WEBVTT Kind: captions Language: en 00:00:02.580 --> 00:00:34.860 Hello students, welcome to lecture 25 00:00:34.860 --> 00:00:40.020 of the online course on Nanophotronics, Plasmonics and Metamaterials. Today's lecture will be on 00:00:40.020 --> 00:00:46.620 metamaterial perfect absorbers. So here is the lecture outline, we will first look into what 00:00:46.620 --> 00:00:52.590 are these metamaterial perfect absorbers and then look into the classifications like narrow 00:00:52.590 --> 00:00:58.920 band perfect absorbers and broadband perfect absorbers. We will also take an example of 00:00:58.920 --> 00:01:06.720 ultra-broad band perfect absorber and then we will see the applications of these metamaterial perfect 00:01:06.720 --> 00:01:17.700 absorbers or MPA in short and their application in solar energy harvesting and as thermal emitters. 00:01:17.700 --> 00:01:23.760 So when we talk about metamaterials, the first thing that comes to our mind is that these are 00:01:23.760 --> 00:01:31.200 basically artificially engineered structures. So where the unit cell is designed in such a way that 00:01:31.200 --> 00:01:38.520 can give rise to some extraordinary properties which are not found in natural materials 00:01:39.300 --> 00:01:46.560 and the first thing that comes to our mind will be the negative index material, right? Negative 00:01:46.560 --> 00:01:52.380 refractive index but that is not all. When you think about other applications 00:01:52.380 --> 00:02:01.740 of metamaterials, one most important application is towards light absorbing. Now light absorbing 00:02:01.740 --> 00:02:09.720 by any artificial structure has always been a matter of research because you want to maximize 00:02:09.720 --> 00:02:17.760 the absorption. Like why you need that? Like if you think of a solar cell where you are harvesting 00:02:17.760 --> 00:02:25.740 solar energy, so you want to absorb the entire solar radiation that is falling on the solar 00:02:25.740 --> 00:02:32.460 cell. You do not want anything to be reflected. What goes back is actually a waste. So you are not 00:02:32.460 --> 00:02:36.960 efficient if you are not able to capture all of the solar radiation coming towards you. 00:02:37.620 --> 00:02:46.560 Similarly, a photodetector or any other such devices which are supposed to harvest light, 00:02:46.560 --> 00:02:54.120 you want them to first act as a perfect absorber so that you can absorb that, okay? And then you 00:02:54.120 --> 00:03:02.280 can convert that absorbed energy into some other form. So light absorption is another eye-catching 00:03:02.280 --> 00:03:09.780 characteristic of these artificial structures which are metamaterials and the metamaterials 00:03:09.780 --> 00:03:17.760 with near perfect light absorption are known as this metamaterial perfect absorbers MPA, okay? 00:03:19.320 --> 00:03:25.020 Now in order to realize perfect absorption, reflectance is suppressed. 00:03:26.400 --> 00:03:32.340 And how do you do that? You do that by matching the effective impedance of the material of the 00:03:32.340 --> 00:03:38.220 metamaterial to that of the incident medium. So whenever there is no impedance mismatch, 00:03:38.220 --> 00:03:46.740 there is no reflection. And when light falls on an interface, there are three phenomena 00:03:46.740 --> 00:03:51.000 that takes place. One is reflection, one is absorption, the other one is transmission, 00:03:51.780 --> 00:03:59.940 okay? So you have to somehow in absorbers, you want the entire light to be absorbed. So you 00:03:59.940 --> 00:04:08.100 are trying to cancel the reflection by using some metamaterial absorber and also you are trying to 00:04:08.100 --> 00:04:15.360 get rid of the transmittance. And that you can do by introducing another metallic plate which 00:04:15.360 --> 00:04:21.480 acts as a mirror or by using similar mechanism that we have seen in the multilayer system. 00:04:21.480 --> 00:04:28.920 So metamaterial perfect absorbers, they do not have any transmission, they also do not have any 00:04:28.920 --> 00:04:36.600 reflection. So whatever falls on them should get absorbed. Now what do you mean by perfect 00:04:36.600 --> 00:04:44.940 absorption? So we typically call a perfect absorber when it is close to 99% absorption, 00:04:44.940 --> 00:04:54.060 okay? And it remains highly absorptive over a wide range of incident angle, okay? For 00:04:54.060 --> 00:05:02.940 both TE and TM polarizations, okay? So if you recall our discussion, previous discussions, 00:05:03.720 --> 00:05:12.420 at larger angle, at gazing angle, you usually have larger reflection. But in that case, your 00:05:13.740 --> 00:05:19.380 light will not be that strongly absorbed. So at larger angle, there is a possibility 00:05:19.380 --> 00:05:28.860 of absorption to get reduced than 99%. But still it should be like at least more 00:05:28.860 --> 00:05:37.440 than 95% or so, okay? So what do you basically take to make this kind of absorbers? Now when I 00:05:37.440 --> 00:05:43.980 tell you that this material is lossy, the first thing that comes to your mind is that okay, 00:05:43.980 --> 00:05:48.840 this is a lossy material, so it is not very good for waveguiding