The hybridization situation you describe could be found in ring species, and is partially related to this concept.
For instance, the three species A, B and C could have partially overlapping distributions, such as:

In such a situation (assuming that the relative distributions have been stable over evolutionary time), A and C might be able to produce offspring, as well as B and C, while A and B might not produce viable offspring if they would meet.
Here is a picture from the wikipedia page for a similar situation, showing interbreeding between seven gull species in the genus Larus:

However, I also know that the ring species concept has been challanged, mainly by the fact that there are very few good empirical examples of ring species. See for instance the nice blog post "There are no ring species" by Jerry Coyne, which may also include a couple of examples of species groups that are relevant to your question.
Also, the hybridization patterns that you describe can be caused by other processes, for instance specific reproductive barriers between groups of species (molecular or behavioural). My answer is therefore merely mentioning one possible explanation for the hybridization patterns you ask for, but there are many other processes to consider.