WEBVTT 00:00:00.000 --> 00:00:01.352 align:middle line:90% 00:00:01.352 --> 00:00:02.810 align:middle line:84% JUDY HOYT: Just remind people where 00:00:02.810 --> 00:00:06.620 align:middle line:84% we are on the course schedule, we just finished up 00:00:06.620 --> 00:00:09.350 align:middle line:84% chapter 9 on deposition and epitaxy, 00:00:09.350 --> 00:00:12.080 align:middle line:84% and we're going to plow ahead, moving along 00:00:12.080 --> 00:00:18.170 align:middle line:84% to chapter 10, which we'll cover in the next couple of lectures. 00:00:18.170 --> 00:00:23.240 align:middle line:84% Chapter 10 is on etching, which is a very important capability 00:00:23.240 --> 00:00:26.090 align:middle line:90% for CMOS processing. 00:00:26.090 --> 00:00:29.090 align:middle line:84% And today is the 18th, and I guess 00:00:29.090 --> 00:00:31.250 align:middle line:90% homework number 5 is due today. 00:00:31.250 --> 00:00:35.480 align:middle line:84% So I brought a new folder here up front, this yellow folder, 00:00:35.480 --> 00:00:37.333 align:middle line:84% we'll use to hold your homework number 5. 00:00:37.333 --> 00:00:39.500 align:middle line:84% I'm going to have-- homework number 4's been graded, 00:00:39.500 --> 00:00:41.100 align:middle line:90% but I need to go through it. 00:00:41.100 --> 00:00:43.115 align:middle line:84% So I'll have that back to you next week. 00:00:43.115 --> 00:00:47.190 align:middle line:90% 00:00:47.190 --> 00:00:50.910 align:middle line:84% So let's take a look at today's handout, which 00:00:50.910 --> 00:00:59.620 align:middle line:84% is handout number 33, and we're starting chapter 10 on etching. 00:00:59.620 --> 00:01:02.542 align:middle line:84% We've talked a lot about, now, a lot of different processes. 00:01:02.542 --> 00:01:04.209 align:middle line:84% We spent the last three lectures talking 00:01:04.209 --> 00:01:05.950 align:middle line:90% about depositing thin films. 00:01:05.950 --> 00:01:07.900 align:middle line:84% After they're deposited, typically they 00:01:07.900 --> 00:01:10.750 align:middle line:84% have to be etched to make some kind of a circuit pattern 00:01:10.750 --> 00:01:13.240 align:middle line:84% or whatever you're trying to pattern. 00:01:13.240 --> 00:01:14.977 align:middle line:84% An important requirement of etching 00:01:14.977 --> 00:01:16.810 align:middle line:84% that we're going to talk about is that films 00:01:16.810 --> 00:01:18.460 align:middle line:90% need to be etched selectively. 00:01:18.460 --> 00:01:20.080 align:middle line:84% You want to etch one film, but you 00:01:20.080 --> 00:01:22.840 align:middle line:84% don't want to etch what's underneath it typically, 00:01:22.840 --> 00:01:25.210 align:middle line:84% or sometimes, you don't want to etch the mask. 00:01:25.210 --> 00:01:27.370 align:middle line:84% So selectivity is very important, 00:01:27.370 --> 00:01:29.410 align:middle line:90% and we'll talk about that. 00:01:29.410 --> 00:01:32.680 align:middle line:84% Besides etching thin films, sometimes we 00:01:32.680 --> 00:01:34.660 align:middle line:84% want to etch down into the substrate itself 00:01:34.660 --> 00:01:37.540 align:middle line:90% for dRAMs these days to-- 00:01:37.540 --> 00:01:39.310 align:middle line:84% so you don't use so much chip area. 00:01:39.310 --> 00:01:41.110 align:middle line:84% The capacitors, the storage capacitors, 00:01:41.110 --> 00:01:42.985 align:middle line:84% are actually sometimes in the trench process. 00:01:42.985 --> 00:01:45.277 align:middle line:84% They're actually etched deep trenches into the silicon. 00:01:45.277 --> 00:01:47.110 align:middle line:84% They make capacitors to store the memory-- 00:01:47.110 --> 00:01:49.400 align:middle line:84% to store the charge and the memory that way. 00:01:49.400 --> 00:01:53.485 align:middle line:84% So sometimes you need to etch silicon substrate. 00:01:53.485 --> 00:01:56.290 align:middle line:84% Dry etching, as it's called, or plasma etching, 00:01:56.290 --> 00:01:58.912 align:middle line:84% is really critical to CMOS fabrication. 00:01:58.912 --> 00:02:00.370 align:middle line:84% I'm going to emphasize dry etching. 00:02:00.370 --> 00:02:02.500 align:middle line:84% I only have two or three slides on wet etching. 00:02:02.500 --> 00:02:05.140 align:middle line:84% Wet etching is still used in research and development, 00:02:05.140 --> 00:02:08.139 align:middle line:84% but it's very rarely used in a manufacturing process, 00:02:08.139 --> 00:02:10.330 align:middle line:84% and we'll see a little bit why that is. 00:02:10.330 --> 00:02:12.460 align:middle line:84% So the next two lectures, what I want to try to do 00:02:12.460 --> 00:02:14.740 align:middle line:84% is give you a basic introduction to etching, 00:02:14.740 --> 00:02:19.460 align:middle line:84% mostly plasma etching, some specific examples. 00:02:19.460 --> 00:02:21.940 align:middle line:84% The next lecture, I'm going to have some specific emphasis 00:02:21.940 --> 00:02:23.740 align:middle line:84% on gate etching, polysilicon gate etching, 00:02:23.740 --> 00:02:27.140 align:middle line:84% since that's really very important to the CMOS process. 00:02:27.140 --> 00:02:31.020 align:middle line:84% And we'll talk a little bit, next lecture as well, 00:02:31.020 --> 00:02:35.010 align:middle line:84% on modeling, how we model etching processes. 00:02:35.010 --> 00:02:38.020 align:middle line:84% Whoops, I forgot to go to full screen mode here. 00:02:38.020 --> 00:02:39.270 align:middle line:90% Let me do that. 00:02:39.270 --> 00:02:42.640 align:middle line:84% So let's move on to slide number 2 in your handout. 00:02:42.640 --> 00:02:45.280 align:middle line:84% Basic introduction, this is what we 00:02:45.280 --> 00:02:47.150 align:middle line:84% talked about in the beginning of the course. 00:02:47.150 --> 00:02:51.310 align:middle line:84% Just to remind you, the planar process, what it consists of, 00:02:51.310 --> 00:02:54.280 align:middle line:84% usually, you'll have a deposited film 00:02:54.280 --> 00:02:57.250 align:middle line:84% here, which is shown in blue, and then to pattern it, 00:02:57.250 --> 00:03:00.235 align:middle line:84% you need to put down a photosensitive film, which 00:03:00.235 --> 00:03:02.860 align:middle line:84% is called photoresist, that can be put down at room temperature 00:03:02.860 --> 00:03:04.150 align:middle line:90% or spun on. 00:03:04.150 --> 00:03:08.830 align:middle line:84% You then expose it through a mask using ultraviolet light, 00:03:08.830 --> 00:03:11.430 align:middle line:90% or the right wavelength light. 00:03:11.430 --> 00:03:15.240 align:middle line:84% That either causes the resist to stay wherever 00:03:15.240 --> 00:03:18.268 align:middle line:84% it's struck by the light or to be removed or etched 00:03:18.268 --> 00:03:19.560 align:middle line:90% where it's struck by the light. 00:03:19.560 --> 00:03:22.800 align:middle line:84% Either way is a positive or negative resist, 00:03:22.800 --> 00:03:26.572 align:middle line:84% and you end up with, here, after developing the resist, 00:03:26.572 --> 00:03:27.780 align:middle line:90% you end up with an etch mask. 00:03:27.780 --> 00:03:29.970 align:middle line:84% So I have photoresist shown here in gray, 00:03:29.970 --> 00:03:35.610 align:middle line:84% and you're ready now to etch the blue thin film below it, 00:03:35.610 --> 00:03:37.737 align:middle line:84% and you do that either in a dry or a wet process. 00:03:37.737 --> 00:03:39.570 align:middle line:84% Then you remove the resist, and you're done. 00:03:39.570 --> 00:03:42.600 align:middle line:84% So now, you've transferred the pattern from the mask 00:03:42.600 --> 00:03:46.390 align:middle line:84% to the resist and from the resist to the film. 00:03:46.390 --> 00:03:48.608 align:middle line:84% So we won't talk about the lithographic process 00:03:48.608 --> 00:03:50.400 align:middle line:84% of transferring to the masks to the resist. 00:03:50.400 --> 00:03:52.572 align:middle line:84% That's a whole class, Hank Smith's course. 00:03:52.572 --> 00:03:54.030 align:middle line:84% What we're going to talk about here 00:03:54.030 --> 00:03:55.590 align:middle line:90% is how we etch this thin film. 00:03:55.590 --> 00:03:58.598 align:middle line:90% 00:03:58.598 --> 00:04:00.390 align:middle line:84% The most important things people care about 00:04:00.390 --> 00:04:03.600 align:middle line:84% are selectivity and directionality, 00:04:03.600 --> 00:04:06.750 align:middle line:84% and I'll talk a little bit more about what they mean. 00:04:06.750 --> 00:04:09.330 align:middle line:84% Selectivity comes from the chemistry. 00:04:09.330 --> 00:04:11.820 align:middle line:84% That is, how selective is the etchant 00:04:11.820 --> 00:04:14.070 align:middle line:90% to a particular material. 00:04:14.070 --> 00:04:17.070 align:middle line:84% The directionality, that is how the etches vertically 00:04:17.070 --> 00:04:19.110 align:middle line:84% versus horizontally, that usually comes 00:04:19.110 --> 00:04:21.899 align:middle line:84% from more physical processes, like energetic ions 00:04:21.899 --> 00:04:24.150 align:middle line:90% bombarding the wafer. 00:04:24.150 --> 00:04:25.882 align:middle line:84% And modern etching techniques have 00:04:25.882 --> 00:04:27.840 align:middle line:84% a little bit of directionality and a little bit 00:04:27.840 --> 00:04:30.510 align:middle line:90% of selectivity, both of those. 00:04:30.510 --> 00:04:33.390 align:middle line:84% Simulation tools are only more recently being developed. 00:04:33.390 --> 00:04:36.150 align:middle line:84% They are starting to play an important role, 00:04:36.150 --> 00:04:37.410 align:middle line:90% just like in deposition. 00:04:37.410 --> 00:04:41.130 align:middle line:84% Again, you will not find etching modeled in SUPREM-IV. 00:04:41.130 --> 00:04:43.530 align:middle line:84% There is no topographic modeling in that. 00:04:43.530 --> 00:04:46.680 align:middle line:84% You need a topography simulator like Speedy, 00:04:46.680 --> 00:04:48.480 align:middle line:84% and usually, topography simulators 00:04:48.480 --> 00:04:50.820 align:middle line:84% do both deposition and etching because they are really 00:04:50.820 --> 00:04:51.780 align:middle line:90% inverse processes. 00:04:51.780 --> 00:04:55.260 align:middle line:84% In one case, you have things coming down, reacting 00:04:55.260 --> 00:04:56.378 align:middle line:90% and sticking. 00:04:56.378 --> 00:04:58.170 align:middle line:84% In other case, you have things coming down, 00:04:58.170 --> 00:05:01.120 align:middle line:84% causing a removal reaction, and something coming off, 00:05:01.120 --> 00:05:01.620 align:middle line:90% an etching. 00:05:01.620 --> 00:05:04.110 align:middle line:84% But they're very much mathematically 00:05:04.110 --> 00:05:07.710 align:middle line:90% very similar processes. 00:05:07.710 --> 00:05:11.320 align:middle line:84% So slide number 3 shows a little bit about these characteristics, 00:05:11.320 --> 00:05:12.990 align:middle line:84% some of the important characteristics 00:05:12.990 --> 00:05:15.990 align:middle line:84% that I just talked about, and illustrate them. 00:05:15.990 --> 00:05:18.540 align:middle line:84% Up at the top, we have a situation 00:05:18.540 --> 00:05:20.970 align:middle line:84% where we start with photoresist or a mask. 00:05:20.970 --> 00:05:23.700 align:middle line:84% Let's say, that looks like these dark regions on top 00:05:23.700 --> 00:05:26.250 align:middle line:84% are the pattern or the mask, and we 00:05:26.250 --> 00:05:29.220 align:middle line:84% want to transfer that into this hatched region, which is 00:05:29.220 --> 00:05:31.215 align:middle line:90% the film we're trying to etch. 00:05:31.215 --> 00:05:35.380 align:middle line:84% After etching, shown on the upper right, what you have is, 00:05:35.380 --> 00:05:37.750 align:middle line:84% here, you see the mask, which is retained. 00:05:37.750 --> 00:05:40.060 align:middle line:84% Good fidelity, the mask was not attacked, 00:05:40.060 --> 00:05:41.500 align:middle line:90% so there's good selectivity. 00:05:41.500 --> 00:05:43.450 align:middle line:84% But we have what we call undercutting. 00:05:43.450 --> 00:05:45.520 align:middle line:84% You can see what happened is not only 00:05:45.520 --> 00:05:48.640 align:middle line:84% did we etch the film vertically, but we also etched laterally 00:05:48.640 --> 00:05:53.060 align:middle line:84% on each side, and so the feature size has been distorted. 00:05:53.060 --> 00:05:54.850 align:middle line:84% It's not exactly what the feature size 00:05:54.850 --> 00:05:58.420 align:middle line:84% was on the mask because of this undercutting. 00:05:58.420 --> 00:06:03.220 align:middle line:84% So that's the top two pictures illustrate undercutting. 00:06:03.220 --> 00:06:06.850 align:middle line:84% The second one here, labeled B, illustrates 00:06:06.850 --> 00:06:09.850 align:middle line:84% two things, both undercutting and poor selectivity. 00:06:09.850 --> 00:06:11.410 align:middle line:90% Up here, I had good selectivity. 00:06:11.410 --> 00:06:14.020 align:middle line:84% The resist was not attacked, and the layer below, 00:06:14.020 --> 00:06:16.450 align:middle line:84% whatever it might be, was not attacked by the etch. 00:06:16.450 --> 00:06:19.240 align:middle line:84% You noticed, it went down, and it stopped. 00:06:19.240 --> 00:06:22.060 align:middle line:84% Here, I have both undercutting and poor selectivity. 00:06:22.060 --> 00:06:23.800 align:middle line:84% Indeed, you can see what's happened. 00:06:23.800 --> 00:06:25.750 align:middle line:84% I've etched laterally left and right, 00:06:25.750 --> 00:06:28.240 align:middle line:84% so my holes end up being bigger than I had wanted them 00:06:28.240 --> 00:06:31.577 align:middle line:84% to be on the mask, and I've also attacked the layer underneath, 00:06:31.577 --> 00:06:33.160 align:middle line:84% so I don't have very good selectivity. 00:06:33.160 --> 00:06:36.460 align:middle line:84% The etchant continues to etch this layer-- 00:06:36.460 --> 00:06:40.660 align:middle line:84% it continues to etch the bottom layer as it's in the solution, 00:06:40.660 --> 00:06:42.910 align:middle line:84% or maybe in the plasma, and not only that, 00:06:42.910 --> 00:06:44.860 align:middle line:84% the etch also attacked the photoresist. 00:06:44.860 --> 00:06:48.250 align:middle line:84% And we caused it to get rounded, and that also 00:06:48.250 --> 00:06:49.510 align:middle line:90% can distort the feature size. 00:06:49.510 --> 00:06:52.390 align:middle line:90% 00:06:52.390 --> 00:06:55.180 align:middle line:84% Usually, what we want most of the time, 00:06:55.180 --> 00:06:57.550 align:middle line:84% although it's not always the case, most of the time, 00:06:57.550 --> 00:06:59.650 align:middle line:84% we want what we call anisotropic etching. 00:06:59.650 --> 00:07:03.587 align:middle line:84% We want the etching to go down perfectly straight. 00:07:03.587 --> 00:07:05.170 align:middle line:84% Generally maybe a slight angle, but we 00:07:05.170 --> 00:07:07.900 align:middle line:84% want to be able to control the angle in any case, 00:07:07.900 --> 00:07:09.610 align:middle line:90% almost vertical. 00:07:09.610 --> 00:07:11.830 align:middle line:84% And usually, we want highly selective, 00:07:11.830 --> 00:07:15.850 align:middle line:84% so we'd like to have what we call a selectivity 00:07:15.850 --> 00:07:17.570 align:middle line:90% on the order of 25 to 50. 00:07:17.570 --> 00:07:19.210 align:middle line:84% So that means the etch would etch 00:07:19.210 --> 00:07:20.950 align:middle line:84% this top film, the film we're trying 00:07:20.950 --> 00:07:23.620 align:middle line:84% to etch, 25 to 50 times faster than it 00:07:23.620 --> 00:07:25.330 align:middle line:90% etches the film underneath it. 00:07:25.330 --> 00:07:29.380 align:middle line:84% Both of these, anisotropy and selectivity, 00:07:29.380 --> 00:07:30.280 align:middle line:90% are usually desired. 00:07:30.280 --> 00:07:31.810 align:middle line:84% Unfortunately, it's very hard to get 00:07:31.810 --> 00:07:33.477 align:middle line:84% both, the best of both worlds, but there 00:07:33.477 --> 00:07:35.110 align:middle line:84% are systems that try to optimize this, 00:07:35.110 --> 00:07:37.220 align:middle line:84% and that's what we'll talk about. 00:07:37.220 --> 00:07:39.610 align:middle line:84% So as we move on the bottom of slide 3, 00:07:39.610 --> 00:07:44.620 align:middle line:84% as you move from the left panel A to B to C, what you're moving 00:07:44.620 --> 00:07:47.380 align:middle line:84% is from completely isotropic etching, 00:07:47.380 --> 00:07:50.140 align:middle line:84% as you can see by the little vectors here, 00:07:50.140 --> 00:07:52.670 align:middle line:84% that the film etches-- this particular etching is 00:07:52.670 --> 00:07:55.160 align:middle line:84% etching this material vertically at the same rate it 00:07:55.160 --> 00:07:56.520 align:middle line:90% is laterally. 00:07:56.520 --> 00:07:58.160 align:middle line:84% Here's a little more anisotropic. 00:07:58.160 --> 00:08:00.560 align:middle line:84% The vertical etch rate is faster than it is laterally, 00:08:00.560 --> 00:08:02.227 align:middle line:84% but still, there's some lateral etching, 00:08:02.227 --> 00:08:04.640 align:middle line:84% and this is completely anisotropic. 00:08:04.640 --> 00:08:07.530 align:middle line:84% There is no lateral etch rate, only vertical etching. 00:08:07.530 --> 00:08:10.190 align:middle line:84% So you're moving from completely isotropic 00:08:10.190 --> 00:08:11.750 align:middle line:90% to completely anisotropic-- 00:08:11.750 --> 00:08:14.900 align:middle line:84% I'm sorry, from left to right, so you're getting more and more 00:08:14.900 --> 00:08:15.920 align:middle line:90% directionality. 00:08:15.920 --> 00:08:18.590 align:middle line:90% 00:08:18.590 --> 00:08:21.188 align:middle line:84% On slide 4, I've listed some things that 00:08:21.188 --> 00:08:22.730 align:middle line:84% are fairly obvious to you, but I just 00:08:22.730 --> 00:08:25.610 align:middle line:84% wanted to make note of some of the practical requirements 00:08:25.610 --> 00:08:27.440 align:middle line:90% on etching. 00:08:27.440 --> 00:08:29.430 align:middle line:84% First of all, we want to get the right profile, 00:08:29.430 --> 00:08:30.950 align:middle line:84% whether we want sloped or vertical. 00:08:30.950 --> 00:08:32.510 align:middle line:90% That's obvious. 00:08:32.510 --> 00:08:34.309 align:middle line:90% We don't want much undercutting. 00:08:34.309 --> 00:08:37.400 align:middle line:84% Undercutting or etch bias is bad because the feature 00:08:37.400 --> 00:08:39.576 align:middle line:84% size on the mask is not maintained on the wafer, 00:08:39.576 --> 00:08:41.659 align:middle line:84% and that's not what the circuit designer designed. 00:08:41.659 --> 00:08:44.690 align:middle line:84% So that's not good, and you have lack of control. 00:08:44.690 --> 00:08:47.480 align:middle line:84% You want good selectivity to other exposed films 00:08:47.480 --> 00:08:48.200 align:middle line:90% and to resist. 