J0216−0021 : left panels (a): velocity plot of a sub-set of the absorbing species. The vertical dashed lines indicate the positions of the C I components. Fits to the absorption profiles are over-plotted with the data. For C I transitions, the red dashed line corresponds to the true ground state, the blue dashed line is for the C I* absorption, and the orange dashed line is for the C I^✻✻ absorption. The redshift is taken in Table 1 to give the zero velocity. Right panels (b): upper panel is the spectrum at the expected position of Na Iλλ5891,5897, the 1 and 2 indicate the Na Iλ5891 and Na Iλ5897 lines respectively; lower panel is the spectrum at the expected positions of the Ca IIλλ3934,3969 lines, the H and K notations indicate the Ca IIλ3969 and Ca IIλ3934 lines respectively. The red curve is the telluric spectrum template of X-shooter: In this system the Ca II K line is clearly detected but we indicate an upper limit for the H line since the data around the H line is noisy (see Fig. A.1b lower panel). The Na I D is detected clearly even though there are strong spikes close to the two absorption lines (see Fig. A.1b upper panel). We subtracted the equivalent width of the weak telluric contamination. The C Iλ1560 line is not used in the fit of the C I absorption as it is strongly blended. There are two components at z = 1.735283 and 1.735888, while the strongest component of Fe II, Si II, and Mg I is seen in-between these two components (see also J1248+2848). This could be an artefact due to the spectral resolution. To discuss the details of the structure of the absorbing cloud, higher resolution data are needed.