because there is high 00:05:48.840 --> 00:05:56.580 loss. Also it is not a very good material for creating a resonator cavity because it will 00:05:56.580 --> 00:06:08.160 have low Q because it is lossy material, so the full width of maximum will be large. But then 00:06:09.120 --> 00:06:16.980 lossy materials are useful, very very useful in case of metamaterial absorbers because 00:06:16.980 --> 00:06:23.100 these materials can then significantly enhance the efficiency of absorption, 00:06:24.060 --> 00:06:29.400 okay? So recently metamaterial perfect absorbers have grabbed a lot of 00:06:30.480 --> 00:06:40.380 attention because of their near unity absorption capability over narrow band or broadband, okay? 00:06:41.040 --> 00:06:47.400 So that takes us to the classification of this metamaterial perfect absorbers. So as you can 00:06:47.400 --> 00:06:54.240 see the based on the wavelength range they cater to, you can categorize them into two buckets, 00:06:54.900 --> 00:06:59.580 narrow band metamaterial perfect absorbers and then you also have 00:06:59.580 --> 00:07:06.240 broadband metamaterial perfect absorbers. So if you carefully look at the narrow band structure, 00:07:06.840 --> 00:07:14.460 here you see figure A, you can see that these are basically overlapping rings and column 00:07:15.060 --> 00:07:24.240 and there is a gap between this column and the center of the ring that is called delta x. 00:07:24.900 --> 00:07:31.320 So when delta x equals 0 you get almost like 90 percent of absorption, right? In this particular 00:07:31.320 --> 00:07:40.680 case. When you change delta x to 200 nanometer or so you get almost 95-99 percent of reflection, 00:07:41.280 --> 00:07:50.880 okay? Sorry, absorption. When you further increase it to 270 or 350 what is happening? The resonance 00:07:50.880 --> 00:08:00.420 is even getting sharper, so you are getting a high Q absorption peak and you are closing towards that 00:08:00.420 --> 00:08:07.560 perfect absorption mark which is very very close to unity absorption, right? So this is typically 00:08:07.560 --> 00:08:13.260 what I am not going to describe immediately the physics behind this, okay? These are kind of 00:08:13.260 --> 00:08:21.180 symmetry breaking high Q modes, okay? And I will just look into the features here because this is 00:08:21.180 --> 00:08:27.600 the narrow band. We will see in this lecture only how to develop a structure that can give you a 00:08:27.600 --> 00:08:31.980 narrow band metamaterial perfect absorber and we will also look into this particular 00:08:31.980 --> 00:08:37.920 structures which are able to give you broadband metamaterial perfect absorber. So here you can 00:08:37.920 --> 00:08:47.700 see that it is basically a broadband absorber. So you can see it ranges from 3 to say 5.5 micron 00:08:47.700 --> 00:08:54.420 where the absorption is maintained over say 95 or 99 percent like that. 00:08:55.080 --> 00:08:59.640 Now what are these three arrows showing here is that a different different regime, 00:09:00.300 --> 00:09:08.040 different portion of the structure is playing the role of absorbing the light that is falling 00:09:08.040 --> 00:09:14.700 onto the structure. So here is the structure this is basically a this is the tapered shape, 00:09:14.700 --> 00:09:22.560 okay? You can also call it tapered tooth kind of a structure. So a different different wavelength, 00:09:23.580 --> 00:09:28.980 different different part of the structure is resonating and that is able to absorb 00:09:28.980 --> 00:09:32.460 the electromagnetic radiation. Here you see a short wavelength the 00:09:32.460 --> 00:09:38.280 top portion of the structure is absorbing most of the incident electric field, magnetic field, 00:09:38.280 --> 00:09:44.340 okay? And at longer wavelength the bottom most part of the structure which is the 00:09:44.340 --> 00:09:49.920 widest portion of the structure is responsible for the absorption. Now let us look into some of 00:09:49.920 --> 00:09:56.700 these techniques of how do you design narrow band metamaterial based perfect absorbers, 00:09:56.700 --> 00:10:06.480 okay? So here is the schematic. So you need metal, lossy metal to design absorbers. So 00:10:06.480 --> 00:10:16.020 here is a schematic of a 2D array of gold discs, okay? So the diameter is 352 nanometer and the 00:10:16.020 --> 00:10:23.940 thickness is 20 nanometer and the periodicity along both X and Y directions are 600 nanometer. 00:10:23.940 --> 00:10:34.320 Then there is a spacer layer below this gold discs that is MgF2 that is magnesium fluoride and it has 00:10:34.320 --> 00:10:40.080 got a thickness of 30 nanometer and then you have a gold mirror which is 200 nanometer thick. 