Versatile Coordination Modes of Multidentate Neutral Amine Ligands with Group 1 Metal Cations.

This work comprehensively investigated the coordination chemistry of a hexa-dentate neutral amine ligand, namely, N,N',N"-tris-(2-N-diethylaminoethyl)-1,4,7-triaza-cyclononane (DETAN), with group-1 metal cations (Li+, Na+, K+, Rb+, Cs+). Versatile coordination modes were observed, from four-coordinate trigonal pyramidal to six-coordinate trigonal prismatic, depending on the metal ionic radii and metal's substituent. For comparison, the coordination chemistry of a tetra-dentate tris-[2-(dimethylamino)ethyl]amine (Me6Tren) ligand was also studied. This work defines the available coordination modes of two multidentate amine ligands (DETAN and Me6Tren), guiding future applications of these ligands for pursuing highly reactive and elusive s-block and rare-earth metal complexes.

Introduction

Ligand design is at the center stage of coordination chemistry and plays an essential role in catalysis. A well-designed ligand is a prerequisite for isolating and analyzing highly reactive and elusive metal coordination species. The knowledge, in return, enables chemists to design new catalytic reactions. Several recent breakthroughs in coordination chemistry highlighted the importance of ligand design, such as a tripodal tris-anionic amide ligand stabilized uranium terminal nitride,[1] and N-heterocyclic carbene (NHC) or cyclo-amino alkyl carbene (cAAC)-stabilized low-valent boron[2,3] and beryllium[4,5] complexes.

As in the d-/f-/p-block metal chemistry, ligand design is equally crucial in s-block Group 1 metal chemistry. Highly reactive Group 1 metal species play essential roles in numerous catalytic processes and act as ubiquitous reagents, such as organolithium reagents,[6] Lochman Schlosser superbases,[7] and the plethora of Group 1 metal salts used as additives in organic synthesis.[8,9] These systems are highly complicated, and they usually involve aggregating/clustering of multimetallic species. To understand their mechanisms, coordination chemists used several ligands to “trap” the highly reactive and elusive species and synthesize model complexes.

Therefore, designing new bespoke ligands and understanding their applicable range (such as metal ionic radii range) and corresponding coordination modes is essential to synthesizing such highly reactive and elusive metal complexes. However, designing bespoke ligands for Group 1 metals is more difficult than for the d-/f-/p-block and Group 2 metals. The difficulties are caused by two factors. (1) From a charge balance perspective, the monovalent Group 1 metal cation rules out the usage of anionic building blocks for heteroleptic complexes. For example, the massively successful anionic cyclopentadienyl (Cp) and beta-diketiminate (BDI) families are of little use in Group 1 chemistry. (2) The Group 1 metal cations feature an ns0 valence shell electronic structure, and form ionic bonds with ligand atoms: the metal-to-ligand backdonation is very weak, if there is any. Therefore, they cannot take advantage of donor–acceptor building blocks, such as the popular N-heterocyclic carbenes (NHCs)[10] and cAACs.[11]

Multidentate neutral amine ligands are arguably the most successful ligand family in Group 1 metal chemistry.[12−16] Their two key advantages are (1) synthetic availability and tuneability; and (2) chemical robustness (the C–N bond is more stable than the C–O bond). The denticity of the multidentate amine ligands plays an essential role in their coordination chemistry. The most-used ligands are bis- and tris-dentate, such as tetra-methyl ethylenediamine (TMEDA),[17] (−)-sparteine,[18] (R,R)-N,N,N′,N′-tetramethyl-1,2-diaminocyclohexane [(R,R)-TMCDA][19] and N,N,N′,N″,N′′-pentamethyldiethylenetriamine (PMDTA).[20,21] The bis- and tris-dentate amine ligands (L) succeeded in isolating monomers of sterically bulky lithium alkyl complexes [LiR(L)].[17−20]Tetra-dentate amine ligands, such as tris-[2-(dimethylamino)ethyl]amine (Me6Tren), also has been used in Group 1 chemistry.[22−31] Efforts were also made to combine multidentate neutral amine donors and anionic donors, such as a cyclen-derived tetra-dentate (NNNN) macrocylic anionic ligand developed by Okuda and co-workers.[32] However, the bis-, tris- and tetra-dentate ligands could not provide sufficient kinetic protection for isolating more reactive species.[17] Higher-dentate amine ligands are necessary. For example, a nona-dentate per-aza cryptand[2,2,2] was designed by Dye and co-workers to isolate a thermally stable organic electride.[33] Recently, we designed a hexa-dentate ligand, namely, N,N′,N′′-tris-(2-N-diethylaminoethyl)-1,4,7-triaza-cyclononane (DETAN), and isolated the first methyllithium (MeLi) monomer.[34]

