Emotional health is important for overall wellbeing. Expressing our emotions brings about a lot more benefits like helps to see problems in a new light, makes decision making and problem solving easier, get rid of the power of the feeling, reduces anxiety and eases depression. But it is very difficult to have a close friend or family member to whom can be trusted always to vent out any problems you are dealing with. This raises an idea of developing a trusted friend like conversational system using AI. The revolutionary potential of emotional conversational system (chatbots) in supporting human emotional health is developed in this work. In this study the operation of various chatbots were analyzed in order to reveal important qualities necessary for the development of highly efficient conversational systems. This examination, which ranges from rule-based to machine learning-based models, provides light on the intricacies of each kind, leading to a better understanding of their mechanics. This highlights critical factors for creating emotionally intelligent chatbots, emphasizing the necessity of adaptive learning, natural language processing, and context awareness. This contribution serves as a starting point for developers, providing insights into the inner workings of various chatbot designs and paving the way for the creation of smart, user-centric, and efficient chatbots poised to revolutionize emotional healthcare support.