WEBVTT 00:00:00.000 --> 00:00:02.420 align:middle line:84% The following content is provided under a Creative 00:00:02.420 --> 00:00:03.600 align:middle line:90% Commons license. 00:00:03.600 --> 00:00:06.050 align:middle line:84% Your support will help MIT OpenCourseWare 00:00:06.050 --> 00:00:09.650 align:middle line:84% continue to offer high quality educational resources for free. 00:00:09.650 --> 00:00:12.620 align:middle line:84% To make a donation or to view additional materials 00:00:12.620 --> 00:00:16.520 align:middle line:84% from hundreds of MIT courses, visit MIT OpenCourseWare 00:00:16.520 --> 00:00:17.750 align:middle line:90% at ocw.mit.edu. 00:00:17.750 --> 00:00:20.095 align:middle line:90% 00:00:20.095 --> 00:00:20.720 align:middle line:90% AUDIENCE: Yeah. 00:00:20.720 --> 00:00:24.620 align:middle line:84% So you have basically two semiconductors to start with-- 00:00:24.620 --> 00:00:26.270 align:middle line:90% one n and one t-- 00:00:26.270 --> 00:00:28.850 align:middle line:90% one p-type and-- p-type. 00:00:28.850 --> 00:00:32.330 align:middle line:84% So for example, on the left-hand side, you have a p-type. 00:00:32.330 --> 00:00:34.420 align:middle line:84% Right-hand side, you have an n-type. 00:00:34.420 --> 00:00:40.040 align:middle line:84% And since you have a high concentration of electrons 00:00:40.040 --> 00:00:43.250 align:middle line:84% on this side, the electrons will basically jump to this side. 00:00:43.250 --> 00:00:44.810 align:middle line:84% And you have a higher concentration 00:00:44.810 --> 00:00:47.540 align:middle line:90% of holes on the p-type. 00:00:47.540 --> 00:00:50.410 align:middle line:84% Holes will basically jump to the other side. 00:00:50.410 --> 00:00:53.380 align:middle line:84% So you will end up with this space charge 00:00:53.380 --> 00:00:56.220 align:middle line:84% that [INAUDIBLE] talked about last week. 00:00:56.220 --> 00:01:01.475 align:middle line:84% And this is basically the characteristic of [INAUDIBLE]. 00:01:01.475 --> 00:01:03.870 align:middle line:90% And I just throw them there. 00:01:03.870 --> 00:01:06.618 align:middle line:90% And I see these [INAUDIBLE]. 00:01:06.618 --> 00:01:11.990 align:middle line:84% So you have a neutral [INAUDIBLE]. 00:01:11.990 --> 00:01:17.255 align:middle line:84% So [INAUDIBLE] you have your [INAUDIBLE]. 00:01:17.255 --> 00:01:19.682 align:middle line:84% And then you have the contact [INAUDIBLE] which basically 00:01:19.682 --> 00:01:22.950 align:middle line:90% [INAUDIBLE]. 00:01:22.950 --> 00:01:23.450 align:middle line:90% OK? 00:01:23.450 --> 00:01:26.300 align:middle line:84% So you impose [INAUDIBLE] or something 00:01:26.300 --> 00:01:29.900 align:middle line:84% to this junction to get rid of all the drivers 00:01:29.900 --> 00:01:33.820 align:middle line:84% or close the circuit to generate [INAUDIBLE] power. 00:01:33.820 --> 00:01:36.110 align:middle line:84% I just talked about the open circuit voltage. 00:01:36.110 --> 00:01:38.840 align:middle line:84% So this is the first thing that basically happens when you 00:01:38.840 --> 00:01:40.550 align:middle line:90% [INAUDIBLE]. 00:01:40.550 --> 00:01:42.604 align:middle line:90% This space just [INAUDIBLE]. 00:01:42.604 --> 00:01:48.360 align:middle line:90% 00:01:48.360 --> 00:01:55.080 align:middle line:84% For [INAUDIBLE], and this basically 00:01:55.080 --> 00:02:00.240 align:middle line:84% also includes [INAUDIBLE], both in voltage [INAUDIBLE]. 00:02:00.240 --> 00:02:10.870 align:middle line:84% So you have the space charge region that's right here. 00:02:10.870 --> 00:02:17.870 align:middle line:84% And then basically, for example, the [INAUDIBLE] concentration. 00:02:17.870 --> 00:02:26.120 align:middle line:90% 00:02:26.120 --> 00:02:33.326 align:middle line:84% And so for the electron, this is p-type and n-type electron 00:02:33.326 --> 00:02:34.990 align:middle line:90% [INAUDIBLE] concentration. 00:02:34.990 --> 00:02:42.590 align:middle line:84% And so you're basically saying you have neutral [INAUDIBLE]. 00:02:42.590 --> 00:02:53.010 align:middle line:84% You have a field generator to do [INAUDIBLE]. 00:02:53.010 --> 00:02:58.790 align:middle line:84% So in case you have-- for example, a photon comes in. 00:02:58.790 --> 00:03:00.802 align:middle line:84% And we'll talk about this later at some point. 00:03:00.802 --> 00:03:02.260 align:middle line:84% When a photon comes in, in general, 00:03:02.260 --> 00:03:04.220 align:middle line:90% it's [INAUDIBLE] exciton. 00:03:04.220 --> 00:03:06.590 align:middle line:84% So they're still very close to each other. 00:03:06.590 --> 00:03:09.630 align:middle line:84% These excitons are just generally-- 00:03:09.630 --> 00:03:12.330 align:middle line:90% 00:03:12.330 --> 00:03:17.100 align:middle line:84% they have to use [INAUDIBLE] space [INAUDIBLE], 00:03:17.100 --> 00:03:20.640 align:middle line:90% but it's separate. 00:03:20.640 --> 00:03:27.416 align:middle line:84% So for this case, for example, we have [INAUDIBLE]. 00:03:27.416 --> 00:03:31.280 align:middle line:90% 00:03:31.280 --> 00:03:34.770 align:middle line:84% So basically, the electron will basically 00:03:34.770 --> 00:03:39.780 align:middle line:84% track through an electron's force in this direction. 00:03:39.780 --> 00:03:45.900 align:middle line:84% So we're having to [INAUDIBLE] to basically get 00:03:45.900 --> 00:03:47.160 align:middle line:90% separate from the [INAUDIBLE]. 00:03:47.160 --> 00:03:48.780 align:middle line:90% And it jumps on the other side. 00:03:48.780 --> 00:03:52.290 align:middle line:84% And you end up with just a positive charge here. 00:03:52.290 --> 00:03:55.440 align:middle line:84% And due to the neutral region from here, 00:03:55.440 --> 00:03:59.190 align:middle line:84% first, you have the diffusion of the electron [INAUDIBLE] two-way 00:03:59.190 --> 00:04:00.720 align:middle line:90% contacts in here. 00:04:00.720 --> 00:04:04.930 align:middle line:84% So further, so this is the carrier concentration. 00:04:04.930 --> 00:04:12.780 align:middle line:84% And then you can also draw the charge in there. 00:04:12.780 --> 00:04:17.200 align:middle line:84% So [INAUDIBLE] charge here, positive charge here. 00:04:17.200 --> 00:04:24.730 align:middle line:84% And the new field looks like [INAUDIBLE]. 00:04:24.730 --> 00:04:26.140 align:middle line:90% So if you have-- 00:04:26.140 --> 00:04:30.100 align:middle line:84% this generator decided to [INAUDIBLE] about this, 00:04:30.100 --> 00:04:34.115 align:middle line:84% or more accurately, it's called a [INAUDIBLE]. 00:04:34.115 --> 00:04:41.210 align:middle line:90% 00:04:41.210 --> 00:04:43.190 align:middle line:90% And I think this was also-- 00:04:43.190 --> 00:04:45.860 align:middle line:84% when he talked about the [INAUDIBLE] last week, 00:04:45.860 --> 00:04:54.140 align:middle line:84% he said something that-- this is important [INAUDIBLE] talk 00:04:54.140 --> 00:04:54.890 align:middle line:90% about the buyback. 00:04:54.890 --> 00:05:00.620 align:middle line:84% The buyback was just the spatial expansion 00:05:00.620 --> 00:05:02.060 align:middle line:90% of distribution [INAUDIBLE]. 00:05:02.060 --> 00:05:07.702 align:middle line:84% So he basically said this [INAUDIBLE]. 00:05:07.702 --> 00:05:14.600 align:middle line:90% 00:05:14.600 --> 00:05:17.500 align:middle line:84% So just say up to a certain point 00:05:17.500 --> 00:05:21.970 align:middle line:84% it basically drops down to a certain level [INAUDIBLE], 00:05:21.970 --> 00:05:23.480 align:middle line:90% especially [INAUDIBLE]. 00:05:23.480 --> 00:05:25.400 align:middle line:90% Up to this point, the exciton-- 00:05:25.400 --> 00:05:29.480 align:middle line:84% when it comes, it will start feeding the space charge region, 00:05:29.480 --> 00:05:32.704 align:middle line:84% and will start to separate the electron [INAUDIBLE]. 00:05:32.704 --> 00:05:39.210 align:middle line:90% 00:05:39.210 --> 00:05:41.230 align:middle line:84% And then I wanted to talk about-- 00:05:41.230 --> 00:05:44.050 align:middle line:84% so in the [INAUDIBLE] case, when we actually 00:05:44.050 --> 00:05:52.120 align:middle line:84% have a photon coming up to [INAUDIBLE] electron, 00:05:52.120 --> 00:05:53.070 align:middle line:90% [INAUDIBLE]. 00:05:53.070 --> 00:05:56.400 align:middle line:90% 00:05:56.400 --> 00:06:06.170 align:middle line:84% So when we joined [INAUDIBLE] environment 00:06:06.170 --> 00:06:09.640 align:middle line:84% since the [INAUDIBLE] have to be the same. 00:06:09.640 --> 00:06:13.446 align:middle line:90% So [INAUDIBLE] somehow. 00:06:13.446 --> 00:06:15.786 align:middle line:84% So this would be the [INAUDIBLE]. 00:06:15.786 --> 00:06:19.062 align:middle line:90% 00:06:19.062 --> 00:06:20.686 align:middle line:90% So that's [INAUDIBLE]. 00:06:20.686 --> 00:06:23.524 align:middle line:90% 00:06:23.524 --> 00:06:28.170 align:middle line:84% And see, what happens is most of the times, 00:06:28.170 --> 00:06:33.150 align:middle line:84% an electron is basically excited [INAUDIBLE]. 00:06:33.150 --> 00:06:37.070 align:middle line:90% 00:06:37.070 --> 00:06:42.130 align:middle line:84% But at the same time, first, you have the [INAUDIBLE] state. 00:06:42.130 --> 00:06:45.250 align:middle line:84% And then when the particle goes through the state [INAUDIBLE], 00:06:45.250 --> 00:06:48.334 align:middle line:90% et cetera, that's [INAUDIBLE]. 00:06:48.334 --> 00:06:51.540 align:middle line:90% 00:06:51.540 --> 00:06:57.880 align:middle line:84% And this basically changes the [INAUDIBLE]. 00:06:57.880 --> 00:07:01.570 align:middle line:84% So what you end up with is the [INAUDIBLE] 00:07:01.570 --> 00:07:04.470 align:middle line:84% will actually increase this to [INAUDIBLE]. 00:07:04.470 --> 00:07:08.083 align:middle line:90% 00:07:08.083 --> 00:07:11.820 align:middle line:84% And when we remove the electron here to end up with a hole 00:07:11.820 --> 00:07:15.350 align:middle line:84% here-- so we're basically keeping this model here. 00:07:15.350 --> 00:07:23.425 align:middle line:84% So we end up with some type of region just [INAUDIBLE]. 00:07:23.425 --> 00:07:31.890 align:middle line:90% 00:07:31.890 --> 00:07:34.230 align:middle line:84% This is basically the whole circuit board. 00:07:34.230 --> 00:07:36.480 align:middle line:84% So the difference between these two [INAUDIBLE] 00:07:36.480 --> 00:07:40.210 align:middle line:84% is due to the exchange of [INAUDIBLE]. 00:07:40.210 --> 00:07:42.312 align:middle line:90% This is basically [INAUDIBLE]. 00:07:42.312 --> 00:07:47.760 align:middle line:84% So the idea is, obviously, you could, basically, 00:07:47.760 --> 00:07:48.740 align:middle line:90% go all the way. 00:07:48.740 --> 00:07:51.870 align:middle line:90% So it's maximum open circuit. 00:07:51.870 --> 00:07:53.331 align:middle line:90% And just [INAUDIBLE]. 00:07:53.331 --> 00:08:00.912 align:middle line:90% 00:08:00.912 --> 00:08:03.120 align:middle line:84% AUDIENCE: So the electron goes through the conduction 00:08:03.120 --> 00:08:04.190 align:middle line:90% band [INAUDIBLE], right? 00:08:04.190 --> 00:08:06.776 align:middle line:90% Is it because that's easier? 00:08:06.776 --> 00:08:08.190 align:middle line:90% AUDIENCE: Yeah. 00:08:08.190 --> 00:08:09.300 align:middle line:90% What was the question? 00:08:09.300 --> 00:08:13.870 align:middle line:84% AUDIENCE: So the [INAUDIBLE] band [INAUDIBLE]? 00:08:13.870 --> 00:08:14.370 align:middle line:90% Yeah. 00:08:14.370 --> 00:08:16.605 align:middle line:90% 00:08:16.605 --> 00:08:17.980 align:middle line:84% AUDIENCE: That's a good question. 00:08:17.980 --> 00:08:20.370 align:middle line:84% I was thinking about that, too-- whether it's actually 00:08:20.370 --> 00:08:23.405 align:middle line:84% a step like this that goes-- and then it goes down. 00:08:23.405 --> 00:08:24.030 align:middle line:90% AUDIENCE: Yeah. 00:08:24.030 --> 00:08:26.150 align:middle line:90% 00:08:26.150 --> 00:08:27.400 align:middle line:90% GANG CHEN: Very good question. 00:08:27.400 --> 00:08:28.840 align:middle line:90% I think I can provide an answer. 00:08:28.840 --> 00:08:32.160 align:middle line:90% 00:08:32.160 --> 00:08:34.729 align:middle line:90% AUDIENCE: You want to answer? 00:08:34.729 --> 00:08:37.260 align:middle line:84% GANG CHEN: Well, it's a distributor. 00:08:37.260 --> 00:08:42.590 align:middle line:84% So at this point, you should not look at the spatial. 00:08:42.590 --> 00:08:43.970 align:middle line:90% So it's a locally-- 00:08:43.970 --> 00:08:48.215 align:middle line:84% it could be anywhere along that curve-- 00:08:48.215 --> 00:08:50.620 align:middle line:90% the induction curve. 00:08:50.620 --> 00:08:53.220 align:middle line:90% So I think that's a good place. 00:08:53.220 --> 00:08:57.912 align:middle line:84% I think back to the same point that I [INAUDIBLE]. 00:08:57.912 --> 00:09:08.190 align:middle line:90% 00:09:08.190 --> 00:09:11.190 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:09:11.190 --> 00:09:15.445 align:middle line:90% 00:09:15.445 --> 00:09:19.740 align:middle line:84% GANG CHEN: So yeah, that's where we stopped. 00:09:19.740 --> 00:09:28.360 align:middle line:84% The last lecture is about the, apparently, [INAUDIBLE] 00:09:28.360 --> 00:09:30.440 align:middle line:90% junction diode. 00:09:30.440 --> 00:09:32.590 align:middle line:90% It's a Schottky barrier. 00:09:32.590 --> 00:09:33.550 align:middle line:90% You can think about-- 00:09:33.550 --> 00:09:35.590 align:middle line:84% on the left there's a Schottky barrier. 00:09:35.590 --> 00:09:38.230 align:middle line:90% On the right is the induction. 00:09:38.230 --> 00:09:42.340 align:middle line:84% And when you look at the current voltage characteristics, 00:09:42.340 --> 00:09:45.580 align:middle line:84% it looks very similar except that 00:09:45.580 --> 00:09:49.240 align:middle line:84% the coefficient-- because remember, the [INAUDIBLE] 00:09:49.240 --> 00:09:51.190 align:middle line:90% for [INAUDIBLE]. 00:09:51.190 --> 00:09:55.030 align:middle line:84% And saturation [INAUDIBLE] quality-- 00:09:55.030 --> 00:09:58.480 align:middle line:84% this is a metal semiconductor junction-- 00:09:58.480 --> 00:10:01.030 align:middle line:90% a Schottky junction. 00:10:01.030 --> 00:10:06.340 align:middle line:84% And in the case of a metal semiconductor Schottky 00:10:06.340 --> 00:10:11.650 align:middle line:84% barrier, the saturation current [INAUDIBLE] the barrier module. 00:10:11.650 --> 00:10:16.570 align:middle line:84% And in the case of the [INAUDIBLE] junction, 00:10:16.570 --> 00:10:20.230 align:middle line:84% the saturation current include the band gap 00:10:20.230 --> 00:10:24.250 align:middle line:84% and include the [INAUDIBLE] of hole. 00:10:24.250 --> 00:10:27.310 align:middle line:84% From the t side is the [INAUDIBLE] hole 00:10:27.310 --> 00:10:29.140 align:middle line:90% that carries the current. 00:10:29.140 --> 00:10:31.340 align:middle line:90% On the inside is the electron. 00:10:31.340 --> 00:10:33.760 align:middle line:84% So if you see an electron on the inside, 00:10:33.760 --> 00:10:36.460 align:middle line:90% you can see the problem inside. 