Pick up the book Audio-Vision by Michel Chion.
It describes one version of the foundations of "using sound to evoke emotions in motion pictures". And draws examples and example film scenes to argue them.
Arguing how frequency content "defines" emotion is not straightforward. Acoustically we can say that certain frequencies have certain qualities (e.g. soft, harsh, piercing, honky, dark, bright). But in soundtracks there aren't individual frequencies, not necessarily even individual instruments or sounds. Thus you'd need to argue starting from the grounding of "what makes a certain sound sound in a particular way (e.g. soft, harsh, piercing, honky, dark, bright ...)" and generalize that to how the same properties may be used in any sound, which can be e.g. in a film scene and which has similar properties than the "pure" example sounds that you drew earlier. Additionally you may want to test it on subjects. At the same time the conclusion is trivial, we all know that sounds with certain properties make us feel in a particular way, but we may not specifically know and cannot know, why it happens (i.e. is it really deducible from some particular frequency content or whether there's much more to it, or whether the results vary from person to person and how much, or if the properties are context-dependent like they are in a soundtrack mix). Thus arguing from the point of "frequency content" might be too much of a simplification. A sound quality or property does not necessarily translate to "emotion" (or a particular type of emotion).