0 00:00:00,000 --> 00:00:30,000 Dear viewer, these subtitles were generated by a machine via the service Trint and therefore are (very) buggy. If you are capable, please help us to create good quality subtitles: https://c3subtitles.de/talk/866 Thanks! 1 00:00:18,210 --> 00:00:20,549 Ex fcl 2 00:00:20,550 --> 00:00:22,979 x ray free electron 3 00:00:22,980 --> 00:00:25,379 lasers, what's 4 00:00:25,380 --> 00:00:26,380 what's that? 5 00:00:27,740 --> 00:00:30,499 Yeah, something something like that, 6 00:00:30,500 --> 00:00:32,149 it's just a little bit of a shorter 7 00:00:32,150 --> 00:00:34,999 wavelength, and it doesn't definitely 8 00:00:35,000 --> 00:00:37,489 fit the size of a pen. 9 00:00:37,490 --> 00:00:39,919 Well, Preston Hallatt 10 00:00:39,920 --> 00:00:42,119 is a physicist working at 11 00:00:42,120 --> 00:00:44,899 the Deutche Electronic Synchrotron 12 00:00:44,900 --> 00:00:47,119 in Hamburg, and he 13 00:00:47,120 --> 00:00:48,409 knows the answers to that. 14 00:00:50,270 --> 00:00:52,339 And he's going to introduce 15 00:00:52,340 --> 00:00:54,439 us to his world of free electron 16 00:00:54,440 --> 00:00:56,539 lasers and their applications 17 00:00:56,540 --> 00:00:58,159 warm. Welcome to Toreson. 18 00:01:04,459 --> 00:01:06,029 Yeah, thank you. 19 00:01:07,220 --> 00:01:09,319 I have to admit, I'm quite nervous, 20 00:01:09,320 --> 00:01:11,179 not only because of this thing, but also 21 00:01:11,180 --> 00:01:13,519 because my computer broke and I got this 22 00:01:13,520 --> 00:01:15,619 about one hour ago, and I don't know if 23 00:01:15,620 --> 00:01:17,959 the presentation works on that as 24 00:01:17,960 --> 00:01:20,269 well, but let's hope for the best. 25 00:01:20,270 --> 00:01:22,309 Despite that, I'm very happy to see that 26 00:01:22,310 --> 00:01:24,139 so many of you are interested in particle 27 00:01:24,140 --> 00:01:25,399 accelerators. 28 00:01:25,400 --> 00:01:27,079 And I have to say that this is not a one 29 00:01:27,080 --> 00:01:28,339 directional relation. 30 00:01:28,340 --> 00:01:30,139 I talked to many colleagues at DC and 31 00:01:30,140 --> 00:01:32,239 also in the US and all of them 32 00:01:32,240 --> 00:01:34,339 literally knew about this Congress, 33 00:01:34,340 --> 00:01:35,959 and most of them even knew that it was 34 00:01:35,960 --> 00:01:38,149 going to happen in Leipzig this year. 35 00:01:38,150 --> 00:01:40,249 So I think I can say that at least 36 00:01:40,250 --> 00:01:42,379 every particle accelerator physicist I 37 00:01:42,380 --> 00:01:44,869 know likes the CCC as well 38 00:01:44,870 --> 00:01:47,629 and is interested in this Congress. 39 00:01:47,630 --> 00:01:48,630 OK, but. 40 00:01:52,330 --> 00:01:54,609 Maybe enough small talk for now. 41 00:01:54,610 --> 00:01:56,649 Let's get to some science. 42 00:01:56,650 --> 00:01:58,749 So while you're watching this talk, 43 00:01:58,750 --> 00:02:00,969 your neurons firing incessantly, 44 00:02:00,970 --> 00:02:03,069 sending electrical impulses to 45 00:02:03,070 --> 00:02:04,539 the neighboring neurons. 46 00:02:04,540 --> 00:02:06,129 But how does this work? 47 00:02:06,130 --> 00:02:08,819 I mean, what are the neurons made of? 48 00:02:08,820 --> 00:02:10,418 Well, artists, you from Harvard 49 00:02:10,419 --> 00:02:13,089 University. Let us have a look inside. 50 00:02:13,090 --> 00:02:14,739 So inside each neuron, there's an 51 00:02:14,740 --> 00:02:17,049 enormous variety of proteins, big 52 00:02:17,050 --> 00:02:19,089 macromolecules, each consisting of 53 00:02:19,090 --> 00:02:20,620 hundreds of thousands of atoms 54 00:02:21,820 --> 00:02:24,069 of up to 40 percent of the volume 55 00:02:24,070 --> 00:02:25,749 of each cell is occupied by these 56 00:02:25,750 --> 00:02:27,769 proteins and by the DNA. 57 00:02:27,770 --> 00:02:29,379 It serves as a blueprint. 58 00:02:29,380 --> 00:02:31,329 The proteins are manufactured somewhere 59 00:02:31,330 --> 00:02:33,129 inside the cell and then have to be 60 00:02:33,130 --> 00:02:34,989 transported to the destination where they 61 00:02:34,990 --> 00:02:35,889 are needed. 62 00:02:35,890 --> 00:02:38,919 For example, membrane proteins have to be 63 00:02:38,920 --> 00:02:40,599 transported to the outer shell of the 64 00:02:40,600 --> 00:02:41,600 cell. Right. 65 00:02:42,520 --> 00:02:44,049 And this is done by the so-called 66 00:02:44,050 --> 00:02:45,849 vesicles like the Blue Fellow you see 67 00:02:45,850 --> 00:02:48,129 over there. So the proteins stick on them 68 00:02:48,130 --> 00:02:50,259 and then motor proteins like this. 69 00:02:50,260 --> 00:02:52,929 Kinnison here pulls the 70 00:02:52,930 --> 00:02:55,029 vesicle along these long molecular 71 00:02:55,030 --> 00:02:57,279 strands, which sponsor of the cell 72 00:02:57,280 --> 00:02:58,280 here, the green one. 73 00:02:59,300 --> 00:03:00,699 I don't know if you've seen such an 74 00:03:00,700 --> 00:03:01,959 animation before. 75 00:03:01,960 --> 00:03:03,939 When I saw this movie for the first time 76 00:03:03,940 --> 00:03:05,919 and when I realized about the enormous 77 00:03:05,920 --> 00:03:07,959 complexity of the molecular basis of 78 00:03:07,960 --> 00:03:10,089 life, this was literally 79 00:03:10,090 --> 00:03:11,979 breathtaking to me. 80 00:03:11,980 --> 00:03:13,749 But have you ever wondered, how do we 81 00:03:13,750 --> 00:03:15,519 know about this? I mean, how do we know 82 00:03:15,520 --> 00:03:17,079 about the structure of this kind of 83 00:03:17,080 --> 00:03:18,309 protein? 84 00:03:18,310 --> 00:03:20,559 And the answer is synchrotron 85 00:03:20,560 --> 00:03:22,719 light source. So the vast majority 86 00:03:22,720 --> 00:03:24,939 of these proteins have been resolved 87 00:03:24,940 --> 00:03:26,829 in the latest third generation 88 00:03:26,830 --> 00:03:28,359 synchrotron light sources. 89 00:03:28,360 --> 00:03:29,889 And in this talk, I will show you what it 90 00:03:29,890 --> 00:03:31,989 takes to build such a machine and how 91 00:03:31,990 --> 00:03:33,549 to make a picture. 92 00:03:33,550 --> 00:03:35,169 But then the next question is, how do we 93 00:03:35,170 --> 00:03:36,699 know about the dynamics? 94 00:03:36,700 --> 00:03:38,799 So how do we know how how these proteins 95 00:03:38,800 --> 00:03:39,939 fold? 96 00:03:39,940 --> 00:03:42,069 And to be honest, we have no 97 00:03:42,070 --> 00:03:43,519 fucking clue. 98 00:03:43,520 --> 00:03:45,639 So don't get fooled 99 00:03:45,640 --> 00:03:47,299 by the name Harvard University. 100 00:03:47,300 --> 00:03:49,449 This is just an artist you and we have no 101 00:03:49,450 --> 00:03:51,279 idea how a protein folds. 102 00:03:51,280 --> 00:03:53,229 No one has ever seen something like this 103 00:03:53,230 --> 00:03:55,629 or witnessed a chemical reaction. 104 00:03:55,630 --> 00:03:57,249 But by the end of this talk, I will have 105 00:03:57,250 --> 00:03:59,049 shown you that just by now we have a 106 00:03:59,050 --> 00:04:01,149 machine at hand, the X-ray free electron 107 00:04:01,150 --> 00:04:03,519 laser, which might be able to resolve 108 00:04:03,520 --> 00:04:05,199 these proteins within their natural 109 00:04:05,200 --> 00:04:07,720 timescale of a couple of femtoseconds. 110 00:04:09,160 --> 00:04:11,199 But OK, to start and bring everyone on 111 00:04:11,200 --> 00:04:12,819 the same page, I have to recall the 112 00:04:12,820 --> 00:04:15,129 electromagnetic spectrum. 113 00:04:15,130 --> 00:04:16,989 We are surrounded by a variety of 114 00:04:16,990 --> 00:04:18,669 electromagnetic waves which can be 115 00:04:18,670 --> 00:04:20,859 categorized according to their wavelength 116 00:04:20,860 --> 00:04:22,689 into different ways. 117 00:04:22,690 --> 00:04:24,789 For example, radio waves with a couple of 118 00:04:24,790 --> 00:04:26,829 meters or more in wavelengths. 119 00:04:26,830 --> 00:04:28,449 Then we have the microwaves with some 120 00:04:28,450 --> 00:04:30,699 centimeters and then the infrared 121 00:04:30,700 --> 00:04:32,319 and the visible light with a couple of 122 00:04:32,320 --> 00:04:34,660 hundred of nanometers and wavelength. 123 00:04:36,620 --> 00:04:38,619 If we decrease the wavelength further, we 124 00:04:38,620 --> 00:04:39,999 will get to the ultraviolet. 125 00:04:40,000 --> 00:04:42,369 And then finally, at 0.01 nanometer 126 00:04:42,370 --> 00:04:45,039 or one angstrom, we have x rays, 127 00:04:45,040 --> 00:04:47,409 OK? And there exists the fundamental 128 00:04:47,410 --> 00:04:48,789 limitation if you want to observe 129 00:04:48,790 --> 00:04:50,799 something with electromagnetic radiation, 130 00:04:50,800 --> 00:04:52,869 namely the diffraction limit. 131 00:04:52,870 --> 00:04:55,179 So it says basically 132 00:04:55,180 --> 00:04:57,189 that if you want to resolve two objects 133 00:04:57,190 --> 00:04:59,199 with the distance D, then you have to use 134 00:04:59,200 --> 00:05:01,089 a wavelength, which is in the order of 135 00:05:01,090 --> 00:05:03,099 that distance or smaller. 136 00:05:03,100 --> 00:05:05,169 So if you want to study an end or a 137 00:05:05,170 --> 00:05:07,299 bacteria, you can use visible light 138 00:05:07,300 --> 00:05:09,099 because visible light has a wavelength 139 00:05:09,100 --> 00:05:11,679 smaller than the size of these objects. 140 00:05:11,680 --> 00:05:13,749 But if we want to study viruses or 141 00:05:13,750 --> 00:05:16,029 the proteins we just saw or even smaller 142 00:05:16,030 --> 00:05:18,759 molecules, we have to use X-rays. 143 00:05:18,760 --> 00:05:20,949 But actually our ways of making a picture 144 00:05:20,950 --> 00:05:22,509 from something that small is quite 145 00:05:22,510 --> 00:05:24,129 different from what you are used to with 146 00:05:24,130 --> 00:05:25,809 your eyes or your camera. 147 00:05:25,810 --> 00:05:28,299 So we do X-ray diffraction images, 148 00:05:28,300 --> 00:05:30,129 and before I can show you how that works, 149 00:05:30,130 --> 00:05:31,419 I have to tell you something about 150 00:05:31,420 --> 00:05:32,979 coherence. 151 00:05:32,980 --> 00:05:35,139 So we start with an ordinary light source 152 00:05:35,140 --> 00:05:36,669 which emits light in different 153 00:05:36,670 --> 00:05:38,709 wavelengths, which is here indicated by 154 00:05:38,710 --> 00:05:40,869 the different colors and the origin of 155 00:05:40,870 --> 00:05:42,829 these wavefront is spread out. 156 00:05:42,830 --> 00:05:45,099 OK, so we have no fixed Face-To-Face 157 00:05:45,100 --> 00:05:47,259 relation at one point in space 158 00:05:47,260 --> 00:05:49,179 and this we call incoherent light. 159 00:05:49,180 --> 00:05:50,529 That's the kind of light we are all 160 00:05:50,530 --> 00:05:51,530 surrounded by. 161 00:05:52,630 --> 00:05:54,219 You may know from high school physics 162 00:05:54,220 --> 00:05:56,649 that if I place an aperture in here, then 163 00:05:56,650 --> 00:05:58,839 the wavefront propagate as they would be 164 00:05:58,840 --> 00:06:01,059 emitted by a point source in the whole 165 00:06:01,060 --> 00:06:02,289 of this aperture. 166 00:06:02,290 --> 00:06:04,389 And now we have a fixed face relation at 167 00:06:04,390 --> 00:06:06,009 one point in space and we call this 168 00:06:06,010 --> 00:06:08,049 spatially coherent light. 169 00:06:08,050 --> 00:06:09,999 And the next step towards coherence is 170 00:06:10,000 --> 00:06:12,519 then to put a filter in 171 00:06:12,520 --> 00:06:14,679 which lets through only one particular 172 00:06:14,680 --> 00:06:15,739 wavelength. 173 00:06:15,740 --> 00:06:18,159 OK, so now this is coherent light, 174 00:06:18,160 --> 00:06:19,959 and if we consider ourselves to be very 175 00:06:19,960 --> 00:06:22,419 far away from the source, we can consider 176 00:06:22,420 --> 00:06:24,789 these ways to be plain ways. 177 00:06:24,790 --> 00:06:26,859 And then if I play something, 178 00:06:26,860 --> 00:06:28,959 for example, a double slit in here, I 179 00:06:28,960 --> 00:06:30,369 will get an interference pattern 180 00:06:30,370 --> 00:06:32,499 downstream and on the screen I would be 181 00:06:32,500 --> 00:06:34,569 able to detect the 182 00:06:34,570 --> 00:06:36,129 diffraction pattern. 183 00:06:36,130 --> 00:06:38,319 And the clue is now that 184 00:06:38,320 --> 00:06:40,599 mathematically there exists the relation 185 00:06:40,600 --> 00:06:42,039 between the diffraction pattern and the 186 00:06:42,040 --> 00:06:44,379 physical arrangement of these objects. 187 00:06:44,380 --> 00:06:46,539 So I am able, if I know the diffraction 188 00:06:46,540 --> 00:06:48,639 pattern and I know the distance between 189 00:06:48,640 --> 00:06:50,799 the screen and the object, I'm 190 00:06:50,800 --> 00:06:52,499 able. Calculate from the diffraction 191 00:06:52,500 --> 00:06:54,749 pattern, the physical arrangement 192 00:06:54,750 --> 00:06:56,849 of these objects, and in our case, 193 00:06:56,850 --> 00:06:59,009 we're doing X-ray diffraction, 194 00:06:59,010 --> 00:07:01,049 so we we don't have double slits, but we 195 00:07:01,050 --> 00:07:03,049 have electrons on which the photons get 196 00:07:03,050 --> 00:07:04,409 scattered. 197 00:07:04,410 --> 00:07:06,479 And to give you an example, this is 198 00:07:06,480 --> 00:07:08,699 a microscopic image from a sample which 199 00:07:08,700 --> 00:07:10,919 was hit by an X-ray laser pulse. 200 00:07:10,920 --> 00:07:12,569 And this is the diffraction pattern you 201 00:07:12,570 --> 00:07:14,850 record at the screen at your detector. 202 00:07:16,380 --> 00:07:18,419 So it's a bit more difficult than the 203 00:07:18,420 --> 00:07:19,529 previous example. 