00:08:48.200 --> 00:08:50.360 align:middle line:84% You don't want the etchant attacking something 00:08:50.360 --> 00:08:52.525 align:middle line:84% that you don't want it to attack. 00:08:52.525 --> 00:08:54.650 align:middle line:84% You'd like the etch to be uniform across the wafer, 00:08:54.650 --> 00:08:56.570 align:middle line:84% so as you etch a trench on the left of the wafer 00:08:56.570 --> 00:08:58.403 align:middle line:84% and the right of the wafer, the trench depth 00:08:58.403 --> 00:09:00.583 align:middle line:90% is exactly the same. 00:09:00.583 --> 00:09:03.000 align:middle line:84% This is an important one that people-- number five, people 00:09:03.000 --> 00:09:04.550 align:middle line:90% often forget. 00:09:04.550 --> 00:09:06.290 align:middle line:84% We don't want to damage the surface, 00:09:06.290 --> 00:09:08.960 align:middle line:84% and we don't want to damage the circuit electrically. 00:09:08.960 --> 00:09:11.810 align:middle line:84% And we are bombarding, in a lot of these plasma etches, 00:09:11.810 --> 00:09:15.320 align:middle line:84% we're bombarding, with hundreds of electron volts, 00:09:15.320 --> 00:09:17.060 align:middle line:90% the surface with ions. 00:09:17.060 --> 00:09:19.768 align:middle line:84% And they those ions can constitute a current 00:09:19.768 --> 00:09:21.560 align:middle line:84% to the wafer, and that current can actually 00:09:21.560 --> 00:09:24.530 align:middle line:84% damage sensitive electronic components. 00:09:24.530 --> 00:09:26.412 align:middle line:84% Obviously, we want the etch to be clean. 00:09:26.412 --> 00:09:28.370 align:middle line:84% We don't want to introduce a lot of impurities. 00:09:28.370 --> 00:09:30.290 align:middle line:84% We don't want to introduce a lot of metals 00:09:30.290 --> 00:09:32.960 align:middle line:90% into the wafer, and safe. 00:09:32.960 --> 00:09:36.470 align:middle line:84% So here's an example, on the bottom part of page 00:09:36.470 --> 00:09:38.750 align:middle line:84% four, what I mean by plasma damage. 00:09:38.750 --> 00:09:41.720 align:middle line:84% Let's just say you have a MOSFET. 00:09:41.720 --> 00:09:45.380 align:middle line:84% You have a field oxide FOx here on the left, another one 00:09:45.380 --> 00:09:47.540 align:middle line:84% on the right, and this thin region in the middle 00:09:47.540 --> 00:09:49.640 align:middle line:84% is meant to represent the gate oxide. 00:09:49.640 --> 00:09:53.155 align:middle line:84% And let's say-- and I have a polysilicon gate 00:09:53.155 --> 00:09:54.030 align:middle line:90% that looks like this. 00:09:54.030 --> 00:09:56.480 align:middle line:84% There's a large contact pad, which we often 00:09:56.480 --> 00:09:58.280 align:middle line:90% have perhaps for contact. 00:09:58.280 --> 00:10:01.243 align:middle line:84% And then it goes down, and it's a very narrow gate, 00:10:01.243 --> 00:10:02.660 align:middle line:84% which is what we were doing today, 00:10:02.660 --> 00:10:04.800 align:middle line:90% say 90 nanometer technology. 00:10:04.800 --> 00:10:06.600 align:middle line:84% Now, I go to put this in a plasma etcher. 00:10:06.600 --> 00:10:09.090 align:middle line:84% I'm etching something on the wafer, maybe not the gate, 00:10:09.090 --> 00:10:10.838 align:middle line:90% but something else. 00:10:10.838 --> 00:10:12.630 align:middle line:84% So I have all these arrows going down here. 00:10:12.630 --> 00:10:16.110 align:middle line:84% It's supposed to represent ions because I have a plasma 00:10:16.110 --> 00:10:18.240 align:middle line:84% current, so I have ions that come down, 00:10:18.240 --> 00:10:21.150 align:middle line:84% strike this-- they strike everywhere, but in particular, 00:10:21.150 --> 00:10:25.020 align:middle line:90% they strike this large area. 00:10:25.020 --> 00:10:28.770 align:middle line:84% Because this is a conductor, this current, 00:10:28.770 --> 00:10:30.430 align:middle line:90% with all these ions-- 00:10:30.430 --> 00:10:33.300 align:middle line:84% the flux of ions to this large area, constitutes a current. 00:10:33.300 --> 00:10:35.580 align:middle line:84% This current can then flow down the conductor, 00:10:35.580 --> 00:10:36.870 align:middle line:90% and where does it go? 00:10:36.870 --> 00:10:38.880 align:middle line:84% Well, if this is a thin oxide, it 00:10:38.880 --> 00:10:40.380 align:middle line:90% can tunnel through that oxide. 00:10:40.380 --> 00:10:43.740 align:middle line:84% So all of a sudden now, I could be forcing a very large current 00:10:43.740 --> 00:10:46.440 align:middle line:84% through a thin oxide, and you can end up 00:10:46.440 --> 00:10:49.090 align:middle line:84% with charges buried in that oxide and damage. 00:10:49.090 --> 00:10:51.300 align:middle line:84% And this is called an antenna because this thing can 00:10:51.300 --> 00:10:55.200 align:middle line:84% collect ions from a large area and funnel them down, 00:10:55.200 --> 00:10:58.095 align:middle line:84% acting like an antenna, and you can do a lot of damage. 00:10:58.095 --> 00:10:59.490 align:middle line:90% You have to be very careful. 00:10:59.490 --> 00:11:00.730 align:middle line:90% Yes? 00:11:00.730 --> 00:11:03.555 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:11:03.555 --> 00:11:04.680 align:middle line:90% JUDY HOYT: Not necessarily. 00:11:04.680 --> 00:11:05.972 align:middle line:90% It depends on what I'm etching. 00:11:05.972 --> 00:11:07.340 align:middle line:90% This may not be the gate etched. 00:11:07.340 --> 00:11:10.940 align:middle line:84% For example, after you do a gate etch, you need to open up-- 00:11:10.940 --> 00:11:14.900 align:middle line:84% you may need to open up some other holes somewhere else. 00:11:14.900 --> 00:11:17.090 align:middle line:84% This could be happening during the gate etch. 00:11:17.090 --> 00:11:19.070 align:middle line:84% It could be happening during another step, 00:11:19.070 --> 00:11:20.960 align:middle line:84% but there are, nevertheless, ions 00:11:20.960 --> 00:11:24.050 align:middle line:84% coming in, in this particular example, into the gate. 00:11:24.050 --> 00:11:27.440 align:middle line:84% And so this may not actually be a gate etch step. 00:11:27.440 --> 00:11:31.490 align:middle line:84% It's just an example where you have the gate exposed, 00:11:31.490 --> 00:11:33.800 align:middle line:90% potentially collecting current. 00:11:33.800 --> 00:11:36.162 align:middle line:90% It can go through the oxide. 00:11:36.162 --> 00:11:38.120 align:middle line:84% So that's something we need to worry about when 00:11:38.120 --> 00:11:39.335 align:middle line:90% we have high density plasmas. 00:11:39.335 --> 00:11:42.670 align:middle line:90% 00:11:42.670 --> 00:11:45.910 align:middle line:84% Slide 5, so let's just talk about some 00:11:45.910 --> 00:11:49.360 align:middle line:84% of the basic concepts, and I'll go through the wet etching 00:11:49.360 --> 00:11:50.513 align:middle line:90% relatively rapidly. 00:11:50.513 --> 00:11:52.430 align:middle line:84% As I said, there's two types of etching-- dry, 00:11:52.430 --> 00:11:54.590 align:middle line:84% which means you use a plasma, and wet, 00:11:54.590 --> 00:11:58.910 align:middle line:84% which means you use a beaker or something in a solution. 00:11:58.910 --> 00:12:00.520 align:middle line:84% So you typically, for wet etching, 00:12:00.520 --> 00:12:04.455 align:middle line:84% you submerge the wafers in some specific bath. 00:12:04.455 --> 00:12:05.830 align:middle line:84% The good thing about wet etching, 00:12:05.830 --> 00:12:07.420 align:middle line:90% it tends to be highly selective. 00:12:07.420 --> 00:12:08.678 align:middle line:90% It's all based on chemistry. 00:12:08.678 --> 00:12:09.970 align:middle line:90% There's no physical components. 00:12:09.970 --> 00:12:12.590 align:middle line:84% You have no bombardment, so as a result, 00:12:12.590 --> 00:12:14.350 align:middle line:84% because it's based on chemistry, you 00:12:14.350 --> 00:12:17.890 align:middle line:84% can select the etchants in the solution 00:12:17.890 --> 00:12:20.860 align:middle line:84% that they will etch exactly what you want. 00:12:20.860 --> 00:12:22.210 align:middle line:90% It's usually isotropic. 00:12:22.210 --> 00:12:25.330 align:middle line:84% There are a couple of what we call crystallographically 00:12:25.330 --> 00:12:26.260 align:middle line:90% dependent etches. 00:12:26.260 --> 00:12:30.100 align:middle line:84% KOH, potassium hydroxide etching silicon is an example of that. 00:12:30.100 --> 00:12:32.410 align:middle line:84% There aren't a whole lot of cases of that, though. 00:12:32.410 --> 00:12:35.230 align:middle line:84% Most of the time, the etches end up being selective. 00:12:35.230 --> 00:12:36.985 align:middle line:90% Here's an example. 00:12:36.985 --> 00:12:40.570 align:middle line:84% Here's a wet etching of SiO2, but by HF. 00:12:40.570 --> 00:12:43.630 align:middle line:90% So you add HF to SiO2. 00:12:43.630 --> 00:12:48.430 align:middle line:84% You end up with some extra water, and the SiO2 goes away. 00:12:48.430 --> 00:12:51.730 align:middle line:84% You can etch silicon in a combination of an oxidizer, 00:12:51.730 --> 00:12:54.370 align:middle line:90% like nitric acid, and HF. 00:12:54.370 --> 00:12:56.590 align:middle line:84% Sometimes, you add acetic, so HF nitric acetic, 00:12:56.590 --> 00:12:57.790 align:middle line:90% but here's an example. 00:12:57.790 --> 00:13:00.700 align:middle line:84% Etching silicon wet etching with nitric NHF, 00:13:00.700 --> 00:13:02.950 align:middle line:84% so the nitric oxidizes the silicon. 00:13:02.950 --> 00:13:05.440 align:middle line:84% The HF strips the oxide, and you end up 00:13:05.440 --> 00:13:07.870 align:middle line:90% with overall etching of silicon. 00:13:07.870 --> 00:13:10.540 align:middle line:84% HF nitric, though, is not that easy to mask 00:13:10.540 --> 00:13:12.528 align:middle line:84% because it also etches any oxide on the wafer. 00:13:12.528 --> 00:13:13.570 align:middle line:90% So you have to watch out. 00:13:13.570 --> 00:13:15.400 align:middle line:84% You'd have to use nitride to mask it, 00:13:15.400 --> 00:13:18.730 align:middle line:84% but those are some examples of wet etches. 00:13:18.730 --> 00:13:21.830 align:middle line:90% 00:13:21.830 --> 00:13:27.230 align:middle line:84% Slide 6, let's talk a little bit about isotropic etching 00:13:27.230 --> 00:13:29.150 align:middle line:90% and undercut. 00:13:29.150 --> 00:13:31.010 align:middle line:84% If we're doing isotropic etching, 00:13:31.010 --> 00:13:34.100 align:middle line:84% as I'm showing here in panel A at the top, 00:13:34.100 --> 00:13:35.570 align:middle line:90% one is supposed to be my mask. 00:13:35.570 --> 00:13:37.430 align:middle line:90% Layer 1 is the mask. 00:13:37.430 --> 00:13:39.260 align:middle line:90% Material 2 is what I'm etching. 00:13:39.260 --> 00:13:44.000 align:middle line:84% It's isotropic, so it's etching vertically and laterally 00:13:44.000 --> 00:13:45.650 align:middle line:90% at the same time. 00:13:45.650 --> 00:13:47.180 align:middle line:84% I'm going to have some undercutting. 00:13:47.180 --> 00:13:50.870 align:middle line:84% And in fact, we can sort of mathematically 00:13:50.870 --> 00:13:53.450 align:middle line:84% express the amount of anisotropy, 00:13:53.450 --> 00:13:59.450 align:middle line:84% how anisotropic is the etch, by defining this number here called 00:13:59.450 --> 00:14:00.920 align:middle line:90% a sub f. 00:14:00.920 --> 00:14:05.180 align:middle line:84% If you define this anisotropy factor, a sub f, 00:14:05.180 --> 00:14:09.050 align:middle line:84% to be 1 minus the lateral dimension that you go, 00:14:09.050 --> 00:14:11.900 align:middle line:84% the lateral in a given time, etched 00:14:11.900 --> 00:14:13.530 align:middle line:90% amount divided by the vertical. 00:14:13.530 --> 00:14:17.690 align:middle line:84% So in my diagram here, this would be 1 minus b over a, 00:14:17.690 --> 00:14:20.300 align:middle line:84% so I etch laterally by distance b-- 00:14:20.300 --> 00:14:22.970 align:middle line:90% I'm sorry, b over d. 00:14:22.970 --> 00:14:26.090 align:middle line:84% Laterally, by distance b and vertically by a distance 00:14:26.090 --> 00:14:30.170 align:middle line:84% d, and so that would be the anisotropy factor. 00:14:30.170 --> 00:14:33.160 align:middle line:84% So this anisotropy factor is 0 if you're 00:14:33.160 --> 00:14:35.710 align:middle line:84% isotropic etching because then b equals 00:14:35.710 --> 00:14:39.550 align:middle line:84% d where r lat equals our vertical, and then af goes to 0. 00:14:39.550 --> 00:14:46.210 align:middle line:84% If you're totally anisotropic, this thing basically-- 00:14:46.210 --> 00:14:48.970 align:middle line:84% b goes to 0 because you don't etch laterally at all, 00:14:48.970 --> 00:14:52.600 align:middle line:90% and this number goes to 1. 00:14:52.600 --> 00:14:54.940 align:middle line:84% Now, typically, if you're etching a film, 00:14:54.940 --> 00:14:57.790 align:middle line:84% what's a practical thing you need to think about? 00:14:57.790 --> 00:14:59.350 align:middle line:84% Well, you need to know the etch rate. 00:14:59.350 --> 00:15:00.045 align:middle line:90% That's fine. 00:15:00.045 --> 00:15:01.420 align:middle line:84% Let's say you know the etch rate, 00:15:01.420 --> 00:15:03.250 align:middle line:84% and I give you the film things, you 00:15:03.250 --> 00:15:04.992 align:middle line:84% can calculate how long you need to etch. 00:15:04.992 --> 00:15:06.950 align:middle line:84% You just divide the thickness by the etch rate, 00:15:06.950 --> 00:15:08.020 align:middle line:90% and that gives you the time. 00:15:08.020 --> 00:15:09.490 align:middle line:84% But that's not the time you should 00:15:09.490 --> 00:15:11.470 align:middle line:90% use for the etch in practice. 00:15:11.470 --> 00:15:14.840 align:middle line:84% In practice, you almost always do something called overetching. 00:15:14.840 --> 00:15:17.470 align:middle line:84% So you typically leave it in the bath 00:15:17.470 --> 00:15:22.080 align:middle line:84% longer than what you would need to etch the film thickness. 00:15:22.080 --> 00:15:23.220 align:middle line:90% Why is that? 00:15:23.220 --> 00:15:25.620 align:middle line:84% Well, basically, you need to ensure that you actually 00:15:25.620 --> 00:15:28.080 align:middle line:84% do completely remove the film because the film-- 00:15:28.080 --> 00:15:30.368 align:middle line:84% if I tell you the film is 1 micron, first of all, 00:15:30.368 --> 00:15:31.410 align:middle line:90% you shouldn't believe me. 00:15:31.410 --> 00:15:32.640 align:middle line:84% I may not know what I'm talking about. 00:15:32.640 --> 00:15:33.598 align:middle line:90% It may not be a micron. 00:15:33.598 --> 00:15:36.480 align:middle line:84% And even if I do know what I'm talking about, films in reality 00:15:36.480 --> 00:15:38.760 align:middle line:84% are never exactly one thickness across the wafer. 00:15:38.760 --> 00:15:40.390 align:middle line:90% There's always variation. 00:15:40.390 --> 00:15:42.705 align:middle line:84% So you may have a slight high spot in the center 00:15:42.705 --> 00:15:43.830 align:middle line:90% or a high spot at the etch. 00:15:43.830 --> 00:15:45.690 align:middle line:90% Who knows? 00:15:45.690 --> 00:15:47.550 align:middle line:84% So you almost always do overetching, 00:15:47.550 --> 00:15:50.130 align:middle line:84% and that's what we're picturing in the upper right. 00:15:50.130 --> 00:15:53.250 align:middle line:84% What we've done here is, in showing you 00:15:53.250 --> 00:15:55.440 align:middle line:84% the contour of what the etched region looks 00:15:55.440 --> 00:15:57.210 align:middle line:84% like for different times, so this is 00:15:57.210 --> 00:15:58.750 align:middle line:90% after a certain amount of time. 00:15:58.750 --> 00:16:01.590 align:middle line:84% This other dashed line is a slightly longer time, 00:16:01.590 --> 00:16:03.390 align:middle line:84% and now you've already broken through. 00:16:03.390 --> 00:16:05.730 align:middle line:84% You've hit that, but maybe somewhere else on the wafer, 00:16:05.730 --> 00:16:07.320 align:middle line:84% you haven't quite etched the film. 00:16:07.320 --> 00:16:09.270 align:middle line:84% So you go a little bit longer, and you end up 00:16:09.270 --> 00:16:12.610 align:middle line:84% with the profile shown in the solid line here. 00:16:12.610 --> 00:16:14.490 align:middle line:84% So that's because we have done what 00:16:14.490 --> 00:16:16.710 align:middle line:84% we call overetching, and overetching 00:16:16.710 --> 00:16:19.170 align:middle line:84% has to be done in any practical process, 00:16:19.170 --> 00:16:21.970 align:middle line:84% either in research or in manufacturing, 00:16:21.970 --> 00:16:24.605 align:middle line:84% to take into account variations in not 00:16:24.605 --> 00:16:26.980 align:middle line:84% only the film thickness, but variations in the etch rate. 00:16:26.980 --> 00:16:29.500 align:middle line:84% The etch rate is not exactly uniform necessarily, 00:16:29.500 --> 00:16:34.870 align:middle line:84% depending on the etcher or even depending on the solution. 00:16:34.870 --> 00:16:37.840 align:middle line:84% For wet etching, the selectivity is usually excellent. 00:16:37.840 --> 00:16:42.070 align:middle line:84% I mean, we usually choose our etchants 00:16:42.070 --> 00:16:45.820 align:middle line:84% so that we get good selectivity, and again, selectivity 00:16:45.820 --> 00:16:49.720 align:middle line:84% is defined as the ratio of the etch rate of the film we want 00:16:49.720 --> 00:16:52.510 align:middle line:84% to etch divided by the etch rate of the film we 00:16:52.510 --> 00:16:54.580 align:middle line:90% don't want to etch. 00:16:54.580 --> 00:16:59.880 align:middle line:84% And chemical reactions tend to be very, very selective. 00:16:59.880 --> 00:17:03.960 align:middle line:84% So just to illustrate this idea of overetch and selectivity, 00:17:03.960 --> 00:17:06.150 align:middle line:84% I wanted to go through a quick example 00:17:06.150 --> 00:17:08.579 align:middle line:84% that I took from your text, and that's 00:17:08.579 --> 00:17:11.010 align:middle line:90% what's shown on slide 7 and 8. 00:17:11.010 --> 00:17:15.150 align:middle line:84% Here's an example of a question that you can consider. 00:17:15.150 --> 00:17:17.339 align:middle line:84% You have a relatively simple setup. 00:17:17.339 --> 00:17:21.119 align:middle line:84% You have a half micron layer of silicon dioxide on a silicon 00:17:21.119 --> 00:17:26.780 align:middle line:84% substrate, and you need to etch it down to the silicon. 00:17:26.780 --> 00:17:28.700 align:middle line:84% Now, you assumed that the nominal etch 00:17:28.700 --> 00:17:32.070 align:middle line:84% rate-- so you're putting this in some etcher, whatever it is. 00:17:32.070 --> 00:17:35.150 align:middle line:84% Let's say it's not HF, but it's some other type of etch. 00:17:35.150 --> 00:17:37.820 align:middle line:84% Has a nominal etch rate of r sub Ox, 00:17:37.820 --> 00:17:40.070 align:middle line:84% so that's a certain number of microns per minute 00:17:40.070 --> 00:17:44.540 align:middle line:84% that you've been told that the oxide etch is in this etchant. 