Pulsar timing arrays offer a probe of the low-frequency gravitational wave spectrum (1–100 nHz), which is intimately connected to a number of markers that can uniquely trace the formation and evolution of the Universe. We present the dataset and the results of the timing analysis from the second data release of the European Pulsar Timing Array (EPTA). The dataset contains high-precision pulsar timing data from 25 millisecond pulsars collected with the five largest radio telescopes in Europe, as well as the Large European Array for Pulsars. The dataset forms the foundation for the search for gravitational waves by the EPTA, presented in associated papers. We describe the dataset and present the results of the frequentist and Bayesian pulsar timing analysis for individual millisecond pulsars that have been observed over the last ~25 yr. We discuss the improvements to the individual pulsar parameter estimates, as well as new measurements of the physical properties of these pulsars and their companions. This data release extends the dataset from EPTA Data Release 1 up to the beginning of 2021, with individual pulsar datasets with timespans ranging from 14 to 25 yr. These lead to improved constraints on annual parallaxes, secular variation of the orbital period, and Shapiro delay for a number of sources. Based on these results, we derived astrophysical parameters that include distances, transverse velocities, binary pulsar masses, and annual orbital parallaxes.

Context. PSR J1528−3146 is a 60.8 ms pulsar orbiting a heavy white dwarf (WD) companion, with an orbital period of 3.18 d. The pulsar was discovered in the early 2000 s in a survey at 1.4 GHz of intermediate Galactic latitudes conducted with the Parkes radio telescope. The initial timing analysis of PSR J1528−3146, using data recorded from 2001 and 2004, did not reveal any relativistic perturbations to the orbit of the pulsar or to the propagation of its pulses. However, with an orbital eccentricity of ∼0.0002 and a large companion mass on the order of 1 M⊙, this system has been deemed likely to exhibit measurable perturbations. Aims. This work is aimed at characterizing the pulsar’s astrometric, spin, and orbital parameters by analyzing timing measurements conducted at the Parkes, MeerKAT, and Nançay radio telescopes over nearly two decades. The measurement of post-Keplerian perturbations to the pulsar’s orbit can be used to constrain the masses of the two component stars of the binary and, in turn, to offer insights into the history of the system. Methods. We analyzed timing data from the Parkes, MeerKAT, and Nançay radio telescopes collected over about 16 yr, obtaining a precise rotation ephemeris for PSR J1528−3146. A Bayesian analysis of the timing data was carried out to constrain the masses of the two components and the orientation of the orbit. We further analyzed the polarization properties of the pulsar to constrain the orientation of the magnetic axis and of the line of sight with respect to the spin axis. Results. We measured a significant rate of advance of periastron, for the first time, and we set constraints on the Shapiro delay in the system and on the rate of change of the projected semi-major axis of the pulsar’s orbit. The Bayesian analysis yielded measurements for the pulsar and companion masses of Mp = 1.61−0.13+0.14 M⊙ and Mc = 1.33−0.07+0.08 M⊙ (68% C.L.), respectively, confirming that the companion is indeed massive. This companion mass as well as other characteristics of PSR J1528−3146 indicate that this pulsar is very similar to PSR J2222−0137, a 32.8 ms pulsar orbiting a WD whose heavy mass (∼1.32 M⊙) has been considered unique among pulsar-WD systems until now. Our measurements suggest common evolutionary scenarios for PSRs J1528−3146 and J2222−0137.