204 00:07:19,530 --> 00:07:21,809 But the key is this is the reconstructed 205 00:07:21,810 --> 00:07:23,999 image. So from this, you 206 00:07:24,000 --> 00:07:25,619 are able to calculate this one, these 207 00:07:25,620 --> 00:07:27,839 two, although this is not very intuitive 208 00:07:27,840 --> 00:07:30,009 or mathematically equivalent. 209 00:07:30,010 --> 00:07:32,369 Okay, we can calculate this from 210 00:07:32,370 --> 00:07:34,409 the diffraction pattern without knowing 211 00:07:34,410 --> 00:07:36,839 the original original sample 212 00:07:36,840 --> 00:07:38,759 and this kind of X-ray diffraction images 213 00:07:38,760 --> 00:07:41,189 have been carried out for many decades. 214 00:07:41,190 --> 00:07:43,259 Just to give you one example, the 215 00:07:44,700 --> 00:07:46,889 discovery of the DNA structure was only 216 00:07:46,890 --> 00:07:49,319 possible because Rosalind Franklin 217 00:07:49,320 --> 00:07:51,539 made these diffraction images of a DNA 218 00:07:51,540 --> 00:07:53,979 crystal and just a side story. 219 00:07:53,980 --> 00:07:56,219 Guess who got the Nobel Prize for this? 220 00:07:56,220 --> 00:07:58,229 Of course, the two white men. 221 00:07:58,230 --> 00:08:00,359 But this is another nasty story 222 00:08:00,360 --> 00:08:03,089 I recommend you to look up afterwards. 223 00:08:03,090 --> 00:08:05,549 The thing is about these X-ray tubes. 224 00:08:05,550 --> 00:08:07,529 They are very limited in brightness and 225 00:08:07,530 --> 00:08:08,909 this becomes a problem if you want to 226 00:08:08,910 --> 00:08:10,689 study, which moves. 227 00:08:10,690 --> 00:08:11,909 You all know if you want to make a 228 00:08:11,910 --> 00:08:13,559 picture from something which moves, you 229 00:08:13,560 --> 00:08:15,959 have to decrease the shutter speed. 230 00:08:15,960 --> 00:08:17,849 So for running whores, for example, it's 231 00:08:17,850 --> 00:08:19,529 sufficient to have a shutter speed of one 232 00:08:19,530 --> 00:08:21,449 millisecond. But if you want to watch a 233 00:08:21,450 --> 00:08:23,999 bullet smashing a melon, you have to use 234 00:08:24,000 --> 00:08:26,369 like 1000 frames per second more. 235 00:08:26,370 --> 00:08:27,719 And finally, if you want to go to 236 00:08:27,720 --> 00:08:29,789 chemical reactions, the shutter speed is 237 00:08:29,790 --> 00:08:30,790 orders beyond that. 238 00:08:32,039 --> 00:08:33,989 And you may have seen how such a movie is 239 00:08:33,990 --> 00:08:35,759 made. So you need big lamps in order to 240 00:08:35,760 --> 00:08:38,579 get enough light hitting your object 241 00:08:38,580 --> 00:08:40,109 in the very short amount of time where 242 00:08:40,110 --> 00:08:41,339 the shutters open. 243 00:08:41,340 --> 00:08:43,439 Right. So the figure of merit for 244 00:08:43,440 --> 00:08:45,719 a normal lamp is the luminous and 245 00:08:45,720 --> 00:08:48,059 luminous intensity which 246 00:08:48,060 --> 00:08:50,279 is defined as the photons per time 247 00:08:50,280 --> 00:08:51,269 for a solid angle. 248 00:08:51,270 --> 00:08:53,459 So basically the amount of light 249 00:08:53,460 --> 00:08:55,979 which are directed to your target. 250 00:08:55,980 --> 00:08:57,629 But we want to make X-ray diffraction 251 00:08:57,630 --> 00:08:59,219 images, so we need coherent light. 252 00:08:59,220 --> 00:09:00,899 And our figure of merit looks a bit 253 00:09:00,900 --> 00:09:02,999 different, as it is called the brilliance 254 00:09:03,000 --> 00:09:04,709 of a light source. And what we want is 255 00:09:04,710 --> 00:09:07,079 basically we want a lot of photons 256 00:09:07,080 --> 00:09:07,719 per time. 257 00:09:07,720 --> 00:09:10,019 We want them emitted in a small spot size 258 00:09:10,020 --> 00:09:12,089 with a small, angular divergence and 259 00:09:12,090 --> 00:09:14,169 basically only one wavelength. 260 00:09:14,170 --> 00:09:15,959 OK, so this is the brilliance. 261 00:09:15,960 --> 00:09:17,939 This is our key. 262 00:09:17,940 --> 00:09:19,709 But before I want to show you what it 263 00:09:19,710 --> 00:09:21,389 takes for the brilliance to get from here 264 00:09:21,390 --> 00:09:23,549 to here, I want to sensitize you a bit 265 00:09:23,550 --> 00:09:25,079 more about the scale of what we are 266 00:09:25,080 --> 00:09:26,369 talking. 267 00:09:26,370 --> 00:09:28,679 So this is an example of some objects 268 00:09:28,680 --> 00:09:30,749 which I sorted according 269 00:09:30,750 --> 00:09:32,639 to their length on a logarithmic scale. 270 00:09:32,640 --> 00:09:34,439 So we start with a fingertip of some 271 00:09:34,440 --> 00:09:36,809 centimeters over a human hair 272 00:09:36,810 --> 00:09:38,939 down to molecules and atoms, 273 00:09:38,940 --> 00:09:41,139 and we are able to produce plenty 274 00:09:41,140 --> 00:09:42,749 of technology basically on the whole 275 00:09:42,750 --> 00:09:45,429 scale so we can produce a microglia 276 00:09:45,430 --> 00:09:47,729 with the diameter of some micrometer 277 00:09:47,730 --> 00:09:48,899 and even nanotubes. 278 00:09:48,900 --> 00:09:51,059 And although this is rather something 279 00:09:51,060 --> 00:09:53,069 academic yet, but in principle we are 280 00:09:53,070 --> 00:09:54,839 able to arrange metter on an atomic 281 00:09:54,840 --> 00:09:55,840 scale. 282 00:09:56,760 --> 00:09:59,129 The corresponding plot in time domain 283 00:09:59,130 --> 00:10:00,369 could look something like this. 284 00:10:00,370 --> 00:10:02,369 So we start with enabling of a couple of 285 00:10:02,370 --> 00:10:05,129 hundred microseconds down over 286 00:10:05,130 --> 00:10:06,689 the time. It takes a shockwave to 287 00:10:06,690 --> 00:10:09,359 propagate by one atom in a crystal 288 00:10:09,360 --> 00:10:11,459 finally to the chemical reactions or 289 00:10:11,460 --> 00:10:12,929 the ball period. 290 00:10:12,930 --> 00:10:15,149 And it takes a one gigahertz CPU about 291 00:10:15,150 --> 00:10:17,669 one nanosecond to make one computational 292 00:10:17,670 --> 00:10:19,899 step and optical networks, which 293 00:10:19,900 --> 00:10:20,999 is even a bit faster. 294 00:10:21,000 --> 00:10:23,189 But basically we are not really 295 00:10:23,190 --> 00:10:25,139 able to produce technology and that 296 00:10:25,140 --> 00:10:26,249 timescale. 297 00:10:26,250 --> 00:10:28,469 I mean, we are able to produce 298 00:10:28,470 --> 00:10:30,749 a laser pulse in the visible light, which 299 00:10:30,750 --> 00:10:33,119 is as short as one femtosecond, which is 300 00:10:33,120 --> 00:10:34,319 really amazing. 301 00:10:34,320 --> 00:10:36,329 But keep in mind the diffraction limit. 302 00:10:36,330 --> 00:10:38,309 So with this, we can watch macroscopic 303 00:10:38,310 --> 00:10:40,109 objects like, for example, the microglia 304 00:10:40,110 --> 00:10:42,329 and we can watch the microglia within one 305 00:10:42,330 --> 00:10:44,339 to femtosecond and see how it changes. 306 00:10:44,340 --> 00:10:46,559 But macroscopic objects don't change 307 00:10:46,560 --> 00:10:48,119 with an femtoseconds. 308 00:10:48,120 --> 00:10:50,009 Things which changed in femtoseconds are 309 00:10:50,010 --> 00:10:51,929 proteins or molecules, and we are 310 00:10:51,930 --> 00:10:54,209 literally blind at these objects within 311 00:10:54,210 --> 00:10:56,669 their natural timescale. 312 00:10:56,670 --> 00:10:58,619 And to give you a better feeling of the 313 00:10:58,620 --> 00:11:00,779 scaling here, a fingertip is 314 00:11:00,780 --> 00:11:03,029 to an atom is about 315 00:11:03,030 --> 00:11:05,189 two times, two times ten to the 316 00:11:05,190 --> 00:11:07,589 eight times bigger 317 00:11:07,590 --> 00:11:08,519 than an atom. 318 00:11:08,520 --> 00:11:10,709 OK, and that's about the scaling of 319 00:11:10,710 --> 00:11:12,269 the walking distance from here to Tel 320 00:11:12,270 --> 00:11:14,369 Aviv to a fingertip 321 00:11:14,370 --> 00:11:16,559 and in time domain and eyeblink is 322 00:11:16,560 --> 00:11:18,659 to a chemical reaction like one year is 323 00:11:18,660 --> 00:11:20,099 to enabling. 324 00:11:20,100 --> 00:11:21,899 And now keep in mind, when you go to a 325 00:11:21,900 --> 00:11:23,729 hospital and you want to make an X-ray 326 00:11:23,730 --> 00:11:25,819 image with a modern X-ray tube from 327 00:11:25,820 --> 00:11:28,409 the finger, you have to stand still for, 328 00:11:28,410 --> 00:11:29,519 let's say, a second. 329 00:11:29,520 --> 00:11:30,419 Right. 330 00:11:30,420 --> 00:11:32,669 And if you want to scale that to an atom 331 00:11:32,670 --> 00:11:34,889 and to such a timescale, 332 00:11:34,890 --> 00:11:37,019 you immediately see that X-ray tubes are 333 00:11:37,020 --> 00:11:39,569 nowhere near what is needed 334 00:11:39,570 --> 00:11:41,729 for resolving proteins on their 335 00:11:41,730 --> 00:11:43,589 natural timescale. 336 00:11:43,590 --> 00:11:46,289 And I want to relate the 337 00:11:46,290 --> 00:11:48,509 the development of our brilliance with 338 00:11:48,510 --> 00:11:50,849 something new. So, you know, so. 339 00:11:50,850 --> 00:11:52,799 This is the computer speed, and you all 340 00:11:52,800 --> 00:11:54,989 know Moore's Law and you have kind 341 00:11:54,990 --> 00:11:56,879 of a feeling what it means as a figure of 342 00:11:56,880 --> 00:11:59,219 merit increases in 12 orders of magnitude 343 00:11:59,220 --> 00:12:01,529 and 60 kids ride 344 00:12:01,530 --> 00:12:03,659 the extra brilliance increased by 18 345 00:12:03,660 --> 00:12:06,099 orders of magnitude in five decades. 346 00:12:06,100 --> 00:12:07,769 And this was not possible by small 347 00:12:07,770 --> 00:12:10,199 innovations, but very 348 00:12:10,200 --> 00:12:11,439 different steps. 349 00:12:11,440 --> 00:12:13,049 So we have different generations of 350 00:12:13,050 --> 00:12:15,269 synchrotron light sources. 351 00:12:15,270 --> 00:12:17,669 And finally, the fourth generation 352 00:12:17,670 --> 00:12:19,979 we call X-ray free electron lasers. 353 00:12:19,980 --> 00:12:21,839 And in this talk, I will go through the 354 00:12:21,840 --> 00:12:23,279 steps, what it takes to build these 355 00:12:23,280 --> 00:12:24,280 machines. 356 00:12:25,260 --> 00:12:27,449 But before I can tell you how 357 00:12:27,450 --> 00:12:29,549 we can build such a particle accelerator, 358 00:12:29,550 --> 00:12:31,259 I have to tell you why these particles 359 00:12:31,260 --> 00:12:32,159 actually radiate. 360 00:12:32,160 --> 00:12:34,229 And for doing that, I have 361 00:12:34,230 --> 00:12:36,569 to tell you something about relativity. 362 00:12:36,570 --> 00:12:38,699 Maybe you have been on his talk 363 00:12:38,700 --> 00:12:39,689 yesterday. 364 00:12:39,690 --> 00:12:41,669 I will try to summarize it on one side. 365 00:12:43,980 --> 00:12:46,139 So we call 366 00:12:46,140 --> 00:12:48,299 our machines particle accelerators. 367 00:12:48,300 --> 00:12:50,639 Right. But I guess your intuitive 368 00:12:50,640 --> 00:12:52,709 understanding of acceleration is an 369 00:12:52,710 --> 00:12:54,989 increase of velocity, but that's not 370 00:12:54,990 --> 00:12:57,119 really the case, but a 371 00:12:57,120 --> 00:12:59,219 step by step. So maybe you are familiar 372 00:12:59,220 --> 00:13:00,989 with Newton's law of kinematics telling 373 00:13:00,990 --> 00:13:02,789 you that the kinetic energy is one over 374 00:13:02,790 --> 00:13:04,739 two times the mass of a particle times 375 00:13:04,740 --> 00:13:06,599 the velocity squared. 376 00:13:06,600 --> 00:13:08,639 But as Einstein revealed, the speed of 377 00:13:08,640 --> 00:13:10,169 light is a constant which can't be 378 00:13:10,170 --> 00:13:12,089 exceeded by any particle with a finite 379 00:13:12,090 --> 00:13:12,989 mass. 380 00:13:12,990 --> 00:13:14,519 So it turned out that Newton's laws of 381 00:13:14,520 --> 00:13:16,469 mechanics is only a borderline case for 382 00:13:16,470 --> 00:13:17,610 very low velocities, 383 00:13:18,750 --> 00:13:20,819 for Einstein's more general equation of 384 00:13:20,820 --> 00:13:21,899 kinematics. 385 00:13:21,900 --> 00:13:23,789 And in here you have this relativistic 386 00:13:23,790 --> 00:13:25,619 factor. Gamma Gamma is one over the 387 00:13:25,620 --> 00:13:27,839 script, over this square root, and 388 00:13:27,840 --> 00:13:29,609 it basically relates to the energy of a 389 00:13:29,610 --> 00:13:31,619 particle with the thrust mass. 390 00:13:31,620 --> 00:13:33,749 And it's quite an important parameter 391 00:13:33,750 --> 00:13:35,369 for us and it will pop up a couple of 392 00:13:35,370 --> 00:13:36,279 times in this talk. 393 00:13:36,280 --> 00:13:38,309 So let me give you an example. 394 00:13:38,310 --> 00:13:39,869 Let's assume that we accelerate an 395 00:13:39,870 --> 00:13:41,909 electron and the proton with five million 396 00:13:41,910 --> 00:13:44,039 volts, OK, so five megawatts. 397 00:13:44,040 --> 00:13:45,929 And then the kinetic energy of both 398 00:13:45,930 --> 00:13:48,389 particles is five mega electron 399 00:13:48,390 --> 00:13:50,759 balls and the rest mass 400 00:13:50,760 --> 00:13:53,519 for an electron is about 500 KVI 401 00:13:53,520 --> 00:13:55,859 kilo electron balls, while it is about 20 402 00:13:55,860 --> 00:13:57,479 times more for proton. 403 00:13:57,480 --> 00:13:59,039 And this means you can now plug the 404 00:13:59,040 --> 00:14:01,169 numbers in that the gamma factor is about 405 00:14:01,170 --> 00:14:03,359 10 four electrons while it is about one 406 00:14:03,360 --> 00:14:05,309 four protons. 