Our DETAN ligand combines a semirigid 1,4,7-triazacyclononane (TACN) macrocyclic backbone and three flexible coordinative side arms, which feature good thermodynamic robustness (no fragile C–O bond). Compared to Dye’s per-aza cryptand[2,2,2],[33] the DETAN is more flexible and could accommodate reactive M–E metal functional groups with variable sizes, such as metal terminal imides/phosphinidenes (M=NR/M=PR) and nitrides/phosphides (M≡N/M≡P). Moreover, unlike the rigid per-aza cryptand[2,2,2], DETAN’s flexible side arms could decoordinate, allowing Lewis basic organic/small molecular substrates to approach the metal center and facilitate subsequent reactivity studies.

The excellent kinetic protection and thermodynamic robustness, as well as the flexible side arms, make DETAN an attractive ligand for pursuing long sought-after and highly reactive metal complexes, such as monomeric Group 1 metal amide/hydride,[35] Group 2 metal terminal imides,[36] and divalent rare-earth metal terminal imides.[37−39] A prerequisite for these applications is a comprehensive understanding of the DETAN’s coordination modes. In this work, employing Group 1 metal halides and tetraphenylborates as a platform, we mapped out the diversified coordination modes of the DETAN ligand and how it changes in accordance with the metal ionic radii and the M–E functional groups. For comparison, the tetra-dentate Me6Tren ligand was studied as well. The results are reported below.

Synthesis and Characterization

We first treated lithium iodide (LiI) with Me6Tren and DETAN in d8-THF, respectively. The NMR scale reactions were monitored by 1H and 7Li NMR spectra, indicating full conversions within 24 h at room temperature. The reactions were subsequently scaled up, employing diethyl ether (Et2O) or THF as the solvent, to yield complexes 1 and 2 in 96% (1) and 79% (2) yields, respectively (Scheme ). Complexes [Li(I)(κ4-N,N′,N″,N′′′-Me6Tren)] (1) and [Li(I)(κ4-N,N′,N″,N′′′-DETAN)] (2) were obtained as white crystalline solids. It is worth mentioning that complex 2 is more soluble than complex 1 in most organic solvents. Specifically, complex 1 is soluble in THF, sparingly soluble in benzene/toluene, and insoluble in Et2O and hexane. In comparison, complex 2 is soluble in THF/Et2O/toluene/benzene, and insoluble in hexane. We attribute the better solubility of 2 to the side arms of the DETAN ligand. Complexes 1 and 2 are stable at room temperature indefinitely, and they are extremely hygroscopic when exposed to air, to produce intractable mixtures containing free DETAN/Me6Tren ligands.

