The reconstruction of mesons was

The reconstruction of φ-mesons was done using their decay channel into two charged kaons, φ→К+К−. Charged particle tracks were reconstructed in two main tracking detectors of the ALICE experiment: TPC (Time Projection Chamber) and ITS (Inner Tracking System). Because of the high multiplicity of particles produced in central ultra-relativistic heavy ion collisions at LHC energies, signal extraction at low transverse momentum is not possible without particle identification (PID) because of a huge combinatorial background. Meanwhile at high transverse momentum PID does not play a considerable role. That is why all results presented in this article were obtained with PID at < 3 GeV/c and without PID at > 3 GeV/c. The PID was provided by using TPC (identification through ionization dna pk losses of charged particles) and TOF (Time of Flight – identification through time of flight and particle momentum). A set of additional cuts was applied and optimized for better signal-to-background ratio: For decay signal extraction invariant mass distributions of oppositely charged pairs of tracks (К+К−) were accumulated. These distributions contained both signal and combinatorial background. A mixed-event technique was used for estimating the uncorrelated background. Events selected for mixing had to have similar centralities ( ΔC < 10%) and z-coordinates of the interaction vertex ( Δz < 2 cм). After the uncorrelated background had been subtracted, the remaining invariant mass distributions were fitted to a complex function consisting of the convolution of the Breit–Wigner (for signal) and Gaussian (for detector mass resolution) functions and a second order polynomial function to take into account the remaining background. Some examples of the invariant mass distributions before and after combinatorial background subtraction are shown in Fig. 1. The Monte Carlo simulation was used for estimating the φ-meson reconstruction efficiency and the mass resolution of the tracking system. Pythia and HIJING were used as event generators for (p+p) and (Pb+Pb) collisions, respectively. Tracking of generated particles through the detector subsystems was simulated using the GEANT3 package. Mass resolution was estimated via a comparison of generated and reconstructed masses of φ-mesons separately for each analyzed pT bin. Distributions with the difference of generated and reconstructed masses were fit to a Gaussian function, while the distribution width obtained from the approximation was taken as equal to the mass resolution. The reconstruction efficiency of φ-mesons for each analyzed range was calculated as Nrec/Ngen, where Ngen is the number of generated φ-mesons in the analyzed transverse momentum and rapidity bin, and Nrec is the number of reconstructed φ-mesons after all selections. The extracted reconstruction efficiency rapidly increases at low transverse momentum. At > 8 GeV/c the efficiency saturates at ∼0.45 and no longer depends on the transverse momentum. Invariant differential production spectra for φ-mesons were built using the following formula: where is the number of events for analysis; is the center of the analyzed transverse momentum range; BR is the probability for a φ-meson to decay into two charged kaons (taken as 48.9%, according to [20]); Nφ is the measured number of φ-mesons in the analyzed transverse momentum and rapidity range; eff is the reconstruction efficiency per acceptance; ɛ is the trigger efficiency correction.
Measurement results Invariant differential production spectra of φ-mesons were measured in p+p collisions at = 2.76 TeV and in Pb+Pb collisions with different centralities at = 2.76 TeV. Results for Pb+Pb collisions are presented in Fig. 2. All measurements are provided in a wide range of transverse momenta up to 21 GeV/c for p+p collisions and (semi)central Pb+Pb collisions and up to 6 GeV/c for peripheral Pb+Pb collisions. In this and all subsequent figures statistical and systematic measurement uncertainties are shown with error bars and boxes, respectively. The spectra for different centralities were scaled by factors of 10 for better visualization.