00:10:40.080 --> 00:10:45.660 So this is basically that portion which is completely cancelling out any chance 00:10:45.660 --> 00:10:52.560 of transmission through this structure. So 200 nanometer gold film behaves like a bulk gold 00:10:52.560 --> 00:10:58.920 film and it is giving you zero transmittance and it will be able to act like a good mirror, 00:10:58.920 --> 00:11:06.120 okay? And this entire structure is placed on top of a glass substrate. Now you can make narrow band 00:11:06.120 --> 00:11:13.320 MPAs, okay? Metamaterial perfect absorbers for which the top metallic layer is either 00:11:13.320 --> 00:11:20.040 patterned or unpatterned. So here we are taking a patterned metallic layer, okay? Now in this case 00:11:20.040 --> 00:11:25.800 what we are using? You are using a normally incident light with X polarization. However, 00:11:25.800 --> 00:11:30.360 X and Y polarization hardly make any difference here because the structure is symmetric 00:11:30.960 --> 00:11:37.020 along X and Y, right? And you have taken the permittivity of magnesium fluoride is 1. 00:11:37.020 --> 00:11:43.320 9 and these are the parameters that describe the bulk gold permittivity near the near infrared 00:11:43.320 --> 00:11:49.920 wavelength range. You can use Drude model with a plasma frequency of 1.37 into 10 to the power 16 00:11:49.920 --> 00:11:58.920 hertz, okay? And damping constant is this one, okay? So with that you can figure out that one, 00:11:59.520 --> 00:12:06.420 so these are basically simulated reflection spectrum which is plotted when the damping 00:12:06.420 --> 00:12:12.240 constant is considered to be 1 times, 3 times and 5 times of the bulk gold. 00:12:12.240 --> 00:12:19.740 So here you see when you are basically, there is significant difference in the reflection spectrum 00:12:21.240 --> 00:12:29.760 or reflection dip, okay? And you can see that when it is 3 times when the damping constant is 00:12:29.760 --> 00:12:39.120 basically 3 times of that of the bulk gold you are able to get negligible reflection, okay? So it is 00:12:39.120 --> 00:12:47.280 0.28 percent, okay? And this is a scale up to 1 so you can understand, so this is basically 0. 00:12:47.280 --> 00:12:55.560 0028. So it is almost 0 reflection, right? And as I told you in this perfect absorber A is basically 00:12:55.560 --> 00:13:03.900 calculated from what is reflected and then what is transmitted these two are taken out from 1. So 00:13:03.900 --> 00:13:09.540 because of this bulk gold there is 0 transmission so if there is no reflection the entire thing 00:13:09.540 --> 00:13:16.680 is getting absorbed, okay? So here it shows the reflection with 0 intensity is achieved using a 00:13:16.680 --> 00:13:24.060 damping constant that is equal to 3 times of that of the bulk gold. So that actually tells us that 00:13:24.060 --> 00:13:31.800 what should be the ideal thickness of those discs, okay? So if you are able to use very thin discs 00:13:31.800 --> 00:13:37.980 so you can actually get this high damping constant that can give you this perfect absorption. So that 00:13:37.980 --> 00:13:44.220 is where the design of this metamaterial comes into picture. At resonance a strong enhancement 00:13:44.220 --> 00:13:49.260 of the localized electromagnetic field takes place between the two layers the two metallic layers. 00:13:49.260 --> 00:13:53.700 So there is a gold disc and then there is a gold film and in between there is 00:13:53.700 --> 00:14:01.140 a thin spacer layer which is a dielectric. So this electromagnetic energy can be efficiently 00:14:01.140 --> 00:14:07.740 confined in this intermediate spacer layer and that ensures that no light is reflected back. 00:14:08.880 --> 00:14:15.420 So you are basically trapping the energy. So this gives rise to profound reflectance dip in 00:14:15.420 --> 00:14:23.460 the spectrum with nearly 0 intensity and that in turn give rise to nearly 100 percent absorption. 00:14:24.600 --> 00:14:32.340 So in fact this kind of devices work as perfect absorber over a wide range of incident angles. 00:14:32.880 --> 00:14:39.900 Now to better understand the nature what is happening in this particular perfect absorber 00:14:39.900 --> 00:14:45.720 we should look into the current distribution at resonance which was simulated and this is 00:14:45.720 --> 00:14:52.860 the figure that shows that, okay? So here you can see that there are basically anti-parallel current 00:14:52.860 --> 00:14:59.760 distribution in the gold disc and the bottom gold layer. So here it goes like this and here it is 00:14:59.760 --> 00:15:06.240 like this. So you can think of a circulating current like this and this is basically 00:15:07.080 --> 00:15:14.520 can be thought of as a magnetic resonance which comes from the circulating current 00:15:15.180 --> 00:15:21.300 and this current basically results in a magnetic moment which strongly interacts with the magnetic 00:15:21.300 --> 00:15:27.120 field of the incident light. Now at resonance what will happen? A strong enhancement of the 00:15:27.120 --> 00:15:33.720 localized electromagnetic field is established between these two layers, okay And that is the 00:15:33.720 --> 00:15:43.440 reason why you will be having a trap of energy in this magnesium fluoride spacer layer and no light 00:15:43.440 --> 00:15:52.440 is going back. Now simulation study has also been conducted to see the angular dispersion 00:15:52.440 --> 00:15:58.740 of the absorption peak because right now the peak is very very attractive it is giving you almost 00:15:58.740 --> 00:16:04.140 100 percent absorption but you have done only for a normal incidence that is theta equals 0. 