00:10:36.460 --> 00:10:40.060 align:middle line:90% And then we have this one-- 00:10:40.060 --> 00:10:42.400 align:middle line:90% recombination time. 00:10:42.400 --> 00:10:46.930 align:middle line:84% This is, say, when you create an electron hole pair, 00:10:46.930 --> 00:10:51.580 align:middle line:84% you excite the electron from valence 00:10:51.580 --> 00:10:54.220 align:middle line:84% to conduction in the conduction band. 00:10:54.220 --> 00:10:58.030 align:middle line:90% How long they can be separated-- 00:10:58.030 --> 00:11:00.190 align:middle line:84% and after a certain amount of time, 00:11:00.190 --> 00:11:05.470 align:middle line:84% they will recombine and give off either photon, as it were, 00:11:05.470 --> 00:11:08.840 align:middle line:84% laser or light emitting-- that, you want-- 00:11:08.840 --> 00:11:14.000 align:middle line:90% or give off the general forms. 00:11:14.000 --> 00:11:18.020 align:middle line:84% So this is the recombination time for hole electron. 00:11:18.020 --> 00:11:21.350 align:middle line:84% You can see that the difference, really, 00:11:21.350 --> 00:11:24.290 align:middle line:84% in this saturation [INAUDIBLE] current 00:11:24.290 --> 00:11:30.260 align:middle line:84% can sometimes [INAUDIBLE] mathematics. 00:11:30.260 --> 00:11:33.330 align:middle line:84% But I want to emphasize another fundamental difference. 00:11:33.330 --> 00:11:34.028 align:middle line:90% Yes? 00:11:34.028 --> 00:11:35.900 align:middle line:90% AUDIENCE: Sorry. 00:11:35.900 --> 00:11:39.050 align:middle line:84% I would have thought intuitively that the longer 00:11:39.050 --> 00:11:42.050 align:middle line:84% the recombination time, the more current you would have. 00:11:42.050 --> 00:11:46.880 align:middle line:84% GANG CHEN: The longer the recombination time-- yes, 00:11:46.880 --> 00:11:48.020 align:middle line:90% let's see. 00:11:48.020 --> 00:11:59.120 align:middle line:84% Recombination is also related to longer recombination. 00:11:59.120 --> 00:12:09.490 align:middle line:84% Well, longer recombination-- smaller [INAUDIBLE] current. 00:12:09.490 --> 00:12:19.990 align:middle line:84% But I see your point, where if the longer recombination-- 00:12:19.990 --> 00:12:22.870 align:middle line:84% essentially, recombination happens in this region. 00:12:22.870 --> 00:12:23.770 align:middle line:90% OK? 00:12:23.770 --> 00:12:30.010 align:middle line:84% So I think in that case, what happens is the diffusion 00:12:30.010 --> 00:12:34.180 align:middle line:90% supply is also limited-- 00:12:34.180 --> 00:12:36.730 align:middle line:90% your current flow. 00:12:36.730 --> 00:12:38.950 align:middle line:84% So this is where I want to emphasize. 00:12:38.950 --> 00:12:43.780 align:middle line:84% The important difference between these two formulas 00:12:43.780 --> 00:12:45.940 align:middle line:90% is the Schottky barrier. 00:12:45.940 --> 00:12:50.475 align:middle line:84% When electrons go from here to here, it's electron. 00:12:50.475 --> 00:12:53.170 align:middle line:90% It's still electron. 00:12:53.170 --> 00:12:54.640 align:middle line:90% The way I draw it-- 00:12:54.640 --> 00:12:56.260 align:middle line:90% here is the electron. 00:12:56.260 --> 00:12:58.720 align:middle line:84% And the metal side is the electron. 00:12:58.720 --> 00:13:00.610 align:middle line:90% Whole problem. 00:13:00.610 --> 00:13:04.930 align:middle line:84% Now, if you look at the t junction, 00:13:04.930 --> 00:13:08.440 align:middle line:90% on the inside is electrons. 00:13:08.440 --> 00:13:10.840 align:middle line:90% And then p-type's the hole. 00:13:10.840 --> 00:13:14.670 align:middle line:84% So if electrons go [INAUDIBLE] here, 00:13:14.670 --> 00:13:18.460 align:middle line:84% electron goes from here to here, go through this space charge 00:13:18.460 --> 00:13:23.710 align:middle line:84% region, there's no more electron. 00:13:23.710 --> 00:13:25.990 align:middle line:90% You have two p-type [INAUDIBLE]. 00:13:25.990 --> 00:13:29.140 align:middle line:90% So what happens? 00:13:29.140 --> 00:13:31.720 align:middle line:84% And you have to have current continuity. 00:13:31.720 --> 00:13:35.620 align:middle line:84% The charge has to be supplied, right? 00:13:35.620 --> 00:13:40.870 align:middle line:84% So what happens is that the hole will 00:13:40.870 --> 00:13:44.970 align:middle line:90% recombine with this electron. 00:13:44.970 --> 00:13:46.230 align:middle line:90% OK? 00:13:46.230 --> 00:13:50.910 align:middle line:84% So the hole recombines with the electron. 00:13:50.910 --> 00:13:54.350 align:middle line:84% That means you have to supply both from this side. 00:13:54.350 --> 00:13:56.640 align:middle line:84% So you draw a hole from this side, 00:13:56.640 --> 00:13:59.460 align:middle line:84% recombine with whatever electron comes to this side. 00:13:59.460 --> 00:14:02.368 align:middle line:84% And this is where the recombination really happens. 00:14:02.368 --> 00:14:03.660 align:middle line:90% It's not only the space charge. 00:14:03.660 --> 00:14:07.350 align:middle line:84% It's slightly just offset the space charge. 00:14:07.350 --> 00:14:10.890 align:middle line:84% So it's a minority carrier device. 00:14:10.890 --> 00:14:13.500 align:middle line:84% That's a really fundamental difference 00:14:13.500 --> 00:14:16.080 align:middle line:84% between a majority carrier device and a minority carrier 00:14:16.080 --> 00:14:17.010 align:middle line:90% device. 00:14:17.010 --> 00:14:20.730 align:middle line:84% Minority carrier device-- recombination is crucial. 00:14:20.730 --> 00:14:23.220 align:middle line:84% And later on, I'm going to talk about solar cells. 00:14:23.220 --> 00:14:27.000 align:middle line:84% All people are trying to refine solar silicon. 00:14:27.000 --> 00:14:32.670 align:middle line:84% It's trying to increase this recombination. 00:14:32.670 --> 00:14:33.540 align:middle line:90% OK? 00:14:33.540 --> 00:14:36.200 align:middle line:90% So that's your problem. 00:14:36.200 --> 00:14:37.650 align:middle line:90% This is the point. 00:14:37.650 --> 00:14:42.570 align:middle line:84% That's really important to appreciate the minority 00:14:42.570 --> 00:14:46.030 align:middle line:90% carrier [INAUDIBLE]. 00:14:46.030 --> 00:14:47.416 align:middle line:90% OK? 00:14:47.416 --> 00:14:50.830 align:middle line:84% And of course, I mentioned-- so this is what I drew on the card. 00:14:50.830 --> 00:14:56.870 align:middle line:84% You can see electron diffusion is on this side. 00:14:56.870 --> 00:15:00.490 align:middle line:84% So the hole diffuses with the-- once it comes to this side, 00:15:00.490 --> 00:15:02.140 align:middle line:84% there's no more electrons, right? 00:15:02.140 --> 00:15:04.390 align:middle line:84% So the electrons will recombine with holes. 00:15:04.390 --> 00:15:07.450 align:middle line:84% So there's less electron flux forward. 00:15:07.450 --> 00:15:08.480 align:middle line:90% So there is more. 00:15:08.480 --> 00:15:11.920 align:middle line:84% So if I look at electron carbon, the decrease 00:15:11.920 --> 00:15:16.942 align:middle line:84% in this recombination regime, and then I 00:15:16.942 --> 00:15:18.025 align:middle line:90% look at the hole current-- 00:15:18.025 --> 00:15:20.650 align:middle line:90% 00:15:20.650 --> 00:15:23.620 align:middle line:84% on this side, the hole difference in this region. 00:15:23.620 --> 00:15:28.270 align:middle line:84% That's where the supply recombines with electrons. 00:15:28.270 --> 00:15:32.530 align:middle line:84% So the electron goes through this [INAUDIBLE] 00:15:32.530 --> 00:15:35.080 align:middle line:90% region recombination. 00:15:35.080 --> 00:15:37.120 align:middle line:90% This is as approximation. 00:15:37.120 --> 00:15:40.540 align:middle line:84% So the concept-- and then after space charge, 00:15:40.540 --> 00:15:42.430 align:middle line:90% they recombine with the hole. 00:15:42.430 --> 00:15:44.870 align:middle line:90% So the electrons diminish. 00:15:44.870 --> 00:15:48.230 align:middle line:84% Now at this side, it's just overflow. 00:15:48.230 --> 00:15:53.960 align:middle line:84% So that's how we go from an electron to a hole. 00:15:53.960 --> 00:15:56.450 align:middle line:90% So the PN junction theory-- 00:15:56.450 --> 00:15:58.820 align:middle line:90% Shockley-- let's see. 00:15:58.820 --> 00:16:00.745 align:middle line:90% The PN junction with the-- 00:16:00.745 --> 00:16:01.245 align:middle line:90% let's see. 00:16:01.245 --> 00:16:04.640 align:middle line:84% It's three people who got work with [INAUDIBLE]. 00:16:04.640 --> 00:16:08.360 align:middle line:84% One is Shockley, and the other is Bardeen. 00:16:08.360 --> 00:16:10.820 align:middle line:90% And the third is Brattain. 00:16:10.820 --> 00:16:18.980 align:middle line:84% And I think it was Shockley and Brattain who did the invention. 00:16:18.980 --> 00:16:25.910 align:middle line:84% And Bardeen was the manager at [INAUDIBLE] building. 00:16:25.910 --> 00:16:30.340 align:middle line:84% And then if you look at the history, 00:16:30.340 --> 00:16:36.380 align:middle line:84% Bardeen was one of the guys that [INAUDIBLE]. 00:16:36.380 --> 00:16:38.250 align:middle line:84% Superconductivity-- also [INAUDIBLE]. 00:16:38.250 --> 00:16:38.780 align:middle line:90% Right? 00:16:38.780 --> 00:16:40.430 align:middle line:90% So he was very-- 00:16:40.430 --> 00:16:42.620 align:middle line:84% the history of the [INAUDIBLE] invention-- 00:16:42.620 --> 00:16:44.240 align:middle line:84% Bardeen had the [INAUDIBLE] field. 00:16:44.240 --> 00:16:48.640 align:middle line:84% This was so important, he had to do something [INAUDIBLE] theory, 00:16:48.640 --> 00:16:49.540 align:middle line:90% OK? 00:16:49.540 --> 00:16:53.110 align:middle line:84% But this is really a very ingenious theory. 00:16:53.110 --> 00:16:58.140 align:middle line:84% Think about how electrons on this side and those 00:16:58.140 --> 00:16:59.350 align:middle line:90% on the other side-- 00:16:59.350 --> 00:17:01.990 align:middle line:84% by the end, the your current has to be continued. 00:17:01.990 --> 00:17:05.859 align:middle line:84% The summation of the electron current, the hole current-- 00:17:05.859 --> 00:17:07.329 align:middle line:90% that's constant, right? 00:17:07.329 --> 00:17:09.960 align:middle line:90% 00:17:09.960 --> 00:17:14.520 align:middle line:84% And in the last lecture, I also said 00:17:14.520 --> 00:17:20.130 align:middle line:84% if you look at the energy flow and when electrons right here-- 00:17:20.130 --> 00:17:25.829 align:middle line:84% yes, I would like to know what happens [INAUDIBLE] 00:17:25.829 --> 00:17:27.030 align:middle line:90% because [INAUDIBLE]. 00:17:27.030 --> 00:17:35.222 align:middle line:90% 00:17:35.222 --> 00:17:37.110 align:middle line:90% It's a matter of-- 00:17:37.110 --> 00:17:38.010 align:middle line:90% there's no hope. 00:17:38.010 --> 00:17:39.945 align:middle line:90% Again, there's no hope. 00:17:39.945 --> 00:17:44.295 align:middle line:84% So typically, contact, people want [INAUDIBLE] contact. 00:17:44.295 --> 00:17:48.570 align:middle line:90% So what happens is that when-- 00:17:48.570 --> 00:17:53.280 align:middle line:84% so the whole motion is really majority of electron motion 00:17:53.280 --> 00:17:54.870 align:middle line:90% in [INAUDIBLE] region. 00:17:54.870 --> 00:17:57.330 align:middle line:84% So when [INAUDIBLE] move this way-- 00:17:57.330 --> 00:18:02.040 align:middle line:84% so to move from here to here, that 00:18:02.040 --> 00:18:06.360 align:middle line:84% means this electron goes out to the metal. 00:18:06.360 --> 00:18:09.882 align:middle line:90% So that's what happens. 00:18:09.882 --> 00:18:13.590 align:middle line:84% And they have [INAUDIBLE] potential-- 00:18:13.590 --> 00:18:14.850 align:middle line:90% electrochemical potential. 00:18:14.850 --> 00:18:17.640 align:middle line:84% That's what the [INAUDIBLE] junction is-- 00:18:17.640 --> 00:18:19.710 align:middle line:84% sort of like [INAUDIBLE] barrier, but 00:18:19.710 --> 00:18:25.170 align:middle line:90% [INAUDIBLE] at those sites. 00:18:25.170 --> 00:18:29.200 align:middle line:84% And if you think about the counter effect, 00:18:29.200 --> 00:18:32.190 align:middle line:84% counter [INAUDIBLE] is the kinetic energy part, 00:18:32.190 --> 00:18:38.580 align:middle line:84% which is a 2 KG of the charge average [INAUDIBLE] minus 00:18:38.580 --> 00:18:46.320 align:middle line:90% [INAUDIBLE] minus ef-- 00:18:46.320 --> 00:18:49.050 align:middle line:90% the chemical energy. 00:18:49.050 --> 00:18:51.420 align:middle line:90% And you look at this-- 00:18:51.420 --> 00:18:54.360 align:middle line:90% [INAUDIBLE] minus ef is growing. 00:18:54.360 --> 00:18:58.170 align:middle line:84% Here, on the other side, [INAUDIBLE] minus ef is here. 00:18:58.170 --> 00:19:01.680 align:middle line:84% And the space charge region is growing larger and larger. 00:19:01.680 --> 00:19:03.570 align:middle line:90% So electrons carry more heat. 00:19:03.570 --> 00:19:06.945 align:middle line:84% Per charge, it's carrying more heat. 00:19:06.945 --> 00:19:11.410 align:middle line:84% It's carrying heat out of this space charge region. 00:19:11.410 --> 00:19:17.550 align:middle line:84% That's why the space charge region is actually [INAUDIBLE]. 00:19:17.550 --> 00:19:21.610 align:middle line:84% And then immediately outside space charge region, 00:19:21.610 --> 00:19:24.534 align:middle line:90% recombination happens. 00:19:24.534 --> 00:19:27.190 align:middle line:84% Recombination dumps the heat just immediately 00:19:27.190 --> 00:19:28.790 align:middle line:90% outside the space charge region. 00:19:28.790 --> 00:19:31.420 align:middle line:84% It will flow back when the space charge region heats up. 00:19:31.420 --> 00:19:34.120 align:middle line:84% The physics is it's not the junction-- 00:19:34.120 --> 00:19:36.940 align:middle line:84% the space charge that really heats up. 00:19:36.940 --> 00:19:41.590 align:middle line:84% The heat is actually generated outside the space charge 00:19:41.590 --> 00:19:43.240 align:middle line:90% here in this recombination. 00:19:43.240 --> 00:19:47.880 align:middle line:90% This is where heat is generated. 00:19:47.880 --> 00:19:50.654 align:middle line:84% That's a fact that most people probably don't know. 00:19:50.654 --> 00:19:55.255 align:middle line:90% 00:19:55.255 --> 00:20:01.120 align:middle line:84% So to finish what I was talking last lecture, the [INAUDIBLE] 00:20:01.120 --> 00:20:02.460 align:middle line:90% more research. 00:20:02.460 --> 00:20:05.740 align:middle line:90% This is the basic physics. 00:20:05.740 --> 00:20:10.030 align:middle line:84% And so we're talking about the Richardson 00:20:10.030 --> 00:20:12.370 align:middle line:90% formula, thermionic emission. 00:20:12.370 --> 00:20:15.970 align:middle line:84% So there are people doing, for example, 00:20:15.970 --> 00:20:17.920 align:middle line:84% in the-- this is the conduction band. 00:20:17.920 --> 00:20:21.280 align:middle line:84% If you don't have the Fermi level around this place, 00:20:21.280 --> 00:20:24.190 align:middle line:84% the energy electron below the Fermi level 00:20:24.190 --> 00:20:26.440 align:middle line:84% is actually carrying negative energy, 00:20:26.440 --> 00:20:29.140 align:middle line:84% and the electron above carries positive energy. 00:20:29.140 --> 00:20:35.650 align:middle line:84% So my idea is if I do a Schottky barrier or potential barrier, 00:20:35.650 --> 00:20:40.690 align:middle line:84% and I just scatter this part of the electron, that's 00:20:40.690 --> 00:20:42.768 align:middle line:84% called the energy field term, or it depends on 00:20:42.768 --> 00:20:43.810 align:middle line:90% if you have an interface. 