Scheme 1

Reactions To Produce [Li(I)(κ-N,N′,N″,N′′′-Me6Tren)] (1) and [Li(I)(κ4-N,N′,N″,N′′′-DETAN)] (2)

Single crystals of complexes 1 and 2 suitable for SCXRD studies were obtained from THF (1) and Et2O (2) solutions at room temperature (1) or −35 °C (2). Their structures are exhibited in Figures and 2, respectively. Both 1 and 2 feature a trigonal bipyramidal coordination geometry surrounding the five-coordinate Li+ center. Three equatorial and one apical positions are occupied by neutral N donors, while the other apical position is occupied by an anionic iodide (I–) donor. In complex 1, all four N donors of the Me6Tren ligand coordinate to the Li+ center. In comparison, in complex 2, only four out of the six N donors coordinate to the Li+ center. A closer examination of the structures of 1 and 2 reveals that, although both are trigonal bipyramidal, their Li–I bond lengths are significantly different (Chart ). The Li–I bond in 1 (2.98(2) Å) is significantly shorter than the one in 2 (3.110(5) Å). The Li–I bonds in 1 and 2 are among the longest ones of their kind. They are longer than most reported terminal, nonbridging Li–I bonds (2.67–2.87 Å) by over 0.1 Å.[40−44] Only one example of a longer terminal Li–I bond (3.233(14) Å) was reported, but the data was low quality (Rint = 10.94%; dmin = 45.8° 2θ).[45] The Li–I bond lengths in 1 (2.98(2) Å) sits at the boundary of the sum of the ionic radii of Li+ (0.9 Å; coordination number (CN) = 6) and I– (2.06 Å),[46] and the Li–I bond length in 2 (3.110(5) Å) is ca. 0.15 Å longer than the sum of their ionic radii. Given the similar level of steric congestion around the Li–I unit in 1 and 2 (as demonstrated by the space-filling models in Figures b and 2b), the difference in the Li–I bond length is intriguing.

Figure 1

(a) X-ray crystal structure of [Li(I)(κ4-N,N′,N″,N′′′-Me6Tren)] (1) at 150 K with 50% probability ellipsoids. Hydrogen atoms are omitted for clarity. Only atoms of the crystallographically independent fragment are labeled. (b) Top-view space-filling presentation for 1 against the Li1–I1 bond. The selected bond distances are Li1–I1, 2.98(2) Å; Li1–N1, 2.19(3) Å; Li1–N2 2.221(6) Å. The selected bond angles are I1–Li1–N1, 180.0°; I1–Li1–N2, 98.4(5)°; N2–Li1–N1, 81.6(5)°; N2–Li1–N2, 117.9(2)°. The atomic color codes: Li (forest green); C (gray); N (blue); and I (magenta).

Figure 2

(a) X-ray crystal structure of [Li(I)(κ4-N,N′,N″,N′′′-DETAN)] (2) at 150 K with 50% probability ellipsoids. Hydrogen atoms are omitted for the sake of clarity. (b) Top-view space-filling presentation for 2 against the Li1–I1 bond. The selected bond distances (Å) and angles (deg) are Li1–I1, 3.110(5); Li1–N1, 2.173(6); Li1–N2 2.149(6); Li1–N3 2.180(6); Li1–N4 2.179(6); I1–Li1–N1, 175.6(2); N2–Li1–N3, 86.3(2); N3–Li1–N4 131.2(3); N4–Li1–N2 137.2(3); N1–Li1–N3, 81.1(2); N1–Li1–N4, 82.7(2); N1–Li1–N2 84.4(2); N4–Li1–I1, 93.40(19); N2–Li1–I1, 99.9(2); N3–Li1–I1, 100.1(2). The atomic color codes: Li (forest green); C (gray); N (blue); I (magenta).

Chart 1

(a) Schematic Representations of the Structures of 1 and 2; (b) The Equatorial Angle Distributions (Apical View) of 1 and 2

Comparing schematic structural representations of 1 and 2 could help readers understand their structural differences (Chart ). It is noticeable that the Li–N bonds in 2 are shorter than those in 1 (Chart a), and the equatorial angle distributions are different (Chart b), although the sum of equatorial angles (∑∠eq) are both close to 360° (353.7° for 1, and 354.8° for 2).

