The phrase πρώτῃ σαββάτου (v. 9) occurs nowhere else in the NT (but see Jub. 3:1; the superscription to Psalm 47 uses δευτέρᾳ σαββάτου, “the second day of the week”) though a similar construction, τῇ πρώτῃ ἡμέρᾳ τῶν ἀζύμων, occurs in 14:12. It might be wondered, however, if such a reference in regards to the first day of the week is not part of “standard usage,” and in that case the standard collocation with σαββάτου/ων seems to be μία σαββάτου/ων (an elliptical expression for μία ἡμέρα σαββάτου/ων; see Matt 28:1; Mark 16:2; Luke 24:1; John 20:1, 19; Acts 20:7; 1 Cor 16:2; see also the superscription of Psalm 23 [Eng., 24]). LXX usage typically employs πρώτη ἡμέρα in regard to a feast (e.g., Exod 12:15) or of a month (e.g., Ezra 10:17). The Pseudepigrapha uses πρώτῃ ἡμέρᾳ (Jub. 2:2) Josephus typically uses πρώτη ἡμέρα (Ant. 1.29), or in the similar construction, τῇ πρώτῃ τῆς ἑορτῆς ἡμέρᾳ (Ant. 5.22). Philo, likewise uses πρώτην ἡμέραν (Spec. Laws 2.162, in regard to a feast). It appears that the normal pattern is to use the ordinal (πρώτη) with ἡμέρα, but the cardinal (μία) in the elliptical expression μία [ἡμέρα] σαββάτου/ων, though the use with σαββάτου/ων appears in our literature almost exclusively in the NT; the OT and related texts are more concerned with the seventh day, typically ἡ ἡμέρα ἡ ἑβδόμη (e.g., Exod 16:26, 27)—also an ordinal. Also of note is the use of the singular σαββάτου; the only other NT uses of the singular in a temporal sense of “week” are δὶς τοῦ σαββάτου (Luke 18:12, “twice a week”) and κατὰ μίαν σαββάτου (1 Cor 16:2, “on the first day of the week”). In the LXX we find τὸ σάββατον (“the Sabbath,” usually genitive or accusative, e.g., 2 Kings 11:5; Neh 13:19), but almost never in the sense of “week” (the superscription to Psalm 47 [Eng., 48] is the only exception). The use of the singular by Josephus and Philo is the same, as it is in the Pseudepigrapha and the Apostolic Fathers.
There are two contrasting uses here: τῇ μιᾷ τῶν σαββάτων in 16:2 and πρώτῃ σαββάτου in 16:9—odd for being used divergently only a few verses apart if Mark were the author of both when usage almost everywhere else is so consistent. These differences in themselves are not adequate to demonstrate a difference of authorship between the Long Ending and Mark (i.e., between 16:9–20 and 1:1–16:8), but it does suggest that this is very unusual usage since πρώτῃ σαββάτου can be paralleled exactly, so far as I can determine, in only one pseudepigraphal text (plus one other similar expression).
I am wondering if this could reflect later usage (i.e., post 1st C AD or at least post-NT), though I do not have TLG access from off campus to check that hypothesis.