00:17:44.540 --> 00:17:47.720 align:middle line:84% But there is about a plus or minus 5% variation 00:17:47.720 --> 00:17:48.770 align:middle line:90% in the oxide thickness. 00:17:48.770 --> 00:17:50.090 align:middle line:90% The oxide's not perfect. 00:17:50.090 --> 00:17:53.900 align:middle line:84% It was grown in a tool that gives you 00:17:53.900 --> 00:17:56.450 align:middle line:90% a 5% variation across the wafer. 00:17:56.450 --> 00:17:59.270 align:middle line:84% There's also a plus or minus 5% variation 00:17:59.270 --> 00:18:00.680 align:middle line:90% in the oxide etch rate. 00:18:00.680 --> 00:18:02.570 align:middle line:84% You're putting it in a solution, or let's say 00:18:02.570 --> 00:18:04.910 align:middle line:84% you're using a plasma etcher, it's not perfect, 00:18:04.910 --> 00:18:09.520 align:middle line:84% so it varies plus or minus 5% across the surface of the wafer. 00:18:09.520 --> 00:18:12.010 align:middle line:84% So the first thing we want to figure out 00:18:12.010 --> 00:18:15.260 align:middle line:84% is how much of an overetch is required, 00:18:15.260 --> 00:18:17.200 align:middle line:84% and we usually represent overetch 00:18:17.200 --> 00:18:19.310 align:middle line:90% as a percent of the total time. 00:18:19.310 --> 00:18:23.830 align:middle line:84% So if I'm going to etch for a minute, to get rid of the film, 00:18:23.830 --> 00:18:26.140 align:middle line:84% and then leave it in an extra minute 00:18:26.140 --> 00:18:29.050 align:middle line:84% to make sure that I've taken care of non-uniformities, 00:18:29.050 --> 00:18:31.210 align:middle line:84% That's called 100% overetch because it's 00:18:31.210 --> 00:18:33.040 align:middle line:90% 100% of the main etch time. 00:18:33.040 --> 00:18:37.210 align:middle line:84% I'm contributing 100% of that to the overetch step. 00:18:37.210 --> 00:18:38.710 align:middle line:84% So we want to know how much overetch 00:18:38.710 --> 00:18:40.720 align:middle line:84% is required as a percent in order 00:18:40.720 --> 00:18:42.700 align:middle line:84% to make sure that all the oxide on the wafer 00:18:42.700 --> 00:18:47.450 align:middle line:84% that you etched all the way down at every point in the wafer. 00:18:47.450 --> 00:18:49.460 align:middle line:84% So let's just go through part a first 00:18:49.460 --> 00:18:52.010 align:middle line:84% and how we would think about that. 00:18:52.010 --> 00:18:54.890 align:middle line:84% Well, I've just drawn a little cartoon down here. 00:18:54.890 --> 00:18:57.260 align:middle line:84% Basically, this green material is supposed 00:18:57.260 --> 00:18:58.250 align:middle line:90% to represent the oxide. 00:18:58.250 --> 00:19:01.070 align:middle line:84% It's nominally half micron on the wafer, 00:19:01.070 --> 00:19:04.820 align:middle line:84% but at the thickest point, it's thicker than that. 00:19:04.820 --> 00:19:10.030 align:middle line:84% It's 5% thicker, so it's 0.525 microns. 00:19:10.030 --> 00:19:14.830 align:middle line:84% So basically, what we want to do is, 00:19:14.830 --> 00:19:17.230 align:middle line:84% the overetch has to be done to make sure 00:19:17.230 --> 00:19:20.380 align:middle line:84% that you remove the oxide at the thickest point for the slowest 00:19:20.380 --> 00:19:22.670 align:middle line:84% etch rate, so you take a worst case. 00:19:22.670 --> 00:19:25.990 align:middle line:84% So I take the thickest point is 0.525, 00:19:25.990 --> 00:19:30.050 align:middle line:84% and then the slowest etch rate is 0.95 times the nominal etch 00:19:30.050 --> 00:19:30.550 align:middle line:90% rate. 00:19:30.550 --> 00:19:35.380 align:middle line:84% It's 5% slower than the nominal etch rate. 00:19:35.380 --> 00:19:39.370 align:middle line:84% So the time to etch this, worst case, the thickest region, 00:19:39.370 --> 00:19:41.200 align:middle line:84% assuming you had the slowest etch rate, 00:19:41.200 --> 00:19:45.310 align:middle line:84% is just this 0.525 divided by the slowest etch 00:19:45.310 --> 00:19:49.640 align:middle line:84% rate, so divided by rOx times 0.95. 00:19:49.640 --> 00:19:54.820 align:middle line:84% So the worst case etch time would be 0.553 divided 00:19:54.820 --> 00:19:57.710 align:middle line:90% by rOx minutes. 00:19:57.710 --> 00:20:02.170 align:middle line:84% So now, again, we express the overetch as a percent in time. 00:20:02.170 --> 00:20:03.670 align:middle line:84% So that's just going to be the worst 00:20:03.670 --> 00:20:05.900 align:middle line:84% case etch time, which we just calculated, 00:20:05.900 --> 00:20:07.840 align:middle line:84% divided by the nominal etch time. 00:20:07.840 --> 00:20:09.670 align:middle line:84% The nominal etch time would be the time 00:20:09.670 --> 00:20:12.280 align:middle line:84% to remove half micron at the nominal etch rate. 00:20:12.280 --> 00:20:16.720 align:middle line:84% So the nominal etch time is given by 0.5 divided by rOx. 00:20:16.720 --> 00:20:19.960 align:middle line:84% So we just take the ratio of the worst case etch, the longest 00:20:19.960 --> 00:20:23.930 align:middle line:84% etch, to the nominal etch, and we get a 1.1, 00:20:23.930 --> 00:20:31.250 align:middle line:84% so that's about a 10% overetch, which isn't bad. 00:20:31.250 --> 00:20:36.020 align:middle line:84% 10%-- to do a 10% overetch these days is pretty darn good. 00:20:36.020 --> 00:20:39.890 align:middle line:84% Most etch-- most films have a greater variation 00:20:39.890 --> 00:20:43.080 align:middle line:84% than plus or minus 5% in a lot of plasma tools. 00:20:43.080 --> 00:20:45.990 align:middle line:84% Plus or minus 5% variation is conservative, 00:20:45.990 --> 00:20:51.230 align:middle line:84% so it's not unusual to see overetches of that length. 00:20:51.230 --> 00:20:54.650 align:middle line:90% Now, the second thing-- 00:20:54.650 --> 00:20:57.500 align:middle line:84% basically, I'm leaving that wafer in the plasma etcher 00:20:57.500 --> 00:21:00.260 align:middle line:84% or in the solution 10% longer than I need 00:21:00.260 --> 00:21:04.640 align:middle line:90% to, to remove the nominal film. 00:21:04.640 --> 00:21:06.740 align:middle line:84% The second question is, now we know 00:21:06.740 --> 00:21:11.570 align:middle line:84% that, what selectivity of the oxide etch rate to the silicon 00:21:11.570 --> 00:21:13.910 align:middle line:84% etch rate is required so that, at most, 00:21:13.910 --> 00:21:18.830 align:middle line:84% we remove 5 nanometers of silicon? 00:21:18.830 --> 00:21:20.990 align:middle line:84% Assuming we do the overetch that we calculate, 00:21:20.990 --> 00:21:24.120 align:middle line:84% assuming we do a 10% overetch in part a, 00:21:24.120 --> 00:21:26.480 align:middle line:84% so what we're asking here is this. 00:21:26.480 --> 00:21:29.630 align:middle line:84% I'm going to be etching down, and in some regions 00:21:29.630 --> 00:21:32.300 align:middle line:84% where the oxide is thinner, and because I have an overetch, 00:21:32.300 --> 00:21:34.490 align:middle line:84% I'm going to actually be exposing the silicon 00:21:34.490 --> 00:21:36.740 align:middle line:90% surface to the etchant. 00:21:36.740 --> 00:21:39.530 align:middle line:84% And it has a finite etch rate, but I don't want to remove-- 00:21:39.530 --> 00:21:41.488 align:middle line:84% I want to make sure I don't go into the silicon 00:21:41.488 --> 00:21:43.200 align:middle line:90% by any more than 5 nanometers. 00:21:43.200 --> 00:21:46.310 align:middle line:84% So that sets a requirement on my selectivity. 00:21:46.310 --> 00:21:48.440 align:middle line:84% So the selectivity is how etch how fast am I 00:21:48.440 --> 00:21:51.050 align:middle line:84% etching through the oxide divided by how fast I'm 00:21:51.050 --> 00:21:52.310 align:middle line:90% etching through the silicon. 00:21:52.310 --> 00:21:54.830 align:middle line:84% If that selectivity is too low, if it's 1 to 1, 00:21:54.830 --> 00:21:57.020 align:middle line:84% then during certain regions during the overetch, 00:21:57.020 --> 00:21:59.330 align:middle line:84% you'll be zipping through that silicon 00:21:59.330 --> 00:22:01.980 align:middle line:84% substrate at the same rate that you're etching here. 00:22:01.980 --> 00:22:03.680 align:middle line:84% So obviously-- and this is exactly 00:22:03.680 --> 00:22:05.930 align:middle line:84% the kind of practical calculation you'll need to know. 00:22:05.930 --> 00:22:07.930 align:middle line:84% If you have to say, oh, I don't want to attack-- 00:22:07.930 --> 00:22:09.890 align:middle line:84% I don't want to eat away the source drains when 00:22:09.890 --> 00:22:11.810 align:middle line:90% I'm etching the contact holes. 00:22:11.810 --> 00:22:13.790 align:middle line:84% Well, how much selectivity do I need 00:22:13.790 --> 00:22:15.830 align:middle line:84% if I'm only allowed to eat away, say, 00:22:15.830 --> 00:22:20.750 align:middle line:84% 5 nanometers, something like that, given a 10% overetch? 00:22:20.750 --> 00:22:24.980 align:middle line:84% So if we go to slide 8, uh-oh, we 00:22:24.980 --> 00:22:27.110 align:middle line:84% can think through that here in solution b. 00:22:27.110 --> 00:22:30.800 align:middle line:84% So what's the most amount of silicon you would remove? 00:22:30.800 --> 00:22:35.885 align:middle line:84% The worst case will occur under the thinnest oxide being 00:22:35.885 --> 00:22:37.010 align:middle line:90% etched at the fastest rate. 00:22:37.010 --> 00:22:39.440 align:middle line:84% That's the worst case, so I have the thinnest oxide on the wafer, 00:22:39.440 --> 00:22:41.360 align:middle line:84% and let's say, at that point, it happens 00:22:41.360 --> 00:22:44.300 align:middle line:84% to be etching the oxide at the very fastest rate. 00:22:44.300 --> 00:22:46.730 align:middle line:84% That gives you the most amount-- the quickest 00:22:46.730 --> 00:22:48.200 align:middle line:90% that you'll clear the oxide. 00:22:48.200 --> 00:22:51.020 align:middle line:84% At that very point in time then, the etchant 00:22:51.020 --> 00:22:53.200 align:middle line:90% starts attacking the silicon. 00:22:53.200 --> 00:22:58.120 align:middle line:84% So the thinnest oxide is 0.95 times half micron, 00:22:58.120 --> 00:23:01.000 align:middle line:90% so it's 0.475 microns. 00:23:01.000 --> 00:23:05.210 align:middle line:84% The fastest etch rate is 5% faster than the nominal rOx, 00:23:05.210 --> 00:23:07.610 align:middle line:90% so that's 1.05 times our rOx. 00:23:07.610 --> 00:23:10.960 align:middle line:84% So the time to etch through that oxide 00:23:10.960 --> 00:23:14.770 align:middle line:84% is 0.475, which is the thickness, divided 00:23:14.770 --> 00:23:16.360 align:middle line:90% by the fastest etch rate. 00:23:16.360 --> 00:23:20.570 align:middle line:84% So it's just this number here, 0.4523 divided by rOx. 00:23:20.570 --> 00:23:23.800 align:middle line:84% So after that amount of minutes, all of a sudden, 00:23:23.800 --> 00:23:28.240 align:middle line:84% in that one part of the wafer, the silicon is exposed. 00:23:28.240 --> 00:23:33.180 align:middle line:84% So now, the time that the silicone is exposed to the etch 00:23:33.180 --> 00:23:37.110 align:middle line:84% is equal to the total time, which 00:23:37.110 --> 00:23:38.610 align:middle line:84% is this-- what we calculated before, 00:23:38.610 --> 00:23:43.010 align:middle line:84% 0.553 over rOx in part a, minus the time it took 00:23:43.010 --> 00:23:44.743 align:middle line:90% to etch the thinnest oxide. 00:23:44.743 --> 00:23:46.160 align:middle line:84% So that's the total amount of time 00:23:46.160 --> 00:23:52.100 align:middle line:84% then that the silicon itself would receive the etch. 00:23:52.100 --> 00:23:54.920 align:middle line:84% Assume it's etching at the fastest rate. 00:23:54.920 --> 00:23:56.780 align:middle line:84% So the time the silicon is exposed 00:23:56.780 --> 00:23:59.630 align:middle line:84% is just what we had from part a, the total etch time, 00:23:59.630 --> 00:24:03.080 align:middle line:84% 0.553 over rOx minus the time it took 00:24:03.080 --> 00:24:06.890 align:middle line:84% us to get the silicon opened up in the worst case. 00:24:06.890 --> 00:24:10.580 align:middle line:84% So we end up with a silicon exposure time of 0.1 00:24:10.580 --> 00:24:14.230 align:middle line:90% divided by rOx in minutes. 00:24:14.230 --> 00:24:16.230 align:middle line:84% That's how long the silicone, in the worst case, 00:24:16.230 --> 00:24:17.400 align:middle line:90% would be exposed. 00:24:17.400 --> 00:24:19.830 align:middle line:84% So now, we know what we have to calculate. 00:24:19.830 --> 00:24:22.860 align:middle line:84% We want to etch only 5 nanometers, which 00:24:22.860 --> 00:24:26.338 align:middle line:90% is 0.005 microns of silicon. 00:24:26.338 --> 00:24:28.630 align:middle line:84% That gives me an idea of what the etch rate of silicon, 00:24:28.630 --> 00:24:29.950 align:middle line:90% the maximum, can be. 00:24:29.950 --> 00:24:32.740 align:middle line:84% It's going to be that 0.005 divided 00:24:32.740 --> 00:24:36.290 align:middle line:84% by the time it's exposed, which we just calculated. 00:24:36.290 --> 00:24:39.520 align:middle line:84% So the maximum etch rate of silicon ends up-- 00:24:39.520 --> 00:24:43.040 align:middle line:84% do the math, ends up to be 0.05 times 00:24:43.040 --> 00:24:48.850 align:middle line:84% rOx microns per minute, so this is always in terms of rOx. 00:24:48.850 --> 00:24:51.530 align:middle line:84% And so now I can calculate-- given this, 00:24:51.530 --> 00:24:54.310 align:middle line:84% I can calculate the selectivity of oxide etch rate 00:24:54.310 --> 00:24:55.940 align:middle line:84% to silicon, just like we did before. 00:24:55.940 --> 00:24:58.450 align:middle line:84% So I just take this rate rOx divided 00:24:58.450 --> 00:25:02.260 align:middle line:84% by this etch rate of the silicon and divide it out, 00:25:02.260 --> 00:25:04.640 align:middle line:84% and you end up with a rate-- the rOx's go out, 00:25:04.640 --> 00:25:07.810 align:middle line:84% you end up with this ratio has to be 20 to 1 or better. 00:25:07.810 --> 00:25:08.860 align:middle line:90% Yeah? 00:25:08.860 --> 00:25:10.030 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:25:10.030 --> 00:25:13.365 align:middle line:90% 00:25:13.365 --> 00:25:14.240 align:middle line:90% JUDY HOYT: Which one? 00:25:14.240 --> 00:25:15.830 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:25:15.830 --> 00:25:18.900 align:middle line:90% 00:25:18.900 --> 00:25:22.740 align:middle line:84% JUDY HOYT: Oh, yeah, the time silicon is exposed? 00:25:22.740 --> 00:25:24.430 align:middle line:90% All right, well, that's fine. 00:25:24.430 --> 00:25:27.480 align:middle line:84% You can do the math to kind of get the idea. 00:25:27.480 --> 00:25:32.410 align:middle line:84% But what this says is, any time you need to do an etch, 00:25:32.410 --> 00:25:35.680 align:middle line:84% you typically have a maximum requirement on the film 00:25:35.680 --> 00:25:38.150 align:middle line:84% below that you can remove, and then you 00:25:38.150 --> 00:25:39.650 align:middle line:84% have a certain overetch requirement, 00:25:39.650 --> 00:25:42.040 align:middle line:84% which is set by non-uniformities and other things that 00:25:42.040 --> 00:25:44.080 align:middle line:90% can be set by topography. 00:25:44.080 --> 00:25:46.270 align:middle line:84% That determines what kind of selectivity you need, 00:25:46.270 --> 00:25:48.723 align:middle line:84% so knowing what kind of selectivity you need, 00:25:48.723 --> 00:25:50.140 align:middle line:84% you can go out and figure out what 00:25:50.140 --> 00:25:55.090 align:middle line:84% etch it is you need to find in the plasma etcher. 00:25:55.090 --> 00:25:57.350 align:middle line:84% So typical selectivity requirements, 00:25:57.350 --> 00:26:00.700 align:middle line:84% we'll talk about this later next time, but for gate etch, 00:26:00.700 --> 00:26:03.850 align:middle line:84% typically you need 50 to 1, so something like that. 00:26:03.850 --> 00:26:06.850 align:middle line:84% 50 times faster etching the polysilicon 00:26:06.850 --> 00:26:09.040 align:middle line:84% compared to etching the oxide, and it can even 00:26:09.040 --> 00:26:11.410 align:middle line:84% be higher than that depending on topography and other 00:26:11.410 --> 00:26:12.310 align:middle line:90% requirements. 00:26:12.310 --> 00:26:13.550 align:middle line:90% That's a simple example. 00:26:13.550 --> 00:26:15.370 align:middle line:84% It just gives you an idea of how one 00:26:15.370 --> 00:26:18.790 align:middle line:84% thinks through these types of problems. 00:26:18.790 --> 00:26:24.360 align:middle line:90% 00:26:24.360 --> 00:26:27.090 align:middle line:84% So one problem that we haven't talked about too much here, 00:26:27.090 --> 00:26:32.880 align:middle line:84% but shown on page 9, slide 9, is something called mask erosion. 00:26:32.880 --> 00:26:35.190 align:middle line:84% The fact that, during the etch, the mask 00:26:35.190 --> 00:26:39.400 align:middle line:84% itself may get etched, not just the layer underneath. 00:26:39.400 --> 00:26:42.150 align:middle line:84% So the mask can be eroded during the etch, 00:26:42.150 --> 00:26:45.480 align:middle line:84% and this can happen in both isotropic and anisotropic 00:26:45.480 --> 00:26:47.130 align:middle line:90% etching. 00:26:47.130 --> 00:26:51.090 align:middle line:84% And so in this example, a, you can 00:26:51.090 --> 00:26:54.780 align:middle line:84% see that the mask was eroded by a certain width 00:26:54.780 --> 00:26:57.270 align:middle line:84% or etched off delta m here, and that 00:26:57.270 --> 00:26:59.980 align:middle line:84% can happen if you're doing isotropic etching or anisotropic 00:26:59.980 --> 00:27:00.480 align:middle line:90% etching. 00:27:00.480 --> 00:27:04.447 align:middle line:84% Again, mask erosion is bad because the feature size 00:27:04.447 --> 00:27:06.780 align:middle line:84% that you get ends up being different from what you would 00:27:06.780 --> 00:27:10.020 align:middle line:84% wanted on the mask, and it can even 00:27:10.020 --> 00:27:13.050 align:middle line:84% end up affecting the shape sometimes of the feature 00:27:13.050 --> 00:27:14.280 align:middle line:90% that you get. 00:27:14.280 --> 00:27:17.780 align:middle line:84% So mask erosion is something we want to avoid. 00:27:17.780 --> 00:27:20.720 align:middle line:84% If you want to know about chemical etchants, 00:27:20.720 --> 00:27:24.530 align:middle line:84% you can look in table 10.1 in your text, 00:27:24.530 --> 00:27:26.450 align:middle line:84% and there are some common ones listed. 