00:20:43.810 --> 00:20:46.000 align:middle line:84% You also call it thermionic emission. 00:20:46.000 --> 00:20:49.630 align:middle line:84% So some people call this thermionic emission [INAUDIBLE] 00:20:49.630 --> 00:20:53.420 align:middle line:84% thermoelectric [INAUDIBLE] thermionic effect. 00:20:53.420 --> 00:20:55.630 align:middle line:84% The so-called solid state thermionic. 00:20:55.630 --> 00:21:01.400 align:middle line:84% But say essentially, it's cutting out the low energy 00:21:01.400 --> 00:21:04.130 align:middle line:84% electron, or sometimes it carries latent heat 00:21:04.130 --> 00:21:08.300 align:middle line:84% because the electron below the [INAUDIBLE]. 00:21:08.300 --> 00:21:19.010 align:middle line:84% And if I [INAUDIBLE] states, most [INAUDIBLE] states 00:21:19.010 --> 00:21:21.390 align:middle line:90% need sharp feature. 00:21:21.390 --> 00:21:25.450 align:middle line:84% The sharp feature helps increase the thermoelectric effect. 00:21:25.450 --> 00:21:28.940 align:middle line:84% So the way I view is, in this region, 00:21:28.940 --> 00:21:30.990 align:middle line:90% I cut it off by transport. 00:21:30.990 --> 00:21:33.230 align:middle line:84% I cut out the transport electrons. 00:21:33.230 --> 00:21:35.970 align:middle line:84% I still have the electron, but they are scattered. 00:21:35.970 --> 00:21:36.770 align:middle line:90% They don't move. 00:21:36.770 --> 00:21:43.520 align:middle line:84% So then the middle ones up here, and I can [INAUDIBLE]. 00:21:43.520 --> 00:21:47.090 align:middle line:84% So that's the way I say, OK, it's all really coming back 00:21:47.090 --> 00:21:49.490 align:middle line:84% to the idea, if you can create the sharp features 00:21:49.490 --> 00:21:52.400 align:middle line:84% in these states, [INAUDIBLE] a lot of electrons there, 00:21:52.400 --> 00:21:56.480 align:middle line:84% and you can potentially improve the electron power 00:21:56.480 --> 00:21:58.190 align:middle line:90% factor [INAUDIBLE] sigma. 00:21:58.190 --> 00:22:01.030 align:middle line:84% But see, the key is you don't want that. 00:22:01.030 --> 00:22:02.540 align:middle line:84% You only want to scatter this part. 00:22:02.540 --> 00:22:05.210 align:middle line:84% You won't affect this part [INAUDIBLE]. 00:22:05.210 --> 00:22:08.480 align:middle line:84% That's [INAUDIBLE] lot of states. 00:22:08.480 --> 00:22:11.150 align:middle line:84% I think that this idea [INAUDIBLE] 00:22:11.150 --> 00:22:15.960 align:middle line:90% fully convincing yet to me. 00:22:15.960 --> 00:22:18.290 align:middle line:90% And I made a few slides on this. 00:22:18.290 --> 00:22:21.660 align:middle line:90% That's all my own [INAUDIBLE]. 00:22:21.660 --> 00:22:25.830 align:middle line:90% I published a paper before. 00:22:25.830 --> 00:22:29.550 align:middle line:84% There's no experiment, and maybe some of you will be interested. 00:22:29.550 --> 00:22:32.610 align:middle line:84% But basically, an issue I'm wrestling-- 00:22:32.610 --> 00:22:35.040 align:middle line:90% what is the evaporation? 00:22:35.040 --> 00:22:38.970 align:middle line:84% I said that the molecule coming out 00:22:38.970 --> 00:22:43.080 align:middle line:84% of the liquid surface, that's like an evaporation. 00:22:43.080 --> 00:22:47.130 align:middle line:84% That's equivalent to thermionic emission. 00:22:47.130 --> 00:22:53.190 align:middle line:84% Let's say, how about I have a vacuum here. 00:22:53.190 --> 00:22:57.060 align:middle line:84% I got a liquid here, and the [INAUDIBLE] of liquid drop 00:22:57.060 --> 00:23:02.070 align:middle line:84% into the vacuum side area and suddenly expand. 00:23:02.070 --> 00:23:05.250 align:middle line:84% Is that corresponding to evaporation? 00:23:05.250 --> 00:23:08.850 align:middle line:84% And I have a solid state case, where I'm 00:23:08.850 --> 00:23:10.920 align:middle line:90% comparing these two examples. 00:23:10.920 --> 00:23:13.620 align:middle line:90% And these two example are-- 00:23:13.620 --> 00:23:17.035 align:middle line:84% why is the thermionic barrier, as everybody 00:23:17.035 --> 00:23:20.590 align:middle line:84% have described before, with electron [INAUDIBLE] this way. 00:23:20.590 --> 00:23:22.960 align:middle line:84% The low energy electron in this region, 00:23:22.960 --> 00:23:26.470 align:middle line:84% [INAUDIBLE] surrounding [INAUDIBLE]. 00:23:26.470 --> 00:23:28.270 align:middle line:90% The other case-- 00:23:28.270 --> 00:23:31.570 align:middle line:90% I consider [INAUDIBLE] physics-- 00:23:31.570 --> 00:23:34.510 align:middle line:84% is what about I flow electron this way 00:23:34.510 --> 00:23:37.390 align:middle line:90% and suddenly drop a [INAUDIBLE]? 00:23:37.390 --> 00:23:40.240 align:middle line:84% And that means to me it's more like I take a droplet 00:23:40.240 --> 00:23:43.165 align:middle line:84% and put it in vacuum and suddenly evaporate. 00:23:43.165 --> 00:23:46.900 align:middle line:84% And this one is more like I have a water surface and the water 00:23:46.900 --> 00:23:50.890 align:middle line:84% molecule coming out of the surface like [INAUDIBLE] water. 00:23:50.890 --> 00:23:56.740 align:middle line:84% So that's a solid state analogy that we're trying to make. 00:23:56.740 --> 00:24:00.250 align:middle line:84% So what I did is I did some energy balance 00:24:00.250 --> 00:24:05.680 align:middle line:84% analysis, which is essentially a recursive formula [INAUDIBLE]. 00:24:05.680 --> 00:24:09.190 align:middle line:84% And you can do-- you can assume the two sides has 00:24:09.190 --> 00:24:13.030 align:middle line:84% different temperature on both sides and different type 00:24:13.030 --> 00:24:14.800 align:middle line:90% of potential. 00:24:14.800 --> 00:24:17.170 align:middle line:84% The real way to do this is [INAUDIBLE]. 00:24:17.170 --> 00:24:19.490 align:middle line:90% Very complicated how to do it. 00:24:19.490 --> 00:24:22.000 align:middle line:84% And so you solve this, actually, problem-- 00:24:22.000 --> 00:24:27.410 align:middle line:84% even when I try to [INAUDIBLE] a very, very detailed, very hard 00:24:27.410 --> 00:24:29.570 align:middle line:90% computation. 00:24:29.570 --> 00:24:33.670 align:middle line:84% [INAUDIBLE] you can do your derivation. 00:24:33.670 --> 00:24:37.240 align:middle line:84% No [INAUDIBLE] statistics, energy balance, 00:24:37.240 --> 00:24:39.430 align:middle line:84% current balance at the interface. 00:24:39.430 --> 00:24:43.810 align:middle line:84% And what I found is that this is the electron 00:24:43.810 --> 00:24:44.900 align:middle line:90% temperature on this side. 00:24:44.900 --> 00:24:48.160 align:middle line:84% The electron temperatures one and two just right 00:24:48.160 --> 00:24:49.660 align:middle line:90% at the interface. 00:24:49.660 --> 00:24:53.020 align:middle line:84% It turns out that when I go down a step, 00:24:53.020 --> 00:24:56.290 align:middle line:84% there is an amplification factor. 00:24:56.290 --> 00:24:59.050 align:middle line:84% This is the potential high [INAUDIBLE], 00:24:59.050 --> 00:25:00.500 align:middle line:90% and this is the mean free path. 00:25:00.500 --> 00:25:03.310 align:middle line:84% So normally, this is very small because the mean free path 00:25:03.310 --> 00:25:06.130 align:middle line:84% time [INAUDIBLE] just a temperature drop within 00:25:06.130 --> 00:25:07.000 align:middle line:90% [INAUDIBLE]. 00:25:07.000 --> 00:25:08.770 align:middle line:84% That's a very beautiful inversion. 00:25:08.770 --> 00:25:14.230 align:middle line:84% But what I thought interesting is if I have this structure, 00:25:14.230 --> 00:25:17.500 align:middle line:90% suddenly, an electron expand-- 00:25:17.500 --> 00:25:19.470 align:middle line:84% actually reach this temperature-- 00:25:19.470 --> 00:25:20.995 align:middle line:84% or create a large temperature drop. 00:25:20.995 --> 00:25:23.130 align:middle line:90% There's an amplification factor. 00:25:23.130 --> 00:25:26.340 align:middle line:84% I can amplify how much electron temperature 00:25:26.340 --> 00:25:28.320 align:middle line:90% drop at the interface. 00:25:28.320 --> 00:25:31.590 align:middle line:84% So that's what I call potential amplified. 00:25:31.590 --> 00:25:36.750 align:middle line:84% On the other hand, if I put this structure, same analysis. 00:25:36.750 --> 00:25:40.125 align:middle line:90% Didn't have that amplification. 00:25:40.125 --> 00:25:41.803 align:middle line:90% So I follow up. 00:25:41.803 --> 00:25:43.220 align:middle line:84% I said this [INAUDIBLE] statistic. 00:25:43.220 --> 00:25:46.260 align:middle line:84% I could have Fermi [INAUDIBLE] statistics 00:25:46.260 --> 00:25:48.060 align:middle line:90% and solve some equations. 00:25:48.060 --> 00:25:49.960 align:middle line:90% And this is the equation. 00:25:49.960 --> 00:25:52.920 align:middle line:84% This is the electron temperature at the interface. 00:25:52.920 --> 00:25:57.120 align:middle line:84% And you can see you can actually drive thermally-- 00:25:57.120 --> 00:26:00.930 align:middle line:84% with a temperature gradient, you can drive electron and proton 00:26:00.930 --> 00:26:02.970 align:middle line:90% out of equilibrium. 00:26:02.970 --> 00:26:06.210 align:middle line:84% And this is the electron temperature jump 00:26:06.210 --> 00:26:07.330 align:middle line:90% at the interface. 00:26:07.330 --> 00:26:09.930 align:middle line:84% But if you do the reverse structure, 00:26:09.930 --> 00:26:13.260 align:middle line:84% there's no discontinuity at the interface. 00:26:13.260 --> 00:26:18.480 align:middle line:84% So this way, since it can help the efficiency of cooling. 00:26:18.480 --> 00:26:20.740 align:middle line:90% But so far, I have not-- 00:26:20.740 --> 00:26:23.050 align:middle line:84% like I said, this is just a theory. 00:26:23.050 --> 00:26:23.990 align:middle line:90% No experiments. 00:26:23.990 --> 00:26:27.390 align:middle line:90% 00:26:27.390 --> 00:26:28.390 align:middle line:90% Now I'll move on. 00:26:28.390 --> 00:26:31.220 align:middle line:90% That's just an example. 00:26:31.220 --> 00:26:34.810 align:middle line:84% Sometimes it's interesting to think whether you can 00:26:34.810 --> 00:26:39.610 align:middle line:90% have new ideas in this field. 00:26:39.610 --> 00:26:42.940 align:middle line:84% Now, coming back-- now we talk about the theory-- 00:26:42.940 --> 00:26:45.250 align:middle line:90% the photovoltaic cells. 00:26:45.250 --> 00:26:48.010 align:middle line:84% And this is the PN junction I just showed, 00:26:48.010 --> 00:26:52.240 align:middle line:90% and the saturation current. 00:26:52.240 --> 00:26:55.090 align:middle line:84% Sometimes, people [INAUDIBLE] current [INAUDIBLE] 00:26:55.090 --> 00:26:57.610 align:middle line:90% detector current. 00:26:57.610 --> 00:27:03.460 align:middle line:84% And now, we have a photon comes in-- generates electron hole 00:27:03.460 --> 00:27:04.540 align:middle line:90% pairs. 00:27:04.540 --> 00:27:09.130 align:middle line:84% So if the photon energy is larger than the bandgap, 00:27:09.130 --> 00:27:10.930 align:middle line:90% the bandgap is always the same. 00:27:10.930 --> 00:27:13.000 align:middle line:90% Everywhere, bandgap is the same. 00:27:13.000 --> 00:27:16.180 align:middle line:84% So if the photon energy is larger than bandgap, 00:27:16.180 --> 00:27:20.620 align:middle line:84% you generate-- lift the one electron from the valence band 00:27:20.620 --> 00:27:22.240 align:middle line:90% to the conduction band. 00:27:22.240 --> 00:27:26.650 align:middle line:84% And this could happen anywhere in the device. 00:27:26.650 --> 00:27:29.260 align:middle line:84% That depends on how vertical [INAUDIBLE]. 00:27:29.260 --> 00:27:32.336 align:middle line:90% We'll talk about that later. 00:27:32.336 --> 00:27:34.000 align:middle line:90% So what I'm doing-- 00:27:34.000 --> 00:27:35.050 align:middle line:90% OK. 00:27:35.050 --> 00:27:39.430 align:middle line:84% My photon happened to be absorbed within the space charge 00:27:39.430 --> 00:27:43.210 align:middle line:84% region, then the electron-hole pair 00:27:43.210 --> 00:27:49.195 align:middle line:84% is generated within the space charge region [INAUDIBLE]. 00:27:49.195 --> 00:27:52.770 align:middle line:84% Now, what happens after this generation? 00:27:52.770 --> 00:27:57.790 align:middle line:84% Of course, they were-- they have a tendency to recombine. 00:27:57.790 --> 00:28:01.860 align:middle line:84% And if you do recombine, your photocell is [INAUDIBLE]. 00:28:01.860 --> 00:28:04.110 align:middle line:84% You don't want them to be combined. 00:28:04.110 --> 00:28:06.840 align:middle line:90% We want to separate them. 00:28:06.840 --> 00:28:10.440 align:middle line:84% And now, if this PN junction [INAUDIBLE] the space charge 00:28:10.440 --> 00:28:14.100 align:middle line:84% region, there's one [INAUDIBLE] because electron-- 00:28:14.100 --> 00:28:16.425 align:middle line:84% now there is an electric field inside. 00:28:16.425 --> 00:28:24.630 align:middle line:84% So the electron wants to go lower energy this way. 00:28:24.630 --> 00:28:28.380 align:middle line:84% The higher energy as a whole, positive is a higher energy. 00:28:28.380 --> 00:28:31.200 align:middle line:84% Electron [INAUDIBLE] higher energy. 00:28:31.200 --> 00:28:35.140 align:middle line:90% So hole go this way [INAUDIBLE]. 00:28:35.140 --> 00:28:36.390 align:middle line:90% OK? 00:28:36.390 --> 00:28:40.036 align:middle line:84% It just go down the potential field. 00:28:40.036 --> 00:28:42.930 align:middle line:84% Now, this is the potential [INAUDIBLE] hole. 00:28:42.930 --> 00:28:47.730 align:middle line:84% This is potential field for the electrons, going down here. 00:28:47.730 --> 00:28:49.980 align:middle line:90% That's natural. 00:28:49.980 --> 00:28:51.610 align:middle line:90% And this natural tendency. 00:28:51.610 --> 00:28:57.390 align:middle line:84% So now, you've got more electron on this side than, say-- 00:28:57.390 --> 00:29:00.030 align:middle line:84% normally at equilibrium, there's no voltage. 00:29:00.030 --> 00:29:02.500 align:middle line:84% Now you accumulate more electron on this side, 00:29:02.500 --> 00:29:05.320 align:middle line:84% accumulate more hole so you generate a voltage. 00:29:05.320 --> 00:29:08.890 align:middle line:90% 00:29:08.890 --> 00:29:11.790 align:middle line:84% And once you generate voltage, this voltage 00:29:11.790 --> 00:29:15.110 align:middle line:90% will drive your diode. 00:29:15.110 --> 00:29:18.330 align:middle line:84% Just like on this side, if you got a voltage difference, 00:29:18.330 --> 00:29:21.960 align:middle line:90% you have a diode current. 00:29:21.960 --> 00:29:26.430 align:middle line:84% So now, if I have light comes in, this is the current-- 00:29:26.430 --> 00:29:28.440 align:middle line:84% the electron-hole current generated 00:29:28.440 --> 00:29:30.510 align:middle line:90% due to absorption of photon. 00:29:30.510 --> 00:29:33.240 align:middle line:84% And now, once I do the voltage, this is the current-- moves 00:29:33.240 --> 00:29:34.460 align:middle line:90% forward and balance it. 00:29:34.460 --> 00:29:37.590 align:middle line:90% 00:29:37.590 --> 00:29:38.860 align:middle line:90% Now I'm going to have voltage. 