This is a rambling brain-dump from an e-mail today.
Historically, I'm not completely clear on the division. (I have this vague memory of T. Noone mentioning something by Ashworth on it.) It establishes itself by the 16th century and is, of course, present incipiently in texts like Aquinas's introduction to the Posterior Analytics commentary. (Yet, that's a dangerous text to wield if you're not careful - because his language goes down the lines that the Suarezians want to take. Even Giorio Pini read him this way. It's not the way other texts clearly go; also, it's not the way the Thomist school went.)
Still, it's perplexing to me, and I'm a little bit on the fence. Later on, you get the weird division of Minor and Major Logic. This is just a pedagogical splitting though. It usually divides along the lines of the old formal and material logic. (There are other more odd things in the 19th and 20th century because of "Criteriology " and "Critique". Nothing to be said here about those phenomena. Critique is valid as a defensive office of metaphysics.)
The general line is this. Allow me to ramble in what basically is two sentences with one very long paragraph. As always, I write in haste, so it's a bit rambling....!
De interpretatione and Prior Analytics deal with the formal structure of the syllogistic: the formal structures of the parts of syllogisms (propositions) and those of syllogisms themselves (considered in their figures and the immediate properties of those figures). Beyond that valid structure, you have questions of truth and certitude (with the latter comes the basic question of scientific knowability and certitude - they coined scibilitas in distinction from cognoscibilitas; hence, the Posterior Analytics concern with discussing not syllogistic-figure-related issues but with the conditions for science: e.g., per se nota propositions, and definition / division as regards the middle term of demonstration [book 2]; one then sticks the concerns of the Categories in here to this latter bit, but this seems a bit fabricated to me because there is a formal structure in defining [superiority, inferiority, etc.; the issue is, though, and here the tradition is mostly on to something that very quickly the predicables come up, and when you have to deal with accidental vs. proper definitions vs. essential definitions, you are dealing with a "material" issue]).
Thus, the general analogy (and it's only that) is:
f. logic : m. logic :: conditions of validity : conditions of soundness
Except, material logic is more than the posterior analytics. It includes the Topics (the bastard child of the Organon, but very important; I am going to be doing some work on this eventually) and the Sophistical Refutations (think of it like "crap matter"). For Aquinas and generally for the school, one wants to put Rhetorics in here with the Poetics. I think it is actually the correct position. But, this is just in need of some more development. Remember, this represents a huge battle in the Aristotelian world; Deborah Black's monograph will give you some thoughts here.
Now, a quick few other bits:
For the Thomists, 2nd intentions are those relations made in things as known and ordered by the natural process of the 3 acts of the intellect. Remember, each act has its own unique character:
1st. Defining: This is a unique work, not syllogistic. The predicables fall here for example. Genus, species, etc. But also "extension" and "intention" etc. Nous trying to get clear on a idea. That's how to think of it. (I can get you an excellent remark by Garrigou on this.)
2nd. Proposition formation. Remember, propositions bear on complexes (Man-being-risible); they aren't merely the semi-simple definition by property "political animal" (=first act of intellect). Thus, you can even say: "The political animal is a being ordered to a common good materially defined by a legitimate authority." In any case, the 2nd act of the intellect forms its own relations in these complexes. (Subject-predicate; etc.) There are various properties here regarding opposition, equipolence, etc. Remember, I stress this as a Thomist: the 2nd act of the intellect forms its own verbum. Not completely clear in Aquinas, but it's there enough that the school picks up on it.
3rd. Syllogizing: Here you deal with all the other sorts of higher level relations so familiar to those focusing on the Prior and Posterior Analytics. No new verbum; perhaps qualitatively altered (because to know a proposition as a conclusion adds a characteristic to that sort of knowledge).
Again, these are relations in things insofar as they objectively / intentionally exist. This differentiates them from real relations. They are reflexively known.
There is some disagreement late in the school, and I can't put my finger on it, that says that logic must be divided according to the formal character of its object. If its object is divided according to the 3 acts, which uniquely create different kinds of relations, then one should divide logic this way instead. (Now, perhaps there would be formal and material sub-branches; I don't know. It's not that easy; just trust me. The various acts interact a bit. For example, suppositio only befalls terms because they are in enunciations; and enunciations only are called propositions if they are in syllogizing; etc.)