00:27:26.450 --> 00:27:29.180 align:middle line:84% There are websites you can go to for common etches, 00:27:29.180 --> 00:27:32.150 align:middle line:84% but again, as I mentioned, because they're isotropic, 00:27:32.150 --> 00:27:36.260 align:middle line:84% what chemical etches are very rarely used in mainstream IC, 00:27:36.260 --> 00:27:39.680 align:middle line:84% maybe for manufacturing some MEMS devices or some very 00:27:39.680 --> 00:27:41.747 align:middle line:84% specialized devices, but for mainstream CMOS, 00:27:41.747 --> 00:27:42.830 align:middle line:90% you don't typically find-- 00:27:42.830 --> 00:27:46.250 align:middle line:84% The only wet etch that you typically do find is HF, 00:27:46.250 --> 00:27:48.500 align:middle line:84% but that's not used for pattern transfer. 00:27:48.500 --> 00:27:52.430 align:middle line:84% HF is usually used as a dip during the RCA clean to remove 00:27:52.430 --> 00:27:53.960 align:middle line:90% a very thin layer of oxide. 00:27:53.960 --> 00:27:55.103 align:middle line:90% But that's not patterning. 00:27:55.103 --> 00:27:56.645 align:middle line:84% That's just cleaning off the surface. 00:27:56.645 --> 00:27:59.560 align:middle line:90% 00:27:59.560 --> 00:28:02.740 align:middle line:84% So we're going to spend most of the discussion of chapter 10 00:28:02.740 --> 00:28:04.990 align:middle line:84% then on this plasma etching, or this gas etching. 00:28:04.990 --> 00:28:08.515 align:middle line:90% 00:28:08.515 --> 00:28:11.060 align:middle line:84% It's fast and simple in most cases, 00:28:11.060 --> 00:28:13.840 align:middle line:84% and the real benefit is it's more directional. 00:28:13.840 --> 00:28:16.870 align:middle line:90% You can get anisotropic etching. 00:28:16.870 --> 00:28:20.180 align:middle line:84% Plasma etching can have both chemical effects, 00:28:20.180 --> 00:28:23.770 align:middle line:84% so very highly reactive species can be created in the plasma, 00:28:23.770 --> 00:28:26.500 align:middle line:84% and it can also have ionic effects, which 00:28:26.500 --> 00:28:29.290 align:middle line:84% would be directional so a sputtering action 00:28:29.290 --> 00:28:32.120 align:middle line:84% as you accelerate ions towards the surface. 00:28:32.120 --> 00:28:33.700 align:middle line:84% So both of these can play a role, 00:28:33.700 --> 00:28:36.220 align:middle line:84% and depending on their ratio of the chemical effect 00:28:36.220 --> 00:28:37.870 align:middle line:84% to the physical effect, you'll either 00:28:37.870 --> 00:28:42.190 align:middle line:90% get less or more anisotropy. 00:28:42.190 --> 00:28:44.380 align:middle line:84% Here's an example of a plasma system. 00:28:44.380 --> 00:28:47.320 align:middle line:84% This should remind you very much of the plasma enhanced CBD. 00:28:47.320 --> 00:28:49.480 align:middle line:84% The only difference is we're not depositing. 00:28:49.480 --> 00:28:52.700 align:middle line:84% We're actually removing material. 00:28:52.700 --> 00:28:55.690 align:middle line:84% Here's a typical RF powered system. 00:28:55.690 --> 00:28:59.680 align:middle line:84% Here, you have a gas coming in, and you have an etchant gas, 00:28:59.680 --> 00:29:04.360 align:middle line:84% for example, could be CF4 or O2 or some combination coming in, 00:29:04.360 --> 00:29:08.560 align:middle line:84% creates ions in the plasma that are highly reactive that 00:29:08.560 --> 00:29:13.570 align:middle line:84% can go down and end up etching off species 00:29:13.570 --> 00:29:14.665 align:middle line:90% from the wafer surface. 00:29:14.665 --> 00:29:17.750 align:middle line:90% 00:29:17.750 --> 00:29:19.812 align:middle line:84% Here's just an example shown on slide 11. 00:29:19.812 --> 00:29:21.520 align:middle line:84% Again, this should be very much analogous 00:29:21.520 --> 00:29:26.080 align:middle line:84% to what you read about in chapter 9 on plasma enhanced 00:29:26.080 --> 00:29:27.700 align:middle line:90% deposition. 00:29:27.700 --> 00:29:29.530 align:middle line:84% Again, we have-- when we have a plasma, 00:29:29.530 --> 00:29:31.990 align:middle line:84% the plasma tends to self biased because 00:29:31.990 --> 00:29:34.930 align:middle line:84% of the differences between the mobility of ions and electrons. 00:29:34.930 --> 00:29:37.600 align:middle line:84% It tends to be self biased to be-- the potential to plasma 00:29:37.600 --> 00:29:40.190 align:middle line:90% tends to be positive. 00:29:40.190 --> 00:29:45.040 align:middle line:84% And so also, the smaller electrode 00:29:45.040 --> 00:29:48.670 align:middle line:84% tends to have a higher voltage drop near it to maintain 00:29:48.670 --> 00:29:51.070 align:middle line:84% current continuity, so the trick in the plasma 00:29:51.070 --> 00:29:53.020 align:middle line:84% is you put the-- if you're doing etching, 00:29:53.020 --> 00:29:57.100 align:middle line:84% you put the wafer here, the target here, 00:29:57.100 --> 00:29:58.580 align:middle line:90% to be the smaller electrode. 00:29:58.580 --> 00:30:02.695 align:middle line:84% There's a large voltage drop then. 00:30:02.695 --> 00:30:04.320 align:middle line:84% On the electrode towards the left here. 00:30:04.320 --> 00:30:07.200 align:middle line:84% This voltage drop will accelerate the ions towards it, 00:30:07.200 --> 00:30:08.430 align:middle line:90% and you can get etching. 00:30:08.430 --> 00:30:10.038 align:middle line:84% If you're doing sputter deposition, 00:30:10.038 --> 00:30:11.080 align:middle line:90% you do just the opposite. 00:30:11.080 --> 00:30:13.560 align:middle line:84% You put the wafer over here on the large electrode. 00:30:13.560 --> 00:30:15.118 align:middle line:84% You sputter away the aluminum target, 00:30:15.118 --> 00:30:17.160 align:middle line:84% and it goes to the wafer, but I'm trying-- again, 00:30:17.160 --> 00:30:20.260 align:middle line:84% now, I'm trying to etch material on the wafer, 00:30:20.260 --> 00:30:22.560 align:middle line:84% so you put the wafer on the smaller electrode 00:30:22.560 --> 00:30:25.050 align:middle line:90% during RF plasma etching. 00:30:25.050 --> 00:30:27.770 align:middle line:90% 00:30:27.770 --> 00:30:31.460 align:middle line:84% Slide 12 shows some typical reactions and species that 00:30:31.460 --> 00:30:33.080 align:middle line:90% might be present in a plasma. 00:30:33.080 --> 00:30:36.770 align:middle line:84% For plasma etching, a particular example, assuming we're 00:30:36.770 --> 00:30:40.250 align:middle line:84% flowing CF4 which is a very common species used 00:30:40.250 --> 00:30:42.200 align:middle line:90% to do etching. 00:30:42.200 --> 00:30:45.050 align:middle line:84% Here, there are different types of processes. 00:30:45.050 --> 00:30:48.210 align:middle line:84% You can get ionization, dissociation. 00:30:48.210 --> 00:30:52.680 align:middle line:84% So the CF4 can be banging around into electrons, 00:30:52.680 --> 00:30:58.040 align:middle line:84% and you can dissociate it into a free radical, CF3, and free 00:30:58.040 --> 00:31:02.120 align:middle line:84% fluorine, both of which can be highly reactive. 00:31:02.120 --> 00:31:04.880 align:middle line:84% And then this, you can have ionization of those species 00:31:04.880 --> 00:31:09.390 align:middle line:84% to create CF3 ions or fluorine ions, 00:31:09.390 --> 00:31:12.500 align:middle line:84% so you have all these different both neutral and ionized 00:31:12.500 --> 00:31:16.200 align:middle line:84% and very reactive species present. 00:31:16.200 --> 00:31:17.700 align:middle line:84% Typically, we have something like 10 00:31:17.700 --> 00:31:21.240 align:middle line:84% to the 15th per cubic centimeter of neutrals. 00:31:21.240 --> 00:31:24.070 align:middle line:84% 1% to 10% of those neutrals may be free radicals. 00:31:24.070 --> 00:31:25.890 align:middle line:84% Remember, free radical is highly reactive. 00:31:25.890 --> 00:31:29.370 align:middle line:84% It's got an unsatisfied bond, and so it's 00:31:29.370 --> 00:31:30.540 align:middle line:90% a highly reactive species. 00:31:30.540 --> 00:31:33.308 align:middle line:90% It's very important for etching. 00:31:33.308 --> 00:31:35.100 align:middle line:84% And then you have a certain number of, say, 00:31:35.100 --> 00:31:38.110 align:middle line:84% 10 of the 12 or so per cubic centimeter ions and electrons. 00:31:38.110 --> 00:31:42.770 align:middle line:84% This isn't a standard parallel plate plasma system. 00:31:42.770 --> 00:31:45.590 align:middle line:84% In a standard system like this, just two parallel plates 00:31:45.590 --> 00:31:48.890 align:middle line:84% with an RF voltage, the plasma density, 00:31:48.890 --> 00:31:55.100 align:middle line:84% so the number of these reactive species per unit volume, 00:31:55.100 --> 00:31:57.860 align:middle line:84% is usually determined by the voltage drop, the sheath 00:31:57.860 --> 00:31:59.760 align:middle line:90% voltage, so to speak. 00:31:59.760 --> 00:32:03.050 align:middle line:84% So increasing the power of the plasma 00:32:03.050 --> 00:32:04.830 align:middle line:90% increases the plasma density. 00:32:04.830 --> 00:32:06.650 align:middle line:84% So you more species and a higher etching, 00:32:06.650 --> 00:32:10.070 align:middle line:84% but you also get a higher voltage 00:32:10.070 --> 00:32:13.560 align:middle line:84% drop between the plasma and the wafer, so you get more damage. 00:32:13.560 --> 00:32:15.450 align:middle line:90% So there's sort of-- 00:32:15.450 --> 00:32:17.692 align:middle line:84% in this type of reactor, you don't have as many knobs 00:32:17.692 --> 00:32:19.650 align:middle line:84% as you might like, so there's a trade off here. 00:32:19.650 --> 00:32:24.570 align:middle line:84% You can etch faster by tweaking up the RF power, 00:32:24.570 --> 00:32:29.270 align:middle line:84% but you have to be careful because you're also increasing-- 00:32:29.270 --> 00:32:31.070 align:middle line:84% tendency to increase the ion energy. 00:32:31.070 --> 00:32:33.650 align:middle line:84% And increasing ion energy may or may not 00:32:33.650 --> 00:32:35.510 align:middle line:84% be desirable from a damage point of view. 00:32:35.510 --> 00:32:38.510 align:middle line:90% 00:32:38.510 --> 00:32:42.430 align:middle line:84% So I wanted to go on slide 13 and talk 00:32:42.430 --> 00:32:45.200 align:middle line:84% about the different plasma etching mechanisms, 00:32:45.200 --> 00:32:47.290 align:middle line:90% and there are basically three. 00:32:47.290 --> 00:32:51.820 align:middle line:84% There is chemical etching, which just like in wet etches, 00:32:51.820 --> 00:32:54.800 align:middle line:84% there are chemicals or species present, 00:32:54.800 --> 00:32:57.160 align:middle line:90% which do isotropic etching. 00:32:57.160 --> 00:32:59.470 align:middle line:84% They etch the same in both directions. 00:32:59.470 --> 00:33:00.730 align:middle line:90% These would tend to be-- 00:33:00.730 --> 00:33:04.660 align:middle line:90% typically not ionized, not ions. 00:33:04.660 --> 00:33:07.030 align:middle line:84% But they give very good selectivity because they 00:33:07.030 --> 00:33:09.010 align:middle line:90% have chemical specificity. 00:33:09.010 --> 00:33:12.610 align:middle line:84% There is physical etching, which is anisotropic. 00:33:12.610 --> 00:33:15.640 align:middle line:84% It tends to be ions that are accelerated down 00:33:15.640 --> 00:33:19.100 align:middle line:84% towards the wafer, and so they have a directionality associated 00:33:19.100 --> 00:33:20.350 align:middle line:90% with them, but less selective. 00:33:20.350 --> 00:33:22.000 align:middle line:84% If you're just banging down on the wafer, 00:33:22.000 --> 00:33:23.833 align:middle line:84% it doesn't matter what material you bang on. 00:33:23.833 --> 00:33:26.330 align:middle line:84% You're going to remove some of it. 00:33:26.330 --> 00:33:28.820 align:middle line:84% Ion enhanced etching is sort of a combination 00:33:28.820 --> 00:33:32.780 align:middle line:84% that we'll talk about with our ionic species that 00:33:32.780 --> 00:33:35.160 align:middle line:84% are present that also have chemical action. 00:33:35.160 --> 00:33:36.320 align:middle line:90% So these are the keys. 00:33:36.320 --> 00:33:38.390 align:middle line:84% This is where we get to reactive ion etching. 00:33:38.390 --> 00:33:41.090 align:middle line:84% You can get both anisotropy and selectivity 00:33:41.090 --> 00:33:43.640 align:middle line:84% to a certain extent at the same time. 00:33:43.640 --> 00:33:46.520 align:middle line:84% Most etches today use reactive ion etching or ion 00:33:46.520 --> 00:33:49.490 align:middle line:84% enhanced etching, but let's go through these three 00:33:49.490 --> 00:33:50.430 align:middle line:90% different mechanisms. 00:33:50.430 --> 00:33:52.550 align:middle line:90% The first is chemical. 00:33:52.550 --> 00:33:54.800 align:middle line:84% As you might imagine, it's typically 00:33:54.800 --> 00:33:57.990 align:middle line:84% etching by neutral species, free radicals, 00:33:57.990 --> 00:34:03.790 align:middle line:84% so for example, you're creating this extra CF4, which 00:34:03.790 --> 00:34:07.150 align:middle line:84% is very reactive, and CF3, which is very reactive, 00:34:07.150 --> 00:34:09.940 align:middle line:84% and fluorine, which can then-- this free fluorine 00:34:09.940 --> 00:34:13.630 align:middle line:84% can come down, hit the surf-- or come to the surface, 00:34:13.630 --> 00:34:15.460 align:middle line:84% react with the silicon at the surface, 00:34:15.460 --> 00:34:19.929 align:middle line:84% and create C4, which is volatile. 00:34:19.929 --> 00:34:24.159 align:middle line:84% You can add things to this to etch silicon a little bit more 00:34:24.159 --> 00:34:24.760 align:middle line:90% efficiently. 00:34:24.760 --> 00:34:29.320 align:middle line:84% You can add oxygen, O2, and that helps react with the CF3 00:34:29.320 --> 00:34:32.770 align:middle line:84% and reduces the recombination of CF3 plus fluorine 00:34:32.770 --> 00:34:34.300 align:middle line:90% to go back to CF4. 00:34:34.300 --> 00:34:35.889 align:middle line:84% So that helps-- adding oxygen tends 00:34:35.889 --> 00:34:38.500 align:middle line:84% to push this more to the right, so you get a higher etch rate. 00:34:38.500 --> 00:34:42.730 align:middle line:84% So a very common chemical etch that is fairly isotropic 00:34:42.730 --> 00:34:46.000 align:middle line:90% is to flow CF4 plus O2. 00:34:46.000 --> 00:34:48.730 align:middle line:84% That etches silicon in a reasonably isotropic manner 00:34:48.730 --> 00:34:50.739 align:middle line:84% because it doesn't depend much on ions. 00:34:50.739 --> 00:34:53.500 align:middle line:84% It depends just on this surface chemistry. 00:34:53.500 --> 00:34:56.330 align:middle line:90% 00:34:56.330 --> 00:35:02.010 align:middle line:84% So as I indicate on slide 14, these processes, 00:35:02.010 --> 00:35:04.340 align:middle line:84% these chemical equations, are really just chemical. 00:35:04.340 --> 00:35:07.290 align:middle line:84% They are isotropic and selective like wet etching. 00:35:07.290 --> 00:35:11.420 align:middle line:84% So here's an example of, you have some species 00:35:11.420 --> 00:35:13.580 align:middle line:90% being created in the plasma. 00:35:13.580 --> 00:35:16.190 align:middle line:84% They are transferred to the surface. 00:35:16.190 --> 00:35:17.720 align:middle line:90% They adsorb on the surface. 00:35:17.720 --> 00:35:19.100 align:middle line:90% We have this free radical. 00:35:19.100 --> 00:35:22.010 align:middle line:84% They react with the surface or the film, 00:35:22.010 --> 00:35:24.050 align:middle line:84% and then they come off as a byproduct. 00:35:24.050 --> 00:35:27.820 align:middle line:90% 00:35:27.820 --> 00:35:30.100 align:middle line:84% Just like we talked about in deposition, 00:35:30.100 --> 00:35:33.340 align:middle line:84% we usually characterize the arrival angle distribution 00:35:33.340 --> 00:35:36.520 align:middle line:84% of these species coming down the etchants. 00:35:36.520 --> 00:35:38.890 align:middle line:84% Usually, there's some kind of cosine theta to the end 00:35:38.890 --> 00:35:42.520 align:middle line:84% if it's isotropic, n equals 1, and 00:35:42.520 --> 00:35:45.670 align:middle line:84% a certain sticking coefficient usually very low. 00:35:45.670 --> 00:35:47.660 align:middle line:84% The surface reaction takes some time, 00:35:47.660 --> 00:35:49.090 align:middle line:90% and you can get some desorption. 00:35:49.090 --> 00:35:51.550 align:middle line:84% But you can use this type of picture 00:35:51.550 --> 00:35:54.640 align:middle line:84% to set up mathematical models of the process. 00:35:54.640 --> 00:35:58.650 align:middle line:90% 00:35:58.650 --> 00:36:01.053 align:middle line:84% So that's chemical etching, and again, you're 00:36:01.053 --> 00:36:02.220 align:middle line:90% just creating free radicals. 00:36:02.220 --> 00:36:04.170 align:middle line:84% It's not that much different, in some ways, 00:36:04.170 --> 00:36:09.950 align:middle line:84% from putting the etchant in a beaker of wet chemical. 00:36:09.950 --> 00:36:13.880 align:middle line:84% On slide 15, I talk about physical etching, so what 00:36:13.880 --> 00:36:14.750 align:middle line:90% is physical etching? 00:36:14.750 --> 00:36:19.280 align:middle line:84% Well, ion etching, for example, on the right, you could have, 00:36:19.280 --> 00:36:22.250 align:middle line:84% in the plasma, there are ionic species. 00:36:22.250 --> 00:36:24.170 align:middle line:84% The plasma-- there's a voltage drop 00:36:24.170 --> 00:36:25.680 align:middle line:84% between the plasma and the surface, 00:36:25.680 --> 00:36:29.280 align:middle line:84% so these ions can be accelerated towards the surface. 00:36:29.280 --> 00:36:31.880 align:middle line:84% So it tends to be more directional. 00:36:31.880 --> 00:36:34.010 align:middle line:84% There's an electric field or a voltage drop, 00:36:34.010 --> 00:36:35.780 align:middle line:84% as we mentioned, across the plasma sheath 00:36:35.780 --> 00:36:38.320 align:middle line:90% right near the surface. 00:36:38.320 --> 00:36:41.490 align:middle line:84% We often model this as a sticking coefficient close to 1. 00:36:41.490 --> 00:36:43.650 align:middle line:90% The ions don't go [INAUDIBLE]. 00:36:43.650 --> 00:36:44.400 align:middle line:90% They just come in. 00:36:44.400 --> 00:36:45.400 align:middle line:90% You assume they come in. 00:36:45.400 --> 00:36:47.700 align:middle line:84% They react, or they come in, and they knock off 00:36:47.700 --> 00:36:48.825 align:middle line:90% whatever's on the surface. 00:36:48.825 --> 00:36:51.330 align:middle line:90% 00:36:51.330 --> 00:36:53.850 align:middle line:84% Typical, in a physical etching mechanism, 00:36:53.850 --> 00:36:56.400 align:middle line:84% you might have species like CF4 plus. 00:36:56.400 --> 00:36:57.930 align:middle line:90% Again, it's not really chemical. 00:36:57.930 --> 00:37:00.528 align:middle line:84% It's just coming in as a large massive molecule 00:37:00.528 --> 00:37:02.070 align:middle line:84% comes in and can knock something off, 00:37:02.070 --> 00:37:05.130 align:middle line:84% or argon itself you can remove by sputtering. 00:37:05.130 --> 00:37:08.330 align:middle line:90% 00:37:08.330 --> 00:37:11.300 align:middle line:84% So it's not-- it tends to be not very selective. 