1 00:00:02,129 --> 00:00:04,049 And hear my voice Yes. 2 00:00:05,250 --> 00:00:11,640 Okay, thanks. Oh, I will be talking about the uhm particle identification using silicon em 3 00:00:11,700 --> 00:00:12,840 timing detectors. 4 00:00:15,120 --> 00:00:20,760 I'm talking on behalf of the Alice collaboration. So the precision timing has 5 00:00:20,760 --> 00:00:24,810 always been a prominent topic in the instrumentation for high energy physics 6 00:00:24,810 --> 00:00:27,150 experiment both for applications ehm 7 00:00:28,290 --> 00:00:33,720 to trigger detectors, but also for for particle identification via time of flight 8 00:00:34,290 --> 00:00:40,410 detectors. And in the last two decades, resistive clay chambers have been widely 9 00:00:40,410 --> 00:00:46,740 used for applications of large area detector, kind of replacing the scintillators 10 00:00:46,740 --> 00:00:52,500 for this type of use. But recently, the silicon sensors have become very popular 11 00:00:52,500 --> 00:00:57,750 for timing and notably for pile-up rejection at the high lumi LHC and there 12 00:00:57,750 --> 00:00:59,970 is also rapid progress for consumer application. 13 00:01:00,359 --> 00:01:05,699 In the field of imaging the leader are 3D scanners. So let me introduce the Alice 14 00:01:05,699 --> 00:01:11,009 experiment which is dedicated to heavy ion physics at the LHC. It's currently being 15 00:01:11,009 --> 00:01:16,619 upgraded for Run three and run four and is designed on purpose for heavy ion physics 16 00:01:16,619 --> 00:01:22,439 such to cope with high multiplicity be able to tracking down to very low pT. And 17 00:01:22,499 --> 00:01:27,029 particle identification identification is an essential feature it uses all the known 18 00:01:27,029 --> 00:01:32,519 PID technique, but in particular among the PID detectors in the central barrel, 19 00:01:32,549 --> 00:01:34,649 there is a large time of flight array. 20 00:01:35,880 --> 00:01:40,770 So, the time of fly detector of Alice is designed for the identification of 21 00:01:40,770 --> 00:01:50,220 hadrons up to five GeV in pT. It consists of multigap resistive, multigap RPC 22 00:01:50,220 --> 00:01:56,130 strip detectors that provide excellent timing and performance with better than 50 23 00:01:56,130 --> 00:01:59,610 picoseconds time resolution measured in beam tests. 24 00:02:00,510 --> 00:02:06,120 Better than 60 Pico seconds deliver for physics. Now the detector is doing physics 25 00:02:06,120 --> 00:02:11,100 at the LHC since 10 years with impact on a large and diverse set of physics 26 00:02:11,100 --> 00:02:16,200 measurements and will be running for the next 10 years at higher rates in tun 27 00:02:16,200 --> 00:02:17,280 three and run four. 28 00:02:18,570 --> 00:02:25,080 I said the ALICE experiment is being upgraded for operation in run three and 29 00:02:25,080 --> 00:02:32,760 run four with luminosity is a factor of ten a hundred larger than then so far. And eventually 30 00:02:32,760 --> 00:02:40,200 this will be the maximum rate that TPC based the experiment can can operate. And 31 00:02:40,230 --> 00:02:47,490 in fact, if if we want to do another step in luminosity, the TPC should be abandoned 32 00:02:47,490 --> 00:02:51,810 and a new technology should should be used. And for this reason, it has been 33 00:02:51,810 --> 00:02:52,620 proposed 34 00:02:53,670 --> 00:02:59,340 to have a completely new detector to be installed in the long shut down for these 35 00:02:59,340 --> 00:02:59,970 proposals. 36 00:03:01,050 --> 00:03:05,970 is discussed in this paper that has been submitted as input to the European 37 00:03:05,970 --> 00:03:11,700 strategy. It would consist of full silicone detector with very high spatial 38 00:03:11,700 --> 00:03:16,890 and time resolution and that will enable to fulfill a rich heavier in physics 39 00:03:16,890 --> 00:03:23,730 program in the 2030. Going from studies of heavy flavors and quarkonia, low mass 40 00:03:23,730 --> 00:03:29,910 dileptons, soft and ultra soft photons and so on and so forth. Now, of course, 41 00:03:29,940 --> 00:03:36,480 those ambitious physics goals are the main drivers