407 00:14:05,310 --> 00:14:07,079 And if you calculate the speed now from 408 00:14:07,080 --> 00:14:09,389 this, you will see that 409 00:14:09,390 --> 00:14:11,429 electrons, while being accelerated with 410 00:14:11,430 --> 00:14:13,499 five million volts, travel with ninety 411 00:14:13,500 --> 00:14:15,179 nine point five percent of the speed of 412 00:14:15,180 --> 00:14:17,219 light, while protons only travel 10 413 00:14:17,220 --> 00:14:19,289 percent or so electrons and 414 00:14:19,290 --> 00:14:21,359 protons, or in general, light and 415 00:14:21,360 --> 00:14:23,279 heavy particles have a very different 416 00:14:23,280 --> 00:14:24,719 relation between the energy and the 417 00:14:24,720 --> 00:14:26,099 velocity. 418 00:14:26,100 --> 00:14:27,929 And in our case, as in synchrotron light 419 00:14:27,930 --> 00:14:29,789 sources, we are always interested in a 420 00:14:29,790 --> 00:14:30,809 very high gamma. 421 00:14:30,810 --> 00:14:32,669 So it's obvious that we are only using 422 00:14:32,670 --> 00:14:34,079 electrons. 423 00:14:34,080 --> 00:14:36,359 OK, so the next 424 00:14:36,360 --> 00:14:38,429 step is why do they radiate? 425 00:14:38,430 --> 00:14:39,959 This is an electron and here I plotted 426 00:14:39,960 --> 00:14:42,179 the electric field lines and you may be 427 00:14:42,180 --> 00:14:44,159 familiar with the relativistic effect 428 00:14:44,160 --> 00:14:45,959 called length contraction or lorence 429 00:14:45,960 --> 00:14:46,949 contraction. 430 00:14:46,950 --> 00:14:49,019 A very neat example is a ruler which 431 00:14:49,020 --> 00:14:51,089 travels close to the speed of light 432 00:14:51,090 --> 00:14:53,249 and it gets compressed with respect to 433 00:14:53,250 --> 00:14:54,779 an observer and rest. 434 00:14:54,780 --> 00:14:57,659 And if you apply this length contraction 435 00:14:57,660 --> 00:14:59,909 to the electric field lines, we will see 436 00:14:59,910 --> 00:15:02,069 that where the speed of 437 00:15:02,070 --> 00:15:04,079 the particle is increased, the electric 438 00:15:04,080 --> 00:15:05,819 field lines are compressed into a very 439 00:15:05,820 --> 00:15:07,949 narrow cone perpendicular to the speed of 440 00:15:07,950 --> 00:15:09,429 the particle. 441 00:15:09,430 --> 00:15:11,699 OK, and now consider 442 00:15:11,700 --> 00:15:13,799 you want to change the velocity from here 443 00:15:13,800 --> 00:15:16,109 to here so you accelerate the particle 444 00:15:16,110 --> 00:15:17,819 and the electric field configuration has 445 00:15:17,820 --> 00:15:19,959 to change from that set up to this one. 446 00:15:19,960 --> 00:15:22,259 But this can not happen infinitely fast, 447 00:15:22,260 --> 00:15:24,449 but only with the speed of light. 448 00:15:24,450 --> 00:15:26,669 So you have a time varying electric field 449 00:15:26,670 --> 00:15:29,369 and this is basically radiation 450 00:15:29,370 --> 00:15:31,529 maybe come a bit more clear 451 00:15:31,530 --> 00:15:32,530 on this slide. 452 00:15:33,540 --> 00:15:35,309 I make this simulation, you can download 453 00:15:35,310 --> 00:15:38,309 the simulator from the center. 454 00:15:38,310 --> 00:15:40,019 This is, again, a point charge and I drag 455 00:15:40,020 --> 00:15:42,329 it now with the mouse and 456 00:15:42,330 --> 00:15:43,679 increase its velocity. 457 00:15:43,680 --> 00:15:45,089 And you can see as I increase the 458 00:15:45,090 --> 00:15:47,159 velocity, the field lines are compressed 459 00:15:47,160 --> 00:15:48,840 into this very narrow cone 460 00:15:50,100 --> 00:15:51,659 and the radiation pattern gets more 461 00:15:51,660 --> 00:15:53,339 obvious. If I change the direction of 462 00:15:53,340 --> 00:15:55,079 motion, for example, by running it on a 463 00:15:55,080 --> 00:15:56,080 circle. 464 00:15:56,700 --> 00:15:59,129 And if you think you sit here 465 00:15:59,130 --> 00:16:01,079 and watch the electron, you will get hit 466 00:16:01,080 --> 00:16:03,299 by narrow flashes of electromagnetic 467 00:16:03,300 --> 00:16:04,709 radiation. 468 00:16:04,710 --> 00:16:06,749 Right. And this is basically a 469 00:16:06,750 --> 00:16:07,859 synchrotron light source. 470 00:16:09,300 --> 00:16:11,309 But I would like to look a bit more 471 00:16:11,310 --> 00:16:13,529 detail on the 472 00:16:13,530 --> 00:16:14,969 radiation properties. 473 00:16:14,970 --> 00:16:17,609 So here again, this is our electron. 474 00:16:17,610 --> 00:16:19,709 And I calculated the radiation 475 00:16:19,710 --> 00:16:21,959 pattern for this motion and I plotted 476 00:16:21,960 --> 00:16:23,909 the angle of distribution here with this 477 00:16:23,910 --> 00:16:25,019 surface plot. 478 00:16:25,020 --> 00:16:27,749 So you see that most of the radiation 479 00:16:27,750 --> 00:16:30,419 is directed in the forward direction. 480 00:16:30,420 --> 00:16:32,489 Yeah, and the opening angle here 481 00:16:32,490 --> 00:16:34,559 of this radiation cone scales with 482 00:16:34,560 --> 00:16:36,959 one over Gamma and the overall power 483 00:16:36,960 --> 00:16:39,119 which is emitted scales with gamma to 484 00:16:39,120 --> 00:16:41,609 the four is directly proportional 485 00:16:41,610 --> 00:16:43,799 to the energy. So if we have a very high 486 00:16:43,800 --> 00:16:46,469 energy, basically 487 00:16:46,470 --> 00:16:48,689 all the radiation is emitted into 488 00:16:48,690 --> 00:16:50,809 a very narrow cone and forward direction. 489 00:16:50,810 --> 00:16:52,549 And in our cases, Gamma is something like 490 00:16:52,550 --> 00:16:54,619 10000, so it's really 491 00:16:54,620 --> 00:16:56,809 small and a nice property of 492 00:16:56,810 --> 00:16:58,429 this radiation is that it covers a 493 00:16:58,430 --> 00:17:00,319 relatively wide range and frequency 494 00:17:00,320 --> 00:17:02,359 domain and you can easily tune it by 495 00:17:02,360 --> 00:17:04,338 changing the gamma or the energy of your 496 00:17:04,339 --> 00:17:05,419 particle. 497 00:17:05,420 --> 00:17:07,189 And this kind of radiation was first 498 00:17:07,190 --> 00:17:09,139 observed in a particle accelerator called 499 00:17:09,140 --> 00:17:11,118 synchrotron, and that's why we call this 500 00:17:11,119 --> 00:17:12,319 synchrotron radiation. 501 00:17:13,970 --> 00:17:15,499 So coming back to that picture, 502 00:17:15,500 --> 00:17:17,838 synchrotron radiation is very suited 503 00:17:17,839 --> 00:17:20,209 to study small things like proteins 504 00:17:20,210 --> 00:17:21,348 or molecules. 505 00:17:21,349 --> 00:17:23,088 Now, the question is, how can we put this 506 00:17:23,089 --> 00:17:25,009 into technology? So how can we make use 507 00:17:25,010 --> 00:17:26,779 of it? And, of course, you know, it's 508 00:17:26,780 --> 00:17:28,609 particle accelerators. 509 00:17:28,610 --> 00:17:30,559 So what are the principles of a light 510 00:17:30,560 --> 00:17:32,539 source? Well, first of all, we have to 511 00:17:32,540 --> 00:17:34,009 generate our electrons. 512 00:17:34,010 --> 00:17:36,139 So we need a device which 513 00:17:36,140 --> 00:17:37,729 serves as an electron source. 514 00:17:37,730 --> 00:17:39,529 Then we need something to increase the 515 00:17:39,530 --> 00:17:41,569 energy. And finally, we need a device to 516 00:17:41,570 --> 00:17:42,829 make them radiate. 517 00:17:42,830 --> 00:17:44,419 And with this radiation, we can then 518 00:17:44,420 --> 00:17:46,549 perform our X-ray experiments. 519 00:17:46,550 --> 00:17:48,589 It's as simple as that. 520 00:17:48,590 --> 00:17:51,079 And it's not too ambitious 521 00:17:51,080 --> 00:17:53,569 analogy to think of such a light source 522 00:17:53,570 --> 00:17:54,949 as a radio station. 523 00:17:54,950 --> 00:17:56,689 There also you have your input signal, 524 00:17:56,690 --> 00:17:58,939 then you have a high power amplification, 525 00:17:58,940 --> 00:18:00,799 and then you put this high power signal 526 00:18:00,800 --> 00:18:02,539 through a device which is designed to 527 00:18:02,540 --> 00:18:04,639 produce electromagnetic radiation, 528 00:18:04,640 --> 00:18:06,559 of which only a tiny fraction hits your 529 00:18:06,560 --> 00:18:07,560 receiver. 530 00:18:08,510 --> 00:18:09,979 OK, so in the following, I want to go 531 00:18:09,980 --> 00:18:12,529 through these different devices, 532 00:18:12,530 --> 00:18:14,899 starting with the acceleration. 533 00:18:14,900 --> 00:18:16,759 You may know that if I connect a 534 00:18:16,760 --> 00:18:19,159 capacitor with a DC voltage source, 535 00:18:19,160 --> 00:18:20,989 I will get an electric field between the 536 00:18:20,990 --> 00:18:23,059 plates and if I place 537 00:18:23,060 --> 00:18:25,219 a negatively charged electron in here, 538 00:18:25,220 --> 00:18:26,889 it will get accelerated. 539 00:18:26,890 --> 00:18:27,859 Right. 540 00:18:27,860 --> 00:18:29,599 And we have these kind of accelerators, 541 00:18:29,600 --> 00:18:31,669 we call them fundido of accelerators 542 00:18:31,670 --> 00:18:33,769 and modern ones like this one is 543 00:18:33,770 --> 00:18:36,349 up to 10 meters long and reach 544 00:18:36,350 --> 00:18:38,059 can accelerate the particles by six 545 00:18:38,060 --> 00:18:40,129 million volts, which is not 546 00:18:40,130 --> 00:18:41,749 bad. But the problem is we can't really 547 00:18:41,750 --> 00:18:44,119 put them in serious and we can't increase 548 00:18:44,120 --> 00:18:45,889 the voltage either because we would 549 00:18:45,890 --> 00:18:47,539 simply get a discharge between the two 550 00:18:47,540 --> 00:18:49,369 plates. So the problem with this 551 00:18:49,370 --> 00:18:51,889 technology is it doesn't scale. 552 00:18:51,890 --> 00:18:53,959 So what we do is we replace 553 00:18:53,960 --> 00:18:55,909 our capacitor with an empty metallic 554 00:18:55,910 --> 00:18:58,249 resonator called cavity, and we connect 555 00:18:58,250 --> 00:19:00,349 this cavity with a waveguide to an 556 00:19:00,350 --> 00:19:01,849 AC voltage source. 557 00:19:01,850 --> 00:19:04,009 And this water source is operated usually 558 00:19:04,010 --> 00:19:05,449 in the radio frequency domain. 559 00:19:05,450 --> 00:19:07,189 So some gigahertz, that's why we call 560 00:19:07,190 --> 00:19:08,419 this RF. 561 00:19:08,420 --> 00:19:10,189 And the nice thing about such a resonator 562 00:19:10,190 --> 00:19:12,589 is that a relatively small r a field 563 00:19:12,590 --> 00:19:14,899 will start to resonate 564 00:19:14,900 --> 00:19:17,149 inside. So we will get a relatively 565 00:19:17,150 --> 00:19:19,309 high oscillating electric field and 566 00:19:19,310 --> 00:19:21,589 we can easily put these in serious. 567 00:19:21,590 --> 00:19:23,659 And if we put the face or set up the 568 00:19:23,660 --> 00:19:25,639 face relation between adjacent cells 569 00:19:25,640 --> 00:19:27,739 correctly, we will get an alternating, 570 00:19:27,740 --> 00:19:30,109 oscillating electric field. 571 00:19:30,110 --> 00:19:31,789 And the real cool thing is now that we 572 00:19:31,790 --> 00:19:34,009 can put holes in here without really 573 00:19:34,010 --> 00:19:36,139 changing the geometry and now the 574 00:19:36,140 --> 00:19:38,269 cells are coupled so we can remove 575 00:19:38,270 --> 00:19:40,409 all the power sources except of one. 576 00:19:40,410 --> 00:19:42,439 And if you put the beam pipe in here, an 577 00:19:42,440 --> 00:19:44,479 electron there, and if we synchronize 578 00:19:44,480 --> 00:19:46,129 everything correctly, you will see that 579 00:19:46,130 --> 00:19:48,289 we get an acceleration in each cell 580 00:19:48,290 --> 00:19:49,969 of the cavity. 581 00:19:49,970 --> 00:19:52,159 Right. And of course, I mean, the devil's 582 00:19:52,160 --> 00:19:53,839 in the details, but this is the basic 583 00:19:53,840 --> 00:19:55,549 principle of an RF cavity. 584 00:19:55,550 --> 00:19:56,550 And 585 00:19:58,400 --> 00:19:59,959 there was no gel content that. 586 00:20:07,570 --> 00:20:09,939 And basically, every particle accelerator 587 00:20:09,940 --> 00:20:12,009 on earth is operated of these kind of 588 00:20:12,010 --> 00:20:14,139 devices, just to give you 589 00:20:14,140 --> 00:20:16,299 one example, this is a Tesla cavity 590 00:20:16,300 --> 00:20:18,639 we have in our linear accelerator that 591 00:20:18,640 --> 00:20:20,829 here we have these nine cells and 592 00:20:20,830 --> 00:20:22,389 it's a superconducting technology. 593 00:20:22,390 --> 00:20:24,489 So everything has to be assembled in the 594 00:20:24,490 --> 00:20:26,409 clean room, which is very challenging. 595 00:20:26,410 --> 00:20:28,419 And then we put eight of them into one of 596 00:20:28,420 --> 00:20:30,489 these cryo vessels with a lot of support. 597 00:20:30,490 --> 00:20:32,409 And then we plug it in these yellow 598 00:20:32,410 --> 00:20:34,539 things here and put it down in 599 00:20:34,540 --> 00:20:36,399 the tunnel. Then we cool it down with 600 00:20:36,400 --> 00:20:38,469 liquid helium to two Kelvin. 601 00:20:38,470 --> 00:20:40,899 And in these cavities, we can reach 602 00:20:40,900 --> 00:20:43,299 something like 30 million volts within 603 00:20:43,300 --> 00:20:44,379 one meter. 604 00:20:44,380 --> 00:20:46,659 So this is 50 times more 605 00:20:46,660 --> 00:20:47,979 what you can get in a Vandergriff 606 00:20:47,980 --> 00:20:50,319 accelerator. And I mean, think of this 30 607 00:20:50,320 --> 00:20:52,029 million walls between these two hands. 608 00:20:53,110 --> 00:20:54,519 To me, this is really an impressive 609 00:20:54,520 --> 00:20:55,520 technology. 610 00:20:57,220 --> 00:20:58,220 Seriously. 