We conducted density functional theory (DFT) calculations to optimize the structures of complexes 1 and 2, employing four types of functionals (PBE, B3LYP, B3PW91, and PBE0).[47−49] The relativistic effective core potential basis set SDD was used for the iodine,[50] while def-TZVPP or 6-31G/6-31+G[50,51] were used for other nuclei. Dispersion forces, weak intermolecular interactions, and THF solvent effects were also considered (see the Supporting Information). All the calculation methods reproduced the experimental structures with good accuracies, including the differences in the Li–I and Li–N bonds between 1 and 2. The NPA charge calculations and the NLMO calculations indicate that the Li–I bonds in complexes 1 and 2, although they differ by ca. 0.13 Å in bond lengths, both are predominantly ionic and have similar electrostatic environments. However, the energy decomposition analysis (EDA) of the Li–I bonds reveal that, in complex 2, the Li–I bond is slightly more covalent than that in 1 (see Tables S1–S4 in the Supporting Information). Given the similarity of the underlying electronic structures between 1 and 2, we attribute the difference in their Li–I bond lengths to slightly different steric environments.

A comparison between the structures of complex 2 and our previous reported [Li(CH3)(κ3-N,N′,N′′-DETAN)] (3) is intriguing (see Chart ).[34] The DETAN ligand in 2 adopts a κ4-N,N′,N″,N′′′ mode, while in 3 it is in a κ3-N,N′,N′′mode. The key difference between 2 and 3 is that one out of the three side arms of the DETAN coordinates in 2, while none of them coordinates in 3. We attribute the difference to the increased steric congestion of 3, compared to 2. The methyl (−CH3) functional group is slightly bigger than the iodide (−I), and the Li–C bond is much shorter than the Li–I bond (2.099(5) Å vs 3.110(5) Å). Similar decoordinating of side arm is observed for the Me6Tren ligand as well. In the [Li(I)(κ4-N,N′,N″,N′′′-Me6Tren)] (1), all the three side arms coordinate to form a five-coordinate, trigonal bipyramidal geometry. In comparison, with the presence of a bulkier Li-CH2SiMe3 group, one of the side arms decoordinates to form a four-coordinate distorted trigonal pyramidal geometry in [Li(CH2SiMe3)(κ3-N,N′,N′′- Me6Tren)] (4) (Chart ), where 4 was recently reported by us.[31]

Chart 2

Comparison between Complexes 2 and 3a

Chart 3

Comparison between Complexes 1 and 4a

Data taken from ref (34). The DETAN ligand is in κ4-N,N′,N″,N′′′ (2) and κ3-N,N′,N′′ (3) mode, respectively.

Data taken from ref (31). The Me6Tren ligand is in κ4-N,N′,N″,N′′′ (1) and κ3-N,N′,N′′ (4) mode, respectively.

In addition to lithium iodide, lithium tetraphenylborate (LiBPh4) also reacts with the DETAN ligand, to produce a separated ion pair (SIP) complex [Li(κ4-N,N′,N″,N′′′-DETAN)][BPh4] (5) (Scheme ). Complex 5 was obtained in 84% yield as a white crystalline solid. Single crystals suitable for SCXRD study were obtained from a Et2O/THF mixed solution at −35 °C. The crystal structure of 5 is displayed in Figure . The coordination geometry of the Li+ center in 5 is best described as a trigonal pyramidal (Li1–N1–N3–N5-N2) capped by a weak N···Li interaction (N4···Li1) (Figure b). The N4···Li1 distance (3.022 Å) is significantly longer than other Li–N distances (2.09–2.21 Å) in the molecule, but shorter than the sum of the van der Waals radii for Li and N (3.36 Å[52]), which support that the N4···Li1 is a weak interaction, while the other Li–N interactions are dative bonds. The coordination geometry of the [Li(DETAN)]+ cationic fragment is similar to a Me6Tren-Li SIP complex [Li(Me6Tren)][AlH2(HMDS)2] reported by Mulvey and co-workers in 2018.[53]

Scheme 2

Reaction To Produce [Li(κ4-N,N′,N″,N′′′-DETAN)][BPh4] (5)

