Efficient metal-catalyzed dimerization of nitrenes (derived from organoazides) constitutes a useful alternative to the above processes. 6-11 Selected transition metal catalysts for azoarene formation are presented in Figure 1. Most of the existing catalysts demonstrate limitations in the types of azoarenes they can produce. Uyeda's catalyst (B) exhibits the broadest scope of substrates, with the only limitation being bulky substrates (larger than mesityl). 11 Our group investigates group-transfer chemistry at [M(OR) 2 ] platforms, generally featuring two bulky monodentate alkoxide ligands. 12 We have reported that Fe(OR) 2 (THF) 2 12a-c (C, OR = OC t Bu 2 Ph) served as an efficient catalyst for nitrene dimerization, although its reactivity was Fe(OR) 2 (THF) 2 /Fe(OR') 2 (THF) 2 (C/D, Figure 1), allowing efficient dimerization of non-bulky aryl nitrenes.

Bis(alkoxide) complexes featuring two bulky monodentate ligands generally exhibit a relatively broad angle between the alkoxides (RO-M-OR ~140 °, Figure 1), which leads to metal geometries intermediate between tetrahedral and seesaw. To model this disposition, we decided to employ the 1,4terphenyl structural motif, which was utilized by Agapie and coworkers in the synthesis of chelating ligands with a wide intra-ligand angle. 13 Chelating bis(alkoxide

was synthesized in a three-step procedure (ESI) that involved initial synthesis of 1,4-bis(2-bromophenyl)benzene, 14 followed by lithiation with t BuLi, and finally reaction with benzophenone (72% yield, Scheme S1 in ESI). Phenyl substituents were chosen for the initial foray into the chelating bis(alkoxide) chemistry due to the anticipated crystallinity and slightly less steric hindrance compared with tert-butyl substituents. 1 was characterized by NMR spectroscopy, highresolution mass spectrometry, and X-ray crystallography (see ESI). X-ray crystallography confirms the overall connectivity and indicates anti geometry in the solid state with the alcohols on different sides of the central phenyl (Figure 2). S3), which suggests [OO] Ph may allow metal-phenyl interactions in some species. The reactivity of 2 in nitrene coupling was investigated and the results are presented in Figure 4. All reactions were carried out in C 6 D 6 at 60 °C for 24 h using 10 mol% catalyst. The reactions were monitored by NMR spectroscopy, using trimethoxybenzene, hexamethylbenzene, or hexafluorobenzene as an internal standard. The products were identified by 1 H NMR spectroscopy and confirmed by GC-MS; selected azoarenes were isolated and further characterized (ESI). Most notably, nitrene coupling reactivity exhibited by Fe[OO] Ph (THF) 2 ( 2) is almost perfectly complementary to the reactivity exhibited by Fe(OR) 2 (THF) 2 or Fe(OR') 2 (THF) 2 : while the previous catalysts were selective for the coupling of bulky aryl nitrenes, the present catalyst (2) demonstrates high to moderate yields for non-bulky substrates featuring para alkyl, trifluoro, or halo substituents; the conversion is quantitative for the phenyl azide (Figure 4). The formation of halosubstituted azoarenes is particularly noteworthy, as these groups can be later functionalized via cross-coupling. Although lower yields were observed for the p-nitro, p-acetyl, or pmethoxy groups, these are uncommon functional groups for metal-mediated formation of azoarenes.

Please do not adjust margins In contrast, Fe(OR) 2 (THF) 2 did not exhibit any reactivity with phenyl azide or para-substituted azides, and Fe(OR') 2 (THF) 2 exhibited low reactivity at best. Meta-substituted 3,5dimethylphenyl azide and 3-trifluoromethylphenyl azide exhibited relatively high yields (69% and 70% yields, respectively). In a sharp contrast, no reactivity with orthosubstituted (e.g. mesityl) azides was observed for Fe [OO] Ph (THF) 2 (2), whereas both Fe(OR) 2 (THF) 2 and Fe(OR') 2 (THF) 2 demonstrated quantitative yield with mesityl azide.

What is the origin of the observed reactivity differences between Fe[OO]

Ph and Fe(OR) 2 /Fe(OR') 2 (C/D) systems? Due to the overall similarity of the three iron bis(alkoxide) systems, it is probable that the reaction of Fe[OO] Ph with aryl azides also proceeds via a metal-nitrene intermediate, whose formation and reactivity were computationally interrogated for the Fe(OR) 2 /Fe(OR') 2 systems. Putative iron nitrenes with Ar = Ph and Ar = 2,6-Me 2 Ph (Mes surrogate) were optimized as quintets, based on our earlier modeling efforts, 10 for [OO]

Ph and OR (Figure 5). In the less sterically bulky NPh species, the N of the nitrene is exposed for reactivity whereas it is blocked by one of the alkoxide phenyl groups for the ortho-substituted aryl substituent when the iron nitrene is supported by the [OO] Ph ligand (Figure 5, top). In contrast, the nitrene N remains accessible for both aryl substituents when FeNAr is supported instead by two OR groups (Figure 5, bottom). Therefore, we hypothesize that the inaccessibility of the nitrene when orthosubstitutents are present may be responsible for the lack of diazoarene formation in the presence of our new chelating ligand. Moreover, the Fe-N bond is mostly accessible by an incoming substrate from either face of the plane define by the alkoxides and nitrene in Fe(OR) 2 (NAr), whereas the back face is blocked by the bridging terphenyl group and the front face is at least partially blocked by the alkoxide phenyl groups in Fe[OO] Ph NAr. Even though these ligands give a similar trigonal planar coordination environments in both bis(alkoxide) environments, the spatial constraints imposed by the chelating ligand may provide an opportunity to control reactivity by limiting the approach of reactants to the nitrene moiety.

In conclusion, we have designed new chelating bis(alkoxide) ligand H 2 [OO] Ph with the 1,4-terphenyl bridge linking [OCPh 3 ]-like donors. Reaction of the ligand with an ironamide precursor led to the formation of a mononuclear complex Fe[OO] Ph (THF) 2 closely related to the previously reported Fe(OR) 2 (THF) 2 complexes with bulky monodentate alkoxides (OR = OC t Bu 2 Ph and OC t Bu 2 (3,5-Ph 2 Ph)). Whereas the overall donors disposition at the iron(II) center is maintained, tying the alkoxides with the terphenyl bridge led to a pronounced difference in the steric nature at the metal site, which is illustrated by a nearly linear OAr-Fe-OAr angle (156°), as compared with ~140° in the more relaxed structures with monodentate ligands. Whereas the present "OCPh 3 " motif is not bulkier than the previously utilized OC t Bu 2 Ar, it is likely the rigidity of the chelating system that prevents the rearrangement (i.e. rotation or geometry change) of the alkoxides. This difference in the nature of the active site is manifested in the reactivity of the resulting nitrene coupling catalysts. While Fe(OR) 2 (THF) 2 complexes were able to couple bulky aryl nitrenes only (for OR = OC Ph (THF) 2 is selective for coupling