss, but still there's always a little bit 00:54:18.810 --> 00:54:20.172 align:middle line:90% of residual optical loss. 00:54:20.172 --> 00:54:22.130 align:middle line:84% This residual optical loss means that if you're 00:54:22.130 --> 00:54:24.270 align:middle line:84% passing laser with very high energy through it, 00:54:24.270 --> 00:54:26.812 align:middle line:84% it's going to actually heat up the material, which eventually 00:54:26.812 --> 00:54:30.480 align:middle line:84% damage it, versus if you have such a hollow core, 00:54:30.480 --> 00:54:33.760 align:middle line:84% then the attenuation or the absorption 00:54:33.760 --> 00:54:35.560 align:middle line:84% can actually be even lower, such that there 00:54:35.560 --> 00:54:38.512 align:middle line:84% is no such photothermal effect, handle much higher power. 00:54:38.512 --> 00:54:40.720 align:middle line:84% So these are some applications where you can actually 00:54:40.720 --> 00:54:43.600 align:middle line:84% use this kind of fiber for different kind 00:54:43.600 --> 00:54:45.760 align:middle line:90% of various implementations. 00:54:45.760 --> 00:54:47.380 align:middle line:90% Now, a couple of questions here. 00:54:47.380 --> 00:54:48.980 align:middle line:84% One is that how this fiber is made. 00:54:48.980 --> 00:54:50.980 align:middle line:84% You can see this actually manufacturing process. 00:54:50.980 --> 00:54:53.560 align:middle line:84% So what you do, you start out by essentially stacking 00:54:53.560 --> 00:54:55.640 align:middle line:84% a lot of silica tubes, little tubes, 00:54:55.640 --> 00:54:57.700 align:middle line:90% in a hexagonal arrangement. 00:54:57.700 --> 00:55:01.200 align:middle line:84% And then you essentially do a little pre-draw 00:55:01.200 --> 00:55:04.120 align:middle line:84% to draw this kind of a-- to essentially fuse together 00:55:04.120 --> 00:55:06.520 align:middle line:84% all the individual silica rods and then 00:55:06.520 --> 00:55:10.000 align:middle line:84% pull them into essentially a little preform that in this case 00:55:10.000 --> 00:55:12.720 align:middle line:84% is a few millimeters in diameter. 00:55:12.720 --> 00:55:15.720 align:middle line:84% So this is what it looks like once all the tubes are 00:55:15.720 --> 00:55:16.800 align:middle line:90% fused together. 00:55:16.800 --> 00:55:20.080 align:middle line:84% And then you can then subject it to a fiber drawing process 00:55:20.080 --> 00:55:22.420 align:middle line:84% to get the final microstructured optical fiber. 00:55:22.420 --> 00:55:23.420 align:middle line:90% So that's how it's made. 00:55:23.420 --> 00:55:25.962 align:middle line:84% Now, the second question is that how you actually guide light 00:55:25.962 --> 00:55:27.300 align:middle line:90% in the kind of structure here. 00:55:27.300 --> 00:55:30.080 align:middle line:84% So this is not through total internal reflection, 00:55:30.080 --> 00:55:31.740 align:middle line:90% because air has a low index. 00:55:31.740 --> 00:55:33.820 align:middle line:84% So the way this works is basically 00:55:33.820 --> 00:55:35.820 align:middle line:84% through a so-called photonic bandgap effect. 00:55:35.820 --> 00:55:39.100 align:middle line:84% So you can see this kind of periodic arrangement of glass 00:55:39.100 --> 00:55:40.820 align:middle line:84% and air is essentially correspond 00:55:40.820 --> 00:55:43.980 align:middle line:84% to periodic modulation of the refractive index. 00:55:43.980 --> 00:55:46.940 align:middle line:84% So this is basically what people call the photonic crystal. 00:55:46.940 --> 00:55:49.700 align:middle line:84% You can create basically photonic bandgaps 00:55:49.700 --> 00:55:52.500 align:middle line:84% where certain frequency of light doesn't propagate. 00:55:52.500 --> 00:55:54.760 align:middle line:84% This is similar to electronic bandgap. 00:55:54.760 --> 00:55:57.420 align:middle line:84% Electronic bandgap, what it means is that electronic wave 00:55:57.420 --> 00:56:01.740 align:middle line:84% functions of certain energy cannot exist within that bandgap 00:56:01.740 --> 00:56:02.320 align:middle line:90% region. 00:56:02.320 --> 00:56:05.300 align:middle line:84% Same idea, so light of certain frequency cannot propagate 00:56:05.300 --> 00:56:09.500 align:middle line:84% in this kind of synthetic photonic crystal media. 