00:29:38.860 --> 00:29:41.370 align:middle line:84% I drive the diode, and this is where 00:29:41.370 --> 00:29:44.550 align:middle line:84% I have an electron-hole pair generation-- the counter 00:29:44.550 --> 00:29:45.630 align:middle line:90% going this way. 00:29:45.630 --> 00:29:49.140 align:middle line:84% And of course, in a solar cell, I want to go this way. 00:29:49.140 --> 00:29:50.910 align:middle line:90% I don't want-- 00:29:50.910 --> 00:29:53.890 align:middle line:90% I want this less than this. 00:29:53.890 --> 00:29:57.980 align:middle line:90% 00:29:57.980 --> 00:30:02.100 align:middle line:84% So that's-- if you understand that, that's a solar cell. 00:30:02.100 --> 00:30:05.848 align:middle line:90% 00:30:05.848 --> 00:30:11.710 align:middle line:84% And so this one, I want a [INAUDIBLE] short circuit 00:30:11.710 --> 00:30:12.850 align:middle line:90% current. 00:30:12.850 --> 00:30:15.260 align:middle line:84% For the short circuit, the voltage difference is 0. 00:30:15.260 --> 00:30:17.960 align:middle line:90% You shorted this. 00:30:17.960 --> 00:30:21.970 align:middle line:84% There is no resistance, then this has to equal this. 00:30:21.970 --> 00:30:23.440 align:middle line:90% No voltage drop. 00:30:23.440 --> 00:30:26.080 align:middle line:90% That's the short current-- 00:30:26.080 --> 00:30:27.160 align:middle line:90% short circuit. 00:30:27.160 --> 00:30:30.460 align:middle line:84% Short circuit voltage [INAUDIBLE] for 0. 00:30:30.460 --> 00:30:32.200 align:middle line:90% That's your current. 00:30:32.200 --> 00:30:33.910 align:middle line:84% jl is your short circuit current. 00:30:33.910 --> 00:30:39.660 align:middle line:90% 00:30:39.660 --> 00:30:42.890 align:middle line:84% Now, I'm going to, with that-- starting from there, 00:30:42.890 --> 00:30:43.750 align:middle line:90% the rest is simple. 00:30:43.750 --> 00:30:44.610 align:middle line:90% It's math. 00:30:44.610 --> 00:30:47.865 align:middle line:84% I can see what's my open circuit voltage. 00:30:47.865 --> 00:30:50.520 align:middle line:84% And my open circuit voltage is just the same. 00:30:50.520 --> 00:30:51.870 align:middle line:90% There's 0 current. 00:30:51.870 --> 00:30:55.500 align:middle line:84% So if I set this side equal 0, I find out 00:30:55.500 --> 00:30:57.810 align:middle line:90% what's the open circuit voltage. 00:30:57.810 --> 00:31:00.330 align:middle line:90% That's the open circuit voltage. 00:31:00.330 --> 00:31:03.150 align:middle line:84% And you can further simplify because we know 00:31:03.150 --> 00:31:06.000 align:middle line:90% [INAUDIBLE] to dark current. 00:31:06.000 --> 00:31:10.910 align:middle line:84% Look at-- this is called dark current or saturation current. 00:31:10.910 --> 00:31:13.760 align:middle line:84% And then [INAUDIBLE] high open circuit voltage. 00:31:13.760 --> 00:31:15.255 align:middle line:90% What do you do? 00:31:15.255 --> 00:31:17.490 align:middle line:90% Smaller is better. 00:31:17.490 --> 00:31:18.790 align:middle line:90% [INAUDIBLE] smaller is better. 00:31:18.790 --> 00:31:20.720 align:middle line:90% You look at it-- 00:31:20.720 --> 00:31:23.530 align:middle line:84% smaller js, higher open circuit voltage. 00:31:23.530 --> 00:31:27.220 align:middle line:90% 00:31:27.220 --> 00:31:33.890 align:middle line:84% So it's js-- how recombination time is longer, js smaller. 00:31:33.890 --> 00:31:36.520 align:middle line:90% 00:31:36.520 --> 00:31:38.360 align:middle line:84% So I don't want the recombination. 00:31:38.360 --> 00:31:41.230 align:middle line:84% I want the long lifetime for the electron. 00:31:41.230 --> 00:31:44.510 align:middle line:84% Basically, I want long lifetime so I can stretch it out. 00:31:44.510 --> 00:31:48.355 align:middle line:90% 00:31:48.355 --> 00:31:50.710 align:middle line:84% And you put again the js [INAUDIBLE] 00:31:50.710 --> 00:31:54.190 align:middle line:84% related to [INAUDIBLE] diffusivity recombination time, 00:31:54.190 --> 00:31:56.560 align:middle line:90% and I'll write it down out here. 00:31:56.560 --> 00:32:00.460 align:middle line:84% Basically, you substitute in [INAUDIBLE] approximate 00:32:00.460 --> 00:32:01.540 align:middle line:90% [INAUDIBLE]. 00:32:01.540 --> 00:32:02.620 align:middle line:90% That's what you get. 00:32:02.620 --> 00:32:07.720 align:middle line:84% That's just electronics [INAUDIBLE] on this. 00:32:07.720 --> 00:32:11.990 align:middle line:84% Now, of course, open circuit is useless, 00:32:11.990 --> 00:32:14.030 align:middle line:90% because there's no power output. 00:32:14.030 --> 00:32:17.720 align:middle line:84% In real operation, we always have a load. 00:32:17.720 --> 00:32:20.045 align:middle line:90% We have current flowing. 00:32:20.045 --> 00:32:27.110 align:middle line:84% So when we have a current flowing, [INAUDIBLE] power. 00:32:27.110 --> 00:32:29.750 align:middle line:84% And so before I do that, I want to show you, 00:32:29.750 --> 00:32:32.700 align:middle line:84% if you draw this current voltage-- 00:32:32.700 --> 00:32:35.690 align:middle line:84% so this is your short circuit voltage for 0-- 00:32:35.690 --> 00:32:38.910 align:middle line:90% short circuit current negative. 00:32:38.910 --> 00:32:41.780 align:middle line:90% 00:32:41.780 --> 00:32:47.045 align:middle line:84% When current is 0, open circuit voltage. 00:32:47.045 --> 00:32:51.750 align:middle line:84% So it's more likely you're shifting [INAUDIBLE] down. 00:32:51.750 --> 00:32:56.390 align:middle line:84% If you don't have a light illumination, 00:32:56.390 --> 00:32:58.550 align:middle line:90% it's a diode equation, right? 00:32:58.550 --> 00:32:59.750 align:middle line:90% That's a diode equation. 00:32:59.750 --> 00:33:08.730 align:middle line:84% [INAUDIBLE] to the negative [INAUDIBLE]. 00:33:08.730 --> 00:33:13.610 align:middle line:90% 00:33:13.610 --> 00:33:17.960 align:middle line:84% And of course, you can look at this. 00:33:17.960 --> 00:33:21.080 align:middle line:84% The power output is the voltage times current. 00:33:21.080 --> 00:33:24.800 align:middle line:84% So somewhere, your voltage [INAUDIBLE]. 00:33:24.800 --> 00:33:26.025 align:middle line:90% That's [INAUDIBLE]. 00:33:26.025 --> 00:33:28.380 align:middle line:84% It depends on which one you operate in. 00:33:28.380 --> 00:33:34.010 align:middle line:84% You'll have maximum, and that's how you optimize. 00:33:34.010 --> 00:33:39.350 align:middle line:84% You change your load resistance so 00:33:39.350 --> 00:33:45.340 align:middle line:84% that you can control the current and the voltage by changing 00:33:45.340 --> 00:33:46.560 align:middle line:90% external load resistance. 00:33:46.560 --> 00:33:47.990 align:middle line:90% So that's a little magic. 00:33:47.990 --> 00:33:52.148 align:middle line:84% We do that for thermoelectrics, you do that for solar cells. 00:33:52.148 --> 00:33:52.690 align:middle line:90% All the same. 00:33:52.690 --> 00:33:54.440 align:middle line:90% You need to optimize. 00:33:54.440 --> 00:34:00.010 align:middle line:84% So some people just design their circuit to optimize. 00:34:00.010 --> 00:34:02.740 align:middle line:84% And the [INAUDIBLE] this way just 00:34:02.740 --> 00:34:05.120 align:middle line:90% reverse it just for convenience. 00:34:05.120 --> 00:34:07.990 align:middle line:84% So that would be another way to [INAUDIBLE]. 00:34:07.990 --> 00:34:11.770 align:middle line:84% That's the [INAUDIBLE] actually [INAUDIBLE]. 00:34:11.770 --> 00:34:16.659 align:middle line:90% 00:34:16.659 --> 00:34:20.365 align:middle line:84% Now, we can [INAUDIBLE] maximum power. 00:34:20.365 --> 00:34:23.050 align:middle line:84% When the power output is maximized, 00:34:23.050 --> 00:34:24.889 align:middle line:84% this is the current times voltage. 00:34:24.889 --> 00:34:27.505 align:middle line:90% That's the power. 00:34:27.505 --> 00:34:30.699 align:middle line:84% So you just take a derivative of your voltage 00:34:30.699 --> 00:34:32.139 align:middle line:90% and say when it's maximized. 00:34:32.139 --> 00:34:35.409 align:middle line:84% When you take derivative, physically, what I do 00:34:35.409 --> 00:34:39.610 align:middle line:84% is I'm changing my external load to find out 00:34:39.610 --> 00:34:45.070 align:middle line:84% where is the optimal voltage so that power is maximized. 00:34:45.070 --> 00:34:48.340 align:middle line:84% So that if you take this derivative, 00:34:48.340 --> 00:34:51.130 align:middle line:84% you'll find out where the optimal current, 00:34:51.130 --> 00:34:52.395 align:middle line:90% where is optimal voltage. 00:34:52.395 --> 00:34:57.478 align:middle line:90% 00:34:57.478 --> 00:35:02.200 align:middle line:84% And again, see, this is just a mathematic detail-- not 00:35:02.200 --> 00:35:03.370 align:middle line:90% important. 00:35:03.370 --> 00:35:05.500 align:middle line:84% But very often in the terminology, 00:35:05.500 --> 00:35:08.350 align:middle line:84% people say, what's your fill factor? 00:35:08.350 --> 00:35:10.225 align:middle line:90% What is that? 00:35:10.225 --> 00:35:12.130 align:middle line:84% So this is the rectangular-- this 00:35:12.130 --> 00:35:17.660 align:middle line:84% is where you optimize-- you run your optimum current voltage 00:35:17.660 --> 00:35:18.910 align:middle line:90% point. 00:35:18.910 --> 00:35:22.460 align:middle line:84% So that's the actual maximum power you have. 00:35:22.460 --> 00:35:28.240 align:middle line:84% And if you draw another rectangle, where you draw the-- 00:35:28.240 --> 00:35:35.570 align:middle line:84% this set up, the [INAUDIBLE] short circuit current and open 00:35:35.570 --> 00:35:37.650 align:middle line:84% circuit voltage, you draw another rectangle. 00:35:37.650 --> 00:35:40.946 align:middle line:90% 00:35:40.946 --> 00:35:45.650 align:middle line:84% This ratio of this extra power divided by the outer rectangle, 00:35:45.650 --> 00:35:47.360 align:middle line:90% that's your fill factor. 00:35:47.360 --> 00:35:50.328 align:middle line:84% And you want to fill as much as you can-- you can see-- 00:35:50.328 --> 00:35:51.370 align:middle line:90% to get the maximum power. 00:35:51.370 --> 00:35:59.280 align:middle line:90% 00:35:59.280 --> 00:36:01.225 align:middle line:84% Now let's do the source [INAUDIBLE]. 00:36:01.225 --> 00:36:06.640 align:middle line:84% Source, this is the electron-hole generation. 00:36:06.640 --> 00:36:10.680 align:middle line:84% So I have the deletion coming in. 00:36:10.680 --> 00:36:14.640 align:middle line:84% Only electrons-- the photons with energy larger than bandgap 00:36:14.640 --> 00:36:18.606 align:middle line:84% can generate electron-hole pairs. 00:36:18.606 --> 00:36:22.080 align:middle line:84% And also let's suppose this is blackbody. 00:36:22.080 --> 00:36:25.670 align:middle line:84% Let's treat the sun as a blackbody. 00:36:25.670 --> 00:36:28.440 align:middle line:84% So the solar radiation comes from the sun, 00:36:28.440 --> 00:36:30.840 align:middle line:84% and this is the fraction of solar radiation 00:36:30.840 --> 00:36:34.560 align:middle line:84% actually-- so you have to look at the distance, the fraction 00:36:34.560 --> 00:36:35.860 align:middle line:90% of solar radiation. 00:36:35.860 --> 00:36:41.910 align:middle line:84% Because the blackbody of the sun is 5,800 Kelvin. 00:36:41.910 --> 00:36:43.950 align:middle line:90% It's not a-- it's close object-- 00:36:43.950 --> 00:36:47.310 align:middle line:90% 5,800 Kelvin [INAUDIBLE]. 00:36:47.310 --> 00:36:48.300 align:middle line:90% It's very far. 00:36:48.300 --> 00:36:49.890 align:middle line:90% It's diluted. 00:36:49.890 --> 00:36:56.160 align:middle line:84% So that's this fraction in relation for the effect. 00:36:56.160 --> 00:37:00.180 align:middle line:84% But it's not the radiation intensity, 00:37:00.180 --> 00:37:06.210 align:middle line:84% it's the number of photons, because each photon generated 00:37:06.210 --> 00:37:09.720 align:middle line:90% one pair of electron-hole. 00:37:09.720 --> 00:37:12.840 align:middle line:90% So I'm looking at the energy-- 00:37:12.840 --> 00:37:14.220 align:middle line:90% because this is the energy. 00:37:14.220 --> 00:37:16.200 align:middle line:90% That's the energy per photon. 00:37:16.200 --> 00:37:17.200 align:middle line:90% So divide that. 00:37:17.200 --> 00:37:21.270 align:middle line:90% That's-- lots of photons coming. 00:37:21.270 --> 00:37:22.920 align:middle line:90% It's not a flux of energy. 00:37:22.920 --> 00:37:26.220 align:middle line:84% It's a flux of photon coming from a solar cell. 00:37:26.220 --> 00:37:28.890 align:middle line:90% And this is the reflectivity. 00:37:28.890 --> 00:37:32.430 align:middle line:84% I assume that the [INAUDIBLE] cell is absorbed and generate 00:37:32.430 --> 00:37:33.480 align:middle line:90% electron-hole pair. 00:37:33.480 --> 00:37:35.170 align:middle line:84% That's, of course, an idealization. 00:37:35.170 --> 00:37:36.990 align:middle line:84% You [INAUDIBLE] different losses. 00:37:36.990 --> 00:37:39.430 align:middle line:90% This is just best case. 00:37:39.430 --> 00:37:43.660 align:middle line:84% And of course, the maximum best is no reflection for that one. 00:37:43.660 --> 00:37:46.571 align:middle line:90% 00:37:46.571 --> 00:37:49.530 align:middle line:90% So this is source term. 00:37:49.530 --> 00:37:52.110 align:middle line:84% And the good thing is that the source term 00:37:52.110 --> 00:37:55.340 align:middle line:84% depends on only the temperature if you idealize. 00:37:55.340 --> 00:37:58.635 align:middle line:84% It depends only on the temperature of the source. 00:37:58.635 --> 00:38:01.875 align:middle line:90% 00:38:01.875 --> 00:38:05.490 align:middle line:84% And when I calculate efficiency, this 00:38:05.490 --> 00:38:09.570 align:middle line:84% is the energy of the photon coming towards the cell 00:38:09.570 --> 00:38:10.720 align:middle line:90% and this is the power. 00:38:10.720 --> 00:38:13.665 align:middle line:84% So of course, I want to maximize the efficiency. 00:38:13.665 --> 00:38:17.160 align:middle line:84% Everybody wants maximum efficiency. 00:38:17.160 --> 00:38:21.030 align:middle line:84% And then the question is, what's the maximum? 00:38:21.030 --> 00:38:21.940 align:middle line:90% How much is in there? 00:38:21.940 --> 00:38:24.742 align:middle line:90% 00:38:24.742 --> 00:38:26.010 align:middle line:90% Second law. 00:38:26.010 --> 00:38:29.640 align:middle line:84% First, the solar cell is limited by the second law column. 00:38:29.640 --> 00:38:32.550 align:middle line:90% You can always say [INAUDIBLE]. 00:38:32.550 --> 00:38:35.490 align:middle line:84% The maximum efficiency you can get [INAUDIBLE] 00:38:35.490 --> 00:38:38.130 align:middle line:90% between the sun and the Earth. 00:38:38.130 --> 00:38:40.972 align:middle line:90% So T [INAUDIBLE] is 300 Kelvin. 00:38:40.972 --> 00:38:46.017 align:middle line:90% T hat is the 5,800. 00:38:46.017 --> 00:38:47.350 align:middle line:90% So that's [INAUDIBLE] efficient. 00:38:47.350 --> 00:38:49.860 align:middle line:84% That's an upper limit [INAUDIBLE] upper limit. 00:38:49.860 --> 00:38:52.890 align:middle line:84% You can't get anything beyond that. 00:38:52.890 --> 00:38:53.520 align:middle line:90% Yes? 00:38:53.520 --> 00:38:55.200 align:middle line:84% AUDIENCE: So I'm just curious about-- so the way 00:38:55.200 --> 00:38:56.908 align:middle line:84% that the picture you have on that slide-- 00:38:56.908 --> 00:39:00.690 align:middle line:84% so it's actually instead on the p-type [INAUDIBLE]. 00:39:00.690 --> 00:39:07.260 align:middle line:84% And then don't we want it to hit the space charge region ideally? 