You could possibly argue that transformational leadership is a mediator of the relationship between EI and OCB: People high in EI probably do more transformational leadership behaviors. People with more transformational leadership behaviors might do more OCBs than people low in transformational leadership (but they might not).
I'm skeptical about such a model. Neither EI nor OCBs require any leadership, nor are they especially higher in leaders vs non-leaders. I'm sure that you'd find that they are all positively correlated, but a person's evaluation of an other's transformational leadership seems to be positively correlated to all the positive qualities of the other. I think this is mainly due to the halo effect: If I like somebody, I think they have transformational leadership qualities. For example, people who like Donald Trump see him as a transformational leader. People who do not like Donald Trump do not see him as a transformational leader.
I do not think I can understand your second model. How can a variable be an antecedent to a relationship? Conceivably, TL could be an antecedent to OCBs. I don't see how TL could be an antecedent to EI.