611 00:21:01,170 --> 00:21:03,299 OK, so the next step is 612 00:21:03,300 --> 00:21:05,189 the electron source. 613 00:21:05,190 --> 00:21:06,719 This is a movie made from the photo 614 00:21:06,720 --> 00:21:08,609 injector test sent home in Sweden. 615 00:21:08,610 --> 00:21:11,549 But the electron source, we have a DC 616 00:21:11,550 --> 00:21:13,139 look basically the same. 617 00:21:13,140 --> 00:21:15,119 So you see it's a very complicated device 618 00:21:15,120 --> 00:21:16,679 and there are whole laboratories only 619 00:21:16,680 --> 00:21:18,899 building these electron sources. 620 00:21:18,900 --> 00:21:21,299 But this movie shows basic principles. 621 00:21:21,300 --> 00:21:23,729 On the inside, you have a copper cavity, 622 00:21:23,730 --> 00:21:25,409 which with its connection to the 623 00:21:25,410 --> 00:21:27,150 waveguide and from the inside 624 00:21:29,940 --> 00:21:32,009 you have a photo Cassard sitting here 625 00:21:32,010 --> 00:21:33,959 and this photo Cassard is impinged by a 626 00:21:33,960 --> 00:21:36,119 UV laser pulse and why the UV laser 627 00:21:36,120 --> 00:21:38,519 hit this photo that there are 628 00:21:38,520 --> 00:21:40,619 electrons emitted due to the photo 629 00:21:40,620 --> 00:21:41,099 effect. 630 00:21:41,100 --> 00:21:42,930 So each of these red 631 00:21:43,950 --> 00:21:46,229 things are about one billion or 10 632 00:21:46,230 --> 00:21:47,909 billion electrons and we call this an 633 00:21:47,910 --> 00:21:48,989 electron bunch. 634 00:21:48,990 --> 00:21:51,239 And then again we have two cells of RF 635 00:21:51,240 --> 00:21:53,099 cavity and everything is synchronized in 636 00:21:53,100 --> 00:21:55,919 a way that accelerates the electrons 637 00:21:55,920 --> 00:21:58,530 immediately as they are generated. 638 00:22:02,240 --> 00:22:04,309 OK, finally, we need 639 00:22:04,310 --> 00:22:05,959 a device to make them radiate, and I 640 00:22:05,960 --> 00:22:08,089 already told you, we just have to 641 00:22:08,090 --> 00:22:10,089 bend them around the circle, right. 642 00:22:10,090 --> 00:22:11,929 And we can do this most easily. 643 00:22:11,930 --> 00:22:13,939 And dipole magnets, you may know from 644 00:22:13,940 --> 00:22:15,829 high school physics or whatever the left 645 00:22:15,830 --> 00:22:17,419 hand rule, if we have an electron with 646 00:22:17,420 --> 00:22:19,579 speed, we and the magnetic 647 00:22:19,580 --> 00:22:21,289 field perpendicular to it will get the 648 00:22:21,290 --> 00:22:23,509 Lawrence force in the third direction. 649 00:22:23,510 --> 00:22:25,699 So the whole thing is 650 00:22:25,700 --> 00:22:28,789 bent around a circle. 651 00:22:28,790 --> 00:22:30,529 So now we have to have everything 652 00:22:30,530 --> 00:22:32,749 together to build our storage ring 653 00:22:32,750 --> 00:22:33,679 electron source. 654 00:22:33,680 --> 00:22:35,869 We need an RF cavity and then a big type 655 00:22:35,870 --> 00:22:38,029 magnet so the particle will move 656 00:22:38,030 --> 00:22:40,099 on a circle and continuously emit 657 00:22:40,100 --> 00:22:41,929 synchrotron radiation. 658 00:22:41,930 --> 00:22:43,789 But it's not that easy because we have 659 00:22:43,790 --> 00:22:45,169 energy conservation and while the 660 00:22:45,170 --> 00:22:47,149 particle emits power, it will lose 661 00:22:47,150 --> 00:22:49,489 kinetic energy. So it will finally spiral 662 00:22:49,490 --> 00:22:51,079 in and get lost. 663 00:22:51,080 --> 00:22:53,389 So we have to replace that and 664 00:22:53,390 --> 00:22:55,999 in insert straight sections 665 00:22:56,000 --> 00:22:57,799 where we can place an RF cavity to 666 00:22:57,800 --> 00:22:59,629 compensate for the power loss in the 667 00:22:59,630 --> 00:23:00,919 dipole magnets. 668 00:23:00,920 --> 00:23:03,109 And then we have to put some focusing 669 00:23:03,110 --> 00:23:04,249 elements in here. 670 00:23:04,250 --> 00:23:06,649 We use quadripolar magnets 671 00:23:06,650 --> 00:23:09,109 to stabilize the system. 672 00:23:09,110 --> 00:23:11,179 And this particle accelerator is 673 00:23:11,180 --> 00:23:13,729 called a synchrotron. 674 00:23:13,730 --> 00:23:15,769 And originally, these kind of devices 675 00:23:15,770 --> 00:23:17,329 were built for high energy physics 676 00:23:17,330 --> 00:23:19,609 applications like, for example, 677 00:23:19,610 --> 00:23:21,709 LHC, the Large Hadron Collider 678 00:23:21,710 --> 00:23:22,789 and CERN is nothing more. 679 00:23:22,790 --> 00:23:24,859 But this, of course, but the basic 680 00:23:24,860 --> 00:23:27,559 principle is the synchrotron. 681 00:23:27,560 --> 00:23:29,899 And this could be your Atlas detector. 682 00:23:29,900 --> 00:23:32,089 Okay. And then the, uh, in 683 00:23:32,090 --> 00:23:34,339 the early 50s, when they started to build 684 00:23:34,340 --> 00:23:36,019 these kind of accelerators, the 685 00:23:36,020 --> 00:23:37,609 synchrotron radiation was found to be 686 00:23:37,610 --> 00:23:39,619 nothing more but a nuisance, which made 687 00:23:39,620 --> 00:23:41,659 everything more complicated. 688 00:23:41,660 --> 00:23:44,119 But in the 60s, uh, 689 00:23:44,120 --> 00:23:45,919 X-ray diffraction became a thing and 690 00:23:45,920 --> 00:23:48,019 scientists started to realize about 691 00:23:48,020 --> 00:23:49,849 the capabilities of this radiation. 692 00:23:49,850 --> 00:23:52,219 So they placed some X-ray optics 693 00:23:52,220 --> 00:23:54,349 in here which guided the synchrotron 694 00:23:54,350 --> 00:23:56,119 radiation to the experiments. 695 00:23:56,120 --> 00:23:58,249 And these kind of devices are considered 696 00:23:58,250 --> 00:24:00,199 as the first generation synchrotron light 697 00:24:00,200 --> 00:24:01,489 sources. 698 00:24:01,490 --> 00:24:03,229 And as an example, this is the Tantalus 699 00:24:03,230 --> 00:24:05,809 one accelerator in the late 60s. 700 00:24:05,810 --> 00:24:07,129 Here is the accelerator. 701 00:24:07,130 --> 00:24:08,989 So this is the RF cavity and there are 702 00:24:08,990 --> 00:24:10,399 some dipole magnets. 703 00:24:10,400 --> 00:24:13,069 You see, it's a fairly small device. 704 00:24:13,070 --> 00:24:15,199 And very soon scientists started to want 705 00:24:15,200 --> 00:24:17,419 to have more power in their radiation. 706 00:24:17,420 --> 00:24:19,849 So in a bending magnet, each electron 707 00:24:19,850 --> 00:24:21,949 radiates. So the intensity or the 708 00:24:21,950 --> 00:24:23,749 brilliance scales with the number of 709 00:24:23,750 --> 00:24:25,459 electrons, right? 710 00:24:25,460 --> 00:24:28,069 Double the electrons, double the power. 711 00:24:28,070 --> 00:24:29,479 And starting from that, if you want to 712 00:24:29,480 --> 00:24:31,099 increase the power, the first obvious 713 00:24:31,100 --> 00:24:33,319 step is to put more dipole magnets in. 714 00:24:33,320 --> 00:24:35,509 So this is an insurgent 715 00:24:35,510 --> 00:24:37,249 device called Wikler, and it's basically 716 00:24:37,250 --> 00:24:39,499 nothing else but a series of dipo magnets 717 00:24:39,500 --> 00:24:41,329 with alternating polarity. 718 00:24:41,330 --> 00:24:43,189 So the electrons will move on a slalom 719 00:24:43,190 --> 00:24:45,439 trajectory. And in each curve you 720 00:24:45,440 --> 00:24:47,659 will get synchrotron radiation as from 721 00:24:47,660 --> 00:24:49,519 a single dipole magnet. 722 00:24:49,520 --> 00:24:51,379 And by doing that, you will increase the 723 00:24:51,380 --> 00:24:53,269 brilliance by a factor of the number of 724 00:24:53,270 --> 00:24:54,609 magnets. 725 00:24:54,610 --> 00:24:57,139 Right. So nothing more but that 726 00:24:57,140 --> 00:24:59,299 then the next generation or 727 00:24:59,300 --> 00:25:01,459 the next step towards 728 00:25:01,460 --> 00:25:03,589 brighter synchrotron light source, as 729 00:25:03,590 --> 00:25:05,239 was the invention of Ananda later. 730 00:25:05,240 --> 00:25:07,309 And Undulated is very similar device 731 00:25:07,310 --> 00:25:08,329 that a wikler. 732 00:25:08,330 --> 00:25:10,399 The only difference is that now 733 00:25:10,400 --> 00:25:12,829 the bending radius is so small that the, 734 00:25:12,830 --> 00:25:15,139 um, that the education can 735 00:25:15,140 --> 00:25:17,269 basically always point in the direction 736 00:25:17,270 --> 00:25:18,589 of the experiment. 737 00:25:18,590 --> 00:25:20,659 And the mathematical details of this 738 00:25:20,660 --> 00:25:22,789 radiation is a bit are a bit 739 00:25:22,790 --> 00:25:24,889 complicated. But the idea is that now 740 00:25:24,890 --> 00:25:26,959 you have interference of the light 741 00:25:26,960 --> 00:25:28,609 emitted in each of these curves. 742 00:25:28,610 --> 00:25:30,469 And by doing this, you compress the 743 00:25:30,470 --> 00:25:32,629 overall power here from a Regla into 744 00:25:32,630 --> 00:25:35,149 very narrow spikes and frequency domain. 745 00:25:35,150 --> 00:25:37,069 And this is great because remember, we 746 00:25:37,070 --> 00:25:39,379 want to make X-ray diffraction images, 747 00:25:39,380 --> 00:25:40,549 so we need coherent light. 748 00:25:40,550 --> 00:25:42,739 So we need only one wavelength anyway. 749 00:25:42,740 --> 00:25:44,509 So we put a filter in somewhere. 750 00:25:44,510 --> 00:25:46,519 And if you place a filter at that 751 00:25:46,520 --> 00:25:48,739 frequency, you will gain a huge 752 00:25:48,740 --> 00:25:49,919 amount of brilliance. 753 00:25:49,920 --> 00:25:52,879 Right. And these kind of devices 754 00:25:52,880 --> 00:25:55,189 are considered as third generation 755 00:25:55,190 --> 00:25:57,499 synchrotrons. So facilities 756 00:25:57,500 --> 00:25:59,419 which are dedicatedly built to produce as 757 00:25:59,420 --> 00:26:01,579 much synchrotron radiation as possible 758 00:26:01,580 --> 00:26:03,949 with many beamlines and many experiments. 759 00:26:05,630 --> 00:26:07,939 And as you can see here, there are plenty 760 00:26:07,940 --> 00:26:09,979 of them operated in the industrialized 761 00:26:09,980 --> 00:26:12,139 countries around the world right now. 762 00:26:12,140 --> 00:26:14,509 And as an example, I want to show you 763 00:26:14,510 --> 00:26:16,699 the three accelerator. 764 00:26:16,700 --> 00:26:18,979 We have a DC in Hamburg. 765 00:26:18,980 --> 00:26:21,169 But let me 766 00:26:21,170 --> 00:26:22,170 drink something. 767 00:26:27,910 --> 00:26:30,009 OK, so 768 00:26:30,010 --> 00:26:32,379 this is a D.C. campus, and here this ring 769 00:26:32,380 --> 00:26:34,539 is Petcare three, it has 770 00:26:34,540 --> 00:26:36,399 a circumference of about two point three 771 00:26:36,400 --> 00:26:38,139 kilometers. So it's a fairly large 772 00:26:38,140 --> 00:26:40,209 device, including this 300 773 00:26:40,210 --> 00:26:42,279 meter long experimental hall, of which 774 00:26:42,280 --> 00:26:44,319 a schematic sketch you can see here. 775 00:26:44,320 --> 00:26:46,719 And each of these lines is an X-ray 776 00:26:46,720 --> 00:26:49,449 beam line with dedicated experiments 777 00:26:49,450 --> 00:26:51,039 from the inside. It looks like this. 778 00:26:51,040 --> 00:26:53,109 So you can't really see the accelerator 779 00:26:53,110 --> 00:26:54,729 because everything has to be shielded 780 00:26:54,730 --> 00:26:56,769 with these concrete walls because of the 781 00:26:56,770 --> 00:26:57,729 radiation. 782 00:26:57,730 --> 00:26:59,739 But the accelerators here on the inner 783 00:26:59,740 --> 00:27:01,989 ring, this is a picture from the inside. 784 00:27:01,990 --> 00:27:04,149 And here you have the beamlines with 785 00:27:04,150 --> 00:27:07,059 the experimental chambers at the end. 786 00:27:07,060 --> 00:27:08,319 OK. 787 00:27:08,320 --> 00:27:09,819 As I said, this is a picture from the 788 00:27:09,820 --> 00:27:11,769 inside. So these are the quadripolar 789 00:27:11,770 --> 00:27:13,209 magnets. And we have some steering 790 00:27:13,210 --> 00:27:15,219 magnets and the yellow devices here. 791 00:27:15,220 --> 00:27:17,349 These are the undulates which produce 792 00:27:17,350 --> 00:27:18,789 the radiation. 793 00:27:18,790 --> 00:27:20,919 And beamline at these facilities is 794 00:27:20,920 --> 00:27:21,909 very expensive. 795 00:27:21,910 --> 00:27:24,249 So most of the beamlines have to be 796 00:27:24,250 --> 00:27:26,019 optimized. 797 00:27:26,020 --> 00:27:28,209 For example, at this one here 798 00:27:28,210 --> 00:27:30,309 we have a robot arm which takes the 799 00:27:30,310 --> 00:27:32,889 crystal samples out of the viewer 800 00:27:32,890 --> 00:27:34,779 here and then mounted on the sample 801 00:27:34,780 --> 00:27:35,949 holder. 802 00:27:35,950 --> 00:27:37,989 And the accuracy is very impressive. 803 00:27:37,990 --> 00:27:39,349 I mean, we have X-ray 804 00:27:40,510 --> 00:27:42,729 sample crystals as small as 100 805 00:27:42,730 --> 00:27:44,349 nanometer, and then they are rotated 806 00:27:44,350 --> 00:27:46,599 around the axis inside a photon 807 00:27:46,600 --> 00:27:48,789 beam, which is as small as 100 nanometer 808 00:27:48,790 --> 00:27:49,790 as well. 809 00:27:51,310 --> 00:27:53,979 But why do we need crystals at all? 810 00:27:53,980 --> 00:27:55,599 The reason for this is that the 811 00:27:55,600 --> 00:27:57,969 cross-section between our x rays 812 00:27:57,970 --> 00:27:59,679 and metal is very low. 813 00:27:59,680 --> 00:28:01,629 So statistically, we need one million 814 00:28:01,630 --> 00:28:03,549 atoms in a row to get one single 815 00:28:03,550 --> 00:28:05,019 refracted photon. 