00:37:11.300 --> 00:37:13.280 align:middle line:84% If you're doing purely physical mechanisms, 00:37:13.280 --> 00:37:14.780 align:middle line:84% it tends to remove the photoresist 00:37:14.780 --> 00:37:16.580 align:middle line:90% or erode the photoresist. 00:37:16.580 --> 00:37:20.720 align:middle line:84% It's hard to get selectivity because most sputter rates 00:37:20.720 --> 00:37:22.350 align:middle line:84% of all the elements are about the same. 00:37:22.350 --> 00:37:23.120 align:middle line:90% They're not that much different. 00:37:23.120 --> 00:37:24.920 align:middle line:84% They can vary by a factor of 2 to 4, 00:37:24.920 --> 00:37:26.990 align:middle line:84% but we just did an example where we 00:37:26.990 --> 00:37:29.180 align:middle line:84% needed a selectivity requirement of 20 to 1. 00:37:29.180 --> 00:37:32.090 align:middle line:84% And other etches, as I mentioned, need 50 or 100 00:37:32.090 --> 00:37:36.090 align:middle line:84% to 1, so purely physical etching is not going to give you that. 00:37:36.090 --> 00:37:38.390 align:middle line:90% But it is very directional. 00:37:38.390 --> 00:37:39.750 align:middle line:90% It is very anisotropic. 00:37:39.750 --> 00:37:42.267 align:middle line:90% 00:37:42.267 --> 00:37:44.100 align:middle line:84% The other problem with it is, of course, you 00:37:44.100 --> 00:37:45.642 align:middle line:84% can damage the surface because you're 00:37:45.642 --> 00:37:47.710 align:middle line:84% putting very energetic species in ion 00:37:47.710 --> 00:37:49.585 align:middle line:84% implanting them into the near surface region. 00:37:49.585 --> 00:37:52.330 align:middle line:90% 00:37:52.330 --> 00:37:54.580 align:middle line:84% So the third type-- so we did we did the chemical etch 00:37:54.580 --> 00:37:56.260 align:middle line:84% mechanisms, the physical mechanisms. 00:37:56.260 --> 00:38:00.370 align:middle line:84% The third type is this sort of ion enhanced etching, 00:38:00.370 --> 00:38:03.388 align:middle line:84% and it's the most confusing one, but it's also 00:38:03.388 --> 00:38:04.180 align:middle line:90% the most important. 00:38:04.180 --> 00:38:06.370 align:middle line:84% It's been observed that, a lot of times, 00:38:06.370 --> 00:38:09.370 align:middle line:84% that the chemical and the physical components of plasma 00:38:09.370 --> 00:38:12.340 align:middle line:84% etching are not independent, that they interact 00:38:12.340 --> 00:38:16.870 align:middle line:84% with each other in terms of the etch rate and the resulting etch 00:38:16.870 --> 00:38:18.460 align:middle line:90% profile. 00:38:18.460 --> 00:38:20.660 align:middle line:84% And there are a couple of-- many examples of this. 00:38:20.660 --> 00:38:22.150 align:middle line:90% I'm just showing one here. 00:38:22.150 --> 00:38:25.600 align:middle line:84% This is a particular example of etching silicon, 00:38:25.600 --> 00:38:28.540 align:middle line:84% and this is a plot of the etch rate in silicon 00:38:28.540 --> 00:38:30.820 align:middle line:84% on the vertical axis in nanometers per minute 00:38:30.820 --> 00:38:32.210 align:middle line:90% as a function of time. 00:38:32.210 --> 00:38:36.800 align:middle line:84% And there are different gases flowed into the plasma etcher. 00:38:36.800 --> 00:38:40.310 align:middle line:84% In the beginning, you're flowing xenon difluoride gas only. 00:38:40.310 --> 00:38:44.210 align:middle line:84% You do have a silicon etch rate, but it's slow. 00:38:44.210 --> 00:38:46.460 align:middle line:84% Now, you add an argon ion beam, so you're 00:38:46.460 --> 00:38:48.410 align:middle line:90% adding some physical component. 00:38:48.410 --> 00:38:49.933 align:middle line:84% You're adding some argon in there, 00:38:49.933 --> 00:38:51.350 align:middle line:84% and all of a sudden, the etch rate 00:38:51.350 --> 00:38:54.720 align:middle line:84% goes up by a factor of 5 or 10, something like that. 00:38:54.720 --> 00:38:57.050 align:middle line:84% So you get a large increase in etch rate 00:38:57.050 --> 00:39:00.800 align:middle line:84% when you have both an argon plasma ion beam plus the xenon 00:39:00.800 --> 00:39:01.670 align:middle line:90% difluoride. 00:39:01.670 --> 00:39:03.380 align:middle line:84% And then you turn the argon off-- 00:39:03.380 --> 00:39:05.747 align:middle line:90% 00:39:05.747 --> 00:39:08.330 align:middle line:84% or you turn the xenon difluoride off, and you have argon only. 00:39:08.330 --> 00:39:11.010 align:middle line:84% And you get a little bit sputtering, but not much. 00:39:11.010 --> 00:39:12.920 align:middle line:84% So really, you only get fast etching, 00:39:12.920 --> 00:39:15.470 align:middle line:84% in this particular case, when you have both the ion beam 00:39:15.470 --> 00:39:18.720 align:middle line:84% and the xenon difluoride present at the same time. 00:39:18.720 --> 00:39:22.110 align:middle line:84% So there's a synergistic effect going on, 00:39:22.110 --> 00:39:23.615 align:middle line:90% and this is not uncommon. 00:39:23.615 --> 00:39:24.365 align:middle line:90% It's very typical. 00:39:24.365 --> 00:39:27.460 align:middle line:90% 00:39:27.460 --> 00:39:28.900 align:middle line:84% We'll talk about some mechanisms. 00:39:28.900 --> 00:39:32.350 align:middle line:84% The nice thing about this reactive ion etching, 00:39:32.350 --> 00:39:34.900 align:middle line:84% or as people call it, is the etch 00:39:34.900 --> 00:39:37.360 align:middle line:84% profiles can be very anisotropic, 00:39:37.360 --> 00:39:39.010 align:middle line:84% and you can also get selectivity. 00:39:39.010 --> 00:39:42.880 align:middle line:84% So you can try-- you can get the best of both worlds. 00:39:42.880 --> 00:39:44.107 align:middle line:90% How does it work? 00:39:44.107 --> 00:39:45.940 align:middle line:84% Well, there's a lot of different mechanisms. 00:39:45.940 --> 00:39:49.910 align:middle line:84% I don't think people know a lot of the details, 00:39:49.910 --> 00:39:52.450 align:middle line:84% but there are-- people try to model it. 00:39:52.450 --> 00:39:56.170 align:middle line:84% But some examples of this synergistic mechanisms 00:39:56.170 --> 00:40:02.000 align:middle line:84% are shown below, for example, here on page 17 on the left. 00:40:02.000 --> 00:40:05.500 align:middle line:84% So you have some reactive neutral species that's in there. 00:40:05.500 --> 00:40:07.750 align:middle line:84% You have the ionic species at the same time. 00:40:07.750 --> 00:40:11.200 align:middle line:84% Well, what might be happening is there 00:40:11.200 --> 00:40:14.432 align:middle line:84% could be a chemical etch that's going on at the surface based 00:40:14.432 --> 00:40:16.640 align:middle line:84% on, say, free fluorine, or whatever it happens to be. 00:40:16.640 --> 00:40:18.140 align:middle line:84% So there's a chemical etch going on, 00:40:18.140 --> 00:40:21.700 align:middle line:84% but the etch rate that the chemical equation, 00:40:21.700 --> 00:40:23.590 align:middle line:90% the chemical-- 00:40:23.590 --> 00:40:29.460 align:middle line:84% the reaction may depend on the bombardment by ionic species. 00:40:29.460 --> 00:40:31.550 align:middle line:84% So there's sort of a synergy going here. 00:40:31.550 --> 00:40:33.080 align:middle line:84% In the presence of this bombardment, 00:40:33.080 --> 00:40:36.550 align:middle line:90% the etch rate goes up. 00:40:36.550 --> 00:40:39.490 align:middle line:84% On the right, this is a little bit easier to understand. 00:40:39.490 --> 00:40:44.260 align:middle line:84% You might imagine that, during this etching, 00:40:44.260 --> 00:40:48.140 align:middle line:84% very often what happens is inhibitor layers form. 00:40:48.140 --> 00:40:51.550 align:middle line:84% So you'll have reactions that take place on the surface, 00:40:51.550 --> 00:40:53.680 align:middle line:84% but you have a lot of carbon, the CF4. 00:40:53.680 --> 00:40:55.540 align:middle line:84% A lot of the etchings have carbon in them. 00:40:55.540 --> 00:40:58.300 align:middle line:84% They tend to form some kind of polymer 00:40:58.300 --> 00:41:00.880 align:middle line:84% that sits-- a monolayer or two of polymers 00:41:00.880 --> 00:41:02.440 align:middle line:90% that sit on the surface. 00:41:02.440 --> 00:41:06.370 align:middle line:84% These polymers are there, but they inhibit the next reaction 00:41:06.370 --> 00:41:07.610 align:middle line:90% from taking place. 00:41:07.610 --> 00:41:10.390 align:middle line:84% In some ways, they inhibit the etch rate locally 00:41:10.390 --> 00:41:11.540 align:middle line:90% wherever they exist. 00:41:11.540 --> 00:41:13.420 align:middle line:84% But now, if you have ion species around, 00:41:13.420 --> 00:41:17.230 align:middle line:84% they can come in and bash up, they can bombard the polymers, 00:41:17.230 --> 00:41:19.180 align:middle line:84% and remove them only where they hit right 00:41:19.180 --> 00:41:20.630 align:middle line:90% at the bottom of the trench. 00:41:20.630 --> 00:41:24.500 align:middle line:84% The inhibitor layer can still be left on the sidewall. 00:41:24.500 --> 00:41:25.970 align:middle line:84% So here's an example where you have 00:41:25.970 --> 00:41:27.830 align:middle line:90% both chemical etching going on. 00:41:27.830 --> 00:41:29.840 align:middle line:84% You have the formation of inhibitor layer, 00:41:29.840 --> 00:41:32.540 align:middle line:84% but only where the ions come down and hit 00:41:32.540 --> 00:41:34.250 align:middle line:84% do you remove the inhibitor, and then 00:41:34.250 --> 00:41:36.830 align:middle line:84% you get etching only at the bottom of the trench. 00:41:36.830 --> 00:41:40.880 align:middle line:84% So here's an example of reactive ion etching. 00:41:40.880 --> 00:41:43.250 align:middle line:84% The presence of these little byproducts, 00:41:43.250 --> 00:41:45.620 align:middle line:84% which you might think are, oh, these 00:41:45.620 --> 00:41:47.390 align:middle line:84% are just parasitics to the whole process. 00:41:47.390 --> 00:41:48.765 align:middle line:84% They're the name of the game when 00:41:48.765 --> 00:41:51.200 align:middle line:84% it comes to doing reactive ion etching and getting 00:41:51.200 --> 00:41:53.150 align:middle line:90% a very anisotropic profile. 00:41:53.150 --> 00:41:56.027 align:middle line:84% You want inhibitor layers to form on the sidewall 00:41:56.027 --> 00:41:57.860 align:middle line:84% because you don't want any sideways etching, 00:41:57.860 --> 00:41:59.840 align:middle line:84% but you don't want them to form on the bottom. 00:41:59.840 --> 00:42:01.260 align:middle line:84% Well, they'll still form on the bottom, 00:42:01.260 --> 00:42:03.140 align:middle line:84% but if you have ionic species coming down, 00:42:03.140 --> 00:42:05.770 align:middle line:90% you can remove them. 00:42:05.770 --> 00:42:08.250 align:middle line:84% So this is how people do reactive ion etching 00:42:08.250 --> 00:42:11.523 align:middle line:84% and get such both selectivity, and the reason you 00:42:11.523 --> 00:42:14.190 align:middle line:84% get selectivity, because there's still a chemical reaction going 00:42:14.190 --> 00:42:14.890 align:middle line:90% on here. 00:42:14.890 --> 00:42:17.800 align:middle line:84% So it's not just etching-- it's not just sputtering. 00:42:17.800 --> 00:42:20.160 align:middle line:84% It's a chemical reaction and a little bit 00:42:20.160 --> 00:42:23.840 align:middle line:84% of removal of the inhibitor together that count. 00:42:23.840 --> 00:42:29.440 align:middle line:84% So as you see on slide 18, this ion enhanced etching or reactive 00:42:29.440 --> 00:42:32.840 align:middle line:84% ion etching is hopefully the best of both worlds. 00:42:32.840 --> 00:42:37.170 align:middle line:84% You get good selectivity, and you can get good directionality. 00:42:37.170 --> 00:42:40.560 align:middle line:84% So although we may not know the exact mechanism, 00:42:40.560 --> 00:42:42.660 align:middle line:84% I kind of prefer, in my own mind, 00:42:42.660 --> 00:42:48.930 align:middle line:84% to have this inhibitor layer removal mechanism. 00:42:48.930 --> 00:42:51.930 align:middle line:84% The two components, that is the physical bombardment 00:42:51.930 --> 00:42:58.750 align:middle line:84% part and the chemical reaction etch part, they act in series. 00:42:58.750 --> 00:43:00.990 align:middle line:84% So you get this anisotropic etching 00:43:00.990 --> 00:43:03.840 align:middle line:84% with very little lateral undercutting because 00:43:03.840 --> 00:43:06.130 align:middle line:90% of the directed ion flux. 00:43:06.130 --> 00:43:09.630 align:middle line:84% So RIE is one of the most common forms 00:43:09.630 --> 00:43:11.760 align:middle line:84% of etching that you will see today 00:43:11.760 --> 00:43:13.860 align:middle line:84% because you can get, to a certain extent, the best 00:43:13.860 --> 00:43:14.520 align:middle line:90% of both worlds. 00:43:14.520 --> 00:43:17.880 align:middle line:90% 00:43:17.880 --> 00:43:21.540 align:middle line:84% Slide 19, talk a little bit about how to control the shape 00:43:21.540 --> 00:43:23.760 align:middle line:90% or what are some of the issues. 00:43:23.760 --> 00:43:27.720 align:middle line:84% Even with ion enhanced etching and this chemical etching, 00:43:27.720 --> 00:43:29.640 align:middle line:84% the slope of the resulting sidewall 00:43:29.640 --> 00:43:31.320 align:middle line:84% is not always perfectly vertical. 00:43:31.320 --> 00:43:33.810 align:middle line:90% There's some tricks. 00:43:33.810 --> 00:43:36.330 align:middle line:84% It needs to be adjusted empirically. 00:43:36.330 --> 00:43:39.990 align:middle line:84% The ion flux may not always be perfectly normal to the surface. 00:43:39.990 --> 00:43:42.900 align:middle line:84% The ion flux may be coming at a slight angle. 00:43:42.900 --> 00:43:44.610 align:middle line:84% You could have bowing of-- that can 00:43:44.610 --> 00:43:47.580 align:middle line:84% cause bowing of the side walls of whatever you're 00:43:47.580 --> 00:43:50.190 align:middle line:90% trying to etch. 00:43:50.190 --> 00:43:51.690 align:middle line:84% You can also get sloped sidewalls 00:43:51.690 --> 00:43:53.700 align:middle line:84% when the inhibitors, formed during the etch, 00:43:53.700 --> 00:43:56.382 align:middle line:84% have a high deposition rate relative to the etch rate 00:43:56.382 --> 00:43:57.840 align:middle line:84% of the inhibitor and the substrate. 00:43:57.840 --> 00:43:59.757 align:middle line:84% That's a little tricky, but on the next slide, 00:43:59.757 --> 00:44:03.120 align:middle line:84% I'll show you an example of how we get sloped sidewalls. 00:44:03.120 --> 00:44:07.470 align:middle line:84% So you need if you change the relative inhibitor deposition 00:44:07.470 --> 00:44:11.370 align:middle line:84% and etch rates and substrate etch rate 00:44:11.370 --> 00:44:13.260 align:middle line:84% by changing the chemistry or changing 00:44:13.260 --> 00:44:15.030 align:middle line:84% the specifics of the plasma conditions, 00:44:15.030 --> 00:44:17.160 align:middle line:84% you can control the slope of that sidewall. 00:44:17.160 --> 00:44:20.440 align:middle line:90% 00:44:20.440 --> 00:44:22.780 align:middle line:84% So I just want to show, on slide 20, 00:44:22.780 --> 00:44:27.820 align:middle line:84% an example of how that works, and we kind of 00:44:27.820 --> 00:44:29.060 align:middle line:90% walk through these. 00:44:29.060 --> 00:44:29.930 align:middle line:90% There are two cases. 00:44:29.930 --> 00:44:33.040 align:middle line:84% One case on the left, I'm assuming-- 00:44:33.040 --> 00:44:35.530 align:middle line:84% we're assuming that the inhibitor deposition 00:44:35.530 --> 00:44:39.250 align:middle line:84% rate is very fast compared to the etch rate. 00:44:39.250 --> 00:44:41.260 align:middle line:84% On the right, the inhibitor deposition rate 00:44:41.260 --> 00:44:45.590 align:middle line:84% is relatively slow compared to the etch rate of the film, 00:44:45.590 --> 00:44:47.470 align:middle line:84% so let's see what happens in both cases. 00:44:47.470 --> 00:44:49.700 align:middle line:84% If we start on the left, here's my mask. 00:44:49.700 --> 00:44:51.190 align:middle line:84% Here's the film I'm trying to etch. 00:44:51.190 --> 00:44:54.290 align:middle line:84% Now, just don't get confused here. 00:44:54.290 --> 00:44:57.340 align:middle line:84% The inhibitor deposition process and the etching process 00:44:57.340 --> 00:44:58.720 align:middle line:90% are happening simultaneously. 00:44:58.720 --> 00:45:01.750 align:middle line:84% I'm just showing them in separate views, separate steps, 00:45:01.750 --> 00:45:03.620 align:middle line:90% to help us think through it. 00:45:03.620 --> 00:45:05.500 align:middle line:84% So let's just say, in the first time step, 00:45:05.500 --> 00:45:10.240 align:middle line:84% the first couple of seconds, the inhibitor deposits everywhere. 00:45:10.240 --> 00:45:12.860 align:middle line:84% So you have this inhibitor formation. 00:45:12.860 --> 00:45:18.890 align:middle line:84% Now, in the next time step, the inhibitor is removed, 00:45:18.890 --> 00:45:20.870 align:middle line:90% and you etch. 00:45:20.870 --> 00:45:23.542 align:middle line:84% So here, the inhibitor DEP rate is fast compared to the etch, 00:45:23.542 --> 00:45:25.250 align:middle line:84% so I don't etch very much of the silicon. 00:45:25.250 --> 00:45:28.430 align:middle line:84% Just a little bit, whereas on the right, 00:45:28.430 --> 00:45:31.610 align:middle line:84% I etched quite a bit because this etch rate of the silicon 00:45:31.610 --> 00:45:35.990 align:middle line:84% is much faster than it is in this case. 00:45:35.990 --> 00:45:37.700 align:middle line:84% Now, I form, again, the next step, 00:45:37.700 --> 00:45:39.320 align:middle line:84% a little bit of inhibitor deposition 00:45:39.320 --> 00:45:42.620 align:middle line:84% here on both the left and the right. 00:45:42.620 --> 00:45:44.820 align:middle line:84% Here, the inhibitor DEP rate is pretty fast, 00:45:44.820 --> 00:45:46.970 align:middle line:90% so I have a thick inhibitor. 00:45:46.970 --> 00:45:50.060 align:middle line:84% Now, during the etch part, I etch just a little ways, 00:45:50.060 --> 00:45:53.415 align:middle line:84% whereas here, look, I etched a deep ways. 00:45:53.415 --> 00:45:55.040 align:middle line:84% And you continue on doing that, and you 00:45:55.040 --> 00:45:56.690 align:middle line:84% can see that, in the left case, you're 00:45:56.690 --> 00:45:59.330 align:middle line:84% going to end up with a more sloped sidewall 00:45:59.330 --> 00:46:03.260 align:middle line:84% because the inhibitor is going down much faster than you 00:46:03.260 --> 00:46:04.760 align:middle line:90% are etching into the silicon. 00:46:04.760 --> 00:46:06.440 align:middle line:90% So you end up with this-- 00:46:06.440 --> 00:46:09.500 align:middle line:84% there's a certain amount of lateral etching that takes 00:46:09.500 --> 00:46:13.220 align:middle line:84% place, and so you end up with this kind of sloped sidewall. , 00:46:13.220 --> 00:46:16.790 align:middle line:84% Whereas in this case, where you've adjusted the inhibitor 00:46:16.790 --> 00:46:21.017 align:middle line:84% DEP rate, it's relatively slow compared to the etch rate, 00:46:21.017 --> 00:46:23.600 align:middle line:84% or the etch rate is relatively fast if you want to think of it 00:46:23.600 --> 00:46:24.150 align:middle line:90% that way. 00:46:24.150 --> 00:46:26.630 align:middle line:84% You end up with a much more vertical profile. 