Figure 3

(a) X-ray crystal structure of [Li(κ4-N,N′,N″,N′′′-DETAN)][BPh4] (5) at 150 K with 50% probability ellipsoids. Hydrogen atoms, the minor disorder component and the [BPh4]− anion are omitted for the sake of clarity. (b) Schematic representation of the coordination geometry of the Li+ center. The selected bond distances are as follows: Li1–N1, 2.104(4) Å; Li1–N2, 2.091(4) Å; Li1–N3, 2.135(4) Å; Li1–N5, 2.206(6) Å; and Li1···N5, 3.022 Å. The selected bond angles are as follows: N1–Li1–N3, 87.41(15)°; N3–Li1–N5, 131.7(2)°; and N5–Li1–N1, 138.5(2)°. The atomic color codes: Li (forest green), C (gray), N (blue).

Beyond lithium, we tested the reactions of the Me6Tren and DETAN ligands to other Group 1 metals. As the immediate neighbor of Li+, Na+ has a significantly larger ionic radius (r) compared to Li+ (coordination number (CN) = 4: r = 0.99 Å for Na+ vs r = 0.59 Å for Li+).[46] However, despite the significantly larger ionic radius, we found that Na+ coordinates to the Me6Tren and DETAN ligands in a similar manner to Li+. Like LiI, sodium iodide (NaI) reacted with Me6Tren and DETAN to produce [Na(I)(κ4-N,N′,N″,N′′′-Me6Tren)] (6) and [Na(I)(DETAN)] (7), respectively (Scheme ). Note that the sodium complexes 6 and 7 are more challenging to crystallize than the corresponding Li complexes: only 6 was obtained as single crystals suitable for the SCXRD study. Despite a decent crystalline yield (76%), we could only obtain microcrystals of 7, which are too small even for the synchrotron X-ray source. On the other hand, the 1H and 13C{1H} NMR spectrum of 7 is similar to its lithium analogue 2 (Figures S5/S7 in the Supporting Information for 2, Figures S17/S20 in the Supporting Information for 7; Figure S19 in the Supporting Information for comparisons). To further confirm the monomeric structure of 7, we conducted comparative diffusion-ordered NMR spectroscopy (DOSY)[54] studies of 7 and its SCXRD-characterized Li analogue 2. The protocols reported by Neufeld and Stalke[55] were employed to determine the diffusion coefficient (D), which reflects the hydrodynamic radius of a molecule in the solution (see ref (55) and the Supporting Information for details). During the studies, adamantane was employed as the internal standard.[55] For the monomeric [Li(I)(κ4-N,N′,N″,N′′′-DETAN)] (2), the D value is 9.89 × 10–10 m2 s–1; while for 7, the D value is 9.93 × 10–10 m2 s–1. The similar D values confirm that 2 and 7 have similar hydrodynamic radii, i.e., they are both monomers in solution.

Scheme 3

Reactions To Produce [Na(I)(κ-N,N′,N″,N′′′-Me6Tren)] (6) and [Na(I)(DETAN)] (7)

The single-crystal structure of 6 is exhibited in Figure . Structural comparison between 6 and its Li analogue 1 would be informative. Both 1 and 6 have a trigonal bipyramidal coordination geometry, but the Na–N and Na–I bonds in 6 are ca. 0.4 Å longer than the corresponding Li–N and Li–I bonds in 1, which reflects the larger Na+ ionic radius.[46] As a result, in 6, the Na+ cation sits above the plane defined by the three equatorial N atoms by 0.70 Å. In comparison, in 1, the Li+ cation is much more in the plane with an out-of-plane distance of 0.32 Å.

Figure 4

X-ray crystal structure of [Na(I)(κ4-N,N′,N″,N′′′-Me6Tren)] (6) at 150 K with 50% probability ellipsoids. Hydrogen atoms are omitted for the sake of clarity. Only crystallographically independent noncarbon atoms are labeled. The selected bond distances are as follows: Na1–N1, 2.490(4) Å; Na1–N2, 2.462(8) Å; Na1–I1, 3.004(6) Å. The selected bond angles are as follows: N1–Na1–I1, 180.0°; N1–Na1–N2, 73.6(2)°; N2–Na1–I1, 106.4(2)°. The atomic color codes: Na (forest green); C (gray); N (blue); I (magenta).