for the requirements for such a 42 00:03:36,480 --> 00:03:41,940 detector that must be low in material budget, it should be fast and high 43 00:03:41,940 --> 00:03:42,960 precision as well. 44 00:03:44,640 --> 00:03:48,570 Now, you see in this slide the concept of the proposed em, 45 00:03:50,010 --> 00:03:56,190 detector for after LS4 it consists of a barrel plus and endcap tracking with 46 00:03:56,190 --> 00:03:59,730 approximately 10 layers and has 47 00:04:00,000 --> 00:04:01,140 a time of flight 48 00:04:02,159 --> 00:04:07,259 silicone detector that will provide hydron identification and electronic 49 00:04:07,259 --> 00:04:14,069 identification at low pT. So it's a compact layout detector features very low mass 50 00:04:14,099 --> 00:04:20,039 tracking layers, I mean down to zero point zero point five porcent of radiation length for the vertex in layers. 51 00:04:20,069 --> 00:04:26,549 This thanks to the current R&D for the ITS three upgrade or at least it would be a 52 00:04:26,549 --> 00:04:31,919 large accessed acceptance down to very low pT. And it will be capable to reach very 53 00:04:31,919 --> 00:04:37,829 high luminosity is up to 10 to the 34, nuclear nuclear average luminosity. 54 00:04:38,940 --> 00:04:44,400 So the performance is excellent with high resolution board for position and for time 55 00:04:44,400 --> 00:04:49,470 of flight. So the expectation is to have a time of flight layer with 20 picosecond 56 00:04:49,500 --> 00:04:49,800 time resolution. 57 00:04:51,540 --> 00:04:57,210 Now the technology for doing timing in the time of flight layers are multiple if you 58 00:04:57,210 --> 00:04:59,850 want to. So here's a list of some of them. 59 00:05:00,000 --> 00:05:06,390 LGAD are devices that have been developed for high energy physics and they are going 60 00:05:06,390 --> 00:05:11,640 to be part of the LS3 upgrades of ATLAS and CMS. The good news is that 61 00:05:11,640 --> 00:05:16,290 already now, they have been proved to be able to, em, 62 00:05:18,270 --> 00:05:23,670 provide a very good time resolution 20 or 30 Picoseconds with radiation loads 63 00:05:25,020 --> 00:05:32,370 pretty high up to 10 to the 30, 10 to the 14 neutron equivalent. SPAD are also another 64 00:05:32,370 --> 00:05:34,020 possibility for timing. 65 00:05:35,400 --> 00:05:40,200 There is a bit of a limitation because of the fact that you have to quench them and 66 00:05:40,200 --> 00:05:44,100 this will reduce the fill factor which has to be 67 00:05:45,450 --> 00:05:51,240 taken care in a different way. They are also high dark count rates devices, but as a 68 00:05:51,240 --> 00:05:56,130 matter of fact, that might be options to lower the gain and operate them in a more 69 00:05:56,130 --> 00:05:59,940 that would not be, let's say sensitive to single photons anymore. 70 00:06:00,000 --> 00:06:06,690 still be okay for MIPS. The problem SPAD is that they have been produced in quite 71 00:06:06,690 --> 00:06:12,300 some time in the industrialized facilities and therefore a possibility to have them 72 00:06:12,300 --> 00:06:15,900 integrated in a monolithic sensor is also quite attractive. 73 00:06:17,100 --> 00:06:21,750 There are also timing sensors without gain. An example is this 74 00:06:22,860 --> 00:06:28,140 monolithic pixel prototype that has been developed by the University of Geneva that 75 00:06:28,140 --> 00:06:32,850 has reached a time resolution of 50 picoseconds. And now the team is exploring 76 00:06:32,850 --> 00:06:38,220 the path doors the one picosecond approximately device adding gain to 77 00:06:38,220 --> 00:06:40,650 this sensor. Now, one point to be 78 00:06:42,660 --> 00:06:48,450 to be added here is that the fact that we expect to have a lower radiation load than 79 00:06:48,480 --> 00:06:54,720 ATLAS and CMS opens some new possibilities such that we could explore new technology 80 00:06:54,720 --> 00:06:58,260 options for timing and therefore perform a complimentary R&D. 81 00:07:00,000 --> 00:07:02,430 From the physics point of view, there are 82 00:07:03,570 --> 00:07:07,110 many points where the TOF will will provide the em 83 00:07:09,900 --> 00:07:14,280 extremely good information just to list a few the measurement of low mass dileptons 84 00:07:14,280 --> 00:07:19,260 in the plus and minus invariant mass, which is where we need high purity 85 00:07:19,260 --> 00:07:24,090 for electron ID at low momentum or the measurement of multi heavy flavored barions 86 00:07:24,090 --> 00:07:29,820 or mesons, many of them have not been seen notably, the Omega Triple C is a 87 00:07:29,820 --> 00:07:34,440 state that is expected to be hugely enhancement in nucleus nucleus collisions. 