00:46:26.630 --> 00:46:29.210 align:middle line:84% So the angle of the vertical profile 00:46:29.210 --> 00:46:32.840 align:middle line:84% depends on this ratio of how quickly the inhibitor forms 00:46:32.840 --> 00:46:38.030 align:middle line:84% the polymer, and how quickly it ends up being etched away, 00:46:38.030 --> 00:46:40.115 align:middle line:84% and the etch rate of the silicon, 00:46:40.115 --> 00:46:42.240 align:middle line:84% or whatever you're trying to etch at the same time. 00:46:42.240 --> 00:46:44.300 align:middle line:90% So very small changes. 00:46:44.300 --> 00:46:47.180 align:middle line:84% Usually, you do this by flowing multiple different types 00:46:47.180 --> 00:46:49.730 align:middle line:90% of gas and things. 00:46:49.730 --> 00:46:51.500 align:middle line:84% Very small changes in those gas flows 00:46:51.500 --> 00:46:54.325 align:middle line:84% can then change the angle which you achieve, 00:46:54.325 --> 00:46:55.700 align:middle line:84% and obviously, people try to tune 00:46:55.700 --> 00:46:58.760 align:middle line:84% this to get the right angle for their particular application. 00:46:58.760 --> 00:47:02.040 align:middle line:90% 00:47:02.040 --> 00:47:05.332 align:middle line:84% So that's just an example of how RIE works 00:47:05.332 --> 00:47:06.790 align:middle line:84% and how you can control that angle, 00:47:06.790 --> 00:47:13.390 align:middle line:84% and that's a pretty critical thing to control these days. 00:47:13.390 --> 00:47:15.850 align:middle line:84% So slide 21, now, I want to go through a couple 00:47:15.850 --> 00:47:18.100 align:middle line:84% of different examples of the types of etching systems, 00:47:18.100 --> 00:47:22.118 align:middle line:84% and then we'll get back more into how we can do the modeling. 00:47:22.118 --> 00:47:24.160 align:middle line:84% Over the years, a lot of different configurations 00:47:24.160 --> 00:47:25.960 align:middle line:84% have been developed, some of which 00:47:25.960 --> 00:47:27.970 align:middle line:84% make use primarily of chemical mechanisms, 00:47:27.970 --> 00:47:31.510 align:middle line:84% some physical, and then a lot of them, primarily, these days, 00:47:31.510 --> 00:47:34.360 align:middle line:84% are this ion enhanced or RIE mechanism. 00:47:34.360 --> 00:47:39.490 align:middle line:84% The most old fashioned type you will find in pretty much 00:47:39.490 --> 00:47:41.822 align:middle line:84% any thin film lab is an old fashioned 00:47:41.822 --> 00:47:43.030 align:middle line:90% what they call barrel etcher. 00:47:43.030 --> 00:47:44.988 align:middle line:84% It's called a barrel etcher because it's shaped 00:47:44.988 --> 00:47:47.560 align:middle line:90% like a barrel, a cylinder. 00:47:47.560 --> 00:47:54.130 align:middle line:84% And you have here one of the outer circumference 00:47:54.130 --> 00:47:58.360 align:middle line:84% of the etcher is one of the electrodes, 00:47:58.360 --> 00:48:03.640 align:middle line:84% and then you have some kind of shield on the center as well. 00:48:03.640 --> 00:48:07.390 align:middle line:84% And you have an RF bias then here between one side 00:48:07.390 --> 00:48:10.180 align:middle line:84% of the barrel and the other, and what you end up with, 00:48:10.180 --> 00:48:13.630 align:middle line:84% this is purely because of the geometry and the chemistry 00:48:13.630 --> 00:48:15.590 align:middle line:84% of what you're using, purely chemical etching. 00:48:15.590 --> 00:48:17.920 align:middle line:84% So you're primarily creating the free radicals 00:48:17.920 --> 00:48:20.500 align:middle line:84% that we talked about earlier that participate 00:48:20.500 --> 00:48:25.360 align:middle line:84% in these chemical reactions, and barrel electrodes are used today 00:48:25.360 --> 00:48:27.590 align:middle line:84% a lot in fabs, but not for anything critical. 00:48:27.590 --> 00:48:29.920 align:middle line:84% They're usually photoresist strippers, 00:48:29.920 --> 00:48:32.560 align:middle line:84% so a common way to strip photoresist off of wafers 00:48:32.560 --> 00:48:36.910 align:middle line:84% is put them in a barrel etcher and put an oxygen plasma. 00:48:36.910 --> 00:48:39.610 align:middle line:84% Photoresist reacts as an organic. 00:48:39.610 --> 00:48:43.000 align:middle line:84% It reacts very quickly with the oxygen, 00:48:43.000 --> 00:48:46.150 align:middle line:84% and this process strips the resist 00:48:46.150 --> 00:48:48.350 align:middle line:84% without attacking much else on the wafer. 00:48:48.350 --> 00:48:51.620 align:middle line:84% And this is called ashing, just commonplace. 00:48:51.620 --> 00:48:55.330 align:middle line:84% So nothing-- the shape and design are not real critical. 00:48:55.330 --> 00:48:59.360 align:middle line:84% It's just a way to strip off certain species 00:48:59.360 --> 00:49:03.200 align:middle line:90% such as a photoresist. 00:49:03.200 --> 00:49:07.460 align:middle line:84% The next level of complexity in a fab that you might find 00:49:07.460 --> 00:49:10.970 align:middle line:84% is shown on page 22, a system called the parallel plate 00:49:10.970 --> 00:49:14.610 align:middle line:84% system operated in the plasma mode as opposed to an RIE mode. 00:49:14.610 --> 00:49:17.990 align:middle line:84% So here's a plasma mode parallel plate system. 00:49:17.990 --> 00:49:21.830 align:middle line:84% We have two parallel plates that form the electrodes, one 00:49:21.830 --> 00:49:25.490 align:middle line:84% with an RF power input, and then there's 00:49:25.490 --> 00:49:28.760 align:middle line:84% another electrode on which the wafers sit. 00:49:28.760 --> 00:49:34.350 align:middle line:84% And you put in some gases like CF4 and O2. 00:49:34.350 --> 00:49:37.050 align:middle line:84% They have roughly equal areas, or maybe the wafer, 00:49:37.050 --> 00:49:38.340 align:middle line:90% is grounded in the chamber. 00:49:38.340 --> 00:49:41.250 align:middle line:90% It might be slightly larger. 00:49:41.250 --> 00:49:46.950 align:middle line:84% Here the sheath voltage may only be 10 to 100 volts, relatively 00:49:46.950 --> 00:49:51.090 align:middle line:84% low, so the energy in which these things come down 00:49:51.090 --> 00:49:54.600 align:middle line:84% is relatively moderate, say 10 to 100 electron volts. 00:49:54.600 --> 00:49:56.760 align:middle line:84% So there's not a whole lot of ionic component. 00:49:56.760 --> 00:50:00.840 align:middle line:84% You're not getting very high acceleration of the ions. 00:50:00.840 --> 00:50:03.870 align:middle line:84% It's primarily chemical, so in a parallel plate 00:50:03.870 --> 00:50:06.480 align:middle line:84% system in the plasma mode is pretty much still chemical. 00:50:06.480 --> 00:50:09.060 align:middle line:84% The etching tends to be fairly isotropic. 00:50:09.060 --> 00:50:11.670 align:middle line:84% It gets us down and sideways about the same, 00:50:11.670 --> 00:50:13.530 align:middle line:84% and it tends to be quite selective. 00:50:13.530 --> 00:50:16.770 align:middle line:84% Again, it's mostly chemical type of etching, 00:50:16.770 --> 00:50:19.550 align:middle line:84% so that's a parallel plate system in a plasma mode. 00:50:19.550 --> 00:50:22.630 align:middle line:90% 00:50:22.630 --> 00:50:25.630 align:middle line:84% Now, there's a second way to operate the parallel plate 00:50:25.630 --> 00:50:31.060 align:middle line:84% system shown here on slide 23, parallel plate in the RIE 00:50:31.060 --> 00:50:34.030 align:middle line:90% mode, Reactive Ion Etching. 00:50:34.030 --> 00:50:38.890 align:middle line:84% In order to do this, we need to get more directed etching. 00:50:38.890 --> 00:50:41.530 align:middle line:84% We need to have a stronger ion bombardment. 00:50:41.530 --> 00:50:44.680 align:middle line:84% We need to have a higher voltage drop between the plasma 00:50:44.680 --> 00:50:45.680 align:middle line:90% and the wafers. 00:50:45.680 --> 00:50:49.090 align:middle line:84% So here, you make the wafers electrode much smaller 00:50:49.090 --> 00:50:55.150 align:middle line:84% than the outer electrode, than the other electrode. 00:50:55.150 --> 00:50:58.600 align:middle line:84% So the wafers sit on the smaller electrode. 00:50:58.600 --> 00:51:00.400 align:middle line:84% Here, the voltage drop across the sheet, 00:51:00.400 --> 00:51:04.310 align:middle line:84% instead of being 10 electron volts, is like 100 to 700 100 00:51:04.310 --> 00:51:08.020 align:middle line:84% to 500 electron volts, so 10 times higher at least, 00:51:08.020 --> 00:51:10.160 align:middle line:90% maybe more, 20 times higher. 00:51:10.160 --> 00:51:14.710 align:middle line:84% So you have much greater energy of the incoming ions. 00:51:14.710 --> 00:51:19.210 align:middle line:84% You also have lower pressures to get more directional etching, 00:51:19.210 --> 00:51:20.290 align:middle line:90% maybe 10 millitorr. 00:51:20.290 --> 00:51:23.380 align:middle line:84% Lower pressures means, what, a longer mean free path, less 00:51:23.380 --> 00:51:25.210 align:middle line:84% randomization, so the ions again get 00:51:25.210 --> 00:51:28.090 align:middle line:84% accelerated towards the wafer, very directional. 00:51:28.090 --> 00:51:31.360 align:middle line:84% So this tends to be a more physical component 00:51:31.360 --> 00:51:33.310 align:middle line:90% than for the plasma mode. 00:51:33.310 --> 00:51:36.340 align:middle line:90% There's better directionality. 00:51:36.340 --> 00:51:39.310 align:middle line:84% A little bit less selectivity, however, for the RIE mode, 00:51:39.310 --> 00:51:41.050 align:middle line:84% but you can-- this is what you trade off. 00:51:41.050 --> 00:51:43.450 align:middle line:84% You trade off directionality for selectivity 00:51:43.450 --> 00:51:46.990 align:middle line:84% to a certain extent in these systems. 00:51:46.990 --> 00:51:48.318 align:middle line:90% Damage is going to be worse. 00:51:48.318 --> 00:51:50.860 align:middle line:84% A little bit more damage because you have higher energy ions. 00:51:50.860 --> 00:51:53.700 align:middle line:90% 00:51:53.700 --> 00:51:56.400 align:middle line:84% Oh, on slide 24 is just an example of a photograph. 00:51:56.400 --> 00:52:00.210 align:middle line:84% It's a very old fashioned, way back 20 years ago, 00:52:00.210 --> 00:52:05.730 align:middle line:84% but still found in some labs is the Applied Materials 8100 Dry 00:52:05.730 --> 00:52:07.350 align:middle line:90% Etcher. 00:52:07.350 --> 00:52:10.620 align:middle line:84% This gives you anisotropic etching, this particular one, 00:52:10.620 --> 00:52:14.850 align:middle line:84% of films like silicon nitride, silicon dioxide, 00:52:14.850 --> 00:52:16.650 align:middle line:90% and polymer layers. 00:52:16.650 --> 00:52:19.720 align:middle line:84% You can relatively control the oxide slope, 00:52:19.720 --> 00:52:22.560 align:middle line:84% the slope of a contact hole, slope angle, 00:52:22.560 --> 00:52:25.230 align:middle line:84% by using sidewall polymer deposition, 00:52:25.230 --> 00:52:28.658 align:middle line:84% just we talked about in that example of the inhibitor layer. 00:52:28.658 --> 00:52:30.450 align:middle line:84% The inhibitor layers are typically polymer, 00:52:30.450 --> 00:52:33.000 align:middle line:84% so sometimes people will call them sidewall polymers. 00:52:33.000 --> 00:52:35.710 align:middle line:84% The chemistry is fluorine based in this particular unit, 00:52:35.710 --> 00:52:39.720 align:middle line:84% so you would use things like CF3, oxygen, SF6, 00:52:39.720 --> 00:52:44.070 align:middle line:84% anything with fluorine in it, creating very reactive fluorine 00:52:44.070 --> 00:52:46.020 align:middle line:90% species. 00:52:46.020 --> 00:52:50.280 align:middle line:84% The inside electrode where the wafers sit is a hexag-- 00:52:50.280 --> 00:52:51.540 align:middle line:90% hexagonally shaped. 00:52:51.540 --> 00:52:53.760 align:middle line:84% The outside barrel, which is shown here, 00:52:53.760 --> 00:52:55.780 align:middle line:90% is the outer electrode. 00:52:55.780 --> 00:52:58.870 align:middle line:84% So the wafers sit on the inside electrode, which is smaller, 00:52:58.870 --> 00:53:00.640 align:middle line:84% and so you operate in the RIE mode. 00:53:00.640 --> 00:53:02.340 align:middle line:84% There's a fairly large voltage drop 00:53:02.340 --> 00:53:04.950 align:middle line:84% across the sheath near the wafer surface. 00:53:04.950 --> 00:53:08.340 align:middle line:90% 00:53:08.340 --> 00:53:10.460 align:middle line:84% So that's kind of old fashioned these days, 00:53:10.460 --> 00:53:13.580 align:middle line:84% although you'll see them in a lot of university labs, 00:53:13.580 --> 00:53:15.050 align:middle line:90% and even some fabs. 00:53:15.050 --> 00:53:20.720 align:middle line:84% There are RIE etchers here at MIT in the clean room here. 00:53:20.720 --> 00:53:23.240 align:middle line:84% The most sophisticated type of system 00:53:23.240 --> 00:53:24.740 align:middle line:84% today is a little different, though. 00:53:24.740 --> 00:53:27.410 align:middle line:90% It's shown here on page 25. 00:53:27.410 --> 00:53:33.050 align:middle line:84% It's called the High Density Plasma etch system, or an HDP, 00:53:33.050 --> 00:53:38.210 align:middle line:84% and what this does is it uses a remote non capacitively coupled 00:53:38.210 --> 00:53:42.170 align:middle line:84% plasma source, such as a microwave electron cyclotron 00:53:42.170 --> 00:53:45.590 align:middle line:84% resonance or inductively coupled plasma. 00:53:45.590 --> 00:53:52.190 align:middle line:84% So the mechanism by which you create the ions, that mechanism 00:53:52.190 --> 00:53:54.770 align:middle line:84% is separate from the mechanism by which you bring them 00:53:54.770 --> 00:53:55.800 align:middle line:90% to the wafer. 00:53:55.800 --> 00:53:57.380 align:middle line:90% So you have different voltages. 00:53:57.380 --> 00:54:02.045 align:middle line:84% Before, if you need-- in the old fashioned here RIE type, 00:54:02.045 --> 00:54:03.920 align:middle line:84% basically, if you want to get a higher plasma 00:54:03.920 --> 00:54:07.340 align:middle line:84% density and a higher etch rate, you drive up the voltage here, 00:54:07.340 --> 00:54:09.740 align:middle line:84% but that also means you have a bigger voltage 00:54:09.740 --> 00:54:10.760 align:middle line:90% drop to the wafer. 00:54:10.760 --> 00:54:13.550 align:middle line:84% So the plasma density and the damage 00:54:13.550 --> 00:54:18.620 align:middle line:84% to the wafer-- or the voltage drop here are directly related. 00:54:18.620 --> 00:54:21.590 align:middle line:84% You cannot decouple them very effectively. 00:54:21.590 --> 00:54:25.020 align:middle line:84% In HDP, High Density Plasma, these things 00:54:25.020 --> 00:54:29.970 align:middle line:84% are somewhat decoupled, so you have a separate RF source here 00:54:29.970 --> 00:54:31.750 align:middle line:90% that biases the wafer. 00:54:31.750 --> 00:54:33.930 align:middle line:84% So this separates the plasma power 00:54:33.930 --> 00:54:36.960 align:middle line:84% or the density of the plasma from the wafer bias, 00:54:36.960 --> 00:54:38.195 align:middle line:90% the accelerating field. 00:54:38.195 --> 00:54:40.320 align:middle line:84% And so now, you can get reasonably high etch rates, 00:54:40.320 --> 00:54:43.425 align:middle line:84% but not to completely damage the wafer surface. 00:54:43.425 --> 00:54:45.940 align:middle line:90% 00:54:45.940 --> 00:54:48.090 align:middle line:84% So on some of the characteristics 00:54:48.090 --> 00:54:50.940 align:middle line:90% are listed here on slide 26. 00:54:50.940 --> 00:54:52.950 align:middle line:84% You can get very high density plasmas, 00:54:52.950 --> 00:54:56.370 align:middle line:84% maybe 100 times higher density than you can achieve. 00:54:56.370 --> 00:54:58.470 align:middle line:84% Otherwise, you get faster etching. 00:54:58.470 --> 00:55:01.020 align:middle line:84% The pressures tend to be lower even in the 1 millitorr 00:55:01.020 --> 00:55:03.780 align:middle line:90% range, 1 to 10. 00:55:03.780 --> 00:55:06.150 align:middle line:84% To get higher ionization efficiency, 00:55:06.150 --> 00:55:09.270 align:middle line:84% longer mean free paths, more anisotropic, 00:55:09.270 --> 00:55:11.910 align:middle line:84% more directional etching, so things do come down 00:55:11.910 --> 00:55:15.488 align:middle line:90% to the surface perpendicular. 00:55:15.488 --> 00:55:17.280 align:middle line:84% The nice thing is, about these systems, you 00:55:17.280 --> 00:55:21.660 align:middle line:84% get a high etch rate, pretty good selectivity, good 00:55:21.660 --> 00:55:25.510 align:middle line:84% directionality, so very vertical side walls. 00:55:25.510 --> 00:55:28.360 align:middle line:84% At the same time, trying to keep the iron energy and the damage 00:55:28.360 --> 00:55:30.280 align:middle line:84% low because you have a separate power 00:55:30.280 --> 00:55:32.620 align:middle line:84% supply to create the plasma from the power 00:55:32.620 --> 00:55:34.970 align:middle line:84% supply that accelerates everything towards the wafer. 00:55:34.970 --> 00:55:38.050 align:middle line:84% So if you need to etch deep trenches, this type of thing, 00:55:38.050 --> 00:55:39.670 align:middle line:90% high density plasma is perfect. 00:55:39.670 --> 00:55:41.710 align:middle line:84% If you are doing optoelectronics, 00:55:41.710 --> 00:55:44.530 align:middle line:84% and you want to make waveguides, high density plasma where 00:55:44.530 --> 00:55:48.070 align:middle line:84% you need to etch through microns of material extremely 00:55:48.070 --> 00:55:50.430 align:middle line:84% vertical, perfectly vertical sidewalls, 00:55:50.430 --> 00:55:52.180 align:middle line:84% high density plasmas are perfect for that. 00:55:52.180 --> 00:55:53.230 align:middle line:90% Very high etch rate. 00:55:53.230 --> 00:55:55.060 align:middle line:84% Good control of the directionality, 00:55:55.060 --> 00:55:57.115 align:middle line:84% and you don't damage the surface too much. 00:55:57.115 --> 00:56:00.100 align:middle line:90% 00:56:00.100 --> 00:56:04.450 align:middle line:84% Here's an example, actually, slide 10-7, of an HDP system. 00:56:04.450 --> 00:56:05.560 align:middle line:90% It's fairly old one now. 00:56:05.560 --> 00:56:07.210 align:middle line:84% This was back in the mid 90s, this 00:56:07.210 --> 00:56:12.850 align:middle line:84% was popular, made by a company called Lam in California 00:56:12.850 --> 00:56:13.600 align:middle line:90% on the West Coast. 00:56:13.600 --> 00:56:16.570 align:middle line:90% It was called the TCP 9400. 