00:56:09.500 --> 00:56:12.540 align:middle line:84% So if you send light in the core because it cannot propagate 00:56:12.540 --> 00:56:17.260 align:middle line:84% in the region, they have to be trapped in air through this 00:56:17.260 --> 00:56:18.880 align:middle line:90% photonic crystal bandgap effect. 00:56:18.880 --> 00:56:23.770 align:middle line:90% 00:56:23.770 --> 00:56:26.730 align:middle line:84% So that's one example about how this kind of fiber 00:56:26.730 --> 00:56:29.290 align:middle line:90% could be constructed and used. 00:56:29.290 --> 00:56:31.930 align:middle line:84% The figure I'm showing on the right-hand side here, 00:56:31.930 --> 00:56:34.610 align:middle line:84% this is actually from Professor Yoel Fink's group. 00:56:34.610 --> 00:56:37.290 align:middle line:84% His group is really one of the pioneers 00:56:37.290 --> 00:56:39.290 align:middle line:84% of the so-called multi-material fibers. 00:56:39.290 --> 00:56:43.930 align:middle line:84% So in this particular example here, it's a hollow core fiber. 00:56:43.930 --> 00:56:48.410 align:middle line:84% The cladding consists of alternating layers of a polymer, 00:56:48.410 --> 00:56:51.130 align:middle line:84% PES, as well as arsenic selenide, which 00:56:51.130 --> 00:56:53.010 align:middle line:90% is the chalcogenide glass. 00:56:53.010 --> 00:56:57.290 align:middle line:84% So the way it operates is that if you have alternating layers 00:56:57.290 --> 00:57:00.250 align:middle line:84% of a high- and low-index material, in this case 00:57:00.250 --> 00:57:02.490 align:middle line:84% low-index is polymer, high-index is glass, 00:57:02.490 --> 00:57:04.650 align:middle line:84% you can also create effectively-- this 00:57:04.650 --> 00:57:07.490 align:middle line:84% is a one-dimensional photonic crystal that 00:57:07.490 --> 00:57:11.170 align:middle line:84% act as a perfect mirror that reflect 100% of light. 00:57:11.170 --> 00:57:13.050 align:middle line:84% So then light will essentially also 00:57:13.050 --> 00:57:16.210 align:middle line:84% be trapped just inside this hollow core. 00:57:16.210 --> 00:57:18.490 align:middle line:84% Besides these two materials, they were also 00:57:18.490 --> 00:57:20.950 align:middle line:84% able to incorporate, in this case, 00:57:20.950 --> 00:57:23.390 align:middle line:84% they have a thin layer of gassed. 00:57:23.390 --> 00:57:25.290 align:middle line:84% This is [INAUDIBLE] chalcogenide glass. 00:57:25.290 --> 00:57:27.190 align:middle line:84% It's actually amorphous semiconductor. 00:57:27.190 --> 00:57:30.950 align:middle line:84% They also put metal in there, in this case metal tin. 00:57:30.950 --> 00:57:33.590 align:middle line:84% So metal tin, turns out it has a melting point that's actually 00:57:33.590 --> 00:57:36.030 align:middle line:84% comparable to the essentially fiber drawing temperature 00:57:36.030 --> 00:57:38.790 align:middle line:84% window of this chalcogenide glass and the polymer there. 00:57:38.790 --> 00:57:41.470 align:middle line:84% So they were able to essentially draw all four 00:57:41.470 --> 00:57:43.270 align:middle line:84% materials simultaneously together 00:57:43.270 --> 00:57:44.790 align:middle line:90% to form this fiber here. 00:57:44.790 --> 00:57:47.110 align:middle line:84% So now you may wonder, why do you actually 00:57:47.110 --> 00:57:49.890 align:middle line:84% need to put this metal lines in there? 00:57:49.890 --> 00:57:53.113 align:middle line:84% So turns out they can be used essentially as metal electrodes 00:57:53.113 --> 00:57:55.030 align:middle line:84% to actually measure the electrical conductance 00:57:55.030 --> 00:57:58.110 align:middle line:84% of this amorphous semiconductor embedded in the fiber. 00:57:58.110 --> 00:57:59.748 align:middle line:90% So why is it useful? 00:57:59.748 --> 00:58:01.790 align:middle line:84% Actually, let me jump ahead and talk a little bit 00:58:01.790 --> 00:58:03.870 align:middle line:84% about that first, before coming back to how you actually 00:58:03.870 --> 00:58:04.570 align:middle line:90% make the fiber. 00:58:04.570 --> 00:58:08.310 align:middle line:84% So they essentially leverage the fact 00:58:08.310 --> 00:58:11.350 align:middle line:84% that the electrical conductivity of this amorphous semiconductor 00:58:11.350 --> 00:58:13.790 align:middle line:84% is a very sensitive function of temperature. 00:58:13.790 --> 00:58:17.990 align:middle line:84% And the mechanism for electrical conduction in this case 00:58:17.990 --> 00:58:19.890 align:middle line:84% is basically through extended states. 