Emo TalkGen: Evaluating Multimodal Synthesis for Emotionally Expressive Talking Head Generation using Diffusion Models, Dual Decoders and 3DMM