Practical and Scalable Two-Step Process for 6-(2-Fluoro-4-nitrophenyl)-2-oxa-6-azaspiro[3.3]heptane: A Key Intermediate of the Potent Antibiotic Drug Candidate TBI-223

A low-cost, protecting group-free route to 6-(2-fluoro-4-nitrophenyl)-2-oxa-6-azaspiro[3.3]heptane (1), the starting material for the in-development tuberculosis treatment TBI-223, is described. The key bond forming step in this route is the creation of the azetidine ring through a hydroxide-facilitated alkylation of 2-fluoro-4-nitroaniline (2) with 3,3-bis(bromomethyl)oxetane (BBMO, 3). After optimization, this ring formation reaction was demonstrated at 100 g scale with isolated yield of 87% and final product purity of >99%. The alkylating agent 3 was synthesized using an optimized procedure that starts from tribromoneopentyl alcohol (TBNPA, 4), a commercially available flame retardant. Treatment of 4 with sodium hydroxide under Schotten–Baumann conditions closed the oxetane ring, and after distillation, 3 was recovered in 72% yield and >95% purity. This new approach to compound 1 avoids the previous drawbacks associated with the synthesis of 2-oxa-6-azaspiro[3,3]heptane (5), the major cost driver used in previous routes to TBI-223. The optimization and multigram scale-up results for this new route are reported herein.

The three fractions were collected, with their composition shown in Table S1 based on the GCMS data shown in Figure S1. The major impurity, 3-bromo-2-bromomethyl-1-propene (RT 2.4 min, matched to a commercial sample), is separated from the rest of the mixture in the initial stage of the distillation. The third major fraction was pure product (>96% wt% purity by GC). Yield of the pure fraction based on purity was 72% (99.7% GC A% purity).

Purity after distillation Fraction
No.   Figure S1: GCMS chromatograms of the different fractions collected during the distillation of crude BBMO (for chromatography conditions, see below).

S5
Step 2: Optimization, Impurity, and Additional Reaction Information 1 General reaction conditions: NaOH was added to a solution of aniline 2 (100 mg, 1.0 eq.) and 3

DOE Optimization of Alkylation Conditions: Data and Analysis Information
in sulfolane. This mixture was then heated to the specified temperature for 3 h. 2 These are HPLC area% data for IPC samples taken of the reaction mixture. The area% are not corrected for the response factors of each compound.
These DOE data were first coded and then analyzed using ordinary least squares (OLS) methods as implemented by Statsmodels, 1 a statistics packs for the Python programming language. Initially, S6 the data were fit with a full model consisting of all the main effects and their interaction terms.
However, the interactions terms were not found to be statistically insignificant (p value > 0.05), so they were ignored. The OLS results for the main effect term models are shown found in Table S3 for 1 and Table S4 for 6.  A sample of crude step 2 product (10 g, 90 area% of product 1 and 10 area% of impurity 6 by HPLC, Figure S5)

Crystal Data and Experimental on 1.
ORTEP of 1. Only unique and disordered positions with hydrogens omitted, (all data) and R 1 was 0.0412 (I≥2

(I)).
Experimental. Single clear yellow prism-shaped crystals of AJA03A-Qu were used as supplied. A suitable crystal with dimensions 0.38 × 0.24 × 0.12 mm 3 was selected and mounted on a XtaLAB Synergy R, DW system, HyPix diffractometer. The crystal was kept at a steady T = 100.01(10) K during data collection. The structure was solved with the ShelXT 2014/5 (Sheldrick, 2014) solution program using dual methods and by using Olex2 1.3-alpha  as the graphical interface. The model was refined with ShelXL 2018/3  using full matrix least squares minimisation on F 2 .  Model. The raw data and refinement model were of good quality and excellent fit. The molecule sits on a crystallographic mirror plane that includes the azetidine and phenyl rings, the nitrogroup, the fluorine atom, and the terminal ether oxygen of the oxetane ring. The orientation of the fluorophenyl moiety was modeled with disorder between two positions. One orientation (36%) placed the fluorine in the vicinity of a nitro-oxygen (rinter-atomic = 2.62 Å) while the second orientation (64%) placed the fluorine into a small void adjacent the oxetane ring of a neighboring molecule (see Figure 2X). No evidence was found for a super-lattice that might eliminate disorder in the model. Figure 2X: Planar packing diagram showing F-disorder.

Structure Quality Indicators
Reflections:
Data were measured using  scans using Mo K  radiation. The diffraction pattern was indexed and the total number of runs and images was based on the strategy calculation from the program The structure was solved and the space group P2 1 /m (# 11) determined by the ShelXT 2014/5 (Sheldrick, 2014) structure solution program using using dual methods and refined by full matrix least squares minimisation on F 2 using version 2018/3 of ShelXL 2018/3 . All non-hydrogen atoms were refined anisotropically. Hydrogen atom positions were calculated geometrically and refined using the riding model. Hydrogen atom positions were calculated geometrically and refined using the riding model.

_refine_special_details:
The molecule was found to be disordered by a 180 degree rotation around the C-N bond. It was conveniently modeled as F atom 2-component positional disorder. No restraints or constraints were applied.
The value of Z' is 0.5. This means that only half of the formula unit is present in the asymmetric unit, with the other half consisting of symmetry equivalent atoms.