00:56:16.570 --> 00:56:18.340 align:middle line:84% It's, again, a little bit obsolete, 00:56:18.340 --> 00:56:20.860 align:middle line:84% but it was developed for high density plasma 00:56:20.860 --> 00:56:23.680 align:middle line:84% etching of polysilicon gates, specifically 00:56:23.680 --> 00:56:26.960 align:middle line:84% to etch the gate electrode and to stop on silicon. 00:56:26.960 --> 00:56:29.020 align:middle line:84% So it was designed to have very, very-- 00:56:29.020 --> 00:56:31.930 align:middle line:84% give you very vertical side walls for gate etching, 00:56:31.930 --> 00:56:34.240 align:middle line:84% but not etch too much of the gate oxide underneath, 00:56:34.240 --> 00:56:36.790 align:middle line:84% so you get pretty good selectivity. 00:56:36.790 --> 00:56:39.130 align:middle line:84% So this etcher tends to be dedicated. 00:56:39.130 --> 00:56:42.040 align:middle line:84% In many fabs, they have a dedicated etcher 00:56:42.040 --> 00:56:44.350 align:middle line:84% just for gate etching and not for anything else. 00:56:44.350 --> 00:56:48.060 align:middle line:90% 00:56:48.060 --> 00:56:49.230 align:middle line:90% So that's the RIE-- 00:56:49.230 --> 00:56:52.180 align:middle line:84% example of RIE etchers and high density plasma. 00:56:52.180 --> 00:56:54.390 align:middle line:84% They're all over the place in fabs. 00:56:54.390 --> 00:56:57.990 align:middle line:84% There is one other type of etching, not that common, 00:56:57.990 --> 00:57:01.050 align:middle line:84% but people do it, shown here on slide 28, 00:57:01.050 --> 00:57:04.080 align:middle line:84% and that's sputter etching or even ion milling. 00:57:04.080 --> 00:57:06.600 align:middle line:90% And this is purely physical. 00:57:06.600 --> 00:57:07.770 align:middle line:90% It's highly directional. 00:57:07.770 --> 00:57:09.178 align:middle line:84% It doesn't have much selectivity. 00:57:09.178 --> 00:57:11.220 align:middle line:84% As I mentioned, the sputter rates of the elements 00:57:11.220 --> 00:57:13.930 align:middle line:90% don't vary very much. 00:57:13.930 --> 00:57:15.730 align:middle line:84% There's not a lot of chemistry going on. 00:57:15.730 --> 00:57:19.020 align:middle line:84% The nice thing about it, though, is you can etch almost anything. 00:57:19.020 --> 00:57:20.700 align:middle line:84% For certain materials, it's actually 00:57:20.700 --> 00:57:25.360 align:middle line:84% hard to remove them because they don't always have the right gas. 00:57:25.360 --> 00:57:27.900 align:middle line:84% But if you need to get the film off somehow, 00:57:27.900 --> 00:57:31.960 align:middle line:84% if you use a physical mechanism, you can sputter it off. 00:57:31.960 --> 00:57:36.100 align:middle line:84% Sputter etching almost always uses a heavy molecule 00:57:36.100 --> 00:57:40.510 align:middle line:84% or a heavy species like argon, an argon ion. 00:57:40.510 --> 00:57:42.190 align:middle line:84% The bad thing about it, of course, 00:57:42.190 --> 00:57:43.870 align:middle line:90% it damages the wafer surface. 00:57:43.870 --> 00:57:46.390 align:middle line:90% It can damage devices. 00:57:46.390 --> 00:57:49.750 align:middle line:84% It can also do things like produce trenching, which is 00:57:49.750 --> 00:57:53.230 align:middle line:90% a non-desirable sort of effect. 00:57:53.230 --> 00:57:56.690 align:middle line:84% Here in panel a, I show what trenching is. 00:57:56.690 --> 00:58:02.900 align:middle line:84% So this could be, here, let's say you're masking this, 00:58:02.900 --> 00:58:05.900 align:middle line:84% and you're trying to etch down here through this layer. 00:58:05.900 --> 00:58:09.470 align:middle line:84% Well, because, if you're using sputtering right near an edge, 00:58:09.470 --> 00:58:12.350 align:middle line:84% you can get the ions kind of bouncing off near this edge, 00:58:12.350 --> 00:58:15.468 align:middle line:84% and they can then come at a certain angle 00:58:15.468 --> 00:58:17.510 align:middle line:84% where they actually produce a little trench right 00:58:17.510 --> 00:58:18.710 align:middle line:90% near the edge of a feature. 00:58:18.710 --> 00:58:20.315 align:middle line:84% So you get enhanced etching right 00:58:20.315 --> 00:58:21.440 align:middle line:90% near the edge of a feature. 00:58:21.440 --> 00:58:22.940 align:middle line:90% You don't necessarily want that. 00:58:22.940 --> 00:58:25.650 align:middle line:84% You'd like to have the etch be straight across, 00:58:25.650 --> 00:58:28.430 align:middle line:90% so trenching is common. 00:58:28.430 --> 00:58:32.780 align:middle line:84% Ion bombardment damage, you can get redeposition, 00:58:32.780 --> 00:58:35.030 align:middle line:84% so this can come in here, sputter off this film, 00:58:35.030 --> 00:58:37.310 align:middle line:84% and redeposit those atoms down here, 00:58:37.310 --> 00:58:41.510 align:middle line:84% where you don't necessarily-- you may not want them. 00:58:41.510 --> 00:58:43.550 align:middle line:84% There can be charging that happens if you 00:58:43.550 --> 00:58:44.780 align:middle line:90% have an insulator up here. 00:58:44.780 --> 00:58:48.330 align:middle line:84% And then as these ions come in, that can distort the ion path, 00:58:48.330 --> 00:58:53.000 align:middle line:84% and so your trenches can bow out, for example. 00:58:53.000 --> 00:58:56.570 align:middle line:84% So the reason we bring up these sputter etching 00:58:56.570 --> 00:58:59.060 align:middle line:84% is not so-- and ion milling, not because people use them 00:58:59.060 --> 00:59:00.710 align:middle line:84% that much, the pure sputter etcher, 00:59:00.710 --> 00:59:03.350 align:middle line:84% but in any etcher system where they have 00:59:03.350 --> 00:59:04.733 align:middle line:84% a very strong physical component, 00:59:04.733 --> 00:59:06.900 align:middle line:84% a certain amount of this sputtering will take place. 00:59:06.900 --> 00:59:08.240 align:middle line:84% And if you're not careful, you can 00:59:08.240 --> 00:59:10.032 align:middle line:84% end up with trenching and things like that. 00:59:10.032 --> 00:59:13.220 align:middle line:84% So this, if you turn up the physical knob too much 00:59:13.220 --> 00:59:15.230 align:middle line:84% on your etch, you can start having things 00:59:15.230 --> 00:59:17.150 align:middle line:84% like this, which are indicative of that you're 00:59:17.150 --> 00:59:19.550 align:middle line:84% having a little bit too much sputter etching going on. 00:59:19.550 --> 00:59:22.570 align:middle line:90% 00:59:22.570 --> 00:59:25.810 align:middle line:84% So slide 29 is kind of a bit of a summary 00:59:25.810 --> 00:59:28.930 align:middle line:84% of the different plasma etching types of systems. 00:59:28.930 --> 00:59:33.970 align:middle line:84% Here at the top are the most physical processes listed. 00:59:33.970 --> 00:59:36.080 align:middle line:84% At the bottom are purely chemical processes, 00:59:36.080 --> 00:59:39.610 align:middle line:84% so we go from sputter etching and ion milling to-- 00:59:39.610 --> 00:59:43.060 align:middle line:84% which is completely physical, to high density plasma, which 00:59:43.060 --> 00:59:44.290 align:middle line:90% is physical plus chemical. 00:59:44.290 --> 00:59:46.210 align:middle line:90% RIE, a little more chemical. 00:59:46.210 --> 00:59:48.310 align:middle line:84% Plasma etching, totally chemical. 00:59:48.310 --> 00:59:50.200 align:middle line:90% Wet etching is totally chemical. 00:59:50.200 --> 00:59:54.020 align:middle line:84% There's no physical process-- no bombardment inside the beaker. 00:59:54.020 --> 00:59:58.420 align:middle line:84% So if we go from here to here, we think about it, 00:59:58.420 --> 01:00:03.310 align:middle line:84% the pressure is going down if I'm going in this direction. 01:00:03.310 --> 01:00:07.640 align:middle line:84% Anisotropy, so more anisotropic, it goes in this way. 01:00:07.640 --> 01:00:09.820 align:middle line:84% So wet etching is total isotropic. 01:00:09.820 --> 01:00:13.740 align:middle line:84% This sputter etching tends to be very anisotropic. 01:00:13.740 --> 01:00:17.620 align:middle line:84% Selectivity actually increases down this way, 01:00:17.620 --> 01:00:19.800 align:middle line:84% so wet etching is much more selective. 01:00:19.800 --> 01:00:22.020 align:middle line:84% Sputter etching is totally non selective, 01:00:22.020 --> 01:00:23.940 align:middle line:84% and the energy of the process tends 01:00:23.940 --> 01:00:25.230 align:middle line:90% to be higher here for these. 01:00:25.230 --> 01:00:28.470 align:middle line:84% A lot more energy and damage than in wet chemical etching. 01:00:28.470 --> 01:00:31.320 align:middle line:84% The energies are just a few electron, or half electron 01:00:31.320 --> 01:00:33.502 align:middle line:90% volts, just chemical reactions. 01:00:33.502 --> 01:00:34.960 align:middle line:84% Just keep that picture in mind when 01:00:34.960 --> 01:00:36.918 align:middle line:84% you're looking at the different types of plasma 01:00:36.918 --> 01:00:39.220 align:middle line:84% etchers in the fab and when you choose which 01:00:39.220 --> 01:00:40.510 align:middle line:90% plasma etcher you want to use. 01:00:40.510 --> 01:00:44.130 align:middle line:90% 01:00:44.130 --> 01:00:46.980 align:middle line:84% This slide just summarizes the different types 01:00:46.980 --> 01:00:50.040 align:middle line:84% of processes, all of which can be occurring during plasma 01:00:50.040 --> 01:00:52.140 align:middle line:84% etching, depending on the particular etcher 01:00:52.140 --> 01:00:54.430 align:middle line:84% that you're using, just as an example. 01:00:54.430 --> 01:00:57.300 align:middle line:84% So I'm assuming this top layer is my mask, 01:00:57.300 --> 01:01:00.870 align:middle line:84% and this is the film that we're trying to etch. 01:01:00.870 --> 01:01:03.060 align:middle line:84% There can be ionic species present, 01:01:03.060 --> 01:01:10.590 align:middle line:84% of course, that can cause some sputtering here and trenching. 01:01:10.590 --> 01:01:13.530 align:middle line:84% There are reactive neutral species, like free radicals. 01:01:13.530 --> 01:01:16.110 align:middle line:84% They cause the chemical part of the etching right 01:01:16.110 --> 01:01:19.500 align:middle line:84% here, which can be isotropic and lead to undercutting. 01:01:19.500 --> 01:01:22.770 align:middle line:84% There can be mask erosion because the mask is attacked 01:01:22.770 --> 01:01:26.040 align:middle line:84% by the etchant, and that can change the shape eventually 01:01:26.040 --> 01:01:28.050 align:middle line:90% of what you end up etching. 01:01:28.050 --> 01:01:29.490 align:middle line:84% Probably the most important-- one 01:01:29.490 --> 01:01:31.770 align:middle line:84% of the most important features on this whole slide, 01:01:31.770 --> 01:01:34.395 align:middle line:84% you can barely see, I guess we should have highlighted it here, 01:01:34.395 --> 01:01:37.680 align:middle line:84% is this little thin slim area called sidewall inhibitor 01:01:37.680 --> 01:01:38.430 align:middle line:90% deposition. 01:01:38.430 --> 01:01:40.650 align:middle line:84% That's probably the most mysterious and yet 01:01:40.650 --> 01:01:44.110 align:middle line:84% the most important part of reactive ion etching today. 01:01:44.110 --> 01:01:48.430 align:middle line:84% Without that, we could not do CMOS as we know it today. 01:01:48.430 --> 01:01:52.030 align:middle line:84% The sidewall inhibitor deposition, what is it made of? 01:01:52.030 --> 01:01:53.994 align:middle line:84% Well, people don't actually know. 01:01:53.994 --> 01:01:57.370 align:middle line:84% Maybe some of the etched byproducts. 01:01:57.370 --> 01:01:59.800 align:middle line:84% It actually also depends on mask erosion. 01:01:59.800 --> 01:02:01.990 align:middle line:84% It turns out that the rate at which you 01:02:01.990 --> 01:02:05.350 align:middle line:84% form this inhibitor depends on the total amount of photoresist 01:02:05.350 --> 01:02:07.510 align:middle line:90% exposed that is on the wafer. 01:02:07.510 --> 01:02:11.170 align:middle line:84% So people will develop a process using resist as the mask. 01:02:11.170 --> 01:02:12.010 align:middle line:90% It works perfectly. 01:02:12.010 --> 01:02:13.660 align:middle line:84% They get nice vertical sidewalls, 01:02:13.660 --> 01:02:15.693 align:middle line:84% good sidewall inhibitor formation. 01:02:15.693 --> 01:02:17.110 align:middle line:84% Now, they go and they say, I don't 01:02:17.110 --> 01:02:18.490 align:middle line:90% want to use resist as my mask. 01:02:18.490 --> 01:02:22.090 align:middle line:84% I want to transfer the mask pattern to an oxide, 01:02:22.090 --> 01:02:23.980 align:middle line:90% and use oxide as the mask. 01:02:23.980 --> 01:02:24.890 align:middle line:90% You might try that. 01:02:24.890 --> 01:02:27.190 align:middle line:84% So they no longer have resist in the etcher 01:02:27.190 --> 01:02:29.290 align:middle line:84% when they're etching the poly or whatever. 01:02:29.290 --> 01:02:30.940 align:middle line:84% All of a sudden, the sidewall inhibitor 01:02:30.940 --> 01:02:35.560 align:middle line:84% doesn't form because there's not enough carbon in the etcher. 01:02:35.560 --> 01:02:37.000 align:middle line:90% You don't have photoresist. 01:02:37.000 --> 01:02:39.430 align:middle line:84% So sometimes, the presence of photoresist, 01:02:39.430 --> 01:02:41.590 align:middle line:84% you want a little bit of mask erosion 01:02:41.590 --> 01:02:45.330 align:middle line:84% just to contribute to the side wall inhibitor. 01:02:45.330 --> 01:02:47.130 align:middle line:84% The key about the sidewall inhibitor 01:02:47.130 --> 01:02:49.110 align:middle line:84% is that it's removed on a flat surface 01:02:49.110 --> 01:02:51.870 align:middle line:84% because the ions come down, and they can knock it off. 01:02:51.870 --> 01:02:55.410 align:middle line:84% It tends to not be removed on a sidewall. 01:02:55.410 --> 01:02:59.400 align:middle line:84% That's why it's called a sidewall inhibitor. 01:02:59.400 --> 01:03:02.340 align:middle line:84% This is a possible mechanism why ion enhanced etching gives you 01:03:02.340 --> 01:03:04.950 align:middle line:84% such nice vertical profiles, because of the formation 01:03:04.950 --> 01:03:07.940 align:middle line:90% of that sidewall inhibitor. 01:03:07.940 --> 01:03:10.220 align:middle line:84% So those are some examples of all the different things 01:03:10.220 --> 01:03:12.965 align:middle line:84% that can happen during plasma etching. 01:03:12.965 --> 01:03:16.180 align:middle line:90% 01:03:16.180 --> 01:03:18.850 align:middle line:84% Slide 31, I'm listing some manufacturing issues. 01:03:18.850 --> 01:03:23.110 align:middle line:84% These are fairly practical things, some general plasma 01:03:23.110 --> 01:03:25.960 align:middle line:90% etching conditions. 01:03:25.960 --> 01:03:28.750 align:middle line:84% So parameters that you have-- if you're 01:03:28.750 --> 01:03:30.280 align:middle line:84% going into the lab, an MTL, and you 01:03:30.280 --> 01:03:32.080 align:middle line:84% need to take an etcher, what knobs 01:03:32.080 --> 01:03:35.440 align:middle line:84% can you turn that you can directly 01:03:35.440 --> 01:03:39.040 align:middle line:84% control in a standard parallel plate system? 01:03:39.040 --> 01:03:41.038 align:middle line:90% Well, you have the RF power. 01:03:41.038 --> 01:03:42.580 align:middle line:84% That's generally a knob you can turn. 01:03:42.580 --> 01:03:44.530 align:middle line:84% You have the pressure to a certain extent 01:03:44.530 --> 01:03:46.180 align:middle line:90% in the plasma etcher. 01:03:46.180 --> 01:03:48.550 align:middle line:84% You can control the gas composition, 01:03:48.550 --> 01:03:52.670 align:middle line:84% which species you decide to put in and their flow rates. 01:03:52.670 --> 01:03:54.520 align:middle line:84% So in a standard parallel plate system, 01:03:54.520 --> 01:03:57.825 align:middle line:84% you may have power densities like in this range, 0.1 to 5 01:03:57.825 --> 01:04:00.200 align:middle line:90% watts per square centimeter. 01:04:00.200 --> 01:04:01.970 align:middle line:84% If you want to increase the etch rate, 01:04:01.970 --> 01:04:06.410 align:middle line:84% you can turn up the RF power, so you get a higher density plasma. 01:04:06.410 --> 01:04:08.333 align:middle line:90% You'll get a higher self bias. 01:04:08.333 --> 01:04:10.250 align:middle line:84% That means the voltage drop between the plasma 01:04:10.250 --> 01:04:12.705 align:middle line:84% and the wafer or the electrodes will increase. 01:04:12.705 --> 01:04:15.080 align:middle line:84% That's going to increase the ion energy at the same time. 01:04:15.080 --> 01:04:17.770 align:middle line:90% 01:04:17.770 --> 01:04:21.430 align:middle line:84% An HDP system, you have an extra knob. 01:04:21.430 --> 01:04:24.190 align:middle line:84% You have a separate power supply for the plasma 01:04:24.190 --> 01:04:27.730 align:middle line:84% and for the wafer bias, so you can get a very high plasma 01:04:27.730 --> 01:04:30.010 align:middle line:90% density with a more gentle-- 01:04:30.010 --> 01:04:33.010 align:middle line:84% without necessarily getting such high ion energies. 01:04:33.010 --> 01:04:35.860 align:middle line:84% Plasma densities are much higher here, up to 10 watts 01:04:35.860 --> 01:04:37.870 align:middle line:90% per square centimeter. 01:04:37.870 --> 01:04:41.080 align:middle line:84% The power density at the wafer, there, though, and the ion 01:04:41.080 --> 01:04:42.430 align:middle line:90% energies, look at them. 01:04:42.430 --> 01:04:45.400 align:middle line:84% They can be much lower, maybe only 10 to 100 electron 01:04:45.400 --> 01:04:46.490 align:middle line:90% volts here. 01:04:46.490 --> 01:04:49.960 align:middle line:84% So faster etch rate with less damage 01:04:49.960 --> 01:04:52.000 align:middle line:90% to your wafer in an HDP system. 01:04:52.000 --> 01:04:55.500 align:middle line:90% 01:04:55.500 --> 01:04:57.300 align:middle line:90% How about the pressure ranges? 01:04:57.300 --> 01:05:01.100 align:middle line:84% These are very important here, shown on slide 32. 01:05:01.100 --> 01:05:04.910 align:middle line:84% This reactive ion etching, you typically 01:05:04.910 --> 01:05:06.530 align:middle line:90% would use something between-- 01:05:06.530 --> 01:05:09.980 align:middle line:84% the reactors are designed for 10 to 100 millitorr, 01:05:09.980 --> 01:05:13.520 align:middle line:84% and HDP system would drop that to the range 01:05:13.520 --> 01:05:17.450 align:middle line:84% of 1 to 10 millitorr, so maybe a factor of 10 lower. 01:05:17.450 --> 01:05:19.283 align:middle line:84% So if you have a higher pressure, 01:05:19.283 --> 01:05:21.200 align:middle line:84% you're going to get more gas phase collisions. 01:05:21.200 --> 01:05:24.080 align:middle line:84% This decreases the directionality, or the other way 01:05:24.080 --> 01:05:27.110 align:middle line:84% around, HDP has a lower pressure, 01:05:27.110 --> 01:05:29.840 align:middle line:84% longer mean free path, more directionality. 01:05:29.840 --> 01:05:32.490 align:middle line:90% 01:05:32.490 --> 01:05:36.295 align:middle line:84% But you also can pay a price in terms of the plasma density. 