816 00:28:05,020 --> 00:28:06,609 And you can imagine we need much more 817 00:28:06,610 --> 00:28:08,799 than one single photon to get an 818 00:28:08,800 --> 00:28:10,659 image on our detector, which we can 819 00:28:10,660 --> 00:28:12,249 calculate something from. 820 00:28:12,250 --> 00:28:13,869 Right. So what we do is we have to 821 00:28:13,870 --> 00:28:15,249 increase the amount of photons. 822 00:28:15,250 --> 00:28:17,649 But this is limited by 823 00:28:17,650 --> 00:28:19,389 some constraints of our particle 824 00:28:19,390 --> 00:28:21,459 accelerators. So we have to increase the 825 00:28:21,460 --> 00:28:23,589 amount of atoms in our sample and we 826 00:28:23,590 --> 00:28:25,779 do this by growing growing 827 00:28:25,780 --> 00:28:27,849 crystals. So this is a protein 828 00:28:27,850 --> 00:28:30,069 and we have to find proteins which 829 00:28:30,070 --> 00:28:32,349 we can form unit cells out of and 830 00:28:32,350 --> 00:28:34,229 then grow a crystal from. 831 00:28:35,950 --> 00:28:37,989 So we need many of these and then we can 832 00:28:37,990 --> 00:28:40,389 put the crystal in our light and our 833 00:28:40,390 --> 00:28:42,519 X-ray beam and get some diffraction 834 00:28:42,520 --> 00:28:44,379 spots, and while turning the crystal 835 00:28:44,380 --> 00:28:46,089 around its axis, we'll get a 3D 836 00:28:46,090 --> 00:28:47,319 diffraction pattern. 837 00:28:47,320 --> 00:28:49,149 And from this, we can then calculate the 838 00:28:49,150 --> 00:28:51,849 Fredi electron density map of our sample. 839 00:28:51,850 --> 00:28:54,189 And if we know the electron density map, 840 00:28:54,190 --> 00:28:55,190 we know the structure. 841 00:28:57,240 --> 00:28:59,309 Here you can see the cumulative number 842 00:28:59,310 --> 00:29:01,139 of structures which are available in the 843 00:29:01,140 --> 00:29:03,329 protein database, and 844 00:29:03,330 --> 00:29:05,009 you can see that within the last 20 845 00:29:05,010 --> 00:29:06,989 years, there's been an astonishing 846 00:29:06,990 --> 00:29:09,329 increase, most importantly made possible 847 00:29:09,330 --> 00:29:11,429 by X-ray diffraction images 848 00:29:11,430 --> 00:29:13,229 in these modern third generation 849 00:29:13,230 --> 00:29:14,759 synchrotron light sources. 850 00:29:14,760 --> 00:29:17,069 And right now, we are not only able to 851 00:29:17,070 --> 00:29:19,649 make pictures of small proteins 852 00:29:19,650 --> 00:29:21,479 like the myoglobin, but even very big 853 00:29:21,480 --> 00:29:23,039 ones like the ribosome. 854 00:29:23,040 --> 00:29:25,019 But this is not really by far not 855 00:29:25,020 --> 00:29:27,599 trivial. So, for example, the ribosome, 856 00:29:27,600 --> 00:29:29,489 the first X-ray diffraction pattern of 857 00:29:29,490 --> 00:29:31,709 ribosome was made in 1980, 858 00:29:31,710 --> 00:29:33,899 but it took scientists 20 years 859 00:29:33,900 --> 00:29:35,909 to calculate the structure from this. 860 00:29:37,050 --> 00:29:39,299 And although this number here 861 00:29:39,300 --> 00:29:41,459 seems very high today, less than two 862 00:29:41,460 --> 00:29:44,069 percent of the human protein is known. 863 00:29:44,070 --> 00:29:46,649 So 89, 98 864 00:29:46,650 --> 00:29:48,839 percent of the proteins of our 865 00:29:48,840 --> 00:29:50,489 body are unknown. 866 00:29:50,490 --> 00:29:52,619 And the reason for this, the bottleneck 867 00:29:52,620 --> 00:29:54,239 is the crystal growth. 868 00:29:54,240 --> 00:29:56,609 So it's very hard to get most 869 00:29:56,610 --> 00:29:58,749 of the proteins to form big crystals. 870 00:29:58,750 --> 00:30:00,099 Some of them might even not. 871 00:30:00,100 --> 00:30:02,159 It's just not possible to crystallize 872 00:30:02,160 --> 00:30:03,659 them at all, for example, membrane 873 00:30:03,660 --> 00:30:04,679 proteins. 874 00:30:04,680 --> 00:30:06,479 But for others, it's very difficult to 875 00:30:06,480 --> 00:30:07,509 grow large crystals. 876 00:30:07,510 --> 00:30:09,599 So what we ideally want is to make 877 00:30:09,600 --> 00:30:11,669 be able to make a picture from a 878 00:30:11,670 --> 00:30:13,439 very small crystal or even a single 879 00:30:13,440 --> 00:30:14,759 molecule. 880 00:30:14,760 --> 00:30:16,679 But in order to do that, we have to 881 00:30:16,680 --> 00:30:18,539 increase the amount of photons by 882 00:30:18,540 --> 00:30:20,819 something like 100 million. 883 00:30:20,820 --> 00:30:21,929 And this is really not easy. 884 00:30:21,930 --> 00:30:23,849 But let's consider for now that we would 885 00:30:23,850 --> 00:30:25,619 be able to make a storage ring bright 886 00:30:25,620 --> 00:30:27,989 enough so 100 million times brighter 887 00:30:27,990 --> 00:30:29,309 than it is. 888 00:30:29,310 --> 00:30:31,649 And to make an 889 00:30:31,650 --> 00:30:34,019 diffraction image of a single lysosome, 890 00:30:34,020 --> 00:30:35,759 what would happen? 891 00:30:35,760 --> 00:30:38,369 Well, this this is 892 00:30:38,370 --> 00:30:40,289 a simulation published a couple of years 893 00:30:40,290 --> 00:30:41,939 ago. And what you see is the Coulomb 894 00:30:41,940 --> 00:30:43,859 explosion of lysozyme. 895 00:30:43,860 --> 00:30:46,199 So as the X-ray beam hits the hits 896 00:30:46,200 --> 00:30:48,599 the sample, it immediately blows 897 00:30:48,600 --> 00:30:50,939 away all the electrons in the molecule. 898 00:30:50,940 --> 00:30:52,769 And what is left are the positively 899 00:30:52,770 --> 00:30:54,959 charged nuclei which repel 900 00:30:54,960 --> 00:30:56,009 each other. 901 00:30:56,010 --> 00:30:58,799 So the whole molecule blows apart. 902 00:30:58,800 --> 00:31:00,509 And the problem is now that because a 903 00:31:00,510 --> 00:31:02,489 fundamental particle beam dynamics, it's 904 00:31:02,490 --> 00:31:04,859 not possible to make an X-ray storage 905 00:31:04,860 --> 00:31:07,079 ring smaller or shorter than 906 00:31:07,080 --> 00:31:08,729 about a picosecond. 907 00:31:08,730 --> 00:31:10,529 So even if we would be able to make a 908 00:31:10,530 --> 00:31:12,539 storage ring pulsed bright enough to 909 00:31:12,540 --> 00:31:14,639 watch a single molecule, we would just be 910 00:31:14,640 --> 00:31:16,589 able to see a blurry picture of an 911 00:31:16,590 --> 00:31:17,590 explosion. 912 00:31:18,240 --> 00:31:20,489 And this is where free electron laser 913 00:31:20,490 --> 00:31:22,259 came into play, because in a linear 914 00:31:22,260 --> 00:31:24,749 accelerator, it is fundamentally possible 915 00:31:24,750 --> 00:31:26,789 to produce an X-ray pulse as short as 916 00:31:26,790 --> 00:31:28,379 some femtosecond. 917 00:31:28,380 --> 00:31:30,959 But as I told you, we have to put 918 00:31:30,960 --> 00:31:33,269 100 million times more photons 919 00:31:33,270 --> 00:31:35,189 in that short putts. 920 00:31:35,190 --> 00:31:36,689 So this is not easy. 921 00:31:36,690 --> 00:31:38,699 And what we do is, well, first of all, 922 00:31:38,700 --> 00:31:40,679 let me rescale this picture. 923 00:31:40,680 --> 00:31:41,790 We replace 924 00:31:43,470 --> 00:31:45,569 we replace the underlaid 925 00:31:45,570 --> 00:31:47,819 by a very long Angelito and 926 00:31:47,820 --> 00:31:48,820 by doing. 927 00:31:51,290 --> 00:31:53,809 And now comes to mind, because 928 00:31:53,810 --> 00:31:55,879 if we set up everything correctly, we 929 00:31:55,880 --> 00:31:58,249 will get on top of this radiation pattern 930 00:31:58,250 --> 00:31:59,689 from a long, unrelated, we will get 931 00:31:59,690 --> 00:32:02,209 narrow spikes of coherent radiation. 932 00:32:02,210 --> 00:32:03,889 And this is what makes a free electron 933 00:32:03,890 --> 00:32:05,239 laser so important. 934 00:32:05,240 --> 00:32:07,459 So mathematically, the radiation 935 00:32:07,460 --> 00:32:09,259 scales now with a number of electrons 936 00:32:09,260 --> 00:32:11,299 squared and the number of electrons in 937 00:32:11,300 --> 00:32:13,369 our bunches is something like 100 million 938 00:32:13,370 --> 00:32:15,379 or a billion. So this is really a huge 939 00:32:15,380 --> 00:32:18,049 number. But let's have a look 940 00:32:18,050 --> 00:32:20,449 inside what's happening in the undulate. 941 00:32:20,450 --> 00:32:22,219 So this is an electron bunch and all the 942 00:32:22,220 --> 00:32:24,499 red circles are supposed to be electrons 943 00:32:24,500 --> 00:32:25,849 and the whole bunch moves in the 944 00:32:25,850 --> 00:32:26,850 undulate, OK? 945 00:32:27,890 --> 00:32:29,779 And they exist. The resonance condition 946 00:32:29,780 --> 00:32:31,909 between the undulate period and the 947 00:32:31,910 --> 00:32:33,889 period of the emitted light. 948 00:32:33,890 --> 00:32:35,779 So here you have the underlaid period, 949 00:32:35,780 --> 00:32:38,089 the the emitted light, 950 00:32:38,090 --> 00:32:39,829 the relativistic gamma factor in this 951 00:32:39,830 --> 00:32:41,809 Cavalia, which corroborate some 952 00:32:41,810 --> 00:32:43,399 information about the magnetic field. 953 00:32:43,400 --> 00:32:45,199 But it's not important for now. 954 00:32:45,200 --> 00:32:47,419 So I'm only looking at the wavelength 955 00:32:47,420 --> 00:32:49,429 of the emitted light, which satisfies 956 00:32:49,430 --> 00:32:50,899 this condition. 957 00:32:50,900 --> 00:32:51,900 OK. 958 00:32:53,080 --> 00:32:54,399 Now, let's have a look. 959 00:32:54,400 --> 00:32:57,309 So this is the electromagnetic wave 960 00:32:57,310 --> 00:32:59,709 which is emitted by that electron 961 00:32:59,710 --> 00:33:02,019 and the whole bunch is moving up and down 962 00:33:02,020 --> 00:33:03,249 in that picture. 963 00:33:03,250 --> 00:33:05,049 So some of the electrons move in the 964 00:33:05,050 --> 00:33:07,509 direction of the electric field. 965 00:33:07,510 --> 00:33:09,249 Sorry, this is the electric field line, 966 00:33:09,250 --> 00:33:10,719 which I plotted here. 967 00:33:10,720 --> 00:33:12,849 So some of the electrons move in the 968 00:33:12,850 --> 00:33:14,529 direction of the electric field while 969 00:33:14,530 --> 00:33:15,969 some of them move in the opposite 970 00:33:15,970 --> 00:33:18,069 direction. OK, so some of 971 00:33:18,070 --> 00:33:20,139 them will gain transverse momentum 972 00:33:20,140 --> 00:33:22,299 while others will lose it. 973 00:33:22,300 --> 00:33:24,429 And if we hit this resonance condition, 974 00:33:24,430 --> 00:33:26,169 both the direction of the motion of the 975 00:33:26,170 --> 00:33:28,269 electrons and the electromagnetic 976 00:33:28,270 --> 00:33:30,579 waves flip side at the same 977 00:33:30,580 --> 00:33:32,889 time. So this process repeats 978 00:33:32,890 --> 00:33:33,879 itself. 979 00:33:33,880 --> 00:33:35,529 And while all of this is happening, we 980 00:33:35,530 --> 00:33:37,689 are in a magnetic chicane, meaning that 981 00:33:37,690 --> 00:33:40,089 there exists a dispersion and dispersion 982 00:33:40,090 --> 00:33:42,489 means that the bending 983 00:33:42,490 --> 00:33:44,319 radius depends on the energy. 984 00:33:44,320 --> 00:33:46,419 So if you have a larger energy and the 985 00:33:46,420 --> 00:33:47,979 bending radius is bigger, if you have a 986 00:33:47,980 --> 00:33:49,749 smaller energy, the bending radius of 987 00:33:49,750 --> 00:33:51,099 smaller. 988 00:33:51,100 --> 00:33:53,289 So some of these particles have 989 00:33:53,290 --> 00:33:55,299 larger transverse momentum, so larger 990 00:33:55,300 --> 00:33:57,459 transverse energy, so to say, and they 991 00:33:57,460 --> 00:33:59,619 will move, they will fall 992 00:33:59,620 --> 00:34:01,989 back and others will overtake 993 00:34:01,990 --> 00:34:03,819 the bunch. So we have a self ordering 994 00:34:03,820 --> 00:34:05,880 effect which repeats itself 995 00:34:07,420 --> 00:34:09,789 now coming back to the big picture. 996 00:34:09,790 --> 00:34:11,859 So from at the beginning, we start 997 00:34:11,860 --> 00:34:13,299 with an incoherent radiation. 998 00:34:13,300 --> 00:34:14,799 So all the electrons, while they are 999 00:34:14,800 --> 00:34:16,448 moving out of where they are bent around 1000 00:34:16,449 --> 00:34:19,388 the circle, radiate, 1001 00:34:19,389 --> 00:34:21,129 but there is no fixed face relation 1002 00:34:21,130 --> 00:34:23,589 between them. OK, so this is incoherent 1003 00:34:23,590 --> 00:34:25,689 radiation and the intensity of that such 1004 00:34:25,690 --> 00:34:27,729 kind of radiation scales with the number 1005 00:34:27,730 --> 00:34:30,849 of emitters in this kind, in this 1006 00:34:30,850 --> 00:34:33,009 example number of electrons. 1007 00:34:33,010 --> 00:34:35,499 And now as the bands move through 1008 00:34:35,500 --> 00:34:38,259 the undulate are the self altering effect 1009 00:34:38,260 --> 00:34:40,419 leads to a micro bunching on exactly 1010 00:34:40,420 --> 00:34:43,119 that length scale of that radiation. 1011 00:34:43,120 --> 00:34:45,638 So for the wavelength satisfying 1012 00:34:45,639 --> 00:34:47,799 this condition, we will get a coherent 1013 00:34:47,800 --> 00:34:50,019 radiation and coherent radiation scales 1014 00:34:50,020 --> 00:34:51,499 with the number of electrons quit. 1015 00:34:52,929 --> 00:34:55,178 But it's not easy to get from 1016 00:34:55,179 --> 00:34:57,129 incoherent to coherent radiation, 1017 00:34:57,130 --> 00:34:59,199 especially if you want to 1018 00:34:59,200 --> 00:35:01,000 want to have x rays here. 1019 00:35:02,530 --> 00:35:03,530 Sorry. 