Assume that you made an XRD measurement and observed two types of profile that fits to your experimental pattern. What is the criteria for second or third phases to mark them as "impurity" or a "mixture" of multiple phases?
Please be aware sometimes at least in catalysis an "impurity" phase may be turn out to be the catalytically active phase. Consider each phase closely before discarding as an impurity.
Clearly there are also instances where a material may have over time hydrated, oxidised etc and up to the whole of the material may therefore now be an 'impurity' phase.
Should good 'search match' software not come up with matches you could have an entirely new phase or possibly phases from solid solutions. High symmetry ones are easy to detect lower symmetries will be much harder to analyse.
To decide that your sample is a mixture of phases you should have performed an accurate phase analysis to find a correct match with your experimental pattern. The impurity phases can be typically defined as the phases present in lower amount or as everything that moves away from your objective. If you want to sinthesize SrTiO3, TiO2 could be an impurity phase for your purpose. In some cases however, in the electrochemical field for example, the phases present in low amount can allow the diffusion or transport process so it is difficult to consider them as impurity phases.
I agree with the discussions above. The XRD pattern of a composite material for example may show signatures of different materials, none of which could be identified as "impurity phase". However, if the relative amount of the phase in the sample (as reflected by the relative intensities of the corresponding reflections) is implied in the original question, then I would refer to phases with small relative fractions (or traces) as "secondary phase", rather than "impurity phases".
Please check the fully-ready preprint article DOI: 10.13140/RG.2.2.27720.65287/3, Title: Qualitative Analyses of Thin Film-Based Materials Validating New Structures of Atoms