I would say that you can totally go ahead and refer to the other chapters of your thesis in your situation if you are willing and able to still change the third manuscript when you are actually going to publish it, if ever (and I don't really see why you wouldn't).
If so, when preparing your manuscript for submission to a journal, simply replace the relevant citations to either references:
- to the published articles, if you have managed to publish them by then (best case scenario).
- to your PhD thesis if you haven't.
- a combination of the two if you were able to one article, but not the other.
In my reading, this is further acceptable, because referring to a different chapters is exactly how you would go about the problem if your thesis was not prepared in a "manuscript-based" style. And given that no part of your work has, in fact, been published yet, the style/format is really all that sets your thesis apart from a monograph-style thesis. And so I don't
The other answers given so far are flawed, in my opinion:
- Roger V. advises to add a fully qualified citation to your thesis. However, at the time of writing you are not able to do so, because you probably don't know the exact DOI, volume, issue or even URL where your thesis will be available. And even if you were, some people might well regard citing your own work within the same work as an abusive form of self citation, as it will increase the citation count of said work. Roger V. also suggests that a PhD thesis is equivalent to a peer-reviewed journal article. I don't necessarily agree with that, because PhD theses are frequently reviewed only by members of the thesis committee, and it could be argued that there may be a conflict of interest, as is demonstrated by the much lower rejection rates of PhD thesis compared to journal article submissions (at least in the universities I know). Finally, at least at the time of drafting your thesis and adding the citations, your thesis hasn't been reviewed at all, so that argument doesn't even apply.
- HEITZ doesn't explicitly recommend anything, but their answer implicitly could be read as advice not to reference to your other chapters because (1) readers of your work would not be able to find the information you are referencing, and (2) a citation implies peer review. Both are wrong, in my opinion. Given that your entire work is contained in the same volume (the PhD thesis/dissertation) and not published elsewhere, all information will be available to any reader (assuming that they have access to the entire work and not just the chapter or a page, but we would typically make that very same or even a stronger assumption for readers of a manuscript, who we also believe to have access to the entire manuscript, plus every other published article). And merely citing "citable resources" does not imply peer review, and neither should it. You can and should cite published works and resources that are not peer reviewed, such as preprints and datasets, as long as there is no better source available to refer to (e.g., prefer to cite a peer-reviewed article over its preprint, once available).
- It appearst to me that Christian Hennig misunderstood the question (or perhaps I did?), understanding that you would like to publish manuscript 3, while not having published manuscripts 1 and 2 outside of your PhD thesis. But even if that were the case, I still think it is acceptable and not unusual to cite a PhD thesis as long as it is publicly available (as it usually is). In fact, I recently saw a case where an editor of a journal specifically asked to include a citation to a "manuscript-based" PhD thesis in a manuscript that included parts of the submitted work (though, in that case, the publication of the thesis was actually predated by publication in a preprint server, which was already cited).
Note, however, that Roger V. actually gives what is probably the best advice: Try to publish all of your articles on a preprint server like arXiv, or whatever is appropriate in your field. This not only avoids your problem (just publish manuscripts 1 and 2 first, then manuscript 3 after), but also gives your thesis the feel of a more conventional manuscript-based thesis. But of course, given your timelines, this may not be possible anymore at this point, as it would still probably take a week or up to a month before you get a citable identifier / DOI to include in manuscript 3. But maybe for next time :-D