Context. Infrared observations of water ice in the interstellar medium are hindered by uncertainties in the band strength, density, and porosity, which introduce considerable errors in the estimated column density of the ice on a line of sight. Aims. We revise the infrared band strength values and band positions of water ice grown under simulated interstellar conditions at different deposition and warm-up temperatures. We also explore other physical ice parameters: density, refractive index, and porosity. Methods. We grew water ice in simulated interstellar conditions in ultra-high vacuum with temperatures between 10 and 150 K. We used infrared spectroscopy and laser interferometry to obtain updated values for the band strengths. With these updated values, we calculated and report the density, refractive index, porosity, and infrared band position for water ice. Results. Previous measurements of these properties were inaccurate because ice was considered non-porous. Our results show that the band strength for the O-H stretching vibration varies significantly with deposition temperature and should not be considered constant. There is also measurable variation of the band strength during the warm-up of the ice after deposition. Conclusions. Previous ice density values were overestimated due to inaccurate band strengths and the omission of porosity, which can only be considered negligible at very slow growth rates and deposition temperatures close to sublimation. The infrared band position shifts with varying porosity.

Context. The surfaces of icy moons are primarily composed of water ice that can be mixed with other compounds, such as carbon dioxide. The carbon dioxide (CO₂) stretching fundamental band observed on Europa and Ganymede appears to be a combination of several bands that are shifting location from one moon to another. Aims. We investigate the cause of the observed shift in the CO₂ stretching absorption band experimentally. We also explore the spectral behaviour of CO₂ ice by varying the temperature and concentration. Methods. We analyzed pure CO₂ ice and ice mixtures deposited at 10 K under ultra-high vacuum conditions using Fourier-transform infrared (FTIR) spectroscopy and temperature programmed desorption (TPD) experiments. Laboratory ice spectra were compared to JWST observation of Europa's and Ganymede's leading hemispheres. The simulated IR spectra were calculated using density functional theory (DFT) methods, exploring the effect of porosity in CO₂ ice. Results. Pure CO₂ and CO₂-water ice show distinct spectral changes and desorption behaviours at different temperatures, revealing intricate CO₂ and H₂O interactions. The number of discernible peaks increases from two in pure CO₂ to three in CO₂-water mixtures. Conclusions. The different CO₂ bands were assigned to ν_3,1 (2351 cm^-1, 4.25 μm) caused by CO₂ dangling bonds (CO₂ found in pores or cracks) and ν_3,2 (2345 cm^-1, 4.26 μm) due to CO₂ segregated in water ice, whereas ν_3,3 (2341 cm^-1, 4.27 μm) is due to CO₂ molecules embedded in water ice. The JWST NIRSpec CO₂ spectra for Ganymede and for Europa can be fitted with two Gaussians attributed to ν_3,1 and ν_3,3. For Europa, ν_3,1 is located at lower wavelengths due to a lower temperature. The Ganymede data reveal latitudinal variations in CO₂ bands, with ν_3,3 dominating in the pole and ν_3,1 prevalent in other regions. This shows that CO₂ is embedded in water ice at the poles and it is present in pores or cracks in other regions. Ganymede longitudinal spectra reveal an increase of the CO₂ ν_3,1 band throughout the day, possibly due to ice cracks or pores caused by large temperature fluctuations.