The reaction between NaBPh4 and DETAN produced a white crystalline product (8) (Scheme ). Complex 8 is a DETAN-coordinated separated ion-pair complex. The cation [Na(κ6-DETAN)]+ features a six-coordinate Na+ center (Figure a): all the six N atoms of the DETAN ligand coordinate, forming a distorted trigonal prismatic geometry (Figure b). This is in sharp contrast with the four-coordinate trigonal pyramidal [Li(κ4-N,N′,N″,N′′′-DETAN)]+ cation in complex 5 (Figure b). The six Na–N bonds in 8 divided into two groups: (1) the N atoms in the macrocycle (Ncyc) coordinate to the Na+ via short Na–N dative bonds, ca. 2.47 Å (Na–N1/N2/N3); (2) the N atoms in the side arms (Narm) coordinate to the Na+ via long Na–N dative bonds, ca. 2.80 Å (Na–N4/N5/N6). However, even the longer Na–Narm bonds are much shorter than the weakly coordinated Li···N distance (3.022 Å) in 5. We attribute the structural differences between the [Li(κ4-N,N′,N″,N′′′-DETAN)]+ (5) and the [Na(κ6-DETAN)]+ (8) to their different ionic radii of Li+ (r = 0.59 Å, CN = 4) and Na+ (r = 1.02 Å, CN = 6).[46]

Scheme 4

Reaction To Produce [Na(κ6-N6-DETAN)][BPh4] (8)

Figure 5

(a) X-ray crystal structure of [Na(κ6-N6-DETAN)][BPh4] (8) at 150 K with 50% probability ellipsoids. Hydrogen atoms, the minor disorder component and the [BPh4]− anion are omitted for the sake of clarity. (b) The schematic representation of the distorted trigonal prismatic coordination geometry of the Na+ center. The selected bond distances are as follows: Na1–N1, 2.480(7) Å; Na1–N2 2.471(8) Å; Na1–N3 2.459(9) Å; Na1–N4 2.892(8) Å; Na1–N5 2.747(8) Å; Na1–N6 2.798(8) Å. The selected bond angles are as follows: N1–Na1–N2, 72.8(2)°; N2–Na1–N3, 73.5(3)°; N3–Na1–N4, 139.4(3)°; N4–Na1–N5, 106.6(3)°; N5–Na1–N6, 111.4(3)°; N6–Na1–N1, 100.4(3)°. [Atomic color codes: Li (forest green); C (gray); N (blue).]

Since the DETAN and Me6Tren ligands exhibited versatile coordination chemistry to Li and Na cations as demonstrated in the complexes 1–8, we tested their coordinating reactions with larger Group 1 metal halides and tetraphenylborates, i.e., K+, Rb+, and Cs+. However, the reactions between a variety of K+/Rb+/Cs+ reagents (KI, KBPh4, RbI, CsI) and the DETAN/Me6Tren ligands did not proceed at room temperature nor 60 °C within 2 days (Scheme ). Harsher conditions were examined for KBPh4, which is supposed to be the most reactive one among the K/Rb/Cs substrates for its small K+ cation and more-likely soluble BPh4– anion. After heating at 100 °C in toluene for 24 h, there was no reaction between KBPh4 and 1 equiv of Me6Tren or DETAN. We attribute the inertness of these larger Group 1 metal iodides/tetraphenylborates to two possible reasons: (1) their larger ionic radii may not be suitable for the ligands, although there were reports of K[23]/Rb[56]/Cs[56,57] Me6Tren complexes; (2) the MX’s poor solubility (where M = K, Rb, Cs; X = I, BPh4) in THF/toluene may prevent the reactions, although the LiX/NaX are comparably insoluble in the above-mentioned solvents.

