HIP 114328: a new refractory-poor and Li-poor solar twin

[Context]. The standard solar model fails to predict the very low lithium abundance in the Sun, which is much lower than the proto-solar nebula. This Li problem has been debated for decades, and it has been ascribed either to planet formation or to secular stellar depletion. In order to test the evolution of Li, it is important to find solar twins in a range of ages. Also, the study of stars similar to the Sun is relevant in relation to the signature of terrestrial planet formation around the Sun. [Methods]. We acquired high-resolution (R=110,000), high S/N (~300) ESO/VLT UVES spectra of several solar twin candidates and the Sun (as reflected from the asteroid Juno). Among the solar twin candidates we identify HIP 114328 as a solar twin and perform a differential line-by-line abundance analysis of this star relative to the Sun. [Results]. HIP 114328 has stellar parameters Teff = 5785+/-10 K, log g = 4.38+/-0.03, [Fe/H] = -0.022+/-0.009, and a microturbulent velocity 0.05+/-0.03 km/s higher than solar. The differential analysis shows that this star is chemically very similar to the Sun. The refractory elements seem even slightly more depleted than in the Sun, meaning that HIP 114328 may be as likely to form terrestrial planets as the Sun. HIP 114328 is about 2 Gyr older than the Sun, and is thus the second oldest solar twin analyzed at high precision. It has a Li abundance of A(Li)NLTE <= 0.46, which is about 4 times lower than in the Sun (A(Li)NLTE = 1.07 dex), but close to the oldest solar twin known, HIP 102152. [Conclusions]. Based on the lower abundances of refractory elements when compared to other solar twins, HIP 114328 seems an excellent candidate to host rocky planets. The low Li abundance of this star is consistent with its old age and fits very well the emerging Li-age relation among solar twins of different ages.