00:16:04.140 --> 00:16:13.140 Now when you do the different angle for both the polarization these are the plots that shows you 00:16:13.140 --> 00:16:23.040 with frequency and angle how the absorption peak is changing. So this is for TE polarization and 00:16:23.040 --> 00:16:29.400 this one is for TM polarization. So if you look at the TM polarization first the one on the right the 00:16:29.400 --> 00:16:35.700 absorption peak is seen to be nearly independent of the incident angle. So here on the y-axis you 00:16:35.700 --> 00:16:41.340 are varying the incident angle so it varies from 0 to 80 degrees and here you see that more or less 00:16:41.340 --> 00:16:49.260 it is independent of the incident angle then that is amazing and even at 80 degree you are able to 00:16:49.260 --> 00:16:57.120 at the same wavelength you are able to heat almost 80 percent of absorption sorry 96 percent of 00:16:57.120 --> 00:17:05.280 absorption and that is that is tremendous, okay? That is really working well and this is because 00:17:06.360 --> 00:17:12.420 of the fact that the direction of the magnetic field of the incident light remains unchanged 00:17:12.420 --> 00:17:18.000 with various incident angles and it can effectively drive the circulating currents 00:17:18.000 --> 00:17:24.240 at all angles of incidence and that is why for TM polarization you hardly see any difference. 00:17:25.380 --> 00:17:31.860 However when you look into this particular figure on the left the contour plot here the 00:17:31.860 --> 00:17:37.800 magnetic field cannot drive the circulating currents efficiently at very large angles so 00:17:37.800 --> 00:17:44.340 any anywhere above 50 degree or so you see the absorption has dropped significantly, 00:17:45.000 --> 00:17:54.600 okay? So at 80 degree you actually land up having like 50 percent of absorption so that is a lot 00:17:54.600 --> 00:18:02.220 of drop from the perfect absorber to that. So this is one study that tells you how to design 00:18:03.240 --> 00:18:10.380 metamaterial perfect absorber and always remember that when you change the size of the disk and the 00:18:10.380 --> 00:18:19.260 periodicity that would help you tune the position of this absorption peak, okay? Right now it is 00:18:19.260 --> 00:18:28.440 shown at a particular wavelength and it is or it is in frequency scale so it is close to say 188 00:18:28.440 --> 00:18:38.580 terahertz or so and if you change the periodicity or the size of the disk you should be able to 00:18:38.580 --> 00:18:46.320 tune that, okay? So that is how you can design application specific narrowband perfect absorbers. 00:18:47.640 --> 00:18:53.760 Now moving on to the broadband perfect absorbers let us see how you do that. Now the name itself 00:18:53.760 --> 00:19:00.480 tells you that you are designing something broadband so you should have kind of a possibility 00:19:01.260 --> 00:19:08.880 where you can cater to multiple peaks that can overlap spectrally and that can give 00:19:08.880 --> 00:19:14.040 rise to this broadband nature, right? So broadband definitely you are catering to 00:19:14.040 --> 00:19:20.100 a much wider wavelength range right now. So here is a schematic of that broadband 00:19:21.960 --> 00:19:28.260 absorber but surprisingly it is more or less similar kind of a structure but then 00:19:28.260 --> 00:19:37.260 here the material property is different you are basically using a titanium disk, okay? On top of 00:19:38.220 --> 00:19:47.340 a gold film with a silica layer on the top acting as a spatial layer, right? So 00:19:47.340 --> 00:19:53.100 this particular schematic shows a broadband polarization insensitive and omnidirectional 00:19:53.100 --> 00:20:01.140 absorber working in near infrared range. So that is the range here it has been targeted to and it 00:20:01.140 --> 00:20:07.680 is based on a simple traditional metal dielectric metal or metal insulator metal configuration. 00:20:08.400 --> 00:20:11.820 So here what happens the highly efficient absorption 00:20:12.720 --> 00:20:20.520 is mainly coming from the excitation of low Q localized surface plus bond resonance 00:20:21.120 --> 00:20:27.060 which are supported by the titanium nanodisks and you are also generating the propagating surface 00:20:27.060 --> 00:20:37.080 plus bond resonance in the interface between gold and gold film and silica. So let us look into the 00:20:37.080 --> 00:20:43.800 spectral characteristics here. So this is the black line shows experimental result and the 00:20:43.800 --> 00:20:49.380 dotted line over there shows the simulation result and then they are matching very very 00:20:50.760 --> 00:20:57.720 closely and you can say this like a perfect match between simulation and experiment. 