1020 00:35:04,540 --> 00:35:06,789 So what we need is a small beam. 1021 00:35:06,790 --> 00:35:08,859 This is just to give you an idea of 1022 00:35:08,860 --> 00:35:10,599 the order, so don't take these well to 1023 00:35:10,600 --> 00:35:13,269 serious that can be affected two or three 1024 00:35:13,270 --> 00:35:15,519 between them. But we need a small beam, 1025 00:35:15,520 --> 00:35:17,049 something like 10 micrometer and 1026 00:35:17,050 --> 00:35:18,039 Transperth size. 1027 00:35:18,040 --> 00:35:20,019 We have to have it as short as 10 1028 00:35:20,020 --> 00:35:22,089 micrometer and we have to 1029 00:35:22,090 --> 00:35:24,609 get it on the energy, something like 10 1030 00:35:24,610 --> 00:35:26,229 billion electron loads. 1031 00:35:26,230 --> 00:35:28,119 And we need a very long until later some 1032 00:35:28,120 --> 00:35:29,169 hundred meters. 1033 00:35:29,170 --> 00:35:31,269 And within that undelete undue 1034 00:35:31,270 --> 00:35:33,939 later. We have to align the electrons 1035 00:35:33,940 --> 00:35:36,159 to better than 10 micrometer in order 1036 00:35:36,160 --> 00:35:38,019 to have an overlap between the electrons 1037 00:35:38,020 --> 00:35:40,539 and their and their light. 1038 00:35:40,540 --> 00:35:41,949 So this is really challenging. 1039 00:35:43,150 --> 00:35:45,009 This is a sketch of such a free electron 1040 00:35:45,010 --> 00:35:46,809 laser. So usually we have different 1041 00:35:46,810 --> 00:35:48,849 acceleration stages and in between we 1042 00:35:48,850 --> 00:35:50,469 have magnetic chicanes. 1043 00:35:50,470 --> 00:35:52,479 We call them bunched compressors, and in 1044 00:35:52,480 --> 00:35:54,399 them we are able to produce the very 1045 00:35:54,400 --> 00:35:56,529 short bunches and we have a long until 1046 00:35:56,530 --> 00:35:59,019 later. And finally we dump the electrons 1047 00:35:59,020 --> 00:36:01,510 in our light, gets to our experiments. 1048 00:36:04,000 --> 00:36:06,489 As you can see here, they are just now, 1049 00:36:06,490 --> 00:36:09,099 five of them in operation 1050 00:36:09,100 --> 00:36:11,469 and at least 1051 00:36:11,470 --> 00:36:13,319 five of them operating in the heart X 1052 00:36:13,320 --> 00:36:15,399 X-ray regime. And just another example I 1053 00:36:15,400 --> 00:36:17,949 would like to show you the European SFL, 1054 00:36:17,950 --> 00:36:19,809 which is the largest free electron laser 1055 00:36:19,810 --> 00:36:21,399 we have on earth. 1056 00:36:21,400 --> 00:36:23,559 This is a map from Hamburg. 1057 00:36:23,560 --> 00:36:25,779 You can see with its overall a length 1058 00:36:25,780 --> 00:36:27,069 of about three kilometers. 1059 00:36:27,070 --> 00:36:28,449 It reaches out the D.C. 1060 00:36:28,450 --> 00:36:30,909 campus and 1061 00:36:30,910 --> 00:36:33,219 reaches the adjacent federal state of 1062 00:36:33,220 --> 00:36:36,189 the college town where the experiments, 1063 00:36:36,190 --> 00:36:38,559 the experimental hall is built, 1064 00:36:38,560 --> 00:36:39,939 where you can't see much from above 1065 00:36:39,940 --> 00:36:42,279 because everything is just beneath 1066 00:36:42,280 --> 00:36:43,089 the Earth. 1067 00:36:43,090 --> 00:36:44,559 They would like to show you that movie 1068 00:36:44,560 --> 00:36:46,119 which was made. 1069 00:36:46,120 --> 00:36:47,109 Hmm. 1070 00:36:47,110 --> 00:36:49,629 Which was made by the accelerator, 1071 00:36:49,630 --> 00:36:50,949 was still under construction. 1072 00:36:50,950 --> 00:36:53,379 So right now, it would not be possible to 1073 00:36:53,380 --> 00:36:54,799 to walk down there. 1074 00:36:54,800 --> 00:36:56,139 It would just die. 1075 00:36:56,140 --> 00:36:58,149 But that was possible. 1076 00:36:58,150 --> 00:36:59,150 And I think, 1077 00:37:00,460 --> 00:37:02,529 yeah, it was really impressive to 1078 00:37:02,530 --> 00:37:04,359 be down there and see all this high tech 1079 00:37:04,360 --> 00:37:07,599 next to you and it just never stops. 1080 00:37:07,600 --> 00:37:09,669 But anyway, you can see this is 1081 00:37:09,670 --> 00:37:10,839 now the main accelerator. 1082 00:37:10,840 --> 00:37:12,849 It goes on for another kilometer. 1083 00:37:12,850 --> 00:37:13,749 You see where we are. 1084 00:37:13,750 --> 00:37:15,279 And this goes on for two minutes. 1085 00:37:15,280 --> 00:37:17,259 I think it's a bit boring, but you can 1086 00:37:17,260 --> 00:37:20,259 watch this movie if you want at home, 1087 00:37:20,260 --> 00:37:22,020 I think at double the speed anyway. 1088 00:37:23,260 --> 00:37:25,449 But I want to give you some numbers. 1089 00:37:25,450 --> 00:37:27,519 So in average, we we 1090 00:37:27,520 --> 00:37:29,679 take about nine point five megawatts 1091 00:37:29,680 --> 00:37:30,729 from the grid. 1092 00:37:30,730 --> 00:37:32,799 This is about the energy consumption of 1093 00:37:32,800 --> 00:37:34,299 a small city. 1094 00:37:34,300 --> 00:37:36,549 And from that, due to the utilization 1095 00:37:36,550 --> 00:37:38,589 of the superconducting RF technology, we 1096 00:37:38,590 --> 00:37:40,749 are able to put 10 percent into our beam. 1097 00:37:40,750 --> 00:37:42,909 So we have an average beam power of 1098 00:37:42,910 --> 00:37:45,009 900 kilowatts, which is really 1099 00:37:45,010 --> 00:37:47,679 impressive for linear accelerator. 1100 00:37:47,680 --> 00:37:50,319 From this, we get 0.01 1101 00:37:50,320 --> 00:37:52,659 percent into our X-ray beam, 1102 00:37:52,660 --> 00:37:55,179 but finally, only less than one percent 1103 00:37:55,180 --> 00:37:57,339 hitting or 1104 00:37:57,340 --> 00:37:59,379 getting the diffraction spots. 1105 00:37:59,380 --> 00:38:01,689 So you could argue that the overall 1106 00:38:01,690 --> 00:38:03,549 efficiency of this machine is terrible. 1107 00:38:04,910 --> 00:38:05,980 And I would agree. 1108 00:38:07,060 --> 00:38:09,369 And also, 900 watts of X-ray 1109 00:38:09,370 --> 00:38:11,439 beam power seems not very impressive. 1110 00:38:11,440 --> 00:38:13,389 But what makes this machine worth a 1111 00:38:13,390 --> 00:38:15,489 billion euro is its ability 1112 00:38:15,490 --> 00:38:17,619 to compress that power into very narrow 1113 00:38:17,620 --> 00:38:18,669 space. 1114 00:38:18,670 --> 00:38:20,739 So what is interesting is the peak 1115 00:38:20,740 --> 00:38:22,329 power in average. 1116 00:38:22,330 --> 00:38:24,759 We have a repetition rate of 27 1117 00:38:24,760 --> 00:38:27,099 kilohertz or 27000 1118 00:38:27,100 --> 00:38:29,679 X-ray pulses per second are produced 1119 00:38:29,680 --> 00:38:31,209 with a wavelength of about four point 1120 00:38:31,210 --> 00:38:33,429 five angstrom pulse energy 1121 00:38:33,430 --> 00:38:35,649 of one molecule and a pulse duration down 1122 00:38:35,650 --> 00:38:37,629 to three femtosecond. 1123 00:38:37,630 --> 00:38:39,699 I mean, this is the time it 1124 00:38:39,700 --> 00:38:41,909 takes light to travel one micrometer. 1125 00:38:42,940 --> 00:38:44,739 This is really short 1126 00:38:45,820 --> 00:38:48,069 and we can focus 1127 00:38:48,070 --> 00:38:50,139 this X-ray beam down to a very 1128 00:38:50,140 --> 00:38:51,999 narrow spot. And in this spot, in the 1129 00:38:52,000 --> 00:38:54,159 focal point, we have a power tent city 1130 00:38:54,160 --> 00:38:56,229 of about 10 to 17 watts per square 1131 00:38:56,230 --> 00:38:57,819 centimeter. 1132 00:38:57,820 --> 00:38:59,919 I guess you don't know what 2010 1133 00:38:59,920 --> 00:39:02,109 to the 17 watts per square centimeter is. 1134 00:39:02,110 --> 00:39:03,609 But let me give you an example. 1135 00:39:03,610 --> 00:39:05,709 Is this about the power density, as 1136 00:39:05,710 --> 00:39:07,509 is you if you would take the total 1137 00:39:07,510 --> 00:39:09,609 sunlight hitting the earth on 1138 00:39:09,610 --> 00:39:11,709 one square centimeter. 1139 00:39:11,710 --> 00:39:13,869 So this is really intense and 1140 00:39:13,870 --> 00:39:15,339 you have to be careful because if you 1141 00:39:15,340 --> 00:39:16,659 accidentally hit something. 1142 00:39:20,170 --> 00:39:22,239 Another thing I would like to 1143 00:39:22,240 --> 00:39:24,579 show you is that it's really 1144 00:39:24,580 --> 00:39:26,649 not easy to build or to operate such a 1145 00:39:26,650 --> 00:39:29,109 machine, like just for the European 1146 00:39:29,110 --> 00:39:31,449 SFL, we have nine million control 1147 00:39:31,450 --> 00:39:32,679 system variables. 1148 00:39:32,680 --> 00:39:33,969 This is the picture I made from the 1149 00:39:33,970 --> 00:39:35,649 control room at DC. 1150 00:39:35,650 --> 00:39:36,849 And you see that there are a lot of 1151 00:39:36,850 --> 00:39:38,949 screens and you have access to all of 1152 00:39:38,950 --> 00:39:41,109 them. So it's really not easy to design 1153 00:39:41,110 --> 00:39:43,119 a control system which can be operated by 1154 00:39:43,120 --> 00:39:45,399 many people and give 1155 00:39:45,400 --> 00:39:46,869 you access to all of these. 1156 00:39:46,870 --> 00:39:48,609 And I made an animation or screen 1157 00:39:48,610 --> 00:39:49,899 recording because once I had a 1158 00:39:49,900 --> 00:39:51,849 measurement shift at Flesche, which is 1159 00:39:51,850 --> 00:39:53,799 another electronic free electron laser, 1160 00:39:53,800 --> 00:39:54,879 we have it easy. 1161 00:39:54,880 --> 00:39:57,369 And I had to take out toroidal signal, 1162 00:39:57,370 --> 00:39:58,899 which was not on the top layer of the 1163 00:39:58,900 --> 00:39:59,979 control system. 1164 00:39:59,980 --> 00:40:02,709 It took me quite some time to find it. 1165 00:40:02,710 --> 00:40:05,159 So this is the this is 1166 00:40:05,160 --> 00:40:07,519 the top panel of the control system. 1167 00:40:07,520 --> 00:40:09,669 And as you can see, as you press some 1168 00:40:09,670 --> 00:40:11,559 of these buttons there, will open up a 1169 00:40:11,560 --> 00:40:14,439 panel with a lot of buttons. 1170 00:40:14,440 --> 00:40:17,169 And if you press one of these buttons, 1171 00:40:17,170 --> 00:40:19,989 another panel opens and 1172 00:40:19,990 --> 00:40:20,990 please. 1173 00:40:21,970 --> 00:40:22,929 OK. 1174 00:40:22,930 --> 00:40:25,269 And please note the saphenous over here 1175 00:40:25,270 --> 00:40:26,270 and here. 1176 00:40:28,870 --> 00:40:29,870 But finally. 1177 00:40:38,940 --> 00:40:41,159 So really, we need a lot of experts 1178 00:40:41,160 --> 00:40:42,959 working together because no one is able 1179 00:40:42,960 --> 00:40:44,399 to keep all of that in mind. 1180 00:40:47,060 --> 00:40:49,579 Another interesting number I find 1181 00:40:49,580 --> 00:40:51,949 is the data production rate, so 1182 00:40:51,950 --> 00:40:53,569 now I'm not talking about the machine, 1183 00:40:53,570 --> 00:40:55,009 I'm just talking about the x ray 1184 00:40:55,010 --> 00:40:57,139 detector. OK, and then there 1185 00:40:57,140 --> 00:40:58,939 we have about one megapixel with the 1186 00:40:58,940 --> 00:41:00,739 resolution of sixteen bit. 1187 00:41:00,740 --> 00:41:03,499 And we want to record this 27000 1188 00:41:03,500 --> 00:41:04,669 times per second. 1189 00:41:04,670 --> 00:41:06,709 And this means we have 60 gigabytes per 1190 00:41:06,710 --> 00:41:07,729 second. 1191 00:41:07,730 --> 00:41:09,889 Just to give you a number, the 1192 00:41:09,890 --> 00:41:12,259 LHC, after filtering has about 600 1193 00:41:12,260 --> 00:41:13,339 megabits per second. 1194 00:41:13,340 --> 00:41:15,319 So you can imagine that we also need very 1195 00:41:15,320 --> 00:41:17,569 sophisticated, uh, trigger 1196 00:41:17,570 --> 00:41:19,699 levels in order to deal with this amount 1197 00:41:19,700 --> 00:41:21,919 of data, because no one is able to record 1198 00:41:21,920 --> 00:41:24,889 or manage 60 gigabytes per second. 1199 00:41:24,890 --> 00:41:27,019 And as an example, this is the 1200 00:41:27,020 --> 00:41:29,419 amount of stored data in the first 1201 00:41:29,420 --> 00:41:31,909 weeks of operation of the European SFL. 1202 00:41:31,910 --> 00:41:33,889 So you see, we are hundreds of terabytes. 1203 00:41:33,890 --> 00:41:35,779 And keep in mind, within that period, the 1204 00:41:35,780 --> 00:41:37,579 machine was working with less than 10 1205 00:41:37,580 --> 00:41:40,039 percent of its full capacity. 1206 00:41:40,040 --> 00:41:41,869 So we are talking about petabytes here. 1207 00:41:42,920 --> 00:41:45,229 So this is also not that easy 1208 00:41:45,230 --> 00:41:46,789 to control. 1209 00:41:46,790 --> 00:41:48,769 But finally, I would like to close this 1210 00:41:48,770 --> 00:41:50,989 talk with a unique application, which 1211 00:41:50,990 --> 00:41:52,999 you can only do at these free electron 1212 00:41:53,000 --> 00:41:55,909 lasers. And it's about molecular movies. 1213 00:41:55,910 --> 00:41:58,250 So, for example, this 1214 00:41:59,750 --> 00:42:01,549 Iren complex in etc.. 1215 00:42:01,550 --> 00:42:03,319 Nitrile solution, if you hit it with the 1216 00:42:03,320 --> 00:42:06,049 UV laser or UV light in general, 1217 00:42:06,050 --> 00:42:08,179 then it will perform a chemical reaction 1218 00:42:08,180 --> 00:42:10,369 and will lead to an acid agent 1219 00:42:10,370 --> 00:42:11,899 and the building of such a solvent 1220 00:42:11,900 --> 00:42:14,209 molecule. OK, this is chemistry. 1221 00:42:14,210 --> 00:42:16,189 We know this for many decades. 1222 00:42:16,190 --> 00:42:18,079 But the problem is that basically all of 1223 00:42:18,080 --> 00:42:20,299 our knowledge of chemistry is 1224 00:42:20,300 --> 00:42:22,189 equilibriums science. 1225 00:42:22,190 --> 00:42:24,349 So we know the reactants and we know 1226 00:42:24,350 --> 00:42:26,479 the reaction products, but we don't 1227 00:42:26,480 --> 00:42:28,669 know what's happening in between. 