Scheme 5

K+/Rb+/Cs+ Iodides/Tetraphenylborates Do Not React with the DETAN/Me6Tren Ligands

Results and Discussion

The coordination modes of the Li+/Na+ DETAN/Me6Tren complexes are summarized in Table . The multidentate neutral Me6Tren and DETAN ligands coordinate to Li+ (r = 0.59 Å, CN = 4) and Na+ (r = 0.99 Å, CN = 4), exhibiting versatile coordination modes, from trigonal bipyramidal (1, 2, 6), tetrahedral (3), trigonal pyramidal (4, 5) to trigonal prismatic (8) (Table ). The coordination mode is dependent on the metal ionic radii and the steric profile of the metal center. Complexes with less steric congestion, i.e., longer Li–E bond and smaller E groups, such as the Li–I complexes 1 and 2, have a tendency to form five-coordinated trigonal bipyramidal geometry, while the side arms of the Me6Tren or DETAN remain coordinated. The larger Li–E groups, such as −CH3 (3) and −CH2SiMe3 (4), have a tendency to cause the decoordination of the side arms. Without any E group, the cationic Li+ center, on the other hand, formed an N-capped four-coordinate trigonal pyramidal geometry in complex 5, instead of a five-coordinated trigonal bipyramidal geometry, probably due to the geometric strains of the DETAN ligand. In comparison, the larger Na+ cation forms a six-coordinate distorted trigonal prismatic geometry in complex 8, where all the six N atoms of the DETAN ligand coordinate.

Table 1

Diversified Solid-State Coordination Modes of the Me6Tren/DETAN Ligands with Li+/Na+

We would like to draw our reader’s attention to the fact that all the previous discussions about coordination modes are based on the SCXRD structures, i.e., in solid state. In solutions, the situations could be more complicated. The discrepancy between solid-state and solution-state structures is a long-lasting debateful topic in coordination chemistry, particularly in s-block metal chemistry.[58,59] It is possible that the DETAN and Me6Tren ligands exhibit a rapid coordination/decoordination equilibrium, involving the entire ligand or a part of it. This hypothesis is proved by deliberately introducing an extra amount of free DETAN ligand to the isolated [Na(I)(DETAN)] (7). The 1H NMR spectrum of a mixture of [Na(I)(DETAN)] (7) (0.0288 g, 0.05 mmol), free DETAN ligand (0.0085 g, 0.02 mmol), and an internal NMR integration standard adamantane (0.0068 g, 0.05 mmol) (see Figure S20) exhibit new 1H NMR signals, which are different from both the free DETAN and 7. However, the complicated solution-state behaviors will not compromise our previous solid-state structural discussions. A comprehensive and systematic solution-state NMR study (including temperature-/concentration-dependent behaviors) about this series of DETAN/Me6Tren complexes, which would complement this work, is out of the scope of this Article and will be published in short due.

Conclusions and Outlook

This work demonstrated the capability and versatility of two multidentate neutral amine ligands: the tetra-dentate Me6Tren and the hexa-dentate DETAN, in Group 1 metal chemistry. The side arms of the Me6Tren/DETAN ligands exhibited flexible coordinating capabilities, which could act as on-demand internal Lewis bases to promote desired reactions and to compensate the reduction of coordination numbers upon the formation of a desired highly reactive species, such as from a heteroleptic metal alkyl amide complex to a terminal metal imido complex via alkane elimination.[37,39]

As a closing remark, this work will serve as a guidebook for future adventures employing the DETAN/Me6Tren ligands to isolate highly reactive Group 1 and Group 2 metal complexes and their reactivity studies, such as small molecular activations. Given the similar ionic M–E bonding character, the knowledge of DETAN’s coordination modes could also be extrapolated into rare-earth metal chemistry, to pursue high-value targets such as divalent rare-earth terminal imides[37] and phosphinidenes.[60] This work could also help the coordination chemistry and catalysis communities to choose suitable multidentate neutral amine ligands to design new metal complexes and stoichiometric/catalytic reactions.