00:20:58.320 --> 00:21:05.340 So here you can see that under normal incidence so here we are only considering theta equals 0, okay? 00:21:05.340 --> 00:21:14.220 The measured absorption of this fabricated sample is over 90% in the spectrum ranging from 900 to 00:21:14.220 --> 00:21:24.300 1825. So this is the range till which it is more than 90%, okay? So this is the experimental one 00:21:24.300 --> 00:21:31.320 and when you do console multi physics numerical simulation for the same structure you also see 00:21:31.320 --> 00:21:39.720 very very similar result and in the simulation you are basically considering a single unit cell 00:21:39.720 --> 00:21:46.380 which has got periodic boundary conditions on both sides to repeat it in both X and Y 00:21:46.380 --> 00:21:53.160 direction and that can give you this particular structure, okay? So numerical simulation also 00:21:53.160 --> 00:22:00.960 shows that it is very very close and because the structure is symmetric along X and Y so it will be 00:22:00.960 --> 00:22:10.140 independent of the polarization of the incident light and you can get very high absorption when 00:22:10.140 --> 00:22:18.540 the incident angle is less than 40 degree. So beyond that there will be drop in the 00:22:18.540 --> 00:22:26.280 absorption. So here are some simulation results that shows you how it works, okay? 00:22:26.880 --> 00:22:34.740 So to reveal the physics physical mechanism in the perfect absorber you can actually plot the 00:22:34.740 --> 00:22:40.800 electric and the magnetic fields, okay? At the two absorption peaks so if you carefully see that 00:22:40.800 --> 00:22:46.200 there is basically one peak and then there is another peak here, okay? So there are basically 00:22:46.200 --> 00:22:52.320 two peaks which are spectrally overlapping one is a narrow peak and another one is a pretty broad 00:22:52.320 --> 00:23:02.580 peak. So there is a peak at 914 and another one at 1468 nanometer, okay? And if you look into the 00:23:03.960 --> 00:23:10.860 electric field distribution these two resonances both look like electric dipolar resonances, 00:23:10.860 --> 00:23:18.360 okay? On the nanodiscs when you are considering the TM polarization. However, 00:23:18.360 --> 00:23:24.540 the magnetic field distribution that you see here is pretty different in these two cases. 00:23:25.140 --> 00:23:33.000 And if you also analyze that what is the origin of this peak then you can say that the short 00:23:33.000 --> 00:23:39.300 wavelength resonance that you are seeing at 914 nanometer is basically considered to be 00:23:39.300 --> 00:23:47.580 the propagating surface plus bond PSP, okay? Resonance that comes between the continuous 00:23:47.580 --> 00:23:55.560 gold film and the silica spacer where the magnetic field is not only strongly confined in 00:23:55.560 --> 00:24:03.540 the gap region between the nanodiscs but also strongly enhanced between the nanodiscs. 00:24:04.500 --> 00:24:13.860 And the other case which is the long wavelength mode so at you can call this as lambda 2 this 00:24:13.860 --> 00:24:19.920 one is lambda 1. So you can say the long wavelength mode which is at lambda 2 1468 00:24:19.920 --> 00:24:26.400 nanometer this is basically a localized surface plus bond resonance peak where the magnetic field 00:24:27.840 --> 00:24:36.420 is mainly concentrated within the gap between the topmost nano antennas and the gold underlay, 00:24:36.420 --> 00:24:43.860 okay? And since titanium is dispersive it is very dispersive and has a relatively large imaginary 00:24:43.860 --> 00:24:49.440 part okay if you look into the dispersion relation you will get to know that the intrinsic absorption 00:24:49.440 --> 00:24:58.140 coefficient of titanium is very large, okay? So you can actually understand that the quality 00:24:58.140 --> 00:25:05.220 factor of the resonance of this localized surface plus bond resonance peak here is rather low and 00:25:05.220 --> 00:25:14.340 that gives you this broadening of the absorption and that is how and this is the difference between 00:25:15.180 --> 00:25:22.440 that gold discs and titanium discs and why we actually opted for titanium discs here when we 00:25:22.440 --> 00:25:29.160 are planning to make a broadband absorber, okay? So once again we can attribute this 00:25:29.160 --> 00:25:36.420 broadband absorption to the combination effect of propagating surface plus bond and low Q localized 00:25:36.420 --> 00:25:47.940 surface plus bond resonance. Now you can see here some of the parametric sweep result it means like 00:25:47.940 --> 00:25:55.860 the influence of some parameters on the absorption performance and how the absorption spectra looks 00:25:55.860 --> 00:26:03.840 like for different materials. So this one is a simulated result where d is the disc diameter that 00:26:03.840 --> 00:26:12.420 has been changed from 380 to 400 to 420 nanometer. The same thing is also seen experimentally and you 00:26:12.420 --> 00:26:19.200 can see how it changes so they are sensitive to the dimensions as I mentioned so you can actually 00:26:19.200 --> 00:26:27.780 design the discs based on your requirement. So again the periodicity is changed here and 00:26:27.780 --> 00:26:36.120 the periodicity also tells you it actually gives you the range over which you want to have this 00:26:36.120 --> 00:26:43.140 broadband absorption. So here periodicity of 580, 600 and 620 nanometer has been studied 00:26:43.140 --> 00:26:47.640 and simulated results and experimental results they are very very close to each other, 00:26:47.640 --> 00:26:54.660 right? And here also you can see if you take the contribution to this absorption coming from gold 00:26:54.660 --> 00:27:02.640 and titanium you can clearly see that titanium is contributing to the most of the absorption. 