1228 00:42:28,670 --> 00:42:30,199 And usually there is not only one 1229 00:42:30,200 --> 00:42:32,389 reaction, PAF, but there are many of 1230 00:42:32,390 --> 00:42:33,919 different probabilities. 1231 00:42:33,920 --> 00:42:35,419 And you can imagine if we don't know 1232 00:42:35,420 --> 00:42:37,639 anything in between, it's very hard 1233 00:42:37,640 --> 00:42:40,069 for us to design a drug or a catalyst 1234 00:42:40,070 --> 00:42:41,089 or something like this. 1235 00:42:41,090 --> 00:42:43,539 This is basically nothing more than 1236 00:42:43,540 --> 00:42:45,679 an applied science 1237 00:42:45,680 --> 00:42:48,169 alchemy. I mean, we just try and error. 1238 00:42:48,170 --> 00:42:50,269 So it would we would really benefit from 1239 00:42:50,270 --> 00:42:52,459 knowing what's happening in between and 1240 00:42:52,460 --> 00:42:53,629 with the FCL. 1241 00:42:53,630 --> 00:42:55,069 We can do this. 1242 00:42:55,070 --> 00:42:57,229 This is a picture of the experimental 1243 00:42:57,230 --> 00:42:58,879 hall and Shoenfeld here we have these 1244 00:42:58,880 --> 00:43:01,189 five beamlines and now we are watching 1245 00:43:01,190 --> 00:43:02,190 one of them. 1246 00:43:06,120 --> 00:43:07,120 So here 1247 00:43:08,190 --> 00:43:10,769 come our X-ray beams. 1248 00:43:10,770 --> 00:43:12,689 This is a photon diagnostic section where 1249 00:43:12,690 --> 00:43:14,399 you can analyze the properties of our 1250 00:43:14,400 --> 00:43:16,349 X-ray beams and here finally we have the 1251 00:43:16,350 --> 00:43:17,429 target. 1252 00:43:17,430 --> 00:43:19,919 This is a liquid jet 1253 00:43:19,920 --> 00:43:20,819 target. 1254 00:43:20,820 --> 00:43:23,069 And it's not easy to to design 1255 00:43:23,070 --> 00:43:25,679 because we want a single molecule 1256 00:43:25,680 --> 00:43:27,629 to get hit by our X-ray beam. 1257 00:43:27,630 --> 00:43:29,159 We don't want to have to and we don't 1258 00:43:29,160 --> 00:43:30,239 want to have zero. 1259 00:43:30,240 --> 00:43:31,739 And all of this has to happen in a 1260 00:43:31,740 --> 00:43:32,699 vacuum. 1261 00:43:32,700 --> 00:43:34,769 And it's really not trivial to build 1262 00:43:34,770 --> 00:43:37,329 these kind of experimental chambers. 1263 00:43:37,330 --> 00:43:39,449 OK, now how can we get to 1264 00:43:39,450 --> 00:43:40,899 a molecular movie from this? 1265 00:43:40,900 --> 00:43:42,509 So first of all, we have to be able to 1266 00:43:42,510 --> 00:43:44,639 trigger our reaction and we can do 1267 00:43:44,640 --> 00:43:46,739 this with a movie Laser Pulse. 1268 00:43:46,740 --> 00:43:48,989 So we hit our our molecules with the 1269 00:43:48,990 --> 00:43:51,119 laser and the reaction starts and then 1270 00:43:51,120 --> 00:43:52,919 we can make a snapshot with our X-ray 1271 00:43:52,920 --> 00:43:53,969 laser. 1272 00:43:53,970 --> 00:43:56,009 And by setting up the delay time between 1273 00:43:56,010 --> 00:43:57,869 the movie and the X-ray laser, we can 1274 00:43:57,870 --> 00:43:59,969 make a snapshot from different stages 1275 00:43:59,970 --> 00:44:02,099 of this reaction. 1276 00:44:02,100 --> 00:44:04,349 And that's basically everything. 1277 00:44:04,350 --> 00:44:06,479 But also the readout of this 1278 00:44:06,480 --> 00:44:08,699 detector is very sophisticated. 1279 00:44:08,700 --> 00:44:11,699 So between there are different layers 1280 00:44:11,700 --> 00:44:13,979 because between each pulse, there's only 1281 00:44:13,980 --> 00:44:16,259 200 nanoseconds and then the defector 1282 00:44:16,260 --> 00:44:17,789 has to be ready for the next picture. 1283 00:44:17,790 --> 00:44:19,289 So it's really not trivial to build 1284 00:44:19,290 --> 00:44:20,909 these. And this is basically the most 1285 00:44:20,910 --> 00:44:23,399 powerful X-ray detector we have on Earth. 1286 00:44:23,400 --> 00:44:25,889 But finally, we get our 1287 00:44:25,890 --> 00:44:27,959 images. And from each image, we 1288 00:44:27,960 --> 00:44:30,869 can calculate, um, 1289 00:44:30,870 --> 00:44:32,849 the structure of our molecule. 1290 00:44:32,850 --> 00:44:35,039 And by putting them all together, we 1291 00:44:35,040 --> 00:44:37,289 are able to make a molecular movie 1292 00:44:37,290 --> 00:44:38,549 from a chemical reaction. 1293 00:44:39,880 --> 00:44:41,549 You see what it takes to make something 1294 00:44:41,550 --> 00:44:43,829 like this. And you guys, I guess you 1295 00:44:43,830 --> 00:44:46,469 you understand that it's a long way 1296 00:44:46,470 --> 00:44:47,949 to get to something like this. 1297 00:44:47,950 --> 00:44:48,950 Right. 1298 00:44:49,350 --> 00:44:51,779 But in principle, I think 1299 00:44:51,780 --> 00:44:53,909 I have shown you not only how we are 1300 00:44:53,910 --> 00:44:55,829 able to resolve the structures of these 1301 00:44:55,830 --> 00:44:58,529 proteins, but also as 1302 00:44:58,530 --> 00:45:00,839 how free electron lasers may 1303 00:45:00,840 --> 00:45:02,939 enable us in a couple 1304 00:45:02,940 --> 00:45:05,099 of years, maybe decades, to watch 1305 00:45:05,100 --> 00:45:07,109 these kind of movies, but not as artists 1306 00:45:07,110 --> 00:45:09,269 used, but as real experimental 1307 00:45:09,270 --> 00:45:10,649 data. 1308 00:45:10,650 --> 00:45:12,959 So thank you very much. 1309 00:45:12,960 --> 00:45:14,249 And if there are any questions. 1310 00:45:43,970 --> 00:45:46,219 Person to person, thank 1311 00:45:46,220 --> 00:45:48,649 you very much for this highly educational 1312 00:45:48,650 --> 00:45:50,809 talk, if anything goes 1313 00:45:50,810 --> 00:45:52,999 wrong with your postdoc in 1314 00:45:53,000 --> 00:45:55,069 Berkeley, I recommend you go to 1315 00:45:55,070 --> 00:45:56,269 science communication. 1316 00:46:04,700 --> 00:46:06,919 OK, we already have a question from 1317 00:46:06,920 --> 00:46:07,939 the Internet. I have heard 1318 00:46:09,800 --> 00:46:11,749 yeah, there's actually one question from 1319 00:46:11,750 --> 00:46:13,069 Galkin. 1320 00:46:13,070 --> 00:46:15,769 How well do the experiments replicate? 1321 00:46:19,300 --> 00:46:22,449 I've seen the talk yesterday as well, 1322 00:46:22,450 --> 00:46:24,820 and I think, uh. 1323 00:46:27,280 --> 00:46:29,229 You mean in general, the x ray 1324 00:46:29,230 --> 00:46:31,630 experiments are from the European SFL. 1325 00:46:33,400 --> 00:46:34,869 It's on the Internet, right? 1326 00:46:34,870 --> 00:46:35,870 OK. 1327 00:46:37,390 --> 00:46:39,519 I, I 1328 00:46:39,520 --> 00:46:42,039 would say they replicate 1329 00:46:42,040 --> 00:46:44,199 quite well, there are experiments 1330 00:46:44,200 --> 00:46:46,419 made at different X-ray sources 1331 00:46:46,420 --> 00:46:48,579 and from time to time they try to 1332 00:46:48,580 --> 00:46:50,649 cross-check at other X-ray sources or 1333 00:46:50,650 --> 00:46:53,199 try to make the experiment 1334 00:46:53,200 --> 00:46:54,459 a bit different. 1335 00:46:54,460 --> 00:46:56,229 And I think this is kind of replicating 1336 00:46:56,230 --> 00:46:57,729 it right. 1337 00:46:57,730 --> 00:46:59,949 But I'm not a photon photon expert, 1338 00:46:59,950 --> 00:47:02,229 so I don't I'm not an I built 1339 00:47:02,230 --> 00:47:04,389 the machine. I don't really care 1340 00:47:04,390 --> 00:47:06,119 about the images. 1341 00:47:09,250 --> 00:47:10,250 So I'm sorry, 1342 00:47:11,420 --> 00:47:13,659 OK, microphone one, 1343 00:47:13,660 --> 00:47:14,769 please. 1344 00:47:14,770 --> 00:47:15,939 OK. 1345 00:47:15,940 --> 00:47:17,659 Yeah, really amazing talk. 1346 00:47:17,660 --> 00:47:20,229 I have to admit that what's 1347 00:47:20,230 --> 00:47:22,809 the current status of the X Files? 1348 00:47:22,810 --> 00:47:24,699 Because he showed now at the end, just as 1349 00:47:24,700 --> 00:47:26,799 this procedure, how you would do a movie. 1350 00:47:26,800 --> 00:47:28,899 How far are we actually to do that 1351 00:47:28,900 --> 00:47:30,520 for a simple example? 1352 00:47:35,050 --> 00:47:37,239 Something like a year, maybe, 1353 00:47:37,240 --> 00:47:39,369 I mean, it really depends for I didn't 1354 00:47:39,370 --> 00:47:41,619 told you how difficult it is to make 1355 00:47:41,620 --> 00:47:43,959 how many pictures you have to to combine 1356 00:47:43,960 --> 00:47:44,979 to make such a movie. 1357 00:47:44,980 --> 00:47:47,199 So you have to combine 1358 00:47:47,200 --> 00:47:49,569 several hundred of thousands of X-ray 1359 00:47:49,570 --> 00:47:51,939 images or diffraction 1360 00:47:51,940 --> 00:47:53,199 images to make such a movie. 1361 00:47:53,200 --> 00:47:55,389 So you need a lot of a lot of boom 1362 00:47:55,390 --> 00:47:58,089 time. And especially right now, I think 1363 00:47:58,090 --> 00:47:59,919 it's more difficult to prepare the 1364 00:47:59,920 --> 00:48:02,169 samples and and 1365 00:48:02,170 --> 00:48:03,999 and to get the full capacity because of 1366 00:48:04,000 --> 00:48:06,189 some issues of the accelerator, I 1367 00:48:06,190 --> 00:48:08,409 would guess something like one year 1368 00:48:08,410 --> 00:48:10,029 to get to something in general. 1369 00:48:10,030 --> 00:48:12,159 The machine is ready and operational 1370 00:48:12,160 --> 00:48:13,869 and it operates and could start doing 1371 00:48:13,870 --> 00:48:14,769 that. Yeah. 1372 00:48:14,770 --> 00:48:16,479 OK, thanks. 1373 00:48:16,480 --> 00:48:18,369 It's just not that all subsystems are 1374 00:48:18,370 --> 00:48:20,589 working, you know, like some of the some 1375 00:48:20,590 --> 00:48:22,089 of the experimental chambers are not 1376 00:48:22,090 --> 00:48:24,519 ready or some properties 1377 00:48:24,520 --> 00:48:26,529 can't be hit right now. 1378 00:48:26,530 --> 00:48:27,530 Yeah. 1379 00:48:28,720 --> 00:48:30,789 OK, microphone number four, please. 1380 00:48:31,900 --> 00:48:33,339 OK. 1381 00:48:33,340 --> 00:48:35,469 How do you stop the molecules degrading 1382 00:48:35,470 --> 00:48:37,539 by when they're 1383 00:48:37,540 --> 00:48:39,639 hit by the free electron laser? 1384 00:48:39,640 --> 00:48:40,719 Sorry again please. 1385 00:48:40,720 --> 00:48:42,789 So yourself before that, if you 1386 00:48:42,790 --> 00:48:44,839 don't have a crystal of molecules of the 1387 00:48:44,840 --> 00:48:46,469 degrades instantly. 1388 00:48:46,470 --> 00:48:48,189 And how do we stop it with the free 1389 00:48:48,190 --> 00:48:49,509 electron laser? 1390 00:48:49,510 --> 00:48:52,149 You mean how do we stop the molecule 1391 00:48:52,150 --> 00:48:53,379 from exploding. 1392 00:48:53,380 --> 00:48:54,729 Yes. Aren't we. 1393 00:48:54,730 --> 00:48:56,050 Don't. Oh OK. 1394 00:48:57,230 --> 00:48:59,389 It just it gets obliterated in each 1395 00:48:59,390 --> 00:49:01,549 shot, so that's why we have 1396 00:49:01,550 --> 00:49:03,829 to make 100000 pictures, 1397 00:49:03,830 --> 00:49:06,139 because after each week 1398 00:49:06,140 --> 00:49:08,299 maybe let me show you 1399 00:49:08,300 --> 00:49:10,339 this, maybe so. 1400 00:49:10,340 --> 00:49:12,529 So each shot, this is our 1401 00:49:12,530 --> 00:49:14,959 molecule and gets hit by this laser in 1402 00:49:14,960 --> 00:49:16,849 each shot gets destroyed. 1403 00:49:16,850 --> 00:49:18,199 And it's more difficult because the 1404 00:49:18,200 --> 00:49:20,419 orientation of this sample is 1405 00:49:20,420 --> 00:49:22,019 random in each shot. 1406 00:49:22,020 --> 00:49:23,959 So we need very sophisticated software to 1407 00:49:23,960 --> 00:49:26,299 calculate the 3D diffraction image 1408 00:49:26,300 --> 00:49:28,699 from this to 1409 00:49:28,700 --> 00:49:30,859 finally resolve the structure. 1410 00:49:30,860 --> 00:49:32,389 This is much more difficult than if you 1411 00:49:32,390 --> 00:49:34,069 have a crystal because they're, you know, 1412 00:49:34,070 --> 00:49:36,019 your orientation and you can rotate it in 1413 00:49:36,020 --> 00:49:37,489 a defined way. 1414 00:49:37,490 --> 00:49:39,829 But finally, each shot is 1415 00:49:39,830 --> 00:49:41,929 is you know, we need to get the data 1416 00:49:41,930 --> 00:49:42,930 from one shot. 1417 00:49:46,820 --> 00:49:49,369 OK, microphone number one, please. 1418 00:49:49,370 --> 00:49:51,469 So this is more of a technicality. 1419 00:49:53,090 --> 00:49:55,279 How is the 1420 00:49:55,280 --> 00:49:57,559 power on the electron beam dump and 1421 00:49:57,560 --> 00:49:59,239 what are you using for the electron beam 1422 00:49:59,240 --> 00:50:01,369 dump to get the 1423 00:50:01,370 --> 00:50:03,469 amount of damage along 1424 00:50:03,470 --> 00:50:05,569 with it to an acceptable level that 1425 00:50:05,570 --> 00:50:07,249 you don't destroy everything with that? 1426 00:50:07,250 --> 00:50:09,199 Yeah, that's basically the limitation of 1427 00:50:09,200 --> 00:50:11,399 this 900 kilowatts is the specification. 1428 00:50:11,400 --> 00:50:13,639 What we get finished should be heard 1429 00:50:13,640 --> 00:50:14,640 on. 1430 00:50:15,580 --> 00:50:17,019 To operate these machines, 1431 00:50:19,540 --> 00:50:21,909 we use big blocks of, 1432 00:50:21,910 --> 00:50:24,699 uh, what is it, graphene, I think, 1433 00:50:24,700 --> 00:50:26,949 and kind of rotating magnitude 1434 00:50:26,950 --> 00:50:29,109 to such that the beam doesn't 1435 00:50:29,110 --> 00:50:32,019 hit the same spot every time. 1436 00:50:32,020 --> 00:50:34,329 And but it's basically just a big, 1437 00:50:34,330 --> 00:50:36,099 big block, very long. 1438 00:50:36,100 --> 00:50:37,479 Like, how long is it? 1439 00:50:37,480 --> 00:50:39,639 Maybe eight meters, like this big. 1440 00:50:39,640 --> 00:50:41,379 And then we have several of them, which 1441 00:50:41,380 --> 00:50:43,449 can be changed and then 1442 00:50:43,450 --> 00:50:45,669 they have to put 1443 00:50:45,670 --> 00:50:48,099 away for some decades 1444 00:50:48,100 --> 00:50:49,100 to cool down. 1445 00:50:51,930 --> 00:50:52,980 I took my. 1446 00:50:59,130 --> 00:51:00,380 Microphone four, please. 