00:39:07.260 --> 00:39:11.550 align:middle line:84% GANG CHEN: So the photon can be absorbed along 00:39:11.550 --> 00:39:14.010 align:middle line:90% the [INAUDIBLE] absorption. 00:39:14.010 --> 00:39:16.730 align:middle line:90% So some of them penetrate. 00:39:16.730 --> 00:39:20.760 align:middle line:84% Before the [INAUDIBLE] here in the space charge region. 00:39:20.760 --> 00:39:23.940 align:middle line:84% They have-- less probability goes on the other side. 00:39:23.940 --> 00:39:28.950 align:middle line:84% Electrons [INAUDIBLE] there are some diffusion happening here. 00:39:28.950 --> 00:39:33.180 align:middle line:84% Just to take a more [INAUDIBLE] chance of getting there is less. 00:39:33.180 --> 00:39:36.570 align:middle line:84% So that's why we want it that way and not [INAUDIBLE]. 00:39:36.570 --> 00:39:38.360 align:middle line:84% AUDIENCE: You see how that is [INAUDIBLE]? 00:39:38.360 --> 00:39:41.610 align:middle line:90% 00:39:41.610 --> 00:39:42.360 align:middle line:90% GANG CHEN: Uh-huh. 00:39:42.360 --> 00:39:43.950 align:middle line:90% AUDIENCE: So-- 00:39:43.950 --> 00:39:48.420 align:middle line:84% GANG CHEN: So you want to have a longer potential field 00:39:48.420 --> 00:39:50.960 align:middle line:84% to drive the-- actually [INAUDIBLE] field [INAUDIBLE]. 00:39:50.960 --> 00:39:53.460 align:middle line:84% AUDIENCE: No, I'm talking about this is a picture right here 00:39:53.460 --> 00:39:58.140 align:middle line:84% [INAUDIBLE] the really thin and flat [INAUDIBLE]. 00:39:58.140 --> 00:40:00.150 align:middle line:90% GANG CHEN: Really thin PN-- 00:40:00.150 --> 00:40:04.650 align:middle line:84% of course, the good thing is your electrode is very close. 00:40:04.650 --> 00:40:09.210 align:middle line:84% You can get to the electron, but your photon may not get through. 00:40:09.210 --> 00:40:11.490 align:middle line:90% So this is another [INAUDIBLE]. 00:40:11.490 --> 00:40:12.670 align:middle line:90% That's your backup. 00:40:12.670 --> 00:40:14.570 align:middle line:84% Your solar cells [INAUDIBLE] better. 00:40:14.570 --> 00:40:17.255 align:middle line:90% [INAUDIBLE] absorbed. 00:40:17.255 --> 00:40:19.870 align:middle line:84% Two, you want to get the charge to the electron 00:40:19.870 --> 00:40:21.340 align:middle line:90% before they recombine. 00:40:21.340 --> 00:40:25.120 align:middle line:84% One is thinner, the other one's thicker. 00:40:25.120 --> 00:40:28.400 align:middle line:84% This is essentially the thing people are trying to solve. 00:40:28.400 --> 00:40:31.690 align:middle line:90% 00:40:31.690 --> 00:40:35.655 align:middle line:84% But the question we want to ask is, what's the maximum possible? 00:40:35.655 --> 00:40:39.800 align:middle line:90% 00:40:39.800 --> 00:40:43.460 align:middle line:84% What's the maximum possible efficiency? 00:40:43.460 --> 00:40:46.500 align:middle line:84% And this one [INAUDIBLE] is two solar cells. 00:40:46.500 --> 00:40:52.152 align:middle line:84% The first [INAUDIBLE] by Shockley-- 00:40:52.152 --> 00:40:56.020 align:middle line:90% the Shockley-Queisser limit. 00:40:56.020 --> 00:41:00.220 align:middle line:84% And then that's the Shockley-Queisser limit 00:41:00.220 --> 00:41:02.110 align:middle line:90% I will discuss next. 00:41:02.110 --> 00:41:10.030 align:middle line:84% And of course, the real diode is hard to get what's maximum, 00:41:10.030 --> 00:41:13.570 align:middle line:84% because the saturation current depends on the field material 00:41:13.570 --> 00:41:18.040 align:middle line:84% recombination time, resistivity and electron-hole. 00:41:18.040 --> 00:41:20.890 align:middle line:90% It's hard to get your maximum. 00:41:20.890 --> 00:41:27.460 align:middle line:84% And the ingenuity of Shockley [INAUDIBLE] very, very smart 00:41:27.460 --> 00:41:28.330 align:middle line:90% guy. 00:41:28.330 --> 00:41:30.727 align:middle line:90% And [INAUDIBLE]. 00:41:30.727 --> 00:41:34.575 align:middle line:90% 00:41:34.575 --> 00:41:41.370 align:middle line:84% So this is what Shockley [INAUDIBLE]. 00:41:41.370 --> 00:41:45.660 align:middle line:84% So if you think about a PN junction, 00:41:45.660 --> 00:41:47.625 align:middle line:90% put it on the other side-- 00:41:47.625 --> 00:41:49.310 align:middle line:90% open circuit. 00:41:49.310 --> 00:41:52.500 align:middle line:84% Of course, the solar radiation will create electron-holes 00:41:52.500 --> 00:41:53.410 align:middle line:90% continuously. 00:41:53.410 --> 00:41:56.190 align:middle line:84% Every photon will generate electron-hole pairs. 00:41:56.190 --> 00:41:59.800 align:middle line:84% If there is no recombination, no defects will combine. 00:41:59.800 --> 00:42:01.110 align:middle line:90% What happens? 00:42:01.110 --> 00:42:03.755 align:middle line:84% You can't have an [INAUDIBLE] recombination process-- 00:42:03.755 --> 00:42:06.558 align:middle line:90% 00:42:06.558 --> 00:42:08.250 align:middle line:90% spontaneous recombination. 00:42:08.250 --> 00:42:09.630 align:middle line:90% [INAUDIBLE] 00:42:09.630 --> 00:42:12.450 align:middle line:90% So that's your maximum limit. 00:42:12.450 --> 00:42:19.430 align:middle line:84% No other recombination, only spontaneous recombination. 00:42:19.430 --> 00:42:23.840 align:middle line:84% But what's [INAUDIBLE] spontaneous recombination 00:42:23.840 --> 00:42:27.395 align:middle line:90% and coupling that up. 00:42:27.395 --> 00:42:29.470 align:middle line:90% And I'm really impressed. 00:42:29.470 --> 00:42:30.720 align:middle line:90% I read a few papers. 00:42:30.720 --> 00:42:33.500 align:middle line:84% This is where every time, I do this, 00:42:33.500 --> 00:42:36.350 align:middle line:84% I have a place where I don't feel comfortable to tell you. 00:42:36.350 --> 00:42:40.340 align:middle line:84% And I want to tell you because I want [INAUDIBLE]. 00:42:40.340 --> 00:42:43.160 align:middle line:84% And this is not just I read it yesterday. 00:42:43.160 --> 00:42:45.270 align:middle line:84% For many years, I tried to understand this. 00:42:45.270 --> 00:42:50.970 align:middle line:84% I still don't feel fully confident. 00:42:50.970 --> 00:42:54.150 align:middle line:84% This expression where it came up, 00:42:54.150 --> 00:42:56.640 align:middle line:84% and this is actually another Shockley paper. 00:42:56.640 --> 00:43:02.010 align:middle line:84% This is Henry's paper, and Henry had another paper [INAUDIBLE] 00:43:02.010 --> 00:43:04.350 align:middle line:84% and we all discussed different-- three different-- 00:43:04.350 --> 00:43:06.760 align:middle line:84% Shockley's paper, Henry's paper, Ross paper. 00:43:06.760 --> 00:43:10.990 align:middle line:84% Three different ways to write this expression. 00:43:10.990 --> 00:43:17.430 align:middle line:84% And so when a PN junction has a voltage [INAUDIBLE]-- 00:43:17.430 --> 00:43:22.170 align:middle line:84% and what's the-- so we see the quantum state. 00:43:22.170 --> 00:43:26.725 align:middle line:84% The Fermi-Dirac-- the Bose-Einstein statistics. 00:43:26.725 --> 00:43:30.300 align:middle line:84% But what [INAUDIBLE] the photons which 00:43:30.300 --> 00:43:33.165 align:middle line:90% obey Bose-Einstein statistics? 00:43:33.165 --> 00:43:37.860 align:middle line:84% In the PN junction, there is, it looks like, a Bose-Einstein, 00:43:37.860 --> 00:43:42.120 align:middle line:90% except that we add [INAUDIBLE]. 00:43:42.120 --> 00:43:44.670 align:middle line:84% [INAUDIBLE] is if you look at it, 00:43:44.670 --> 00:43:48.930 align:middle line:84% [INAUDIBLE] is really the Fermi level here in the [INAUDIBLE] 00:43:48.930 --> 00:43:50.160 align:middle line:90% in this region. 00:43:50.160 --> 00:43:53.510 align:middle line:90% That's the voltage [INAUDIBLE]. 00:43:53.510 --> 00:43:57.410 align:middle line:84% Expanding this term that I just feel I 00:43:57.410 --> 00:44:00.590 align:middle line:90% can comfortably derive it. 00:44:00.590 --> 00:44:03.680 align:middle line:84% You can go to expand so each paper 00:44:03.680 --> 00:44:06.210 align:middle line:90% you start from different place. 00:44:06.210 --> 00:44:09.380 align:middle line:90% And I encourage you to read. 00:44:09.380 --> 00:44:13.760 align:middle line:84% And again, if you have any idea to make my self comfortable 00:44:13.760 --> 00:44:18.290 align:middle line:84% [INAUDIBLE], because I, for many years, 00:44:18.290 --> 00:44:20.390 align:middle line:84% we were actually looking into this, 00:44:20.390 --> 00:44:24.620 align:middle line:84% seeing how we can really make [INAUDIBLE] derive some 00:44:24.620 --> 00:44:25.150 align:middle line:90% near-field-- 00:44:25.150 --> 00:44:28.930 align:middle line:84% [INAUDIBLE] near-field where it is surrounded [INAUDIBLE]. 00:44:28.930 --> 00:44:35.855 align:middle line:84% But see, so I can only just say take it from here [INAUDIBLE]. 00:44:35.855 --> 00:44:40.730 align:middle line:84% And the [INAUDIBLE] recombination 00:44:40.730 --> 00:44:48.530 align:middle line:84% in PN junction and [INAUDIBLE] you do later or [INAUDIBLE]. 00:44:48.530 --> 00:44:49.550 align:middle line:90% This is very important. 00:44:49.550 --> 00:44:53.005 align:middle line:84% This is very often where we will start. 00:44:53.005 --> 00:44:59.300 align:middle line:84% But I haven't [INAUDIBLE] thermodynamics very easy. 00:44:59.300 --> 00:45:05.440 align:middle line:84% So that's the [INAUDIBLE] how mechanically [INAUDIBLE]. 00:45:05.440 --> 00:45:10.530 align:middle line:90% 00:45:10.530 --> 00:45:15.810 align:middle line:84% Now, starting from here, you can derive 00:45:15.810 --> 00:45:18.210 align:middle line:90% what [INAUDIBLE] recombination. 00:45:18.210 --> 00:45:20.280 align:middle line:90% You have a voltage-- 00:45:20.280 --> 00:45:20.860 align:middle line:90% voltage. 00:45:20.860 --> 00:45:22.590 align:middle line:90% What's the recombination part? 00:45:22.590 --> 00:45:29.295 align:middle line:84% So you just integrate it again from the bandgap to any energy. 00:45:29.295 --> 00:45:31.230 align:middle line:84% And if you do [INAUDIBLE] statistics, 00:45:31.230 --> 00:45:35.340 align:middle line:84% you just elected one factor, one. 00:45:35.340 --> 00:45:37.380 align:middle line:90% And this is the-- 00:45:37.380 --> 00:45:40.110 align:middle line:84% I believe I didn't write the equation right-- 00:45:40.110 --> 00:45:42.990 align:middle line:90% copied from [INAUDIBLE] paper. 00:45:42.990 --> 00:45:45.160 align:middle line:84% And this is solid angle integration, 00:45:45.160 --> 00:45:49.780 align:middle line:84% so that's essentially the photon density of states 00:45:49.780 --> 00:45:52.700 align:middle line:90% inside the [INAUDIBLE]. 00:45:52.700 --> 00:45:56.110 align:middle line:84% So the Fermi-Dirac Bose-Einstein statistics 00:45:56.110 --> 00:45:58.765 align:middle line:90% integrated [INAUDIBLE]. 00:45:58.765 --> 00:46:03.330 align:middle line:84% So this is the spontaneous emission due to the current 00:46:03.330 --> 00:46:04.120 align:middle line:90% [INAUDIBLE]. 00:46:04.120 --> 00:46:08.800 align:middle line:90% 00:46:08.800 --> 00:46:12.550 align:middle line:84% And everything is very similar to what we had before, 00:46:12.550 --> 00:46:16.300 align:middle line:84% except that [INAUDIBLE] is no longer due to-- if you 00:46:16.300 --> 00:46:22.420 align:middle line:84% look at the diode right here, this is the a for the diode. 00:46:22.420 --> 00:46:26.230 align:middle line:84% That's determined by actual material recombination. 00:46:26.230 --> 00:46:30.640 align:middle line:84% And here, the ingenuity is now this 00:46:30.640 --> 00:46:35.294 align:middle line:84% is the [INAUDIBLE] dielectric constant refractive index 00:46:35.294 --> 00:46:36.530 align:middle line:90% bandgap of the material. 00:46:36.530 --> 00:46:39.040 align:middle line:84% So this is the maximum you can have-- 00:46:39.040 --> 00:46:40.630 align:middle line:90% the best you can have. 00:46:40.630 --> 00:46:43.760 align:middle line:84% [INAUDIBLE] really do recombination. 00:46:43.760 --> 00:46:47.290 align:middle line:90% no other defects [INAUDIBLE]. 00:46:47.290 --> 00:46:50.850 align:middle line:84% So because of that, it's free of any material. 00:46:50.850 --> 00:46:54.200 align:middle line:84% So now, you can derive the [INAUDIBLE]. 00:46:54.200 --> 00:46:57.590 align:middle line:84% But it's still one material that's [INAUDIBLE]. 00:46:57.590 --> 00:47:01.970 align:middle line:84% This is the only other material you can have. 00:47:01.970 --> 00:47:04.400 align:middle line:84% But while I talk, you can also say, 00:47:04.400 --> 00:47:06.170 align:middle line:84% ah, how about the photonic crystal-- 00:47:06.170 --> 00:47:08.310 align:middle line:90% inside the photonic crystal? 00:47:08.310 --> 00:47:14.270 align:middle line:84% In this chapter, we checked [INAUDIBLE] spontaneous form. 00:47:14.270 --> 00:47:17.360 align:middle line:84% And maybe you've got a different [INAUDIBLE]. 00:47:17.360 --> 00:47:21.146 align:middle line:90% Maybe [INAUDIBLE]. 00:47:21.146 --> 00:47:25.685 align:middle line:84% So now, I have translated that if you look at the term, 00:47:25.685 --> 00:47:29.285 align:middle line:84% this is the current term, the diode equation. 00:47:29.285 --> 00:47:32.240 align:middle line:84% One is the diode current, the other 00:47:32.240 --> 00:47:34.460 align:middle line:90% is the photon generated current. 00:47:34.460 --> 00:47:36.150 align:middle line:84% This is the photon generated current. 00:47:36.150 --> 00:47:37.910 align:middle line:90% In this case, this is 1. 00:47:37.910 --> 00:47:40.640 align:middle line:90% So I put in the best scenario. 00:47:40.640 --> 00:47:43.910 align:middle line:84% And the [INAUDIBLE] is the independent material except 00:47:43.910 --> 00:47:44.840 align:middle line:90% the bandgap-- 00:47:44.840 --> 00:47:45.840 align:middle line:90% only the bandgap. 00:47:45.840 --> 00:47:49.520 align:middle line:84% So that's where Shockley could derive 00:47:49.520 --> 00:47:51.770 align:middle line:90% what's the maximum efficiency. 00:47:51.770 --> 00:47:56.240 align:middle line:84% And this is the curve not from Shockley, 00:47:56.240 --> 00:47:58.340 align:middle line:90% but from [INAUDIBLE] paper. 00:47:58.340 --> 00:48:03.560 align:middle line:84% And using this [INAUDIBLE] recombination current, 00:48:03.560 --> 00:48:05.580 align:middle line:84% photon generated electron-hole pair, 00:48:05.580 --> 00:48:12.260 align:middle line:84% so this is [INAUDIBLE] photon generated current. 00:48:12.260 --> 00:48:14.960 align:middle line:84% And you do the rest [INAUDIBLE] we did before. 00:48:14.960 --> 00:48:17.630 align:middle line:84% We can find out what's the maximum power 00:48:17.630 --> 00:48:20.100 align:middle line:84% and what's the maximum efficiency. 00:48:20.100 --> 00:48:24.750 align:middle line:84% And this is what [INAUDIBLE] calculated. 00:48:24.750 --> 00:48:28.530 align:middle line:90% And here, c is a concentration. 00:48:28.530 --> 00:48:29.820 align:middle line:90% What concentration? 00:48:29.820 --> 00:48:31.170 align:middle line:90% [INAUDIBLE] concentration. 00:48:31.170 --> 00:48:34.065 align:middle line:90% So what does is change jl. 00:48:34.065 --> 00:48:36.750 align:middle line:84% If you concentrate more, you generate 00:48:36.750 --> 00:48:39.850 align:middle line:84% more electron-hole pairs per [INAUDIBLE]. 