WEBVTT 00:00:00.000 --> 00:00:01.000 For now, since we're sort of on time, I think we can move on to the next talk with electron identification using eco-friendly gas mixtures for the Alice 3B rich detector at LAC. 00:00:01.000 --> 00:00:03.000 Nicola, if you're ready. Yes. 00:00:03.000 --> 00:00:10.000 Yes, sir. Can you hear me? Okay. 00:00:10.000 --> 00:00:16.000 Okay, so let's start the good morning to everyone. I'm Nikola Nicasio from Dialis Collaboration. 00:00:16.000 --> 00:00:26.000 In his contribution, we'll discuss some results on our studies on performing electron identification using a possibly eco-friendly cosmixture for the future barrel-rich detector. 00:00:26.000 --> 00:00:37.000 As the LHCA. So I will provide an outline on the specification of this detector then I will fork you soon on this possibility to achieve a plus C minus identification announcement using S. 00:00:37.000 --> 00:00:43.000 And finally, I will show you some projections for the achievable reconstruction and physics performance. 00:00:43.000 --> 00:00:48.000 So this detector is in the context of this upgrade we are planning for our apparatus. 00:00:48.000 --> 00:01:01.000 So basically the main goal of our collaboration is to address the dynamics of the strongly interacting matter produced in AVI young collisions. But despite the huge planned physics program up to the end of round four, fundamental questions will remain 00:01:01.000 --> 00:01:19.000 Open question demanding for excellent vertexing tracking particle identification that now is out of reach. So demanding for a next generation experiment This is what we call the LS3 upgrade here you can see a schematic of the state of the art detector concept 00:01:19.000 --> 00:01:31.000 With only subsystems and among them for the identification of charged particles, we have a ring image in Cherenkov detector covering the This is the main focus of this presentation. 00:01:31.000 --> 00:01:56.000 So it's a proximity focusing reach using aerogel as a chunk of radiator separated from an expansion gap from a photosensitive surface equipped with silicon photomultipliers and each of these modules is part of this projective geometry With modules oriented toward the nominal collision vertex. 00:01:56.000 --> 00:02:16.000 And with the aerial sites optimize to achieve full coverage to the impinging charged particles without other labs. With such a baseline configuration, we are able to achieve some target particle identification requirements, among which the electron ion separation larger than 3 sigma 00:02:16.000 --> 00:02:30.000 Up to 2 GB. That is the baseline for this detector However, some physics observables, including Waconia delectrons, would benefit a lot from extending these electroidentification capabilities. 00:02:30.000 --> 00:02:48.000 Above 2 gv up to let's say 4 gv or more in principle and so there is a dedicated randy line on this possibility, the strategy we are planning to adopt without making any modification in the detector geometry consists in filling the rich expansion gap, the rich vessel basically 00:02:48.000 --> 00:03:04.000 With the proper gas, with the target refractive index larger than this value to perform a chunk of threshold based discrimination So the idea is that the electrons are above the threshold for chunk of emission. Actually, it rules to saturation. 00:03:04.000 --> 00:03:17.000 At a very small moment, actually. And the muons and the weird particles are below this threshold so that the identification of clusters from gas can be used to univocally label electrons. 00:03:17.000 --> 00:03:22.000 In this context, the refractive index is crucial for the mixture we plan to use. 00:03:22.000 --> 00:03:40.000 Because a low index like this one is good because means high emission threshold for every particle species but at the same time a limited photon yield and cluster size and from dedicated simulation studies, we see that the optimal reflective index we plan to achieve 00:03:40.000 --> 00:04:05.000 Is of the order of 1.0006. So how do we get it? Well, first of all, in order to address a proper mixture we have to rule out some options such as saturated fluorocarbons like CF4, C4F10 that are commonly used in currently ring imaging Sharkov systems, but very large global warming potential 00:04:05.000 --> 00:04:10.000 Thousands of times the one of the CO2, and the production is also being discontinued. 00:04:10.000 --> 00:04:28.000 So they must certainly be avoided. And so how do we get the target index Well, as a backup option, we see that the lower possible index is achieved using CO2, but to get higher indices we need to employ mixtures involving heavy gases 00:04:28.000 --> 00:04:36.000 And the option, the most promising option in terms of limited global warming potential consists in blinding. 00:04:36.000 --> 00:04:54.000 Low moral concentration. Fluorocetones like gases like this c5 of 10 or with the light gases like n2 or co2 So to get the required optical performance but with the global warming potential consistent or even smaller than the one of CO2. 00:04:54.000 --> 00:05:07.000 Making the mixture of course means monitoring the molar concentrations of these components but this is something done already in the past, for instance in the LST detector for the chart uncovering imaging. 00:05:07.000 --> 00:05:18.000 Detector using basically measuring the speed of sound in the mixture to monitor and adjust on the run the concentration of the components. 00:05:18.000 --> 00:05:32.000 But a criticality I would like to mention is the boiling point of gases like this C5F10 or about 27 degrees compared with the CPM operation temperature that is of minus 40 degrees. 