1447 00:51:02,040 --> 00:51:04,439 First, thank you again for this really 1448 00:51:04,440 --> 00:51:05,440 amazing talk. 1449 00:51:07,200 --> 00:51:09,569 This is a very grizzly question, 1450 00:51:09,570 --> 00:51:11,639 but is it a it is 1451 00:51:11,640 --> 00:51:13,769 anticipated that the growth 1452 00:51:13,770 --> 00:51:16,139 in the ability of these will continue 1453 00:51:16,140 --> 00:51:18,359 to go beyond what free 1454 00:51:18,360 --> 00:51:20,129 electron lasers have achieved. 1455 00:51:20,130 --> 00:51:22,349 And is there a glimpse into what 1456 00:51:22,350 --> 00:51:24,599 the fifth generation of synchrotrons 1457 00:51:24,600 --> 00:51:25,889 would be? 1458 00:51:25,890 --> 00:51:28,379 I asked a couple of guys 1459 00:51:28,380 --> 00:51:30,959 in the scope of this talk and depending 1460 00:51:30,960 --> 00:51:33,029 on where they are, they 1461 00:51:33,030 --> 00:51:34,289 answer different things. 1462 00:51:34,290 --> 00:51:36,449 So some of them answer no. 1463 00:51:37,770 --> 00:51:39,359 It will be different techniques. 1464 00:51:39,360 --> 00:51:41,249 So free electron lasers have the unique 1465 00:51:41,250 --> 00:51:43,409 ability to make very short pulses and 1466 00:51:43,410 --> 00:51:44,799 this may be become even better. 1467 00:51:44,800 --> 00:51:46,559 So less than one femtosecond. 1468 00:51:46,560 --> 00:51:48,719 But there are other tools like electron 1469 00:51:48,720 --> 00:51:51,569 diffraction or also 1470 00:51:51,570 --> 00:51:53,699 electron microscopy, which are maybe 1471 00:51:53,700 --> 00:51:55,769 suited better for different samples. 1472 00:51:55,770 --> 00:51:57,959 But I actually I don't know what's 1473 00:51:57,960 --> 00:51:59,699 really gonna be the next step in 1474 00:51:59,700 --> 00:52:01,169 synchrotron radiation sources. 1475 00:52:02,420 --> 00:52:03,420 Thank you. 1476 00:52:04,300 --> 00:52:06,549 OK, let's be fair to the Internet, 1477 00:52:06,550 --> 00:52:07,550 is there any question? 1478 00:52:08,650 --> 00:52:11,019 Yeah, we have some more questions, 1479 00:52:11,020 --> 00:52:13,539 but is it all right? 1480 00:52:13,540 --> 00:52:15,639 Bucking Sheep is asking, how long does it 1481 00:52:15,640 --> 00:52:17,739 take to run an experiment, as 1482 00:52:17,740 --> 00:52:19,779 in riding the specter of the experiment, 1483 00:52:19,780 --> 00:52:21,489 sending the beam, collecting all the 1484 00:52:21,490 --> 00:52:23,709 images and producing, producing 1485 00:52:23,710 --> 00:52:25,869 a picture, a beam 1486 00:52:25,870 --> 00:52:26,870 time is 1487 00:52:28,360 --> 00:52:30,949 something like so unflashy 1488 00:52:30,950 --> 00:52:32,859 or other free electron laser. 1489 00:52:32,860 --> 00:52:35,019 The typical beam time slot is 1490 00:52:35,020 --> 00:52:37,119 eight hours. And so the machine runs 1491 00:52:37,120 --> 00:52:39,459 24/7. But some experiments 1492 00:52:39,460 --> 00:52:41,709 take eight, some 16, some two days. 1493 00:52:41,710 --> 00:52:43,569 But that's the order. 1494 00:52:43,570 --> 00:52:46,119 So let's say 10 hours. 1495 00:52:46,120 --> 00:52:48,699 And setting up the experiment 1496 00:52:48,700 --> 00:52:50,079 is actually the bottleneck. 1497 00:52:50,080 --> 00:52:52,639 So this can take up to one week. 1498 00:52:52,640 --> 00:52:55,359 So I don't unfortunately, 1499 00:52:55,360 --> 00:52:56,469 I don't have a picture from the 1500 00:52:56,470 --> 00:52:58,479 experiment. Experimental Hollett Flash. 1501 00:52:58,480 --> 00:53:02,109 But we have different timelines and 1502 00:53:02,110 --> 00:53:04,239 there are ten people working there to 1503 00:53:04,240 --> 00:53:05,979 build up the experiment for one week. 1504 00:53:05,980 --> 00:53:08,109 And then they have like eight hours of 1505 00:53:08,110 --> 00:53:10,269 X-ray beam and then they work half 1506 00:53:10,270 --> 00:53:12,669 a year on on reading 1507 00:53:12,670 --> 00:53:15,099 the data and then combining 1508 00:53:15,100 --> 00:53:16,100 these images. 1509 00:53:17,500 --> 00:53:19,719 So the prime time they are making 1510 00:53:19,720 --> 00:53:21,429 the images is the smallest part. 1511 00:53:23,950 --> 00:53:26,139 OK, microphone one, please. 1512 00:53:26,140 --> 00:53:27,819 Yes, thank you for the great talk as 1513 00:53:27,820 --> 00:53:28,719 well. 1514 00:53:28,720 --> 00:53:30,969 My question is, I'm sure you are aware 1515 00:53:30,970 --> 00:53:33,369 of these protein 1516 00:53:33,370 --> 00:53:35,769 folding software projects which 1517 00:53:35,770 --> 00:53:38,679 try to make these images by calculation. 1518 00:53:38,680 --> 00:53:40,809 How well do these work and 1519 00:53:40,810 --> 00:53:42,639 how much do you benefit from these 1520 00:53:42,640 --> 00:53:44,259 approaches? I mean, that's the point. 1521 00:53:44,260 --> 00:53:47,019 We don't know how well they work, right? 1522 00:53:47,020 --> 00:53:48,549 I mean, we have these simulations. 1523 00:53:48,550 --> 00:53:51,039 We can find them on YouTube and 1524 00:53:51,040 --> 00:53:52,040 they are nice. But 1525 00:53:54,340 --> 00:53:55,340 I what? You know. 1526 00:53:56,500 --> 00:53:57,500 Thank you. 1527 00:53:58,390 --> 00:54:00,219 All right. Another microphone. 1528 00:54:00,220 --> 00:54:01,149 One, please. 1529 00:54:01,150 --> 00:54:03,219 Yeah. And this was an amazing talk. 1530 00:54:04,600 --> 00:54:06,549 Can I talk a little bit more about how to 1531 00:54:06,550 --> 00:54:08,719 focus on the X-ray pulse? 1532 00:54:08,720 --> 00:54:10,809 Yes, but I don't know if I can answer 1533 00:54:10,810 --> 00:54:12,069 your question. 1534 00:54:12,070 --> 00:54:13,070 I should talk more. 1535 00:54:14,740 --> 00:54:15,939 I think for questions. 1536 00:54:15,940 --> 00:54:17,469 If you were have discussions, I think 1537 00:54:17,470 --> 00:54:20,019 that means to do that probably outside. 1538 00:54:20,020 --> 00:54:22,210 So Internet question. 1539 00:54:25,050 --> 00:54:27,329 And restricted Eve would like to know 1540 00:54:27,330 --> 00:54:29,579 if you could tell some more details about 1541 00:54:29,580 --> 00:54:31,739 how the X-ray camera manages to hold 1542 00:54:31,740 --> 00:54:33,809 so many data in such a 1543 00:54:33,810 --> 00:54:35,160 short period of time. 1544 00:54:36,360 --> 00:54:37,889 OK, to the Internet question. 1545 00:54:37,890 --> 00:54:39,179 No, I can't really. 1546 00:54:39,180 --> 00:54:41,459 I, uh, I wanted to ask 1547 00:54:41,460 --> 00:54:43,619 the guy who designed the detector 1548 00:54:43,620 --> 00:54:45,809 or who was responsible for 1549 00:54:45,810 --> 00:54:47,609 designing the detector, but he was on 1550 00:54:47,610 --> 00:54:50,309 holiday already in the last week before 1551 00:54:50,310 --> 00:54:51,780 Christmas. So I couldn't really 1552 00:54:53,550 --> 00:54:54,899 get an answer to this question. 1553 00:54:54,900 --> 00:54:56,129 I don't know it exactly. 1554 00:54:56,130 --> 00:54:58,439 I just know that there are several layers 1555 00:54:58,440 --> 00:55:00,919 and it's not, 1556 00:55:00,920 --> 00:55:03,689 uh, know, I would talk bullshit, I think, 1557 00:55:03,690 --> 00:55:05,789 but I guess 1558 00:55:05,790 --> 00:55:08,129 very soon they wanted to write a 1559 00:55:08,130 --> 00:55:10,319 big, uh, comprehensive, 1560 00:55:10,320 --> 00:55:11,320 um. 1561 00:55:12,090 --> 00:55:14,159 Some comprehensive stuff about the x ray 1562 00:55:14,160 --> 00:55:16,229 detector on on their home page of the 1563 00:55:16,230 --> 00:55:18,599 European X Files, so I would recommend 1564 00:55:18,600 --> 00:55:21,149 you to look it up there. 1565 00:55:21,150 --> 00:55:23,369 But to come to your question, 1566 00:55:23,370 --> 00:55:25,739 we do this with basically diamonds 1567 00:55:25,740 --> 00:55:28,109 or some some diamond like crystals. 1568 00:55:28,110 --> 00:55:30,659 This is an x ray mirror we have 1569 00:55:30,660 --> 00:55:32,759 and we have this 1570 00:55:32,760 --> 00:55:34,649 crazing incident angle. 1571 00:55:34,650 --> 00:55:36,959 So that's how we focus 1572 00:55:36,960 --> 00:55:38,699 these these beams. 1573 00:55:38,700 --> 00:55:39,700 And it's very 1574 00:55:42,270 --> 00:55:44,669 I don't know, it was in the news. 1575 00:55:44,670 --> 00:55:46,559 The flatness of this mirror is really 1576 00:55:46,560 --> 00:55:48,449 amazing. But I don't I don't have the 1577 00:55:48,450 --> 00:55:49,769 numbers right now. 1578 00:55:49,770 --> 00:55:51,329 But look it up. It's crazy. 1579 00:55:54,410 --> 00:55:56,570 And going microphone one, yeah, 1580 00:55:57,930 --> 00:55:59,669 this is, of course, an amazing piece of 1581 00:55:59,670 --> 00:56:01,799 hardware, but as you mentioned, 1582 00:56:01,800 --> 00:56:03,179 you showed us the controlled software. 1583 00:56:03,180 --> 00:56:05,339 It's also an amazing piece of 1584 00:56:05,340 --> 00:56:07,919 software and amount of software. 1585 00:56:07,920 --> 00:56:10,409 Can you give us some numbers on a 1586 00:56:10,410 --> 00:56:12,449 number of programs, lines of code, 1587 00:56:12,450 --> 00:56:14,939 manyas, whatever, because 1588 00:56:14,940 --> 00:56:17,629 you spent a billion in software. 1589 00:56:17,630 --> 00:56:19,709 But software is also 1590 00:56:19,710 --> 00:56:22,079 probably. Yes, yes, yes, yes, 1591 00:56:22,080 --> 00:56:23,369 for sure. 1592 00:56:23,370 --> 00:56:25,109 That would be an interesting no, no, I 1593 00:56:25,110 --> 00:56:27,569 don't have the number of lines involved 1594 00:56:27,570 --> 00:56:28,469 in this code. 1595 00:56:28,470 --> 00:56:30,719 I know that the amount of compute, uh, 1596 00:56:30,720 --> 00:56:32,849 of a CPU power 1597 00:56:32,850 --> 00:56:34,349 we need is not that much. 1598 00:56:34,350 --> 00:56:36,479 So it's more the most difficult 1599 00:56:36,480 --> 00:56:38,609 thing is to get all these channels 1600 00:56:38,610 --> 00:56:40,209 appear on our controlled system. 1601 00:56:40,210 --> 00:56:42,689 So the graphical, uh, the graphical 1602 00:56:42,690 --> 00:56:45,179 interfaces is more challenging 1603 00:56:45,180 --> 00:56:48,059 than, uh, than 1604 00:56:48,060 --> 00:56:49,319 work with the data. 1605 00:56:49,320 --> 00:56:51,199 But I really don't know how many. 1606 00:56:51,200 --> 00:56:53,849 Oh, no, I can't really tell you this, but 1607 00:56:53,850 --> 00:56:56,039 if you write me at the end 1608 00:56:56,040 --> 00:56:58,049 of the slide, I have my email address. 1609 00:56:58,050 --> 00:57:00,269 I could ask some some guy 1610 00:57:00,270 --> 00:57:01,270 said easy. 1611 00:57:05,510 --> 00:57:07,369 OK, microphone two, please. 1612 00:57:08,600 --> 00:57:10,489 I also have a question about the control 1613 00:57:10,490 --> 00:57:12,799 software. Do you have a query 1614 00:57:12,800 --> 00:57:14,929 language to find the controls you need 1615 00:57:14,930 --> 00:57:16,309 instead of having to step through all 1616 00:57:16,310 --> 00:57:17,310 those windows? 1617 00:57:18,240 --> 00:57:19,489 Oh, yes, of course. 1618 00:57:19,490 --> 00:57:21,739 Of course. But usually when you 1619 00:57:21,740 --> 00:57:23,089 when you have no clue of what you are 1620 00:57:23,090 --> 00:57:25,159 looking, it's sometimes easier if 1621 00:57:25,160 --> 00:57:27,379 you have a GUI where you can get at least 1622 00:57:27,380 --> 00:57:27,769 sorted. 1623 00:57:27,770 --> 00:57:29,329 But of course you can. 1624 00:57:29,330 --> 00:57:31,849 You can have access read and write also 1625 00:57:31,850 --> 00:57:34,039 by by by just writing 1626 00:57:34,040 --> 00:57:35,040 lines in your. 1627 00:57:38,700 --> 00:57:39,700 Internet questions. 1628 00:57:40,680 --> 00:57:42,149 No more questions. 1629 00:57:42,150 --> 00:57:43,409 OK, microphone one, please. 1630 00:57:44,530 --> 00:57:46,679 Yeah, my question is, is there 1631 00:57:46,680 --> 00:57:48,449 any policy in place 1632 00:57:49,860 --> 00:57:52,259 for, like, publishing stuff 1633 00:57:52,260 --> 00:57:54,269 like only open access or something like 1634 00:57:54,270 --> 00:57:55,270 that 1635 00:57:56,730 --> 00:57:58,859 at some risk to come 1636 00:57:58,860 --> 00:58:00,929 into your facility applying for time? 1637 00:58:00,930 --> 00:58:02,939 Do I have some policy to fulfill that? 1638 00:58:02,940 --> 00:58:05,429 Only I yeah, you have to publish. 1639 00:58:05,430 --> 00:58:07,079 I mean, you have to publish and then. 1640 00:58:08,640 --> 00:58:09,749 Is it open access? 1641 00:58:09,750 --> 00:58:11,369 That's the question. Yeah, that's a good 1642 00:58:11,370 --> 00:58:13,529 point. I think it doesn't have to be. 1643 00:58:13,530 --> 00:58:15,899 OK, so you have to make sure that 1644 00:58:15,900 --> 00:58:17,639 your results are published and. 1645 00:58:18,770 --> 00:58:20,899 But since it's not. 1646 00:58:24,230 --> 00:58:25,549 A good point. 1647 00:58:25,550 --> 00:58:27,619 I know that a private company can 1648 00:58:27,620 --> 00:58:29,449 also come and ask for prime time, but 1649 00:58:29,450 --> 00:58:31,549 they have to pay a lot of money 1650 00:58:31,550 --> 00:58:34,759 to get that. But if you are a scientific 1651 00:58:34,760 --> 00:58:37,429 researcher or university or whatever, 1652 00:58:37,430 --> 00:58:38,430 you get it for free. 1653 00:58:42,570 --> 00:58:43,570 Thanks.