1 00:00:05,859 --> 00:00:12,389 The last two weeks we have discussed the first and second generation photovoltaic technology. 2 00:00:12,389 --> 00:00:16,550 This week we will discuss some third generation PV concepts. 3 00:00:16,550 --> 00:00:23,550 In addition, we will look at alternative technologies to convert the energy in the sunlight into heat, 4 00:00:24,090 --> 00:00:28,700 solar fuels and again electricity. 5 00:00:28,700 --> 00:00:33,620 First, we will start with third generation PV technology. 6 00:00:33,620 --> 00:00:40,410 As discussed in week 1, third generation technology is technology based on concepts, which are 7 00:00:40,410 --> 00:00:46,190 able to surpass the so-called Shockley-Queisser limit of a single junction solar cell. 8 00:00:46,190 --> 00:00:51,760 The Shockley-Queisser limit as discussed in week 3, is a kind of thermodynamic approach 9 00:00:51,760 --> 00:00:56,329 to estimate the maximum performance of a single junction solar cell. 10 00:00:56,329 --> 00:01:00,670 The AM1.5 spectrum is incident on a solar cell. 11 00:01:00,670 --> 00:01:04,320 We don't allow the solar cell to increase in temperature. 12 00:01:04,320 --> 00:01:11,600 This means that all energy in the incoming AM1.5 spectrum can escape the system of the solar cell 13 00:01:11,600 --> 00:01:18,400 by either the current density generated or by the radiative recombination of charge carriers. 14 00:01:19,070 --> 00:01:26,070 As a result the maximum efficiency which can be achieved is around 33% in the band gap 15 00:01:26,150 --> 00:01:33,150 range from 1.0 eV up to 1.8 eV as indicated by the black area in the shown graph. 16 00:01:33,960 --> 00:01:40,510 Now we are going to look at a very fundamental limitation of the photovoltaic effect as we 17 00:01:40,510 --> 00:01:43,049 have discussed it so far. 18 00:01:43,049 --> 00:01:47,800 What physical principles are limiting the extent of photogeneration? 19 00:01:47,800 --> 00:01:56,000 First, one problem is, that in a single junction only one band gap material is used. 20 00:01:56,080 --> 00:02:03,200 A large fraction of the energy in the most energetic photons is lost as heat as illustrated here. 21 00:02:03,369 --> 00:02:08,039 The arrows represent photons with different energies. 22 00:02:08,039 --> 00:02:14,040 The arrows on the right indicate how much energy is wasted as heat. 23 00:02:14,040 --> 00:02:21,040 Secondly, most solar cells concepts are based on irradiance incidence of 1 sun. 24 00:02:21,230 --> 00:02:24,980 Higher irradiance means more current generation. 25 00:02:24,980 --> 00:02:32,500 Thirdly, every photon only excites one electron in the conduction band creating only one electron-hole pair. 26 00:02:33,770 --> 00:02:39,760 The energy of high energetic photons could be utilized better if they would create more 27 00:02:39,760 --> 00:02:43,819 than one excited electron in the conduction band. 28 00:02:43,819 --> 00:02:48,920 Fourthly, the photons below the band gap are not used. 29 00:02:48,920 --> 00:02:52,790 They do not result in charge carrier excitation. 30 00:02:52,790 --> 00:02:59,390 Finally, the charge carriers populate single energy levels. 31 00:02:59,390 --> 00:03:05,160 Light absorption excites electrons and holes up into the conduction or down into the 32 00:03:05,160 --> 00:03:06,550 valence band. 33 00:03:06,550 --> 00:03:11,890 However, the charge carriers relax very fast, the electron to the bottom of the conduction 34 00:03:11,890 --> 00:03:16,410 band and the hole to the top of the valence band. 35 00:03:16,410 --> 00:03:21,700 The potential difference between the charge carriers right after the excitation is not utilized. 36 00:03:22,050 --> 00:03:28,260 In addition, in the band are no states, so the light below the band gap does not excite 37 00:03:28,260 --> 00:03:29,720 any charge carriers. 38 00:03:29,720 --> 00:03:36,100 Tackling these fundamental limitations means that we can develop PV concepts with conversion 39 00:03:36,100 --> 00:03:41,849 efficiencies that can surpass the Shockley-Queisser limit. 