00:05:32.000 --> 00:05:41.000 So it is maybe not the final choice and further studies on the optimization of this measure and Identifying the proper gas are of course still ongoing. 00:05:41.000 --> 00:05:47.000 Nevertheless, in the following would I like to show you some projections on the expected physics performance. 00:05:47.000 --> 00:05:52.000 And to do so, I will assume the optical properties of this specific mixtures. 00:05:52.000 --> 00:06:11.000 So starting from the features of the cluster, we plan to observe, here you can see a simulation for electrons and basically biomes that whereas a function of true rapidity so to get the full acceptance of the barn rich detector and counting the mean number of fired cpms 00:06:11.000 --> 00:06:19.000 In the two cases. So you can see that we have a pretty good signal strength with approximately 10 third CPMs per electron in a cluster. 00:06:19.000 --> 00:06:35.000 Is assuming the baseline photon detection efficiency of the CPMs. Then you see these structures corresponding one to each of the sectors of the apparatus this mean number follows the thickness of the expansion gap we get in each of the projective sector of the system. Of course, the larger the gap 00:06:35.000 --> 00:06:39.000 The larger the number of emitted photons and fired the CPMs. 00:06:39.000 --> 00:06:44.000 And we just have some minor losses close to the sector boundaries, the points you see here. 00:06:44.000 --> 00:06:48.000 That's because in this region, part of the cluster is simply lost. 00:06:48.000 --> 00:06:53.000 And to give a more detailed focus on the geometry of these clusters. 00:06:53.000 --> 00:07:08.000 First of all, one thing to take into account is that our system is embedded in a two Tesla superconducting solenoid And this affects the shape of the cluster in our, and so there is an impact on the reconstruction algorithm we have to implement 00:07:08.000 --> 00:07:24.000 Then as you can see for instance at 3GB, we have clusters with a radius of the order of eight millimeter around the expected impact point with the CPMs featuring more than one photo electron In the first enabled CPMs to the impact point. 00:07:24.000 --> 00:07:33.000 And single photoelectron signals in the nearby SAS. Of course, without a canting, crosstalks and CPM itself. 00:07:33.000 --> 00:07:49.000 On the right you can see a map for the expected fraction of events with the minimum number of third CPMs And the minimum number of total required photoelectrons convoluted in the full detector acceptance that is then crucial to define a proper strategy 00:07:49.000 --> 00:07:56.000 For a cluster reconstruction and an identification and to evaluate the impact on the physics performance in the end. 00:07:56.000 --> 00:08:16.000 So here you can see a projection on the identification performance in physics events In particular, proton proton, and then I will show lead lead collisions The main challenge in this context is rejecting the plus c minus like signal Due to the random clustering of background heats, where we have a pyon to electron ratio that is 00:08:16.000 --> 00:08:20.000 Of the order of hundreds. So we need a good rejection. 00:08:20.000 --> 00:08:25.000 For this study, I developed a dedicated machine learning algorithm based on extreme gradient posting. 00:08:25.000 --> 00:08:36.000 Taking as a feature to track parameters as well as the number of third CPMs and the photoelectrons in four regions in the middle angle relative to the impinging track direction. 00:08:36.000 --> 00:08:43.000 In the plot below, you can see the resulting purity and efficiency, in particular the red dots for electrons, is a function of momentum. 00:08:43.000 --> 00:08:54.000 And you can see that even counting the CPM and dark count rate, we're able to ensure excellent identification performance Up to them you want to buy on threshold in the gas. 00:08:54.000 --> 00:09:12.000 You see these results look very promising, but are we happy about them Actually, no, this is not the end of the story because in, as I said, the main target of the strategics is our lead-led collisions. And in most central collisions, we have a track multiplicity and it multiplicity is large 00:09:12.000 --> 00:09:23.000 Of course, the reconstruction is much more complex. Contamination is very large, so this kind of reconstruction techniques has to be optimized even further. 00:09:23.000 --> 00:09:28.000 But as you can see, using this approach, even in these more complex situation. 00:09:28.000 --> 00:09:35.000 We're able to ensure reconstruction purity and efficiency larger than 90% using this gas mixture. 00:09:35.000 --> 00:09:50.000 In the whole momentum region of interest. So this is pretty good, but which is the impact on the physics Well, here you can see, for instance, one of the cases we mentioned starting from our letter of intent dating back in 2022. 00:09:50.000 --> 00:10:08.000 Where basically we want to prove the same evolution of the medium produced in the collision of nuclei, the curcum plasma And this is done by reconstructing the electron pairs as a function of reconstructed invariant mass and transverse momentum. 00:10:08.000 --> 00:10:20.000 So we need to achieve a result like this. This, of course, a simulation where we need to achieve excellent coverage in terms of MEE and PTE. 00:10:20.000 --> 00:10:43.000 Exceeding the 3GB level. So this is considering only the baseline configuration, so using only the original where no gas is included So what did it was then evaluated the impact on the electron reconstruction efficiency Combining the gas option to the other identification subsystems foreseen for the LS3 apparatus. 