00:27:02.640 --> 00:27:09.600 So that is the case here this absorber is mainly based on this titanium nano discs. 00:27:09.600 --> 00:27:16.140 Now we understood that how we can make narrow band we understood how we can make broadband absorbers 00:27:16.140 --> 00:27:23.700 now we have to understood we have to understand how we can make ultra broadband metamaterial 00:27:23.700 --> 00:27:31.380 based perfect absorbers. When I say ultra broadband I am thinking of a window something 00:27:31.380 --> 00:27:39.480 as large as 300 to say 4500 nanometer. So it starts from typically you know UV 00:27:39.480 --> 00:27:48.540 visible near infrared and then short and mid IR something like that. So it is typically catering 00:27:48.540 --> 00:27:56.700 to a very very broadband and this is the kind of structures we call them as ultra broadband 00:27:56.700 --> 00:28:01.860 absorbers. So how do you make it? You can actually design them using 2D infinite array of 00:28:01.860 --> 00:28:07.680 hemi ellipsoid shaped metallo-dielectric multilayer structures. So here is the top view 00:28:07.680 --> 00:28:14.220 of the structure so these and this is the side view so you can see it is like a hemi ellipsoid 00:28:14.220 --> 00:28:20.580 so half of the ellipsoid and it is alternating metal dielectric metal dielectric structure. So 00:28:20.580 --> 00:28:28.680 it is based on a silicon substrate but then you have a ground metal ok it is it can be gold it 00:28:28.680 --> 00:28:35.100 or silver it can it has to be 300 nanometer so it blocks light completely and it will reflect ok. 00:28:35.100 --> 00:28:40.740 And then you have dielectric metal dielectric metal you have these are the simulation setup 00:28:40.740 --> 00:28:47.100 basically with PML perfectly matched layers. So this allows you to do a simulation of this 00:28:47.100 --> 00:28:53.280 particular structure for TM polarization ok. Here the H field is considered to be along Y 00:28:53.280 --> 00:29:01.800 axis and the wave propagation is considered to be along Z axis ok and port 1 is the excitation port 00:29:01.800 --> 00:29:08.520 and port 2 is the other port. So if you calculate S11 you can get the reflection 00:29:08.520 --> 00:29:13.980 characteristics and if you calculate S21 you can find out the transmission characteristics 00:29:13.980 --> 00:29:19.980 from the S parameter matrix ok. And here is the periodicity P that is basically the size 00:29:19.980 --> 00:29:28.980 D and the gap G between the hemi ellipsoids. So how many layers we have considered 20 layers 00:29:28.980 --> 00:29:36.540 of metal dielectric alternate structures here molybdenum and germanium are considered and 00:29:37.200 --> 00:29:46.320 not gold we are using tungsten as a ground metal and this is standing over a silicon substrate ok. 00:29:46.320 --> 00:29:51.540 So a perfectly matched layer as I mentioned has been applied here on the top and bottom 00:29:51.540 --> 00:29:56.340 of the unit cell. So this is the side view of the unit cell this is the top view of the unit 00:29:56.340 --> 00:30:03.960 cell that you can see here. d is the diameter of this hemi ellipsoid the base diameter that 00:30:03.960 --> 00:30:11.580 is 400 nanometer and then gap is 20 nanometer ok. The periodicity is 420 nanometer and these 00:30:11.580 --> 00:30:16.860 are the parameters that we have used and other parameters are mentioned in this particular 00:30:16.860 --> 00:30:25.020 figure I will not read out each of them. And then we have then using the RF module of the 00:30:25.020 --> 00:30:33.120 console multi physics software the absorption spectrum has been calculated for 300 to 500 00:30:33.120 --> 00:30:40.560 nanometer spectral window ok. And this allows one to get absorption reflectance and transmittance 00:30:40.560 --> 00:30:46.680 for this structure at normal incidence. So this is that particular structure absorption 00:30:46.680 --> 00:30:53.340 reflectance and transmittance and as you can see here the transmittance the red one is completely 00:30:53.340 --> 00:30:59.520 flat. So there is zero transmittance across the structure and you can also see that the absorption 00:30:59.520 --> 00:31:10.260 reflectance sorry absorption is almost 100 percent of other than this few ringing effects that come 00:31:10.260 --> 00:31:19.200 from multiple resonance here ok. And more or less it is a very flat wide band absorption ok 00:31:19.200 --> 00:31:26.400 and what is not absorbed is kind of reflected. So the green curve shows you the reflection curve 00:31:26.400 --> 00:31:33.000 and this is the plot for the two polarization TE and TM polarization for normal incidence and they 00:31:33.000 --> 00:31:38.820 are perfectly matching because the structure is also symmetric right. So this particular 00:31:38.820 --> 00:31:46.260 structure as I mentioned it is a ultra broadband metamaterial based perfect absorber or this kind 00:31:46.260 --> 00:31:52.320 of absorbers are also called super absorbers because they give you almost 99 percent average 00:31:52.320 --> 00:32:01.440 absorption and that is a big thing 99 percent average absorption between 300 to 400 nanometer 00:32:01.440 --> 00:32:07.920 spectral range at normal incidence. And this particular spectral range it comprises as I 00:32:07.920 --> 00:32:14.760 mentioned earlier UV visible near infrared short wave and mid wave infrared wavelength. 