88 00:07:34,440 --> 00:07:40,320 So giving rise to spectacular phenomena. And in this case, PID is of course, very 89 00:07:40,320 --> 00:07:45,000 useful to enhance the significance in the in the environment massively construction 90 00:07:45,000 --> 00:07:51,360 of these things. Many more measurements from nuclear exotic hyper nuclear can be 91 00:07:51,360 --> 00:07:56,130 performed and down to very ultra low pT in the measurement of coherent pion 92 00:07:56,130 --> 00:07:59,730 production would benefit from this time of flight system. 93 00:08:00,810 --> 00:08:04,770 Now I'm going to show you an example study of the performance of the time of fligth 94 00:08:04,770 --> 00:08:10,680 detector that it has to be taken just as initial study to guide future 95 00:08:10,680 --> 00:08:15,840 optimizations with the detector towards the achievement of the requirements for 96 00:08:15,840 --> 00:08:16,680 for physics 97 00:08:17,760 --> 00:08:23,520 one start from an analytical estimate of the performance you can plot the 98 00:08:23,550 --> 00:08:29,070 separation power as a function of transverse momentum for electrons, kaons 99 00:08:29,340 --> 00:08:36,450 and protons in an approximation that uses straight tracks and an idea of tracking 100 00:08:36,450 --> 00:08:42,240 performance. And and and this plus here, sure your what would be the expected 101 00:08:42,240 --> 00:08:48,030 performance of a 20 picoseconds time of flight they are at at one meter from the 102 00:08:48,030 --> 00:08:53,460 beam pipe. When you can you can read from the plus the number for simplicity they 103 00:08:53,460 --> 00:08:58,110 are written here, this system would give you a three sigma separation for the 104 00:08:58,110 --> 00:08:59,970 electrons and pions up to 750 GeV 105 00:09:00,000 --> 00:09:06,990 K to pi up to 2.5, GeV and proton to K up to about four GeV oversee. 106 00:09:08,130 --> 00:09:13,500 Now, to have a more realistic estimate of the performance of the full system one has 107 00:09:13,500 --> 00:09:19,170 to follow the in also the performance of the tracker. So, what we have done was 108 00:09:19,170 --> 00:09:21,090 to estimate the, em, 109 00:09:22,320 --> 00:09:27,870 the efficiency of momentum resolution of the tracker via studies with a fast MonteCarlo 110 00:09:27,870 --> 00:09:33,600 tool. This take into account the detector configuration material effects and the 111 00:09:33,600 --> 00:09:39,030 occupancy of the detector and therefore provides a reliab a reliable estimate of the 112 00:09:39,030 --> 00:09:45,030 performance, you see the plots here on the left hand side, the efficiency on the on 113 00:09:45,030 --> 00:09:50,220 the top part for electrons, pions and protons as a function of pT. And in the 114 00:09:50,220 --> 00:09:55,830 lower part, the momentum resolution for electron, pions and protons as well. So 115 00:09:55,830 --> 00:09:59,970 these are examples for magnetic field configuration of point two Tesla, 116 00:10:00,000 --> 00:10:09,000 which is what would enable the measurement of particles down to low pT. And em as a 117 00:10:09,000 --> 00:10:13,260 matter of fact, with this configuration electrons could be measured down to about 118 00:10:13,260 --> 00:10:14,460 50 MeV over c. 119 00:10:15,690 --> 00:10:21,630 So, these curves have been used for input as an input for the simulations of the 120 00:10:21,630 --> 00:10:26,910 time of flight said to get a more realistic performance estimate of the 121 00:10:26,940 --> 00:10:28,260 particular identification 122 00:10:29,760 --> 00:10:35,760 and, and they have been fed to the Delphes fast simulation tool for the following results. 