40 00:03:41,849 --> 00:03:45,730 Tackling the first problem has been discussed last week. 41 00:03:45,730 --> 00:03:51,830 Both the III-V semiconductor PV technology and the thin-film silicon technology use the 42 00:03:51,830 --> 00:03:58,599 concept of multi-junctions - several solar cell junctions stacked upon each other with 43 00:03:58,599 --> 00:04:02,129 a response to different parts of the solar spectrum. 44 00:04:02,129 --> 00:04:09,000 The second problem, that concepts are based on 1-sun irradiance, is tackled using concentrator solar cells. 45 00:04:09,700 --> 00:04:16,700 As discussed last week, the concentrator technology is applied on multi-junctions based on III-V semiconductor materials. 46 00:04:17,940 --> 00:04:23,749 As a result, the highest conversion efficiencies of 44% have been achieved. 47 00:04:23,749 --> 00:04:29,689 This exceeds the Shockley-Queisser limit by more than 10%. 48 00:04:29,689 --> 00:04:36,499 The fact that every photon can only generate at maximum one collected electron can in theory 49 00:04:36,499 --> 00:04:38,909 be tackled by two approaches. 50 00:04:38,909 --> 00:04:44,499 The first one is down-conversion, which is a spectral conversion approach. 51 00:04:44,499 --> 00:04:51,469 The high energetic photons are split in two or even more low energetic photons, before 52 00:04:51,469 --> 00:04:54,369 being absorbed in the PV active part. 53 00:04:54,369 --> 00:05:01,369 As a result high energetic photons can result in more than one electron being collected. 54 00:05:02,930 --> 00:05:08,909 The second approach to enhance the charge carrier excitation by a single energetic photon 55 00:05:08,909 --> 00:05:14,409 is called multiple exciton generation (MEG). 56 00:05:14,409 --> 00:05:21,409 Here nanostructure semiconductor materials might be able to convert the energy of a photon 57 00:05:21,610 --> 00:05:26,309 in two or more excited electron-hole pairs. 58 00:05:26,309 --> 00:05:31,249 The problem of the non-use of the photon below the band gap can be tackled with a spectral 59 00:05:31,249 --> 00:05:33,419 conversion approach as well. 60 00:05:33,419 --> 00:05:39,289 Here the same low energetic photons, which are transmitted through the solar cell, are 61 00:05:39,289 --> 00:05:46,300 converted into one photon with an energy above that of the band gap of the semiconductor material. 62 00:05:46,550 --> 00:05:53,270 If this photon is reflected back in the material it can be absorbed. 63 00:05:53,270 --> 00:05:58,199 The last problem in the list is that of a single population of each charge carrier. 64 00:05:58,199 --> 00:06:02,979 In theory, this can be tackled by hot carrier solar cells. 65 00:06:02,979 --> 00:06:08,379 Here the charge carriers are collected, just after light excitation, before they are relaxed 66 00:06:08,379 --> 00:06:11,509 back to the edges of the electronic bands. 67 00:06:11,509 --> 00:06:14,699 This improves the band gap energy utilization. 68 00:06:14,699 --> 00:06:19,909 Another, theoretical solution is the concept of intermediate band solar cells. 69 00:06:19,909 --> 00:06:25,779 This is the concept in which intentionally an electronic band within the band gap is 70 00:06:25,779 --> 00:06:33,000 engineered to enable photons below the band gap to help additional electrons to be excited as well. 71 00:06:33,610 --> 00:06:39,889 Note, that besides multi-junction and the concentrator approach, none of these concepts 72 00:06:39,889 --> 00:06:46,039 have resulted in high-efficiency solar cells or even been demonstrated yet. 73 00:06:46,039 --> 00:06:50,999 These other concepts are still in a fundamental research phase. 74 00:06:50,999 --> 00:06:57,999 Here I will give you a quick introduction to down-conversion, multiple exciton generation, 75 00:06:58,699 --> 00:07:04,949 up-conversion, hot carrier solar cells and intermediate band solar cells. 76 00:07:04,949 --> 00:07:10,889 Let's start with spectral conversion, which results in splitting one photon in multiple 77 00:07:10,889 --> 00:07:13,179 lower energetic photons. 78 00:07:13,179 --> 00:07:19,490 It might tackle the problem that per photon only one electron is excited. 79 00:07:19,490 --> 00:07:26,490 A down convertor is a magic material that absorbs a high energetic photon and converts 80 00:07:26,550 --> 00:07:30,520 this photon in at least two lower energetic photons. 