00:48:39.850 --> 00:48:44.340 align:middle line:84% So if you concentrate, you get this larger-- 00:48:44.340 --> 00:48:46.890 align:middle line:84% you get a larger open circuit voltage. 00:48:46.890 --> 00:48:52.690 align:middle line:84% So concentration actually can help you increase efficiency. 00:48:52.690 --> 00:48:57.700 align:middle line:84% So not too much increase, but reasonable increase. 00:48:57.700 --> 00:48:59.200 align:middle line:84% And what's interesting when you look 00:48:59.200 --> 00:49:03.060 align:middle line:84% at it, under one [INAUDIBLE] condition, where 00:49:03.060 --> 00:49:06.160 align:middle line:90% is the optimum bandgap? 00:49:06.160 --> 00:49:09.450 align:middle line:90% About 1.5. 00:49:09.450 --> 00:49:12.880 align:middle line:84% If you look at the material, where is [INAUDIBLE] material? 00:49:12.880 --> 00:49:14.960 align:middle line:90% 1.5. 00:49:14.960 --> 00:49:18.910 align:middle line:84% A lot of MOSFET is actually very good [INAUDIBLE]. 00:49:18.910 --> 00:49:21.640 align:middle line:90% 00:49:21.640 --> 00:49:23.290 align:middle line:90% 1.1. 00:49:23.290 --> 00:49:25.510 align:middle line:90% Not too far from the optimum. 00:49:25.510 --> 00:49:26.643 align:middle line:90% So this is about 31%. 00:49:26.643 --> 00:49:30.200 align:middle line:90% 00:49:30.200 --> 00:49:34.990 align:middle line:84% So that tells you, if you build a [INAUDIBLE] cell, 00:49:34.990 --> 00:49:39.480 align:middle line:84% this is about the best you can do in terms of efficiency. 00:49:39.480 --> 00:49:42.980 align:middle line:90% [INAUDIBLE] 00:49:42.980 --> 00:49:48.380 align:middle line:84% AUDIENCE: [INAUDIBLE] energetically focused 00:49:48.380 --> 00:49:50.192 align:middle line:90% compared to the [INAUDIBLE]. 00:49:50.192 --> 00:49:52.900 align:middle line:90% 00:49:52.900 --> 00:49:55.000 align:middle line:90% GANG CHEN: You mean the-- 00:49:55.000 --> 00:49:59.050 align:middle line:84% how many photon [INAUDIBLE] compared to bandgap or? 00:49:59.050 --> 00:50:01.933 align:middle line:84% AUDIENCE: I mean, you have [INAUDIBLE]. 00:50:01.933 --> 00:50:05.040 align:middle line:90% 00:50:05.040 --> 00:50:08.350 align:middle line:84% What happens, for example, [INAUDIBLE]. 00:50:08.350 --> 00:50:15.315 align:middle line:84% They are absorbed [INAUDIBLE] energy as the [INAUDIBLE] 1,100. 00:50:15.315 --> 00:50:18.150 align:middle line:84% So the [INAUDIBLE] is [INAUDIBLE]. 00:50:18.150 --> 00:50:19.230 align:middle line:90% GANG CHEN: Exactly. 00:50:19.230 --> 00:50:22.620 align:middle line:90% So this is a blackbody curve. 00:50:22.620 --> 00:50:25.890 align:middle line:84% This is your bandgap somewhere here. 00:50:25.890 --> 00:50:30.065 align:middle line:84% And really, what's usable [INAUDIBLE] energy 00:50:30.065 --> 00:50:32.850 align:middle line:84% is because this is above the bandgap energy [INAUDIBLE]. 00:50:32.850 --> 00:50:34.650 align:middle line:90% So if I look at-- 00:50:34.650 --> 00:50:36.660 align:middle line:90% this is the fine structure. 00:50:36.660 --> 00:50:40.350 align:middle line:84% If I have photon generated electron-hole pair this way, 00:50:40.350 --> 00:50:44.790 align:middle line:84% this energy electron relaxes [INAUDIBLE] very fast. 00:50:44.790 --> 00:50:47.265 align:middle line:90% Hole relaxes here very fast. 00:50:47.265 --> 00:50:49.695 align:middle line:90% So this is a key. 00:50:49.695 --> 00:50:53.790 align:middle line:84% And in fact, the people have been theorizing 00:50:53.790 --> 00:50:55.620 align:middle line:90% how the electron [INAUDIBLE]. 00:50:55.620 --> 00:50:59.850 align:middle line:84% If you can extract this electron-hole with this kind 00:50:59.850 --> 00:51:02.670 align:middle line:90% of energy, that's [INAUDIBLE]. 00:51:02.670 --> 00:51:04.930 align:middle line:84% Question of how you get it out before you relax. 00:51:04.930 --> 00:51:07.350 align:middle line:90% This is about picosecond. 00:51:07.350 --> 00:51:09.660 align:middle line:90% So picosecond is very fast. 00:51:09.660 --> 00:51:13.560 align:middle line:84% If you times 10 to the minus 12 times diffusivity, 00:51:13.560 --> 00:51:15.660 align:middle line:84% [INAUDIBLE] travel nanometers of-- 00:51:15.660 --> 00:51:17.160 align:middle line:84% [INAUDIBLE] nanometers [INAUDIBLE] 00:51:17.160 --> 00:51:19.470 align:middle line:90% have no way to get it out. 00:51:19.470 --> 00:51:22.350 align:middle line:84% And let's say if you have a good idea, 00:51:22.350 --> 00:51:25.460 align:middle line:84% that's something people have been thinking since the '70s. 00:51:25.460 --> 00:51:32.590 align:middle line:90% 00:51:32.590 --> 00:51:37.230 align:middle line:84% So that's a single junction maximum you can get. 00:51:37.230 --> 00:51:45.300 align:middle line:84% And exactly on this issue, how about we do multi-junction? 00:51:45.300 --> 00:51:51.180 align:middle line:84% One cell-- so all the junction shown here-- 00:51:51.180 --> 00:51:55.320 align:middle line:84% the first one absorbs blue, so larger bandgap-- 00:51:55.320 --> 00:51:58.050 align:middle line:90% the topmost layer. 00:51:58.050 --> 00:52:00.120 align:middle line:84% And the lower-- longer wavelengths, 00:52:00.120 --> 00:52:04.890 align:middle line:84% lower energy [INAUDIBLE], so gets absorbed in second layer 00:52:04.890 --> 00:52:06.780 align:middle line:90% and get absorbed third. 00:52:06.780 --> 00:52:08.400 align:middle line:90% [INAUDIBLE] idea. 00:52:08.400 --> 00:52:12.300 align:middle line:84% And of course, you can see with this stacking, 00:52:12.300 --> 00:52:16.738 align:middle line:90% each one generates a current. 00:52:16.738 --> 00:52:18.780 align:middle line:84% If you look at this, the key is that your current 00:52:18.780 --> 00:52:22.140 align:middle line:84% has to be continuous because you put [INAUDIBLE] on top, 00:52:22.140 --> 00:52:23.520 align:middle line:90% [INAUDIBLE] bottom. 00:52:23.520 --> 00:52:26.790 align:middle line:84% So this still has to generate the same amount of current 00:52:26.790 --> 00:52:28.280 align:middle line:90% as a design [INAUDIBLE]. 00:52:28.280 --> 00:52:32.484 align:middle line:90% 00:52:32.484 --> 00:52:34.860 align:middle line:90% But people have done that. 00:52:34.860 --> 00:52:39.540 align:middle line:84% It's a material nightmare and the design nightmare. 00:52:39.540 --> 00:52:42.420 align:middle line:84% Design in the sense you have to get a current match. 00:52:42.420 --> 00:52:45.150 align:middle line:90% So the electron goes-- 00:52:45.150 --> 00:52:47.130 align:middle line:84% come from the top layer, has to go 00:52:47.130 --> 00:52:49.920 align:middle line:84% to the hole of the next layer, or hole of the top layer 00:52:49.920 --> 00:52:54.030 align:middle line:84% has to merge with the electron-hole of the next layer. 00:52:54.030 --> 00:52:55.710 align:middle line:90% The current is continuous. 00:52:55.710 --> 00:52:58.560 align:middle line:84% And you have to group different materials, 00:52:58.560 --> 00:53:02.820 align:middle line:84% because the recombination is the killer. 00:53:02.820 --> 00:53:05.040 align:middle line:84% So that's why people do molecular beam 00:53:05.040 --> 00:53:06.660 align:middle line:90% epitaxy and lattice-match. 00:53:06.660 --> 00:53:10.140 align:middle line:84% And you generate defects, you generate dislocations. 00:53:10.140 --> 00:53:13.320 align:middle line:84% These were just the centers for the recombination. 00:53:13.320 --> 00:53:19.650 align:middle line:84% And then you kill the electrons before you can take it out. 00:53:19.650 --> 00:53:22.360 align:middle line:84% So to group those material, people 00:53:22.360 --> 00:53:26.460 align:middle line:84% have done molecular beam epitaxy like spectral diode-- 00:53:26.460 --> 00:53:28.385 align:middle line:90% spectral diode lab. 00:53:28.385 --> 00:53:29.760 align:middle line:84% And they were doing laser before. 00:53:29.760 --> 00:53:34.270 align:middle line:84% Now, they're doing probably 40%, 41% photocells-- 00:53:34.270 --> 00:53:38.890 align:middle line:84% [INAUDIBLE] cells, three junctions, and that rule 00:53:38.890 --> 00:53:41.890 align:middle line:90% match this. 00:53:41.890 --> 00:53:43.320 align:middle line:90% And so this is-- 00:53:43.320 --> 00:53:46.540 align:middle line:84% I got one graph from the internet. 00:53:46.540 --> 00:53:50.250 align:middle line:84% I think it gives some idea [INAUDIBLE] in real world, 00:53:50.250 --> 00:53:55.570 align:middle line:90% about 15% to 18% [INAUDIBLE]. 00:53:55.570 --> 00:54:00.100 align:middle line:90% And then [INAUDIBLE] 30%. 00:54:00.100 --> 00:54:03.040 align:middle line:84% And now, this is a triple junction indium gallium 00:54:03.040 --> 00:54:05.800 align:middle line:84% phosphide, gallium arsenide, germanium bandgap. 00:54:05.800 --> 00:54:10.232 align:middle line:90% 1.8, 1.4, 1.7 [INAUDIBLE]. 00:54:10.232 --> 00:54:12.190 align:middle line:84% You have to look at what are the materials that 00:54:12.190 --> 00:54:14.680 align:middle line:84% match your current and that you can grow. 00:54:14.680 --> 00:54:18.570 align:middle line:84% So you have to look at both material and bandgap device. 00:54:18.570 --> 00:54:21.610 align:middle line:90% 00:54:21.610 --> 00:54:26.695 align:middle line:84% And if you [INAUDIBLE] much [INAUDIBLE] infinitely. 00:54:26.695 --> 00:54:34.840 align:middle line:84% So this is a paper that shows the 72% [INAUDIBLE] 00:54:34.840 --> 00:54:39.235 align:middle line:84% multi-spectral [INAUDIBLE] 36-state. 00:54:39.235 --> 00:54:45.260 align:middle line:84% And if you have an infinite states, the theory is about 86%. 00:54:45.260 --> 00:54:48.490 align:middle line:84% So that's the best you can do [INAUDIBLE]. 00:54:48.490 --> 00:54:53.320 align:middle line:90% 00:54:53.320 --> 00:54:59.410 align:middle line:84% That's still lower than [INAUDIBLE] 300 and 1,500. 00:54:59.410 --> 00:55:00.760 align:middle line:90% That's about the line. 00:55:00.760 --> 00:55:09.180 align:middle line:84% If you do 1 minus 300 divide by 1,500. 00:55:09.180 --> 00:55:13.260 align:middle line:84% So that's the-- what people can do. 00:55:13.260 --> 00:55:15.240 align:middle line:84% Because in theory, it says now you 00:55:15.240 --> 00:55:17.560 align:middle line:90% look at [INAUDIBLE] in the lab. 00:55:17.560 --> 00:55:18.800 align:middle line:90% This is the best in the lab. 00:55:18.800 --> 00:55:21.600 align:middle line:90% It's not the best deployed. 00:55:21.600 --> 00:55:26.280 align:middle line:84% Silicon [INAUDIBLE] crystalline silicon. 00:55:26.280 --> 00:55:28.590 align:middle line:90% Crystalline silicon down here. 00:55:28.590 --> 00:55:32.580 align:middle line:90% This is a function [INAUDIBLE]. 00:55:32.580 --> 00:55:36.480 align:middle line:90% The best is about 95%, 96%. 00:55:36.480 --> 00:55:45.360 align:middle line:84% Remember, [INAUDIBLE] probably the biggest [INAUDIBLE] lab. 00:55:45.360 --> 00:55:47.010 align:middle line:90% They told me about [INAUDIBLE]. 00:55:47.010 --> 00:55:50.580 align:middle line:90% 00:55:50.580 --> 00:55:55.140 align:middle line:84% Remember, Shockley-Queisser gives you 30%, 31%. 00:55:55.140 --> 00:56:00.480 align:middle line:84% So [INAUDIBLE] really well compared to [INAUDIBLE]. 00:56:00.480 --> 00:56:02.670 align:middle line:90% Amazing. 00:56:02.670 --> 00:56:09.380 align:middle line:84% And also, that's why your potential is also limited. 00:56:09.380 --> 00:56:11.422 align:middle line:90% That's the best you can do. 00:56:11.422 --> 00:56:14.980 align:middle line:84% But the fact that you can really do that at low cost, 00:56:14.980 --> 00:56:16.630 align:middle line:90% I think your problem's solved. 00:56:16.630 --> 00:56:25.660 align:middle line:84% [INAUDIBLE] you can always work in [INAUDIBLE] is only 10% 00:56:25.660 --> 00:56:27.150 align:middle line:90% [INAUDIBLE]. 00:56:27.150 --> 00:56:28.700 align:middle line:90% Of course, [INAUDIBLE]. 00:56:28.700 --> 00:56:31.870 align:middle line:90% 00:56:31.870 --> 00:56:39.700 align:middle line:84% And so the best multijunction cell [INAUDIBLE] lab. 00:56:39.700 --> 00:56:41.140 align:middle line:84% Here, it's less than [INAUDIBLE]. 00:56:41.140 --> 00:56:43.673 align:middle line:84% I think that now, the best that I heard is about 41%. 00:56:43.673 --> 00:56:48.950 align:middle line:90% 00:56:48.950 --> 00:56:53.780 align:middle line:90% And this is a multicrystalline. 00:56:53.780 --> 00:56:56.780 align:middle line:84% This is a single crystalline, multicrystalline. 00:56:56.780 --> 00:57:02.498 align:middle line:84% And let's look at amorphous silicon [INAUDIBLE]. 00:57:02.498 --> 00:57:06.719 align:middle line:90% 00:57:06.719 --> 00:57:11.080 align:middle line:84% It's-- well, the real world, it's about 6% to 8% [INAUDIBLE]. 00:57:11.080 --> 00:57:14.220 align:middle line:84% If I say the lab, it's about 12%, but that's [INAUDIBLE]. 00:57:14.220 --> 00:57:18.760 align:middle line:90% 00:57:18.760 --> 00:57:24.430 align:middle line:84% And this is a, say, copper indium. 00:57:24.430 --> 00:57:28.780 align:middle line:84% [INAUDIBLE] copper indium [INAUDIBLE]. 00:57:28.780 --> 00:57:34.300 align:middle line:90% And so the best is about 19%. 00:57:34.300 --> 00:57:41.260 align:middle line:84% And [INAUDIBLE] telluride is [INAUDIBLE] here and here. 00:57:41.260 --> 00:57:44.200 align:middle line:84% And because, as I said, it's backwards because most people 00:57:44.200 --> 00:57:46.600 align:middle line:84% talk about 64, and now [INAUDIBLE] 00:57:46.600 --> 00:57:50.355 align:middle line:84% is probably the best most profitable company 00:57:50.355 --> 00:57:53.400 align:middle line:90% in the [INAUDIBLE]. 00:57:53.400 --> 00:57:56.050 align:middle line:90% And their CEO I think is about-- 00:57:56.050 --> 00:57:59.246 align:middle line:84% [INAUDIBLE] from lab to real production [INAUDIBLE]. 00:57:59.246 --> 00:58:03.620 align:middle line:90% 00:58:03.620 --> 00:58:09.930 align:middle line:84% And then this one is the polymer, I think, 00:58:09.930 --> 00:58:12.400 align:middle line:90% with the organic cells. 00:58:12.400 --> 00:58:17.860 align:middle line:84% And then there's a [INAUDIBLE] cell [INAUDIBLE]. 00:58:17.860 --> 00:58:23.060 align:middle line:90% 00:58:23.060 --> 00:58:26.110 align:middle line:84% So this one gives you [INAUDIBLE]. 00:58:26.110 --> 00:58:31.030 align:middle line:84% Not the most up-to-date in years, but pretty close. 00:58:31.030 --> 00:58:34.330 align:middle line:90% What are the problems? 00:58:34.330 --> 00:58:38.550 align:middle line:84% I mentioned the one for solar cell-- 00:58:38.550 --> 00:58:40.420 align:middle line:84% the number one is [INAUDIBLE] device. 00:58:40.420 --> 00:58:42.460 align:middle line:90% That's [INAUDIBLE] cell. 00:58:42.460 --> 00:58:46.450 align:middle line:84% And our device, very sensitive to recombination. 00:58:46.450 --> 00:58:53.410 align:middle line:84% So what are the reasons for recombination? 00:58:53.410 --> 00:59:00.250 align:middle line:90% Those [INAUDIBLE] bandgap. 00:59:00.250 --> 00:59:01.390 align:middle line:90% Deep level. 00:59:01.390 --> 00:59:04.390 align:middle line:84% Those are recombination [INAUDIBLE]. 00:59:04.390 --> 00:59:05.380 align:middle line:90% [INAUDIBLE] 00:59:05.380 --> 00:59:08.410 align:middle line:90% 00:59:08.410 --> 00:59:14.010 align:middle line:84% And [INAUDIBLE] where you've got a lot of [INAUDIBLE]. 00:59:14.010 --> 00:59:19.510 align:middle line:84% So those are regions where recombination happens. 00:59:19.510 --> 00:59:23.110 align:middle line:84% And that's why people purify silicon. 