00:10:43.000 --> 00:11:13.000 And as you can see, while in the baseline configuration. Reconstruction is limited to the region close to 3 gb including these threshold-based identification Using us, we're able to extend the identification range at 5GB and even higher You can better appreciate this from the one-dimensional projection of previous plots for selected electron pair transverse momentum values. 00:11:13.000 --> 00:11:30.000 From one to four gv where as you can see, not only we recovered the region at large the electron transverse momentum but they were able to also to suppress background in the region and of course gaining efficiency in the region at small me 00:11:30.000 --> 00:11:35.000 Simply because we have a cross check with us of the information on the aerogel. 00:11:35.000 --> 00:11:47.000 And so this is a double variable. And in particular, as you can see from these plots, all these results show very good prospects for the whole at least three-day electron physics program. 00:11:47.000 --> 00:12:11.000 So coming to the conclusion, I've shown how filling the barrel reach expansion gap with a proper gas mixture is very promising to announce the electrons plus minus identification capabilities consistently with the requirements for the LS3 physics This can be done using an eco-friendly gas mixture by mixing 00:12:11.000 --> 00:12:35.000 Heavy gases with the likely fluoro ketones like with the light gas latter gas is like N2 and co2 with the limited global warming potential. And in the end, I've shown how the stable identification performance received by keeping the same baseline geometry, the same silicon photo multiplier technology 00:12:35.000 --> 00:12:40.000 And it's from proton-proton collisions. To the most challenging environment of central LED. 00:12:40.000 --> 00:12:48.000 Our next steps in this respect are first of all more detailed studies on the final gas mixtures to be used. 00:12:48.000 --> 00:12:57.000 If this kind of approach becomes a thing. Because, of course, not only it has to provide the required optical properties in terms of reflective index transmitters and so on. 00:12:57.000 --> 00:13:18.000 But also have features select a volume point consistent with the expected barrel reach CPM operation temperature And finally, we are planning to perform the direct measurement on BIM In July and September, testing of various classes and again with the aim of finding a proper option, best option or at least a compromise 00:13:18.000 --> 00:13:24.000 To get the required physics performance consistently with the LS3 physics program. 00:13:24.000 --> 00:13:30.000 So this concludes my presentation. Thank you for your attention. 00:13:30.000 --> 00:13:35.000 Thank you so much, Nicola. Any questions? 00:13:35.000 --> 00:13:39.000 Duane, I can get what was the first half. 00:13:39.000 --> 00:13:58.000 A very interesting talk. Perhaps I missed something i wasn't actually sort of understanding So the airbridge and the gas are different the gas that you tested most of it has an effective index of 1.006. 00:13:58.000 --> 00:14:09.000 What is the refractive index equivalent of the aerogel and what's it made it out of Is it much better? Is it submitting to the gas? I didn't quite understand that. 00:14:09.000 --> 00:14:23.000 Sorry, I didn't get the question. I mean, the average index, as I mentioned here It's in 1.006. This is the average weighted considering the whole molar concentration of the components. 00:14:23.000 --> 00:14:28.000 In this case, for instance, obtained with the 1.003. 00:14:28.000 --> 00:14:33.000 Of the M2 combined with what is about 1.00. 00:14:33.000 --> 00:14:39.000 16550 of the… c5 f10, O. 00:14:39.000 --> 00:14:42.000 So 1.006 is the index of the average in the end. 00:14:42.000 --> 00:14:43.000 Yes, no, my point here is, sorry, just to the conversation, because you said about Everbogel. 00:14:43.000 --> 00:14:48.000 I don't know if… 00:14:48.000 --> 00:15:02.000 And I don't know where i don't know. Where this is here, because this is just the gas mixture, right? So with your airbogel what is the refractive index of that or is that made of the same gases. 00:15:02.000 --> 00:15:13.000 Sorry, I really missed the question. 00:15:13.000 --> 00:15:14.000 Oh, hi. 00:15:14.000 --> 00:15:19.000 On slide 13. You have using aerogel only and Aerogel plus gas We've talked about the gas, but I was wanting to know a bit more about the air gel. I don't necessarily know if it was about what the properties of that was or what it was made out of. 00:15:19.000 --> 00:15:25.000 The problem is that the aerogel is a reflective index of 1.03. 00:15:25.000 --> 00:15:38.000 So it would allow to separate considering this geometry this specification to identify the electrons the single electron at momentum value larger than 2 GB. 00:15:38.000 --> 00:15:48.000 Then, of course, you see values larger because then we have larger the electron in various mass combining both the electron information And what the gas does is simply to enhance this region. 00:15:48.000 --> 00:16:00.000 Covering not only the region that is already covered by that then of course is crucial for Adrance and other species, but extending the identification capabilities up to 4GB and more. 00:16:00.000 --> 00:16:20.000 So this is the gain we have with the gas. So basically adding this 1.0006 to do one of the .03 that is The baseline for the original, and that is of course fixed because it is uh was optimized for studies related to the identification of add-ons. 00:16:20.000 --> 00:16:21.000 Thank you. I was just missing the reflective index of the average. 00:16:21.000 --> 00:16:23.000 That is crucial again for illustrate. 00:16:23.000 --> 00:16:27.000 Okay, okay. 00:16:27.000 --> 00:16:33.000 Thank you so much, Jen. Nikolom, I think we can move on in the interest of time.