81 00:07:30,520 --> 00:07:36,740 If the energy of both photons is still larger than that of the band gap of the photovoltaic 82 00:07:36,740 --> 00:07:42,550 material, both photons can be absorbed and used for exciting charge carriers. 83 00:07:42,550 --> 00:07:48,889 As a result, a high energetic photon, like a photon in the blue visible part, can result 84 00:07:48,889 --> 00:07:52,699 in two excited electrons in the blue part. 85 00:07:52,699 --> 00:08:00,600 In other words, the maximum theoretical EQE of 100% at the wavelength of the blue photon 86 00:08:00,700 --> 00:08:04,100 can be increased to 200%. 87 00:08:05,139 --> 00:08:10,159 If the photon has enough energy to be split into three photons with an energy higher than 88 00:08:10,159 --> 00:08:15,639 the band gap, a theoretical EQE of 300% could be obtained. 89 00:08:15,639 --> 00:08:22,639 So spectral down-conversions first convert photons into lower energetic photons, and 90 00:08:22,729 --> 00:08:26,949 these photons are used in photovoltaic devices. 91 00:08:26,949 --> 00:08:33,949 Multiple exciton generation, abbreviated with MEG, is another approach, which can accomplish 92 00:08:34,560 --> 00:08:39,579 the excitation of more than one electron-hole pair per photon. 93 00:08:39,579 --> 00:08:44,699 The difference with spectral down-conversion is that all fundamental energy conversion 94 00:08:44,699 --> 00:08:47,759 steps occur in the PV active layer. 95 00:08:47,759 --> 00:08:54,459 In a normal semiconductor material, a high energetic photon has some rest energy, which 96 00:08:54,459 --> 00:08:58,110 is not used to excite the electron. 97 00:08:58,110 --> 00:09:02,990 This excess energy is usually lost as heat. 98 00:09:02,990 --> 00:09:08,670 In down-conversion approach this rest energy is transferred as a quantized energy package 99 00:09:08,670 --> 00:09:14,589 within the material, where it can excite a second electron into the conduction band. 100 00:09:14,589 --> 00:09:21,589 It means that the energy in one photon, results in two excited electrons. 101 00:09:22,129 --> 00:09:27,839 The requirement of course is that the energy in the initial photon is at least two times 102 00:09:27,839 --> 00:09:29,829 that of the band gap energy. 103 00:09:29,829 --> 00:09:35,389 In this way, theoretical EQE of 200% can be achieved as well. 104 00:09:35,389 --> 00:09:40,980 If we would have a photon with an energy larger than three times the band gap, a theoretical 105 00:09:40,980 --> 00:09:44,440 EQE of 300% could be achieved. 106 00:09:44,440 --> 00:09:51,440 Both for down-conversion and multiple exciton generation, nanostructured semiconductors are 107 00:09:51,459 --> 00:09:53,620 studied and developed. 108 00:09:53,620 --> 00:09:58,060 These structures are based on so-called quantum dots. 109 00:09:58,060 --> 00:10:05,060 Quantum dots are small spherical nanoparticles made of semiconductor materials with typical diameters 110 00:10:05,900 --> 00:10:07,490 of a few nanometers. 111 00:10:07,490 --> 00:10:13,920 These semiconductor particles still behave like a semiconductor material; however, due 112 00:10:13,920 --> 00:10:19,930 to quantum mechanical effects the band gap can be larger of the semiconductor quantum 113 00:10:19,930 --> 00:10:26,660 dots in reference to the band gap of the semiconductor material in large bulk materials. 114 00:10:26,660 --> 00:10:31,339 The band gap can be tuned by the size of the nanoparticles. 115 00:10:31,339 --> 00:10:35,810 The smaller the particles, the larger the band gap. 116 00:10:35,810 --> 00:10:41,870 This enables interesting band gap engineering possibilities, such as a multi-junction solar cell, 117 00:10:41,870 --> 00:10:47,399 based on junctions with different quantum dot sizes in every junction. 118 00:10:47,399 --> 00:10:53,629 To use QDs for down-conversion or multiple exciton generation, an ensemble of nanoparticles 119 00:10:53,629 --> 00:10:56,649 is embedded in a host material. 120 00:10:56,649 --> 00:11:03,379 How can the properties of an ensemble of nanoparticles be used for down-conversion? 121 00:11:03,379 --> 00:11:08,990 Here we see the electronic band gap diagram of two nanoparticles. 