00:32:14.760 --> 00:32:24.780 So that is pretty wide range ok. So this kind of metamaterial absorbers can be designed. 00:32:24.780 --> 00:32:30.240 Now let us look into the applications of this metamaterial perfect absorbers. The 00:32:30.240 --> 00:32:36.240 first application that comes to our mind is solar energy harvesting and when we talk about solar 00:32:36.240 --> 00:32:41.880 energy harvesting the most important thing for us is to know that how the solar spectrum looks like 00:32:41.880 --> 00:32:51.780 ok. So you can see this is the solar spectrum or solar spectral radiation ok versus wavelength. So 00:32:51.780 --> 00:32:59.940 it actually follows this particular pattern ok. So you can actually have a blackbody radiation that 00:33:02.040 --> 00:33:10.620 blackbody radiation picking around 5500 I believe ok that can match this 5500 Kelvin ok that can 00:33:10.620 --> 00:33:14.940 match this particular solar spectrum. But then what is important here is that 00:33:14.940 --> 00:33:20.100 it tells you the range over which you are supposed to absorb. Now if you do not absorb 00:33:20.100 --> 00:33:26.280 anything beyond say 1800 you are not losing much you are only having a very small portion 00:33:26.280 --> 00:33:33.360 of the solar spectrum which lies beyond 1800. So you can actually design your absorber until 00:33:33.360 --> 00:33:39.240 here ok. There are many atoms basically in the literature to design this kind of 00:33:39.960 --> 00:33:46.860 perfect absorber. One such design has been published in this paper that I am showing here. 00:33:46.860 --> 00:33:55.500 So this is a broadband absorber where you have a periodic square array of silica which are coated 00:33:55.500 --> 00:34:07.140 with iron film ok and this is separated from a bottom iron mirror by another silica film 00:34:07.140 --> 00:34:15.900 and a GST film. So this is the 3D view and this is the 2D view of the structure. Again you can 00:34:16.800 --> 00:34:23.880 calculate what is the spectral absorptivity of this structure. You can see that this structure 00:34:24.780 --> 00:34:33.180 can absorb very strongly from 400 to 2000 that is the entire band that we are looking 00:34:33.180 --> 00:34:39.120 for ok. Now this is the absorption that is from this particular structure. 00:34:39.120 --> 00:34:45.720 Now if we tweak this structure a little bit or you redesign this structure your aim would be 00:34:45.720 --> 00:34:53.580 to have a flat absorption line over this entire window and the discussion we had previously those 00:34:53.580 --> 00:35:02.220 kind of structures can give you almost 100% of absorption over this band. Now the question is 00:35:02.220 --> 00:35:08.040 why then we need to design this one again if that previous structure is giving us all that 00:35:08.040 --> 00:35:12.900 we need. Now if you see the previous structure fabrication wise that structure is very very 00:35:12.900 --> 00:35:19.800 challenging. It is a hemispherical shape with alternating layers of metal dielectric and then 00:35:19.800 --> 00:35:28.500 you have to maintain that reducing dimension as well. So that is a very challenging structure on 00:35:28.500 --> 00:35:33.780 in comparison to that this is much easier structure and that can give you an average 00:35:33.780 --> 00:35:40.500 absorption of almost a 90% or so 90 to 95%. So in some cases where you do not have the 00:35:40.500 --> 00:35:46.860 facility to fabricate those complicated structures you can still be happy with this kind of a 00:35:46.860 --> 00:35:54.720 structure. What are the other applications? So this one is basically an application of 00:35:54.720 --> 00:36:03.420 a broadband metamaterial perfect absorber. Now you can also have applications of narrow 00:36:03.420 --> 00:36:09.960 band metamaterial perfect absorbers as thermal emitters. Now what are these thermal emitters? 00:36:10.620 --> 00:36:17.940 Now if you think of a black body it is basically an idealized body that can absorb all radiation 00:36:19.320 --> 00:36:25.500 which falls on it and it re-radiates energy solely determined by its temperature 00:36:26.400 --> 00:36:34.620 as described by the Planck's law. Now when you develop metamaterials okay and you are 00:36:34.620 --> 00:36:42.540 targeting an application of perfect absorber which exhibits the ability to have near uniform 00:36:42.540 --> 00:36:48.840 or near unity absorption in a frequency range you are basically making it like a black body. 00:36:49.500 --> 00:36:54.840 So this same material can also behave like a thermal emitter right just like 00:36:54.840 --> 00:37:00.600 black body radiates your metamaterial can also radiate. According to Kirchhoff's law of thermal 00:37:00.600 --> 00:37:06.660 radiation at equilibrium the emissivity of a material equals to its absorptivity. 