Study on the discovery of new heavy, neutral gauge bosons with the ATLAS experiment

Abstract

In this thesis, a search for a new, heavy, neutral gauge boson, namely the Z′ boson withStandard Model like couplings, is presented. The study has been performed using the LHC p-p collision data, collected with the ATLAS experiment, at a center of mass energy of √ s = 7 TeV, covering two different time periods: March-November 2010 and March-June 2011. The total integrated luminosity for these periods was of the order of 42 pb−1 and 1.2 fb−1 correspondingly.After a brief theoretical introduction of the models predicting such a new boson and a description of its properties and characteristics, we essentially proceeded to the analysis, by, first of all, presenting analytically the methods of reconstruction of the muon software and the system performance and giving all the details regarding the study methods that were used in this work (reconstruction algorithms, software tools, data quality control, etc.). The Z′ boson is expected to appear as a resonance in the high mass region of the Drell-Yan process (pp → ℓ+ℓ−, where ℓ = e, µ and τ ). The production mechanism chosen indicates that Z′, in analogy with the Standard Model Z boson, is electrically neutral, colorless and self-adjoint. In the present study, we focused on the muonic decay of the new boson (Z′ → µ+µ−), therefore the final state will be characterized by the presence of two, oppositely charged muons. Any Standard Model procedure with the same signature can be a possible background source for the Z′ signal. The dominant and irreducible background is the Drell-Yan process of the Standard Model into muons Z/γ∗ → µµ. Obviously, we are mainly interested in the high mass region, where the rapid decrease of the cross-section leads to small statistics. For this reason, apart from an inclusive sample, several binned samples, generated in different large mass bins, were used, in order to enlarge the statistics. Other procedures that can also constitute background processes to our signal and were examined are the tt¯→ µ+µ−, W + jets (where a jet may be identified incorrectly as a muon), the di-boson production (WW, WZ and ZZ) and finally the QCD processes b¯b → µ+µ− and cc¯ → µ+µ−. The contribution of cosmic particles as another possible source of background is also discussed. The first element that was examined in our analysis were the pile-up conditions. Since we can not know in advance the content of pile-up events in our data, so as to appropriately adjust our Monte Carlo, this has to be estimated a posteriori. The method adopted was to measure the number of primary vertices in the real data and apply a weight on every Monte Carlo event versus its number of vertices “n”, that is extracted by the ratio of the total number of the Data events with n vertices over the total number of Monte Carlo events with n vertices: • MCp.u.W.(n) = Data events with n vrts. / MC events with n vrts. Not the whole amount of each data sample are appropriate for physics’ studies. Before anything else, we had to choose, for each data sample, this part that was taken with all the sub-detectors up on High Voltage. This was done by applying the Good Run Lists (GRLs), a dedicated list that determines the “good” Luminosity Blocks for every data sample.After that, we proceeded to apply the appropriate selection criteria of our analysis, on both data and Monte Carlo samples. These are: 1. Trigger: For the data, we have used different muon triggers, with a muon momentum that increased in time, due to the constantly increasing luminosity (PT varied from 10 to 13 GeV). For the Monte Carlo, we kept one common trigger, of10 GeV PT 2. Primary Vertex: In order to ensure that the particular events that we tested are,indeed, the product of a real collision, we requested the presence of one Primary Vertex, displaced with respect to the interaction point 20 cm maximum, and with at least three different tracks associated to it 3. Two Combined Muons: We requested at least two muons with both Inner Detector and Muon Spectrometer information 4. PT : We only kept muons with a transverse momentum above 25 GeV 5. η: In order for the muon to fall within the 2010 trigger acceptance, we restricted the muon pseudorapidity up to 2.4 6. Inner Detector hits: In order to ensure the quality of the track reconstruction, we requested a minimum number of hits of the muon in the several parts of the Tracker 7. Muon Spectrometer hits: Ensured quality of the reconstruction in the Spectrometer 8. Impact Parameters: We used upper limits on the impact parameters of the muons(z0 and d0), to reject muons coming from cosmic rays -which are expected to appear displaced from the Primary Vertex 9. Isolation: We imposed a limit on the activity (i.e., additional tracks) around the muons, to reject those within jets 10. Opposite Charge: We requested that the two most energetic muons of each event should possess opposite charges 11. Dimuon Mass: Finally, we examined those events that had a dimuon mass greater than 70 GeV/c2 The appropriate normalization of the background processes to the data luminosity is performed via their cross-section. In contrast to the less important backgrounds, where the higher-order cross-sections were used as a single number applied to the whole distribution, for the Drell-Yan background a more dedicated study was done. Emphasis was given to the estimation of a k-factor (that includes both QCD and Electroweak correction) which should vary with mass. The same study was performed for the Z′ signal as well. Assuming that all colorless final states are characterized by the same QCD radiation in their initial state, we can apply the same QCD correction as in the Drell-Yan case. Nevertheless, because the EW ones are model dependent, we ignored them in this case.Part of the study was devoted specifically to QCD background processes, b ¯b and cc¯. This background has little contribution to the analysis; however, due to the large uncertainty on its cross-section, it has to be estimated based on the data distribution. This is done by adjusting the distribution of the non-isolated muons (which is a characteristic of these processes) of the Monte Carlo to this of the data. The final QCD distribution with the isolation requirement included (i.e., after the whole selection) was extracted from the non-isolated one with the use of an appropriate scale factor. Another dedicated study was performed on the background from cosmics. Since a cosmic muon can be reconstructed as two different ones, characterized also by high momentum, it can be an artificial source of background. The impact parameter requirement does remove the vast majority of these muons. The study has shown that the remaining amount is negligible. After the whole selection was applied, the dimuon mass distributions for the data and the Monte Carlo samples were used to search for excesses that could be an indication of the Z′ presence. The data showed no evidence of signal and, thus, limits were set on the σB of Z′, in a mass range from 300 GeV to 1.5 TeV. The limits are evaluated using a likelihood analysis that is taking the shape of the mass into account. With the 2010 ATLAS pp data, the limit on the Z′ was measured equal to 867 GeV/c2 . This is equivalent to a σB exclusion of 0.278 pb. A very much alike analysis was performed on the 2011 data, with only minor changes implemented. In this case, too, the data was in agreement with the Standard Model expectations and no important excess was presented. The 2011 data extended the mass limit on the Z′ σB to the level of 1.678 TeV. The excluded cross-section this time was equal to 7.3 fb. The search on the Z′ existence continues from the ATLAS exotics group. The total amount of data taken in 2011 (∼ 5 fb−1) is studied and, since no excess is yet discovered, the limits are expected to reach the level of 2 TeV. A new publication is anticipatedsoon.