00:59:23.110 --> 00:59:25.720 align:middle line:84% Crystalline silicon, for example, [INAUDIBLE] 00:59:25.720 --> 00:59:27.430 align:middle line:90% large [INAUDIBLE]. 00:59:27.430 --> 00:59:31.250 align:middle line:90% It's about $2 per gram-- 00:59:31.250 --> 00:59:34.040 align:middle line:90% per kilogram [INAUDIBLE]. 00:59:34.040 --> 00:59:35.630 align:middle line:84% If you want to make a solar cell, 00:59:35.630 --> 00:59:39.710 align:middle line:84% you can recrystallize it, purify it and [INAUDIBLE] 00:59:39.710 --> 00:59:42.343 align:middle line:90% $40 per kilogram. 00:59:42.343 --> 00:59:44.010 align:middle line:84% And then, of course, during the process, 00:59:44.010 --> 00:59:49.590 align:middle line:84% you introduce a lot [INAUDIBLE] heat up, melting [INAUDIBLE]. 00:59:49.590 --> 00:59:54.740 align:middle line:84% And in fact, last year, before the market crashed, 00:59:54.740 --> 00:59:56.750 align:middle line:90% everybody was on silicon. 00:59:56.750 --> 01:00:01.915 align:middle line:84% So the silicon was short, and that got $240 per kilogram. 01:00:01.915 --> 01:00:07.190 align:middle line:90% 01:00:07.190 --> 01:00:08.530 align:middle line:90% So that's one. 01:00:08.530 --> 01:00:10.980 align:middle line:84% That's the same [INAUDIBLE] device Achilles heel. 01:00:10.980 --> 01:00:14.760 align:middle line:84% And in that sense, I can argue thermoelectrics is good 01:00:14.760 --> 01:00:16.580 align:middle line:84% because it's a majority carrying device. 01:00:16.580 --> 01:00:20.420 align:middle line:84% It doesn't have severe recombination. 01:00:20.420 --> 01:00:24.480 align:middle line:84% And there are other problems with this [INAUDIBLE]. 01:00:24.480 --> 01:00:26.070 align:middle line:90% In fact, this, to me-- 01:00:26.070 --> 01:00:29.940 align:middle line:90% I say, OK, [INAUDIBLE] absorb. 01:00:29.940 --> 01:00:31.930 align:middle line:84% The second step is to charge it where it goes. 01:00:31.930 --> 01:00:34.110 align:middle line:84% So this is really the first step. 01:00:34.110 --> 01:00:37.200 align:middle line:84% And look at where the whole photon goes. 01:00:37.200 --> 01:00:40.410 align:middle line:84% Of course, the relay interface and reflection [INAUDIBLE] 01:00:40.410 --> 01:00:42.720 align:middle line:90% reflection [INAUDIBLE]. 01:00:42.720 --> 01:00:45.930 align:middle line:84% [INAUDIBLE] coating and surface treatment. 01:00:45.930 --> 01:00:48.390 align:middle line:84% And once you get an input material, 01:00:48.390 --> 01:00:52.880 align:middle line:84% you don't have almost no other ways to control this profile 01:00:52.880 --> 01:00:53.380 align:middle line:90% anymore. 01:00:53.380 --> 01:00:55.770 align:middle line:84% It's material-intrinsic absorption. 01:00:55.770 --> 01:00:59.040 align:middle line:84% So this is the absorption coefficient and-- 01:00:59.040 --> 01:01:02.820 align:middle line:84% because the intensity decays exponentially. 01:01:02.820 --> 01:01:04.380 align:middle line:90% So now, you think about it. 01:01:04.380 --> 01:01:07.380 align:middle line:84% In your [INAUDIBLE] photon generated electron-hole pair, 01:01:07.380 --> 01:01:09.630 align:middle line:90% let's say, in this region. 01:01:09.630 --> 01:01:13.230 align:middle line:84% And you want a space charge in this region 01:01:13.230 --> 01:01:16.050 align:middle line:84% to overlap it so that you can separate them. 01:01:16.050 --> 01:01:17.980 align:middle line:84% You can use your field to drive the electron 01:01:17.980 --> 01:01:22.000 align:middle line:84% from one side of the hole to the other side of the hole. 01:01:22.000 --> 01:01:25.270 align:middle line:84% So the depth of this is 1 over r-- 01:01:25.270 --> 01:01:27.640 align:middle line:90% gives you exponential minus 1. 01:01:27.640 --> 01:01:30.850 align:middle line:84% So that's called the penetration depth. 01:01:30.850 --> 01:01:35.560 align:middle line:84% So this is where you look at the material, what I should choose. 01:01:35.560 --> 01:01:38.110 align:middle line:90% The absorption coefficient. 01:01:38.110 --> 01:01:40.450 align:middle line:84% Absorption happens above the bandgap. 01:01:40.450 --> 01:01:43.060 align:middle line:90% So bandgap is the first step. 01:01:43.060 --> 01:01:45.760 align:middle line:84% And then you look at the material. 01:01:45.760 --> 01:01:49.110 align:middle line:84% You say, oh, silicon, we love so much because that's 01:01:49.110 --> 01:01:50.980 align:middle line:90% the [INAUDIBLE] of silicon. 01:01:50.980 --> 01:01:54.130 align:middle line:90% Why silicon is a problem? 01:01:54.130 --> 01:01:56.350 align:middle line:84% Because it's really just-- firstly, 01:01:56.350 --> 01:01:58.660 align:middle line:90% because this absorption. 01:01:58.660 --> 01:02:02.620 align:middle line:84% You see here, absorption-- this is the photon energy. 01:02:02.620 --> 01:02:06.400 align:middle line:84% Bandgap is about 1.1 electronvolts. 01:02:06.400 --> 01:02:10.630 align:middle line:84% It's a very slow curve, and it's 1 over [INAUDIBLE]. 01:02:10.630 --> 01:02:11.890 align:middle line:90% So this is centimeter. 01:02:11.890 --> 01:02:14.890 align:middle line:90% So here, it's about 10 microns. 01:02:14.890 --> 01:02:16.960 align:middle line:90% So if I use a 10 micron-- 01:02:16.960 --> 01:02:20.510 align:middle line:90% so the energy is below 1.6 eV-- 01:02:20.510 --> 01:02:24.700 align:middle line:84% then those part of the energy is absorbed for very little. 01:02:24.700 --> 01:02:28.462 align:middle line:90% It goes through [INAUDIBLE]. 01:02:28.462 --> 01:02:30.760 align:middle line:90% They don't get absorbed. 01:02:30.760 --> 01:02:35.260 align:middle line:84% So it means [INAUDIBLE] this part can go out. 01:02:35.260 --> 01:02:37.730 align:middle line:90% So that's the absorption. 01:02:37.730 --> 01:02:41.950 align:middle line:84% And that's why [INAUDIBLE] say like a [INAUDIBLE] sigma 01:02:41.950 --> 01:02:46.520 align:middle line:90% six [INAUDIBLE]. 01:02:46.520 --> 01:02:50.920 align:middle line:84% They have very sharp [INAUDIBLE] 5th power. 01:02:50.920 --> 01:02:53.750 align:middle line:84% So if you have 1 micron-- this is the inverse-- 01:02:53.750 --> 01:02:55.930 align:middle line:90% is 1 micron. 01:02:55.930 --> 01:02:58.570 align:middle line:84% 10 to the 4th inverse [INAUDIBLE] 1 microns. 01:02:58.570 --> 01:02:59.574 align:middle line:90% It's a centimeter. 01:02:59.574 --> 01:03:02.150 align:middle line:90% 01:03:02.150 --> 01:03:05.010 align:middle line:84% So if you want to use [INAUDIBLE] the bandgap 01:03:05.010 --> 01:03:08.270 align:middle line:90% [INAUDIBLE]. 01:03:08.270 --> 01:03:13.760 align:middle line:84% And that's why when people make amorphous silicon [INAUDIBLE]. 01:03:13.760 --> 01:03:15.505 align:middle line:90% It's pretty good here. 01:03:15.505 --> 01:03:18.280 align:middle line:90% 01:03:18.280 --> 01:03:19.400 align:middle line:90% This is useless here. 01:03:19.400 --> 01:03:22.270 align:middle line:90% 01:03:22.270 --> 01:03:24.000 align:middle line:84% You can generate that from [INAUDIBLE] 01:03:24.000 --> 01:03:25.710 align:middle line:84% but then you don't have [INAUDIBLE]. 01:03:25.710 --> 01:03:28.200 align:middle line:90% So solar cell is [INAUDIBLE]. 01:03:28.200 --> 01:03:29.520 align:middle line:90% First, you want to absorb it. 01:03:29.520 --> 01:03:32.425 align:middle line:84% Second, you want to get a charge to the electrode. 01:03:32.425 --> 01:03:35.720 align:middle line:90% 01:03:35.720 --> 01:03:38.750 align:middle line:84% So why is it so different from the [INAUDIBLE]? 01:03:38.750 --> 01:03:45.420 align:middle line:90% 01:03:45.420 --> 01:03:47.600 align:middle line:84% This is the one I had mentioned before, 01:03:47.600 --> 01:03:49.540 align:middle line:84% whether it's a direct semiconductor 01:03:49.540 --> 01:03:52.705 align:middle line:90% gap or indirect gap. 01:03:52.705 --> 01:03:59.275 align:middle line:84% [INAUDIBLE] all the others that have sharp increased direct gap. 01:03:59.275 --> 01:04:03.670 align:middle line:84% So the photon goes from here to here. 01:04:03.670 --> 01:04:07.420 align:middle line:84% If the electron goes from valence to conduction band, 01:04:07.420 --> 01:04:11.150 align:middle line:84% you need to satisfy the energy conservation. 01:04:11.150 --> 01:04:13.420 align:middle line:84% So the photon energy [INAUDIBLE] the difference 01:04:13.420 --> 01:04:15.190 align:middle line:90% of electron-hole energy. 01:04:15.190 --> 01:04:18.310 align:middle line:84% So that's an energy conservation. 01:04:18.310 --> 01:04:21.340 align:middle line:84% And second is the momentum conservation. 01:04:21.340 --> 01:04:28.630 align:middle line:90% So momentum is h over lambda. 01:04:28.630 --> 01:04:32.710 align:middle line:90% Lambda photon is a micron. 01:04:32.710 --> 01:04:37.310 align:middle line:84% One level of electron is lattice constant [INAUDIBLE]. 01:04:37.310 --> 01:04:41.990 align:middle line:84% So in this, gives us 1 over [INAUDIBLE]. 01:04:41.990 --> 01:04:45.060 align:middle line:84% But so one more lambda is essentially vertical, 01:04:45.060 --> 01:04:47.420 align:middle line:84% so the photon doesn't give you much momentum. 01:04:47.420 --> 01:04:49.460 align:middle line:90% That's the problem. 01:04:49.460 --> 01:04:54.770 align:middle line:84% In indirect gap semiconductor, the bandgap is here. 01:04:54.770 --> 01:04:59.210 align:middle line:84% When you want, let's say, the momentum 01:04:59.210 --> 01:05:02.690 align:middle line:84% doesn't match between electron-hole. 01:05:02.690 --> 01:05:05.900 align:middle line:84% So if you want to leave this electron here, 01:05:05.900 --> 01:05:07.890 align:middle line:84% the photon don't have enough momentum. 01:05:07.890 --> 01:05:08.710 align:middle line:90% It has an energy. 01:05:08.710 --> 01:05:11.060 align:middle line:90% It doesn't have a momentum. 01:05:11.060 --> 01:05:13.520 align:middle line:90% It has to have something else. 01:05:13.520 --> 01:05:14.970 align:middle line:90% That's photon. 01:05:14.970 --> 01:05:18.570 align:middle line:84% You made a photon absorb a photon. 01:05:18.570 --> 01:05:22.160 align:middle line:84% Photon doesn't have much energy, but it has momentum. 01:05:22.160 --> 01:05:25.000 align:middle line:90% So this is a widely-- 01:05:25.000 --> 01:05:27.740 align:middle line:84% So you've got to have either emission 01:05:27.740 --> 01:05:29.140 align:middle line:90% or absorption of photons. 01:05:29.140 --> 01:05:32.360 align:middle line:84% So you have to have [INAUDIBLE], like I said. 01:05:32.360 --> 01:05:36.050 align:middle line:84% Very much scattered [INAUDIBLE] together. 01:05:36.050 --> 01:05:39.825 align:middle line:84% So the absorption is weaker because when 01:05:39.825 --> 01:05:42.200 align:middle line:84% the photon wants to do it, then there is no photon there. 01:05:42.200 --> 01:05:45.360 align:middle line:90% 01:05:45.360 --> 01:05:47.070 align:middle line:90% So that's the dilemma-- 01:05:47.070 --> 01:05:52.860 align:middle line:84% why they [INAUDIBLE] absorption there and near the bandgap. 01:05:52.860 --> 01:05:56.970 align:middle line:84% And of course, once you move to this region, it goes up. 01:05:56.970 --> 01:05:58.620 align:middle line:90% It goes from here to here. 01:05:58.620 --> 01:06:03.335 align:middle line:90% 01:06:03.335 --> 01:06:06.770 align:middle line:84% So that's when I said solar cell, essentially, 01:06:06.770 --> 01:06:09.650 align:middle line:84% you solve one problem, maybe the other problem. 01:06:09.650 --> 01:06:13.460 align:middle line:84% Thick or thin and how you can address those issues. 01:06:13.460 --> 01:06:16.680 align:middle line:84% It's really, the electron thing is better. 01:06:16.680 --> 01:06:18.020 align:middle line:90% You just take it out. 01:06:18.020 --> 01:06:20.930 align:middle line:84% Before they recombine, you send the electron [INAUDIBLE]. 01:06:20.930 --> 01:06:23.880 align:middle line:90% Photons get absorbed. 01:06:23.880 --> 01:06:29.050 align:middle line:90% 01:06:29.050 --> 01:06:32.370 align:middle line:84% Of course, that's related to efficiency. 01:06:32.370 --> 01:06:34.290 align:middle line:84% And the single cell is directly related 01:06:34.290 --> 01:06:36.420 align:middle line:90% to how much material you use. 01:06:36.420 --> 01:06:40.026 align:middle line:90% That's a cost issue. 01:06:40.026 --> 01:06:42.480 align:middle line:84% To have a low cost, you need a thinner. 01:06:42.480 --> 01:06:47.400 align:middle line:84% Less material, so always the material cost will be better. 01:06:47.400 --> 01:06:49.560 align:middle line:90% And we look at the cost. 01:06:49.560 --> 01:06:52.830 align:middle line:84% Now the lowest [INAUDIBLE] from [INAUDIBLE] and they 01:06:52.830 --> 01:06:58.080 align:middle line:84% have a manufacturing cost of about $1 per watt. 01:06:58.080 --> 01:07:02.970 align:middle line:84% And this is where the [INAUDIBLE] 01:07:02.970 --> 01:07:06.930 align:middle line:84% assuming certain amount of years you're using [INAUDIBLE]. 01:07:06.930 --> 01:07:11.580 align:middle line:84% [INAUDIBLE] you have $1 per watt. 01:07:11.580 --> 01:07:16.845 align:middle line:84% And it's a good curve here in terms of the cost. 01:07:16.845 --> 01:07:19.260 align:middle line:84% But saying, look at a [INAUDIBLE]. 01:07:19.260 --> 01:07:21.780 align:middle line:90% 01:07:21.780 --> 01:07:24.190 align:middle line:90% It takes a long time. 01:07:24.190 --> 01:07:26.670 align:middle line:84% Right now, the worldwide installation per year 01:07:26.670 --> 01:07:29.550 align:middle line:90% is about [? 67 ?] gigawatt. 01:07:29.550 --> 01:07:31.100 align:middle line:90% [? 67 ?] gigawatt. 01:07:31.100 --> 01:07:34.540 align:middle line:84% China put 100 gigawatts steam turbine every year. 01:07:34.540 --> 01:07:37.580 align:middle line:90% 01:07:37.580 --> 01:07:40.570 align:middle line:84% [INAUDIBLE] really just [INAUDIBLE]. 01:07:40.570 --> 01:07:44.520 align:middle line:90% 01:07:44.520 --> 01:07:47.240 align:middle line:84% It will take a lot of effort [INAUDIBLE]. 01:07:47.240 --> 01:07:52.890 align:middle line:90% 01:07:52.890 --> 01:07:56.190 align:middle line:90% Different type of solar cells. 01:07:56.190 --> 01:08:03.150 align:middle line:84% And a silicon cell is the most predominant. 01:08:03.150 --> 01:08:04.770 align:middle line:90% [INAUDIBLE] cost. 01:08:04.770 --> 01:08:06.840 align:middle line:84% Single crystal [INAUDIBLE] crystal [INAUDIBLE] 01:08:06.840 --> 01:08:10.110 align:middle line:90% is considered [INAUDIBLE]. 01:08:10.110 --> 01:08:15.550 align:middle line:84% And what's interesting is all those crystal surface 01:08:15.550 --> 01:08:18.210 align:middle line:90% [INAUDIBLE]. 01:08:18.210 --> 01:08:24.090 align:middle line:84% And that will trap the light so that we use refraction. 01:08:24.090 --> 01:08:26.560 align:middle line:84% We use refraction, not only reduce refraction. 01:08:26.560 --> 01:08:30.189 align:middle line:84% Once you get into the backside with metal, you reflect back. 01:08:30.189 --> 01:08:31.620 align:middle line:90% You don't want to go out. 01:08:31.620 --> 01:08:35.250 align:middle line:84% Some of the photons [INAUDIBLE] more [INAUDIBLE]. 01:08:35.250 --> 01:08:38.790 align:middle line:84% And the difference between single crystal or multicrystal 01:08:38.790 --> 01:08:44.460 align:middle line:84% is single crystal, you can get a regular [INAUDIBLE]. 