00:37:07.260 --> 00:37:13.440 Therefore in principle the metamaterial perfect absorbers can radiate energy as described by their 00:37:13.440 --> 00:37:20.760 absorptivity at a given temperature. Now because of the resonant nature of the metamaterials 00:37:20.760 --> 00:37:27.480 the perfect absorber the narrowband perfect absorbers they yield very sharp resonances 00:37:27.480 --> 00:37:33.720 with high absorption that means they will basically act as very high q thermal emitters 00:37:34.320 --> 00:37:41.460 and they will also have very high emissivity. So that way you can actually make very high q 00:37:41.460 --> 00:37:47.760 high emissivity thermal emitters. So here is one design of infrared 00:37:48.720 --> 00:37:53.640 metamaterial absorber that will also work as a thermal emitter we will see that. So here is a 00:37:53.640 --> 00:38:00.720 structure first you start with a plus type okay plus kind of a structure. So this is a metallic 00:38:00.720 --> 00:38:06.240 structure gold structure on a dielectric and on the back side also you have another gold layer 00:38:06.240 --> 00:38:12.840 okay. So this is the top view of the single band metamaterial absorber and these are the 00:38:12.840 --> 00:38:18.540 dimensions okay all are in microns here length, width, periodicity all are given 00:38:19.140 --> 00:38:25.920 okay and this is a dual band so this is basically a mixture of the small and the large there are 00:38:25.920 --> 00:38:32.100 two structures which resonate at two different frequency band and that is why it is called a 00:38:32.100 --> 00:38:39.900 dual band metamaterial absorber and the dimensions are given here all are in microns again. And these 00:38:39.900 --> 00:38:48.180 two shows the top view these two figure shows the perspective view and in the two cases the 00:38:48.180 --> 00:38:54.600 thickness of the dielectric spacer is 0. 2 micron here for the single band metamaterial 00:38:54.600 --> 00:38:59.700 absorber and it is 0.3 micron in the case of dual band metamaterial absorber. 00:39:00.660 --> 00:39:08.820 Now when you do the experimental absorption study of this structure so here in the inset you can see 00:39:08.820 --> 00:39:17.700 the SEM images. So for the single band structure you get this peak which is pretty good and also 00:39:17.700 --> 00:39:24.240 these are all experimental pictures okay for the dual band you get these two bands which 00:39:24.240 --> 00:39:32.880 are absorbing very strongly right. So that way you can also compare the experimental 00:39:32.880 --> 00:39:38.940 absorptivity with the emissivity and you see that they do follow the law of Kirchhoff that 00:39:38.940 --> 00:39:46.020 we have discussed that the absorption pattern is same as their emission spectrum right. 00:39:47.280 --> 00:39:58.920 So that way you can actually see that you can develop a thermal emitter a high Q thermal 00:39:58.920 --> 00:40:04.680 emitter at a single band a dual band or multiple bands depending on the design of your metamaterial 00:40:04.680 --> 00:40:14.520 absorber okay. So this one shows the absorptivity and emissivity of a single band absorber or you 00:40:14.520 --> 00:40:22.320 can say emitter whereas this one shows a dual band absorber or emitter okay. So that way you 00:40:22.320 --> 00:40:27.780 can actually design the materials. So what is important here you to understand that the 00:40:27.780 --> 00:40:34.020 structures the emission spectrum spectrum is completely dependent on the structure that you 00:40:34.020 --> 00:40:40.080 are designing okay. If you choose a different shape, if you choose a different material, 00:40:40.080 --> 00:40:46.620 if you choose a different periodicity or a thickness the resonance peak can be changed. 00:40:46.620 --> 00:40:53.220 If you want to play with the Q factor of the resonance you can choose if you want 00:40:53.220 --> 00:41:01.920 to lower high Q okay means you want a sharper resonance peak you should select materials which 00:41:01.920 --> 00:41:10.980 are less lossy and then if you want a resonance with a broader Q okay you should actually choose 00:41:10.980 --> 00:41:16.800 materials with high loss that we have seen. Now while choosing the material properties 00:41:16.800 --> 00:41:23.280 you got to be very careful about the dispersion. So you have to choose the material property that 00:41:23.280 --> 00:41:29.820 is suitable for the range the frequency range that you are considering. So usually there are 00:41:29.820 --> 00:41:36.360 websites like refractiveindex.info where you can download and see the dispersion curves of 00:41:36.360 --> 00:41:43.800 different materials and that helps you understand which material could be useful in designing what 00:41:43.800 --> 00:41:52.740 kind of emitters at what frequency band okay. So with that I think we have covered the topics 00:41:52.740 --> 00:42:01.260 and that is all for this lecture we will consider the topics of smart sorry superlens 00:42:01.260 --> 00:42:06.660 and hyperlens in the next lecture and in case you have got any queries on this particular 00:42:06.660 --> 00:42:12.660 lecture you can drop an email to this email address with in the subject line. Thank you.