00:58:19.890 --> 00:58:21.307 align:middle line:84% So if you go to high temperature, 00:58:21.307 --> 00:58:22.890 align:middle line:84% you can excite more electrons and hole 00:58:22.890 --> 00:58:24.670 align:middle line:90% pairs to the extended states. 00:58:24.670 --> 00:58:27.610 align:middle line:84% And as a result, you have a higher electrical conductivity. 00:58:27.610 --> 00:58:30.530 align:middle line:84% So with this kind of embedded thin electrodes, 00:58:30.530 --> 00:58:33.050 align:middle line:84% you can then essentially measure the local temperature 00:58:33.050 --> 00:58:35.837 align:middle line:84% in the fiber, and this is a way to actually create 00:58:35.837 --> 00:58:37.670 align:middle line:84% essentially fiber-based temperature sensing. 00:58:37.670 --> 00:58:39.875 align:middle line:84% So in this demo, we have shown here 00:58:39.875 --> 00:58:42.250 align:middle line:84% they essentially generate this kind of crisscross pattern 00:58:42.250 --> 00:58:43.390 align:middle line:90% of optical fibers. 00:58:43.390 --> 00:58:46.910 align:middle line:84% And if you put your finger at one of the cross points here, 00:58:46.910 --> 00:58:49.210 align:middle line:84% you can immediately pinpoint this is where you actually 00:58:49.210 --> 00:58:50.470 align:middle line:90% have elevated temperature. 00:58:50.470 --> 00:58:52.080 align:middle line:84% Similarly, if you put a piece of ice, 00:58:52.080 --> 00:58:54.330 align:middle line:84% this is where you actually have depressed temperature. 00:58:54.330 --> 00:58:56.050 align:middle line:84% So it's a way to essentially monitor 00:58:56.050 --> 00:58:57.690 align:middle line:90% temperature in this case. 00:58:57.690 --> 00:59:01.250 align:middle line:84% Some other examples of applications 00:59:01.250 --> 00:59:02.770 align:middle line:84% with this kind of fiber here, this 00:59:02.770 --> 00:59:06.450 align:middle line:84% is from Professor Polina Anikeeva's group. 00:59:06.450 --> 00:59:10.250 align:middle line:84% What they have demonstrated is a multifunctional platform 00:59:10.250 --> 00:59:12.990 align:middle line:84% for essentially neural excitation and recording. 00:59:12.990 --> 00:59:16.850 align:middle line:84% So conventionally people can use essentially 00:59:16.850 --> 00:59:20.160 align:middle line:84% a light to actually modulate neural activities. 00:59:20.160 --> 00:59:21.960 align:middle line:84% They can then use metal electrodes 00:59:21.960 --> 00:59:24.220 align:middle line:84% to actually record some of the neural functions, 00:59:24.220 --> 00:59:25.140 align:middle line:90% neural activities. 00:59:25.140 --> 00:59:28.085 align:middle line:84% But the conventional way of doing it 00:59:28.085 --> 00:59:29.460 align:middle line:84% is that you have separate probes. 00:59:29.460 --> 00:59:32.740 align:middle line:84% You have essentially a probe that has electrical functions. 00:59:32.740 --> 00:59:34.360 align:middle line:84% You have a fiber or something else 00:59:34.360 --> 00:59:35.780 align:middle line:90% to actually implement the light. 00:59:35.780 --> 00:59:37.600 align:middle line:84% So what they have demonstrated is 00:59:37.600 --> 00:59:40.140 align:middle line:84% that this is really a multifunctional platform here, 00:59:40.140 --> 00:59:42.348 align:middle line:84% where you can use a fiber to actually transmit light. 00:59:42.348 --> 00:59:44.723 align:middle line:84% You have the electrodes, the metal [INAUDIBLE] electrodes 00:59:44.723 --> 00:59:46.120 align:middle line:90% will record neuron activities. 00:59:46.120 --> 00:59:47.760 align:middle line:84% You also have a hollow core which 00:59:47.760 --> 00:59:50.960 align:middle line:84% allows you to actually inject drugs so that you can also 00:59:50.960 --> 00:59:54.240 align:middle line:84% use that to provide additional functions there, so essentially 00:59:54.240 --> 00:59:56.360 align:middle line:90% as a microfluidic channel. 00:59:56.360 --> 01:00:00.345 align:middle line:90% So now comes back here. 01:00:00.345 --> 01:00:02.720 align:middle line:84% So with all this kind of multi-material, multi-functional 01:00:02.720 --> 01:00:05.560 align:middle line:84% fibers, this slide just shows you some of the examples 01:00:05.560 --> 01:00:07.340 align:middle line:84% how people can actually fabricate them. 01:00:07.340 --> 01:00:09.720 align:middle line:84% It all boils down to how you actually 01:00:09.720 --> 01:00:13.680 align:middle line:84% create a preform that has a geometry that's exactly 01:00:13.680 --> 01:00:16.380 align:middle line:84% the scaled version of what you want to get in the final fiber. 