123 00:10:37,140 --> 00:10:40,770 Now, if you look at the performance of a time of flight detector, you might 124 00:10:40,800 --> 00:10:47,100 generally plot the measured velocity as a function of the particle momentum which is 125 00:10:47,100 --> 00:10:52,290 on the x axis on this plot. And there you can see the separation power and the and the 126 00:10:52,320 --> 00:10:58,860 different particle IDs that show up quite nicely. Now, the truth is that the TOF by 127 00:10:58,860 --> 00:11:00,000 itself is an a, it's a layer 128 00:11:00,000 --> 00:11:04,620 that measures only the time of arrival of the particle. And if 129 00:11:04,620 --> 00:11:08,700 you want to have a performance as good as the one shown in this plot, you have to 130 00:11:08,700 --> 00:11:14,550 know the start time of the particle to be able to measure the time of flight. And in 131 00:11:14,550 --> 00:11:19,980 an environment where the collision time has a jitter of about three, 300 picoseconds, 132 00:11:20,250 --> 00:11:26,610 you have to find a solution. But the good thing is that the TOF detector itself can 133 00:11:26,610 --> 00:11:31,590 measure the event time. This can be done with a combinatorial algorithm, you just 134 00:11:31,590 --> 00:11:36,450 need the two tracks. And eventually you measure the event time resolution, which 135 00:11:36,450 --> 00:11:41,430 scales as one over the square root of the number of tracks that enter into your 136 00:11:41,460 --> 00:11:45,900 algorithm. And you see in this plot the distribution foreseen and seen on 137 00:11:45,900 --> 00:11:51,810 collision of the event time resolution that has a mean better than four 138 00:11:51,810 --> 00:11:56,790 picoseconds, that means being negligible, compared to the 20 picoseconds of the 139 00:11:56,790 --> 00:12:00,000 timing resolution of the array itself, and with a tiny fraction 140 00:12:00,000 --> 00:12:07,920 of events below point five porcent, where the event time cannot be measured, 141 00:12:07,980 --> 00:12:11,430 because only one track has reached the time of flight. 142 00:12:13,860 --> 00:12:20,790 Now, eventually, if you want to quantify the particle identification performance a 143 00:12:20,790 --> 00:12:21,780 bit more, em 144 00:12:23,400 --> 00:12:27,810 you want to try to see what is the efficiency of the, em 145 00:12:29,370 --> 00:12:34,800 detector system and the purity in the in the identification. So, this is shown in this 146 00:12:34,800 --> 00:12:41,490 plot is the blue curves, you see the efficiency for the electrons, kaons and 147 00:12:41,490 --> 00:12:44,520 protons they are primary particles, em, 148 00:12:47,190 --> 00:12:52,860 and in and in red the purity for the particle identification. Now, this plot 149 00:12:52,860 --> 00:12:58,680 here and this study include all the detector effects: start time, resolution, 150 00:12:59,040 --> 00:13:00,000 time resolution, tracking, decays, 151 00:13:00,000 --> 00:13:06,690 everything, and the particle identification strategy which is being 152 00:13:06,690 --> 00:13:12,480 applied here to estimate the purity of the sample is a very simple three sigma cut selection. 153 00:13:12,480 --> 00:13:18,960 So, something that would allow you to have 100% roughly particle 154 00:13:18,960 --> 00:13:25,680 identification efficiency. And you can see that for instance in the electron case, 155 00:13:25,860 --> 00:13:32,040 electron identification is very clean better than 90% up to pT or 500 MeV over c 156 00:13:32,040 --> 00:13:37,470 and the coincide if you want it in very nicely agreement with what you would 157 00:13:37,470 --> 00:13:45,150 expect from the Six Sigma separation from the analytical calculation, you see the 158 00:13:45,150 --> 00:13:52,410 same case also for the kaons and and the protons in this case. So, this Yeah. 159 00:13:52,410 --> 00:13:57,240 Just two minutes left. Yeah, I'm almost done. 160 00:13:57,240 --> 00:14:00,000 So, this system would give you a clean particle identification better than with a better 161 00:14:00,000 --> 00:14:07,710 than 90% purity up to 500 MeV for electrons 1.9 were about two GeV four kaons 162 00:14:07,710 --> 00:14:09,630 and three GeV for protons. 