122 00:11:08,990 --> 00:11:15,629 The particles are at a very close distance from each other, in the order of a nanometer. 123 00:11:15,629 --> 00:11:19,930 In one particle an electron is excited into the conduction band. 124 00:11:19,930 --> 00:11:26,930 It appears that in such nanoparticle systems the quantized rest energy is not necessarily 125 00:11:27,180 --> 00:11:34,100 lost as heat to the lattice, but can be transferred as a quantized energy package 126 00:11:34,180 --> 00:11:37,930 to a neighboring quantum dot. 127 00:11:37,930 --> 00:11:43,829 Here a second electron is excited into conduction band of the second quantum dot. 128 00:11:43,829 --> 00:11:49,810 Now we have generated two electron-hole pairs out of one photon. 129 00:11:49,810 --> 00:11:56,560 If other recombination loss mechanisms like Auger recombination and SRH recombination 130 00:11:56,560 --> 00:12:04,100 can be suppressed, these electron-hole pairs in both quantum dots can radiatively recombine. 131 00:12:04,360 --> 00:12:09,980 It means they send out two red-ish photons. 132 00:12:09,980 --> 00:12:16,980 As a result, one blue-ish photon is converted into two red-ish photons, which can be absorbed 133 00:12:17,439 --> 00:12:21,670 by a PV material. 134 00:12:21,670 --> 00:12:26,800 Multiple exciton generation in an ensemble of quantum dots is quite similar. 135 00:12:26,800 --> 00:12:32,449 Again, in one particle an electron is excited into the conduction band. 136 00:12:32,449 --> 00:12:39,449 This quantized energy package is transferred to a neighboring quantum dot. 137 00:12:39,620 --> 00:12:45,579 Here a second electron is excited into the conduction band of the second quantum dot. 138 00:12:45,579 --> 00:12:50,930 If the charge carriers are separated and collected before they recombine, the result is that 139 00:12:50,930 --> 00:12:57,930 one photon is able to produce more than one collected electron. 140 00:12:58,209 --> 00:13:04,430 Here you see some experimental results on down-conversion based on silicon quantum dots 141 00:13:04,430 --> 00:13:10,389 in a narrow spectral range from a paper of Jursberg et al. 142 00:13:10,389 --> 00:13:14,540 The horizontal axis represents the photon emission wavelength. 143 00:13:14,540 --> 00:13:20,790 At around 790 nm a down-conversion efficiency of 60% is achieved. 144 00:13:20,790 --> 00:13:29,600 Here we see the results out of paper of Semonin demonstrating that EQE above 100% can be achieved. 145 00:13:29,620 --> 00:13:38,600 The absorber layer contains PbSe quantum dots and in the blue region, from 3.1 up to 3.4 eV, 146 00:13:38,629 --> 00:13:43,449 the quantum dots realize an EQE above 100%. 147 00:13:43,449 --> 00:13:49,240 The challenge is to move this spectral response to lower photon energies as the solar spectrum 148 00:13:49,240 --> 00:13:53,499 contains far more photons in this spectral range. 149 00:13:53,499 --> 00:14:00,499 Spectral up-conversion is the process in which two or more low energetic photons excite 150 00:14:00,519 --> 00:14:05,889 electrons many steps from the ground state up to higher excited states. 151 00:14:05,889 --> 00:14:12,230 If the electrons occupy the higher excited states they can fall directly back to the ground 152 00:14:12,230 --> 00:14:17,170 state by sending out higher energetic photons. 153 00:14:17,170 --> 00:14:24,170 As a result many low energetic photons result in one high energetic photon. 154 00:14:24,319 --> 00:14:28,639 This photon can be absorbed in the PV active material. 155 00:14:28,639 --> 00:14:34,709 So many low energetic photons can excite charge carriers as well. 156 00:14:34,709 --> 00:14:40,309 Studies have been performed on materials containing rare-Earth ions, like erbium. 157 00:14:40,309 --> 00:14:47,600 They can pump up electrons by infrared photons to an excited level, which emits in the visible light. 158 00:14:47,779 --> 00:14:53,939 However, the problem is that these systems have low efficiencies and only have a response 159 00:14:53,939 --> 00:14:56,829 in a very narrow spectral range. 160 00:14:56,829 --> 00:15:01,350 Next we look at the intermediate band solar cell. 161 00:15:01,350 --> 00:15:08,350 This concept tries to tackle the problem that each charge carrier only has a single population state. 