01:00:16.380 --> 01:00:19.120 align:middle line:84% So on the left-hand side here, these 01:00:19.120 --> 01:00:20.620 align:middle line:84% are some of the simple examples, how 01:00:20.620 --> 01:00:22.620 align:middle line:84% you create this kind of multi-material platform. 01:00:22.620 --> 01:00:26.700 align:middle line:84% You can do rod-in-tube, sliding one material inside the other. 01:00:26.700 --> 01:00:28.780 align:middle line:84% You can also use the extrusion method. 01:00:28.780 --> 01:00:31.780 align:middle line:84% You start out making this kind of a little pellet here. 01:00:31.780 --> 01:00:34.980 align:middle line:84% By extrusion, you can then generate a preform containing 01:00:34.980 --> 01:00:36.620 align:middle line:90% multiple different materials. 01:00:36.620 --> 01:00:39.400 align:middle line:84% We have already shown you one example about stack-and-draw 01:00:39.400 --> 01:00:41.460 align:middle line:84% in the case of a silica microstructured 01:00:41.460 --> 01:00:42.740 align:middle line:90% optical fiber here. 01:00:42.740 --> 01:00:45.013 align:middle line:84% You can again, by stacking different materials, 01:00:45.013 --> 01:00:47.180 align:middle line:84% different tubes together, you can draw them together 01:00:47.180 --> 01:00:48.480 align:middle line:90% to form optical fiber. 01:00:48.480 --> 01:00:52.812 align:middle line:84% Now, for the kind of optical fiber from Professor Yoel Fink's 01:00:52.812 --> 01:00:55.020 align:middle line:84% group, this is the process they have developed, which 01:00:55.020 --> 01:00:56.840 align:middle line:90% I think is quite ingenious. 01:00:56.840 --> 01:01:00.020 align:middle line:84% So they will start out by having essentially-- this 01:01:00.020 --> 01:01:01.560 align:middle line:90% is like a rod. 01:01:01.560 --> 01:01:04.140 align:middle line:84% You then embed this kind of metal strips in there. 01:01:04.140 --> 01:01:06.220 align:middle line:90% Those are the metal electrodes. 01:01:06.220 --> 01:01:08.700 align:middle line:84% What you do next is that you essentially 01:01:08.700 --> 01:01:12.260 align:middle line:84% wrap it around a thin sheet, polymer sheet. 01:01:12.260 --> 01:01:14.520 align:middle line:84% So this polymer sheet, they actually 01:01:14.520 --> 01:01:16.520 align:middle line:84% deposit a layer of chalcogenide glass 01:01:16.520 --> 01:01:18.760 align:middle line:90% on top of the polymer sheet. 01:01:18.760 --> 01:01:20.680 align:middle line:84% So when you are actually loading it, 01:01:20.680 --> 01:01:23.440 align:middle line:84% it then creates this kind of alternating layers 01:01:23.440 --> 01:01:25.602 align:middle line:84% of chalcogenide glass and polymer here 01:01:25.602 --> 01:01:27.560 align:middle line:84% to form this kind of a one-dimensional photonic 01:01:27.560 --> 01:01:28.400 align:middle line:90% crystal. 01:01:28.400 --> 01:01:31.220 align:middle line:84% So in the end, this is actually what a preform looks like. 01:01:31.220 --> 01:01:33.440 align:middle line:84% When you draw it, you get a microstructured fiber 01:01:33.440 --> 01:01:36.040 align:middle line:84% that I mentioned in the previous slides. 01:01:36.040 --> 01:01:40.740 align:middle line:84% So this is a very intensively investigated area right now. 01:01:40.740 --> 01:01:42.800 align:middle line:84% People have all kinds of very creative pathways 01:01:42.800 --> 01:01:44.640 align:middle line:84% to actually generate diverse geometries 01:01:44.640 --> 01:01:48.560 align:middle line:84% of microstructured fibers for different applications. 01:01:48.560 --> 01:01:49.140 align:middle line:90% All right. 01:01:49.140 --> 01:01:51.422 align:middle line:84% So I'm actually going to stop here for this lecture. 01:01:51.422 --> 01:01:52.880 align:middle line:84% And then in the next lecture, we're 01:01:52.880 --> 01:01:54.338 align:middle line:84% going to switch gears a little bit, 01:01:54.338 --> 01:01:56.920 align:middle line:84% talk about optical waveguides, the on-chip version 01:01:56.920 --> 01:02:00.350 align:middle line:84% of essentially light conducting media. 01:02:00.350 --> 01:03:12.000 align:middle line:90%