163 00:14:10,889 --> 00:14:17,039 Now this brings me to my summary. There is significant interest in precision in 164 00:14:17,039 --> 00:14:22,769 precision timing with silicone sensors. This is of quickly evolving field. And 165 00:14:22,769 --> 00:14:26,579 there are several interesting developments in in 166 00:14:28,590 --> 00:14:34,320 ongoing that are promising and that the timing capabilities of silicone are 167 00:14:34,320 --> 00:14:36,120 still to be fully exploited. 168 00:14:37,140 --> 00:14:42,090 I've shown you the proposal for our next generation LHC experiment for heavy 169 00:14:42,090 --> 00:14:47,340 ion physics. This is the detector conceived for running at about a factor of 170 00:14:47,340 --> 00:14:52,830 50 higher luminosity than they currently upgraded the ALICE experiment and 171 00:14:52,830 --> 00:14:58,620 therefore will enable a rich physics program in 2030s. This detector 172 00:14:58,620 --> 00:15:00,000 will be equipped with a time flight layer 173 00:15:00,000 --> 00:15:06,270 for particle identification made of silicon sensors. The time 174 00:15:06,270 --> 00:15:11,130 resolution is expected to be of the 20 picosecond level, which is something that 175 00:15:11,130 --> 00:15:17,010 is already achieved so far for MIPS. The point is that given the fact that the flux 176 00:15:17,010 --> 00:15:22,920 and the radiation is different, different requirements with respect to ATLAS and CMS 177 00:15:22,950 --> 00:15:27,780 apply for this case, and therefore, this opens the possibility to explore new 178 00:15:27,780 --> 00:15:33,990 technologies, like the use of SPAD or monolithic sensors. Thank you very much 179 00:15:33,990 --> 00:15:34,560 for your attention. 180 00:15:36,270 --> 00:15:37,140 Thank you, Roberto. 181 00:15:38,190 --> 00:15:39,360 Do we have any questions 182 00:15:49,620 --> 00:15:50,310 so I 183 00:15:51,720 --> 00:15:54,510 don't see any raised hands. 184 00:16:04,110 --> 00:16:11,940 So, I have a question just from a technology like feasibility standpoint, 185 00:16:12,210 --> 00:16:17,310 which do you think of these newer technologies is more feasible? Let say. 186 00:16:19,050 --> 00:16:25,530 Well, I cannot I cannot really answer now, I must say, well, for sure, LGADs are 187 00:16:25,530 --> 00:16:31,230 kind of already proven as a as a technology for for a timing layer with 188 00:16:31,470 --> 00:16:38,160 with these capabilities. But I I believe that there is quite a good let's say, 189 00:16:38,700 --> 00:16:45,420 perspectives in in the use of SPAD, these are devices that can be in principle quite 190 00:16:45,420 --> 00:16:47,220 nicely coupled, in a, 191 00:16:48,780 --> 00:16:56,310 in a sensor that will also embed the circuitry, using CMOS technology. 192 00:16:58,500 --> 00:16:59,910 So I would say that, 193 00:17:00,450 --> 00:17:08,910 the the fact that one can move to perform a new R&D that can be complimentary with 194 00:17:08,910 --> 00:17:14,370 respect to the ones that have been carried out for Atlas and CMS timing layers, will 195 00:17:14,370 --> 00:17:21,630 be for sure something to try to to to achieve in order to you know, move forward the field 196 00:17:21,690 --> 00:17:29,220 like the the the field of timing, precision timing with silicone even further, as I 197 00:17:29,220 --> 00:17:35,160 was also discussing, there are these these developments also on monolithic sensors 198 00:17:35,160 --> 00:17:42,540 that are quite, quite promising. And these type of devices, if they are eventually 199 00:17:42,540 --> 00:17:47,610 prove, if you can prove that these devices are really going to be, 200 00:17:49,590 --> 00:17:51,510 you know, capable to give you 201 00:17:53,310 --> 00:17:59,640 picoseconds level high resolution that could be exploited for physics in the 202 00:17:59,640 --> 00:18:00,000 future experiment, that 203 00:18:00,000 --> 00:18:05,220 that is definitely going to be the the most promising one that 204 00:18:05,220 --> 00:18:10,230 would be a game changer I mean a factor of 10 in time resolution would be really ...