01:08:44.460 --> 01:08:49.650 align:middle line:84% Just the manufacturing process [INAUDIBLE] crystal [INAUDIBLE]. 01:08:49.650 --> 01:08:54.279 align:middle line:84% And then you got, let's say, electron and silicon here. 01:08:54.279 --> 01:08:56.600 align:middle line:90% You can see this question here. 01:08:56.600 --> 01:09:00.350 align:middle line:84% If you generate [INAUDIBLE] it has to diffuse. 01:09:00.350 --> 01:09:03.010 align:middle line:90% It can go both ways. 01:09:03.010 --> 01:09:04.479 align:middle line:90% You don't have a field. 01:09:04.479 --> 01:09:09.340 align:middle line:84% You have to design so that the [INAUDIBLE] field [INAUDIBLE] 01:09:09.340 --> 01:09:11.060 align:middle line:90% go one direction. 01:09:11.060 --> 01:09:11.560 align:middle line:90% Yes? 01:09:11.560 --> 01:09:12.477 align:middle line:90% AUDIENCE: [INAUDIBLE]? 01:09:12.477 --> 01:09:17.319 align:middle line:90% 01:09:17.319 --> 01:09:21.580 align:middle line:84% GANG CHEN: It must be related to the material [INAUDIBLE]. 01:09:21.580 --> 01:09:24.255 align:middle line:84% I would imagine probably [INAUDIBLE] recombination, 01:09:24.255 --> 01:09:26.290 align:middle line:90% but I didn't check. 01:09:26.290 --> 01:09:29.220 align:middle line:84% But most of the time, people use [INAUDIBLE]. 01:09:29.220 --> 01:09:30.243 align:middle line:90% AUDIENCE: [INAUDIBLE]. 01:09:30.243 --> 01:09:34.882 align:middle line:90% 01:09:34.882 --> 01:09:38.620 align:middle line:84% GANG CHEN: No, the material is not silicon, then, no. 01:09:38.620 --> 01:09:39.410 align:middle line:90% That I don't know. 01:09:39.410 --> 01:09:40.035 align:middle line:90% I didn't check. 01:09:40.035 --> 01:09:42.750 align:middle line:90% 01:09:42.750 --> 01:09:43.735 align:middle line:90% [INAUDIBLE] 01:09:43.735 --> 01:09:47.735 align:middle line:90% 01:09:47.735 --> 01:09:54.710 align:middle line:84% And you can see here, [INAUDIBLE] contact. 01:09:54.710 --> 01:09:56.065 align:middle line:90% We use that Schottky barrier. 01:09:56.065 --> 01:09:59.360 align:middle line:90% 01:09:59.360 --> 01:10:01.833 align:middle line:90% And you have [INAUDIBLE]. 01:10:01.833 --> 01:10:04.850 align:middle line:90% 01:10:04.850 --> 01:10:06.540 align:middle line:90% That's a shielding effect. 01:10:06.540 --> 01:10:09.200 align:middle line:84% [INAUDIBLE] the photon couldn't get through. 01:10:09.200 --> 01:10:10.250 align:middle line:90% That's not good. 01:10:10.250 --> 01:10:14.240 align:middle line:84% So sometimes, people even bury the grid down or make it 01:10:14.240 --> 01:10:17.240 align:middle line:90% on the back side [INAUDIBLE]. 01:10:17.240 --> 01:10:21.380 align:middle line:84% So there are a lot of [INAUDIBLE] improvement. 01:10:21.380 --> 01:10:23.725 align:middle line:84% A little bit of improvement people done over the years. 01:10:23.725 --> 01:10:29.280 align:middle line:90% 01:10:29.280 --> 01:10:36.480 align:middle line:84% [INAUDIBLE] here say [INAUDIBLE] amorphous silicon [INAUDIBLE]. 01:10:36.480 --> 01:10:41.730 align:middle line:84% Amorphous silicon [INAUDIBLE] in particular, so you can use it. 01:10:41.730 --> 01:10:43.950 align:middle line:84% But the electron amorphous material-- cell 01:10:43.950 --> 01:10:45.780 align:middle line:90% is very hard to move. 01:10:45.780 --> 01:10:49.440 align:middle line:84% A lot of [INAUDIBLE] recombination. 01:10:49.440 --> 01:10:53.280 align:middle line:84% So you've got [INAUDIBLE] the amorphous silicon 01:10:53.280 --> 01:10:56.280 align:middle line:90% cell is about 200 nanometers. 01:10:56.280 --> 01:11:00.780 align:middle line:84% And if you can reduce that even thinner, it's probably better. 01:11:00.780 --> 01:11:04.590 align:middle line:84% Too thin, your manufacturing is a problem [INAUDIBLE] 01:11:04.590 --> 01:11:06.527 align:middle line:90% always [INAUDIBLE]. 01:11:06.527 --> 01:11:09.150 align:middle line:90% 01:11:09.150 --> 01:11:13.800 align:middle line:84% But this is actually showing you just a multijunction. 01:11:13.800 --> 01:11:17.096 align:middle line:84% So if you look at it, this is [INAUDIBLE]. 01:11:17.096 --> 01:11:19.770 align:middle line:90% 01:11:19.770 --> 01:11:22.830 align:middle line:84% So unlike silicon, silicon, you look 01:11:22.830 --> 01:11:25.290 align:middle line:90% at the electrode on the top. 01:11:25.290 --> 01:11:26.820 align:middle line:90% It's spaced apart. 01:11:26.820 --> 01:11:30.520 align:middle line:84% So electrons go here-- actually diffuse to the electrode, 01:11:30.520 --> 01:11:35.260 align:middle line:84% naturally going to the electrode, right? 01:11:35.260 --> 01:11:39.250 align:middle line:84% Silicon is reasonably good in terms of recombination. 01:11:39.250 --> 01:11:41.770 align:middle line:90% Amorphous silicon doesn't work. 01:11:41.770 --> 01:11:44.870 align:middle line:84% So you got to have electrode right on top of it. 01:11:44.870 --> 01:11:47.630 align:middle line:90% 01:11:47.630 --> 01:11:49.805 align:middle line:90% So it's a [INAUDIBLE] conductor. 01:11:49.805 --> 01:11:52.810 align:middle line:90% 01:11:52.810 --> 01:11:58.660 align:middle line:90% And this is [INAUDIBLE]. 01:11:58.660 --> 01:12:01.620 align:middle line:84% The active region is about 6 microns. 01:12:01.620 --> 01:12:06.480 align:middle line:90% 01:12:06.480 --> 01:12:10.150 align:middle line:84% And this is [INAUDIBLE] space after [INAUDIBLE]. 01:12:10.150 --> 01:12:14.040 align:middle line:90% 01:12:14.040 --> 01:12:17.450 align:middle line:84% It's a trade-off between recombination and absorption. 01:12:17.450 --> 01:12:20.550 align:middle line:84% If you increase absorption, [INAUDIBLE] always better 01:12:20.550 --> 01:12:22.530 align:middle line:84% [INAUDIBLE] theoretically work better. 01:12:22.530 --> 01:12:24.930 align:middle line:84% If you can use [INAUDIBLE] the first [INAUDIBLE]. 01:12:24.930 --> 01:12:28.110 align:middle line:90% 01:12:28.110 --> 01:12:31.560 align:middle line:84% You can capitalize on those methods [INAUDIBLE]. 01:12:31.560 --> 01:12:34.110 align:middle line:90% 01:12:34.110 --> 01:12:38.970 align:middle line:84% AUDIENCE: [INAUDIBLE] the number? 01:12:38.970 --> 01:12:43.500 align:middle line:84% GANG CHEN: [INAUDIBLE] here, it is, about 0.5 microns. 01:12:43.500 --> 01:12:47.310 align:middle line:84% I'm not sure whether that's the best way to optimize it. 01:12:47.310 --> 01:12:50.590 align:middle line:84% I'm just copy from where I was trying to show the scale rather 01:12:50.590 --> 01:12:51.970 align:middle line:90% than [INAUDIBLE]. 01:12:51.970 --> 01:12:54.730 align:middle line:90% 01:12:54.730 --> 01:12:57.120 align:middle line:84% So if you are [INAUDIBLE] you're also 01:12:57.120 --> 01:12:58.980 align:middle line:84% trying to improve the conductivity. 01:12:58.980 --> 01:13:01.527 align:middle line:84% Basically, you want a very highly conductive. 01:13:01.527 --> 01:13:03.360 align:middle line:84% So if you're highly conductive, then thinner 01:13:03.360 --> 01:13:06.450 align:middle line:84% is better because the free electrons in this material 01:13:06.450 --> 01:13:10.990 align:middle line:84% also absorb photons, which generally, you don't want that. 01:13:10.990 --> 01:13:15.300 align:middle line:90% 01:13:15.300 --> 01:13:20.400 align:middle line:84% Polymer cell-- polymer, ideally, [INAUDIBLE] polymer 01:13:20.400 --> 01:13:23.460 align:middle line:90% reduces cost [INAUDIBLE]. 01:13:23.460 --> 01:13:26.250 align:middle line:84% The problem with polymer is that it's 01:13:26.250 --> 01:13:29.880 align:middle line:84% even worse than amorphous silicon. 01:13:29.880 --> 01:13:33.420 align:middle line:84% You generate-- it immediately recombines. 01:13:33.420 --> 01:13:36.420 align:middle line:90% Recombination is very fast. 01:13:36.420 --> 01:13:39.510 align:middle line:84% In fact, when you first generate the bonded together, 01:13:39.510 --> 01:13:42.840 align:middle line:84% they are not even a separate electron or hole. 01:13:42.840 --> 01:13:45.900 align:middle line:84% This [INAUDIBLE] state for the exciton. 01:13:45.900 --> 01:13:49.900 align:middle line:84% So it maybe has an exciton center, [INAUDIBLE] center, 01:13:49.900 --> 01:13:51.720 align:middle line:90% [INAUDIBLE]. 01:13:51.720 --> 01:13:57.060 align:middle line:84% I have, let's say, maybe a solar thermal center [INAUDIBLE]. 01:13:57.060 --> 01:13:58.580 align:middle line:84% And there is-- the whole issue is 01:13:58.580 --> 01:14:03.120 align:middle line:84% to find how I can separate this exciton 01:14:03.120 --> 01:14:06.408 align:middle line:84% and make the electron go one way and the hole go the other way. 01:14:06.408 --> 01:14:08.550 align:middle line:90% [INAUDIBLE] 01:14:08.550 --> 01:14:15.880 align:middle line:84% So in that case, you look at this really thin 220 nanometer 01:14:15.880 --> 01:14:19.270 align:middle line:90% [INAUDIBLE] separate. 01:14:19.270 --> 01:14:22.120 align:middle line:84% But that's still difficult, so people are now 01:14:22.120 --> 01:14:24.820 align:middle line:84% doing the so-called bulk junction cells. 01:14:24.820 --> 01:14:27.280 align:middle line:84% So you put a [INAUDIBLE] of multiple. 01:14:27.280 --> 01:14:29.320 align:middle line:84% One conducts electron, one conducts [INAUDIBLE]. 01:14:29.320 --> 01:14:31.540 align:middle line:84% So locally, you generate electron-hole pair. 01:14:31.540 --> 01:14:33.340 align:middle line:84% Like here, you want [INAUDIBLE] separate 01:14:33.340 --> 01:14:37.338 align:middle line:84% so the electron goes one model, the hole goes another model. 01:14:37.338 --> 01:14:43.074 align:middle line:90% 01:14:43.074 --> 01:14:45.022 align:middle line:90% [INAUDIBLE] cell. 01:14:45.022 --> 01:14:47.060 align:middle line:90% So this is not a solid cell. 01:14:47.060 --> 01:14:49.870 align:middle line:90% It's an actual liquid cell. 01:14:49.870 --> 01:14:56.726 align:middle line:84% And what it has is titanium dioxide nanoparticles, 01:14:56.726 --> 01:15:00.460 align:middle line:84% [INAUDIBLE] titanium dioxide nanoparticles [INAUDIBLE] 01:15:00.460 --> 01:15:02.400 align:middle line:90% all three electron-holes. 01:15:02.400 --> 01:15:02.950 align:middle line:90% Too large. 01:15:02.950 --> 01:15:04.646 align:middle line:90% Not much hole [INAUDIBLE]. 01:15:04.646 --> 01:15:07.264 align:middle line:90% 01:15:07.264 --> 01:15:10.720 align:middle line:84% And so what [INAUDIBLE] cell does 01:15:10.720 --> 01:15:14.410 align:middle line:84% is put a dye on top of that titanium dioxide, 01:15:14.410 --> 01:15:17.110 align:middle line:84% and this dye is a photosynthesizer. 01:15:17.110 --> 01:15:18.980 align:middle line:84% So this is where photons are absorbed. 01:15:18.980 --> 01:15:21.630 align:middle line:90% 01:15:21.630 --> 01:15:28.470 align:middle line:84% And photons are absorbed [INAUDIBLE] monolayer generated 01:15:28.470 --> 01:15:31.560 align:middle line:90% electron-hole pair exciton. 01:15:31.560 --> 01:15:34.200 align:middle line:84% Before they dissociate, they are exciton. 01:15:34.200 --> 01:15:36.480 align:middle line:84% And you want them to dissociate-- 01:15:36.480 --> 01:15:38.940 align:middle line:90% separate into electron-hole. 01:15:38.940 --> 01:15:43.590 align:middle line:84% And electron will eventually flow into nanoparticle, 01:15:43.590 --> 01:15:45.560 align:middle line:90% so the nanoparticle diffuses. 01:15:45.560 --> 01:15:47.400 align:middle line:90% Now there is no field. 01:15:47.400 --> 01:15:50.730 align:middle line:90% Really, diffusion process. 01:15:50.730 --> 01:15:52.920 align:middle line:84% So the [INAUDIBLE] run out, so that's 01:15:52.920 --> 01:15:56.520 align:middle line:84% why people try to do nanowires so that gives a better 01:15:56.520 --> 01:16:00.060 align:middle line:84% path to the electron rather than nanoparticles. 01:16:00.060 --> 01:16:02.360 align:middle line:84% But the problem is that if we do nanowire, 01:16:02.360 --> 01:16:05.520 align:middle line:84% your surface area for absorption also changes. 01:16:05.520 --> 01:16:11.790 align:middle line:84% Here, you can have this surface area [INAUDIBLE]. 01:16:11.790 --> 01:16:13.790 align:middle line:84% Your nanowire is only this surface-- 01:16:13.790 --> 01:16:16.443 align:middle line:90% 01:16:16.443 --> 01:16:17.520 align:middle line:90% vertical. 01:16:17.520 --> 01:16:20.460 align:middle line:90% So the surface area [INAUDIBLE]. 01:16:20.460 --> 01:16:22.260 align:middle line:90% The particle has advantages. 01:16:22.260 --> 01:16:24.400 align:middle line:84% It's almost like a 3D wrap-around particle. 01:16:24.400 --> 01:16:27.270 align:middle line:90% [INAUDIBLE] 01:16:27.270 --> 01:16:31.034 align:middle line:84% There's always a [INAUDIBLE] trade-off. 01:16:31.034 --> 01:16:36.900 align:middle line:90% And the [INAUDIBLE] liquid. 01:16:36.900 --> 01:16:38.730 align:middle line:84% And this is a chemical reaction now. 01:16:38.730 --> 01:16:43.080 align:middle line:84% It's not just a hole diffused [INAUDIBLE]. 01:16:43.080 --> 01:16:48.210 align:middle line:84% So the hole oxidize the molecules 01:16:48.210 --> 01:16:51.360 align:middle line:84% in the case of [INAUDIBLE] iodide. 01:16:51.360 --> 01:16:55.980 align:middle line:84% Iodide in the liquid [INAUDIBLE] iodine(III) and this 01:16:55.980 --> 01:16:59.810 align:middle line:84% charged molecule diffuse to [INAUDIBLE]. 01:16:59.810 --> 01:17:01.980 align:middle line:90% So again, diffusion. 01:17:01.980 --> 01:17:05.220 align:middle line:84% So electron diffused [INAUDIBLE] electrode, hole 01:17:05.220 --> 01:17:07.910 align:middle line:84% diffuse to the other electron [INAUDIBLE] reduce, 01:17:07.910 --> 01:17:12.150 align:middle line:90% and the reduced diffuse back. 01:17:12.150 --> 01:17:15.030 align:middle line:84% Sounds like a very cool process, right? 01:17:15.030 --> 01:17:22.020 align:middle line:84% And the idea is, again, you can [INAUDIBLE] pretty good 01:17:22.020 --> 01:17:23.094 align:middle line:90% in [INAUDIBLE]. 01:17:23.094 --> 01:17:27.670 align:middle line:90% 01:17:27.670 --> 01:17:35.890 align:middle line:84% And see, so this is a flat surface versus the [INAUDIBLE]. 01:17:35.890 --> 01:17:40.700 align:middle line:90% 01:17:40.700 --> 01:17:44.360 align:middle line:84% So that's pretty much what I want to say. 01:17:44.360 --> 01:17:46.220 align:middle line:90% And then we're right on time. 01:17:46.220 --> 01:17:51.830 align:middle line:90% And this is the dream. 01:17:51.830 --> 01:17:54.380 align:middle line:84% The dream, in a sense-- here is the cost, 01:17:54.380 --> 01:17:56.030 align:middle line:90% here is the efficiency. 01:17:56.030 --> 01:17:59.360 align:middle line:84% You want a low cost, high efficiency, 01:17:59.360 --> 01:18:04.880 align:middle line:84% Silicon is cheaper in the first generation for gen 1. 01:18:04.880 --> 01:18:08.570 align:middle line:84% Thin film is here-- lower efficiency, but lower cost. 01:18:08.570 --> 01:18:11.240 align:middle line:84% And in fact, they make the business, at least 01:18:11.240 --> 01:18:14.390 align:middle line:90% for first order [INAUDIBLE]. 01:18:14.390 --> 01:18:19.380 align:middle line:84% And people are [INAUDIBLE] gen 3. 01:18:19.380 --> 01:18:21.510 align:middle line:90% No winner in gen 3 yet. 01:18:21.510 --> 01:18:24.861 align:middle line:84% This is where you guys come in later. 01:18:24.861 --> 01:18:27.705 align:middle line:90% 01:18:27.705 --> 01:18:28.740 align:middle line:90% OK? 01:18:28.740 --> 01:18:31.520 align:middle line:90% I think that's it.