162 00:15:10,139 --> 00:15:16,540 Here we see a semiconductor material with valence band and conduction band. 163 00:15:16,540 --> 00:15:22,529 An intermediate band material contains a narrow electronic band in the band gap as well. 164 00:15:22,529 --> 00:15:27,350 Such structure is believed to increase the spectral utilization. 165 00:15:27,350 --> 00:15:32,800 The high energetic photons can excite an electron from the valence band into the conduction band, 166 00:15:32,800 --> 00:15:36,029 just like in a normal semiconductor material. 167 00:15:36,029 --> 00:15:42,540 The difference is, that photons below the band gap can excite an electron from the valence 168 00:15:42,540 --> 00:15:44,319 band into the intermediate band. 169 00:15:44,319 --> 00:15:50,619 A second photon is required to excite the electron from the intermediate band into the 170 00:15:50,619 --> 00:15:53,009 conduction band. 171 00:15:53,009 --> 00:15:59,900 Various ideas exist how these charge carriers are excited, transported and collected in those devices. 172 00:16:00,180 --> 00:16:07,100 I won't go into detail, but in this slide you see some of the ideas. 173 00:16:07,100 --> 00:16:13,610 Most important is that the photons below the band gap also result in an excited charge carrier. 174 00:16:14,300 --> 00:16:21,300 In addition, the two photons with energy smaller than the band gap can effectively result in quasi-Fermi 175 00:16:21,970 --> 00:16:27,559 level splittings larger than the energy of one of the low energetic photons. 176 00:16:27,559 --> 00:16:32,449 The last concept we will discuss is the hot carrier solar cell. 177 00:16:32,449 --> 00:16:37,180 It basically improves the band gap energy utilization. 178 00:16:37,180 --> 00:16:42,689 Here we see the light-excited charge carriers in a semiconductor material. 179 00:16:42,689 --> 00:16:48,490 The population of the charge carrier levels reflects the situation just after the excitation 180 00:16:48,490 --> 00:16:51,279 by the absorption of a photon. 181 00:16:51,279 --> 00:16:55,589 This distribution is not in thermal equilibrium. 182 00:16:55,589 --> 00:17:00,009 The electrons are excited into a position higher in the conduction band. 183 00:17:00,009 --> 00:17:04,860 The holes are excited down to a lower level in the valence band. 184 00:17:04,860 --> 00:17:10,350 Such charge carriers are called hot electrons and hot holes. 185 00:17:10,350 --> 00:17:15,429 It takes only a few picoseconds for the hot charge carriers to relax back to the edges 186 00:17:15,429 --> 00:17:17,399 of the electronic bands. 187 00:17:17,399 --> 00:17:24,029 A hot carrier solar cell is based on the collection of charge carriers when they are still hot. 188 00:17:24,029 --> 00:17:31,600 It means that the energy larger than the band gap energy could be utilized per excited charge carrier. 189 00:17:32,559 --> 00:17:38,159 The fundamental challenge is to collect the hot carriers before they relax back to the 190 00:17:38,159 --> 00:17:42,940 edge of the electronic bands. 191 00:17:42,940 --> 00:17:46,950 Such a concept would require selective contacts. 192 00:17:46,950 --> 00:17:52,900 These are contacts which only select electrons above a particular energy level in the conduction band 193 00:17:52,900 --> 00:18:01,000 and contacts that selectively collect holes below a certain energy level in the valence band. 194 00:18:01,390 --> 00:18:07,000 In theory, the band gap utilization could be higher than the band gap itself. 195 00:18:07,000 --> 00:18:13,370 At the moment the main challenge is to increase the lifetime of the hot charge carriers, such 196 00:18:13,370 --> 00:18:20,370 that they have the time to move from the absorber layer to the selective contacts. 197 00:18:20,679 --> 00:18:27,679 With the third generation PV concepts we finish our introduction into the various PV technologies. 198 00:18:27,730 --> 00:18:33,669 Now we are going to focus on harvesting the heat out of the solar light. 199 00:18:33,669 --> 00:18:38,940 In the next block we will discuss solar thermal technology. 200 00:18:38,940 --> 00:18:39,970 See you in the next block.