01:05:36.295 --> 01:05:37.920 align:middle line:84% Usually, if you have a higher pressure, 01:05:37.920 --> 01:05:41.575 align:middle line:84% you'll get a higher plasma density up to a certain point. 01:05:41.575 --> 01:05:42.950 align:middle line:84% Above a certain pressure, you get 01:05:42.950 --> 01:05:45.890 align:middle line:84% a lot of collisions between the gas molecules and the electrons, 01:05:45.890 --> 01:05:47.540 align:middle line:84% and that limits the energy electrons, 01:05:47.540 --> 01:05:51.140 align:middle line:84% and that limits the overall ionization rate. 01:05:51.140 --> 01:05:53.420 align:middle line:84% So pressure is an important knob that 01:05:53.420 --> 01:05:57.560 align:middle line:84% can be used to control not only etch rate in the plasma etcher, 01:05:57.560 --> 01:06:01.370 align:middle line:84% but also has effect on the sidewall profiles and things 01:06:01.370 --> 01:06:04.170 align:middle line:90% like that, and selectivity. 01:06:04.170 --> 01:06:07.800 align:middle line:90% 01:06:07.800 --> 01:06:09.630 align:middle line:84% One thing we haven't talked about much, 01:06:09.630 --> 01:06:12.180 align:middle line:84% I've sort of implied implicitly here 01:06:12.180 --> 01:06:15.030 align:middle line:84% that, in most people's minds, people 01:06:15.030 --> 01:06:18.960 align:middle line:84% think of plasma etching as being a room temperature process, 01:06:18.960 --> 01:06:22.170 align:middle line:90% and that is more or less true. 01:06:22.170 --> 01:06:25.470 align:middle line:84% But because we don't usually intentionally heat the wafer 01:06:25.470 --> 01:06:28.920 align:middle line:84% except certain exceptions, like aluminum etching, where 01:06:28.920 --> 01:06:31.140 align:middle line:84% the temperature of the etch system, most of the time, 01:06:31.140 --> 01:06:34.210 align:middle line:84% we don't intentionally raise the temperature of the wafer. 01:06:34.210 --> 01:06:39.300 align:middle line:84% However, the plasma puts a lot of energy into the process. 01:06:39.300 --> 01:06:40.950 align:middle line:84% You have a pretty high voltage drop. 01:06:40.950 --> 01:06:42.870 align:middle line:84% You have a large current flowing in there. 01:06:42.870 --> 01:06:45.810 align:middle line:84% V times I is power, so you're dumping power 01:06:45.810 --> 01:06:49.150 align:middle line:84% into this little space where your wafers are sitting. 01:06:49.150 --> 01:06:50.050 align:middle line:90% What does that mean? 01:06:50.050 --> 01:06:52.960 align:middle line:84% V times I power-- that power has to be dissipated somehow, 01:06:52.960 --> 01:06:56.180 align:middle line:84% and it comes out, a lot of the time, as heat. 01:06:56.180 --> 01:07:00.820 align:middle line:84% So the gas can get hot, and the wafer can get hot. 01:07:00.820 --> 01:07:04.460 align:middle line:90% 01:07:04.460 --> 01:07:07.070 align:middle line:84% But we don't usually need to heat it itself 01:07:07.070 --> 01:07:10.070 align:middle line:84% in order to increase the etch rate or improve the process. 01:07:10.070 --> 01:07:14.390 align:middle line:84% For aluminum etching, they sometimes do heat the system up 01:07:14.390 --> 01:07:16.818 align:middle line:84% to maybe 50 degrees C or so, and that's 01:07:16.818 --> 01:07:18.860 align:middle line:84% to keep the species volatile because it turns out 01:07:18.860 --> 01:07:22.250 align:middle line:84% the species that tend to be formed during the aluminum etch 01:07:22.250 --> 01:07:24.350 align:middle line:84% tend to not be very volatile at room temperature, 01:07:24.350 --> 01:07:25.520 align:middle line:90% and you'll never get them off. 01:07:25.520 --> 01:07:26.937 align:middle line:84% You'll get these inhibitor layers, 01:07:26.937 --> 01:07:28.890 align:middle line:90% and the etch will slow down. 01:07:28.890 --> 01:07:31.100 align:middle line:84% So to remove byproducts, sometimes, you 01:07:31.100 --> 01:07:32.750 align:middle line:84% will see people heating the wafer. 01:07:32.750 --> 01:07:34.170 align:middle line:84% This unintentional heating though, 01:07:34.170 --> 01:07:35.450 align:middle line:90% is much more of a problem. 01:07:35.450 --> 01:07:37.190 align:middle line:84% The wafers can get up to 100 degrees 01:07:37.190 --> 01:07:39.380 align:middle line:90% C without too much trouble. 01:07:39.380 --> 01:07:42.380 align:middle line:84% This needs to be controlled very carefully these days 01:07:42.380 --> 01:07:47.330 align:middle line:84% because that side wall inhibitor deposition tends to go down 01:07:47.330 --> 01:07:49.280 align:middle line:90% as the temperature goes up. 01:07:49.280 --> 01:07:51.410 align:middle line:84% So you never want to have the wafer getting 01:07:51.410 --> 01:07:53.810 align:middle line:84% so hot, self-heating, that the side wall inhibitor starts 01:07:53.810 --> 01:07:54.518 align:middle line:90% evaporating away. 01:07:54.518 --> 01:07:56.810 align:middle line:84% Next thing you know, you lose your side wall inhibitor, 01:07:56.810 --> 01:07:59.330 align:middle line:84% and you're-- all of a sudden you have a very poor looking 01:07:59.330 --> 01:08:01.620 align:middle line:90% profile, very isotropic. 01:08:01.620 --> 01:08:04.148 align:middle line:84% And that can happen unintentionally just because you 01:08:04.148 --> 01:08:06.690 align:middle line:84% turn up the power a little bit too much, and all of a sudden, 01:08:06.690 --> 01:08:07.982 align:middle line:90% you're putting enough power in. 01:08:07.982 --> 01:08:11.940 align:middle line:84% The wafers are heating, or you change wafer size or something 01:08:11.940 --> 01:08:14.790 align:middle line:84% in the plasma etcher, and you get too much heating, not enough 01:08:14.790 --> 01:08:16.920 align:middle line:84% sidewall inhibitor, and then your anisotropy 01:08:16.920 --> 01:08:18.180 align:middle line:90% goes through the roof-- 01:08:18.180 --> 01:08:20.200 align:middle line:84% goes to pot, so you're in trouble. 01:08:20.200 --> 01:08:24.029 align:middle line:84% So these days, the trends are to have better wafer temperature 01:08:24.029 --> 01:08:24.779 align:middle line:90% control. 01:08:24.779 --> 01:08:27.689 align:middle line:84% People have heat removal at the chuck. 01:08:27.689 --> 01:08:31.080 align:middle line:84% They have very careful clamping of the wafer, tightly controlled 01:08:31.080 --> 01:08:32.340 align:middle line:90% onto the chuck. 01:08:32.340 --> 01:08:34.080 align:middle line:84% Sometimes, they have helium gas flowing 01:08:34.080 --> 01:08:36.960 align:middle line:84% on the backside of the wafer to try to take heat away, 01:08:36.960 --> 01:08:39.000 align:middle line:84% just so you can control it because you 01:08:39.000 --> 01:08:40.111 align:middle line:90% don't want this etch-- 01:08:40.111 --> 01:08:42.569 align:middle line:84% especially if you're running the reactor through 25 wafers. 01:08:42.569 --> 01:08:44.580 align:middle line:84% The first wafer might be kind of cool, but by the time 01:08:44.580 --> 01:08:46.872 align:middle line:84% you get to the last one, the whole thing has heated up. 01:08:46.872 --> 01:08:49.830 align:middle line:84% And next thing you know, from wafer to wafer, 01:08:49.830 --> 01:08:52.830 align:middle line:84% you're getting a lot of variation in the etch profile. 01:08:52.830 --> 01:08:56.065 align:middle line:84% So controlling the temperature is also more and more important 01:08:56.065 --> 01:08:56.565 align:middle line:90% these days. 01:08:56.565 --> 01:08:59.200 align:middle line:90% 01:08:59.200 --> 01:09:01.420 align:middle line:84% Loading effects, there are things 01:09:01.420 --> 01:09:03.550 align:middle line:84% called macroscopic loading effects, 01:09:03.550 --> 01:09:06.670 align:middle line:84% and this is because you can deplete the etch and species 01:09:06.670 --> 01:09:08.920 align:middle line:84% across the wafer or across the whole etcher 01:09:08.920 --> 01:09:11.609 align:middle line:84% if you're doing more than one wafer at a time, 01:09:11.609 --> 01:09:13.069 align:middle line:90% so you'll see this. 01:09:13.069 --> 01:09:15.590 align:middle line:90% You put one wafer in. 01:09:15.590 --> 01:09:16.850 align:middle line:90% You measure the etch rate. 01:09:16.850 --> 01:09:17.569 align:middle line:90% Everything looks great. 01:09:17.569 --> 01:09:18.069 align:middle line:90% You have it. 01:09:18.069 --> 01:09:20.120 align:middle line:84% Now, you put all 10 of your wafers in. 01:09:20.120 --> 01:09:22.350 align:middle line:84% All of a sudden, the etch rate goes way down, 01:09:22.350 --> 01:09:24.058 align:middle line:84% so now, you screwed up your etch. 01:09:24.058 --> 01:09:25.850 align:middle line:84% So in fact, when you measure the etch rate, 01:09:25.850 --> 01:09:28.790 align:middle line:84% you should have put 10 in, 10 dummies, because otherwise, you 01:09:28.790 --> 01:09:31.550 align:middle line:84% can have this loading effect, and you don't really measure 01:09:31.550 --> 01:09:33.439 align:middle line:90% the etch rate appropriately. 01:09:33.439 --> 01:09:35.630 align:middle line:84% So it's kind of difficult to control. 01:09:35.630 --> 01:09:37.370 align:middle line:84% People get around it just by always 01:09:37.370 --> 01:09:40.189 align:middle line:84% using the same number of wafers with the same amount of material 01:09:40.189 --> 01:09:41.625 align:middle line:90% exposed every time. 01:09:41.625 --> 01:09:44.000 align:middle line:84% If you change that, you need to remeasure your etch rate. 01:09:44.000 --> 01:09:49.210 align:middle line:84% It's a pain, but it's a practical reality. 01:09:49.210 --> 01:09:50.770 align:middle line:90% That's not that hard to solve. 01:09:50.770 --> 01:09:55.180 align:middle line:84% This one, though, on page 35, this micro loading, 01:09:55.180 --> 01:09:57.320 align:middle line:90% it's a little tricky. 01:09:57.320 --> 01:09:59.890 align:middle line:84% And this refers to the fact that the etch rate 01:09:59.890 --> 01:10:02.110 align:middle line:84% can vary over very small distances 01:10:02.110 --> 01:10:05.380 align:middle line:90% on the surface of the wafer. 01:10:05.380 --> 01:10:06.530 align:middle line:90% And why is that? 01:10:06.530 --> 01:10:10.180 align:middle line:84% Well, it's because the density of the open area, the area 01:10:10.180 --> 01:10:11.050 align:middle line:90% that you're etching. 01:10:11.050 --> 01:10:13.360 align:middle line:84% The density of the area that's reacting 01:10:13.360 --> 01:10:16.002 align:middle line:84% can vary over small distances depending on the chip design. 01:10:16.002 --> 01:10:18.460 align:middle line:84% In one area, you may be doing a lot of etching in the chip. 01:10:18.460 --> 01:10:21.400 align:middle line:84% Another area, it might be mostly photoresist. 01:10:21.400 --> 01:10:24.590 align:middle line:84% So you have this sort of micro loading effect, 01:10:24.590 --> 01:10:27.550 align:middle line:84% so you get differences analogous to the macroscopic loading. 01:10:27.550 --> 01:10:30.005 align:middle line:84% Well, what people do is they put in dummy structures 01:10:30.005 --> 01:10:32.380 align:middle line:84% intentionally, where you don't really need to do an etch, 01:10:32.380 --> 01:10:34.172 align:middle line:84% but you're doing etching of the active area 01:10:34.172 --> 01:10:37.960 align:middle line:84% there anyway just to make a more uniform appearance 01:10:37.960 --> 01:10:40.600 align:middle line:90% to the plasma of the surface. 01:10:40.600 --> 01:10:42.820 align:middle line:84% This one is even trickier, number two. 01:10:42.820 --> 01:10:46.910 align:middle line:84% You may have differences in the aspect ratios. 01:10:46.910 --> 01:10:48.890 align:middle line:84% So you may have a portion of the wafer 01:10:48.890 --> 01:10:52.460 align:middle line:84% where you're etching a structure that's very closely spaced, 01:10:52.460 --> 01:10:54.470 align:middle line:84% and then you'll measure the etch rate here, 01:10:54.470 --> 01:10:57.380 align:middle line:90% shown in this SCM micrograph. 01:10:57.380 --> 01:11:01.370 align:middle line:84% The etch rate will be lower when you have a higher aspect ratio 01:11:01.370 --> 01:11:02.210 align:middle line:90% trench. 01:11:02.210 --> 01:11:04.550 align:middle line:84% On the same wafer right next door, 01:11:04.550 --> 01:11:07.220 align:middle line:84% a few microns away, the trench is 01:11:07.220 --> 01:11:10.200 align:middle line:84% etching faster where the aspect ratio is smaller, 01:11:10.200 --> 01:11:12.860 align:middle line:84% so there's a bigger opening between the trenches here. 01:11:12.860 --> 01:11:14.090 align:middle line:90% The ions can get down. 01:11:14.090 --> 01:11:15.570 align:middle line:84% The etch rate is a little faster. 01:11:15.570 --> 01:11:18.470 align:middle line:84% So right next door, depending on the feature, density, 01:11:18.470 --> 01:11:22.830 align:middle line:84% and dimension, you're getting a different etch rate. 01:11:22.830 --> 01:11:25.760 align:middle line:84% This is sometimes called aspect ratio dependent etching, 01:11:25.760 --> 01:11:27.980 align:middle line:90% or people call RIE lag. 01:11:27.980 --> 01:11:31.260 align:middle line:84% This guy is lagging behind this one in some ways, 01:11:31.260 --> 01:11:34.475 align:middle line:84% so something to be aware of when you're etching complex patterns. 01:11:34.475 --> 01:11:38.430 align:middle line:90% 01:11:38.430 --> 01:11:39.820 align:middle line:90% So why might this happen? 01:11:39.820 --> 01:11:43.240 align:middle line:84% Well, you may get depletion or trapping of the reactant species 01:11:43.240 --> 01:11:45.240 align:middle line:84% when you're traveling to a bottom of the trench. 01:11:45.240 --> 01:11:47.392 align:middle line:84% You can imagine these reactive species have to make 01:11:47.392 --> 01:11:48.600 align:middle line:90% their way to the very bottom. 01:11:48.600 --> 01:11:52.890 align:middle line:84% They may react the longer the distance they have to go, 01:11:52.890 --> 01:11:54.870 align:middle line:84% when it's a very tightly controlled tube, 01:11:54.870 --> 01:11:57.360 align:middle line:84% they may not have a chance to get to the bottom. 01:11:57.360 --> 01:12:00.600 align:middle line:84% You may have distortion of the ion paths due to charging. 01:12:00.600 --> 01:12:03.150 align:middle line:84% You may even have shadowing effects. 01:12:03.150 --> 01:12:05.865 align:middle line:84% The point is that the net result is the probability 01:12:05.865 --> 01:12:08.490 align:middle line:84% of a reactive species getting to the bottom of a trench is very 01:12:08.490 --> 01:12:10.980 align:middle line:84% narrow, goes down as you make the trench more and more 01:12:10.980 --> 01:12:13.140 align:middle line:90% deep and more and more narrow. 01:12:13.140 --> 01:12:17.130 align:middle line:84% So that's something we need to take into account when 01:12:17.130 --> 01:12:19.350 align:middle line:84% you're doing these kind of dense patterns. 01:12:19.350 --> 01:12:22.660 align:middle line:90% 01:12:22.660 --> 01:12:25.570 align:middle line:84% So slide 37 is kind of the summary of this introduction. 01:12:25.570 --> 01:12:27.960 align:middle line:84% We talked only a little bit about wet etching. 01:12:27.960 --> 01:12:31.440 align:middle line:84% It's primarily a chemistry problem, chemical reactions. 01:12:31.440 --> 01:12:33.390 align:middle line:84% Very good selectivity, but it etches 01:12:33.390 --> 01:12:35.580 align:middle line:84% the same amount in each direction, the same rate. 01:12:35.580 --> 01:12:37.650 align:middle line:90% It's isotropic. 01:12:37.650 --> 01:12:41.233 align:middle line:84% Dry etching, there are two kind of species 01:12:41.233 --> 01:12:42.900 align:middle line:84% in the plasma that tend to be important, 01:12:42.900 --> 01:12:45.960 align:middle line:84% the reactive neutral species or free radicals, just 01:12:45.960 --> 01:12:47.430 align:middle line:90% free fluorine. 01:12:47.430 --> 01:12:50.020 align:middle line:84% It's neutral, so it doesn't get accelerated, 01:12:50.020 --> 01:12:54.510 align:middle line:84% so it's fairly isotropic, but it's also quite selective. 01:12:54.510 --> 01:12:56.260 align:middle line:84% Ionic species, on the other hand, 01:12:56.260 --> 01:12:58.320 align:middle line:84% can be accelerated towards the surface. 01:12:58.320 --> 01:13:00.720 align:middle line:84% They tend to have more of a physical mechanism. 01:13:00.720 --> 01:13:03.330 align:middle line:84% It can be anisotropic, so they etch vertically 01:13:03.330 --> 01:13:05.700 align:middle line:84% much more than laterally, but therefore, 01:13:05.700 --> 01:13:08.640 align:middle line:90% sometimes not very selective. 01:13:08.640 --> 01:13:11.620 align:middle line:84% In plasma etchers today, there are different mechanisms. 01:13:11.620 --> 01:13:13.590 align:middle line:84% There's purely chemical mechanism, which 01:13:13.590 --> 01:13:15.750 align:middle line:90% is based on neutral species. 01:13:15.750 --> 01:13:19.680 align:middle line:84% There may be sputtering or physical sputtering 01:13:19.680 --> 01:13:23.130 align:middle line:84% by acceleration of the plasma, and the most popular 01:13:23.130 --> 01:13:26.010 align:middle line:84% is the sort of ion enhanced etching, where ions enhance 01:13:26.010 --> 01:13:29.340 align:middle line:84% the chemical processes, or ions remove these inhibitor 01:13:29.340 --> 01:13:34.330 align:middle line:84% layers, which is key to keeping the etch going. 01:13:34.330 --> 01:13:37.090 align:middle line:84% There's a lot of different types of plasma systems, 01:13:37.090 --> 01:13:38.620 align:middle line:90% just literally dozens of them. 01:13:38.620 --> 01:13:39.520 align:middle line:90% There's barrels. 01:13:39.520 --> 01:13:42.610 align:middle line:84% There's the plasma mode, RIE high density plasma, 01:13:42.610 --> 01:13:46.120 align:middle line:84% and there's even sputter etchers and lots of variations. 01:13:46.120 --> 01:13:50.200 align:middle line:84% So these tend to be purely chemical in the barrel range 01:13:50.200 --> 01:13:52.900 align:middle line:84% and purely physical if you're in the sputter etcher. 01:13:52.900 --> 01:13:55.938 align:middle line:84% Most of the systems these days operate in between those two 01:13:55.938 --> 01:13:57.730 align:middle line:84% limits, a little bit of chemistry going on, 01:13:57.730 --> 01:14:00.632 align:middle line:84% and a little bit of the physical etching. 01:14:00.632 --> 01:14:02.590 align:middle line:84% So that's about all I have to say introduction. 01:14:02.590 --> 01:14:04.930 align:middle line:84% Please go ahead, read chapter 10. 01:14:04.930 --> 01:14:06.700 align:middle line:84% Next time, we'll talk in detail about how 01:14:06.700 --> 01:14:09.160 align:middle line:84% you etch polysilicon gates with good profiles 01:14:09.160 --> 01:14:10.630 align:middle line:90% and good selectivity. 01:14:10.630 --> 01:14:13.570 align:middle line:84% Also homework 5, your last homework is due. 01:14:13.570 --> 01:14:15.810 align:middle line:90% You can put that in the folder. 01:14:15.810 --> 01:14:36.000 align:middle line:90%