Παρουσιάζονται τα πρώτα αποτελέσματα της αναζήτησης νέων βαρέων, ουδετέρων μποζονίων στο πείραμα ATLAS, με τη χρήση της διάσπασής τους σε μιόνια (Z’ aμ+μ-). Τα αποτελέσματα αφορούν αφενός μεν τα πρώτα 42pb-1 των δεδομένων από συγκρούσεις πρωτονίων-πρωτονίων που συνελέγησαν κατά τη διάρκεια του 2010 στον επιταχυντή LHC του CERN και αφετέρου δεδομένα που συνελέγησαν μέσα στο 2011, συνολικής φωτεινότητας 1,21 fb-1. Αρχικά παρουσιάζεται μία θεωρητική εισαγωγή πάνω στα μοντέλα που προβλέπουν την ύπαρξη των νέων μποζονίων. Στη συνέχεια περιγράφεται ο ανιχνευτής του πειράματος και κατόπιν οι μέθοδοι ανακατασκευής μιονίων, η απόδοση του συστήματος και οι μέθοδοι μελέτης που χρησιμοποιούνται στην εργασία. Στο τελευταίο μέρος παρουσιάζεται η ανάλυση των πραγματικών δεδομένων. Περιγράφεται η διαδικασία επιλογής των ζευγών μιονίων και τα κριτήρια ποιότητας που εφαρμόζονται, τόσο στα πραγματικά δεδομένα, όσο και στα προσομοιωμένα δείγματα του σήματος και των διαδικασιών υποβάθρου. Βασικό υπόβαθρο είναι η διαδικασία Drell-Yan σε μιόνια, ενώ συνυπολογίστηκαν οι διαδικασίες ttbar, W+jets (όπου ένα jet ενδέχεται να αναγνωριστεί εσφαλμένα ως μιόνιο), η παραγωγή δι-μποζονίων (WW,WZ και ZZ) και τέλος οι QCD διαδικασίες bbbar και ccbar σε μιόνια. Τα επιλεχθέντα γεγονότα χρησιμοποιήθηκαν για να κατασκευαστούν οι κατανομές της αναλλοίωτης μάζας από τα ζεύγη μιονίων στα πραγματικά δεδομένα και στις διαδικασίες υποβάθρου. Με αυτό τον τρόπο εκτιμάται αν τα πραγματικά δεδομένα παρουσιάζουν παραγωγή γεγονότων πάνω από τα αναμενόμενα από το Καθιερωμένο Πρότυπο. Τα δεδομένα και των δύο ετών δεν έδειξαν παραγωγή γεγονότων επιπλέον των αναμενομένων. Με χρήση των δεδομένων του 2010, το όριο στη μάζα του Ζ’ υπολογίστηκε στα 874GeV/c2. Τα δεδομένα του 2011 επέκτειναν αυτό το όριο στα 1,682 TeV/c2 .